NANOPARTICLES OF METAL-ORGANIC FRAMEWORKS: A GENERAL
SYNTHETIC METHOD AND SIZE-DEPENDENT PROPERTIES
by
CHECKERS MARSHALL
A DISSERTATION
Presented to the Department of Chemistry and Biochemistry 
and the Division of Graduate Studies of the University of 
Oregon in partial fulfillment of the requirements
for the degree of
Doctorate of Philosophy
March 2023
DISSERTATION APPROVAL PAGE
Student: Checkers Marshall
Title: Nanoparticles of Metal-Organic Frameworks: A General Synthetic Method
and Size-Dependent Properties
This dissertation has been accepted and approved in partial fulfillment of the
requirements for the Doctorate of Philosophy degree in the Department of
Chemistry and Biochemistry by:
Victoria J. DeRose Chair
Carl K. Brozek Core Member
Jim E. Hutchison Core Member
Matthew L. Polizzotto Outside Member
and
Krista Chronister Vice Provost for Graduate Studies
Original approval signatures are on file with the University of Oregon Division 
of Graduate Studies.
Degree awarded March 2023
ii
© 2023 Checkers Marshall
This work, including text and images of this document, is licensed under a Creative 
Commons Attribution Share-Alike 4.0 International License.
iii
DISSERTATION ABSTRACT
Checkers Marshall
Doctorate of Philosophy
Department of Chemistry and Biochemistry
March 2023
Title: Nanoparticles of Metal-Organic Frameworks: A General Synthetic Method
and Size-Dependent Properties
Metal-organic framework nanoparticles exhibit both high internal surface
area and good colloidal stability in a variety of solvents including biological
media. These materials are sought after for a range of applications, mainly in drug
delivery, catalysis, and separation membranes. (1) Considerable effort has been put
into controlling the size and shape of MOF crystals to develop materials that, due
to small particle size and good colloidal stability, may be solution-processable.
In this thesis, a simple model to help predict size trends in MOF nanoparticle
syntheses is developed, then the model is applied both to well-known and novel
MOF nanoparticle systems. In Chapter 2, I first present a simple equilibrium
model to further our understanding of how to control MOF nanoparticle size. MOF
nanoparticles can be synthesized via several top-down and bottom-up approaches.
One of the most prevalent bottom-up methods is to use “modulators,” molecules
added to the growth media to change the reaction conditions and therefore the
crystals’ growth. This chapter encompasses a literature-based perspective on
how the presence and identity of a modulator will impact the final size of a MOF
nanoparticle and introduces the “Seesaw model” to explain these effects. In
Chapter 3, I then apply this model to the well-known nanoMOF systems Zn(mIm)2
iv
(ZIF-8) and Cu3BTC2 (HKUST-1). We show that, by using a mixture of an
conjugate acid/base pair, that both modulator equivalents and proton activity
play a role in determining final particle size. In Chapter 4, I first develop the
synthesis for a novel nanoMOF system, a family of metal-triazolate MOFs, then
explore the MOF Fe(1,2,3-triazolate)2 in more depth for its size-dependent optical
and electronic properties. Finally, in Chapter 5, the effects of ion identity, solvent
identity, particle size, and film thickness on the redox activity of Fe(TA)2 thin films
is studied in depth.
This dissertation contains both published and unpublished co-authored
material.
v
CURRICULUM VITAE
NAME OF AUTHOR: Checkers Marshall
GRADUATE AND UNDERGRADUATE SCHOOLS ATTENDED:
University of Oregon, Eugene, OR, USA
Fort Lewis College, CO, USA
DEGREES AWARDED:
Doctor of Philosophy, Inorganic Chemistry, 2023, University of Oregon
Master of Science, 2023, University of Oregon
Bachelor of Science, Chemistry, 2016, Fort Lewis College
AREAS OF SPECIAL INTEREST:
Sustainable Energy
Metal-Organic Frameworks
PROFESSIONAL EXPERIENCE:
Graduate Teaching Fellow, University of Oregon, 2017-2020
General & Organic Chemistry Lab TA, Fort Lewis College, 2014-2016
GRANTS, AWARDS AND HONORS:
Rosaria Haugland Fellowship, 2020
Deans First Year Merit Award, University of Oregon, 2017
James Mills Outstanding Performance in Physical Chemistry, 2016
PUBLICATIONS:
vi
Marshall, C.R.; Dvorak, J.D.; Twight, L.P.; Kadota, K.; Chen, L.;
Overland, A.E.; Ericson, T.; Cozzolino, A.F.; Brozek, C.K. “Size-
Dependent Properties of Solution-Processable Conductive MOF
Nanoparticles,” J. Am. Chem. Soc., 2022
Marshall, C.R.; Timmel, E.T; Staudhammer, S.A.; Brozek, C.K.
“Experimental Evidence for a General Model of Modulated MOF
Nanoparticle Growth,” Chem. Sci., 2020
Schaub, T. A.; Prantl, E.; Kohn, J.; Bursch, M.; Marshall, C. R.;
Leonhardt, E. J.; Lovell, T. C.; Zakharov, L. N.; Brozek, C. K.;
Waldvogel, S. R.; Grimme, S.; Jasti, R. “Exploration of the Solid-State
Sorption Properties of Shape-Persistent Macrocyclic Nanocarbons as
Bulk Materials and Small Aggregates,” J. Am. Chem. Soc., 2020.
Van Raden, J.; Leonhardt, E.; Zakharov, L.; Pérez-Guardiola, A.; Pérez-
Jiménez, Á.; Marshall, C.; Brozek, C.; Sancho-Garćıa, J.-C.; Jasti, R.
“Precision Nanotube Mimics via Self-Assembly of Programmed Carbon
Nanohoops,” Chem. Mater., 2020, 32, 5802.
Marshall, C. R.; Staudhammer, S. A.; Brozek, C. K. “Size Control of
Metal-Organic Framework Porous Nanocrystals,” Chem. Sci., 2019
vii
ACKNOWLEDGEMENTS
During my PhD, I faced challenges I thought I never would. I have made
it through this experience with my own hard work, but it would not have been
possible without the support of those around me. Most importantly, I am deeply
grateful to my advisor, Dr. Carl Brozek, who taught me about the innerworkings of
academia and pushed me to be the best scientist I could become. I would also like
to thank my commitee members, Dr. Jim Hutchison, Dr. Vickie DeRose, and Dr.
Matt Polizzotto, for their valuable advice. Special thanks to the Rosaria Haugland
Foundation for granting me a generous Fellowship. I would like to thank the
CAMCOR staff, in particular Valerie Brogden, Stephen Golledge, and Josh Razink,
for their invaluable knowledge and skills that helped me collect beautiful data.
Thank you to my mentors during my first year: Lisa Enman, Brandon Crockett,
Meredith Sharps, and Dani Hamann. Many thanks to the front office staff in the
Chemistry department, without whom this place would certainly fall apart.
My community made my success possible. To my partner, Alex Brunner,
I give my heartfelt thanks for his constant pep talks, incredible cooking, and for
taking me on weekend adventures even when I told him I should be in lab. To
my closest friends, I am deeply indebted: Kai Pham, Courtney Ragle, Remington
Koltes, Sam and Evan West, Jake Anderson, and Quinn Valentine. To those in
my cohort who helped me through the program with study dates and lunches,
particularly Jenna Mancuso and Marisa Choffel. To my parents, who have always
encouraged my creative and scientific pursuits, even though some of them are
rather dangerous. To my sister Laura, for endless LOTR memes, and her partner
Dan, who I can always rely on to share my complaints about working in science. To
viii
Henry Tregillus, who helped me format this giant document properly. And to Jess
Lohrman, who read drafts of my documents and who let me borrow her backyard.
To my lab mates, who inspired me, challenged my thinking, and occasionally
challenged my patience. To every one of my mentees, both rotation students and
undergraduates: you taught me as much as I taught you. I would like to thank
all of my incredible undergraduate researchers individually: Sara Staudhammer,
Maria Anderson, Alexi Overland, Kelsie Heffernan, and Jeff Gombart. To the
lovely and creative people I met through flow arts, blues dancing, and the Oregon
Country Fair. To the staff at Staszak Physical Therapy, who somehow make injury
management fun. To my undergraduate professors, who ignited my passion in
chemistry. And finally to the discipline of chemistry itself, for as we all know, we
stand on the shoulders of giants.
ix
This thesis is dedicated to the potato, the most noble of Earth tubers.
x
TABLE OF CONTENTS
Chapter Page
I. INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . 1
II. SIZE CONTROL OVER MOF NANOPARTICLES . . . . . . . . . . 6
NanoMOF Metadata . . . . . . . . . . . . . . . . . . . . . . 8
Factors Controlling MOF Nanocrystal Sizes . . . . . . . . . . . . 13
Modulators . . . . . . . . . . . . . . . . . . . . . . . . . 14
Linker Equivalents . . . . . . . . . . . . . . . . . . . . . . 18
Metal-Linker Binding Strength . . . . . . . . . . . . . . . . 21
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . 22
The Seesaw Relationship . . . . . . . . . . . . . . . . . . . . . 23
Best Practices and Outstanding Challenges . . . . . . . . . . . . 26
Colloidal Stability . . . . . . . . . . . . . . . . . . . . . . 26
Incorporation of Modulator . . . . . . . . . . . . . . . . . . 26
Measurement Methods . . . . . . . . . . . . . . . . . . . . 28
Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . 29
III. EXPERIMENTAL EVIDENCE FOR THE SEESAW
MODEL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
The Seesaw Model and pKa . . . . . . . . . . . . . . . . . . . 33
ZIF-8 and Cu2BTC3 Nanoparticle Syntheses . . . . . . . . . . . . 35
Applying the Seesaw Model . . . . . . . . . . . . . . . . . . . 46
Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . 53
xi
Chapter Page
IV. SIZE-DEPENDENT PROPERTIES OF CONDUCTIVE
MOF NANOCRYSTALS . . . . . . . . . . . . . . . . . . . . . . 55
Size-tunable Synthesis of Iron Triazolate Nanoparticles . . . . . . . 57
Size-Dependent Optical Properties of Colloidal Fe(TA)2 Nanoparticles . 66
Redox Chemistry and Charge Transport of Fe(TA)2 Nanoparticles . . 74
Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . 82
V. NANOCONFINEMENT EFFECT AND REDOX
CHEMISTRY IN FE(TA)2 NANOPARTICLE THIN FILMS . . . . . . . 83
Identifying Intercalation Chemistry in Fe(TA)2 Films . . . . . . . . 87
Factors Controlling Ion Intercalation Redox Chemistry in Fe(TA)2 . . 93
Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . 101
VI. CONCLUDING REMARKS . . . . . . . . . . . . . . . . . . . . 103
APPENDICES
A. SUPPLEMENTARY INFORMATION FOR CHAPTER 2 . . . . . . . 106
B. SUPPLEMENTARY INFORMATION FOR CHAPTER 3 . . . . . . . 124
Synthetic Details and Analytical Methods . . . . . . . . . . . . . 124
Materials & Methods . . . . . . . . . . . . . . . . . . . . . 124
Cu3BTC2 (HKUST-1) Expanded Synthetic Methods . . . . . . . 124
Scherrer Size Estimations . . . . . . . . . . . . . . . . . . . 125
Tables Detailing Synthetic Conditions and Particle Size . . . . . . 126
PXRD Patterns of Cu3BTC2 and ZIF-8 Products . . . . . . . . 135
xii
Chapter Page
C. SUPPLEMENTARY INFORMATION FOR CHAPTER 4 . . . . . . . 142
General Methods . . . . . . . . . . . . . . . . . . . . . . . . 142
Synthesis of Fe(TA)2 under Different Conditions . . . . . . . . . . 144
Basic Characterization of Fe(TA)2 Nanoparticles . . . . . . . . . . 150
Additional UV-Vis Data . . . . . . . . . . . . . . . . . . . 157
Additional Electrochemical Data . . . . . . . . . . . . . . . . 160
D. SUPPLEMENTARY INFORMATION FOR CHAPTER 5 . . . . . . . 171
General Methods . . . . . . . . . . . . . . . . . . . . . . . . 171
Synthesis and Characterization of Iron Triazolate
(Fe(TA)2) Nanoparticles . . . . . . . . . . . . . . . . . . . 171
REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
xiii
LIST OF FIGURES
Figure Page
1.1. Summary of work presented in this thesis . . . . . . . . . . . . . . 2
2.1. Summary of all MOF materials reported to-date as
nanocrystals with precisely measured particle diameters. . . . . . . . . 8
2.2. Size comparisons of HKUST-1 nanocrystals prepared
by (A) microwave-assisted growth at varying reactant
concentrations and added equivalents of dodecanoic acid
and (B) by solvothermal synthesis . . . . . . . . . . . . . . . . . 9
2.3. Key chemical equilibria controlling nano-MOF growth and termination. . 13
2.4. Nanoscale MOF sizes depend on the equivalents and pKa
values of added modulator reagents. . . . . . . . . . . . . . . . . 17
2.5. ZIF nanocrystal syntheses with varying relative ratios of
metal, linker, modulator, and solvent. . . . . . . . . . . . . . . . . 20
2.6. Heterobimetallic ZIF-8 nanocrystals increase in size as the
Zn2+ atoms are substituted for Co2+ or Cu2+ atoms. . . . . . . . . . 21
2.7. Reaction conditions that favor small or large MOF
nanocrystal sizes when linker or acidic modulators are
present in excess. . . . . . . . . . . . . . . . . . . . . . . . . . 22
2.8. The “Seesaw” relationship between nanocrystal sizes and
added equivalents of acidic ligands. . . . . . . . . . . . . . . . . . 24
3.1. Equilibrium concentrations of MOF linker species
determined from the coupled equilibria of the Seesaw model. . . . . . . 32
3.2. Scheme depicting the dependence of particle sizes on metal-
to-linker solution stoichiometry. . . . . . . . . . . . . . . . . . . 34
3.3. ZIF-8 nanoparticle sizes as a function of excess modulator equivalents. . 38
3.4. Dependence of Cu3BTC2 particle sizes on modulator equivalents. . . . . 41
3.5. Dependence of Cu3BTC2 particle sizes on benzoic acid
content (% BA). . . . . . . . . . . . . . . . . . . . . . . . . . 43
xiv
Figure Page
3.6. Cu3BTC2 particle sizes resulting from variable reactant
concentrations and total modulator equivalents. . . . . . . . . . . . 46
3.7. Dependence of Cu3BTC2 particle morphology on modulator
equivalents for a 33% BA modulator mixture. . . . . . . . . . . . . 48
4.1. Overview of Fe(1,2,3,-triazolate)2 (Fe(TA)2) nanocrystal synthesis. . . . 56
4.2. Particle sizes of Fe(TA)2 resulting from modulated syntheses. . . . . . 58
4.3. Mössbauer spectra of the smallest FeTA2 samples under N2. . . . . . . 60
4.4. SEM images of particles synthesized with different
modulators. a) Fe(TA)2 synthesized with 0.218 eq of n-butylamine. . . . 62
4.5. Synthesis of nanosized Co(TA)2 using 1-mIm as a modulator. . . . . . 64
4.6. Size trends of new Cd(TA)2 products modulated by 1-methylimidazole. . 65
4.7. Solution-state UV-Vis absorption spectra of Fe(TA)2 nanoparticles. . . . 67
4.8. Nitrogen isotherms of bulk iron triazolate compared to
48-nm nanoparticles. . . . . . . . . . . . . . . . . . . . . . . . 70
4.9. Extinction coefficients of Fe(TA)2 nanoparticles. . . . . . . . . . . . 71
4.10. The extinction coefficient and oscillator strength of Fe(TA)2
nanoparticles per formula unit. . . . . . . . . . . . . . . . . . . . 73
4.11. Cyclic voltammetry of Fe(TA)2 nanoparticles analyzed as
colloids or thin films. . . . . . . . . . . . . . . . . . . . . . . . 75
4.12. Cyclic voltammetry with superimposed QCM data of
Fe(TA)2 nanoparticles. . . . . . . . . . . . . . . . . . . . . . . 79
4.13. Charge transport measurements of Fe(TA)2 nanoparticle
thin films. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
5.1. Scheme of ion intercalation and redox chemistry in FeTA2 film. . . . . . 84
5.2. CV scans and corresponding mass change data from EQCM
for three Fe(TA)2 nanoparticle thin films . . . . . . . . . . . . . . 86
5.3. Plotting the change in mass on the QCM electrode over the
voltage range of surface and intercalation features. . . . . . . . . . . 88
xv
Figure Page
5.4. ToF-SIMS of spatial distribution of boron and
tetrabutylammonium (TBA) in Fe(TA)2 film after applying
different voltages. . . . . . . . . . . . . . . . . . . . . . . . . 90
5.5. Mass change vs. charge change in Fe(TA)2 film. . . . . . . . . . . . 93
5.6. AFM images of films of Fe(TA)2 particles on QCM
electrodes to determine film thickness. . . . . . . . . . . . . . . . 94
5.7. Film thickness, particle size, and redox reversibility. . . . . . . . . . . 95
5.8. Tracking the changes in E1/2 in three different Fe(TA)2 thin films. . . . 97
5.9. Electrolyte-dependent redox properties in Fe(TA)2
nanoparticle film. . . . . . . . . . . . . . . . . . . . . . . . . . 98
5.10. Electrolyte titrations show redox properties in Fe(TA)2
nanoparticle film. . . . . . . . . . . . . . . . . . . . . . . . . . 100
B.1. An example of peak fitting in Igor 6.32. The sample being
fit is denoted in table B.9. . . . . . . . . . . . . . . . . . . . . . 125
B.2. Initial experiments were performed to explore the use of
sodium benzoate in Cu3BTC2 synthesis . . . . . . . . . . . . . . . 135
B.3. Representative PXRD patterns of Cu3BTC2 synthesized
with increasing modulator equivalents (50% benzoic acid) . . . . . . . 135
B.4. Representative PXRD patterns of ZIF-8. . . . . . . . . . . . . . . 136
B.5. The two ratios 3L : 1M and 2L : 3M are compared at the
same metal concentration of 0.001 M with a 50% benzoic
acid modulator in Cu3BTC2 synthesis. . . . . . . . . . . . . . . . 137
B.6. Data from several Cu3BTC2 samples is compared to show reproducibility 137
B.7. Additional crystallite size data for Cu3BTC2 with respect to
modulator equivalents. . . . . . . . . . . . . . . . . . . . . . . 138
B.8. SEM images of Cu3BTC2 particles as a function of benzoic
acid content at a constant modulator equivalents. . . . . . . . . . . . 139
B.9. SEM images of Cu3BTC2 particles as a function of
modulator equivalents for a 66% benzoic acid modulator
mixture. Reactant concentrations for these samples can be
found in Table B.7. . . . . . . . . . . . . . . . . . . . . . . . . 140
xvi
Figure Page
B.10.Particle size from SEM compared to Scherrer crystallite
size, as a function of benzoic acid content. . . . . . . . . . . . . . . 140
B.11.Particle size from SEM compared to Scherrer crystallite size
as a function of modulator equivalents for a 33% benzoic
acid modulator mixture. . . . . . . . . . . . . . . . . . . . . . . 141
C.1. PXRD patterns of Fe(TA)2 products synthesized with modulators. . . . 145
C.2. SEM images of Fe(TA)2 products synthesized with different
reaction times and conditions. . . . . . . . . . . . . . . . . . . . 146
C.3. Reaction yield of Fe(TA)2 as a function of 1-
methylimidazole equivalents. . . . . . . . . . . . . . . . . . . . . 147
C.4. Iron triazolate crystallize size trends with respect to
modulator equivalents. . . . . . . . . . . . . . . . . . . . . . . 148
C.5. Le Bail fits for largest and smallest Fe(TA)2 nanoparticles. . . . . . . . 149
C.6. Larger SEM images of iron triazolate products. . . . . . . . . . . . . 150
C.7. SEM images of the two smallest Fe(TA)2 particle sizes. . . . . . . . . 151
C.8. Acid digestion 1H NMR spectra of iron triazolate
nanoparticles compared to the constituent ligands and
the bulk MOF product. . . . . . . . . . . . . . . . . . . . . . . 153
C.9. IR spectra of the bulk material and the smallest Fe(TA)2 nanoparticles. . 154
C.10.DLS data collected for Fe(TA)2 particles synthesized with
0.436 eq 1-methylimidazole. . . . . . . . . . . . . . . . . . . . . 154
C.11.Differential Scanning Calorimetry of FeTA2 nanoparticles
synthesized with 3.63 eq 1-mIm compared to previously
published data of the bulk material. ()3 Cycles were collected
at a scan rate of 10 °C min−1. . . . . . . . . . . . . . . . . . . . 155
C.12.Mössbauer spectra of several sizes of Fe(TA)2 in air. . . . . . . . . . 156
C.13.UV-Vis traces for syntheses employing other modulators. . . . . . . . 157
C.14.Beer’s Law plots used to determine Fe(TA)−2 extinction coefficients. . . 158
C.15.Energetic transitions determined from UV-Vis spectra, from
the raw data and from gaussian fits, plotted against particle
diameter, 1/r, and 1/(r2). . . . . . . . . . . . . . . . . . . . . . 159
xvii
Figure Page
C.16.CV scans of Fe(TA)2 particles of varying size as colloids in
0.1 M TBAPF6 / DMF. . . . . . . . . . . . . . . . . . . . . . 160
C.17.Control CV collected of the triazole ligand in 0.1 M
TBAPF6 / DMF. Only a simple irreversible oxidation is
observed. Current is normalized to the area of the glassy
carbon working electrode. . . . . . . . . . . . . . . . . . . . . . 161
C.18.CV scans of 16-nm Fe(TA)2 particles drop-casted onto the
glassy carbon working electrode in 0.1 M TBAPF6 / MeCN. . . . . . . 162
C.19.CV scans of the bulk Fe(TA)2 material. . . . . . . . . . . . . . . . 163
C.20.CV scans with varying scan rate on QCM electrodes
in TBAPF6 (a) and TBABF4 (b). The lightest grey
corresponds to 10 mV/s, followed by 100 mV/s, 300 mV/s,
and finally 500 mV/s. . . . . . . . . . . . . . . . . . . . . . . . 163
C.21.Microscope images of films used in QCM experiments. a)
SEM image of a film as-deposited. . . . . . . . . . . . . . . . . . 164
C.22.Photos of Fe(TA)2 thin films made in air. a) Films of the
largest sizes of particles were created from 20 mg/mL dispersions. . . . 167
C.23.Photos of Fe(TA)2 thin films made in the glovebox, then
brought out into air. . . . . . . . . . . . . . . . . . . . . . . . 168
C.24.SEM images of Fe(TA)2 thin films of varying particle size
made in air and used for in-air co-linear conductivity measurements. . . 169
C.25.SEM images of Fe(TA)2 thin films of varying particle size
made under N2 and used for Van der Pauw conductivity measurements. . 170
D.1. Basic Characterization of Fe(TA)2. PXRD pattern (a) and
SEM image (b) of 16-nm particles. PXRD pattern (c) and
SEM image (d) of 25-nm particles. PXRD pattern (e) and
SEM image (f) of 16-nm particles. . . . . . . . . . . . . . . . . . 175
D.2. CV traces starting and stopping at different voltages show
that the 1.2 V feature is independent of other features in
the CV. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176
D.3. Charge variation in Fe(TA)2 films during CV measurements. . . . . . . 177
D.4. Tracking the changes in ∆E and E1/2 at the 0.0 V and 1.2
V feature during TBABF6 electrolyte titration. . . . . . . . . . . . . 178
xviii
Figure Page
D.5. CVs collected during a titration experiment where
TBACLO4 was added to the TBABF4 electrolyte. CV
results are collected at the scan rate of 10 mV/s . . . . . . . . . . . 179
D.6. CV measurements of Fe(TA)2 films in acetonitrile (red) and
1,2-difluorobenzene (blue). . . . . . . . . . . . . . . . . . . . . . 180
D.7. CVs of Fe(TA)2 particles in four different solvents. . . . . . . . . . . 181
xix
LIST OF TABLES
Table Page
2.1. A guide to MOF names, their component linkers and metals,
and modulators that have been shown to access nanoscale
particle size. . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
A.2. Typical M:L ratios and L concentrations for MOF
nanocrystals discussed in this perspective. These values are
compared to 1-2 representative examples for bulk syntheses.
Bold values indicate nanoscale syntheses where either excess
linker or a more dilute system was used compared to bulk syntheses. . . 106
A.1. Values for the smallest, median, and average nanocrystal
sizes (all in nm) reported in Figure 1 of the main text.
The smallest MOF nanocrystals made by other methods
are reported along with the method used: metal organic
gel (Gel,) ionic liquid microemulsion (ILM), dual injection
(Inject), and slow addition (SA) . . . . . . . . . . . . . . . . . . 110
A.3. Nanocrystal sizes used to create Fig 1 in the main text. Size
distributions given in nanometers are shown in parentheses.
Dispersity measurements reported as PDI (Polydispersity
Index), standard deviation, relative standard deviation, or a
range of nanometers are reported in Disp. (Other) with the
type of data in parentheses (PDI, SD, RSD, or Range). In
cases where anisotropic particles were presented with both
length and width, they are differentiated here as (L) and
(W). N/A stands for Not Available. . . . . . . . . . . . . . . . . . 111
B.1. Cu3BTC2 synthesis using varying benzoic acid to sodium
benzoate ratios with excess linker and a metal concentration
of 0.001 M. M, L, and Mod refer to the concentrations of the
metal salt, linker, and modulator in M. Sizes are calculated
from the Scherrer equation. These data are depicted in
Figure 3.4 of the main text and Figure B.8. . . . . . . . . . . . . . 126
xx
Table Page
B.2. Cu3BTC2 synthesis using varying benzoic acid to sodium
benzoate ratios with excess linker and a metal concentration
of 0.025 M. M, L, and Mod refer to the concentrations of
the metal salt, linker, and modulator in M.These data are
depicted in Figure 3.4 of the main text. . . . . . . . . . . . . . . . 127
B.3. Cu3BTC2 synthesis using varying benzoic acid to sodium
benzoate ratios with excess linker and a metal concentration
of 0.005 M. M, L, and Mod refer to the concentrations of
the metal salt, linker, and modulator in M. These data are
depicted in Figure 3.4 of the main text. . . . . . . . . . . . . . . . 127
B.4. Cu3BTC2 synthesis using varying benzoic acid to sodium
benzoate ratios with a stoichiometric amount of linker and a
metal concentration of 0.027 M. M, L, and Mod refer to the
concentrations of the metal salt, linker, and modulator in M. . . . . . . 128
B.5. Cu3BTC2 synthesis using varying equivalents of a 33%
benzoic acid modulator, with excess linker and a metal
concentration of 0.001 M. M, L, and Mod refer to the
concentrations of the metal salt, linker, and modulator in
M. These data are depicted in Figure 3.3 of the main text.
SEM images of these samples are shown in figure 3.?, and
the Scherrer crystallite sizes and apparent particle size from
SEM images are compared in figure B.10. . . . . . . . . . . . . . . 128
B.6. Cu3BTC2 synthesis using varying equivalents of a 50%
benzoic acid modulator, with excess linker and a metal
concentration of 0.001 M.M, L, and Mod refer to the
concentrations of the metal salt, linker, and modulator in
M. These data are depicted in Figure 3.3 of the main text. . . . . . . . 129
B.7. Cu3BTC2 synthesis using varying equivalents of a 66%
benzoic acid modulator, with excess linker and a metal
concentration of 0.001 M. M, L, and Mod refer to the
concentrations of the metal salt, linker, and modulator in
M. These data are depicted in Figure 3.3 of the main text
and SEM images are in figure B.10. . . . . . . . . . . . . . . . . . 129
B.8. Cu3BTC2 synthesis using varying equivalents of a 33%
benzoic acid modulator, with a stoichiometric amount of
linker and a metal concentration of 0.0027 M. M, L, and
Mod refer to the concentrations of the metal salt, linker, and
modulator in M. . . . . . . . . . . . . . . . . . . . . . . . . . 130
xxi
Table Page
B.9. Cu3BTC2 synthesis using varying equivalents of a 50%
benzoic acid modulator, with a stoichiometric amount of
linker and a metal concentration of 0.0027 M. M, L, and
Mod refer to the concentrations of the metal salt, linker, and
modulator in M. . . . . . . . . . . . . . . . . . . . . . . . . . 130
B.10.Cu3BTC2 synthesis using varying equivalents of a 66%
benzoic acid modulator, with a stoichiometric amount of
linker and a metal concentration of 0.0027 M. M, L, and
Mod refer to the concentrations of the metal salt, linker, and
modulator in M. . . . . . . . . . . . . . . . . . . . . . . . . . 131
B.11.Cu3BTC2 synthesis using varying equivalents of a 50%
benzoic acid modulator, with a stoichiometric amount of
linker and a metal concentration of 0.001 M. M, L, and Mod
refer to the concentrations of the metal salt, linker, and
modulator in M. . . . . . . . . . . . . . . . . . . . . . . . . . 131
B.12.Cu3BTC2 synthesis with varying total concentration,
employing 7 equivalents of modulator with respect to the
linker. M, L, and Mod refer to the concentrations of the
metal salt, linker, and modulator in M. The ratios of the
reagents and the total volume were kept constant. These
data are depicted in Figure 4 of the main text. . . . . . . . . . . . . 132
B.13.Cu3BTC2 synthesis with varying total concentration,
employing 0.7 equivalents of modulator with respect to the
linker. The ratios of the reagents and the total volume were
kept constant. M, L, and Mod refer to the concentrations of
the metal salt, linker, and modulator in M. These data are
depicted in Figure 4 of the main text. . . . . . . . . . . . . . . . . 132
B.14.Cu3BTC2 synthesis with varying total concentration,
employing 13.34 equivalents of modulator with respect to
the linker. The ratios of the reagents and the total volume
were kept constant. These data are depicted in Figure 4 of
the main text. . . . . . . . . . . . . . . . . . . . . . . . . . . 133
B.15.Cu3BTC2 synthesis scale up of representative samples. The
total volume was kept constant. . . . . . . . . . . . . . . . . . . . 133
B.16.Cu3BTC2 synthesis utilizing copper acetate as a metal
source and benzoic acid as the modulator. . . . . . . . . . . . . . . 133
xxii
Table Page
B.17.ZIF-8 synthesis with increasing linker equivalents: In this
experiment, a modified literature method was used.1 The
total volume changes, allowing linker concentration (0.1901
M) to remain constant. . . . . . . . . . . . . . . . . . . . . . . 134
B.18.ZIF-8 synthesis with increasing linker equivalents: In this
experiment, a modified literature method was used.1 The
total volume changes, allowing linker concentration (0.0988
M) to remain constant. . . . . . . . . . . . . . . . . . . . . . . 134
B.19.ZIF-8 synthesis with HCl: This experiment used the
conditions for 14 Hmim equivalents. A 1 M solution of HCl
was made from concentrated HCl mixed into methanol. . . . . . . . . 134
C.1. Crystallite sizes and strain of 130 nm (0.05 eq 1-mIm) and
5.5 nm (10.9 eq 1-mIm) particles from Le Bail fitting . . . . . . . . . 149
C.2. Iron triazolate particle sizes from syntheses with varying
amounts of 1-methylimizadole (modulator eq is with respect
to FeCl2). We include Scherrer sizes for two separate
batches, followed by the statistical particle distribution as
determined by sizing particles from SEM image data with
the line tool in ImageJ. (2) The particle size distribution was
fit to a weighted gaussian, and the values presented are the
mode ± standard deviation (σ). . . . . . . . . . . . . . . . . . . 151
C.3. Iron triazolate particle sizes from syntheses with varying
amounts of 5-bromo-1-methylimidazole. We include Scherrer
sizes, and a select SEM size for these exploratory modulator
syntheses. The particle size distribution was fit to a weighted
Gaussian curve, and the values presented are the mode ±
standard deviation (σ). The asterisk denotes a significant
phase impurity observed in the sample. . . . . . . . . . . . . . . . 152
C.4. Iron triazolate particle sizes from syntheses with varying
amounts of 1-benzyl-2-methylimidazole. We include Scherrer
sizes, and a select SEM size for these exploratory modulator
syntheses. The particle size distribution was fit to a weighted
gaussian, and the values presented are the mode ± standard
deviation (σ). . . . . . . . . . . . . . . . . . . . . . . . . . . 152
xxiii
Table Page
C.5. Iron triazolate particle sizes from six identical syntheses.
To ensure the synthesis was replicable, seven small scale
syntheses were repeated. Phase purity was confirmed by
PXRD and Scherrer analysis was performed to obtain
particle sizes. . . . . . . . . . . . . . . . . . . . . . . . . . . . 152
C.6. Peak positions of redox features in colloidal Fe(TA)2 in 0.1
M TBAPF6 / DMF. Peaks in the colloids are wide and they
overlap one another, such that determining peak positions
was not always possible to do accurately. . . . . . . . . . . . . . . . 161
C.7. Values of electrons and ions determined from QCM
experiments at 100 mV/s. The mass of the MOF was
calculated using the Sauerbrey relation from the change
in frequency of the dry, bare QCM and the QCM after spin-
coating and the crystal was allowed to dry fully. Moles of
electrons was determined by integrating the CV curves.
Moles of anions was calculated from change in frequency
over a cycle, using first the Saurbrey relationship, then the
molecular weight of the unsolvated anions. . . . . . . . . . . . . . 165
C.8. Raw data for conductivity measurements on Fe(TA)2 thin
films made in air. The films were allowed to sit in ambient
aerobic conditions for over 1 week. Measurements were
performed with a co-linear four point probe method. . . . . . . . . . . 165
C.9. Raw data for conductivity measurements on Fe(TA)2 thin
films made in N2. The films were allowed to sit in ambient
aerobic conditions for 1 day, and measurements were
collected once per day whenever possible. Measurements
were performed with a Van der Pauw four point probe
method. Two measurements are collected in different
orientations (R1 and R2). F is a correction factor. . . . . . . . . . . . 166
xxiv
CHAPTER I
INTRODUCTION
Metal-organic framework nanoparticles exhibit both high internal surface
area and colloidal stability in a variety of solvents, including biological media.
These materials are sought after for a range of applications, mainly in drug
delivery, catalysis, and separation membranes. (1) MOFs are materials that self-
assemble from metal ions and multitopic organic ligands; because the particular
metal ion and ligand can be chosen by the chemist, this class of materials is
astonishingly diverse, with over 90,000 unique structures having been synthesized
to-date. (3) Most often, MOFs are synthesized as “bulk powders,” meaning that
their crystal size and shape is not controlled, and typically the diameteres of the
crystals are hundreds to thousands of nanometers. Hence, the materials are in a
powdered form that can be difficult to process for use in practical applications.
Considerable effort has been put towards controlling the size and shape of MOF
crystals to develop materials that, due to small particle size and good colloidal
stability, may be solution-processable. (4) Smaller crystals additionally boast a
shorter diffusion pathlength for molecules of interest, meaning that catalytic
efficiency and separation efficiency are often inversely related to particle size. (5)
Due to the compositional and structural diversity of MOF materials, each new
nanoparticle synthesis typically requires high-throughput testing methods; this
type of development is time-intensive. In this thesis, an approachable model to help
predict size trends in MOF nanoparticle syntheses is developed, then the model is
applied both to well-known and novel MOF nanoparticle systems.
1
Figure 1.1. Summary of work presented in this thesis. Classic MOF nanoparticles
(Cu3BTC2) and redox-active MOF nanocrystals (Fe(TA)2) are synthesized by
applying the Seesaw model. Fe(TA)2 nanoparticles are processed into thin films and
studied via electrochemical methods.
Chapter 2 encompasses a literature-based perspective on how the presence
and identity of a modulator will impact the final size of a MOF nanoparticle. MOF
nanoparticles can be synthesized via several top-down and bottom-up approaches.
One of the most prevalent bottom-up methods is to use “modulators,” molecules
added to the growth medium to change the reaction conditions and therefore the
crystals’ growth. Most often, modulators are monotopic ligands and are employed
as the acidic form of their conjugate acid/base pair. (4) The modulator method
has facilitated the development of many carboxylate, imidazolate, and azolate
2
MOF nanocrystals. (6) While it is not often clear how modulators affect crystal
growth, there are several in-situ reports that shed light on nanoMOF growth
mechanisms for some species. Rather than relying on complex in-situ analysis for
each individual synthesis, the “Seeesaw model” uses a system of coupled equilibria
to help the chemist intuit whether adding a certain modulator to their reaction will
result in an increase or a decrease in particle size.
In chapter 3, syntheses of prototypical MOF nanoparticles are explored
to support the Seesaw model. One classic example of a modulated nanocrystal
synthesis is that of of Zn(2-mIm)2 (ZIF-8), which simply uses excess of the MOF’s
linker (2-methylimidazole) to achieve smaller crystal sizes. (7) An equilibrium
model was proposed by Cravillon et al in 2009 to explain the effect of excess
ligand. (8) In-situ studies using SAXS found that small amorphous particles are
formed prior to crystallization. (9) Ex-situ studies by ESI-MS and PDF analysis
additionally show many small clusters in the reaction solutions. (10) The example
of ZIF-8 nanoparticles shows that, whereas simple equilibrium models can aid
in our understanding of synthetic conditions that yield nanoparticles, their
growth mechanisms are complicated and individual, requiring advanced analysis
to elucidate. The size control of ZIF-8 is used as an example in this chapter to
show how the system of equilibria can be de-coupled in certain situations. In
the case of the carboxylic acid MOF Cu3BTC2 (HKUST-1), there are a handful
of existing modulated syntheses. Basic modulators such as sodium formate and
trimethylamine have been shown to decrease particle size, while acidic modulators
such as dodecanoic acid increase particle size. (11;12) By using a mixture of an
conjugate acid/base pair, we show that not only can both of these seemingly
conflicting trends can be observed by changing a single variable, but that both
3
modulator equivalents and proton activity play a role in determining final particle
size (Fig. 1.1).
With evidence that the Seesaw model can be used to predict MOF particle
size trends in modulated syntheses, chapter 4 moves forward to develop a novel
class of nanoMOFs. In the realm of redox-active metal-organic frameworks,
there are only limited synthetic studies in which crystal size is controlled.
Despite this, evidence suggests that crystal size and shape influence the bulk
electronic properties of MOF materials. For instance, the electronic conductivity
of Cu3(HHTP)2 and Ni3(HITP)2 was first discussed as being dependent on
crystal size and shape in 2019 (HHTP = 2,3,6,7,10,11-hexahydroxytriphenylene;
HITP = 2,3,6,7,10,11-hexaiminotriphenylene). (13) A single rod-shaped crystal
of Ni3(HITP)2 shows metallic behavior, while a polycrystalline pellet shows
semiconducting behavior. This experiment confirmed computational results
that implied the material is metallic. (14) A variety of defects, including grain
boundaries, in the polycrystalline material create a band-gap that is not present
in the single crystal. Additionally, measurements of rods and exfoliated flakes
of Cu3(HHTP)2 show that both in-plane and out-of-plane conductivity are of a
similar order of magnitude, indicating that charge carriers flow through the 2D
sheets in addition to hopping between them. In a separate study, the conductivity
of single crystals of Ni3(HHTP)2 show an increase of two orders of magnitude over
polycrystalline Ni3(HHTP)2.
(13) Only two other studies delve into the morphology
control of conductive MOFs, in which Cu3(HHTP)2 is modulated with pyridine,
ammonia, and DMF. (15;16) Most studies are limited to top-down methods, by which
a variety of thin film materials have been made. (17;18) Generally, the perception of
the utility of conductive MOFs has been dependent upon the difficulty of single
4
crystal synthesis. This roadblock, however, appeared as an opportunity to create
a novel class of semiconductor nanocrystals that exhibit permanent porosity. The
conductive MOF Fe(1,2,3-triazolate)2 was first reported in 2012.
(19) Since then,
the bulk material has been studied for its redox-hopping conductivity that is
dependent upon the mixed valency of Fe2+/3+ centers, as well as its spin-crossover
behavior. (20;21) This material was identified as a promising nanoMOF candidate
since it is one of few 3-D connected conductive MOFs and the synthesis of the
material is relatively simple and results in a phase-pure product (Fig. 1.1). Finally,
in chapter 5, the redox chemistry of iron triazolate nanoparticles is studied in-
depth, showing how particle size can be used as a tool to tune redox activity. This
work shows that, to consider using conductive MOFs in real-world technological
applications, processability, particle size, and thin film thickness are important
practical considerations.
In this thesis, the development of the Seesaw model led to the synthesis
of a novel family of nanoMOFs that show size-dependent optical and electronic
properties. The materials Fe(1,2,3-triazolate)2, Co(1,2,3-triazolate)2, and Cd(1,2,3-
triazolate)2 were synthesized as nanoparticles of different sizes for the first time.
It was found that Fe(TA)2’s optical absorption, thin film conductivity and redox
properties depend on particle size.
5
CHAPTER II
SIZE CONTROL OVER MOF NANOPARTICLES
This chapter includes an excerpt from previously published and co-authored
material from Marshall, C.R.; Staudhammer, S.A.; Brozek, C.K. Size control over
metal–organic framework porous nanocrystals. Chem. Sci. 2019, 10, 9396-9408.
This work was conceptualized and developed by Checkers R. Marshall and Carl K.
Brozek. The perspective was written by Checkers Marshall, with data collection
assistance from Sara Staudhammer and with editorial assistance from Professor
Carl K. Brozek.
Nanocrystals are distinguished from their bulk counterparts by the extreme
size-dependence of their functional properties. For example, catalytic activities of
metal nanoparticles, (22) nanocrystal plasmon resonance energies, (23;24) and quantum
dot absorption and emission profiles in photovoltaic, solar fuel, and luminescence
technologies (25;26) reflect underlying electronic structures sensitive to sub-nanometer
size variations. Tailoring nanocrystals to a given application therefore relies on
generating particles with precise diameter values and uniform size distributions.
Since the advent of reliable synthetic methods, inorganic nanocrystals of metals (27)
and semiconductors (28;29;30) have found widespread use as advanced materials in
diverse areas, whereas design principles for organic-inorganic hybrid nanomaterials
are just emerging.
Recently, considerable efforts have focused on exploring the nanoscale
synthesis of metal-organic frameworks (nano-MOFs) due to the promise of their
heightened performance in drug delivery, (31;32;33) catalysis, (34) membrane design
for gas storage and separation, (35;36;37) and analyte sensing. (38) As 3D porous
coordination polymers comprised of inorganic clusters bridged by multi-topic
6
organic linkers, MOFs display immense modularity that has given rise to more
than 20,000 unique bulk phases, (39) each with the potential to adopt enhanced
functionalities when prepared as nanocrystals. (40) To advance this research
frontier, we must identify synthetic targets and universal mechanistic principles.
Building on the publication of recent reviews (41;42;43;44) and rigorous mechanistic
studies, (9;45;46;47;48) we identified key open questions: Which MOFs have been
prepared as nanocrystals? Which sizes are achievable? And Which mechanistic
parameters govern nano-MOF sizes? Here, we address these outstanding questions
by compiling experimental parameters and particle sizes from across the nano-
MOF literature; statistically treating reported size data (see Methods section
below); comparing nano-MOF sizes, size-measurement techniques, and synthetic
conditions; and identifying underlying chemical principles from observed trends.
Whereas recent reviews (49;1;41;50) have compared the impacts of varying synthetic
techniques, such as microwave versus solvothermal, and conditions, such as
time and temperature, we target the generalized chemical equilibria and kinetic
pathways universal to nano-MOF syntheses.
7
NanoMOF Metadata
Figure 2.1. Summary of all MOF materials reported to-date as nanocrystals with
precisely measured particle diameters. Average and median sizes are included
using all reported literature values for each MOF material. Smallest known sizes
for each MOF are labelled according to the corresponding synthetic method,
i.e., coordination modulation (CM), metal-organic gel (gel), slow addition (SA),
and ionic liquid microemulsions (ILM). See methods section for details of data
treatment. All tabulated values are included in Table A.1.
Figure 2.1 summarizes all nano-MOFs we identified with quantifiable size
diameters, plotted by average, median, and smallest sizes (listed in Table A.1),
and in Table 2.1 with MOF compositions and experimental details. These data
indicate that while many MOF materials have been accessed as nanocrystals, the
vast majority have not. Furthermore, Fig. 2.1 suggests that typical nano-MOF
sizes exist on the 100 nm scale, with few extending below 20 nm, in contrast to
the 1–10 nm diameters achievable for inorganic nanocrystals. (28) For most MOF
materials, select studies have achieved sub-100 nm diameters, but these cases are
exceptions, as size averages and median values are far larger. For each class of
MOF materials displayed in Fig. 2.1, the smallest size provides the current state-
8
of-the-art in minimizing nanocrystal sizes, median values indicate the most likely
achievable sizes when using the coordination modulation synthetic method, and
average values lend insight into the distribution of reported values for each given
class of MOF nanocrystals.
Figure 2.2. Size comparisons of HKUST-1 nanocrystals prepared by (A) microwave-
assisted growth at varying reactant concentrations and added equivalents of
dodecanoic acid and (B) by solvothermal synthesis at a fixed reactant concentration
of 0.0024 M and varying equivalents of triethylamine (TEA) or acetate (OAc)
modulators. (51;52;53) The nanocrystal sizes in these studies were determined by TEM
(A) and PXRD (B).
Interestingly, compiling size data for a given MOF revealed that often
the most impactful size determinants were those that changed between separate
synthetic investigations, rather than the parameters systematically explored
within isolated studies. For example, Figure 2.2 shows a portion of data compiled
9
for nanoscale HKUST-1 (Cu3(BTC)2(H2O)3). Clearly, the differing reaction
conditions between the results in panel a versus panel b had a greater impact
on the nanocrystal sizes compared to the minor impact of copper salt and added
base identities shown in panel b. Either the differing reactant concentrations (2.34
mmol versus 0.17 mmol), solvent conditions (DMF/H2O/EtOH versus butanol),
or solvothermal versus microwave synthetic routes involved distinct processes that
produced stark size differences. In response to such cases, we focus our mechanistic
analysis on reports that employed “coordination modulators”—typically monotopic
acid ligands—as these represent the bulk of literature examples, although small
particles of MOFs have been generated by many other techniques, such as
preparation via microemulsions, (54;55) dual injection, (52) and metal organic gels. (56)
Reliable preparation of small nano-MOF particles depends on a firm
mechanistic understanding of nano-MOF initiation, growth, and termination.
Typically, nano-MOF syntheses are discussed (43;45;57;50) in terms of the LaMer
model of particle growth, (58) which separates crystal nucleation from growth,
and describes both in terms of thermodynamic driving forces triggered by high
precursor concentrations. In-situ data suggest that MOF-5 (Zn4O(BDC)3) may
follow this model, as nucleation and growth appear to be effectively separated. (45)
However, systems such as HKUST-1 and ZIF-8 (Zn(Hmim)2) behave differently,
exhibiting slow nucleation phases that overlap with growth. (59;60) A collection of
in-situ XRD studies of MOF crystal formation revealed no significant difference in
the time scales between nucleation and growth phases, implying that both processes
can occur simultaneously. (61) Furthermore, the majority of nano-MOF syntheses
occur under dilute conditions (Table A.2).
10
We argue, therefore, that while thermodynamics remain central to
understanding MOF crystal nucleation and growth, nano-MOF sizes are kinetically
controlled by chemical parameters that arrest particle growth. In particular, the
critical conditions for ensuring small nano-MOF sizes involve depleting the local
concentrations of reactant metal ions, thereby allowing linkers and monotopic
modulators to trap nano-MOF particles. Analysis of the literature reveals that
ideal conditions involve excess ligand (linker or modulator), dilute reactant
concentrations, strong metal-ligand bonds, and low proton activities. In this
perspective, we support this kinetic model with literature examples that illustrate
the role performed by each parameter and apply this insight to rationalizing
previously unexplained phenomena.
Table 2.1. A guide to MOF names, their component linkers and metals, and
modulators that have been shown to access nanoscale particle size.
MOF Names Linker Metal Source Effective Modulators
COMOC-4, 2,2’-bipyrimidine-5,5’- GaNO3 * H2O none
(62)
MOF-253 dicarboxylic acid
DUT-23 4,4,4”-benzene-1,3,5- CuNO3 * H2O none
(63)
triyl-tribenzoic acid,
4-4’-bipyrimidine
DyBTC 1,3,5-benzenetricarboxylic acid DyNO3 * H2O acetic acid
sodium acetate (64)
Fe-soc-MOF 3,3’,5,5’-azobenzenetetra- Fe(NO3)3 * 9H2O sorbitan trioleate (tween-85)
(65)
carboxylic acid
HKUST-1, 1,3,5-benzenetricarboxylic acid Cu(NO3)2 or Cu(Oac)2 dodecanoic acid
(51)
Cu-BTC acetic acid
sodium acetate (64)
w/ triethylamine (53)
2-methylimidazole (66)
PAA (67)
IR-MOF-3 2-aminoterepthalic acid Zn(NO ) PVP (67)3 2
PVP + TMAB (68)
MFU-4 1H,5H-benzo(1,2-d:4,5- ZnCl (69)2 lutidene
d’)bistriazole NaOH, KOH (33)
MIL-88A fumaric acid FeCl * 6H O NaOH (70)3 2 NaOH, acetic
acid (71)
MIL-88-Fe 1,4-benzenedicarboxylic acid FeCl3 * 6H2O acetic acid
(72)
MIL-88B-NH2 1,4-benzenedicarboxylic acid FeCl
(73)
3 * 6H2O acetic acid, F127
11
MOF Names Linker Metal Source Effective Modulators
MIL-96 1,3,5-benzenetricarboxylic acid Al(NO3)3 trimethyl-1,3,5-
benzenetricarboxylate (74)
MIL-100-Al 1,3,5-benzenetricarboxylic acid Al(NO3)3 benzoic acid,
trimethyltrimesate (75)
MIL-100-Cr 1,3,5-benzenetricarboxylic acid Cr(NO3) none
(76;75)
3
MIL-100-Fe 1,3,5-benzenetricarboxylic acid FeCl3 none
(75)
MIL-101 1,4-benzenedicarboxylic acid Cr(NO3)3 none, stearic acid,
4-methoxybenzoic acid,
benzoic acid,
4-nitrobenzoic acid,
perfluorobenzoic acid (77)
acetic acid (78)
benzoic acid, HF, TMAOH,
sodium hydroxide (79)
benzoic acid (80)
(Fe) 2-methylimidazole (81)
MIL-125 1,4-benzenedicarboxylic acid Ti(OCH(CH3)2)4 poly(ethylene glycol) diglycidyl
ether (82)
MIL-125-NH2 2-aminoterephthalic acid Ti(OCH(CH3)2)4 benzoic acid
p-toluylic acid
none (83)
MOF-5 / 1,4-benzenedicarboxylic acid Zn(NO3)2 Acetate (counterion)
(64)
IR-MOF-1 decylbenzoic acid (84)
TEA (85) TEA+PVP (86)
TEA+CTAB (87)
p-perfluoromethyl-
benzenecarboxylate (45)
MOF-74, 1,4-benzenedicarboxylic acid Cu(NO3)2, Ni(NO3)2 benzoic acid, acetic acid
(88)
CPO-27 2-methylimidazole (81)
NU-1000 (1,3,6,8-tetrakis(p-benzoic ZrOCl2 benzoic acid +
acid)pyrene trifluoroacetic acid (89)
biphenyl-4-carboxylic acid (90)
acetic acid (91)
NU-1003 1,3,6,8-Tetra(6- ZrOCl2 benzoic acid + trifluoroacetic
carboxynaphthalen-2-yl)pyrene acid (83)
PCN-222, tetrakis(4- ZrOCl (89)2 benzoic acid, dichloroacetic
MOF-525 carboxyphenyl)porphyrin acid (91)
PCN-224 tetrakis(4- ZrOCl2 benzoic acid
(92;57)
carboxyphenyl)porphyrin
UiO-66 1,4-benzenedicarboxylic acid ZrOCl2 trifluoroacetic acid, dichloroacetic
acid, acetic acid, formic acid (93)
UiO-68 Biphenyl-4,40-dicarboxylic acid ZrOCl2 Benzoic acid, acetic acid
(94)
(BPDC)
ZIF-7 benzimidazole Zn(NO3)2 * 6H2O Polyethyleneimine
(95)
ZIF-71 4,5,-dichloroimidazole Zn(NO3)2 * 6H2O n-butylamine + 1-mIm
(96)
ZIF-8 2-methylimidazole Zn(NO3)2 * 6H2O Excess linker
(8)
n-butylamine (59)
12
MOF Names Linker Metal Source Effective Modulators
Zn-BDP 1,4-Bis(1H-pyrazol-4-yl)benzene Zn(OAc)2 * 2H2O None
(97)
Factors Controlling MOF Nanocrystal Sizes
We propose that the kinetic trapping of MOF nanocrystals of particular
sizes depends on the competition between four chemical equilibria (Figure 2.3):
1) linker deprotonation; 2) modulator deprotonation; 3) linker complexation, and
4) termination. Equilibria with fast forward-direction rates and low reversibility
dictate whether MOF particles steadily grow toward bulk phases or arrest quickly
to form small nanocrystals. MOF linkers must deprotonate (Eq. 1) before forming
metal-linker bonds. Modulators are usually acids, and so must also be deprotonated
(Eq. 2). Complexation between metal ions and linkers facilitates particle
growth (Eq. 3). Reports suggest that early in MOF growth, large collections of
molecular complexes and oligomers develop in solution before coalescing into MOF
particles. (10) Subsequent MOF growth is then dominated by the arrival of oligomer
clusters or solvated reactant molecules. (98) During the final termination step (Eq.
4), linker and modulator ligands compete for metal ion coordination sites.
Figure 2.3. Key chemical equilibria controlling nano-MOF growth and termination.
According to our kinetic model, this process continues until the local
concentration of ligands far exceeds the metal ions, thereby arresting particle
growth. In addition to these four chemical processes, the assembly of cluster nodes
13
and solvent decomposition have also been invoked to discuss nano-MOF nucleation
and growth, (99) but we focus on the most general processes that dominate
particle trapping. Critical analysis of nano-MOF sizes and synthetic conditions
reveal the existence of key parameters that may be programmed to deplete local
concentrations of metal ions and generate small particle sizes: modulator identity
and concentration, equivalents of linker or modulator, and metal–ligand bond
strengths.
Modulators. Modulators are typically monotopic carboxylic acids
and occasionally Brønsted bases added to nano-MOF syntheses. The intended
purpose of modulators varies, but we propose that their function is to influence
nano-MOF sizes by affecting linker deprotonation and arresting particle growth. (51)
Modulators also act to prevent particle aggregation. Although modulators produce
size trends that appear complex and contradictory, their role can be rationalized
in terms of the four equilibria outlined above. When strong Brønsted bases
are used as modulators, their primary role is to facilitate ligand deprotonation
(Eq. 1) and enhance metal-linker complexation (Eq. 3) relative to metal-ion
diffusion, thereby depleting local metal ion concentrations and forming small MOF
nanocrystals. For example, nanocrystals of MFU-4 (Zn5Cl4(BBTA)3) decrease in
size with added lutidine or KOH. (69) Similarly, when nanocrystals of NU-1000
(Zr6(µ3–OH)8(OH)8(TBAPy)2) are prepared with the addition of 4-biphenyl-
carboxylic acid, particle sizes decrease further if NaOH is added to the precursor
linker solution. (90) and IR-MOF-3 (Zn4O(TPDC)3) require triethylamine (TEA),
which become more uniform with initial addition of cetyltrimethylammonium
bromide (CTAB). (87) Similarly, including n-butylamine decreases nanocrystal
sizes of ZIF-71 (Zn(Hdcim) ). (10)2 Interestingly, nanoparticles of MIL-101(Cr)
14
(Cr3(H2O)2O[(C6H3)(CO2)3]2) are synthesized without any modulator by simply
decreasing the amount of HF, which is used as a mineralizing agent in the
traditional bulk synthesis. (100;80;78) Adding a strong base to the reaction mixture,
however, results in smaller particle sizes. (79)
When carboxylic acids serve as modulators, their presence can increase
or decrease nano-MOF sizes depending on whether they impede linker
deprotonation (Eq. 1) or act as surface capping ligands (Eq. 4). By interfering
with deprotonation, they slow down metal-linker complexation (Eq. 3) relative to
metal-ion diffusion, resulting in large nano-MOF sizes. On the other hand, they
can terminate particle growth by acting as surface-capping ligands and produce
small sizes. For example, Figure 2.4a shows that while adding 0.33 equivalents of
perfluorobenzoic acid generates larger MIL-101 particles relative to using no HF or
modulator, the addition of more weakly acidic 4-nitrobenzoic acid, benzoic acid,
4-methoxybenzoic acid, and stearic acid decreases particle sizes with increasing
modulator pKa values. (101) The less acidic the modulator, the lower the H+ activity
in solution available to protonate linker molecules (Eq. 1).
Adding small quantities of acidic modulators decreases nano-MOF sizes
until the H+ activity in solution reaches a threshold value that begins to interfere
with linker deprotonation (Eq. 1). Further addition of acid slows metal-ligand
complexation relative to metal-ion diffusion, leading to large particle sizes. For
example, Figure 2.4b serves as a useful comparison to the data in Figure 2.4a.
Both studies were conducted at similar concentrations (0.076 M versus 0.033 M)
and both involve similarly strong metal-ligand bond strengths (Zr4+-carboxylate
and Cr3+-carboxylate) but whereas 0.33 modulator equivalents were employed in
Figure 2.4a, much higher quantities were involved in Figure 2.4b. The data show
15
that UiO-66 (Zr6O6(BDC)6) nanocrystal sizes increase with additional modulator.
Interestingly, modulators with lower pKa values produce larger particle sizes at a
given amount of added modulator. For instance, 15-20 equivalents of trifluoroacetic
acid (TFA) or dichloroacetic acid (DCA) produce 200-nm UiO-66 nanocrystal sizes,
whereas twice that amount of acetic and formic acid are needed. Acidic modulators
hinder metal-ligand complexation (Eq. 3) relative to metal-ion diffusion so that
particles continue to grow. Indeed, adding thousands of equivalents of formic
acid to the synthesis of UiO-66 generates single crystals hundreds of microns in
diameter. (102) This kinetic explanation fits many other studies in which particle
sizes increase with additional acidic modulator, (51;103;94;104) including HKUST-1
modulated by dodecanoic acid, (51) PCN-224 (Zr-TCPP) with benzoic acid, (103)
UiO-66 with benzoic acid (94) and MIL-88B-NH2 ([Fe3O(BDC-NH2)3(H2O)2) with
acetic acid. (104)
Concentrated reaction conditions necessitate the addition of modulator;
otherwise, rapid metal-ion diffusion due to short effective pathlengths outcompetes
growth termination (Eq. 4). Indeed, most nanoscale MOF syntheses rely on dilute
conditions (Table A.2). For example, synthesis of MIL-101-Cr involving high
concentrations (0.2 M H2BDC) produces small particle sizes only with addition
of small quantities of benzoic acid (Fig 2.3c). (80) Benzoic acid has a greater effect
than acetic acid on decreasing particle sizes at such high reactant concentrations,
suggesting that under these reaction conditions, interfering with metal-ligand
complexation is critical to kinetically trapping small MIL-101-Cr nanocrystals.
16
Figure 2.4. Nanoscale MOF sizes depend on the equivalents and pKa values
of added modulator reagents. (A) MIL-101(Cr) nanocrystal sizes decrease
with increasing modulator pKa values. Sizes were determined by TEM. (101)
(B) As modulator equivalents increases, sizes of UiO-66 particles increase.
(TFA: trifluoroacetic acid, DCA: dichloroacetic acid, FA: formic acid, and AA:
acetic acid). Shaded box provided to emphasize sizes below 200 nm. Sizes
were determined with STEM and DLS (DLS not shown). (105) (C) MIL-101-Cr
nanocrystal sizes decrease with increased modulator equivalents, while MIL-88B-
NH2-Fe exhibits the opposite trend. Interestingly, MIL-88B microcrystals are
formed as an impurity at and above 5 benzoic acid equivalents (orange). Sizes
were determined with SEM (orange and pink) and TEM (blue). (80;78;106)
17
Phase purity must be considered when choosing modulator equivalents
and reaction concentrations. For example, while adding few equivalents of either
acetic or benzoic acid in the synthesis of MIL-101 at high concentrations results
in phase-pure MIL-101 nanocrystals, greater equivalents induce the formation
of mixed-phase products (107) because MIL-101 and MIL-88B occupy the same
reaction space, with both arising from Fe3+ or Cr3+ and trimesic acid. (108)
Therefore, at a benzoic acid : linker ratio of 10 : 1, only MIL-88B microcrystals
form. (79) Concentration plays an important role in controlling nanocrystal
phase purity as well. For example, MIL-101-Cr and MIL-88B-Fe nanocrystals
have been obtained with similar equivalents of acetic acid, but the synthesis
of MIL-88B-Fe was an order of magnitude more dilute (Fig. 2.4c). Such phase
transformations with variable modulator equivalents indicate the importance of
nonclassical growth mechanisms. (61) Similar phenomena have been observed for the
phases spaces involving MIL-100-Al (Al3×(H2O)20(BTC)2) / MIL-96-Al (Al12O-
(OH)16(H2O)5(BTC)6 MIL-110-Al (Al8(OH)12(OH)3(H2O)3(BTC)3) and NU-901
(Zr6(µ3-OH)8(OH)8(TBAPy)
(90;109;110)
2) / NU-1000.
Linker Equivalents. Excess linker equivalents shift equilibria toward
enhanced metal-ligand complexation (Eq. 3), thereby depleting local metal ion
concentrations (51) and arresting particle growth without added modulator (Eq. 4).
In other words, excess linkers serve as surface-capping ligands. The excess linker
method was first reported in 2009 for ZIF-8 and has since been used in further ZIF-
8 and ZIF-71 nanocrystal syntheses (Fig. 2.4). (8;10;59) Nano-MOF particle sizes can
be further reduced by adding Brønsted bases to enhance linker deprotonation (Eq.
1). (8) Irreversible ligand deprotonation may lead, however, to unchecked particle
growth through rapid metal-ligand complexation, unless counterbalanced by excess
18
surface-capping ligands—illustrating the intricate kinetic balance of the four key
underlying processes outlined in Eq. 1–4.
Although several chemical parameters may contribute to decreased nano-
MOF sizes, the impact of certain factors may dominate over others. For example,
linker excess was discovered to be the single strongest size determinant of ZIF-
8 nanocrystals through systematic investigations into the role of Brønsted base,
linker excess, and reactant concentrations (Fig. 2.5a,b). (8) Nanocrystals of ZIF-
8 can be synthesized using an excess of the linker 2-methylimidazole (2-mIm), (8)
whereas typical bulk syntheses of ZIF-8 combine the zinc salt and imidazole linker
in a 1:1 - 1:2 ratio. (111) Simply increasing the metal-to-linker ratio to 1:5 results
in nanocrystals sizes of 40 nm (Fig. 2.5a). (8) Reactant concentration was also
studied as a size determinant, with the data in Figure 2.5a showing that more
dilute systems lead to smaller ZIF-8 crystal sizes. In terms of our kinetic model,
the role of dilution is to increase metal-ion diffusion pathlengths, allowing particles
to be terminated in isolation from additional metal ions. The impact of added
base was also investigated, but only the basic modulator n-butylamine resulted in
reduced nanocrystal sizes, whereas less basic 1-methylimidazole and sodium formate
resulted in micrometre-sized crystals. (59) Nevertheless, compared to the impact of
dilution (Fig. 2.5a) and Brønsted base, the most significant decreases in ZIF-8 sizes
were achieved by linker excess (Fig. 2.5b). These systematic comparisons suggest
that growth termination is more important than linker deprotonation in controlling
ZIF-8 nanocrystal sizes.
19
Figure 2.5. ZIF nanocrystal syntheses with varying relative ratios of metal,
linker, modulator, and solvent. Synthetic variables are in bold. (Hmim: 2-
methylimidazole, Hdcim: 4,5-dichloroimidazole). (A) Dilution results in a series
of ZIF-8 nanocrystals sizes. (8) (B) Excess linker exerts a stronger influence than
base on nanocrystal sizes. (59) (C) Addition of n-butylamine rather than linker
excess exhibits biggest impact on ZIF-71 sizes. (10)
If linker deprotonation limits nanocrystal formation kinetics, however,
addition of Brønsted base will produce a greater effect than the equivalents of
excess linker. For example, systematic studies of ZIF-71 nanocrystal synthesis
indicate that in contrast to ZIF-8, the most influential variable is n-butylamine
equivalents (Fig. 2.5c). (10) When the linker-to-metal ratio is doubled from two
to four with base and concentration held constant, particle sizes remain around
80-100 nm. Increasing the proportion of base, however, reduces particle sizes to
approximately 20 nm. The sensitivity of ZIF-71 nanocrystal sizes to the equivalents
of added base results from the less acidic 4,5-dichloroimidazole linker.
Interestingly, rather than follow this excess linker strategy, most reported
nano-MOF syntheses rely on the same linker equivalents used in bulk syntheses
(Table A.2). On the other hand, select studies have shown that excess linker was
ineffective in generating nanoscale particles. Excess trimesic acid does not produce
20
HKUST-1 nanocrystals, for instance. (60) Although excess linker reduces the sizes
of UiO-66 particles, higher water content exerted the greatest size control, perhaps
due to its role in assembling the Zr4+-oxo cluster nodes. (112)
Metal-Linker Binding Strength. Strong metal-ligand interactions
favour small particle size because they enhance rates of both complexation (Eq.
3) and termination (Eq. 4) during nano-MOF growth, thereby depleting the local
concentrations of metal ions relative to linkers or modulators. Systematic studies
varying the metal identities of heterobimetallic materials illustrates the influence of
metal-ligand interactions on nanocrystal size. For example, higher Co2+ contents in
Zn2+-based ZIF-8 nanocrystals results in larger nanocrystals (Fig. 2.6). (113) Using
Cu2+ further accentuates this effect, with comparatively larger sizes produced at
identical dopant metal concentrations. (114) Because linker-to-metal ratios remained
Figure 2.6. Heterobimetallic ZIF-8 nanocrystals increase in size as the Zn2+
atoms are substituted for Co2+ or Cu2+ atoms. Insert: highlighted data at low
equivalents, where identical Co2+ and Cu2+ quantities produce different particle
sizes. Particle sizes were determined by TEM (main) and SEM (inset). (113;114)
constant in these experiments, the increase in size with lower Zn2+ content can
be attributed to the strong Zn2+-imidazolate interactions, which quickly produce
small particles unless harder ions such as Cu2+ interfere. Similarly, differences in
21
metal ion labilities were invoked to explain why MOF-74 (M2(DOBDC)) crystals
nucleate and grow faster with Zn2+ than with Co2+. (98) Surprisingly, cobalt-doped
UiO-66 nanoparticles are smaller in size than their zirconium-only counterparts
when synthesized under otherwise identical conditions. (115) As the strength of the
zirconium–carboxylate bond is expected to be stronger than cobalt-carboxylate
bonds, metal-linker complexation rates may not be the only equilibrium to
consider. For instance, weaker bonds might slow particle growth, allowing diffusing
linkers trap the cobalt variants at smaller sizes. To date, there have been few
studies regarding the effect of mixed metals on MOF nanocrystal size and this area
warrants further exploration.
Figure 2.7. Reaction conditions that favor small or large MOF nanocrystal sizes
when linker or acidic modulators are present in excess.
Summary. The metal-ligand chemistry outlined in Eq. 1–4 provides a
framework for understanding trends in reported nano-MOF sizes. Based on these
insights, Figure 2.6 offers a general guide for designing small MOF nanocrystals.
Excess linker or acidic modulator generally reduce nanocrystal sizes unless either
metal-linker complexation far exceeds termination kinetics or if acid addition
inhibits linker deprotonation. Dilute reactant concentrations paired with low
proton activities ensure small particle sizes by enhancing complexation (Eq. 3) and
22
termination (Eq. 4), while isolating particles from diffusing metal ions to prevent
runaway growth.
The Seesaw Relationship
Seemingly incompatible trends reported for nano-MOF sizes can be
reconciled by viewing nano-MOF growth as a balance between reactant
concentration, linker and modulator deprotonation, metal-ligand interactions,
and metal-ion diffusion. Figure 2.8a summarizes our model into two regimes. In
regime I, small quantities of acidic ligands (either modulators or linkers) decrease
particle sizes by supplying surface-capping ligands, overwhelming local metal ion
concentrations. Higher quantities of acidic ligands further decrease nanocrystal
sizes by increasing the rate of metal-ligand complexation relative to metal-ion
diffusion. This trend continues until reaching minimum nanocrystal sizes α at
threshold values of added acidic ligand ε (Fig. 2.8a). This critical point corresponds
to a minimum of relative ratios between local metal ion-to-ligand concentrations β
and ratios of relative rates of diffusion and metal-ligand complexation σ (Fig 2.8a).
In Regime II, additional equivalents of acidic ligands raise solution proton activities
such that they interfere with linker deprotonation. As a result, nanocrystal sizes
increase with additional acidic ligand because metal-ion diffusion rates outcompete
particle termination. This “Seesaw” relationship between nano-MOF sizes and
relative termination versus diffusion rates strikes a balance precisely where particles
sizes are at a minimum.
23
Figure 2.8. The “Seesaw” relationship between nanocrystal sizes and added
equivalents of acidic ligands. Nanocrystal sizes increase with higher ratios of metal-
to-linker local concentrations (A) particles reach a minimum size α at critical
values of acidic ligand ε and minimum relative ratios local metal ion-to-ligand
concentrations β and relative ratios diffusion and metal–ligand complexation ratios
σ. (B) MIL-125-NH2 and UiO-66 exhibit the full Seesaw relationship curve in
trends between particle sizes and equivalents of p-toluic acid. (116)
Figure 2.8b summarizes data exhibiting a Seesaw curve for NH2-MIL-125
and UiO-66 sizes with varying quantities of p-toluic acid. (116) For both materials,
particle sizes at first decrease, bottom-out at minimum values, and then increase
with higher quantities of modulator. Competition between complexation and
metal-ion diffusion may explain Seesaw curves in related phenomena, such as the
polymorphic balance between MIL-101 and MIL-88B phases achieved by tuning
benzoic acid equivalents (Fig 2.4c). (80) Low benzoic acid equivalents produce
decreasing sizes of MIL-101 until higher equivalents lead to large micron-sized
particles of MIL-88B instead. We propose that most reported trends of nano-
24
MOF sizes capture just portions of the entire curve of the Seesaw relationship.
Type-I behavior, where added acidic ligand decreases particle sizes, is observed
for the MIL-101(Cr) and ZIF-8 syntheses discussed above (Fig. 2.4a, 2.5a). (117;78;8)
Examples of regime II relationships behavior, where particle size increases with
respect to increasing modulator equivalents, have been observed for many MOFs
when monocarboxylic acids are added, including reports on UiO-66, HKUST-1,
PCN-224, and MIL-88B-NH . (51;103;94;104)2 Based on numerous reports exhibiting
regime II behavior, the curvature of the slope region II appears proportional to
modulator acidity such that highly acidic ligands produce larger particle sizes at
fixed equivalents. The impact of highly acidic modulators is so pronounced that
they can halt particle growth entirely, whereas less acidic modulators added in large
excess simply promote large particles.
Whether type-I or type-II behaviors emerge for a given MOF material
depends on the similarities between the particular complexation, termination, and
metal-ion diffusion rates. Although the Seesaw curve involves kinetic trapping,
the extreme limit at the far right of the curve involves particles grow over much
longer time periods due to sluggish metal–ligand complexation that places MOF
crystal growth in an entirely different regime determined by thermodynamics.
According to this model, defects incorporated in nano-MOFs must be kinetically
trapped, whereas defects in macroscopic MOF single crystals arrive through
thermodynamically driven processes. This model helps explains why addition of
strong acid helps to produce large single crystals of MOFs. (118)
25
Best Practices and Outstanding Challenges
Elevating the rigor of MOF nanocrystal synthesis will require addressing
critical challenges. Here, we offer recommendations on synthetic, characterization,
and data-reporting methods to facilitate future MOF nanocrystal investigations.
Colloidal Stability. Applying MOF nanoparticles in applications such
as drug delivery requires that the particles be colloidally stable. A re-dispersed
nanoparticle solution may suffer from significant aggregation or coalescence without
sufficient surfactant ligand coverage. Measurements of zeta potentials provide useful
information on the charge at the nanoparticle surface, such that values far away
from zero indicate that a dispersion is stable. (105) DLS (dynamic light scattering)
measurements may be used to determine colloidal stability, as it is a solution-
phase size measurement method. Aggregating particles observed by DLS display
unusually high hydrodynamic radii. Additionally, further growth or aggregation
causes the apparent sizes to increase over time.
Incorporation of Modulator. Identifying the presence and location
of modulators in nano-MOFs is important in determining whether they serve
as surface-capping ligands or form internal defects. Mirkin et al. found that
while the colloidal stability of UiO-66 crystals correlated to the identity and
amount of modulator, the exact role of modulators at the particles surfaces was
unclear. (105) For example, while small equivalents of weakly acidic modulators
resulted in aggregation, nanoparticles of UiO-66 have been synthesized without
any monocarboxylic acid modulator. (112)
A common method to quantify ligand incorporation in MOFs is to
perform acid digestion NMR studies. The linker-to-modulator ratio in the MOF
can be elucidated through 1H-NMR peak integration. (119) Defects may also be
26
identified as a weight percent by thermal gravimetric analysis (TGA). (87;120) The
relative incorporation of a modulator depends on its function during synthesis.
For instance, in the synthesis of ZIF-8 with n-butylamine, less than 1 percent
incorporation is observed by 1H NMR. The absence of incorporated modulator
indicates its primary role is to deprotonate the linker, rather than cap particles
during termination. (96) When the ratio of modulator is higher than would be
expected for surface passivation, it must either be creating defects in the MOF
particle, or be present as a guest. For example, a reported synthesis of UiO-
66 modulated with benzoic acid revealed an 8:10 benzoic acid-to-linker ratio,
even after extensive washing. (94) The amount of modulator incorporated can
depend on pH, as one NMR digestion study of UiO-66 showed that acetic acid
incorporation first decreased, then increased, with respect to the amount of
triethylamine added. (119) The authors speculated that amount of deprotonated
MOF linker BDC (benzenedicarboxylate) in the reaction was maximized at the
minimum of the acetate incorporation curve. Interestingly, modulator incorporation
observed in this study exhibits a U-shaped curve, indicating that modulator defect
concentrations can be minimized at a critical amount of added modulator. This
U-shape does not correspond to size, however; the size monotonically decreases,
indicating the minimum size and minimum defect concentration occur with
different quantities of added modulator. Due to the insight obtained from these
studies, we recommend acid digestion NMR studies as a standard method to
characterize MOF nanoparticles. We expect synthetic methods to advance toward
finer levels of control as trends emerge from the impact modulators have on defect
incorporation and nanocrystal size.
27
Measurement Methods. Size analysis of MOF nanocrystals relies on
appropriate use of structural characterization methods, as has been discussed in a
previous review. (46) Typical techniques include PXRD (powder X-ray diffraction),
microscopy, DLS, and SAXS (small angle X-ray scattering). In general, we
recommend reporting data from at least two complimentary methods, even when
data contradict.
According to the Scherrer relation, the full-width-at-half-max of a given
PXRD peak relates to the particle size. Although smaller particles will exhibit
broader diffraction peaks in general, peak broadening may result from several
factors, such as lattice stress or instrument effects. (121) Several peaks should
be modelled to determine reliable size estimates. When considering polyhedral
crystals, shape factors should be chosen to match the particle morphology and
specific miller index of the peak under consideration. (121) Crystallographic domain,
not particle, sizes, are estimated by this method. Aggregated particles comprise of
multiple domains, which leads to conflicting data between PXRD and other sizing
techniques. (122)
Dynamic light scattering (DLS) overestimates particle sizes because the
method determines the hydrodynamic radii of particles. Authors often attribute
size overestimates from DLS to aggregation. Recent reports have suggested that
MOF porosity may induce unconventional diffusion behaviour, which would
hamper analysis by DLS. (112) The interpretation of DLS relies on the assumption
that particles are hard spheres that move in solution via Brownian motion. (123)
Irregularly shaped particles or porous particles defy these simplified models. (124;125)
Several advanced models exist that describe hollow nanoparticles, although
these too may be inadequate for describing the complex microporosity of MOF
28
particles. (126) The key utility of DLS is in developing biological applications of large
particles, where it can effectively identify the presence of microscopic aggregates in
solution. (127)
Microscopy is the most common method to determine particle sizes. Both
SEM (scanning electron microscopy) and TEM (transmission electron microscopy)
are widely used, although they rely on high-energy electron beams that can
compromise MOF structural integrity. (46) Microscopy finds its greatest advantage
in probing particle morphology, although the 2D projections of 3D particle
shapes should be considered carefully. (8) Furthermore, analysis must be applied
to statistically relevant ensembles of particles. It is essential to report the size of
the population used to estimate size and size distributions; these values are often
missing in the literature.
SAXS is a less common technique, but it presents several advantages:
SAXS measures solution-state samples without overestimating sizes and it
examines statistically relevant populations. (59) Accurate analysis relies on choosing
appropriate approximations and form factors. (128) Although size and porosity of
hollow nanoparticles can be accurately determined by SAXS, nano-MOFs lack
a generally accepted model due to their complex topology. (126) The model used,
and any other relevant data analysis, should be rigorously reported. In general,
critical treatment of particle size data is essential to rigorous investigations into the
structure–property relationships of MOF nanocrystals.
Conclusions. MOF nanocrystal sizes and synthetic conditions were
critically analysed from across the literature to develop a deeper mechanistic
understanding of nanocrystal formation. A general model was presented that
reconciles seemingly contradictory trends for MOF nanocrystal sizes versus
29
common synthetic parameters: excess ligand, additional acid or base, reactant
concentrations, and metal ion identities. A universal “Seesaw” relationship is
proposed that relates nano-MOF sizes to a competition between particle growth
facilitated by diffusing metal ions and particle termination by depleting metal ion
local concentrations through rapid ligand complexation. Therefore, conditions that
favour high relative concentrations of ligands and that maximize metal-ion diffusion
pathlengths produce the smallest nano-MOF sizes. This model also sheds light
on the mechanism of MOF crystal growth, in general, and provides a framework
for designing macroscopic single crystals. By compiling data for all known MOF
nanocrystals, we define the goalposts for future nano-MOF synthetic targets and
provide a mechanistic model rooted in chemical parameters that may be tuned to
discover the full potential of this emerging class of nanomaterials.
30
CHAPTER III
EXPERIMENTAL EVIDENCE FOR THE SEESAW MODEL
This chapter includes an excerpt from previously published and co-authored
material from Marshall, Timmel, E.T.; C.R.; Staudhammer, S.A.; Brozek, C.K.
Experimental evidence for a general model of modulated MOF nanoparticle growth.
Chem. Sci. 2020, 11, 11539-11547. This work was conceptualized and developed
by Checkers R. Marshall and Carl K. Brozek. The article was written by Checkers
Marshall, with experimental and analytical assistance from Emma Timmel and
Sara Staudhammer, and with editorial assistance from Professor Carl K. Brozek.
Precise size control can yield distinct functional behavior from materials
with seemingly similar compositions. Achieving control at the nanoscale,
in particular, has uncovered remarkable size-dependent properties, such as
the luminescence of quantum dots and the distinct catalytic activities of
metal nanoparticles. (26;22) Recent reports suggest that the rich structural and
compositional diversity of bulk metal-organic frameworks (MOFs) produces
enhanced functional properties when realized on the nanoscale. (44) For example,
advanced MOF-based gas separation technologies use nanoparticulate MOFs
(nanoMOFs) dispersed into mixed matrix membranes (MMMs) to achieve enhanced
efficiencies over bulk phases. Remarkably, MMMs that employ nanoMOFs have
been shown to surpass the Robeson limit—an intrinsic trade-off between selectivity
and permeability in separation membranes. (129;130) While the gas separation
performance of nanoMOFs has attracted industrial interest, their improved
activities as atomically defined catalysts and drug delivery agents has opened
emerging areas of research. (131;132;133) Despite advances in nanoMOF applications,
31
accurate models are still needed to probe fundamental mechanistic details and
reliably control particle sizes.
Several models exist to describe nanocrystal nucleation and growth,
including the classic LaMer model, the Watzky-Finke model, and various statistical
models, yet recent evidence challenges their applicability to MOF growth. (134;135;136)
Whereas the LaMer model describes distinct stages of a burst nucleation induced
by supersaturated monomer concentrations, followed by diffusion-limited particle
growth, (134) nanoMOFs form at dilute concentrations, and in-situ studies reveal
continuous nucleation and growth of MOF particles. (59) Models based on monomer
addition also do not apply to MOFs, as studies suggest MOF formation involves
transiently metastable “primary” phases, aggregative growth, and other non-
classical events. (61;9;137;138) Although these existing models can be modified to
account for non-classical events, nanoMOF research requires a general model
based on the acid-base and coordination chemistry of MOFs to reliably predict
and control particle sizes.
Figure 3.1. Equilibrium concentrations of MOF linker species determined from
the coupled equilibria of the Seesaw model. Linker–metal ([LM]) and protonated
linker ([LH]) concentrations calculated as a function of initial metal ion ([M+]) and
proton ([H+]) concentrations. Inset: [LM] calculated as a function of initial [M+]
for a range of [H+]. Initial concentrations were chosen from typical experimental
conditions reported here.
32
The Seesaw Model and pKa
Previously, we proposed a novel “Seesaw” model of nanoMOF growth
based on a metadata analysis of existing literature. (6) This model specifically
explained why the use of modulators—typically monotopic analogs of MOF
linkers—causes particle sizes to increase in certain cases, but decrease in others.
These trends could be explained by modulators functioning as capping ligands at
low concentrations and as acids at high concentrations (Figure 3.2, regions I and
II, respectively). More broadly, we proposed that MOF nanoparticles result from
excess ligand depleting local concentrations of metal ions and kinetically trapping
particle growth. Accordingly, the trapping process depends on the competition
between coupled equilibria associated with ligand deprotonation and metal-ligand
complexation. In the absence of aggregation, crystallite sizes, therefore, minimize
when monomer ratios in solution are most “off-stoichiometric” relative to the MOF
stoichiometry, which complements established models of bulk polymer and crystal
growth that propose the largest particles are generated by solutions with monomer
ratios that match the intended material stoichiometry (Figure 3.2a), as discussed
below. The curious Seesaw-shaped relationship between particle size and modulator
equivalents observed previously for MIL-125-NH2 and UiO-66 nanoparticles
(139)
(Figure 3.2b) could be explained as exhibiting both region I and II behavior, where
sizes first decrease with additional modulator acting as capping ligand and then
increase as acidic modulators keep linkers protonated and unable to kinetically trap
particles.
33
Figure 3.2. Scheme depicting the dependence of particle sizes on metal-to-linker
solution stoichiometry. (a) In the proposed Seesaw model of MOF nanocrystal
growth (solid line), particle sizes minimize at maximally imbalanced metal-to-
linker ratios and at low proton concentrations (ε). The Seesaw arises from acidic
ligands acting as capping agents in region I and as acids in region II. Conversely,
rates of bulk ionic crystal growth (dashed line) maximizes at monomer solution
stoichiometries that match the bulk crystal (σ). (b) Observation of the Seesaw
relationship between NH2-MIL-125 particle sizes and modulator equivalents,
reproduced from ref. 115.
Although observed previously for just these materials, we posited that the
Seesaw trend could be observed generally for all MOFs through deliberate control
over the parameters outlined by the model. Herein, we provide experimental
evidence for the Seesaw model by demonstrating that solution acidity, ligand
excess, and concentration form independent parameters that can be used to
achieve the first demonstration of a Seesaw relationship and reproducibly control
34
nanoparticle sizes of two iconic MOF materials, Zn(mIm)2 (ZIF-8) and Cu3BTC2
(HKUST-1). With the results presented here, the Seesaw trend appears in at least
four compositionally distinct MOF systems, suggesting that the mechanistic model
outlined here is universal to all modulated MOF syntheses.
ZIF-8 and Cu2BTC3 Nanoparticle Syntheses
In chapter 2, the four key equilibria expressions that form the basis of the
Seesaw model were presented (Fig. 2.3): linker deprotonation (Eq. 1), modulator
deprotonation (Eq. 2), metal-linker complexation (Eq. 3), and metal-modulator
complexation (Eq. 4). According to our model, the competition of these coupled
reactions creates conditions that either produce bulk or nanocrystalline MOFs.
For example, if these coupled equilibria maintain stoichiometric ratios of metal
ions and linkers, then the reaction proceeds to form bulk MOF crystals. If, on the
other hand, the coupled equilibria cause depletion of local metal ion concentrations,
then excess ligand overwhelms particle surfaces, trapping MOFs as nanoparticles.
Coupled equilibria can be studied by several mathematical formalisms. (140;141)
Herein, we simplify the system of equations using assumptions similar to the well-
known Initial-Change-Equilibrium (ICE) table method (Table B.19).
We propose that analysis of these coupled equilibria can be used to predict
whether synthetic conditions produce nanoMOFs by 1) knowing equilibrium
constants for the metal ions and carboxylic acids, 2) knowing the initial reactant
concentrations, and 3) by assuming that nanoparticles arise from reaction
conditions that develop excess concentrations of deprotonated linkers [L–]eq
relative to the concentration of uncoordinated metal ions [M+]eq. In other words,
reactant concentrations and equilibrium constants could be chosen such that
[L–]eq + [Mod
–]eq > [M
+]eq in the distribution of chemical species at equilibrium.
35
For simplicity, we treat the MOF linkers as monotopic ligands and consider
only individual metal–linker bonds rather than the entire coordination sphere.
We also assume Ka of the modulator and MOF linker to be equal, and that the
complexation equilibrium constants are equal.
For a fundamental justification of the Seesaw model, namely the U-shaped
size trend arising from the dual roles of modulator acting as acid or capping
ligand, we solved the system of coupled equilibria in Figure 2.3. Figure 3.1
shows the equilibrium concentration of linker-metal [LM]eq and protonated linker
species [LH]eq as a function of [H
+] donated by the linker and modulator under
typical reaction concentrations (see Appendix B for full details). When initial
concentrations of acid are low, [LM]eq is maximized and [LH]eq is minimized.
In other words, reactions with low equivalents of acidic modulator favor linker
deprotonation and metal-linker bond formation. Critical concentrations of acid,
however, induce steep changes to [LM]eq and [LH]eq. At higher [H
+], [LM]eq is
minimized and [LH]eq is maximized. The plateau of [LM]eq and [LH]eq spanning
four orders of magnitude for low values of [H+] followed by a steep change supports
our model that formation of linker-metal bonds can be favorable even with
additional acid, until reaching a critical concentration threshold that causes linkers
to remain protonated, thereby suppressing metal-linker bond formation and the
trapping of MOF particles. These results therefore justify a key claim of the Seesaw
model that acidic reaction conditions produce large particles by favoring linker
protonation, which improve the equilibria reversibility and maintains solution
stoichiometry required to grow MOF single crystals. Because we expect modulator
to decrease metal ion concentrations by forming modulator-metal species (Eq.
4), we also investigated the relationship between [LM]eq and initial [M
+]. The
36
inset of Figure 3.1 shows that [LM]eq depends directly on the available [M
+] in
solution, suggesting that equivalents of modulator inhibits bulk MOF growth by
modulators competing with linkers for metal ions. Interestingly, the formation of
[LM]eq decreases for high initial [H
+], providing further evidence that modulators in
large excess function more as acids than as surface capping ligands. Taken together,
these results provide a fundamental basis for the key claims of the Seesaw model.
Powerful predictions can be made about the outcome of MOF syntheses
by analyzing the system of coupled equilibria in Figure 3.2. The results of the
analysis suggest that, as long as the presence of excess deprotonated linker is a
key determinant in kinetic trapping, nanoparticles should always result from cases
that employ excess linker and no modulator, as well as cases that use deprotonated
modulators as Brønsted bases or capping ligands. When modulators function solely
as acids, they decrease the availability of L–, which inhibits the kinetic trapping
of nanoMOFs. The outcome depends on the particular equilibrium constants,
initial reactant concentrations, and how the reactant concentrations change during
MOF synthesis. The most complex scenario, which appears most frequently in the
literature, involves modulators functioning as both acids and capping ligands. (4)
Nevertheless, nanoMOF sizes should be tunable through careful manipulation of
acid and ligand binding strengths if indeed nanoMOF growth depends on these two
independent parameters.
With these predictions in hand, we sought experimental evidence for the
Seesaw model and the independent tunability of nanoMOF sizes through acid and
ligand addition. Figure 3.3a plots nanoparticle sizes of ZIF-8 (Zn(mIm)2, mIm =
2-methylimidazolate) as a function of added HmIm ranging from 4 to 14 equivalents
per Zn2+, where conditions above 2 equivalents represent linker in excess. Based on
37
analysis of the coupled equilibria, addition of only HmIm without modulator should
cause the depletion of [M+], resulting in kinetic trapping of smaller nanoMOF sizes.
Indeed, up to a reactant stoichiometry of 16-18 Hmim equivalents per metal, ZIF-
8 particle sizes continue to decrease. Above this linker excess, product is simply
not observed. These results resemble a previous report by Cravillon et al. that
showed excess linker up to eight linker equivalents leads to progressively smaller
ZIF-8 nanoparticles, (59) but the data presented here show that trend progresses
further. We note that particle sizes isolated here are slightly smaller than those
reported by Cravillon et al., which we attribute to the different characterization
techniques: here, crystallite sizes are reported by Scherrer analysis whereas the
previous account reported particle sizes from SEM imaging.
Figure 3.3. ZIF-8 nanoparticle sizes as a function of excess modulator equivalents.
(a) Particle sizes versus excess linker equivalents of 2-methylimidazole (Hmim).
Above 18–20 equivalents, formation of ZIF-8 was not observed. (b) Particle sizes
resulting from increasing HCl equivalents with respect to zinc nitrate. Blue circles
denote samples synthesized with 14 equivalents of Hmim compared to zinc nitrate.
Details of the synthetic conditions can be found in the Experimental methods and
Tables B.16–B.19.
Figure 3.3b plots the size of ZIF-8 particles versus number of HCl
equivalents with respect to Zn2+ at a fixed linker excess of 14 equivalents. As a
non-coordinating species, HCl acts only to increase proton activity. Therefore,
38
HCl addition allowed us to test the hypothesis that nanoMOF sizes increase with
higher proportions of protonated linkers incapable of trapping nanoparticles,
as indicated in Figure 3.1. The data show ZIF-8 nanoparticle sizes predictably
increase with additions up to 0.7 HCl equivalents per Zn2+ ion. (Fig. 3.3b).
These results corroborate previous studies showing that the addition of HCl can
weaken metal-linker bonds in ZIF-8, implying there is a more dynamic equilibrium
for metal-linker bond formation in acidic media. (142) Futher, the use of acids as
modulators helps single-crystal growth for in several MOF systems. (143;94) These
results demonstrate that protons and capping ligands form independent parameters
that control nanoMOF sizes.
The dual role of modulators acting as acids and ligands complicates
the synthetic control of nanoMOFs, but can nevertheless be separated and
demonstrated to control the equilibria that govern particle sizes. Previously, we
proposed that the dual role of modulators could give rise to a Seesaw relationship
between particle sizes and modulator equivalents where sizes decrease with
increasing equivalents and then increase as proton activities become sufficiently
high to inhibit linker deprotonation. For proof of the Seesaw relationship, we
targeted the iconic MOF Cu3BTC2 (BTC = 1,3,5-benzenetricarboxylate) because
many modulated nanoMOF reports have focused on this material. These reports
have shown that increasing equivalents of carboxylic acid modulators increases
particle sizes, whereas Cu3BTC2 sizes decrease with increased amounts of
deprotonated carboxylate, or other basic, modulators. (52;11;51;144) Hypothesizing that
these trends reflect the dual role of modulators acting as acids versus ligands, we
investigated the impact of adding benzoate versus benzoic acid to the synthesis of
Cu3BTC2. Although literature reports often use SEM or light scattering methods
39
to determine sizes of MOF nanoparticles, sizes in this study are reported from
Scherrer analysis because this method gives not particle size, but average size of
coherently scattering domains. This metric is particularly useful for identifying the
sizes of particles independent of aggregation. SEM images of Cu3BTC2 products in
alkaline conditions (Figure B.9) revealed severe aggregation of particles, which is
also prevalent in literature reports. (52;53;46) Because aggregation prevents accurate
statistical analysis of particle sizes, all particle sizes reported here are derived
using the Scherrer equation, with SEM images provided as supplements to discuss
discrepancies in particle size determination techniques as well as the morphology of
the products. Interestingly, MOF product could not be isolated from the synthesis
with benzoic acid by mixing Cu(NO3)2 and trimesic acid at room temperature,
whereas particle sizes strictly decreased by adding additional equivalents of sodium
benzoate (Fig. B.2). Increasing sodium benzoate also leads to extra peaks in the
PXRD patterns, likely arising from fast reaction kinetics causing benzoate ligands
to trap within the MOF structure and cause defects. These results validate the
prediction from the analysis of the coupled equilibria (Eq. 1-4) that deprotonated
modulators acting as either ligands or bases lead to MOF nanoparticles, also
consistent with previous studies. (11)
40
Figure 3.4. Dependence of Cu3BTC2 particle sizes on modulator equivalents.
SEM images of particles synthesized with 7 modulator equivalents show globular
morphology at (a) 33% and (b) 50% benzoic acid (BA), whereas (c) SEM
shows octahedral particles when synthesized with 66% BA. Percentages indicate
the mole fraction of BA, with the remainder added as sodium benzoate. (d)
Dependence of size on modulator equivalents with varying benzoic acid content.
Syntheses were performed with a linker:metal:modulator ratio of 3:1:X, where
modulator corresponds to the sum of benzoic acid and sodium benzoate. The metal
concentration was held constant (1 mM). The grey line at 7 equivalents shows the
data used again in Fig. 3.5. Details of the synthetic conditions can be found in the
Experimental methods and Tables B.5–B.7.
To observe a Seesaw dependence between Cu3BTC2 sizes and modulator
equivalents, we employed a buffer mixture of sodium benzoate and benzoic
acid to balance the opposing trends observed when adding only acid or ligand.
Hypothesizing that particle stability depends on achieving ligand-rich surfaces,
we employed excess linker stoichiometries (3 linkers to 1 metal). Figure 3.4 plots
Cu3BTC2 nanoparticle sizes versus equivalents of modulator mixtures with benzoic
acid (BA) molar contents of 33%, 50%, or 66%, with the remainder comprised
of sodium benzoate. Indeed, Seesaw curves appear in all three cases with similar
qualitative features: a decrease from 7 to 10 equivalents, a plateau, and a gradual
41
increase in sizes from 28 to 40 equivalents. When using 1 equivalent of modulator,
SEM revealed particle sizes that exceeded 1 µm (Fig. 3.7, B.12), showing a steep
decline in size at the beginning of the Seesaw trend. While the full data set is in
the appendices (Fig. B.7a), we chose to include the data from 6 to 40 equivalents
in the main text for clarity. Interestingly, the least acidic modulator mixture
(33% BA) results in the steepest downward slope from 1 to 7 equivalents, which,
according to the Seesaw model, reflects more favorable linker deprotonation or
metal ion complexation (Fig. B.7a). Additionally, the most acidic modulator
mixture produced sizes that were overall largest, whereas while the least acidic
mixture gave the smallest sizes. Whereas Scherrer analysis provides insight into the
impact of modulator on crystalline domain size, SEM provides information about
the impact of modulator on morphology as well. Both techniques show a U-shaped
size dependence at 33% benzoic acid (Fig. B.10), although sizes by SEM analysis
were systematically larger, which is typical for the two instrumental techniques;
there is reasonable agreement between the techniques when the Scherrer size is
below 100 nm (Fig. B.10, Fig. B.11). Particles isolated with the 33% benzoic acid
modulator show globular morphologies, except below 7 modulator equivalents
where particles exhibit faceting (Fig. 3.7). Rather than exhibit minimum sizes
around a narrow range of modulator equivalents, all data sets show a broad flat
region that we attribute to the modulator mixture functioning as a buffer: the
benzoic acid-benzoate pair accepts protons from the excess linkers so that sizes
increase only when the proton activity exceeds the buffer capacity of the modulator
mixture. Indeed, in conditions using copper acetate as a metal source and solely
benzoic acid as a modulator, there is no long plateau and a minimum size occurs at
14 modulator equivalents (Fig. B.7b). Comparing the buffered systems, the most
42
dramatic size increase can be observed with the 66% BA set. In fact, SEM images
of the product at 40 modulator equivalents show regions that resemble bulk crystal
growth (Fig. B.9). While modulators have induced changes to MOF morphology,
PXRD patterns exhibit peaks associated only with the Cu3BTC2 phase (Fig.
B.3). (12) These results therefore validate the Seesaw model we had previously
proposed: by using conjugate acid/base mixtures, nanoMOF sizes decrease with
additional capping ligand until proton activities inhibit ligand from trapping metal
ions, allowing for bulk MOF growth. (6)
Figure 3.5. Dependence of Cu3BTC2 particle sizes on benzoic acid content (% BA).
SEM images of samples from the 3 mM linker set show spherical morphology at 0%
BA (a) large flower-like aggregates at 8.33% BA (b), but indistinct morphology for
BA contents 83.33% and higher (c). (d) Cu3BTC2 grain sizes versus benzoic acid
content of modulator mixtures at constant modulator equivalents. Two trials at
low concentration were completed, and we present the data as the average of the
two. Reactant concentrations are defined with respect to the linker (L). Details
of the synthetic conditions can be found in the Experimental methods and Tables
B.1–B.3.
To further decouple the independent roles of acidity and ligand
complexation, we explored the size dependence of Cu3BTC2 nanoparticles as a
43
function of benzoic acid content in the modulator mixture at fixed modulator
equivalents. Figure 3.5d plots Cu3BTC2 particle sizes versus benzoic acid content
of modulator mixtures for three different reactant concentrations. According to
the Seesaw model, larger crystallites should result from increased proton activities.
Indeed, without adding additional modulator equivalents, Scherrer analysis shows
increasing crystallite size in reactions where BA contents exceed 50%. Interestingly,
higher reactant concentrations lead overall to smaller sizes except at BA contents
near 100%, where sizes exceed the typical range of Scherrer analysis, showing that
in all cases acidic conditions result in products that resemble bulk crystal growth.
Within the context of the Seesaw model, the decreased sizes with higher reactant
concentrations result from efficient metal ion depletion and kinetic trapping of
particles. The experiments in Figures 3.4 and 3.5 thus explore a cross section of
the same multi-dimensional reaction space. Datasets that intersect in this reaction
space could be analyzed for reproducibility.
Figure 3.5d includes three data points from Figure 3.4d for comparison
against the 3-mM dataset, which employs the same reaction conditions, showing
good reproducibility. At low BA contents and low concentrations, average
crystallite sizes increase slightly, whereas morphologies by SEM undergo dramatic
changes. The particles isolated with only sodium benzoate appear spherical (Fig.
3.5a), but an increase in benzoic acid to 8.33% yields large star-like structures
comprised of many aggregated crystallites (Fig. 3.5b). From 16.67% to 50%, BA
contents, the particles show spherical morphologies and the 66% and 75% benzoic
acid mixtures yield the octahedral particles typical of Cu3BTC2. For mixtures with
BA contents as high as 83.33% and 91.67%, the particles no longer appear faceted,
but instead become bulk-like aggregates (Fig. 3.5c). These results reveal that small
44
changes in proton activity can yield dramatic changes in particle morphology, which
we will explore in ongoing studies. Although the Seesaw model at present does not
offer predictions of particle morphologies, we expect that its foundation in acid-
base and metal-ligand chemistry can help explain the dependence of morphology on
solution acidity and modulator composition. Although previous reports have used
modulators in combination with manipulation of pH, (11;145) this report is the first
to systematically manipulate particle sizes with a buffer.
Concentration acts as a third key determinant of nanoMOF sizes because
it controls the impact of the other two parameters, acidity and ligand excess.
Building on the concentration dependence exhibited in Figure 3.5d for modulator
acidity, we explored the impact of concentration in relation to modulator
equivalents. Figure 3.6 plots the concentration dependence of Cu3BTC2 sizes
using a 50% benzoic acid modulator mixture at three different equivalents of
modulator. At both 0.7 and 7 equivalents of modulator, sizes decrease with
increasing concentration. The effect of concentration on size is most pronounced
at 7 modulator equivalents, spanning the size range of 20-80 nm, with the largest
difference occurring between linker concentration of 3 and 5 mM (Fig. 4d), and
little if any difference at 13.34 modulator equivalents (Fig. 3.6c).
45
Figure 3.6. Cu3BTC2 particle sizes resulting from variable reactant concentrations
and total modulator equivalents. Panels a, b, and c show dependence of particle
sizes on concentration at three different modulator:linker ratios. Syntheses were
performed with a linker:metal:modulator ratio of 3:1:X, with a 50% benzoic acid
modulator mixture. Synthetic conditions can be found in Tables B.12–B.14.
Applying the Seesaw Model
These results demonstrate that MOF nanoparticle sizes can be tuned
through independent control over solution acidity, ligand excess, and concentration,
although certain parameters have greater impacts than others. For example,
the data in Figure 3.6 show that while increased concentrations cause Cu3BTC2
particle sizes to decrease over a range of linker concentrations from 5 to 20 mM,
increasing the modular equivalents produces a greater overall decrease in particle
sizes. In terms of the Seesaw model, these results suggest that equilibria shift
less in response to concentration changes compared to changes in stoichiometry.
Furthermore, equilibria become so shifted by excess modulator that they become
nearly insensitive to concentration changes, as indicated by Figure 3.6c. On the
other hand, Figure 3.5d shows that changes to BA content have the greatest impact
on crystallite size when the BA content exceeds 50%, which we propose corresponds
to a critical decrease in the buffer capacity of the modulator mixture. Below
the buffer capacity, concentration affects crystallite size to a lesser extent. Once
the buffer mixture no longer absorbs excess protons, we expect that conditions
46
favor exchange of surface capping ligands for linkers that allow continued bulk
growth, as has been observed in post-synthetic MOF linker exchange (146) and in
reports on metal-linker stability in ZIF-8. (142) These results, taken to the extreme
limit, explain why additional acid aids in the synthesis of large single crystals. (94)
Interestingly, the cross-sectional data in Figure 3.5d indicate that modulator excess
and BA content can produce similar absolute changes to crystallite sizes. These
two variables have different effects on particle morphology, however. Whereas
modulator equivalents have minimal impact on morphology, except in extreme
cases of low or high equivalents (Fig. 3.7), minute changes to BA content can
result in dramatic differences in particle morphology (Fig. B.8). These results
suggest buffered systems may serve as a powerful synthetic tool for tailoring MOF
particle shapes. The impact of changing any of these parameters appears strikingly
nonlinear. Just as reducing BA content from 50% to 33% leads to small overall
changes in sizes, doubling linker concentrations from 7.5 mM to 15 mM in Figure
3.5d has diminishing effects. Such nonlinearity complicates predictions about
nanoparticle sizes, but its existence lends further proof for the Seesaw model, which
relies on nonlinear relationships between coupled equilibria (Fig. 3.1).
47
Figure 3.7. Dependence of Cu3BTC2 particle morphology on modulator equivalents
for a 33% BA modulator mixture. Products isolated from syntheses with: a) 1
eq; b) 14 eq, c) 21 eq, d) 28 eq, e) 35 eq, and f) 40 eq. Details of the synthetic
conditions can be found in the Experimental methods and Tables B.5–B.7.
Overall, these results suggest that ligand excess—of either linker or
modulator—exerts the greatest impact on nanoMOF sizes. Whereas excess linker
generates nanoparticles of ZIF-67, ZIF-7, and ZIF-71, cases for carboxylate MOFs
48
are rare. (147;148;149) While excess trimesic acid increases grain sizes of Cu3BTC2
particles, in NU-1000, another carboxylate MOF, particle sizes decrease as excess
linker is used. (60;150) In the report of NU-1000 particles, a strong base is added
as well, which likely counters any increase in [H+]. We attribute this difference
to the fact that imidazole linkers contain just single protic sites, whereas multi-
topic carboxylates contain several, which, according to the Seesaw model, increase
solution acidity and hinder the ability of ligands to trap metal ions. Additionally,
the greater strength of zinc-imidazolate bonds should facilitate rapid trapping by
excess linker, whereas metal-carboxylate bonds are more dynamic and, hence,
less effective at terminating particle growth. The synthesis of carboxylate-based
MOF nanoparticles therefore depends strongly on modulator excess. Although
solution acidity and reactant concentration influence the kinetic trapping of MOF
nanoparticles, achieving small nanoMOF sizes ultimately relies on the presence of
excess ligands.
According to the Seesaw model, dilute local concentrations of metal
ions overwhelmed by excess ligand leads to kinetic trapping of small particle
sizes. This prediction helps explain previous reports that dilution yields smaller
ZIF-8 nanoparticles, (8) which at first seems in conflict with the concentration
studies presented here. Figures 3.4 and 3.5 both show Cu3BTC2 particle sizes
decreasing with increased concentrations; we suggest two reasonable hypotheses
to explain this apparent discrepancy. Firstly, imidazole and carboxylate MOFs
exhibit rather different metal-linker bond strengths. Whereas ZIF-8 features
strong metal-linker bonds that rapidly form bulk crystals under concentrated
conditions, carboxylate-based MOFs exhibit slower growth kinetics that tend
to form large single crystals through slow and dynamic exchange of ligands. (151)
49
Increasing the local concentration of linker therefore improves kinetic trapping
of carboxylate MOF nanoparticles by shifting the weak metal-linker equilibrium
towards complexation. The concentration dependence also depends on the stages
of MOF growth. For example, it has been demonstrated in injection syntheses
that including an excess of either linker or metal ions at the beginning of the
reaction results in smaller particle sizes, which is attributed to the rapid formation
of small MOF clusters and oligomers. On the other hand, slow addition of MOF
components during the reaction results in the growth of larger particles. (152) In
the context of the Seesaw model, this shows that with excess of either of the MOF
components, rapid depletion can lead to the kinetic trapping of small particles.
Increasing the total concentration during the reaction, however, will inevitably
lead to larger particles because the metal ions and linkers will add onto existing
particles.
These results also highlight the complex role played by ligands in trapping
metal ions as molecular complexes, preventing their incorporation into growing
MOF particles. For example, Figure 3.3 shows that extreme excess of linker
equivalents suppresses ZIF-8 formation. Under these conditions, imidazolate
molecules likely coordinatively saturate Zn2+ ions, shifting the metal-linker binding
equilibrium far toward complexation and inhibiting the dynamic ligand dissociation
needed for monomer attachment and growth. Indeed, a previously reported in-situ
study of ZIF-8 growth under excess linker conditions suggested particle growth
proceeds by linker dissociation from zinc-imidazolate oligomers under concentrated
conditions. (153) Similarly, excess benzoate appears to inhibit Cu3BTC2 formation
by trapping Cu2+ ions in benzoate complexes or small, saturated clusters. This
hypothesis could explain why particle sizes at first decrease with added BA content
50
in Figure 3.5d at 3 mM linker concentration. By increasing the solution acidity,
the modulator mixture becomes less competitive with the trimesate linker for Cu2+
coordination and Cu3BTC2 forms more readily, whereas benzoate-rich conditions
make particle growth reliant on the slow release of Cu2+ ions, leading to larger
particles. Another explanation for the larger sizes under benzoate-rich conditions is
the tendency of deprotonated modulators to induce aggregation. (52;53) The extreme
difference between the moderate crystallite size of the 8.33% BA sample and its
large star-like morphology indicates that aggregative growth is operative (Fig. 3.5).
Previous reports have shown that pH adjustment can manipulate the assembly
and aggregation of MOF-525 particles (154;155) and that strongly acidic modulators
improve the colloidal stability of UiO-66. (60)
More generally, the Seesaw model serves as a complement to well-established
models of bulk crystal growth and polymer formation. Whereas successful growth
of ionic crystals and condensation polymers depends on maintaining stoichiometric
mixtures of reactants, the Seesaw model proposes that MOF nanoparticle sizes
minimize from maximally imbalanced local concentrations of reactants: in
modulated MOF syntheses, ligands always outnumber metal ions. For many
classes of polymers, molecular weights decrease considerably under conditions of
imbalanced monomer stoichiometries, producing oligomers instead of long polymer
chains. Interestingly, molecular weights can also be controlled by terminating
chain growth with the addition of monofunctional monomers, akin to MOF
modulators. (156) For ionic solids, the rate of crystal growth maximizes when the
relative ratios of monomers diffusing to crystal surfaces matches the stoichiometry
of the bulk lattice. (157) Models of ionic crystal growth state that rates of monomer
attachment relates directly to nsite * τm * J, where nsite represents the density of
51
available binding surface sites, τm is the lifetime of monomer units, and J is the flux
of the monomer to the growing crystal. (157) In terms of the Seesaw model, therefore,
small nanoMOF sizes result from rapid depletion of metal ion concentrations
and overwhelming particle surfaces with excess ligand. We note that the model
assumes growth to depend solely on metal-linker bond formation, whereas interlayer
stacking constitutes an important component of 2D MOF growth. We propose
that the Seesaw model could be adjusted for such materials by including the
formation constants for interlayer stacking. (13) Additionally, these models assume
the absence of aggregation, which we expect to play a large role in determining size
after crystallite formation. For MOFs, the relevant diffusing species may be metal
ions, linkers, or even entire clusters, given in-situ studies that suggest Cu3BTC2
grows by increments of individual Cu2BTC4 paddlewheel units, or even the oriented
attachment of small crystals. (158;159) Interestingly, typical representations of this
model of ionic crystal growth plot growth rate versus solution stoichiometry,
with maximum growth centered at balanced ratios in an “upside down Seesaw”
curve. (160) Figure 3.2, therefore, illustrates the complementary relationship of the
Seesaw model to common models of bulk crystal and polymer growth, where sizes
minimize at maximally unbalanced stoichiometries, and vice versa.
Lastly, we propose that the steep downward slope observed in “region I”
of the Seesaw curve can be interpreted in terms of classical collision theory. In
collision theory, the probability of no collision taking place between particles
as a function of time, P(t), equals exp(–t/τ), where τ represents the average
time between collisions. (161) Similarly, we propose that the probability of MOF
nanoparticles not being kinetically trapped by excess ligand decreases exponentially
as more modulator enters the reaction mixture. In other words, the probability of
52
particles colliding with capping ligand increases with higher available equivalents
of excess ligand. This model helps explain why decreasing slopes in Figure
B.7 become shallower from 1 to 7 equivalents with higher acidic content—the
probability of successful particle trapping diminishes as the modulator becomes
more acidified. Higher concentration leading to smaller particles is also discussed
in classical nucleation theory (CNT), in which the energy barrier to nucleation is
overcome only in supersaturated conditions. Here, MOFs grow in dilute conditions,
but increasing concentration leads to smaller particles as there are more collisions
in solution that lead to particle formation. We anticipate that temperature,
concentration, and other factors expected to impact collision probability play
decisive roles in the mechanism of MOF nanoparticle growth.
Conclusions
In conclusion, we offer experimental proof of the Seesaw model of nanoMOF
growth by demonstrating for the first time the existence of a Seesaw relationship
between nanoMOF sizes and modulator excess through deliberate manipulation of
key parameters in the model. Specifically, we show that MOF nanoparticle sizes
can be tuned through independent control over solution acidity, ligand excess,
and reactant concentrations. Demonstrating that these three parameters control
nanoMOF sizes supports the key claim of the model that nanoMOFs result from
kinetic trapping of nanoparticles determined by competition between coupled
equilibria involving metal-ligand complexation and ligand acid-base chemistry.
The relative impact of these parameters on nanoparticle sizes was explored, with
ligand excess showing the greatest overall impact. Sizes generally decease with
lower acidity, greater ligand excess, and, for dynamic metal-linker bonds, higher
concentrations. Importantly, particle sizes were reproducible when approaching
53
similar reaction conditions from different directions on the multi-dimensional
reaction space defined by these three parameters. The Seesaw model represents
a novel perspective for understanding MOF growth in general, and we further
show that it complements well-established models of bulk polymer and crystal
growth. Despite its clear relation to other models of crystal growth, the competing
equilibria should be expanded to account for equilibrium constants associated with
events such as oriented attachment, in which multiple metal-ligand bonds form at
the same time. We further note that thus far the Seesaw curve has been observed
by using aromatic carboxylate modulators with similar pKa values and expect that
altering the size and acidity of modulators to have a large impact on the particle
Seesaw dependence. Although the dependence of morphology on solution acidity
was unexpected, buffer systems may prove powerful tools for deliberate control over
tailoring particle architectures. Taken together, these results demonstrate that the
Seesaw model offers a fundamental platform for advancing the synthesis and basic
understanding of this emerging class of materials.
54
CHAPTER IV
SIZE-DEPENDENT PROPERTIES OF CONDUCTIVE MOF NANOCRYSTALS
This chapter includes an excerpt from previously published and co-
authored material from Marshall, C.R.; Twight, L.P.; Dvorak, J.P.; Overland,
A.E.; Brozek, C.K. Size-Dependent Properties of Solution-Processable Conductive
MOF Nanocrystals. J. Am. Chem. Soc. 2022, 144, 13, 5784-5794. The project
in this chapter was conceptualized and developed by Checkers R. Marshall; the
final manuscript was co-written by Checkers R. Marshall and Carl K. Brozek. The
experimental work in this chapter was performed by Checkers R. Marshall, Josh. D.
Dvorak, Lan Chen, and Alexi E. Overland. The Mössbauer data in this chapter was
collected by Thomas Ericson.
The discovery of colloidal metal and semiconductor nanocrystals has led
to a revolution in materials chemistry. (162) While the class of materials known
as metal-organic frameworks (MOFs) attracts broad interest due to their high
accessible surface areas and wide synthetic tunability, their preparation as stable,
monodisperse nanoparticles remains an open frontier. (163) Nevertheless, recent
research into the few examples of MOF nanoparticles has revealed enhanced
properties compared to their bulk counterparts due to greater mass transport
rates, higher ratios of exposed surface areas, and superior processability, leading
to higher catalytic activities, improved gas permeability in separation membranes,
more uniform integration into composite films, and utility as unique drug delivery
agents. (31;164;165;166;167) The controlled synthesis of MOF nanoparticles with precise
sizes will therefore facilitate their application in myriad applications, while enabling
solution-state analysis to reveal size-dependent physical properties. (168) Currently,
the few MOFs that can be prepared as colloidally stable nanoparticles with narrow
55
Figure 4.1. Overview of Fe(1,2,3,-triazolate)2 (Fe(TA)2) nanocrystal synthesis.
a) General synthetic route to Fe(TA)2 nanoparticles using 1-methylimidazole (1-
mIm) as a modulator. b) Equilibrium between metal ions (M), linkers (L), and
modulators (Mod) that control MOF nanoparticle formation. c) The secondary
building unit (SBU) cluster of Fe(TA)2. d) PXRD patterns of the Fe(TA)2
nanoparticles synthesized with varying amounts of 1-mIm. Red and grey lines
indicate expected reflections for the low-spin (LS) phase and high-spin (HS) phase,
respectively. e) The SBU of Fe(TA)2. e) An idealized representation of a 5.5-nm
Fe(TA)2 particle based on the bulk crystalline structure.
size dispersities are electrical insulators, whereas conductive MOF nanoparticles
will enable electronic technologies that benefit from the microporosity and high
surface areas of MOFs and the processability and size-control of nanoparticles. (169)
Here, we address the long-standing challenge of making solution-processable
conductive MOF nanoparticles (nanoMOFs) with controlled size and narrow
dispersity. Solution processability also enables solution-state spectroscopy for
56
unprecedented analysis of electronic properties and thin film fabrication reveals
unique insight into their redox chemistry and charge transport phenomena.
Size-tunable Synthesis of Iron Triazolate Nanoparticles
Nanoparticles of Fe(TA)2 were prepared with tunable sizes by modifying
the standard bulk synthesis11 through addition of 1-methylimidazole (1-mIm) as
a “modulator” (Fig. 4.1). Although the use of “modulators,” usually monotopic
analogs of the native MOF linkers, (6) is common in MOF nanoparticle syntheses,
their mechanistic role has been debated, especially because increasing modulator
equivalents leads to smaller sizes in some systems, but larger sizes in others.
Chapters 2 and 3 point to a “Seesaw” model where modulators impact particle
sizes by interfering with linker deprotonation (Fig. 4.1, Eq. 1) and metal-linker
complexation (Eq. 3). Based on a meta-data analysis of the literature, (6) this
model contends that MOF particles result from kinetic trapping of metal ions
by excess linker or modulator. Through deliberate control of the key parameters
outlined by this model—proton activity, concentration, and ligand excess—excellent
size reproducibility can be achieved for benchmark MOF materials. Using
these insights, we can devise MOF syntheses previously inaccessible at the
nanoscale. To access small and controllable particle sizes, we hypothesized that
a monotopic imidazole linker, 1-methylimidazole, would work as a modulator
that mimics the native MOF linker. Reversible bonds between Fe and 1-mIm
(Eq. 4) would compete with metal-linker complexation (Eq. 3) to rapidly form
the secondary building blocks (SBUs) of Fe(TA)2, shown in Figure 4.1c. The
presence of the monotopic linker at the surface of a growing particle also inhibits
further growth until it dissociates. Additionally, previous work with other MOF
57
systems have shown promising nanosizing effects with the use of similar imidazole
modulators. (170;171)
Figure 4.2. Particle sizes of Fe(TA)2 resulting from modulated syntheses. (a) –
(e) SEM images of Fe(TA)2 nanoparticles synthesized with 1-methylimidazole
equivalents (with respect to FeCl2) ranging from 0.055 eq (a) to 0.709 eq (e).
Images were set to greyscale using Adobe Illustrator. Insets: histograms of particle
size distributions from >200 particles fitted to weighted gaussian distributions.
f) Particle sizes using 1-mIm (blue squares), 1-benzyl-2-methylimidazole (Benzyl,
green triangle) 5-bromo-1-methylimidazole (Bromo, red triangle), and n-butylamine
(nBA, purple triangle). Filled data points were determined by SEM and open data
points by Scherrer analysis.
In comparison to the bulk synthesis first reported by Yaghi et al in
2012, (19;20;21) the synthesis of nanoparticles was conducted under dilute conditions,
with stirring, and varying modulator equivalents. After heating the reaction
mixture under air-free conditions in DMF at 120 ◦C, the reaction was terminated
after 1.5 h by removing from heat and immediate centrifugation and washing.
These reactions are low yielding (6 – 23%), which is further evidence of arrested
growth of kinetically trapped particles, as has been observed in other nanoMOF
58
systems (Fig. C.3). Indeed, Figure 4.1d shows the expected PXRD pattern of
Fe(TA)2 but with peaks that broaden with greater equivalents of modulator.
Additionally, Figure 4.2 shows that particle sizes determined by Scherrer analysis of
these data and by SEM imaging dramatically decrease with increasing equivalents
of 1-mIm. At low equivalents of 1-mIm, the particle sizes decrease steeply: from
0.055 eq to 0.709 eq, the particle sizes reduce from 130 nm to 16 nm. Beyond
0.709 eq, the particle sizes level off abruptly, decreasing to 5.5 nm only with 10.9
equivalents of modulator (Fig. 4.2f). To the best of our knowledge, the 5.5-nm
particles represent some of the smallest reported MOF particle sizes synthesized
by a facile one-pot modulator method, whereas the synthesis of nanoMOFs below
10 nm typically requires multiphase systems, dropwise additions, or multiple
modulator ligands. (172;173;174) Figure 1e depicts a simulated structure of 5.5-nm
Fe(TA)2, confirming that such a small size still includes many pores due to the
unusually high density of this MOF. SEM analysis of the 16 and 25 nm particles
reveals unimodal and symmetrical size distributions and essentially spherical
morphologies (Fig. 4.2). Although the two smallest particle sizes synthesized were
not well-resolved by SEM, the images clearly reveal small nanoscale particles (Fig.
C.7). Instead, particle sizing was accomplished by Scherrer analysis. To confirm
the domain sizes from PXRD, we additionally performed Le Bail refinements
on patterns of the largest and smallest particles and found the domain sizes
agree with SEM imaging for the 130 nm particles and Scherrer analysis for the
5.5 nm particles (Table C.1, Fig. C.2). Scherrer analysis, as expected, produces
underestimated values, particularly as particle sizes increase. Microscopy of larger
particles (0.055 eq 1-mIm, 130 nm) show octahedral faceting, akin to the bulk
MOF product. Size distributions also skew towards larger sizes when the modulator
59
equivalents are low, indicating that particles may grow by non-classical mechanisms
such as digestive ripening or interparticle coalescence. (175;176) The dispersity of the
particle populations can be quantified by dividing the standard deviation given
from a Gaussian fit by the mean size; these values vary between 0.09 and 0.18.
These dispersity values are considerably lower than the typical dispersity values of
nanoMOFs, which are often as high as 0.6, and rival the archetype of low-dispersity
MOF particles, ZIF-8, which also exhibits values around 0.09. (177;8;32) Mössbauer
spectra of the smallest particles before and after air exposure were collected to
determine the efficacy of our air-free synthesis. In the air-free sample, Fe3+ is not
observed (Fig. 4.3).
Figure 4.3. Mössbauer spectra of the smallest FeTA2 samples under N2. In addition
to the main peak, indicative of low-spin Fe2+ in an octahedral environment, there is
a shoulder at high isomer shift that may be related to undercoordinated iron sites
at the surface.
As the smallest particles exhibit the highest external surface area and would
therefore oxidize readily, these data suggest that all samples synthesized, stored,
and measured air-free exist in the fully ferrous state. Whereas modulators may
incorporate as internal defects or surface ligands in other MOF systems, 1H acid
digestion NMR indicates that 1-mIm does not incorporate in most cases (Fig. C.8).
Remarkably, despite the lack of conventional capping ligands, particles of all sizes
60
exhibit colloidal stability in DMF under anaerobic conditions for at least three
months, even at relatively high concentrations of 10-20 mg/mL. Dynamic light
scattering measurements of 25-nm particles stored for three months gave a solvated
diameter of 60 ± 10 nm, a reasonable increase from the SEM images of dried
particles that did not exhibit aggregation (Fig. C.10). Therefore, these Fe(TA)2
particles exhibit superior long-term stability compared to previously reported MOF
systems, which, with the notable exception of ZIF-8 in methanol, are typically
coated with polymers or surfactants to achieve colloidal stability. (178;179)
To explore the general application of this nanocrystal synthetic route,
we investigated 5-bromo-1-methylimidazole, 1-benzyl-2-methylimidazole, and n-
butylamine (nBA) as alternatives to 1-mIm. Due to their differences in Lewis
and Brønsted-Lowry basicities, this series of modulators offers a platform for
exploring the competing equilibria outlined in Figure 4.1b. Specifically, we expected
that nBA, as the strongest ligand and proton accepter, would yield the smallest
sizes. Figure 4.2f plots the resulting particle sizes using 0.218 equivalents of
each modulator, indicating that each of these alternative modulators produces
nanoparticles, with nBA giving the smallest sizes and the bromo-substituted
variant, i.e., the weakest ligand, producing the largest sizes. This trend correlates
with the strength of the N-Fe bond, as the weakly electron-donating benzyl group
allows the imidazole nitrogen to increase its sigma-donating ability, whereas the
electronegative bromine moiety withdraws electron density from the N-Fe bond.
This size trend therefore provides further evidence that modulators compete with
the triazole MOF linkers. SEM images confirms that the resulting particles are also
uniform (Fig. 4.4)
61
Figure 4.4. SEM images of particles synthesized with different modulators. a)
Fe(TA)2 synthesized with 0.218 eq of n-butylamine. b) Fe(TA)2 synthesized
with 0.218 eq of 1-benzyl-2-methylimdaizole. c) Fe(TA)2 synthesized with 0.218
equivalents of 5-bromo-1-methylimidazole. d) Fe(TA)2 product synthesized with
10.9 eq 5-bromo-1-methylimidazole showing small aggregated particles and long
string-like structures.
High equivalents of n-butylamine (>1 eq) formed an amorphous precipitate
at room temperature (Fig. C.1a). Interestingly, if heated and stirred, this gel-
like reaction mixture still produced phase-pure Fe(TA)2, although samples using
this modulator suffered from poor colloidal stability. High equivalents of 5-
bromo-1-methylimidazole resulted in significant particle aggregation as well as an
unidentified phase impurity (Fig. 4.4d, C.1c,). With 10.9 equivalents of 5-bromo-
1-methylimidazole, the PXRD peaks also appear to shift to slightly lower angles,
62
indicating a small increase in unit cell size. Additionally, particles synthesized with
5-bromo-1-methylimidazole often suffered from poor colloidal stability. The use
of 1-benzylimidazole resulted in small Scherrer sizes, although a possible phase
impurity is also observed (Fig. C.1b).
We therefore chose to move forward with 1-methylimidazole, as the products
were consistently phase-pure and small particles were obtained at reasonably low
equivalents. We also employed more dilute conditions compared to the previously
reported bulk synthesis. To ensure that the results observed were due to the
modulator and short reaction time, we performed our bulk synthesis under the
same concentration (0.0575 M FeCl2, Fig. C.6a). Encouraged by the results with
Fe(TA)2, we explored the synthesis of Co(TA)2 and Cd(TA)2 with the addition
of 1-methylimidazole and observed a similar size trend by Scherrer analysis and
SEM imaging (Fig. 4.5, 4.6). Indeed, this synthetic strategy yielded nanoparticles
of both materials, showing that the addition of 1-mIm as a modulator may be a
generally applicable method to synthesize nanoMOFs comprised of triazole linkers.
63
Figure 4.5. Synthesis of nanosized Co(TA)2 using 1-mIm as a modulator. a) PXRD
patterns of the bulk synthesis of CoTA2 compared to the nanosized synthesis using
varying equivalents of 1-methylimidazole. b) Particle diameter trend with respect
to 1-methylimidazole equivalents. Filled squares are sizing from SEM, measuring
>200 particles. Empty squares are Scherrer crystallite sizes. SEM images and size
histograms for (c) the bulk synthesis with no modulator, (d) 0.05 equivalents of
1-mIm, and (e) 0.3 1-mIm equivalents. f) SEM image for 0.5 1-mIm equivalents;
crystallite sizes from Scherrer analysis follow the expected trend but SEM shows
poor morphology.
64
Figure 4.6. Size trends of new Cd(TA)2 products modulated by 1-methylimidazole.
Particles synthesized with a) 0.5 equivalents, b) 1.0 equivalents, c) 1.25 equivalents,
d) 1.5 equivalents, and e) 1.75 equivalents with respect to Cd in the reaction. f)
Particle sizes determined from SEM imaging vs 1-methylimidazole equivalents.
65
Size-Dependent Optical Properties of Colloidal Fe(TA)2 Nanoparticles
The remarkable colloidal stability and small sizes of Fe(TA)2 nanoparticles
yielded suspensions with minimal light scattering, as seen in the inset photo of
Fig. 4.7a. In fact, diluting colloidal suspensions of Fe(TA)2 results in samples
with sufficient transparency for solution-state UV-Vis absorption spectroscopy. In
general, any spectroscopy of MOF materials involves analysis of solid-state samples,
with UV-Vis analysis demanding diffuse reflectance methods. Figure 4.7a shows
solution-state UV-Vis spectra for all Fe(TA)2 particle sizes. These spectra comprise
some of the only examples of solution-state spectroscopy of MOF materials, which
include isolated reports of MOF-525, TMU-34, and Zn-MOF-74. (180;179;178;112) To
the best of our knowledge, this report is the first to quantify fundamental optical
properties from the absorbance spectra of MOF particles in solution. The spectra
exhibit the same two MLCT bands (ca. 4.2 eV and 3.3 eV) as observed in the bulk
material (Fig. 4.7a, shaded grey). Due to significant scattering observed in the
spectra of larger particle sizes, the low-intensity d-d transition expected at 2.36 eV
for the low-spin-state material from the bulk spectrum is obscured in all spectra.
Notably, the spectra lack an inter-valence-charge-transfer (IVCT) band around 1.24
eV that arises from mixed Fe2+/3+ valency, suggesting the particles exist in a fully
ferrous state, in agreement with the Mössbauer spectra showing only Fe2+ (Fig.
4.3). (20)
66
Figure 4.7. Solution-state UV-Vis absorption spectra of Fe(TA)2 nanoparticles. a)
Normalized spectra of colloidal Fe(TA)2 nanoparticles of varying sizes suspended in
DMF (colored traces) compared with the bulk MOF (filled grey area) as reported
by diffuse reflectance. (20) Smallest and largest particles traces are emphasized. Inset
contains a photo of the smallest and largest particles as solutions in cuvettes. b)
Normalized CT2 band and c) corresponding peak maxima versus particles sizes. d)
Normalized CT1 band and e) corresponding peak maxima versus particles sizes.
Peak maxima are reported for syntheses modulated by 1-mIm except for data
labeled for n-butylamine (nBA) and 1-benzyl-2-methylimidazole (Benzyl).
Close inspection of the solution-state spectra reveal that the peak maxima
of the two charge transfer bands decrease in energy with increasing particle sizes,
while a shoulder emerges at energies below the lower-energy band (Fig. 4.7). For
the higher energy band (CT1), the maximum shifts a total of 0.22 eV, while the
lower band (CT2) maximum shifts 0.11 eV. In the bulk material, the λmax of
CT1 appears at a lower energy of 3.99 eV, and CT2 appears at 3.47 eV, within
range of the λmax of the nanoparticles. Although the CT2 band is split in the bulk
spectrum, the peak-to-peak separation appears more extreme in the nanoparticles,
with the lower-energy shoulder appearing at much lower energies (Fig. C.15).
These data represent the first examples of size-dependent shifts to optical
properties of MOF materials. To determine whether the modulator plays a role in
67
the size dependence, UV-Vis spectra were collected of particles synthesized with n-
butylamine and 1-benzyl-2-methylimidazole. Because modulated MOF syntheses
often introduce defects, we anticipated the modulator identity to influence the
extent of defect incorporation in Fe(TA)2 nanoparticles.
(112) Interestingly, the
λmax values for particles prepared with these alternative modulators are similar
to Fe(TA)2 particles prepared with 1-mIm. Therefore, the size-dependent optical
behavior is unlikely due to modulator-induced defects and is reproducible (Fig.
4.7c, 4.7e).
The shift in the peak maxima suggest that the electronic structure of
Fe(TA)2 changes with particle size. Several physical scenarios might explain this
observation. The first possibility is quantum confinement; particle sizes smaller
than the excitonic Bohr radius of Fe(TA)2 would exhibit blue-shifted absorption
events. Previous electronic structure calculations of Fe(TA)2, however, depict
relatively shallow band curvatures and therefore low charge carrier mobilities, which
would result in an insufficiently large excitonic radii for quantum confinement. (181)
Additionally, band-gap transitions of quantum-confined materials linearly increase
with 1/r2, based on the effective mass approximation, or as 1/r due to Coulombic
electron-hole interactions. (182) However, a linear trend is not observed in this
transition with respect to 1/r2 or 1/r (Fig. C.15). A second explanation may be
that the observed optical shifts arise from a size-dependent magnetic transition.
Previous studies of bulk Fe(TA)2 indicate that low-spin Fe
2+ centers undergo
spin-crossover transitions at 290 ◦C, (21) but this transition temperature might
occur at lower temperatures for smaller particles due to decreased cooperativity
of smaller domain sizes. For example, the spin-crossover temperatures of related
Fe-based 1-D coordination polymers decrease with decreasing particle sizes. (183)
68
The distinct electronic configurations of high- and low-spin states would produce
different sets of spectra, as is well-documented for spin-crossover materials. (184;185)
Preliminary PXRD analysis, however, suggests the nanoparticles exist in the
denser low-spin crystallographic phase (Fig. 4.1d). (21) Additionally, we observe
by differential scanning calorimetry (DSC) that the LS-HS phase transition still
occurs far above room temperature for even the smallest Fe(TA)2 particles (Fig.
C.11). Partial oxidation of iron sites is another possibility, as a change in shape
of the charge transfer bands has been reported for Fe(TA)2 bulk powders after
chemical oxidation. (20) This scenario is unlikely, however, as the colloids are stored
and measured air-free, and the spectra lack an intervalence charge-transfer band
and Mössbauer spectra of the smallest particles exhibit only a ferrous species
(Fig. 4.3). Finally, structural distortions due to the high surface-to-volume ratios
of small particles could be responsible for the size-dependent spectral shifts as
optical shifts arising from surface restructuring has been reported in several metal-
oxide nanoparticle systems. (186) Specifically, the localized orbitals of surface, sub-
surface, and internal species in different geometries could contribute different
absorption bands to UV-vis profile, with surface species playing the greatest role
in the smallest particles. Delocalized wavefunctions, on the other hand, might
respond coherently to structural distortions, contorting the shapes of valence band
and conduction band orbitals and shifting their relative energies. The presence of
surface defects is supported by N2 sorption experiments, which show that the 48-
nm particles display a lower accessible surface area than the bulk material despite
their higher surface area to volume ratio (Fig. 4.8).
69
Figure 4.8. Nitrogen isotherms of bulk iron triazolate compared to 48-nm
nanoparticles. The nanoparticles were synthesized with 0.218 equivalents 1-
methylimidazole. Samples were washed with DMF (2), MeCN (2), and DCM (5)
and evaporated to dryness. Samples were degassed at 120 ◦C for 24-48 hours,
until the degas rate was below 2.5 µtorr/min. a) N2 adsorption (blue circles) and
desorption (red triangles) isotherms for bulk and nano Fe(TA)2. b) BET plot for
bulk Fe(TA)2. c) PXRD of Fe(TA)2 nanoparticles before and after measurement. d)
BET plot for Fe(TA)2 nanoparticles.
This strong size-dependence in the absorption energies has never before been
observed for MOFs. Although the exact cause is yet unknown, we have ruled out
several possibilities and are investigating the size-dependence further in on-going
studies.
Conventional spectroscopy of bulk MOFs inhibits interpretation of
absorption intensities, whereas solution state measurements allow for the
determination of extinction coefficients (ε) for all Fe(TA)2 particle sizes.
Quantitative analysis of extinction coefficients has driven the quantum mechanical
70
understanding of semiconductor nanocrystal optical phenomena by relating
optical oscillator strengths (f ) to detailed information about electronic structure
and by providing a practical estimation of particle concentrations from optical
spectra. (187;188) Such analysis is currently limited for MOF materials; to date
Figure 4.9. Extinction coefficients of Fe(TA)2 nanoparticles. a) CT1 (blue)
and CT2 (black) extinction coefficients plotted as ε per particle versus particle
diameter. Solid curves are fits of ε to cubic functions of diameter. Pre-factors
for CT1 and CT2 fits are 18013 and 15100, respectively. b) UV-Vis traces for all
particle sizes plotted as extinction coefficient ε per particle. c) Oscillator strength
determined per particle for the two charge transfer bands as well as the shoulder
(sh) of CT2.
only one report discusses the dependence of optical properties—in this case
fluorescence—on particle size and defectiveness. (112) Figure 4.9a plots the extinction
coefficients of peak maxima versus particle diameters, which increase by four
71
orders of magnitude, from 106 (cm−1 M−1) for 5.5 nm particles to 1010 (cm−1
M−1) for 130 nm particles. Extinction coefficients were determined using a linear
relationship between concentration and extinction at low concentration, with both
absorption and scattering processes contributing to the overall extinction (Fig.
C.14). It is evident from the spectra alone that the extinction coefficients depend
strongly on particle sizes, with larger sizes absorbing more intensely than smaller
particles, as expected from the increased number of absorber units (Fig 4.9b).
To explain this dramatic increase in absorber strength, the extinction coefficient
data were fitted to cubic functions of particle diameters, as has been demonstrated
for lead chalcogenide semiconductor nanocrystals based on the hypothesis that
absorber strength per particle arises from increased particle volumes. (189) A
cubic relationship produces an excellent fit except for larger sizes, in which the
experimental data appears depressed in relation to the expectic cubic trend, most
likely due to light scattering (Fig. 4.9a). While greater extinction coefficients may
be expected for larger particles, the absorption strength per Fe(TA)2 formula unit
depends on particle diameters as well, ranging from 3000 – 7000 cm−1 M−1, which
are typical values observed for molecular charge transfer bands (Fig. 4.10a).
72
Figure 4.10. The extinction coefficient and oscillator strength of Fe(TA)2
nanoparticles per formula unit. a) The extinction coefficient of Fe(TA)2
nanoparticles per formula unit. Generally, the extinction coefficient per formula
unit increases as size increases, although the 16 nm particles and 130 nm particles
deviate from this trend. b) Oscillator strength determined per formula unit
Fe(TA)2. While the oscillator strength per particle clearly trends upwards with
size, here the value appears to peak for the 25 nm particles. Values do not change
significantly, however, and could be considered consistent across particle sizes.
For a deeper analysis of absorption intensities, oscillator strengths f can
be determined for each of the absorption bands. When calculated per particle,
f increases with particle sizes (Fig. 4.9c), as expected for an increased number
of absorbing units. Determined per formula unit, the oscillator strength changes
minimally with particle sizes, with f values ranging from 0.05 to 0.27 (Fig.
4.10b). In all cases, f per formula unit is on the order of 10−1, as expected for
a charge transfer that is both spin and parity allowed. Combined, these results
show that colloidal stability enables powerful quantitative analysis of the optical
properties of MOF nanocrystals. Although the analysis herein does not yet consider
possible non-linear contributions of scattering to total extinction, this type of
analysis represents a step forward for understanding the optical properties of MOF
structures. (190;191)
73
Redox Chemistry and Charge Transport of Fe(TA)2 Nanoparticles
Although solid-state techniques and additives such as polymer binders
are required to study redox properties of bulk MOFs, the colloidal stability of
Fe(TA)2 nanoparticles enabled characterization by solution-state electrochemistry.
Figure 4.12a shows cyclic voltammogram (CV) traces for several particle sizes
using standard 3-electrode cell configurations. These four smallest-sized particles
were stable in 0.1-M TBAPF6 / DMF electrolyte mixtures under applied bias,
while attempts to analyze 48-nm particles failed, likely due to aggregation, as the
particles precipitated from the stored electrolyte solution over the course of a few
days. Although we expected only a single redox event corresponding to Fe2+/3+
oxidation, all samples exhibited qualitatively similar traces with three broad, quasi-
reversible features appearing at similar applied potentials. The 16-nm colloidal
suspension was further investigated due to the well-defined features of its CV trace.
As summarized in Figure 4.11d, variable scan rate measurements showed a strong
dependence of the peak-to-peak separation ∆E for the lowest-potential redox event
at around -0.61 V (purple closed triangles), ranging from 9.2 mV at 100 mV/s to
56 mV at 130 mV/s, while the second event, centered at -0.29 V (purple open
circles), showed less reversibility, with ∆E of 126 mV at 100 mV/s to 144 mV at
130 mV/s (Fig. 4.11d), while ∆E for the highest-potential feature, centered around
0.13 V, could not be resolved.
74
Figure 4.11. Cyclic voltammetry of Fe(TA)2 nanoparticles analyzed as colloids or
thin films. Initial scan directions are indicated by arrows. a) CV traces collected at
a 130 mV/s scan rate for four particle sizes prepared as colloids in 0.1-M TBAPF6
/ DMF; current density is normalized to the second faradaic event. Scan rate
dependence of 16-nm particles drop-casted onto glassy carbon in 0.1-M TBAPF6 /
MeCN (b) and TBABF4/MeCN (c). Light to dark greyscale traces correspond to
10 mV/s – 500 mV/s. Relevant peaks are marked with filled triangles (first peak
analyzed) or open circles (second peak analyzed) in all panels. d) Peak-to-peak
separation with respect to scan rate. e) Peak current with respect to scan rate for
two reduction peaks in each of the 16 nm particle CVs. Dashed lines correspond to
linear fits.
To investigate the origin of these redox events and to demonstrate that the
solution processability of colloidal Fe(TA)2 nanoparticles will facilitate their use
in MOF-based electrochemical devices, 16-nm Fe(TA)2 were drop-casted onto the
surface of glassy carbon electrodes. Figure 4.11b shows the variable-scan rate CVs
of the particle films in 0.1 M TBAPF6 / MeCN. Despite the difference in solvent,
which was chosen to discourage particle delamination, the CV traces resemble those
of the free-standing colloids in Figure 4.11a. Due to the small pore size of Fe(TA)2,
we hypothesized that the broad waves of the voltammetric response reflect the
75
hindered ability of the bulky PF–6 anion to diffuse through the material. Therefore,
these experiments were repeated on particle films using TBABF4, as previous
computational work has shown that the BF−4 anion is just below the pore size of
the MOF. (20) In contrast to voltammetry with TBAPF6, the use of TBABF4 causes
all peaks to sharpen and induces new voltammetric responses, most notably a
sharp, reversible feature at 1.2 V (Fig. 4.11c). For direct quantitative comparison,
we investigated two well-defined features from each set of electrolyte experiments,
as indicated by open circles and shaded triangles throughout Figure 4.11. Figure
4.11d shows that in both electrolyte media, ∆E is always smaller for the lower-
potential event, supporting the assignment of this peak to a surface Fe2+/3+ species,
which benefits from both better ion-pairing and faster kinetics of ion diffusion.
Additionally, the second event (pink circles) for a film in TBAPF6 shows the
largest ∆E among all conditions, including the colloidal suspension in TBAPF6
/ DMF (purple circles). Interestingly, as shown in Figure 4.11e, the peak current
for the events in all experiments exhibit a linear dependence with respect to scan
rate. Therefore, even if the particles are employed as colloids, the redox events are
adsorption-controlled. (192)
To precisely quantify the role of ions in the redox chemistry of Fe(TA)2
particles, we employed quartz crystal microbalance (QCM) electrodes. Spin-
coating 16-nm particles onto the QCMs yielded uniform multi-layer nanoparticle
films (Fig. C.21), which allowed us to measure their voltammetric responses and
simultaneous mass changes in either TBAPF6 or TBABF4 environments. The
frequency of the quartz crystal oscillation is sensitive to mass changes at the crystal
surface, allowing monitoring of film loading and ion flux while the potential is
scanned. Figures 4.12a and 4.12b show both the CV traces and the number of
76
moles of anion adsorbed to the particles based on the mass change of the QCM
electrode, using the Sauerbrey equation (193) and assuming that all mass change
is due to unsolvated PF6 and BF4 anions. Once again, the redox waves of the
sample analyzed with TBAPF6 exhibits broad features in comparison to the same
size nanoparticles in TBABF4 (Fig. 4.11, Fig. C.20). Additionally, substantially
more current passes to the particles in the presence of TBABF4. Charge integration
shows nearly seven-fold enhancement, with stoichiometric oxidation of nearly all
Fe sites, i.e., e–:Fe = 0.9 when using TBABF4, whereas TBAPF6 yields only
e–:Fe = 0.3. This comparison suggests that more Fe sites are electrochemically
accessible in the TBABF4 experiment, which we attribute mainly to the ability
of the smaller anions to diffuse through the porous MOF particles, while bulky
PF6 anions merely collect at the nanoparticle surface. This result contrasts with
studies of ion diffusion in MOFs with large pores, wherein ion pairing hinders
small ions to a greater degree than larger ions. (194) A further contribution may
be differences in thin thickness and morphology, although films were homogenous
for both cases (Fig. C.24, Table C.7). The mass of the particle films increases in
both experiments with oxidizing potentials, as expected for incorporation of charge-
balancing anions, but far more anions incorporate when using the smaller BF–4 ions.
Approximately four times as many moles of BF– adsorb compared to PF–4 6, yielding
PF6:Fe = 0.1 and BF4:Fe = 0.4 (Table C.7). We expect that surface ions provide
additional charge-balancing anions needed for stoichiometric oxidation of all Fe
sites. Comparing mass changes to redox waves at a given potential provides further
insight into the nature of the redox chemistry. Most notably, the coincidence of a
sharp increase in mass and current at ca. 1.2 V strongly suggest this redox event
corresponds to ion-coupled charge transport to interior Fe sites enabled by the
77
smaller size of the BF−4 anion in comparison to the gradual mass changes and broad
redox waves of the TBAPF6 experiments. None of the other features in the CV
traces are obviously associated with an increase in mass; rather, in TBAPF6 and
at low potentials in TBABF4, the mass increases gradually as the cell is scanned
oxidatively. A gradual increase in mass therefore indicates that for these features,
ions are collecting at the particle surface rather than diffusing through the pores.
The observation of several voltammetric features spanning a wide potential window
is surprising; however, it is expected that surface species would be oxidized at a
lower potential than internal iron sites that require significant ion intercalation.
These data stand in stark contrast to two prior reports of bulk Fe(TA)2 cyclic
voltammetry. In one study, scanning the bulk material in an air-free environment
in LiBF4 / propylene carbonate showed a single redox feature at high potential,
assigned to Fe2+/3+. (20) In the other report, the material was synthesized in aerobic
conditions and scanned in 0.1-M KOH, showing several redox events assigned
to varying Fe surface species. (195) Here, we assign the sharp reversible event at
1.2 V in TBABF to Fe2+/3+4 , while broad events at lower potential in either
electrolyte are assigned to a combination of capacitive charging and surface Fe2+/3+
species. Cyclic voltammetry of the 1,2,3-triazole linker in 0.1-M TBAPF6 / DMF
reveals a chemically irreversible event at 0.31 V vs Fc+/0 (Fig. C.17), supporting
the assignment of redox waves to Fe2+/3+. Although ion intercalation and ion
pairing are frequently invoked to understand MOF redox chemistry, charge storage
and sensing, these data represent some of the only direct measurements of ion
intercalation processes by employing quartz crystal microbalance electrodes. (194;196)
They also show that nanosizing MOFs can improve the availability of redox-active
78
sites, thus lowering redox potentials through improved ion pairing, a key insight
into designing electronic MOF devices.
Figure 4.12. Cyclic voltammetry with superimposed QCM data of Fe(TA)2
nanoparticles. Initial scan directions are indicated by arrows. CV traces collected
at 100 mV/s with a Pt QCM electrode for 16-nm particles in TBAPF6 / MeCN (a)
and TBABF4. (b) Colored open circles refer to the right axis, moles of anions with
respect to moles Fe(TA)2 on the QCMs.
Thin film fabrication and solid-state measurements are enabled by the
long-term colloidal stability and solution-processability of Fe(TA)2 nanoparticles.
Doctor blading high-concentration suspensions onto glass slides afforded uniform
films with smooth surfaces, as shown in the cross-sectional FIB-SEM images in
Figure 4.13. Additional SEM images and photos of particle films can be found
in Appendix C (Fig. C.22-C.25). Figure 4.13b shows conductivities of thin films
under N2 for the three particle sizes: the conductivity increases from 6.33 nS/cm
79
for the bulk sample to 50.6 nA/cm for 84 nm, and finally to 611 nS/cm for a film
comprised of 25-nm particles. A previous report of bulk Fe(TA)2 measured values
of 0.1 nS/cm for a sample kept air-free. (20) Although the values herein are higher,
the samples still lack oxidation, as even the smallest particles (5.5 nm) handled
air-free do not show Fe3+ by Mössbauser spectroscopy (Fig. 4.3). After three days
of air exposure, the conductivity of the samples increases to 1.21 µS/cm (Bulk),
8.25 µS/cm (84 nm), and 91.0 µS/cm (25 nm) (Fig. 4.13c). The values collected in
air agree with the first reported value of 77 µS/cm; the bulk value herein is likely
lower due to the air exposure time only lasting three days. (19) These measurements
suggest a size-dependence in the charge transport behavior of Fe(TA)2, with the
conductivity increasing as the particle size decreases. While this trend may be
unexpected due to smaller particles introducing more grain boundaries, the cross-
sectional SEM images indicate that nanoparticles exhibit denser inter-particle
packing in comparison to the bulk material (Fig. 4.13). Hence, despite the largest
grain size, the overall conductivity of the bulk sample still suffers in comparison
to well-packed nanoparticles. After air exposure, the conductivity of all samples
increases 2 to 3 orders of magnitude, most likely due to oxidation and the creation
of mixed valency. Mössbauser data shows that samples allowed to oxidize in air
varying small amounts of Fe3+ but the degree of oxidation shows no trend with
respect to particle size (Fig. C.12). Therefore, the main cause of size-dependence
in the conductivity values arises from efficient particle packing. These results imply
that grain boundaries in Fe(TA)2 may present relatively shallow barriers to charge
transport. Notably, in polycrystalline organic semiconductor films, smaller grain
size does not always decrease conductivity; transport in such systems depends
extensively on microstructure and grain boundary site energy. (197;198;199) Further
80
investigations are ongoing into the charge transport mechanisms of Fe(TA)2
nanoparticles, including the dependence on size, redox-state, and guest-host
interactions for technologies ranging from charge storage to electrochemical sensing.
Figure 4.13. Charge transport measurements of Fe(TA)2 nanoparticle thin
films. a) Photograph of a thin film prepared under N2 by doctor-blading 84-nm
nanoparticles. b) Van der Pauw DC conductivity values of three Fe(TA)2 thin films
under N2 and after three days of air exposure. Values for samples in air are in
turquoise. Inset shows a zoomed-in view of conductivity values of the films under
N2. Error is derived from the ratio between the probe width and the sample width.
d) FIB-SEM image of a cross-section of a representative bulk film. e) FIB-SEM
cross-section of a representative 84-nm film, showing dense inter-particle packing.
81
Conclusions
In summary, conductive Fe(TA)2 nanoparticles can be prepared reproducibly
with excellent colloidal stability. The solution processability of this unprecedented
class of semiconductor nanocrystals enables solution-state spectroscopy and
electrochemistry, whereas MOF characterization typically requires solid-state
techniques. These measurements reveal an unexpected size dependence to the
optical transitions and enable the first analysis of MOF extinction coefficients,
which scale with particle size. Fe(TA)2 particles can be probed by colloidal and
thin-film voltammetry, revealing show redox chemistry sensitive to ion pairing and
intercalation effects within the porous materials. Finally, we demonstrated that the
particles can be easily processed into thin films for charge transport measurements
that reveal increased conductivity compared to the bulk material.
82
CHAPTER V
NANOCONFINEMENT EFFECT AND REDOX CHEMISTRY IN FE(TA)2
NANOPARTICLE THIN FILMS
This chapter includes unpublished and co-authored material from the
prospective manuscript Huang, J,; Marshall, C.R.; Ojha, K.; Shen, M.; Kadota,
K.; McKenzie, J.; Fabrizio, K.; Brozek, C.K. Giant Redox Entropy in Intercalation
Chemistry of Nanoconfined MOF Nanocrystals. The project in this chapter was
conceptualized and developed by Checkers R. Marshall, Jiawei Huang, and Carl
Brozek. The manuscript was written by Jiawei Huang, with editorial assistance
from Checkers R. Marshall and Carl K. Brozek; the manuscript was edited by
Checkers R. Marshall for this dissertation. The experimental work in this chapter
was performed by Jiawei Huang, Kasinath Ojha, Checkers R. Marshall, Meikun
Shen, and Jacob McKenzie.
Metal-organic frameworks (MOFs) have drawn tremendous scientific interest
in the field of electrochemical energy storage and conversion. Their tunable porous
structures allow facile transportation of electrolyte species, and the large surface
area provides a high density of electroactive sites for driving electrochemical
processes. Distinct from electrochemical reactions on the non-porous electrode
surface, electrolyte species can intercalate into the MOF pores to drive redox events
via ion-coupled charge transport at the interior electrochemical interface. Previous
studies into porous carbon-based materials and two-dimensional layered transition
metal oxides have reported that the nanoconfinement effect inside the pore
environment alters the solvation and intercalation structures of electrolyte species,
significantly impacting the electrochemical properties of porous materials. (200;201)
83
Figure 5.1. Scheme of ion intercalation and redox chemistry in FeTA2 film. (a)
FeTA2 film in TBABF4 electrolyte in acetonitrile under open circuit voltage
(OCV). (b) BF−4 -induced Fe
2+/Fe3+ redox via the partial desolvation and
adsorption of BF−4 anions on the surface Fe sites. (c) BF
−
4 -induced Fe
2+/Fe3+
redox induced by the complete desolvation and intercalation of BF−4 anions into
the nanoconfined FeTA2 pore. Note: green, orange, gray, red, blue, and purple dots
represent Fe2+, Fe3+, C, F, N, and B atoms, respectively. Blue cloud-shaped area
represents the solvation shell of electrolyte ions.
By comparison, a fundamental understanding of the ion intercalation redox
chemistry inside the nanoconfined MOF pore and its implications for practical
applications still remain elusive, mainly due to the difficulty of differentiating
intercalation redox chemistry inside the pore from the reactions on the surface of
the MOF electrode. Previous studies on the ferrocenecarboxylic acid-functionalized
MOF-808/Nu-1000/NU-1003 series of MOFs, which are reticular structures with
varying pore size, show clearly that ion diffusion increases with increasing pore
84
size, while electron transport shows the opposite trend. (202) Meanwhile, in the
MOF Zn(dcph-Oh-NDI), ion diffusion was explored with different ion sizes and
it was found that strong interactions between the ions and the reduced redox-
active napthalenediimide linkers can hinder ion movement. (203) To date, the studies
involving charge transport on MOFs are generally those with large pore size, with
ample room for ion diffusion. Herein we explore a dense MOF, Fe(1,2,3-triazolate)2,
as we expect that a small pore size will lead to severe nanoconfinement effects.
Solvated BF−4 anions have a diameter of 10 Å in acetonitrile, and the Fe(TA)2
pore is merely 4.54 Å. Desolvating the BF4 anions entirely gives a radius of 3
Å, which would allow for ion intercalation, but the desolvaiton process requires
energetic input. Therefore, in the previous chapter, we hypothesized that the
voltammetric response of Fe(TA)2 included both surface redox reactions at low
oxidizing potentials, and intercalation redox chemistry at much higher potentials.
Herein, we unravel how the nanoconfinement effect impacts the ion-intercalation
redox chemistry inside Fe(TA)2 (TA = 1,2,3-triazolate) nanoparticles and assign the
redox features to surface and intercalation events (Fig. 5.1).
Fe(TA)2 nanoparticles with tunable sizes were synthesized by following our
recent work, where different amounts of 1-methylimidazole as a “modulator” were
added into bulk synthesis procedures. (204) Scanning electron microscopy (SEM)
analysis showed that nanoparticles of Fe(TA)2 MOFs with sizes of 16 nm, 25 nm,
and 46 nm were prepared (Fig. D.1). Powder X-ray diffraction (PXRD) patterns
further confirmed the retention of the crystalline structure of obtained Fe(TA)2
nanoparticles (Fig. D.1). Those nanoparticles were spin-coated into a Fe(TA)2 film
on the quartz crystal electrode for electrochemical investigations. Previous reports
indicate that the Fe centers in Fe(TA)2 are redox-active,
(20;195) making this
85
Figure 5.2. CV scans and corresponding mass change data from EQCM for three
Fe(TA)2 nanoparticle thin films: a) 16 nm particles (4.0 µg), b) 25 nm particles
(4.0 µg) and c) 48 nm particles (5.7 µg). Asterisk indicates an irreverible event
arising from the MOF linker. Arrows indicate the direction of the scan from open
circuit potential; for clarity, the end of the QCM loop returning to open circuit has
been removed.
86
small-pore MOF an ideal model to study ion intercalation in well-ordered porous
materials.
Identifying Intercalation Chemistry in Fe(TA)2 Films
THe CV scans of Fe(TA)2 films, composed of particles of varying size,
exhibit similar voltammetric features in the potential range from -0.4 V to 0.2
V vs. Fc0/+ (Fig. 5.2). All films show several broad voltammetric features at
low potential, and a sharp and reversible feature at ca. 1.2 V vs. Fc0/+, which
is independent of other voltammetric features in the CV trace (Fig. D.2). The
difference in voltage between the various features suggests that the Fe sites within
Fe(TA)2 are fundamentally different in nature. Previous studies have shown that
the position of Fe2+/3+ redox can be tuned by manipulating the Fe coordination
sphere. (205) In the case of the ideal Fe(TA)2 material, all Fe sites are octahdral and
coordinated by 6 N-donating triazole ligands. However, nanoparticles often exhibit
significant surface defects, which may contribute to coordination site changes and a
resulting shift in redox potential.
Even though coordination environment can shift redox potential, this
is insufficient to explain the event at 1.2 V vs Fc0/+, which unusally sharp and
reversible. Such a sharp “butterfly” signature has previously been attributed
to adsorbates on the surface of noble metal electrodes undergoing a disorder-
order phase transition; notably, however, these events are not associated with
any increase in mass as the adsorbates are already on the electrode surface. (206)
Accordingly, we monitored the potential-dependent mass variation on Fe(TA)2 film
using in situ electrochemical quartz crystal microbalance (EQCM). A significant
mass increase and decrease was observed at the 1.2 V feature in the anodic and
cathodic direction of CV trace, respectively (Fig. 5.2), excluding the possibility
87
that the 1.2 V voltammetric feature was attributable to the phase change of
absorbates on the Fe(TA)2 film surface. The total mass change per formula unit
Fe(TA)2 only increases slightly as particle size increases. However, if normalized
to particle volume, the mass change of the 1.2 V feature shows a linear trend,
indicative that the feature originates from the intercalation of BF−4 anions into
the confined MOF pores (Fig. 5.3b). Reversibility also suffers in the case of the
48-nm particles; both the voltammetric features as well as the QCM loop are
noticeably wider in this sample, most likley due to an intrinsically longer diffusion
pathlegnth across the particles’ diameter (Fig. 5.2c). Lower-energy redox events
show a mass increase that is proportional to the surface area of the nanoparticles,
further indicating that surface and intercalation redox events are distinct in this
system (Fig. 5.3a).
Figure 5.3. Plotting the change in mass on the QCM electrode over the voltage
range of surface and intercalation features. a) The feature at 1.2 V vs Fc0/+ for
the three films of differently-sized particles vs the particle volume shows a linear
relationship. b) The feature at 0.0 V instead shows a linear relationship with
respect to paricle surface area.
To support the assignment of the 1.2 V feature to BF−4 anion intercalation
redox reaction, time-of-flight secondary ion mass spectrometry (ToF-SIMS) was
88
employed to map the depth distribution of both BF−4 anions and TBA
+ cations
via sputtering through the Fe(TA)2 film. Given that bulky TBA
+ cations cannot
intercalate into the pore of Fe(TA)2 nanoparticles, the distribution of TBA
+ cations
can be used as a reference to identify the surface region of Fe(TA)2 film in ToF-
SIMS results (denoted as the gray area throughout Fig. 5.4). After applying
the reducing potential of -1.0 V vs. Fc0/+ to Fe(TA)2 film, the depth-dependent
boron signal, originating from the bond cleavage of BF−4 anions during the surface
sputtering, was only detected within the surface region (Fig. 5.4a(i)). This is
consistent with the spatial distribution of TBABF4 electrolyte under a negative
applied potential, where both BF−4 anions and TBA
+ cations should remain on the
surface of Fe(TA)2 film. ToF-SIMS further showed that the boron signal remained
on the surface region after applying 0.2 V vs. Fc0/+ to Fe(TA)2 film (Fig. 5.4a(ii)),
illustrating that the voltammetric features in the potential range from -0.4 V to 0.2
V vs. Fc0/+ originated from the adsorption (anodic current) or desorption (cathodic
current) of BF−4 anions on the surface Fe sites. By contrast, the boron signal is still
observable even when sputtering into the Fe(TA)2 nanoparticle film after applying
the potential of 1.4 V vs. Fc0/+ (Fig. 5.4a(iii)). These results clearly confirm the
proposed mechanism depicted in Figure 5.1: at low voltage, surface redox occurs,
then the intercalation of BF−4 anions occurs at the 1.2 V voltammetric feature.
89
Figure 5.4. ToF-SIMS of spatial distribution of boron and tetrabutylammonium
(TBA+) in Fe(TA)2 film after applying different voltages. (a) ToF-SIMS depth
profile of boron within Fe(TA)2 film after applying (i) -1.0 V vs. Fc
0/+, (ii) 0.2 V
vs. Fc0/+, and (iii) 1.4 V vs. Fc0/+. (b) ToF-SIMS depth profile of TBA+ within
Fe(TA)2 film after applying (i) -1.0 V vs. Fc
0/+, (ii) 0.2 V vs. Fc0/+, and (iii) 1.4 V
vs. Fc0/+. Both boron and TBA+ signals are referenced to the Pt signal from the
QCM crystal surface. The gray area represents the surface region of Fe(TA)2 film
that is defined by the distribution of the TBA+ siganl. The spin-coating amount
of 16-nm Fe(TA)2 nanoparticles to those three EQCM electrodes was controlled to
be ca. 5.5 µg (i.e., ca. 60 nm in film thickness) in order to obtain depth-dependent
data.
Investigations into the charge variation across the Fe(TA)2 film in CV
measurements found that the mass increase in the anodic CV direction also
cooperated with the continuous accumulation of positive holes in the Fe(TA)2
film (Fig. D.4). These mass and charge increases of the Fe(TA)2 film suggest
that voltammetric features both from the surface and inside the nanoconfined
90
pore of Fe(TA)2 follow the ion-coupled charge transfer mechanism, where the
oxidation of Fe2+ to Fe3+ requires the adsorption/intercalation of BF−4 anions to
Fe centers to maintain the charge neutrality of the Fe(TA)2 film. Interestingly,
distinct from the mass change (Fig. 5.2a), the charge increased more uniformly
and did not exhibit a significant change at the redox feature of 1.2 V vs. Fc0/+
(Fig. D.3). We propose that holes can be initially stored at C2p and N2p orbitals
of triazolate linkers in Fe(TA) , (207;181)2 and the further intercalation of BF
−
4 anions
to the inner Fe sites induces the spatial redistribution of holes to Fe centers, leading
to Fe2+/Fe3+ redox reactions. The ratio of BF−4 /integrated charge was ca. 0.42
rather than 1 in the CV trace of Fe(TA)2 film; we note that the mass variation
in EQCM is a net value including the mass decrease by the desorption of solvent
and TBA+ cations from the surface and nanopores, and the mass increase due to
the adsorption/intercalation of BF−4 anions to the Fe(TA)2 film. This net effect
might underestimate the number of BF−4 anions that participate in Fe
2+/Fe3+ redox
events inside the Fe(TA)2 film.
To unravel the nature of ion intercalation, the apparent mass-to-charge ratio
can be determined as follows in Equation 1:
M ′
∆m
w = zF ( )∆Q
where z is the number of electrons and F is the Faraday’s constant. The obtained
M’w can be further used to calculate the solvation number (n) of BF
−
4 anions
during the intercalation/de-intercalation process via Equation 2:
M ′w −Mw(BF−n = 4 )
Mw(solvent)
where Mw(BF
−
4 ) is the molecular weight of BF
−
4 anions and Mw (solvent) is the
molecular weight of solvent (e.g., acetonitrile). For complete desolvation during
91
intercalation, a ∆m slope of 0.9 is expected, assuming that n = 0 and M
∆q w
=
M −w(BF4 ). As shown in Figure 5.5, the slope of the intercalation feature (ca 0.8)
matches well with this value, demonstrating that BF4 anions undergo complete
desolvation. By comparison, the slope of ∆m for BF−4 -adsorption-induced surface∆q
redox feature was found ca. 0.2, which is much smaller than the theoretical value
(i.e., 0.9). This illustrates that the adsorption of BF−4 anions must also induce
desorption of other species from the film surface, such as solvent molecules and
TBA+ cations. In the cathodic scan direction, the de-intercalation gives a slope
of 1.9, higher than the theoretical value of 0.9. From Equation 1 and 2, the
corresponding solvation number n is ca. 2, suggesting that BF−4 anions are partially
solvated by acetonitrile once they de-intercalate from the nanopores and move
to the surface. The surface events in the cathodic direction give a slope of 0.1,
indicating that other species (solvent and cations) may replace BF4 anions once
they desorb from the film surface.
Taken together, potential-dependent mass and charge variations and ToF-
SIMS analysis clearly demonstrat that the voltammetric features in the range of
-0.4 to 0.2 V vs Fc0/+ can be assigned to surface Fe2+/3+ reactions, coupled with
partial desolvation of the BF4 anions, whereas the feature at 1.2 V originates from
BF−4 intercalation induced Fe
2+/3+ redox chemistry in the nanoconfined pores of
Fe(TA)2 particles (Fig. 5.1). Given that the solvated BF
−
4 anions (diameter of
ca. 10 Å in acetonitrile) are larger than the Fe(TA)2 cavity (diameter of ca. 4.54
Å), a high potential of 1.2 V vs. Fc0/+ is required as a driving force to achieve the
complete desolvation of BF−4 anions into the nanoconfined pores.
92
Figure 5.5. Mass change vs. charge change in Fe(TA)2 film. The theoretical result
with the slope of ca. 0.9 is shown in the black dashed line. The experimental
data is shown as the blue dots, where the red line stands for the potential
region for BF−4 -intercalation/de-intercalation-induced Fe
2+/Fe3+ redox feature
at 1.2 V vs. Fc0/+ and the green line stands for the potential region for BF−4 -
adsorption/desorption-induced Fe2+/Fe3+ redox feature from -0.4 V to 0.2 V vs.
Fc0/+. The mass and charge changes of Fe(TA)2 film collected at a 10 mV/s scan
rate using 0.1 M TBABF4 / MeCN.
Factors Controlling Ion Intercalation Redox Chemistry in Fe(TA)2
Investigating how different experimental factors will impact redox
behaviors inside the pore and on the surface of Fe(TA)2 nanoparticles help
establish the fundamental picture of ion intercalation redox chemistry within the
nanoconfinement environment of MOF nanoparticles. The role of film thickness
in governing redox properties is usually not considered in MOF electrochemistry.
Film-thickness-dependent studies were conducted by first spin-coating ca. 4.0
µg of 16-nm Fe(TA)2 nanoparticles onto the quartz crystal electrode. Atomic
force microscope (AFM) showed that those 16-nm Fe(TA)2 nanoparticles self-
assembled into a film with the thickness of ca. 21 ± 5 nm (Fig. 5.6a, suggesting
the formation of single-particle-thickness Fe(TA)2 film. CV measurements of this
Fe(TA)2 film were collected in a narrow potential window to isolate surface and
93
Figure 5.6. AFM images of films of Fe(TA)2 particles on QCM electrodes to
determine film thickness. a) A film comprised of 16-nm particles with a thickness of
21 ± 5 nm. A film comprised of 16-nm particles with a thickness of 60 ± 6 nm. c)
A film comprised of 25-nm particles with a thickness of 53 ± 8 nm.
ion-intercalation redox features. It was found that ∆E of both surface and ion-
intercalation redox features did not vary in different CV traces (Fig. 5.7a(ii) and
(iii)). This demonstrates that both surface and intercalation redox reactions on
21-nm-thickness Fe(TA)2 film are kinetically reversible. Further increasing the spin-
coating amount of 16-nm Fe(TA)2 nanoparticles to 5.5 µg led to a Fe(TA)2 film
with a thickness of 60 ± 6 nm (i.e., ca. four layers of particles as shown in Fig.
5.6b). In the thicker film, surface-isolated events remain kinetically reversible (Fig.
5.7b(ii)), whereas the intercalation chemistry at 1.2 V vs. Fc0/+ gradually became
kinetically irreversible as evidenced by the increase of ∆E in subsequent CV scans
(Fig. 5.7b(iii) and 5.7b(iv)). Such ∆E increase together with the decrease of the
current density (Fig. 3b(iii)) of the 1.2 V redox feature illustrates that increasing
film thickness hinders the intercalation dynamics of BF−4 anions and suppresses the
ion intercalation redox chemistry inside the nanoconfined pore environment.
94
Figure 5.7. Film thickness, particle size, and redox reversibility. a) A thin 16-
nm film with thickness of 21 ± 5 nm shows reversible features at both a low
oxidizing potential (a(ii)) and a high oxidizing potential (a(iii, iv)). b) A thicker
film, 60 ± 6 nm, also comprised of 16-nm particles, shows good reversibility only
at lower potentials (b(ii)) and suffers at higher potential (b(iii, iv). c) A film made
with larger 25-nm particles, measuring 53 ± 8 nm in thickness, also shows good
reversibility in the lower range of potential (c(ii) and suffers from poor reversibility
at high potential c(iii, iv)
Further, a 53 ± 8 nm-thickness Fe(TA)2 film (i.e., ca. two particle layers)
was prepared by loading ca. 4.0 µ g of 25-nm Fe(TA)2 nanoparticles onto the
electrode (Fig. 5.6c). As shown in Fig. 5.7c(ii) and Fig. D.5, the use of larger
25-nm nanoparticles did not affect the thermodynamic (i.e., E1/2) and kinetic
properties (i.e., ∆E variation) of surface redox reaction. In contrast, a more
significant ∆E increase (Fig. 5.7c(iv)) and the current density decrease (Fig.
95
5.7c(iii)) were detected in CV scans as compared to those on the similar thickness
film made by 16-nm Fe(TA)2 nanoparticles (Fig. 5.7b(iii) and 4b(iv)). It is possible
that for BF−4 anions, the distance of inner-particle intercalation inside the film
prepared by 25-nm nanoparticles is longer than that inside the similar thickness
film formed by more assembled layers of 16-nm Fe(TA)2 nanoparticles, where more
inter-particle space exists to provide possible ion pathways. A longer inner-particle
intercalation pathway leads to the more sluggish intercalation dynamics of BF−4
anions and severely hinders the intercalation redox reaction inside the pore of 25-
nm Fe(TA)2 nanoparticles. Taken together, film-thickness-dependent studies show
that the intercalation of BF−4 anions to Fe(TA)2 has a more significant impact on
the intercalation redox feature at 1.2 V vs. Fc0/+ rather than the surface redox
reaction since BF−4 anions are still accessible to surface Fe sites regardless of the
film thickness. The stark difference in peak position, shown in Figure 5.8, further
indicates that thicker films and larger particles cause intercalation redox chemistry
to require more energy with each subsequent scan. Practically, these results suggest
that film thickness and particle size are both critical variables to consider in MOF-
based charge storage devices.
96
Figure 5.8. Tracking the changes in E1/2 and ∆E in three different Fe(TA)2 thin
films. (a) E 0/+1/2 of redox feature at 0 V vs. Fc for Fe(TA)2 film with different
film thicknesses. (b) E 0/+1/2 of redox feature at 1.2 V vs. Fc for Fe(TA)2 film with
different film thicknesses. Green dots represent E1/2 of 21-nm-thickness Fe(TA)2
film (4.0 µg of 16-nm nanoparticles). Green stars represent E1/2 of 60-nm-thickness
Fe(TA)2 film (5.5 µg of 16-nm nanoparticles). Blue dots stand for E1/2 of 50-nm-
thickness Fe(TA)2 film (4.0 µg of 25-nm nanoparticles.)
Ion-pairing strength has also been reported to affect the intercalation of
counter ions into the pore of MOF nanocrystals. (203) By keeping the BF−4 anions
constant, but changing the TBA+ cations to the smaller trimethylamine cations
(TMA+), it is expected that the increase of ion-pairing strength will suppress
the BF−4 intercalation and the corresponding redox reactions within the Fe(TA)2
pore. Indeed, as shown in Fig. 5.9a, E1/2 of the intercalation redox feature shifted
positivily by 40 mV from ca. 1.23 V vs. Fc0/+ in TBABF4 electrolyte to ca. 1.27
V vs. Fc0/+ in TMABF4 electrolyte. A positive shift of E1/2 was also detected for
surface redox features (e.g., from ca. 0.02 V vs. Fc0/+ in TBABF4 to ca. 0.04 V
vs. Fc0/+ in TMABF4). Further use of LiBF4 as the electrolyte greatly reduced
the current density of both intercalation and surface redox features (Fig. 5.9a),
97
suggesting that the strongest ion-pairing strength induced by the smallest Li+
cations severely prevented BF−4 anions from neither adsorbing on the surface nor
intercalating into the pore of Fe(TA)2 nanoparticles. These cation-dependent
electrochemical studies discover that the ion-pairing strength plays an essential
role in determining both surface and intercalation redox properties of Fe(TA)2 film.
Figure 5.9. Electrolyte-dependent redox properties in Fe(TA)2 nanoparticle film.
(a) Cation-dependent CV measurements of 16-nm Fe(TA)2 nanoparticle film. (b)
Anion-dependent CV measurements of 16-nm Fe(TA)2 nanoparticle film.
The impact of anion size relative to the pore size of Fe(TA)2 nanoparticles
on controlling the intercalation redox chemistry should also be considered.
Given the similar size between BF−4 anions (i.e., ca. 3 Å in diameter) and the
cavity of Fe(TA)2 nanoparticles (i.e., ca. 4.5 Å in diameter), BF
−
4 anions enable
to intercalate into the nanoconfined pore under the sufficient electrochemical
potential at ca. 1.2 V vs. Fc0/+. Instead, the use of ClO4- anions (ca. 4.8 Å in
diameter) greatly suppressed the ion-intercalation redox feature as compared to the
surface redox features (Fig. 5.9b), suggesting that the intercalation was blocked if
electrolyte ions are larger than the cavity of Fe(TA)2 nanoparticles. Indeed, this
98
ion-intercalation redox feature from the nanoconfined Fe(TA)2 pore completely
disappeared by using even larger PF−6 anions as electrolytes (i.e., ca. 5.0 Å in
diameter) (Fig. 5.9b).
Further titration of small amounts of TBAPF6 (i.e., in the mM scale)
into 0.1 M TBABF4 electrolyte was found to have a more pronounced impact on
hindering the BF−4 intercalation chemistry inside the Fe(TA)2 pore as compared to
the surface redox reaction as evidenced by the faster decrease of current density at
1.2 V vs. Fc0/+ (Fig. 5.10). In addition, the ∆E of BF−4 intercalation redox feature
also increased more significant than that of surface redox features, showing the
sluggish BF−4 intercalation dynamics in the presence of the larger PF
−
6 anions (Fig.
D.4). Similar redox behaviors of Fe(TA)2 film were also observed when titrating
small amounts of TBAClO4 into TBABF4 electrolyte (Fig. D.5). These titration
experiments suggest that larger PF−6 and ClO
−
4 anions block the intercalation of
BF−4 anions into the nanoconfined pore of Fe(TA)2 nanoparticles and therefore
suppress the intercalation redox chemistry. Indeed, EQCM detected a more severe
mass decrease at the intercalation redox feature at 1.2 V vs. Fc0/+ as compared
to the mass decrease at one of the surface redox features at 0 V vs. Fc0/+ (Fig.
5.10b). The blocking effect of larger PF−6 anions on the BF
−
4 intercalation further
prevents the charge (i.e., hole) transfer to oxidize interior Fe2+ centers to Fe3+
(Fig, 5.10b). Taken together, the titration experiments along with anion-size-
dependent and film-thickness-dependent investigations together illustrate that the
rate-determining step of Fe2+/Fe3+ redox reaction inside nanoconfined Fe(TA)2
pore is the BF−4 intercalation process.
99
Figure 5.10. Electrolyte titrations show redox properties in Fe(TA)2 nanoparticle
film. (a) CV measurements of 16-nm Fe(TA)2 nanoparticle film during titrating
TBAPF6 into 0.1 M TBABF4 acetonitrile electrolyte. (b) Mass (hollow circle) and
charge (solid circle) change of 16-nm Fe(TA)2 nanoparticle film during titrating
TBAPF6 into 0.1 M TBABF4 acetonitrile electrolyte. Note: CV scan is collected
at a 10 mV/s scan rate. The loading amount of 16-nm Fe(TA)2 nanoparticle is
controlled to be ca. 4.0 µg on QCM crystal.
Given that the desolvation step occurs in both BF−4 intercalation-induced
redox chemistry and BF−4 adsorption-induced surface reactions, it is necessary
to investigate the importance of solvent effect in determining redox properties of
Fe(TA)2 film. In addition to acetonitrile, the CV of Fe(TA)2 film was collected
in 1,2-difluorobenzene. Since 1,2-difluorobenzene is known as a chemically inert
and non-coordinating solvent, it is expected that a less positive voltage is already
able to drive the desolvation and intercalation of BF−4 anions into the Fe(TA)2
pore as compared to that in acetonitrile. Fig. D.6 indeed detected ca. 150-mV
cathodic shift of E of BF−1/2 4 intercalation redox reaction inside the Fe(TA)2 pore
when switching the solvent from acetonitrile to 1,2-difluorobenzene. Given that
the potential of ca. 1.2 V vs. Fc0/+ is already beyond the potential window of
many commonly used organic electrolytes, we further studied the solvent effect on
100
governing surface redox properties of Fe(TA)2 film. Similar to solvent-dependent
intercalation redox reaction, E1/2 of BF
−
4 adsorption-induced Fe
2+/Fe3+ redox
feature also shifted cathodically by switching acetonitrile to 1,2-difluorobenzene
(Fig. D.6). Interestingly, it was found that those surface redox features were
highly dependent on solvents as evidenced by different E1/2 and current densities
in distinct solvents (e.g., acetonitrile, dichloromethane, tetrahydrofuran, and 1,2-
difluorobenzene in Fig. D.7). We hypothesize that those solvent-dependent surface
redox features are due to the different re-organization structures of solvent and
electrolyte on the Fe(TA)2 film surface induced by the partial desolvation and
adsorption of BF−4 anions. Overall, our studies confirm that the solvent effect
must be considered as an essential factor in the microscopic picture of both surface
reactions and the ion intercalation redox chemistry inside nanoconfined MOF pores.
Conclusions
In summary, our studies unravel how the nanoconfinement effect governs
the ion-intercalation redox chemistry inside the pore of Fe(TA)2 nanoparticles. The
nanoconfinement effect inside the Fe(TA)2 pore requires a higher electrochemical
potential to drive the complete desolvation and intercalation of BF−4 anions into
the pore as compared to the partial desolvation and adsorption of BF−4 anions on
the Fe(TA)2 surface. The further manipulation of experimental factors (e.g., film
thickness, electrolyte species, and solvents) discovers that the redox properties
of ion intercalation into the nanoconfinement environment of Fe(TA)2 pore
behave significantly different from the surface redox. It is found that Fe2+/Fe3+
redox chemistry inside the Fe(TA) pore is determined by the BF−2 4 intercalation
step, while the surface Fe2+/Fe3+ redox is more impacted by solvents. These
fundamental studies establish a microscopic-level description of ion intercalation
101
redox chemistry inside the nanoconfined pore of conductive MOF nanoparticles,
providing the guideline for designing efficient energy storage under confinement in
porous materials.
Additional studies would be required to provide insights into the electrolyte
and solvent ordering on the Fe(TA)2 surface to further understand the solvent-
dependent surface redox features. Meanwhile, the nanoconfinement-induced
complete desolvation in the process of BF−4 intercalation process shortens the
distance (d) between BF−4 and the interior Fe(TA)2 surface inside the pore,
suggesting Fe(TA)2 nanoparticles as a promising material for pseudo-capacitance
applications.
102
CHAPTER VI
CONCLUDING REMARKS
This dissertation presents the Seesaw model of Metal-Organic Framework
(MOF) particle growth, verifies the model by applying it to known MOF systems,
then implements it to develop the new nanoMOFs Co(TA)2, Cd(TA)2, and
Fe(TA)2. The research in this thesis shows that controlling particle size of 1,2,3-
triazole MOFs can be done using the modulator method, and that the resulting
particles are stable as colloids and can be processed into thin films. For these three
MOFs, the modulator 1-methylimidazole can be added in increasing quantities in
order to decrease particle size. Additionally, the development of this new class of
MOF nanoparticles shows that the Seesaw model can guide synthetic chemists
as we target new nanoMOF materials and reduce the amount of high-throughput
testing needed.
This work demonstrates that solution-stable MOF nanoparticles can
be analyzed via methods typically exclusive to smaller clusters and inorganic
nanoparticles. Controlling particle concentration for UV-Vis absorption
spectroscopy allowed for the first determination of a MOF material’s extinction
coefficient. It was found that the extinction coefficients of the charge transfer bands
of Fe(TA)2, per particle, scale with particle volume. In the future, these types of
analyses may allow for a) the rapid determination of particle size and concentration
through UV-Vis and b) the quantitative comparison of MOFs’ optical properties.
Unexpectedly, the charge transfer bands of Fe(TA)2 also show size-dependence,
blueshifting as particle size decreases. This is the first report of size-dependent
optical properties in the absorbance spectra of MOF particles. This dependence is
hypothesized to be due to the variations in coordination environment of Fe centers
103
in the particles as the surface area to volume ratio changes. Although the cause is
not yet certain, this study opens the door to utilizing particle size as another useful
handle to tune the optical properties of MOF materials. As crystal engineering
emerges in the forefront of the MOF field, and the materials are increasingly sought
out for selective sensing and other applications, quantitative UV-Vis will likely
emerge as a standard characterization technique.
Synthesizing porous semiconductors with control over size will be a crucial
step in developing conductive and redox-active MOF materials for technological
applications. While there are no commercial uses of conductive MOFs in devices
to-date, several studies show promise for their use as electrode and transducer
materials. (208;209) By making conductive MOFs as solution-processable particles
with low size dispersity, greater control over the materials’ processing can be
attained. The material Fe(TA)2 shows size-dependent electronic behavior not
previously observed: the redox behavior of the nanoparticles shows superior
reversibility to its bulk counterpart. Additionally, an extremely sharp and reversible
feature is observed which was unambiguously assigned to the complete desolvation
and intercalation of BF−4 anions into the MOF particles coupled with Fe
2+/3+ redox
events. The reversibility of the intercalation chemistry suffers when film thickness
is too high, particles are too large, or electrolyte anions are too large. The ion and
solvent dependence of MOF redox chemistry is essential to their future development
in devices. Further, it was found that the bulk conductivity of Fe(TA)2 thin films
increases as particle size decreases from bulk to 25 nm. This intriguing development
suggests that nanoparticle thin films may be a more practical option for device
applications over single crystals, as grain boundary sites may not impede bulk
conductivity significantly. The advantage of solution processibility, in combination
104
with improving the efficiency of redox processes, will place conducting nanoMOF
materials at the forefront of charge storage research and device implementaiton.
105
APPENDIX A
SUPPLEMENTARY INFORMATION FOR CHAPTER 2
Table A.2. Typical M:L ratios and L concentrations for MOF nanocrystals
discussed in this perspective. These values are compared to 1-2 representative
examples for bulk syntheses. Bold values indicate nanoscale syntheses where either
excess linker or a more dilute system was used compared to bulk syntheses.
Bulk L : Bulk L Nano L : Nano L Nano Bulk
MOF Name
1M Conc. (M) 1M Conc. (M) Ref Ref
[(2-PTZ)2 0.125 – 6, 0.66 – 1 2, 3 0.2, 0.3
(221) (222)
Cd(H2O)2]
0.5 – 4 (M = Zn)
(M = Zn)
COMOC-4 Not available 1.14 0.05 (223)
DUT-23 (Cu) 0.4 0.0076 2, 4 0.00052, (224) (225)
0.0010
Dy-BTC 1 0.01 1.5 0.0125 (64) (226)
Fe-soc-MOF 2 0.49 0.013 (227;228)
HKUST-1 0.67 0.020 0.53 0.04 (64) (229)
0.53 0.012 0.67 0.042 (230) (231)
0.67 0.04 (81)
0.56 variable (232)
0.56 0.099 (233)
1 0.041 (67)
IR-MOF-3 0.34 0.04 0.34 0.043 (68) (234)
0.37 0.02 (67)
MFU-4 0.25 0.0625 0.24 0.03 (69) (235)
MFU-4l 0.048 0.0036 0.05 0.0019 (33) (236)
MIL-100 (Al) 0.67 0.11 0.66 0.241 (75) (237)
MIL-96 (Al) 0.81 0.14 0.081 0.01 (74) (238)
106
MIL-100 (Cr) 0.67 0.14 0.67 0.1333 (75) (239)
(M = Cr(0))
MIL-100 (Fe) 0.66 0.13 0.67 0.1333 (75) (240)
(M = Fe(0))
0.67 0.201 (76)
MIL-101 (Cr) 1 0.21 1 0.2 (78) (241)
1 0.2 (79)
1 0.033 (77)
1 0.2 (242)
MIL-101-NH (Fe) 0.75 0.017 0.51 0.083 (81) (68)2
MIL-88A 1 0.2 1 0.2 (214) (243)
MIL-88B (Cr) 1 0.08 1 0.2 (80) (244)
MIL-88B (Fe) 1 0.046 0.65 0.1 (72) (245)
MIL-88B-NH2 (Fe) 1 0.046 0.5 0.022 (73) (245)
MIL-125 0.67 0.3 1.3 Not available (82)
0.55 1.44 × 10−7 (216) (246)
MIL-125-NH2 2 0.12 1.5 0.277
(83) (246)
1.55 0.0775 (139)
MIL-53 (Al) 0.5 0.347 0.5 0.347 (247) (248)
MOF-5 0.48 0.024 0.4 – 1 0.053 – 0.13 (64) (249)
0.337 0.0367 0.2 0.0067 (81) (250)
0.33 0.01 (251)
0.5 0.0005 (224)
1 0.00166 (84)
3 0.005
0.43 0.000125 - (86)
0.0028
0.33 0.04 (252)
0.5 0.05 (253)
0.33 0.04 (45)
107
MOF-74 (Co) 0.25 0.01 0.28 Not available (81) (81)
MOF-74 (Ni) Not available 0.29 0.012 (81)
0.5 0.0417 (88)
MOF-801 Not available 2 0.18 (213)
NU-1000 0.2 0.0074 0.088 0.0014 (91) (254)
0.061 0.0056 (90)
0.1 0.0018 (89)
0.0054 0.0036 (255)
PCN-222 0.12 0.0038 0.068 0.00053 (91) (256)
0.1 0.0018 (89)
PCN-224 0.14 0.0065 0.14 0.0013 (92)
0.17 0.002 (57)
UiO-66 1 0.0257 (257) (258)
1 1.5 0.0386
2 0.0515
0.0086 3.34 0.075 (155)
1 0.0043 (259)
1 varies (260)
1 0.0172 (218)
1 0.0454 (261)
1 0.0454 (139)
0.74 0.02 (255)
UiO-66 w/ Co 1 0.078 (262)
UiO-66-NH2 1 (X = NH2) 0.0356
UiO-67 Not available 0.74 0.02 (255)
UMCM-150 0.49 0.0070 0.33 0.00051 (224) (263)
ZIF 8 w/ Co Not available 0.34 0.1 (264)
16 0.000099 (265)
ZIF-65-Zn 0.5 0.067 2 0.1 (266) (267)
ZIF-7 0.81 0.027 2 0.05 (266) (268)
108
2 0.015 20
ZIF-71 4 0.053 2 0.2 (266) (269)
ZIF-8 0.91 0.041 1.9-2 0.05 (264)
8 0.20, 0.089 (270) (271)
Zn-BPD-X (X = 1 0.05 1 0.013 (97) (272)
NH2, NO2, OH)
Zn-BPD-OH 1 0.027 (97)
109
Table A.1. Values for the smallest, median, and average nanocrystal sizes (all in
nm) reported in Figure 1 of the main text. The smallest MOF nanocrystals made
by other methods are reported along with the method used: metal organic gel
(Gel,) ionic liquid microemulsion (ILM), dual injection (Inject), and slow addition
(SA)
MOF Smallest Median Average Smallest by Other Methods Smallest Ref
DyBTC 50 60 59.2 (64)
HKUST-1 2.6 60 164.71 2.6 – Gel (173;210;174;211)
1.6 – ILM
24 – Inject
IR-MOF-3 30 (68)
MFU-4 33.66 1200 750.2 (69)
MOF-74 17 (Co) 16.6 (Zn, SA)
13.6 (Mn, SA),
9.4 (Mg)
5.1 (Co, SA)
2.8 (Ni, SA) (172;212)
MOF-801 23 95.5 95.5 (213)
MIL-88-A 60 195 263.64 (214)
MIL-88B-NH2 30 105 100
(73)
MIL-96 930 2100 2795.7 (74)
MIL-100-Al 272 423.5 423.5 (215)
MIL-100-Cr 109.33 119.79 119.79 (75)
MIL-100-Fe 100 323 298.81 (71)
MIL-101-Cr 19 127.5 211.8 (77)
MIL-101-Fe 47 – Inject (211)
MIL-125 85 550 583.8 (216)
MIL-125-NH2 70 220 305.6
(83)
MOF-5 25 137.5 709.8
NU-1000 75 400 1965.5 (89)
NU-1003 300 (217)
PCN-222 190 550 535.8 (91)
PCN-224 49 1141 133.1 (92)
UiO-66 14 117 221.2 35 – Inject (211;218)
UiO-66-NH 16 82 82 (218)2
UiO-67 308 515.5 481 (218)
ZIF-7 30.7 71.4 71.4 (95)
ZIF-8 9 42.3 141.3 32 – Inject (174;211;219)
2.2 - ILM
ZIF-67 80 - Inject (211;174)
2.3 - ILM
ZIF-71 13 37 41.6 (220)
ZIF-90 65 – Inject (211)
Zn-BPD-H 75.5 100 105.1 (97)
Zn-BPD-NH 95 137.5 144.4 (97)2
Zn-BPD-NO2 57.5 79.5 74.6
(97)
Zn-BPD-OH 97 102.5 105.1 (97)
110
111
Table A.3. Nanocrystal sizes used to create Fig 1 in the main text. Size distributions given in nanometers are shown
in parentheses. Dispersity measurements reported as PDI (Polydispersity Index), standard deviation, relative standard
deviation, or a range of nanometers are reported in Disp. (Other) with the type of data in parentheses (PDI, SD, RSD,
or Range). In cases where anisotropic particles were presented with both length and width, they are differentiated here
as (L) and (W). N/A stands for Not Available.
MOF Size (nm) Disp. (Other) Method Size Method Size Method Method Average Citation
[(2-PTZ)2Cd(H2O)2] Not available
(221)
BUT-12 Not available (273)
COMOC-4 (w/ Eu) N/A (223)
DUT-23 Cu N/A (224)
Dy-BTC 70(15) SEM
50(30) SEM
50(25) SEM
65(25) SEM
60(20) SEM
60(20) SEM
Fe-soc-MOF 310(10) SEM
HKUST-1 100 DLS (67)
47 DLS
48 DLS
60 DLS
54 DLS
20000 SEM (274)
300 SEM
85 SEM
100 SEM
115 SEM
600 SEM
2500 SEM
2500 SEM
2500 SEM
Not available SEM (64)
10 SEM (81)
21 19.90% (RSD) TEM (232)
112
32 30% (RSD) TEM
46 22.20% (RSD) TEM
299 15.60% (RSD) TEM
76 10.80% (RSD) TEM
190 19.20% (RSD) TEM
448 16.70% (RSD) TEM
658 25.40% (RSD) TEM
331 20.50% (RSD) TEM
449 16.10% (RSD) TEM
563 17.40% (RSD) TEM
1190 12.40% (RSD) TEM
58.3 PXRD (233)
51.9 PXRD
41.7 PXRD
51.8 PXRD
70 PXRD
80.9 PXRD
60 PXRD
78.6 PXRD
68.5 PXRD
42.4 PXRD
42.1 PXRD
57.4 PXRD
57 PXRD
N/A TEM 2.2 PXRD 3 PXRD 2.6 (173)
1.5 TEM 3.3 PXRD 4.7 PXRD 3.17
2 TEM 4.3 PXRD 6.1 PXRD 4.13
2.5 TEM 6.2 PXRD 8.7 PXRD 5.8
IR-MOF-3 30 DLS (68)
32 DLS
33 DLS
47.5 DLS
46 DLS
N/A (252)
Hf-BUT-12 N/A (273)
Hf-UiO-66-NH2 N/A
(273)
113
Hf-UiO-66-2,6-NDC N/A (273)
MFU-4 >1200 DLS (69)
>1200 DLS
>1200 DLS
>1200 DLS
>1200 DLS
>1200 DLS
>1200 DLS
119 DLS
46(2) DLS 37(3) PXRD 32(11) TEM 38.33
42(1) DLS 34(2) PXRD 25(6) TEM 33.67
48(1) DLS 36(2) PXRD 36(8) TEM 40
43(3) DLS 35(1) PXRD 29(8) TEM 35.67
MIL-53 350(100) (71)
MIL-53 Al N/A (247)
MIL-88B-Cr N/A (80)
MIL-88B-Fe N/A (72)
MIL-88B-Fe-NH 30 TEM (73)2
55 TEM
155 TEM
160 TEM
MIL-88A 100(25) DLS (71)
N/A (275)
255(25) (214)
390(30)
360(35)
550(55)
110(25) DLS
195(15) DLS
230(25) DLS
150(30) DLS
270(20) DLS
290(20) DLS
210(25) DLS
380(35) DLS
360(50) DLS
114
275(65) DLS
885(90) DLS
1050(210) DLS
720(75) DLS
¿1200 DLS
380(35) DLS
480(45) DLS
¿1200 DLS
180(15) DLS
240(15) DLS
390(25) DLS
95(15) DLS
75(10) DLS
190(20) DLS
60(10) DLS
70(10) DLS
180(15) DLS
115(15) DLS
70(10) DLS
75(10) DLS
60(5) DLS
80(5) DLS
85(8) DLS
110(10) DLS
160(10) DLS
270(23) DLS
350(55) DLS
87(15) DLS
110(10) DLS
280(45) DLS
575(85) DLS
110(10) DLS
280(45) DLS
575(85) DLS
110(10) DLS
185(25) DLS
115
275(25) DLS
MIL-89 75(25) DLS (71)
MIL-96 Al 8500 Length 1000 Width (74)
2900 SEM (L) 450 SEM (W)
1600 SEM (L) 410 SEM (W)
980 SEM (L) 510 SEM (W)
2100 SEM (L) 570 SEM (W)
2200 SEM (L) 810 SEM (W)
1290 SEM (L) 610 SEM (W)
MIL-100 Al 585(150) DLS (H2O) 565(75) DLS (EtOH) 575
(75)
249(28) DLS (H2O) 237(41) DLS (EtOH) 311(41) DLS (DMEM) 272
(215)
291(24) DLS (MEM)
MIL-100 Fe 100(50) DLS (71)
452 DLS (Tris) 256 DLS (H2O) 354
(76)
646 DLS (Tris) 323
458 DLS (Tris) 215 DLS (H2O) 336.5
698 DLS (Tris) 698
596 DLS (Tris) 596
592 DLS (Tris) 238 DLS (H2O) 415
139(25) DLS (H O) 168(10) DLS (EtOH) 153.5 (75)2
110(25) DLS (H2O) 119(11) DLS (EtOH) 114.5
141(43) DLS (H2O) 155(61) DLS (PBS) 162(60) DLS (PBS + Alb) 150.75
(276)
145(38) DLS (RMPI)
139(25) DLS(H2O) 168(10) DLS (EtOH) 252(32) DLS (DMEM) 203.5
(215)
255(21) DLS (MEM)
MIL-100 Cr 143(63) DLS (H2O) 80(41) DLS (EtOH) 106(47) DLS (MeOH) 109.33 (75)
142(63) DLS (H2O) 80(41) DLS (EtOH) 146(32) DLS (DMEM) 130.25
(215)
153(49) DLS (MEM)
MIL-101 Cr 50(9) TEM (77)
19(4) TEM
25(6) TEM
28(6) TEM
36(7) TEM
73(8) TEM
387(28) TEM (78)
383(25) TEM
116
346(19) TEM
279(21) TEM
160(16) TEM
148(14) TEM
90(10) TEM
114(13) SEM (79)
1336(174) SEM
141(16) SEM
87(9) SEM
101(11) SEM
214(25) SEM
387(28) (80)
156(17)
100(12)
MIL-101-Fe-NH2 Not available
(81)
MIL-125 Not available (82)
200(50) (216)
MIL-125-NH2 240(70) DLS (DMF) 313.33
(83)
660(250) DLS (H2O)
320(100) DLS (MeOH)
220(60) DLS (EtOH)
230(60) DLS (PBS)
210(100) DLS (PBS FBS)
750 SEM (139)
280 SEM
192 SEM
100 SEM
70 SEM
150 SEM
190 SEM
750 SEM
420 SEM
320 SEM
220 SEM
100 SEM
90 SEM
117
180 SEM
350 SEM
720 SEM
MOF-5 N/A (64)
N/A (81)
70(10) TEM (84)
624(39) (86)
399(23)
220 140-420 (252)
125(25) (253)
37.5(7.5) (45)
150
MOF-74 N/A (81)
17(3) (Co) (212)
100(20) (Mn)
200(50) (Mg)
18(5) (Ni)
MOF-801 168(24) SEM (213)
23(11) SEM
NU-1000 180 140-220 (Range) (91)
N/A (90)
75 (89)
150
500
1200
15000
150(50) (255)
300(50)
600(100)
1500(500)
NU-1003 300 SEM (217)
1000 SEM
2000 SEM
7000 SEM
10000 SEM
PCN-222 190 150-250 (Range) (91)
118
350(50) (74)
475(25)
625(125)
650(50)
925(75)
Not available (273)
PCN-224 33(4) TEM 65(3) DLS 49 (92)
91(5) TEM 114(3) DLS 86.5
91(8) TEM 137(2) DLS 114
144(7) TEM 197(1) DLS 170.5
189(11) TEM 255(2) DLS 222
100 50-120 (Range) (57)
190 150-150 (Range) (91)
UiO-66 106.5(11.9) SEM 547.1(25.1) DLS 326.8 (257)
118.2(18.1) SEM 1327(57.6) DLS 722.6
90.2(10.9) SEM 472.7(182.3) DLS 281.45
66.9(10.2) SEM 288.1(31.4) DLS 177.5
71.4(10.2) SEM 167.7 DLS 119.55
82.5(12.2) SEM 526.2(24.1) DLS 304.35
87.9(15.3) SEM 872.5(9.2) DLS 480.2
66.4(12.3) SEM 374.2(33.9) DLS 220.3
23.7(5.0) SEM 244.7(4.1) DLS 134.2
23.1(3.5) SEM 139.9(8.7) DLS 81.5
91.9(21.0) SEM 443.3(54.7) DLS 267.6
53(9.7) SEM 709.2(40.7) DLS 381.1
29.2(6.8) SEM 388.1(99) DLS 208.65
26.5(5.9) SEM 209.2(33) DLS 117.85
20.0(4.0) SEM 132.2(6.2) DLS 76.2
26.5(5.5) SEM 294.8 DLS 160.65
82.5(12.2) SEM 526.2(24.1) DLS 304.35
23.3(3.1) SEM 575 DLS 299.15
17(2) STEM Not available DLS 17 (155)
34(22) STEM Not available DLS 34
72(18) STEM 158(19) DLS 115
208(56) STEM 173(40) DLS 190.5
270(48) STEM 319(67) DLS 294.5
119
514(88) STEM 677(204) DLS 595.5
721(56) STEM 826(135) DLS 773.5
17(4) STEM 83(38) DLS 50
31(5) STEM 89(25) DLS 60
44(12) STEM 94(19) DLS 69
147(50) STEM 175(31) DLS 161
439(83) STEM 559(110) DLS 499
813(223) STEM 919(144) DLS 866
1174(258) STEM 1208(275) DLS 1191
19(5) STEM 132(72) DLS 75.5
29(8) STEM 74(18) DLS 51.5
88(25) STEM 95(24) DLS 91.5
550(51) STEM 784(70) DLS 667
38(7) STEM 100(29) DLS 69
53(12) STEM 140(30) DLS 96.5
227(46) STEM 277(80) DLS 252
1949 154 (SD) SEM (259)
1060 65 (SD) SEM
783 32 (SD) SEM
583 19 (SD) SEM
522 24 (SD) SEM
502 26 (SD) SEM
220 122-615 (Range) DLS (260)
255 105-396 (Range) DLS
164 142-459 (Range) DLS
164 105-396 (Range) DLS
105 79-255 (Range) DLS
164 122-342 (Range) DLS
190 164-458 (Range) DLS
85 XRD
83 XRD
81 XRD
78 XRD
74 XRD
71 XRD
46 XRD
120
23 XRD
14 XRD
265 SEM (261)
96 SEM
48 SEM
155 SEM
75 SEM
40 SEM
190 SEM
153 SEM
144 SEM
117 SEM
224 SEM
150 SEM
167 SEM
177 SEM
150 SEM
100 SEM
97 SEM
102 SEM
265 SEM
91 SEM
78 SEM
61 SEM
41 SEM
230 SEM (262)
N/A (273)
UiO-66-NH 115 PXRD (218)2
90 PXRD
85 PXRD
82 PXRD
80 PXRD
54 PXRD
21 PXRD
16 PXRD
200 TEM (273)
121
UiO-66-OH N/A (273)
UiO-66-2,6-NDC N/A (273)
UiO-67 641 365-995 (Range) DLS (218)
553 356-859 (Range) DLS
553 356-859 (Range) DLS
413 308-641 (Range) DLS
308 229-478 (Range) DLS
478 356-742 (Range) DLS
Not availble (273)
ZIF-7 112 DLS (266)
30.7(5.9) TEM (95)
40(10) SEM (277)
70(10) SEM
ZIF-8 18 PXRD 17 SLS 17.5 (219)
45 PXRD 39 SLS 42
10 PXRD 9 SLS 9.5
24 PXRD 20 SLS 22
55 PXRD 40 SLS 47.5
9 PXRD 9 SLS 9
16 PXRD 16 SLS 16
43 PXRD 42 SLS 42.5
10 PXRD 8 SLS 9
50(20) (278)
44(17)
34 15-65 (Range)
60(25)
21(4) TEM (265)
744 TEM (264)
601 TEM
490 TEM
271 TEM
93 TEM
43 TEM
27 TEM
40.6 0.09 (PDI) SLS (270)
45.4 SLS
122
28 SLS
31.2 SLS
25 SLS
25 SLS
ZIF-65-Zn 125 DLS (266)
125 DLS
ZIF-71 55 DLS (266)
83 DLS
83 DLS
98 PXRD 100 SEM 99 (279)
11 PXRD 15 SEM 13
94 PXRD 90 SEM 92
44 PXRD 30 SEM 37
24 PXRD 25 SEM 24.5
Zn-BPD-H 105(15) DLS 46 PXRD 75.5 (97)
150(20) DLS 47 PXRD 98.5
145(15) DLS 50 PXRD 97.5
155(30) DLS 51 PXRD 103
140(20) DLS 60 PXRD 100
170(25) DLS 50 PXRD 110
180(25) DLS 52 PXRD 151
Zn-BDP-NH2 190(35) DLS Not available 95
(97)
210(30) DLS 58 PXRD 134
193(40) DLS 71 PXRD 132
220(35) DLS 67 PXRD 143.5
210(35) DLS 65 PXRD 137.5
210(35) DLS 71 PXRD 140.5
275(35) DLS 61 PXRD 120(40) (L), 45(20) (W) SEM 228
Zn-BPD-NO2 125(35) DLS 45 85
(97)
100(30) DLS 43 PXRD 71.5
120(25) DLS 42 PXRD 81
120(30) DLS 39 PXRD 79.5
115(30) DLS Not available 57.5
100(30) DLS 37 PXRD 68.5
120(35) DLS 39 PXRD 65(30) SEM 79.5
Zn-BPD-OH 145(35) DLS 52 PXRD 98.5 (97)
123
145(35) DLS 50 PXRD 97.5
160(30) DLS 45 PXRD 102.5
165(35) DLS 45 PXRD 105
165(35) DLS 40 PXRD 102.5
160(35) DLS 34 PXRD 97
155(40) DLS 50 PXRD 60(20) 132.5
APPENDIX B
SUPPLEMENTARY INFORMATION FOR CHAPTER 3
Synthetic Details and Analytical Methods
Materials & Methods. Zinc nitrate hexahydrate was purchased from
Sigma-Aldrich, 2-methylimidazole was purchased from Aldrich, and ACS grade
methanol was purchased from Fisher Scientific. Copper (II) nitrate trihydrate was
purchased from Acros organics, 1,3,5-benzenetricarbosylic acid from TCI America,
and 200 proof dry ethanol from Fisher Scientific. Sodium benzoate was purchased
from MCB and benzoic acid was purchased from JT Baker. All chemicals were used
without further purification.
Cu3BTC2 (HKUST-1) Expanded Synthetic Methods. Initial
syntheses to explore the use of sodium benzoate as a modulator for Cu3BTC2
were performed on a larger scale. Both ligands, 1,3,5-benzenetricarboxylic acid
(0.1 mmol) and sodium benzoate (0.1-2.0 mmol) were dissolved in 10 mL 1:1
H2O:EtOH. A solution of copper nitrate (0.005 M, 4 mL) was poured in, resulting
in the immediate formation of a blue precipitate. If only the linker and metal are
combined in these conditions, no product forms. If benzoic acid is added without
any sodium benzoate, no product forms. For scale-up syntheses, four key conditions
were selected (refer to table B.14). The stock solutions of 1,3,5-benzenetricarboxylic
acid, benzoic acid, and sodium benzoate in 1:1 H2O:EtOH were combined in 100
mL Schott bottles. The solutions were diluted to 90 mL. Copper nitrate stock
solution was added under stirring. Schott bottles were removed from the stir
plate as soon as the solution was homogenously blue in color—less than 1 second.
Samples were left to sit 25 hours, then washed twice with 1:1 H2O:EtOH.
124
Scherrer Size Estimations. ZIF-8 and Cu3BTC2 particles were
characterized using PXRD and approximate sizes were determined using Scherrer
equation:
Kλ
τ = (B.1)
βcosθ
In this equation, β is the full width at half max of a peak, in radians, θ is
the scattering angle, in radians, and λ is the wavelength in nm. The shape factor K
was taken to be 1, a typical value for spherical particles.
Figure B.1. An example of peak fitting in Igor 6.32. The sample being fit is
denoted in table B.9.
Diffraction peaks were fit in Igor Pro 6.37 with Multipeak fit 2.0. A gaussian
function with a constant baseline was used in cases where the peaks were thin
and well-defined. The baseline function was changed to a log cubic function in
cases where the constant background function did not produce representative fits.
Particle size was estimated using compiled Full Width at Half Max (FWHM) and
peak location data. For ZIF-8, the peaks [100], [111], and [102] were used and the
size values were averaged. For Cu3BTC2, the peaks [200], [222], and [400] were used
125
and the size values were averaged. If not all these peaks were present or could be
adequately fit, they were not included.
Tables Detailing Synthetic Conditions and Particle Size. The
tables below detail the synthetic conditions and the resulting particle sizes. The
amount of each component is given by their final concentration in mol L−1, where
M is the metal, L is the linker, and Mod is the total modulator. Refer to sections
1.2 and 1.3 for the general synthetic methods. Representative PXRD patterns for
some of these sets can be found in section 1.7.
Table B.1. Cu3BTC2 synthesis using varying benzoic acid to sodium benzoate
ratios with excess linker and a metal concentration of 0.001 M. M, L, and Mod
refer to the concentrations of the metal salt, linker, and modulator in M. Sizes are
calculated from the Scherrer equation. These data are depicted in Figure 3.4 of the
main text and Figure B.8.
Abs. Eq. Mod
M L Mod Size (nm) Morphology
M : L : Mod BA : B-
0.001 0.003 0.021 1 : 3 : 21 11 : 1 115.3 Indistinct
0.001 0.003 0.021 1 : 3 : 21 5 : 1 112.1 Indistinct
0.001 0.003 0.021 1 : 3 : 21 3 : 1 125.7 Octahedral
0.001 0.003 0.021 1 : 3 : 21 2 : 1 94.3 Octahedral
0.001 0.003 0.021 1 : 3 : 21 1 : 1 74.0 Globular
0.001 0.003 0.021 1 : 3 : 21 1 : 2 35.5 Globular
0.001 0.003 0.021 1 : 3 : 21 1 : 3 38.6 Globular
0.001 0.003 0.021 1 : 3 : 21 1 : 5 48.9 Globular
0.001 0.003 0.021 1 : 3 : 21 1 : 11 61.0 Aggregates
0.001 0.003 0.021 1 : 3 : 21 All B- 58.0 Spherical
126
Table B.2. Cu3BTC2 synthesis using varying benzoic acid to sodium benzoate
ratios with excess linker and a metal concentration of 0.025 M. M, L, and Mod
refer to the concentrations of the metal salt, linker, and modulator in M.These data
are depicted in Figure 3.4 of the main text.
Abs. Eq. Mod
M L Mod Scherrer Size (nm)
M : L : Mod BA : B-
0.0025 0.0075 0.0525 1 : 3 : 21 11 : 1 113.8
0.0025 0.0075 0.0525 1 : 3 : 21 5 : 1 83.8
0.0025 0.0075 0.0525 1 : 3 : 21 3 : 1 55.3
0.0025 0.0075 0.0525 1 : 3 : 21 1 : 1 22.1
0.0025 0.0075 0.0525 1 : 3 : 21 1 : 3 22.8
0.0025 0.0075 0.0525 1 : 3 : 21 1 : 5 19.5
0.0025 0.0075 0.0525 1 : 3 : 21 1 : 11 19.9
0.0025 0.0075 0.0525 1 : 3 : 21 100% B- 29.8
Table B.3. Cu3BTC2 synthesis using varying benzoic acid to sodium benzoate
ratios with excess linker and a metal concentration of 0.005 M. M, L, and Mod
refer to the concentrations of the metal salt, linker, and modulator in M. These
data are depicted in Figure 3.4 of the main text.
Abs. Eq. Mod
M L Mod Size (nm)
M : L : Mod BA : B-
0.005 0.015 0.105 1 : 3 : 21 11 : 1 117.7
0.005 0.015 0.105 1 : 3 : 21 5 : 1 47.3
0.005 0.015 0.105 1 : 3 : 21 3 : 1 35.8
0.005 0.015 0.105 1 : 3 : 21 1 : 1 21.2
0.005 0.015 0.105 1 : 3 : 21 1 : 3 35.8
0.005 0.015 0.105 1 : 3 : 21 1 : 5 17.0
0.005 0.015 0.105 1 : 3 : 21 1 : 11 26.4
0.005 0.015 0.105 1 : 3 : 21 100% B- 24.1
127
Table B.4. Cu3BTC2 synthesis using varying benzoic acid to sodium benzoate
ratios with a stoichiometric amount of linker and a metal concentration of 0.027 M.
M, L, and Mod refer to the concentrations of the metal salt, linker, and modulator
in M.
Abs. Eq.
M L Mod Mod BA : B- Size (nm)
M : L : Mod
0.0027 0.0018 0.0126 3 : 2 : 14 11 : 1 99.7
0.0027 0.0018 0.0126 3 : 2 : 14 5 : 1 90.2
0.0027 0.0018 0.0126 3 : 2 : 14 3 : 1 83.5
0.0027 0.0018 0.0126 3 : 2 : 14 1 : 1 70.1
0.0027 0.0018 0.0126 3 : 2 : 14 1 : 3 52.2
0.0027 0.0018 0.0126 3 : 2 : 14 1 : 5 56.2
0.0027 0.0018 0.0126 3 : 2 : 14 1 : 11 42.8
0.0027 0.0018 0.0126 3 : 2 : 14 100% B- 54.1
Table B.5. Cu3BTC2 synthesis using varying equivalents of a 33% benzoic acid
modulator, with excess linker and a metal concentration of 0.001 M. M, L, and
Mod refer to the concentrations of the metal salt, linker, and modulator in M.
These data are depicted in Figure 3.3 of the main text. SEM images of these
samples are shown in figure 3.?, and the Scherrer crystallite sizes and apparent
particle size from SEM images are compared in figure B.10.
Abs. Eq.
M L Mod Mod BA : B- Size (nm) Morphology
M : L : Mod
0.001 0.003 0.003 1 : 3 : 3 1 : 2 112.4 octahedral
0.001 0.003 0.021 1 : 3 : 21 1 : 2 35.5 globular
0.001 0.003 0.030 1 : 3 : 30 1 : 2 27.0 globular
0.001 0.003 0.042 1 : 3 : 42 1 : 2 26.2 globular
0.001 0.003 0.060 1 : 3 : 60 1 : 2 52.6 globular
0.001 0.003 0.063 1 : 3 : 63 1 : 2 23.7 globular
0.001 0.003 0.084 1 : 3 : 84 1 : 2 25.1 globular
0.001 0.003 0.105 1 : 3 : 105 1 : 2 33.4 globular
0.001 0.003 0.120 1 : 3 : 120 1 : 2 43.3 globular
128
Table B.6. Cu3BTC2 synthesis using varying equivalents of a 50% benzoic acid
modulator, with excess linker and a metal concentration of 0.001 M.M, L, and Mod
refer to the concentrations of the metal salt, linker, and modulator in M. These
data are depicted in Figure 3.3 of the main text.
Abs. Eq.
M L Mod Mod BA : B- Size (nm) Morphology
M : L : Mod
0.001 0.003 0.006 1 : 3 : 3 1 : 1 104.6
0.001 0.003 0.018 1 : 3 : 18 1 : 1 61.6 globular
0.001 0.003 0.030 1 : 3 : 30 1 : 1 36.0
0.001 0.003 0.042 1 : 3 : 42 1 : 1 33.5
0.001 0.003 0.060 1 : 3 : 60 1 : 1 30.7
0.001 0.003 0.084 1 : 3 : 63 1 : 1 24.0
0.001 0.003 0.105 1 : 3 : 105 1 : 1 36.2
0.001 0.003 0.120 1 : 3 : 120 1 : 1 51.1
Table B.7. Cu3BTC2 synthesis using varying equivalents of a 66% benzoic acid
modulator, with excess linker and a metal concentration of 0.001 M. M, L, and
Mod refer to the concentrations of the metal salt, linker, and modulator in M.
These data are depicted in Figure 3.3 of the main text and SEM images are in
figure B.10.
Abs. Eq.
M L Mod Mod BA : B- Size (nm) Morphology
M : L : Mod
0.001 0.003 0.003 1 : 3 : 3 2 : 1 111.2
0.001 0.003 0.021 1 : 3 : 21 2 : 1 94.3 Octahedral
0.001 0.003 0.030 1 : 3 : 30 2 : 1 51.2 Globular
0.001 0.003 0.042 1 : 3 : 42 2 : 1 59.5
0.001 0.003 0.060 1 : 3 : 63 2 : 1 42.7
0.001 0.003 0.084 1 : 3 : 84 2 : 1 41.5
0.001 0.003 0.105 1 : 3 : 105 2 : 1 55.1
0.001 0.003 0.120 1 : 3 : 120 2 : 1 59.0 Octahedral
129
Table B.8. Cu3BTC2 synthesis using varying equivalents of a 33% benzoic acid
modulator, with a stoichiometric amount of linker and a metal concentration of
0.0027 M. M, L, and Mod refer to the concentrations of the metal salt, linker, and
modulator in M.
Abs. Eq.
M L Mod Mod BA : B- Size (nm)
M : L : Mod
0.0027 0.0018 0.0018 1 : 3 : 1 1 : 2 85.6
0.0027 0.0018 0.0126 1 : 3 : 4.67 1 : 2 46.0
0.0027 0.0018 0.0252 1 : 3 : 9.33 1 : 2 32.0
0.0027 0.0018 0.0378 1 : 3 : 14 1 : 2 31.0
0.0027 0.0018 0.0504 1 : 3 : 18.67 1 : 2 32.0
0.0027 0.0018 0.0630 1 : 3 : 23.33 1 : 2 20.6
0.0027 0.0018 0.0720 1 : 3 : 26.67 1 : 2 23.0
0.0027 0.0018 0.1080 1 : 3 : 40 1 : 2 26.9
Table B.9. Cu3BTC2 synthesis using varying equivalents of a 50% benzoic acid
modulator, with a stoichiometric amount of linker and a metal concentration of
0.0027 M. M, L, and Mod refer to the concentrations of the metal salt, linker, and
modulator in M.
Abs. Eq.
M L Mod Mod BA : B- Size (nm)
M : L : Mod
0.0027 0.0018 0.0018 1 : 3 : 1 1 : 1 126.0
0.0027 0.0018 0.0126 1 : 3 : 4.67 1 : 1 70.1
0.0027 0.0018 0.0252 1 : 3 : 9.33 1 : 1 74.5
0.0027 0.0018 0.0378 1 : 3 : 14 1 : 1 28.0
0.0027 0.0018 0.0504 1 : 3 : 18.67 1 : 1 20.0
0.0027 0.0018 0.0630 1 : 3 : 23.33 1 : 1 22.6*
0.0027 0.0018 0.0720 1 : 3 : 26.67 1 : 1 23.3
0.0027 0.0018 0.1080 1 : 3 : 40 1 : 1 27.7
*This sample was used to show the peak fitting process in Figure S1.
130
Table B.10. Cu3BTC2 synthesis using varying equivalents of a 66% benzoic acid
modulator, with a stoichiometric amount of linker and a metal concentration of
0.0027 M. M, L, and Mod refer to the concentrations of the metal salt, linker, and
modulator in M.
Abs. Eq.
M L Mod Mod BA : B- Size (nm)
M : L : Mod
0.0027 0.0018 0.0018 1 : 3 : 1 2 : 1 99.9
0.0027 0.0018 0.0126 1 : 3 : 4.67 2 : 1 92.4
0.0027 0.0018 0.0252 1 : 3 : 9.33 2 : 1 44.4
0.0027 0.0018 0.0378 1 : 3 : 14 2 : 1 46.8
0.0027 0.0018 0.0504 1 : 3 : 18.67 2 : 1 36.4
0.0027 0.0018 0.063 1 : 3 : 23.33 2 : 1 32.5
0.0027 0.0018 0.072 1 : 3 : 26.67 2 : 1 31.6
0.0027 0.0018 0.108 1 : 3 : 40 2 : 1 31.5
Table B.11. Cu3BTC2 synthesis using varying equivalents of a 50% benzoic acid
modulator, with a stoichiometric amount of linker and a metal concentration of
0.001 M. M, L, and Mod refer to the concentrations of the metal salt, linker, and
modulator in M.
Abs. Eq.
M L Mod Mod BA : B- Size (nm)
M : L : Mod
0.001 0.00067 0.00067 1 : 3 : 1 1 : 1 74.2
0.001 0.00067 0.00467 1 : 3 : 4.67 1 : 1 73.4
0.001 0.00067 0.00933 1 : 3 : 9.33 1 : 1 49.1
0.001 0.00067 0.01407 1 : 3 : 14 1 : 1 32.6†
0.001 0.00067 0.01867 1 : 3 : 18.67 1 : 1 55.8
0.001 0.00067 0.02345 1 : 3 : 23.33 1 : 1 43.8
†Note that a phase impurity was observed in this sample. These peaks were
excluded from Scherrer analysis.
131
Table B.12. Cu3BTC2 synthesis with varying total concentration, employing 7
equivalents of modulator with respect to the linker. M, L, and Mod refer to the
concentrations of the metal salt, linker, and modulator in M. The ratios of the
reagents and the total volume were kept constant. These data are depicted in
Figure 4 of the main text.
Abs. Eq.
M L Mod Mod BA : B- Size (nm)
M : L : Mod
0.00067 0.002 0.0140 1 : 3 : 21 1 : 1 63.9
0.001 0.003 0.0210 1 : 3 : 21 1 : 1 74.0
0.00167 0.005 0.0350 1 : 3 : 21 1 : 1 24.1
0.00250 0.0075 0.0525 1 : 3 : 21 1 : 1 22.1
0.00500 0.015 0.1050 1 : 3 : 21 1 : 1 21.2
0.00667 0.02 0.1400 1 : 3 : 21 1 : 1 14.3
Table B.13. Cu3BTC2 synthesis with varying total concentration, employing 0.7
equivalents of modulator with respect to the linker. The ratios of the reagents and
the total volume were kept constant. M, L, and Mod refer to the concentrations of
the metal salt, linker, and modulator in M. These data are depicted in Figure 4 of
the main text.
Abs. Eq.
M L Mod Mod BA : B- Size (nm)
M : L : Mod
0.00067 0.002 0.0140 1 : 3 : 2.1 1 : 1 104.0
0.00167 0.005 0.0350 1 : 3 : 2.1 1 : 1 87.0
0.00250 0.0075 0.0525 1 : 3 : 2.1 1 : 1 83.6
0.00333 0.01 0.0700 1 : 3 : 2.1 1 : 1 81.4
0.00500 0.015 0.1050 1 : 3 : 2.1 1 : 1 87.8
0.00667 0.02 0.1400 1 : 3 : 2.1 1 : 1 75.0
132
Table B.14. Cu3BTC2 synthesis with varying total concentration, employing 13.34
equivalents of modulator with respect to the linker. The ratios of the reagents and
the total volume were kept constant. These data are depicted in Figure 4 of the
main text.
Abs. Eq.
M Conc L Conc Mod Conc Mod BA : B- Size (nm)
M : L : Mod
0.00067 0.002 0.0140 1 : 3 : 40.02 1 : 1 ‡
0.00167 0.005 0.0350 1 : 3 : 40.02 1 : 1 17.8
0.00250 0.0075 0.0525 1 : 3 : 40.02 1 : 1 17.8
0.00333 0.01 0.0700 1 : 3 : 40.02 1 : 1 12.4
0.00500 0.015 0.1050 1 : 3 : 40.02 1 : 1 17.8
0.00667 0.02 0.1400 1 : 3 : 40.02 1 : 1 17.6
‡This reaction mixture, although blue in color, never centrifuged down. If
any particles were present, they could not be isolated.
Table B.15. Cu3BTC2 synthesis scale up of representative samples. The total
volume was kept constant.
Abs. Eq.
M L Mod Mod BA : B Size (nm)
M : L : Mod
0.001 0.003 0.021 1 : 3 : 21 2 : 1 82.8
0.001 0.003 0.021 1 : 3 : 21 1 : 1 61.0
0.001 0.003 0.021 1 : 3 : 21 1 : 2 50.7
0.001 0.003 0.063 1 : 3 : 63 1 : 1 46.4
Table B.16. Cu3BTC2 synthesis utilizing copper acetate as a metal source and
benzoic acid as the modulator.
Abs. Eq.
M L Mod Scherrer Size (nm)
M : L : Mod
0.001 0.003 0.003 1 : 3 : 3 72.7
0.001 0.003 0.021 1 : 3 : 21 59.0
0.001 0.003 0.042 1 : 3 : 42 45.7
0.001 0.003 0.060 1 : 3 : 63 47.5
0.001 0.003 0.084 1 : 3 : 84 49.2
0.001 0.003 0.120 1 : 3 : 120 56.3
133
Table B.17. ZIF-8 synthesis with increasing linker equivalents: In this experiment,
a modified literature method was used.1 The total volume changes, allowing linker
concentration (0.1901 M) to remain constant.
Linker Eq M Conc L Conc M : L Total Volume (mL) Scherrer Size (nm)
4 0.04752 0.1901 1:4 5 63.8
6 0.0317 0.1901 1:6 7.5 38.0
8 0.0238 0.1901 1:8 10 27.7
10 0.0190 0.1901 1:10 12.5 19.7
12 0.0158 0.1901 1:12 15 16.7
14 0.0136 0.1901 1:14 17.5 14.0
16 0.0119 0.1901 1:16 20 11.5
18 0.0106 0.1901 1:18 22.5 15.0
Table B.18. ZIF-8 synthesis with increasing linker equivalents: In this experiment,
a modified literature method was used.1 The total volume changes, allowing linker
concentration (0.0988 M) to remain constant.
Linker Eq M Conc L Conc M : L Total Volume (mL) Scherrer Size (nm)
6 0.0165 0.0988 1 : 5.85 15 19.8
8 0.0123 0.0988 1 : 7.76 20 16.3
10 0.0099 0.0988 1 : 9.79 25 13.1
12 0.0082 0.0988 1 : 11.6 30 11.2
14 0.0071 0.0988 1 : 13.9 35 7.8
16 0.0062 0.0988 1 : 15.4 40 13.6
Table B.19. ZIF-8 synthesis with HCl: This experiment used the conditions for 14
Hmim equivalents. A 1 M solution of HCl was made from concentrated HCl mixed
into methanol.
HCl Eq M Conc L Conc M : L : HCl Total Volume (mL) Scherrer Size (nm)
0.05 0.0136 0.1901 1:14.0:0.05 17.5 16.1
0.1 0.0136 0.1901 1:14:0.1 17.5 20.4
0.2 0.0136 0.1901 1:14:0.2 17.5 25.5
0.3 0.0136 0.1901 1:14:0.3 17.5 32.6
0.4 0.0136 0.1901 1:14:0.4 17.5 59.6
0.5 0.0136 0.1901 1:14:0.5 17.5 61.9
0.6 0.0136 0.1901 1:14:0.6 17.5 70.4
0.7 0.0136 0.1901 1:14:0.7 17.5 72.3
134
PXRD Patterns of Cu3BTC2 and ZIF-8 Products.
Figure B.2. Initial experiments were performed to explore the use of sodium
benzoate in Cu3BTC2 synthesis. a) PXRD patterns of Cu3BTC2 synthesized using
only sodium benzoate as the modulator. The simulated PXD pattern is shown in
grey. Asterisks indicate peaks that are do not appear in the simulated pattern. b)
Particle size decreases as the amount of sodium benzoate increases. Particle size
was determined by Scherrer analysis.
Figure B.3. Representative PXRD patterns of Cu3BTC2 synthesized with
increasing modulator equivalents (50% benzoic acid) a) With linker in excess
(L:M 3:1), the PXRD peaks first broaden with respect to added modulator up
to 28 equivalents, after which the peaks begin to narrow. Synthetic parameters for
these products can be found in table S6. b) With linker in a stoichiometric ratio
(L:M 2:3), the PXRD peaks broaden with respect to added modulator up to 40
equivalents. At 60 equivalents, there is a slight narrowing of the peaks. Synthetic
parameters for these products can be found in Table B.9.
135
Figure B.4. Representative PXRD patterns of ZIF-8. a) ZIF-8 synthesized with
increasing Hmim equivalents. Synthetic parameters for these products can be found
in table B.16. b) ZIF-8 synthesized with 14 Hmim equivalents and increasing HCl.
Synthetic parameters for these products can be found in table B.18.
136
Figure B.5. The two ratios 3L : 1M and 2L : 3M are compared at the same metal
concentration of 0.001 M with a 50% benzoic acid modulator. From 10 to 30
equivalents, the excess linker samples exhibit smaller sizes. Synthetic parameters
for these samples can be found in Tables B.6 and B.11. Generally, we found
conditions without excess linker to be less reproducible and chose to focus on 3L
: 1M reaction conditions.
Figure B.6. Data from several samples is compared to show reproducibility. The
trendline shown is the same data as in Figure 3 (Table B.1). Open circles are
repeated trials. Black triangles are scale-up syntheses, the synthetic conditions for
which can be found in Table B.14.
137
Figure B.7. Additional crystallite size data for Cu3BTC2 with respect to modulator
equivalents. a) Full data set for modulator equivalents shown in Fig 3.4 in the
main text. At 1 equivalent of modulator, all the crystallites are above 1 micron. b)
Copper acetate can be used as a metal source without an external source of base,
as acetate can deprotonate the MOF linkers. Increasing equivalents of benzoic acid
results in a seesaw trend without the long plateau observed in the buffered systems.
138
Figure B.8. SEM images of Cu3BTC2 particles as a function of benzoic acid
content at constant modulator equivalents. Reaction conditions for these samples
can be found in Table B.1.
139
Figure B.9. SEM images of Cu3BTC2 particles as a function of modulator
equivalents for a 66% benzoic acid modulator mixture. Reactant concentrations
for these samples can be found in Table B.7.
Figure B.10. Particle size from SEM compared to Scherrer crystallite size, as a
function of benzoic acid content. Some samples were excluded from SEM sizing due
to ambiguity when particles overlap or when the edges of particles are not clear
(Fig. S9). SEM sizes are generally above Scherrer sizes. The two methods are in
reasonable agreement until the Scherrer size exceeds 100 nm; grey lines show the
100 nm limit. Error bars on Scherrer sizes are from two batches, while error bars
from SEM sizes are the size dispersity from at least 20 particles. SEM particle size
was determined by using the circle tool and measure function in ImageJ.
140
Figure B.11. Particle size from SEM compared to Scherrer crystallite size as a
function of modulator equivalents for a 33% benzoic acid modulator mixture. SEM
sizes are above Scherrer sizes in all cases, which is typical of the two instrumental
techniques. SEM sizes for the crystals isolated with 1 eq. of modulator were over
1 m, indicated here by an arrow. Error bars from SEM sizes are the size dispersity
from at least 20 particles using the circle and measure tools in ImageJ. () Sizes could
not be determined for the sample at 21 equivalents. SEM images are in Figure B.7,
and reaction conditions can be found in Table B.5.
141
APPENDIX C
SUPPLEMENTARY INFORMATION FOR CHAPTER 4
General Methods
Materials
All commercial chemicals were used as received and handled in inert
conditions unless stated otherwise. All solvents were collected from a solvent
purification system and stored over 4Å molecular sieves, and all liquid reagents
were freeze-pump-thawed four cycles prior to use. N,N-dimethylformamide (DMF,
ACS grade, Fisher Scientific), acetonitrile (MeCN, HPLC grade, Fisher Scientific),
dichloromethane (DCM, ACS grade, Fisher Scientific), iron (II) chloride (98%,
anhydrous, Strem), 1-methylimidazole (99%, Sigma-Aldrich), 1,2,3-triazole (98%,
TCI), tetrabutylammonium hexafluorophosphate (98%, TCI, recrystallized 3 times
from ethanol), tetrabutylammonium tetrafluoroborate (98%, ACROS, recrystallized
once from ethanol and once from ethyl acetate).
All manipulations were performed under N2 unless stated otherwise.
Solution-state UV-Vis spectra were collected using either a Shimazdu Biospec-1601
for visible range measurements, and a Perkin Elmer Lambda-1050 UV/Vis/NIR
spectrophotometer for extended range measurements. For acid-digestion 1H-NMR,
samples were dried under vacuum, digested in 10% DCl / D2O in DMSO-D6 in air,
then filtered through cotton plugs prior to analysis with a Bruker Advance III-HD
600 NMR Spectrometer. IR spectra were recorded on a Bruker Alpha II compact
IR with an ATR attachment in a N2-filled glovebox.
Dynamic Light Scattering (DLS) experiments. DLS data was collected using
a Wyatt Mobius instrument with a custom-made airfree quartz cuvette with a
pathlength of 1 mm. Samples suspended in DMF were filtered through 0.45-µm
142
PTFE filters, and the solvent itself was prepared by filtration through a 0.10-m
PTFE filter. A normal 50-mW laser mode was used, and the samples were diluted
such that the measured counts were between 1 and 8 million, and the correlation
function was reproducible over the course of 6 measurements 1 minute apart.
N2 sorption measurements. For gas sorption measurements, the samples
were further washed with MeCN twice, and DCM five times. A typical washing
process proceeded over the course of 1 week. Samples were dried under vacuum in
tared ASAP tubes. Samples were degassed under high vacuum and 120 °C heat on
an ASAP 2020 instrument; degassing was considered complete when the pressure
in the closed manifold rose less than 2.5 µtorr/min. BET analysis was based on a
linear fit in the BET plot to N2 isotherm data at relative pressures between 10
−5 –
10−1 P/P0. Data for these experiments can be found in Figure C.10.
Crystallite size analysis by Le Bail fitting. Crystallite size of 130 nm
particles and 5.5 nm particles was characterized by Le Bail fitting on the PXRD
patterns in GSAS-II (280) according to the literature. (281) The analysis allows to
deconvolute the sample (Lorentzian) and the instrumental (Gaussian) broadening
parameters. Aluminum oxide (Al2O3, 300 nm) was used to determine the
instrumental broadening parameters (U, V, W, the Gaussian component). The
sample broadening parameters, crystallite size (X from the Lorentzian component)
and crystalline strain (Y from the Lorentzian component) were refined while the
obtained instrumental broadening parameters were fixed. X and Y parameters were
utilized to estimate the crystallite size and crystalline strain with the following
formulas.
Crystallite size:
143
18000Kλ
p = (C.1)
Xπ
Crystalline strain:
π
S = Y 100% (C.2)
18000
where K is the Scherrer constant (0.90), λ is the wavelength (0.15418 nm),
X and Y are the Lorentzian components of the peak profile, and = 3.14. The
obtained parameters are summarized in Table C.1.
Synthesis of Fe(TA)2 under Different Conditions
In the syntheses of iron triazolate particles, several variables were used to
control particle size and dispersity: reaction time, concentration, and the identity
of an added modulator. We found that the calculated Scherrer size of the particles
decreased upon dilution, and upon the addition of sodium formate, n-butylamine,
1-methylimidazole, 5-bromo-1-methylimidazole, and 1-benzyl-2-methylimidazole.
Additionally, particle size decreased upon using further excess 1,2,3-triazole: the
typical synthesis gave 44 ± 7 nm but having 4 equivalents of triazolate rather than
3 further decreased the size to 10 ± 1 nm (Fig, D.2e).
144
Figure C.1. PXRD patterns of Fe(TA)2 products synthesized with modulators.
Expected reflections are shown as grey lines, and possible phase impurities are
denoted by asterisks. a) Syntheses using 1.0 eq of n-butylamine, sodium formate,
and 1-methylimidazole. Photos of reaction solutions prior to heating are inset
next to each trace. b) Fe(TA)2 products created using varying equivalents of
1-benzyl-2-methlimidazole. c) Fe(TA)2 products obtained from syntheses with
varying amounts of 5-bromo-1-methylimidazole. d) Fe(TA)2 products obtained from
syntheses with varying amounts of 1-methylimidazole.
We noticed a dependence of particle size on the batch of iron (II) chloride
stock. Purposeful oxidation of iron (II) chloride by brief oxygen exposure led to
particles of decreased size: the control sample with rigorous air-free conditions
gave 44 ± 7 nm and with air-exposed FeCl2 the particle size was 12 ± 1 (Fig.
C.2f) Tables C.1-C.4 contain reaction conditions and particle sizes for all Fe(TA)2
particles discussed herein. Sodium formate (1 eq) did not have a significant impact
on crystallite size, and the PXRD pattern exhibited a phase impurity peak (Fig.
C.1a).
145
Figure C.2. SEM images of Fe(TA)2 products synthesized with different reaction
times and conditions. a) Polydisperse particles are created when a reaction
with 3.763 eq 1-mIm was left overnight. b) Particles increase only slightly in
polydispersity for 0.218 equivalents when allowed to react for 5 hours instead of
1.5. c) With 0.055 eq 1-mIm and a 5-hour reaction time, the resulting particles
are more polydisperse. d) When allowed to react overnight, the 0.055 eq 1-mIm
reaction will grow long rods in addition to smaller pseudo-octahedral particles.
Initially, reactions were performed without stirring and with a 18-21 hour
reaction time, which resulted in extremely high polydispersity (Fig. C.2a). We
146
found that, at 0.055 equivalents of 1-mIm the particles would grow into bulk-
like structures if allowed to react overnight, while a 5 hour reaction time did not
significantly change the particle size, but did increase the polydispersity (130 ± 10
nm vs 112 ± 20 nm with a 5 hour reaction). A five-hour reaction time for 0.218 eq
1-mIm gave the same size and polydispersity as the 1.5 hour reaction (Fig. C.2b,
Fig 4.2). The particles characterized throughout the main manuscript were all
synthesized with a 1.5 hour reaction time. Exploratory syntheses were performed
on a 2-mL scale; when suitable conditions were found, the reactions were scaled up
to 14 mL, and no changes in the crystallite sizes or SEM sizes were observed. The
particles synthesized with the lowest equivalents were less replicable than others,
possibly due to the larger error in micropippettors at small volumes (Table C.2).
Figure C.3. Reaction yield of Fe(TA)2 as a function of 1-methylimidazole
equivalents. All nanoparticle reactions were halted at 1.5 hours. Low yield is
typical in nanomof synthesis. Trials with 5 hours resulted in a higher polydispersity
and a marginally improved yield.
147
Figure C.4. Iron triazolate crystallize size trends with respect to modulator
equivalents. Sizes were determined by Scherrer analysis. a) Trend of particle
size from Scherrer analysis for three different modulators. Bromo = 5-
bromo-1-methylimidazole; Benzyl = 1-benzyl-2-methylimidazole; 1-mIm = 1-
methylimidazole. b) A zoom-in showing the trend at low equivalents for the three
modulators.
148
Figure C.5. Le Bail fits for largest and smallest Fe(TA)2 nanoparticles. Fitting
conditions and parameters can be found in Section C.1 and Table C.1. a) Le Bail
fitting for 130 nm Fe(TA)2 particles gives a domain size of 128.53 nm. b) Le Bail
fitting for 5.5 nm particles gives a domain size of 6.40 nm.
Table C.1. Crystallite sizes and strain of 130 nm (0.05 eq 1-mIm) and 5.5 nm (10.9
eq 1-mIm) particles from Le Bail fitting
Sample p /nm S / % Rwp / %
5.5-nm 6.40 4.76 8.01
130-nm 128.53 0.21 11.69
149
Basic Characterization of Fe(TA)2 Nanoparticles
Figure C.6. Larger SEM images of iron triazolate products. a) the morphology
resulting from a typical synthesis of bulk iron triazolate, b) – d) nanoparticles of
iron triazolate synthesized with varying equivalents of 1-methylimidazole.
150
Table C.2. Iron triazolate particle sizes from syntheses with varying amounts of
1-methylimizadole (modulator eq is with respect to FeCl2). We include Scherrer
sizes for two separate batches, followed by the statistical particle distribution
as determined by sizing particles from SEM image data with the line tool in
ImageJ. (2) The particle size distribution was fit to a weighted gaussian, and the
values presented are the mode ± standard deviation (σ).
Mod. Eq. Batch 1 Size (nm) Batch 2 Size (nm) Batch 2 SEM Size (nm) Dispersity
0.055 57 63 130 ± 10 0.09
0.109 50 56 84 ± 20 0.18
0.218 34 35 48 ± 3 0.06
0.436 16 19 25 ± 3 0.10
0.709 9.1 9.1 16 ± 2 0.11
3.00 6.8 6.8 — —
10.9 5.4 5.5 — —
Figure C.7. SEM images of the two smallest Fe(TA)2 particle sizes. a) With 3.0
1-mIm equivalents, the Scherrer size is 6.8 nm and although small nanosized
features are observed in the SEM, they cannot be fully resolved. b) With 10.9
1-mIm equivalents, the Scherrer size is 5.5 nm and the SEM image shows small
nanosized features that cannot be resolved.
151
Table C.3. Iron triazolate particle sizes from syntheses with varying amounts of
5-bromo-1-methylimidazole. We include Scherrer sizes, and a select SEM size for
these exploratory modulator syntheses. The particle size distribution was fit to
a weighted Gaussian curve, and the values presented are the mode ± standard
deviation (σ). The asterisk denotes a significant phase impurity observed in the
sample.
Mod. Eq. (to FeCl2) Scherrer Size (nm) SEM Size (nm) Dispersity
0.055 64 — —
0.109 73 — —
0.218 71 78 ± 6 0.08
0.436 69 — —
0.709 45 — —
3.00 26 — —
10.9 56* N/A —
Table C.4. Iron triazolate particle sizes from syntheses with varying amounts of
1-benzyl-2-methylimidazole. We include Scherrer sizes, and a select SEM size for
these exploratory modulator syntheses. The particle size distribution was fit to a
weighted gaussian, and the values presented are the mode ± standard deviation
(σ).
Mod. Eq. (to FeCl2) Scherrer Size (nm) SEM Size (nm) Dispersity
0.055 55 — —
0.109 39 — —
0.218 21 22 ± 3 0.14
0.436 7.3 — —
Table C.5. Iron triazolate particle sizes from six identical syntheses. To ensure the
synthesis was replicable, seven small scale syntheses were repeated. Phase purity
was confirmed by PXRD and Scherrer analysis was performed to obtain particle
sizes.
Equivalents Scherrer Size (nm)
0.709 9.1
0.709 9.0
0.709 8.8
0.709 10
0.709 8.1
0.709 8.1
0.709 8.3
152
Figure C.8. Acid digestion 1H NMR spectra of iron triazolate nanoparticles
compared to the constituent ligands and the bulk MOF product. MOF samples
were digested in 1.5 : 7 DCl : DMSO. Asterisks indicate peaks from DMF. An
arrow indicates a peak coincident with the methyl group of 1-methylimidazole; no
other peaks from the modulator appear.153
Figure C.9. IR spectra of the bulk material and the smallest Fe(TA)2 nanoparticles.
The expected stretches for the triazole ligand appear in both traces. For the
nanoparticles, there are several additional stretched attributed to trapped DMF
solvent.
Figure C.10. DLS data collected for Fe(TA)2 particles synthesized with 0.436
eq 1-methylimidazole. a) Solvated radius, in DMF, of the colloidal particles was
measured six times and the collected data is plotted, then fit to a Gaussian curve,
resulting in a solvated radius of 30 ± 5 nm. The x axis is presented in logarithmic
scale. b) Representative correlation functions from the first and sixth measurement,
showing a reproducible curve.
154
Figure C.11. Differential Scanning Calorimetry of FeTA2 nanoparticles synthesized
with 3.63 eq 1-mIm compared to previously published data of the bulk material. ()3
Cycles were collected at a scan rate of 10 °C min−1.
155
Figure C.12. Mössbauer spectra of several sizes of Fe(TA)2 in air. Spectra of the
smallest nanoparticles shows a slight shoulder at room temperature (RT) (a) and
clear quadropole splitting at 77 K (b). Spectra of medium sized particles (25 nm)
exhibit a second feature similar to that observed for the smallest particles under N2
in Fig 4.3 (c,d). Spectra of large 84-nm particles show clear oxidation (e,f).
156
Figure C.13. UV-Vis traces for syntheses employing other modulators. Absorption
is normalized. All samples were synthesized with 0.218 equivalents of the
modulators: 5-bromo-1-methylimidazole (Bromo), 1-benzyl-2-methylimidazole
(Benzyl), and n-butylamine (nBua). Broadening in the spectra of the Bromo
sample is likely due to poor colloidal stability.
Additional UV-Vis Data.
157
Figure C.14. Beer’s Law plots used to determine Fe(TA)−2 extinction coefficients.
At each concentration, two spectra were collected a few minutes apart to ensure
there were no transient effects. Linear fits were conducted in Igor with the y-
intercept held at 0.0 Abs.
158
Figure C.15. Energetic transitions determined from UV-Vis spectra, from the raw
data and from gaussian fits, plotted against particle diameter, 1/r, and 1/(r2).
The absolute maximum of the two peaks CT1 and CT2 were used to determine
extinction coefficients (solid line). The peaks were additionally fitted to gaussian
curves using the Multipeak 2.0 package in Igor 6.3 (dashed line).
159
Additional Electrochemical Data. Although the particle dispersions
are colloidally stable for months, our CV analysis of the colloids was carefully
conducted 3 days after the particle batch was synthesized to avoid any possible
aggregation effects from long-term storage. All the relevant CV data was collected
within four hours. Even so, there is a noticeable shift to lower potential in the first
(70 mV/s) scan compared to the subsequent scans in the 6.8 nm particle batch (Fig
D.15b).
Figure C.16. CV scans of Fe(TA)2 particles of varying size as colloids in 0.1 M
TBAPF6 / DMF. Scans were collected with rates of 10 mV/s (lightest blue),
40 mV/s, 70 mV/s, 100 mV/s, and 130 mV/s (black). The particle sizes are
5.5 nm (a), 6.8 nm (b), 15 nm (c) and 25 nm (d). The colloids have uncertain
concentration due to the low stability of the particles under bias and in electrolyte
solution; the current is therefore normalized to the area of the GC working
electrode.
160
Table C.6. Peak positions of redox features in colloidal Fe(TA)2 in 0.1 M TBAPF6
/ DMF. Peaks in the colloids are wide and they overlap one another, such that
determining peak positions was not always possible to do accurately.
Particle Size nm Scan Rate mV/s Peak Pos R1 (V) Peak Pos R2 (V) Peak Pos O1 (V)
130 -0.669 -0.382
5.5 100 -0.667 -0.386
70 -0.390
130 -0.661 -0.416
6.8
100 -0.417
130 -0.641 -0.360 -0.585
16.3
100 -0.631 -0.359 -0.622
130 -0.675 -0.395
25.3 100 -0.400
70 -0.353
Figure C.17. Control CV collected of the triazole ligand in 0.1 M TBAPF6 / DMF.
Only a simple irreversible oxidation is observed. Current is normalized to the area
of the glassy carbon working electrode.
161
Figure C.18. CV scans of 16-nm Fe(TA)2 particles drop-casted onto the glassy
carbon working electrode in 0.1 M TBAPF6 / MeCN. a) Scans were collected with
rates of 10 mV/s (lightest green), 100 mV/s, 300 mV/s, and 500 mV/s (darkest
green). Data is normalized with respect to the area of the bare GC electrode. b)
Peak-to-peak separation of the two main faradaic events, denoted by a circle and
a triangle in the CV. c) Peak current with respect to scan rate with a linear fit.
Black and blue symbols correspond to oxidation and reduction events, respectively.
d) Position of the peak (E½) for the second faradaic event (circles). e) Position of
the peak for the first faradaic event (triangles).
162
Figure C.19. CV scans of the bulk Fe(TA)2 material. The bulk material was
dispersed in DMF, sonicated, then dropcast onto a glassy carbon electrode.
Data was collected in 0.1 M TBAPF6 / MeCN. Scans were collected with rates
of 10 mV/s (lightest grey), 100 mV/s, 300 mV/s, and 500 mV/s (black). Data is
normalized with respect to the area of the bare GC electrode.
Figure C.20. CV scans with varying scan rate on QCM electrodes in TBAPF6
(a) and TBABF4 (b). The lightest grey corresponds to 10 mV/s, followed by 100
mV/s, 300 mV/s, and finally 500 mV/s.
163
Figure C.21. Microscope images of films used in QCM experiments. a) SEM image
of a film as-deposited. b) SEM image of a film used for CV scans in TBABF4.
This ex-situ analysis was not performed in the case of TBAPF6 because the
same electrode is cleaned and re-used. c) Optical microscope image of a film
after running CVs in TBAPF6. A few scratches and drying defects are visible. D)
Optical microscope image of a film used to run CVs in TBABF4. A small scratch
and similar drying defects are visible.
164
Table C.7. Values of electrons and ions determined from QCM experiments at 100
mV/s. The mass of the MOF was calculated using the Sauerbrey relation from
the change in frequency of the dry, bare QCM and the QCM after spin-coating
and the crystal was allowed to dry fully. Moles of electrons was determined by
integrating the CV curves. Moles of anions was calculated from change in frequency
over a cycle, using first the Saurbrey relationship, then the molecular weight of the
unsolvated anions.
TBAPF6 TBABF4
Mass Fe(TA) 22 deposited / cm 5.50 µg 4.41 µg
Moles Fe(TA)2 / cm
2 2.86 10−8 2.30 10−8
Average frequency change / cycle 22 Hz 46.8 Hz
Mass change / cm2 3.89 10−7 µg 8.27 10−7 µg
Moles anions / cm2 2.68 10−9 9.52 10−9
Anions per Fe 0.09 0.4
Moles e− (oxidation) 8.26 10−9 2.25 10−8
Moles e− (reduction) 7.20 10−9 1.86 10−8
Moles e− (average) 7.73 10−9 2.05 10−8
Table C.8. Raw data for conductivity measurements on Fe(TA)2 thin films made
in air. The films were allowed to sit in ambient aerobic conditions for over 1 week.
Measurements were performed with a co-linear four point probe method.
Particle Size Additives Film Thickness µm Conductivity S/cm
Bulk material None 4.1 1.08 × 10−5
130 ± 20 nm None 5.5 4.84 × 10−5
84 ± 30 nm None 6.6 1.53 × 10−4
84 ± 30 nm 5% carbon, 5% PVDF 4.1 6.30 × 10−4
165
Table C.9. Raw data for conductivity measurements on Fe(TA)2 thin films made
in N2. The films were allowed to sit in ambient aerobic conditions for 1 day, and
measurements were collected once per day whenever possible. Measurements were
performed with a Van der Pauw four point probe method. Two measurements are
collected in different orientations (R1 and R2). F is a correction factor.
Measurement Thickness µm R1 A/V R2 A/V F Conductivity S/cm
80 nm N2 2.1 36231884058 10615711253 0.8865 5.06 × 10−8
80 nm Air Day 1 2.1 211416490.5 492610837.4 0.94215 3.17 × 10−6
80 nm Air Day 2 2.1 137174211.2 267379679.1 0.96306 5.39 × 10−6
80 nm Air Day 3 2.1 182149362.5 85251491.9 0.95276 8.25 × 10−6
80 nm Air Day 4 2.1 35211267.61 60606060.61 0.97518 2.25 × 10−5
80 nm Air Day 7 2.1 28818443.8 13003901.17 0.9484 5.30 × 10−5
Bulk N2 8.6 15220700152 86206896552 0.79954 6.33 × 10−9
Bulk Air Day 1 8.6 37313432.84 854700854.7 0.5691 1.01 × 10−6
Bulk Air Day 2 8.6 41841004.18 552486187.8 0.6528 1.32 × 10−6
Bulk Air Day 3 8.6 43668122.27 617283950.6 0.6417 1.21 × 10−6
25 nm N2 2.6 3623188406 363636363.6 0.6999 6.11 × 10−7
25 nm Air Day 3 2.6 32268473.7 1298701.299 0.5577 9.10 × 10−5
166
Figure C.22. Photos of Fe(TA)2 thin films made in air. a) Films of the largest
sizes of particles were created from 20 mg/mL dispersions. Such concentrated
dispersions, when used for smaller particles, resulted in films that flaked easily off
the glass b) For smaller particles, 10 mg/mL dispersions were used. c) The bulk
film was fabricated from a 20 mg/ml dispersion. e) The composite film is comprised
of 90% MOF, with 5% PVDF binder and 5% carbon black by weight. The black
dots are sharpie marks, placed after a conductivity measurement, as a protective
carbon cap for SEM-FIB cross-sectioning.
167
Figure C.23. Photos of Fe(TA)2 thin films made in the glovebox, then brought out
into air. a) Bulk film in a clover-like shape with silver contacts at the corners. b)
84-nm film in a square shape with silver contacts at the corners. c) 25-nm film in a
rectangular shape with silver contacts at the corners.
168
Figure C.24. SEM images of Fe(TA)2 thin films of varying particle size made in air
and used for in-air co-linear conductivity measurements. Cross section (CS, a) and
surface (b) of the bulk material. Cross section (c) and close-up cross section (d)
of 130-nm particles. Cross section (e) and close-up cross-section (f) of the 84-nm
particle thin film. g) Cross section (g) and surface (h) of a composite film with
84-nm particles with 5% carbon black and 5% PVDF, with measurements.
169
Figure C.25. SEM images of Fe(TA)2 thin films of varying particle size made under
N2 and used for Van der Pauw conductivity measurements. a) Top of the bulk
material. b) Cross-section (CS) of the bulk film. c) Top of the 84-nm particle
film. d) Cross-section (CS) of the 84-nm particle film. e) Top of the 25-nm film. f)
Cross-section (CS) of the 25-nm film. 170
APPENDIX D
SUPPLEMENTARY INFORMATION FOR CHAPTER 5
General Methods
Materials All commercial chemicals were used as received and handled in
inert conditions unless stated otherwise. All solvents were collected from a solvent
purification system and stored over 4A molecular sieves, and all liquid reagents
were freeze-pump-thawed four cycles prior to use. N,N-dimethylformamide (DMF,
ACS grade, Fisher Scientific), acetonitrile (MeCN, HPLC grade, Fisher Scientific),
dichloromethane (DCM, ACS grade, Fisher Scientific), 1,2-difluorobenzene (98%,
Sigma-Aldrich), tetrahydrofuran (THF, ACS grade, Fisher Scientific), iron (II)
chloride (98%, anhydrous, Strem), 1-methylimidazole (99%, Sigma-Aldrich),
1,2,3-triazole (98%, TCI), tetrabutylammonium hexafluorophosphate (98%,
TCI, recrystallized once from ethanol), tetrabutylammonium tetrafluoroborate
(98%, ACROS, recrystallized once from ethanol), tetrabutylammonium
perchlorate (99+%, Fisher Scientific), lithium tetrafluoroborate (98.0%, TCI), and
tetramethylammonium tetrafluoroborate (98.0%, TCI).
Synthesis and Characterization of Iron Triazolate (Fe(TA)2)
Nanoparticles. Controlling the size of Fe(TA)2 nanoparticles was achieved
by adding different amounts of 1-methyl imidazole (1-mIm) to the synthesis. In
detail, the solution of anhydrous iron(II) chloride in DMF (0.805 mmol, 0.0575
M, 14 mL) was first prepared, followed by adding 45.5 µL (0.709 equiv), 25 µL
(0.436 equiv), and 14 µL (0.218 equiv) of 1-mIm for the synthesis of 16-nm, 25-
nm, and 46-nm nanoparticles. It is noted that all equivalents are with respect to
iron(II) chloride. Vials were further heated under 120 ◦C for 1.5 hours and then
immediately centrifuged and washed twice with DMF.
171
PXRD Measurements and Analysis. PXRD data were collected in the air
using BraggBrentano geometry with a step size of 0.02◦ in the range of 3.5-35◦
2θ with a Bruker D2 Phaser. A variable detector opening was used to reduce air
scattering at low angles. Patterns were matched to the low-spin Fe(TA)2 cif file.
SEM Imaging for Size Analysis Imaging was performed using a Thermo
Fisher Apreo 2 SEM instrument with 10.00 kV energy and 0.8 nA current. SEM
samples were prepared by spin-coating the dispersion of Fe(TA)2 nanoparticles in
DMF onto silicon substrates. Particle sizing was performed in ImageJ. (2)
Atomic Force Microscopy (AFM) for Fe(TA)2 Film Thickness Analysis
A Bruker Dimension ICON with ScanAsyst was used for topographical AFM
measurements. The SCANASYST-AIR probes were used to measure the Fe(TA)2
film thickness on the quartz crystal electrode. Time-of-Flight Secondary Ion Mass
Spectrometry (ToF-SIMS) Experiments. Fe(TA)2 films on QCM electrodes were
analyzed on a ToF-SIMS IV time-of-flight instrument manufactured by ION TOF
GmbH, Münster, Germany. Mass spectra were collected using a beam of Bi3+
primary ions (bunched mode, 25 kV, 0.3 pA @ 10 kHz). The primary ion flux for
each mass spectrum was less than 2.0 x 1012 ions/cm2. An electron flood source
was used for charge compensation. Spectra were acquired from three separate areas
on a single sample. Data processing was done with SurfaceLab 6 software provided
by the manufacturer.
Electrochemical Studies of Fe(TA)2 Films All electrochemical data were
collected using a Biologic SP200. For electrochemical quartz crystal microbalance
(EQCM) experiments, the PT/Ti-coated 5 MHz AT-cut QCM working electrodes
were cleaned using acidic piranha solution, then rinsed copiously with 18.2 MΩ
nanopure water, followed by isopropyl alcohol, and lastly dried under N2 pressure.
172
Different amounts of suspensions of Fe(TA)2 nanoparticles in DMF were spin-
coated onto the EQCM electrodes in the N2 glovebox. The EQCM electrodes with
Fe(TA)2 films were further dried in the vacuum for 3 hours. A standard three-
electrode electrochemical cell was set up with 0.1 M electrolytes (i.e., TBABF4,
TBAPF6, TBAClO4, LiBF4, and TMABF4) in acetonitrile (80 mL), the QCM as
the working electrode, a platinum wire as the counter electrode, and a bare silver
wire as a pseudo-reference electrode. Frequency data were collected simultaneously
with CV scans using a SRS QCM200 apparatus. The frequency was converted to
mass using the Sauerbrey equation below, in which f is the experimental change in
frequency, Cf is the sensitivity factor (56.6 Hz cm2 g−1 for 5 MHz AT-cut crystals),
and m is the change in mass.
∆f = Cf∆m
Cyclic voltammogram (CV) of Fe(TA)2 film was collected under N2
condition at the rate of 10 mV/s. Titration experiments were conducted by
adding certain amounts of either TBAPF6 or TBAClO4 into TBABF4 / MeCN
electrolyte without changing the solvent volume. Impedance data were recorded
at potentials of either surface redox feature or ion-intercalation redox feature. An
AC amplitude of 10mV was applied. The frequency was scanned from 4.5 kHz to
60 mHz with 6 points per decade. Most of the EIS data were fitted using Bio-
Logic EC-Lab V11.33. For solvent-dependent electrochemical studies, CV was
collected in different solvents with 0.1 M TBABF4 as the supporting electrolyte
in a standard three-electrode cell with a glassy carbon (GC) working electrode, a
silver wire pseudo-reference electrode, and a platinum wire counter electrode. To
prepare the working GC electrode, 16-nm Fe(TA)2 nanoparticles were suspended in
DMF at a concentration of 5mg/mL and then drop-casted (7 µL) onto the polished
173
GC surface. The electrode was further dried in the vacuum for 3 hours before
electrochemical studies. In all CV experiments, potential is referenced to the Fc0/+
couple which is determined by adding a small amount of ferrocene to the solution
at the end of the experiment.
174
Figure D.1. Basic Characterization of Fe(TA)2. PXRD pattern (a) and SEM image
(b) of 16-nm particles. PXRD pattern (c) and SEM image (d) of 25-nm particles.
PXRD pattern (e) and SEM image (f) of 16-nm particles.
175
Figure D.2. CV traces starting and stopping at different voltages show that the 1.2
V feature is independent of other features in the CV.
176
Figure D.3. Charge variation in Fe(TA)2 films during CV measurements.(a) CV
(black) and charge change (green) of Fe(TA)2 film collected at a 10 mV/s scan rate.
(b) CV (black) and charge change (green) of Fe(TA)2 film focusing on the potential
region of 1.2 V vs. Fc0/+ redox feature. Scan direction is indicated by arrows and
the charge change is only shown in the anodic CV direction. The loading amount
of 16-nm Fe(TA)2 nanoparticles on EQCM crystal is ca. 4.0 µg; this is the same
sample depicted in Figure 5.2a
177
Figure D.4. Tracking the changes in ∆E and E1/2 at the 0.0 V and 1.2 V feature
during TBABF6 electrolyte titration. (a) ∆E variation of redox features at 0 V and
1.2 V vs. Fc0/+ of Fe(TA)2 film when titrating different amount of TBAPF6 into
0.1 M TBABF4 electrolyte. (b) E1/2 of redox features at 0 V and 1.2 V vs. Fc
0/+
for Fe(TA)2 film when titrating different amount of TBAPF6 into 0.1 M TBABF4
electrolyte.
178
Figure D.5. CVs collected during a titration experiment where TBACLO4 was
added to the TBABF4 electrolyte. CV results are collected at the scan rate of 10
mV/s
179
Figure D.6. CV measurements of Fe(TA)2 films in acetonitrile (red) and 1,2-
difluorobenzene (blue). The E1/2 of BF
−
4 intercalation redox feature is shifted from
ca. 1.26 V in acetonitrile to 1.11 V in 1,2-difluorobenzene.
180
Figure D.7. CVs of Fe(TA)2 particles in four different solvents. (a) CV of Fe(TA)2
film in acetonitrile. (b) CV of Fe(TA)2 film in dichloromethane. (c) CV of Fe(TA)2
film in tetrahydrofuran. (d) CV of Fe(TA)2 film in 1,2-difluorobenzene.
181
REFERENCES
1. Liu, B.; Vellingiri, K.; Jo, S.-H.; Kumar, P.; Ok, Y. S.; Kim, K.-H. Recent
advances in controlled modification of the size and morphology of metal-
organic frameworks. Nano Res. 2018, 11, 4441–4467.
2. Schindelin, J. et al. Fiji: An open-source platform for biological-image
analysis. Nature Methods 2012, 9, 676–682.
3. Moosavi, S. M.; Nandy, A.; Jablonka, K. M.; Ongari, D.; Janet, J. P.;
Boyd, P. G.; Lee, Y.; Smit, B.; Kulik, H. J. Understanding the diversity of
the metal-organic framework ecosystem. Nature Communications 2020, 11,
4068.
4. Forgan, R. S. Modulated self-assembly of metal–organic frameworks. Chem.
Sci. 2020, 11, 4546–4562.
5. Chang, M.; Wei, Y.; Liu, D.; Wang, J.-X.; Chen, J.-F. A General Strategy
for Instantaneous and Continuous Synthesis of Ultrasmall Metal–Organic
Framework Nanoparticles. Angewandte Chemie International Edition 2021,
60, 26390–26396.
6. Marshall, C. R.; Staudhammer, S. A.; Brozek, C. K. Size control over
metal–organic framework porous nanocrystals. Chem. Sci. 2019, 10, 9396–
9408.
7. Troyano, J.; Carné-Sánchez, A.; Avci, C.; Imaz, I.; Maspoch, D. Colloidal
metal–organic framework particles: the pioneering case of ZIF-8. Chem. Soc.
Rev. 2019, 48, 5534–5546.
182
8. Cravillon, J.; Münzer, S.; Lohmeier, S.-J.; Feldhoff, A.; Huber, K.;
Wiebcke, M. Rapid Room-Temperature Synthesis and Characterization of
Nanocrystals of a Prototypical Zeolitic Imidazolate Framework. Chem. Mater.
2009, 21, 1410–1412.
9. Cravillon, J.; Schröder, C. A.; Nayuk, R.; Gummel, J.; Huber, K.;
Wiebcke, M. Fast Nucleation and Growth of ZIF-8 Nanocrystals Monitored
by Time-Resolved In Situ Small-Angle and Wide-Angle X-Ray Scattering.
Angew. Chem. Int. Ed. 2011, 50, 8067–8071.
10. Lim, I. H.; Schrader, W.; Schüth, F. Insights into the Molecular Assembly of
Zeolitic Imidazolate Frameworks by ESI-MS. Chem. Mater. 2015, 27, 3088–
3095.
11. Guo, H.; Zhu, Y.; Wang, S.; Su, S.; Zhou, L.; Zhang, H. Combining
Coordination Modulation with Acid–Base Adjustment for the Control over
Size of Metal–Organic Frameworks. Chem. Mater. 2012, 24, 444–450.
12. Umemura, A.; Diring, S.; Furukawa, S.; Uehara, H.; Tsuruoka, T.;
Kitagawa, S. Morphology Design of Porous Coordination Polymer Crystals
by Coordination Modulation. J. Am. Chem. Soc. 2011, 133, 15506–15513.
13. Day, R. W.; Bediako, D. K.; Rezaee, M.; Parent, L. R.; Skorupskii, G.;
Arguilla, M. Q.; Hendon, C. H.; Stassen, I.; Gianneschi, N. C.; Kim, P.;
Dincă, M. Single Crystals of Electrically Conductive Two-Dimensional
Metal–Organic Frameworks: Structural and Electrical Transport Properties.
ACS Cent. Sci. 2019, 5, 1959–1964.
183
14. Foster, M. E.; Sohlberg, K. Unraveling the semiconducting/metallic
discrepancy in Ni3(HITP)2. J. Phys. Chem. Lett. 2018, 11.
15. Hoppe, B.; Hindricks, K. D. J.; Warwas, D. P.; Schulze, H. A.; Mohmeyer, A.;
Pinkvos, T. J.; Zailskas, S.; Krey, M. R.; Belke, C.; König, S.; Fröba, M.;
Haug, R. J.; Behrens, P. Graphene-like metal–organic frameworks:
morphology control, optimization of thin film electrical conductivity and fast
sensing applications. CrystEngComm 2018, 20, 6458–6471.
16. Gittins, J. W.; Balhatchet, C. J.; Fairclough, S. M.; Forse, A. C. Enhancing
the energy storage performances of metal–organic frameworks by controlling
microstructure. 2022, 13, 9210–9219.
17. Mähringer, A.; Jakowetz, A. C.; Rotter, J. M.; Bohn, B. J.; Stolarczyk, J. K.;
Feldmann, J.; Bein, T.; Medina, D. D. Oriented Thin Films of Electroactive
Triphenylene Catecholate-Based Two-Dimensional Metal–Organic
Frameworks. ACS Nano 2019, 13, 6711–6719.
18. Lahiri, N.; Lotfizadeh, N.; Tsuchikawa, R.; Deshpande, V. V.; Louie, J.
Hexaaminobenzene as a building block for a Family of 2D Coordination
Polymers. J. Am. Chem. Soc. 2017, 139, 19–22.
19. Gándara, F.; Uribe-Romo, F. J.; Britt, D. K.; Furukawa, H.; Lei, L.;
Cheng, R.; Duan, X.; O’Keeffe, M.; Yaghi, O. M. Porous, Conductive Metal-
Triazolates and Their Structural Elucidation by the Charge-Flipping Method.
Chem. Eur. J. 2012, 18, 10595–10601.
20. Park, J. G.; Aubrey, M. L.; Oktawiec, J.; Chakarawet, K.; Darago, L. E.;
Grandjean, F.; Long, G. J.; Long, J. R. Charge Delocalization and Bulk
184
Electronic Conductivity in the Mixed-Valence Metal–Organic Framework
Fe(1,2,3-triazolate)2(BF4)x . J. Am. Chem. Soc. 2018, 140, 8526–8534.
21. Grzywa, M.; Röß-Ohlenroth, R.; Muschielok, C.; Oberhofer, H.;
B lachowski, A.; Żukrowski, J.; Vieweg, D.; von Nidda, H.-A. K.; Volkmer, D.
Cooperative Large-Hysteresis Spin-Crossover Transition in the Iron(II)
Triazolate [Fe(ta)2] Metal–Organic Framework. Inorg. Chem. 2020, 59, 10501–
10511.
22. Liu, L.; Corma, A. Metal Catalysts for Heterogeneous Catalysis: From Single
Atoms to Nanoclusters and Nanoparticles. Chem. Rev. 2018, 118, 4981–5079.
23. Llordés, A.; Garcia, G.; Gazquez, J.; Milliron, D. J. Tunable near-infrared and
visible-light transmittance in nanocrystal-in-glass composites. Nature 2013,
500, 323–326.
24. Schimpf, A. M.; Thakkar, N.; Gunthardt, C. E.; Masiello, D. J.;
Gamelin, D. R. Charge-Tunable Quantum Plasmons in Colloidal
Semiconductor Nanocrystals. ACS Nano 2014, 8, 1065–1072.
25. Kroupa, D. M.; Roh, J. Y.; Milstein, T. J.; Creutz, S. E.; Gamelin, D. R.
Quantum-Cutting Ytterbium-Doped CsPb(Cl1–xBrx )3 Perovskite Thin Films
with Photoluminescence Quantum Yields over 190%. ACS Energy Lett. 2018,
3, 2390–2395.
26. Sanehira, E. M.; Marshall, A. R.; Christians, J. A.; Harvey, S. P.;
Ciesielski, P. N.; Wheeler, L. M.; Schulz, P.; Lin, L. Y.; Beard, M. C.;
Luther, J. M. Enhanced mobility CsPbI3 quantum dot arrays for record-
efficiency, high-voltage photovoltaic cells. Science Advances 2017, 3 .
185
27. Chakraborty, I.; Pradeep, T. Atomically Precise Clusters of Noble Metals:
Emerging Link between Atoms and Nanoparticles. Chem. Rev. 2017, 117,
8208–8271.
28. Owen, J.; Brus, L. Chemical Synthesis and Luminescence Applications of
Colloidal Semiconductor Quantum Dots. J. Am. Chem. Soc. 2017, 139,
10939–10943.
29. Hühn, J. et al. Selected Standard Protocols for the Synthesis, Phase Transfer,
and Characterization of Inorganic Colloidal Nanoparticles. Chem. Mater.
2017, 29, 399–461.
30. Yin, Y.; Alivisatos, A. P. Colloidal nanocrystal synthesis and the
organic–inorganic interface. Nature 2005, 437, 664–670.
31. Wuttke, S.; Braig, S.; Preiß, T.; Zimpel, A.; Sicklinger, J.; Bellomo, C.;
Rädler, J. O.; Vollmar, A. M.; Bein, T. MOF nanoparticles coated by lipid
bilayers and their uptake by cancer cells. Chem. Commun. 2015, 51, 15752–
15755.
32. Wang, S.; Chen, Y.; Wang, S.; Li, P.; Mirkin, C. A.; Farha, O. K. DNA-
Functionalized Metal–Organic Framework Nanoparticles for Intracellular
Delivery of Proteins. J. Am. Chem. Soc. 2019, 141, 2215–2219.
33. Graf, M.; Moreno, N.; Kerl, K.; Bunzen, H. Metal–organic framework
nanoparticles for arsenic trioxide drug delivery. J. Mater. Chem. 2018, 6,
6481–6489.
34. Rogge, S. M. J.; Bavykina, A.; Hajek, J.; Garcia, H.; Olivos-Suarez, A. I.;
Sepúlveda-Escribano, A.; Vimont, A.; Clet, G.; Bazin, P.; Kapteijn, F.;
186
Daturi, M.; Ramos-Fernandez, E. V.; Llabrés i Xamena, F. X.;
Van Speybroeck, V.; Gascon, J. Metal–organic and covalent organic
frameworks as single-site catalysts. Chem. Soc. Rev. 2017, 46, 3134–3184.
35. Adil, K.; Belmabkhout, Y.; Pillai, R. S.; Cadiau, A.; Bhatt, P. M.;
Assen, A. H.; Maurin, G.; Eddaoudi, M. Gas/vapour separation using ultra-
microporous metal–organic frameworks: insights into the structure/separation
relationship. Chem. Soc. Rev. 2017, 46, 3402–3430.
36. He, Y.; Zhou, W.; Qian, G.; Chen, B. Methane storage in metal–organic
frameworks. Chem. Soc. Rev. 2014, 43, 5657–5678.
37. Yu, G.; Liu, Y.; Zou, X.; Zhao, N.; Rong, H.; Zhu, G. A nanosized
metal–organic framework with small pores for kinetic xenon separation. J.
Mater. Chem. A 2018, 6, 11797–11803.
38. Sakata, Y.; Furukawa, S.; Kondo, M.; Hirai, K.; Horike, N.; Takashima, Y.;
Uehara, H.; Louvain, N.; Meilikhov, M.; Tsuruoka, T.; Isoda, S.; Kosaka, W.;
Sakata, O.; Kitagawa, S. Shape-Memory Nanopores Induced in Coordination
Frameworks by Crystal Downsizing. Science 2013, 339, 193–196.
39. Furukawa, H.; Cordova, K. E.; O’Keeffe, M.; Yaghi, O. M. The Chemistry and
Applications of Metal-Organic Frameworks. Science 2013, 341, 1230444.
40. Freund, R.; Lächelt, U.; Gruber, T.; Rühle, B.; Wuttke, S. Multifunctional
Efficiency: Extending the Concept of Atom Economy to Functional
Nanomaterials. ACS Nano 2018-03-27, 12, 2094–2105.
41. Wang, J.; Wan, J.; Ma, Y.; Wang, Y.; Pu, M.; Guan, Z. Metal–organic
frameworks MIL-88A with suitable synthesis conditions and optimal dosage
187
for effective catalytic degradation of Orange G through persulfate activation.
RSC Adv. 2016, 6, 112502–112511.
42. Barros, B. S.; deLimaNeto, O. J.; deOliveiraFrós, A. C.; Kulesza, J. Metal-
Organic Framework Nanocrystals. ChemistrySelect 2018, 3, 7459–7471.
43. Sindoro, M.; Yanai, N.; Jee, A.-Y.; Granick, S. Colloidal-Sized Metal–Organic
Frameworks: Synthesis and Applications. Acc. Chem. Res. 2014, 47, 459–469.
44. Majewski, M. B.; Noh, H.; Islamoglu, T.; Farha, O. K. NanoMOFs: little
crystallites for substantial applications. J. Mater. Chem. A 2018, 6, 7338–
7350.
45. Hermes, S.; Witte, T.; Hikov, T.; Zacher, D.; Bahnmüller, S.; Langstein, G.;
Huber, K.; Fischer, R. A. Trapping metal-organic framework nanocrystals: An
in-situ time-resolved light scattering study on the crystal growth of MOF-5 in
solution. J. Am. Chem. Soc. 2007, 129, 5324–5325.
46. Hirschle, P.; Preiß, T.; Auras, F.; Pick, A.; Völkner, J.; Valdepérez, D.;
Witte, G.; Parak, W. J.; Rädler, J. O.; Wuttke, S. Exploration of MOF
nanoparticle sizes using various physical characterization methods – is what
you measure what you get? CrystEngComm 2016, 18, 4359–4368.
47. Patterson, J. P.; Abellan, P.; Denny, M. S.; Park, C.; Browning, N. D.;
Cohen, S. M.; Evans, J. E.; Gianneschi, N. C. Observing the Growth of
Metal–Organic Frameworks by in Situ Liquid Cell Transmission Electron
Microscopy. J. Am. Chem. Soc. 2015, 137, 7322–7328.
48. Zhu, Y.; Ciston, J.; Zheng, B.; Miao, X.; Czarnik, C.; Pan, Y.; Sougrat, R.;
Lai, Z.; Hsiung, C.-E.; Yao, K.; Pinnau, I.; Pan, M.; Han, Y. Unravelling
188
surface and interfacial structures of a metal–organic framework by
transmission electron microscopy. Nature Mater 2017, 16, 532–536.
49. Khan, N. A.; Hasan, Z.; Jhung, S. H. Beyond pristine metal-organic
frameworks: Preparation and application of nanostructured, nanosized, and
analogous MOFs. Coordination Chemistry Reviews 2018, 376, 20–45.
50. Wang, S.; McGuirk, C. M.; D’Aquino, A.; Mason, J. A.; Mirkin, C. A. Metal-
Organic Framework Nanoparticles. Adv. Mater. 2018, 30, 1800202.
51. Diring, S.; Furukawa, S.; Takashima, Y.; Tsuruoka, T.; Kitagawa, S.
Controlled Multiscale Synthesis of Porous Coordination Polymer in
Nano/Micro Regimes. Chem. Mater. 2010, 22, 4531–4538.
52. Wang, F.; Guo, H.; Chai, Y.; Li, Y.; Liu, C. The controlled regulation of
morphology and size of HKUST-1 by “coordination modulation method”.
Microporous and Mesoporous Materials 2013, 173, 181–188.
53. Xin, C.; Zhan, H.; Huang, X.; Li, H.; Zhao, N.; Xiao, F.; Wei, W.; Sun, Y.
Effect of various alkaline agents on the size and morphology of nano-sized
HKUST-1 for CO2 adsorption. RSC Adv. 2015, 5, 27901–27911.
54. Sun, W.; Zhai, X.; Zhao, L. Synthesis of ZIF-8 and ZIF-67 nanocrystals with
well-controllable size distribution through reverse microemulsions. Chemical
Engineering Journal 2016, 289, 59–64.
55. Ye, R.; Ni, M.; Xu, Y.; Chen, H.; Li, S. Synthesis of Zn-based metal–organic
frameworks in ionic liquid microemulsions at room temperature. RSC Adv.
2018, 8, 26237–26242.
189
56. Qi, Y.; He, C.-T.; Lin, J.; Lin, S.; Liu, J.; Huang, J.; Xue, W.; Yu, G.;
Chao, H.-Y.; Tong, Y.; Qiao, Z. Mild metal-organic-gel route for synthesis
of stable sub-5-nm metal-organic framework nanocrystals. Nano Res. 2017,
10, 3621–3628.
57. Wang, J.; Fan, Y.; Tan, Y.; Zhao, X.; Zhang, Y.; Cheng, C.; Yang, M.
Porphyrinic Metal–Organic Framework PCN-224 Nanoparticles for Near-
Infrared-Induced Attenuation of Aggregation and Neurotoxicity of Alzheimer’s
Amyloid-β Peptide. ACS Appl. Mater Interfaces 2018, 10, 36615–36621.
58. LaMer, V. K.; Dinegar, R. H. Theory, Production and Mechanism of
Formation of Monodispersed Hydrosols. J. Am. Chem. Soc. 1950, 72, 4847–
4854.
59. Cravillon, J.; Nayuk, R.; Springer, S.; Feldhoff, A.; Huber, K.; Wiebcke, M.
Controlling Zeolitic Imidazolate Framework Nano- and Microcrystal
Formation: Insight into Crystal Growth by Time-Resolved In Situ Static Light
Scattering. Chem. Mater. 2011, 23, 2130–2141.
60. Schäfer, P.; Kapteijn, F.; van der Veen, M. A.; Domke, K. F. Understanding
the Inhibiting Effect of BTC on CuBTC Growth through Experiment and
Modeling. Crystal Growth & Design 2017, 17, 5603–5607.
61. Van Vleet, M. J.; Weng, T.; Li, X.; Schmidt, J. In Situ, Time-Resolved, and
Mechanistic Studies of Metal–Organic Framework Nucleation and Growth.
Chem. Rev. 2018, 118, 3681–3721.
62. Liu, Y.-Y.; Decadt, R.; Bogaerts, T.; Hemelsoet, K.; Kaczmarek, A. M.;
Poelman, D.; Waroquier, M.; Van Speybroeck, V.; Van Deun, R.; Van
190
Der Voort, P. Bipyridine-Based Nanosized Metal–Organic Framework with
Tunable Luminescence by a Postmodification with Eu(III): An Experimental
and Theoretical Study. J. Phys. Chem. C 2013, 117, 11302–11310.
63. Qi, Z.-P.; Yang, J.-M.; Kang, Y.-S.; Guo, F.; Sun, W.-Y. Facile water-stability
evaluation of metal–organic frameworks and the property of selective removal
of dyes from aqueous solution. Dalton Trans. 2016, 45, 8753–8759.
64. Guo, H.; Zhu, Y.; Wang, S.; Su, S.; Zhou, L.; Zhang, H. Combining
coordination modulation with acid-base adjustment for the control over size
of metal-organic frameworks. Chem. Mater. 2012, 24, 444–450.
65. Pang, M.; Cairns, A. J.; Liu, Y.; Belmabkhout, Y.; Zeng, H. C.; Eddaoudi, M.
Synthesis and Integration of Fe-soc-MOF Cubes into Colloidosomes via a
Single-Step Emulsion-Based Approach. J. Am. Chem. Soc. 2013, 135, 10234–
10237.
66. Guo, C.; Zhang, Y.; Zhang, L.; Guo, Y.; Akram, N.; Wang, J. 2-
Methylimidazole-Assisted Synthesis of Nanosized Cu3(BTC)2 for Controlling
the Selectivity of the Catalytic Oxidation of Styrene. 2018, 1, 5289–5296.
67. Ranft, A.; Betzler, S. B.; Haase, F.; Lotsch, B. V. Additive-mediated size
control of MOF nanoparticles. CrystEngComm 2013, 15, 9296–9300.
68. Li, D.; Wang, H.; Zhang, X.; Sun, H.; Dai, X.; Yang, Y.; Ran, L.; Li, X.;
Ma, X.; Gao, D. Morphology Design of IRMOF - 3 Crystal by Coordination
Modulation. Cryst. Growth Des. 2014, 14, 5856–5864.
191
69. Bunzen, H.; Grzywa, M.; Hambach, M.; Spirkl, S.; Volkmer, D. From Micro to
Nano: A Toolbox for Tuning Crystal Size and Morphology of Benzotriazolate-
Based Metal–Organic Frameworks. Cryst. Growth Des. 2016, 16, 3190–3197.
70. Amaro-gahete, J.; Klee, R.; Esquivel, D.; Ruiz, J. R.; Jiménez-sanchidrián, C.;
Romero-salguero, F. J. Ultrasonics - Sonochemistry Fast ultrasound-assisted
synthesis of highly crystalline MIL-88A particles and their application as
ethylene adsorbents. Ultrasonics - Sonochemistry 2019, 50, 59–66.
71. Chalati, T.; Horcajada, P.; Couvreur, P.; Serre, C.; Ben Yahia, M.;
Maurin, G.; Gref, R. Porous metal organic framework nanoparticles to address
the challenges related to busulfan encapsulation. Nanomedicine 2011, 6,
1683–1695.
72. Cai, X.; Lin, J.; Pang, M. Facile Synthesis of Highly Uniform Fe-MIL-88B
Particles. Cryst. Growth Des. 2016, 16, 3565–3568.
73. Pham, M.; Vuong, G.; Vu, A.; Do, T. Novel Route to Size-Controlled Fe–MIL-
88B–NH2 Metal–Organic Framework Nanocrystals. Langmuir 2011, 27,
15261–15267.
74. Liu, D.; Liu, Y.; Dai, F.; Zhao, J.; Yang, K.; Liu, C. Size- and morphology-
controllable synthesis of MIL-96 (Al) by hydrolysis and coordination
modulation of dual aluminium source and ligand systems. Dalton Trans.
2015, 96, 16421–16429.
75. Márquez, A. G.; Demessence, A.; Platero-prats, A. E.; Heurtaux, D.;
Horcajada, P.; Serre, C.; Chang, J.; Férey, G.; De, A.; Boissière, C.;
Grosso, D.; Sanchez, C. Green Microwave Synthesis of MIL-100 (Al, Cr, Fe)
192
Nanoparticles for Thin- Film Elaboration. Eur. J. Inorg. Chem. 2012, 100,
5165–5174.
76. Anand, R.; Borghi, F.; Manoli, F.; Manet, I.; Agostoni, V.; Reschiglian, P.;
Gref, R.; Monti, S. Host–Guest Interactions in Fe(III)-Trimesate MOF
Nanoparticles Loaded with Doxorubicin. J. Phys. Chem. B 2014, 118, 8532–
8539.
77. Jiang, D.; Burrows, A. D.; Edler, K. J. Size-controlled synthesis of
MIL-101(Cr) nanoparticles with enhanced selectivity for CO2 over N2.
CrystEngComm 2011, 13, 6916.
78. Zhao, T.; Yang, L.; Feng, P.; Gruber, I.; Janiak, C.; Liu, Y. Facile synthesis
of nano-sized MIL-101(Cr) with the addition of acetic acid. Inorganica Chim.
Acta 2018, 471, 440–445.
79. Zhao, T.; Li, S.; Shen, L.; Wang, Y.; Yang, X. The sized controlled synthesis
of MIL-101 (Cr) with enhanced CO2 adsorption property. Inorg. Chem.
Commun. 2018, 96, 47–51.
80. Yang, L.; Zhao, T.; Boldog, I.; Janiak, C.; Yang, X.-Y.; Li, Q.; Zhou, Y.-J.;
Xia, Y.; Lai, D.-W.; Liu, Y.-J. Benzoic acid as a selector–modulator in the
synthesis of MIL-88B(Cr) and nano-MIL-101(Cr). Dalton Trans. 2019, 48,
989–996.
81. Guo, C.; Zhang, Y.; Guo, Y.; Zhang, Y.; Jide, W. A general and efficient
approach for tuning the crystal morphology of classical MOFs. Chem.
Commun. 2018, 54, 252–255.
193
82. Chi, W. S.; Roh, D. K.; Lee, C. S.; Kim, J. H. A shape- and morphology-
controlled metal organic framework template for high-efficiency solid-state
dye-sensitized solar cells. J. Mater. Chem. A 2015, 3, 21599–21608.
83. Vilela, S.; Salcedo-Abraira, P.; Colinet, I.; Salles, F.; de Koning, M.;
Joosen, M.; Serre, C.; Horcajada, P. Nanometric MIL-125-NH2 Metal–Organic
Framework as a Potential Nerve Agent Antidote Carrier. Nanomaterials 2017,
7, 321.
84. Nayuk, R.; Zacher, D.; Schweins, R.; Wiktor, C.; Fischer, R. A.; Van
Tendeloo, G.; Huber, K. Modulated formation of MOF-5 nanoparticles-A
SANS analysis. Journal of Physical Chemistry C 2012, 116, 6127–6135.
85. Huang, L.; Wang, H.; Chen, J.; Wang, Z.; Sun, J.; Zhao, D.; Yan, Y.
Synthesis, morphology control, and properties of porous metal–organic
coordination polymers. Microporous Mesoporous Mater. 2003, 58, 105–114.
86. Wang, S.; Lv, Y.; Yao, Y.; Yu, H.; Lu, G. Modulated synthesis of
monodisperse MOF-5 crystals with tunable sizes and shapes. Inorg. Chem.
Commun. 2018, 93, 56–60.
87. Ma, M.; Zacher, D.; Zhang, X.; Fischer, R. A.; Metzler-Nolte, N. A Method
for the Preparation of Highly Porous, Nanosized Crystals of Isoreticular
MetalOrganic Frameworks. Crystal Growth & Design 2011, 11, 185–189.
88. Nesse, F. S.; Opitz, M.; Dietzel, P. D. C. Morphology control in modulated
synthesis of metal-organic framework CPO-27. Microporous Mesoporous
Mater. 2019, 275, 207–213.
194
89. Li, P.; Klet, R. C.; Moon, S.-y.; Wang, T. C.; Deria, P.; Peters, A. W.;
Klahr, B. M.; Park, H.-j.; Al-juaid, S. S.; Hupp, J. T.; Farha, O. K. Synthesis
of nanocrystals of Zr-based metal – organic frameworks with csq-net :
significant. Chem. Comm. 2015, 51, 10925–10928.
90. Webber, T. E.; Liu, W.; Desai, S. P.; Lu, C. C.; Truhlar, D. G.; Penn, R. L.
Role of a Modulator in the Synthesis of Phase-Pure NU-1000. ACS Appl.
Mater. Interfaces 2017, 9, 39342–39346.
91. Wang, S.; Chen, Y.; Wang, S.; Li, P.; Mirkin, C. A.; Farha, O. K. DNA-
Functionalized Metal–Organic Framework Nanoparticles for Intracellular
Delivery of Proteins. J. Am. Chem. Soc. 2019, 141, 2215–2219.
92. Park, J.; Jiang, Q.; Feng, D.; Mao, L.; Zhou, H. Size-Controlled Synthesis of
Porphyrinic Metal Organic Framework and Functionalization for Targeted
Photodynamic Therapy. J. Am. Chem. Soc. 2016, 138, 3518–3525.
93. Morris, W.; Wang, S.; Cho, D.; Auyeung, E.; Li, P.; Farha, O. K.;
Mirkin, C. A. Role of modulators in controlling the colloidal stability and
polydispersity of the UiO-66 metal-organic framework. ACS Appl. Mater.
Interfaces 2017, 9, 33413–33418.
94. Schaate, A.; Roy, P.; Godt, A.; Lippke, J.; Waltz, F.; Wiebcke, M.;
Behrens, P. Modulated Synthesis of Zr-Based Metal-Organic Frameworks:
From Nano to Single Crystals. Chem. Eur. J. 2011, 17, 6643–6651.
95. Li, Y. S.; Bux, H.; Feldhoff, A.; Li, G. N.; Yang, W. S.; Caro, J. Controllable
synthesis of metal-organic frameworks: From MOF nanorods to oriented MOF
membranes. Adv. Mater. 2010, 22, 3322–3326.
195
96. Springer, S.; Satalov, A.; Lippke, J.; Wiebcke, M. Nanocrystals and
nanomaterials of isoreticular zeolitic imidazolate frameworks. Microporous
and Mesoporous Materials 2015-11, 216, 161–170.
97. Rojas, S.; Carmona, F. J.; Maldonado, C. R.; Horcajada, P.; Hidalgo, T.;
Serre, C.; Navarro, J. A.; Barea, E. Nanoscaled zinc pyrazolate metal-organic
frameworks as drug-delivery systems. Inorg. Chem. 2016, 55, 2650–2663.
98. Haque, E.; Jhung, S. H. Synthesis of isostructural metal–organic frameworks,
CPO-27s, with ultrasound, microwave, and conventional heating: Effect of
synthesis methods and metal ions. Chemical Engineering Journal 2011, 173,
866–872.
99. Surblé, S.; Millange, F.; Serre, C.; Férey, G.; Walton, R. I. An EXAFS study
of the formation of a nanoporous metal–organic framework: evidence for the
retention of secondary building units during synthesis. Chem. Commun. 2006,
1518.
100. Thi Thanh Chau, V.; Thi MinhThanh, H.; Dinh Du, P.; Thanh Tam Toan, T.;
Ngoc Tuyen, T.; Xuan Mau, T.; Quang Khieu, D. Metal-Organic Framework-
101 (MIL-101): Synthesis, Kinetics, Thermodynamics, and Equilibrium
Isotherms of Remazol Deep Black RGB Adsorption. Journal of Chemistry
2018, 2018, 1–14.
101. Jiang, D.; Burrows, A. D.; Edler, K. J. Size-controlled synthesis of
MIL-101(Cr) nanoparticles with enhanced selectivity for CO2 over N2.
CrystEngComm 2011, 13, 6916.
196
102. Trickett, C. A.; Gagnon, K. J.; Lee, S.; Gándara, F.; Bürgi, H.-B.;
Yaghi, O. M. Definitive Molecular Level Characterization of Defects in UiO-
66 Crystals. Angew. Chem. Int. Ed. 2015, 54, 11162–11167.
103. Park, J.; Jiang, Q.; Feng, D.; Mao, L.; Zhou, H.-C. Size-Controlled Synthesis
of Porphyrinic Metal–Organic Framework and Functionalization for Targeted
Photodynamic Therapy. J. Am. Chem. Soc. 2016, 138, 3518–3525.
104. Pham, M.-H.; Vuong, G.-T.; Vu, A.-T.; Do, T.-O. Novel Route to Size-
Controlled Fe–MIL-88B–NH2 Metal–Organic Framework Nanocrystals.
Langmuir 2011, 27, 15261–15267.
105. Morris, W.; Wang, S.; Cho, D.; Auyeung, E.; Li, P.; Farha, O. K.;
Mirkin, C. A. Role of Modulators in Controlling the Colloidal Stability and
Polydispersity of the UiO-66 Metal–Organic Framework. ACS Appl. Mater.
Interfaces 2017, 9, 33413–33418.
106. Cai, X.; Lin, J.; Pang, M. Facile Synthesis of Highly Uniform Fe-MIL-88B
Particles. Crystal Growth & Design 2016, 16, 3565–3568.
107. Zhao, T.; Yang, L.; Feng, P.; Gruber, I.; Janiak, C.; Liu, Y. Facile synthesis of
nano-sized MIL-101(Cr) with the addition of acetic acid. Inorganica Chimica
Acta 2018, 471, 440–445.
108. Suresh, M.; David Raju, B.; Rama Rao, K. S.; Raveendranath Reddy, K.;
Kantam, M. L.; Srinivasu, P. Metal organic framework MIL-101(Cr) for
dehydration reactions. J Chem Sci 2014, 126, 527–532.
109. Seoane, B.; Dikhtiarenko, A.; Mayoral, A.; Tellez, C.; Coronas, J.;
Kapteijn, F.; Gascon, J. Metal organic framework synthesis in the presence
197
of surfactants: towards hierarchical MOFs? CrystEngComm 2015, 17, 1693–
1700.
110. Garćıa Márquez, A.; Demessence, A.; Platero-Prats, A. E.; Heurtaux, D.;
Horcajada, P.; Serre, C.; Chang, J.-S.; Férey, G.; de la Peña-O’Shea, V. A.;
Boissière, C.; Grosso, D.; Sanchez, C. Green Microwave Synthesis of MIL-
100(Al, Cr, Fe) Nanoparticles for Thin-Film Elaboration. Eur. J. Inorg.
Chem. 2012, 2012, 5165–5174.
111. Lee, Y.-R.; Jang, M.-S.; Cho, H.-Y.; Kwon, H.-J.; Kim, S.; Ahn, W.-S. ZIF-8:
A comparison of synthesis methods. Chemical Engineering Journal 2015, 271,
276–280.
112. Decker, G. E.; Stillman, Z.; Attia, L.; Fromen, C. A.; Bloch, E. D. Controlling
Size, Defectiveness, and Fluorescence in Nanoparticle UiO-66 through Water
and Ligand Modulation. Chem. Mater. 2019, 31, 4831–4839.
113. Zareba, J. K.; Nyk, M.; Samoć, M. Co/ZIF-8 Heterometallic Nanoparticles:
Control of Nanocrystal Size and Properties by a Mixed-Metal Approach.
Crystal Growth & Design 2016, 16, 6419–6425.
114. Schejn, A.; Aboulaich, A.; Balan, L.; Falk, V.; Lalevée, J.; Medjahdi, G.;
Aranda, L.; Mozet, K.; Schneider, R. Cu2+ -doped zeolitic imidazolate
frameworks (ZIF-8): efficient and stable catalysts for cycloadditions and
condensation reactions. Catal. Sci. Technol. 2015, 5, 1829–1839.
115. Cao, J.; Yang, Z.-h.; Xiong, W.-p.; Zhou, Y.-y.; Peng, Y.-r.; Li, X.; Zhou, C.-
y.; Xu, R.; Zhang, Y.-r. One-step synthesis of Co-doped UiO-66 nanoparticle
198
with enhanced removal efficiency of tetracycline: Simultaneous adsorption and
photocatalysis. Chemical Engineering Journal 2018, 353, 126–137.
116. Hu, S.; Liu, M.; Guo, X.; Li, K.; Han, Y.; Song, C.; Zhang, G. Effects of
Monocarboxylic Acid Additives on Synthesizing Metal–Organic Framework
NH2 -MIL-125 with Controllable Size and Morphology. Crystal Growth &
Design 2017, 17, 6586–6595.
117. Yang, L.; Zhao, T.; Boldog, I.; Janiak, C.; Yang, X.-Y.; Li, Q.; Zhou, Y.-J.;
Xia, Y.; Lai, D.-W.; Liu, Y.-J. Benzoic acid as a selector–modulator in the
synthesis of MIL-88B(Cr) and nano-MIL-101(Cr). Dalton Trans. 2019, 48,
989–996.
118. Furukawa, H.; Gándara, F.; Zhang, Y.-B.; Jiang, J.; Queen, W. L.;
Hudson, M. R.; Yaghi, O. M. Water Adsorption in Porous Metal–Organic
Frameworks and Related Materials. J. Am. Chem. Soc. 2014, 136, 4369–4381.
119. Zhao, Y.; Zhang, Q.; Li, Y.; Zhang, R.; Lu, G. Large-Scale Synthesis of
Monodisperse UiO-66 Crystals with Tunable Sizes and Missing Linker Defects
via Acid/Base Co-Modulation. ACS Appl. Mater. Interfaces 2017, 9, 15079–
15085.
120. Hu, Z.; Castano, I.; Wang, S.; Wang, Y.; Peng, Y.; Qian, Y.; Chi, C.;
Wang, X.; Zhao, D. Modulator Effects on the Water-Based Synthesis of
Zr/Hf Metal–Organic Frameworks: Quantitative Relationship Studies between
Modulator, Synthetic Condition, and Performance. Crystal Growth & Design
2016, 16, 2295–2301.
199
121. Burton, A. W.; Ong, K.; Rea, T.; Chan, I. Y. On the estimation of average
crystallite size of zeolites from the Scherrer equation: A critical evaluation of
its application to zeolites with one-dimensional pore systems. Microporous and
Mesoporous Materials 2009, 117, 75–90.
122. Uemura, T.; Ohba, M.; Kitagawa, S. Size and Surface Effects of Prussian Blue
Nanoparticles Protected by Organic Polymers. Inorg. Chem. 2004-11, 43,
7339–7345.
123. Stetefeld, J.; McKenna, S. A.; Patel, T. R. Dynamic light scattering: a
practical guide and applications in biomedical sciences. Biophys Rev 2016-
12, 8, 409–427.
124. Mandø, M.; Yin, C.; Sørensen, H.; Rosendahl, L. On the modelling of motion
of non-spherical particles in two-phase flow. Powder Technology 2007, 15.
125. Abade, G. C.; Cichocki, B.; Ekiel-Jeżewska, M. L.; Nägele, G.; Wajnryb, E.
High-frequency viscosity and generalized Stokes–Einstein relations in dense
suspensions of porous particles. J. Phys.: Condens. Matter 2010, 22, 322101.
126. Chen, Z. H.; Kim, C.; Zeng, X.-b.; Hwang, S. H.; Jang, J.; Ungar, G.
Characterizing Size and Porosity of Hollow Nanoparticles: SAXS, SANS,
TEM, DLS, and Adsorption Isotherms Compared. Langmuir 2012, 28, 15350–
15361.
127. Bellido, E.; Guillevic, M.; Hidalgo, T.; Santander-Ortega, M. J.; Serre, C.;
Horcajada, P. Understanding the Colloidal Stability of the Mesoporous MIL-
100(Fe) Nanoparticles in Physiological Media. Langmuir 2014, 30, 5911–5920.
200
128. Li, T.; Senesi, A. J.; Lee, B. Small Angle X-ray Scattering for Nanoparticle
Research. Chem. Rev. 2016, 116, 11128–11180.
129. Lin, R.; Villacorta Hernandez, B.; Ge, L.; Zhu, Z. Metal organic framework
based mixed matrix membranes: an overview on filler/polymer interfaces. J.
Mater. Chem. A 2018, 6, 293–312.
130. Xiang, L.; Sheng, L.; Wang, C.; Zhang, L.; Pan, Y.; Li, Y. Amino-
Functionalized ZIF-7 Nanocrystals: Improved Intrinsic Separation Ability
and Interfacial Compatibility in Mixed-Matrix Membranes for CO2/CH4
Separation. Adv. Mater. 2017, 29, 1606999.
131. Della Rocca, J.; Liu, D.; Lin, W. Nanoscale Metal–Organic Frameworks for
Biomedical Imaging and Drug Delivery. Acc. Chem. Res. 2011, 44, 957–968.
132. Ettlinger, R.; Sönksen, M.; Graf, M.; Moreno, N.; Denysenko, D.; Volkmer, D.;
Kerl, K.; Bunzen, H. Metal–organic framework nanoparticles for arsenic
trioxide drug delivery. J. Mater. Chem. B 2018, 6, 6481–6489.
133. Ghorbanloo, M.; Safarifard, V.; Morsali, A. Heterogeneous catalysis with a
coordination modulation synthesized MOF: morphology-dependent catalytic
activity. New J. Chem. 2017, 41, 3957–3965.
134. LaMer, V. K.; Dinegar, R. H. Theory, Production and Mechanism of
Formation of Monodispersed Hydrosols. J. Am. Chem. Soc. 1950, 72, 4847–
4854.
135. Watzky, M. A.; Finke, R. G. Transition Metal Nanocluster Formation
Kinetic and Mechanistic Studies. A New Mechanism When Hydrogen Is the
201
Reductant: Slow, Continuous Nucleation and Fast Autocatalytic Surface
Growth. J. Am. Chem. Soc. 1997, 119, 10382–10400.
136. Lazzari, S.; Abolhasani, M.; Jensen, K. F. Modeling of the formation kinetics
and size distribution evolution of II–VI quantum dots. React. Chem. Eng.
2017, 2, 567–576.
137. Griffin, S.; Briuglia, M.; ter Horst, J.; Forgan, R. S. Assessing Crystallisation
Kinetics of Zr Metal-Organic Frameworks through Turbidity Measurements to
Inform Rapid Microwave-Assisted Synthesis. Chem. Eur. J. 2020,
138. Haouas, M.; Volkringer, C.; Loiseau, T.; Férey, G.; Taulelle, F. In Situ
NMR, Ex Situ XRD and SEM Study of the Hydrothermal Crystallization
of Nanoporous Aluminum Trimesates MIL-96, MIL-100, and MIL-110. Chem.
Mater. 2012, 24, 2462–2471.
139. Hu, S.; Liu, M.; Guo, X.; Li, K.; Han, Y.; Song, C.; Zhang, G. Effects of
monocarboxylic acid additives on synthesizing metal-organic framework NH2-
MIL-125 with controllable size and morphology. Cryst. Growth. Des. 2017, 17,
6586–6595.
140. Fishtik, I.; Nagypál, I.; Gutman, I. Response reactions: a way to explain the
unusual behaviour of multiple equilibrium systems. J. Chem. Soc., Faraday
Trans. 1995, 91, 259–267.
141. Smith, W.; Missen, R. Reaction Equilibrium Analysis: Theory and Algorithms ;
Wiley, 1982.
142. Avci, C.; Ariñez-Soriano, J.; Carné-Sánchez, A.; Guillerm, V.; Carbonell, C.;
Imaz, I.; Maspoch, D. Post-Synthetic Anisotropic Wet-Chemical Etching of
202
Colloidal Sodalite ZIF Crystals. Angew. Chem. Int. Ed. 2015-11-23, 54,
14417–14421.
143. Li, L.; Sun, F.; Jia, J.; Borjigin, T.; Zhu, G. Growth of large single MOF
crystals and effective separation of organic dyes. CrystEngComm 2013, 15,
4094.
144. Sanil, E. S.; Cho, K.-H.; Lee, S.-K.; Lee, U.-H.; Ryu, S. G.; Lee, H. W.;
Chang, J.-S.; Hwang, Y. K. Size and morphological control of a metal–organic
framework Cu-BTC by variation of solvent and modulator. J Porous Mater
2015, 22, 171–178.
145. Shan, B.; James, J. B.; Armstrong, M. R.; Close, E. C.; Letham, P. A.;
Nikkhah, K.; Lin, Y. S.; Mu, B. Influences of Deprotonation and Modulation
on Nucleation and Growth of UiO-66: Intergrowth and Orientation. J. Phys.
Chem. C 2018, 122, 2200–2206.
146. Chiu, C.-C.; Shieh, F.-K.; Tsai, H.-H. G. Ligand Exchange in the Synthesis
of Metal–Organic Frameworks Occurs Through Acid-Catalyzed Associative
Substitution. Inorg. Chem. 2019, 58, 14457–14466.
147. Qian, J.; Sun, F.; Qin, L. Hydrothermal synthesis of zeolitic imidazolate
framework-67 (ZIF-67) nanocrystals. Materials Letters 2012-09, 82, 220–223.
148. Saha, S.; Springer, S.; Schweinefuß, M. E.; Pontoni, D.; Wiebcke, M.;
Huber, K. Insight into Fast Nucleation and Growth of Zeolitic Imidazolate
Framework-71 by In Situ Time-Resolved Light and X-ray Scattering
Experiments. Crystal Growth & Design 2016, 16, 2002–2010.
203
149. Tu, M.; Wiktor, C.; Rösler, C.; Fischer, R. A. Rapid room temperature
syntheses of zeolitic-imidazolate framework (ZIF) nanocrystals. Chem.
Commun. 2014, 50, 13258–13260.
150. Webber, T. E.; Desai, S. P.; Combs, R. L.; Bingham, S.; Lu, C. C.;
Penn, R. L. Size Control of the MOF NU-1000 through Manipulation of the
Modulator/Linker Competition. Crystal Growth & Design 2020, 2965–2972.
151. Bosch, M.; Yuan, S.; Zhou, H.-C. In The Chemistry of Metal-Organic
Frameworks: Synthesis, Characterization, and Applications ; Kaskel, S., Ed.;
Wiley-VCH Verlag GmbH & Co. KGaA, 2016; pp 137–170.
152. Wang, X.-G.; Cheng, Q.; Yu, Y.; Zhang, X.-Z. Controlled Nucleation and
Controlled Growth for Size Predicable Synthesis of Nanoscale Metal-Organic
Frameworks (MOFs): A General and Scalable Approach. Angew. Chem. Int.
Ed. 2018, 57, 7836–7840.
153. Yeung, H. H.; Sapnik, A. F.; Massingberd-Mundy, F.; Gaultois, M. W.;
Wu, Y.; Fraser, D. A. X.; Henke, S.; Pallach, R.; Heidenreich, N.;
Magdysyuk, O. V.; Vo, N. T.; Goodwin, A. L. Control of Metal–Organic
Framework Crystallization by Metastable Intermediate Pre-equilibrium
Species. Angew. Chem. Int. Ed. 2019, 58, 566–571.
154. Lai, Y.-C.; Kung, C.-W.; Su, C.-H.; Ho, K.-C.; Liao, Y.-C.; Tsai, D.-H.
Metal–Organic Framework Colloids: Disassembly and Deaggregation.
Langmuir 2016, 32, 6123–6129.
155. Morris, W.; Wang, S.; Cho, D.; Auyeung, E.; Li, P.; Farha, O. K.;
Mirkin, C. A. Role of Modulators in Controlling the Colloidal Stability and
204
Polydispersity of the UiO-66 Metal–Organic Framework. ACS Appl. Mater.
Interfaces 2017, 9, 33413–33418.
156. Flory, P. J. Fundamental Principles of Condensation Polymerization. Chem.
Rev. 1946, 39, 137–197.
157. Hong, M.; Teng, H. H. Implications of solution chemistry effects: Direction-
specific restraints on the step kinetics of calcite growth. Geochimica et
Cosmochimica Acta 2014, 141, 228–239.
158. Wang, Y.; Li, L.; Liang, H.; Xing, Y.; Yan, L.; Dai, P.; Gu, X.; Zhao, G.;
Zhao, X. Superstructure of a Metal–Organic Framework Derived from
Microdroplet Flow Reaction: An Intermediate State of Crystallization by
Particle Attachment. ACS Nano 2019, 13, 2901–2912.
159. Shoaee, M.; Anderson, M.; Attfield, M. Crystal Growth of the Nanoporous
Metal-Organic Framework HKUST-1 Revealed by In Situ Atomic Force
Microscopy. Angew. Chem. Int. Ed. 2008, 47, 8525–8528.
160. Hellevang, H.; Haile, B. G.; Miri, R. A Statistical Approach To Explain the
Solution Stoichiometry Effect on Crystal Growth Rates. Crystal Growth &
Design 2016, 16, 1337–1348.
161. Feynman, C.; Gottlieb, M.; Leighton, R. The Feynman Lectures of Physics ;
California Institute of Technology, 1963.
162. Kim, T.-H.; Cho, K.-S.; Lee, E. K.; Lee, S. J.; Chae, J.; Kim, J. W.;
Kim, D. H.; Kwon, J.-Y.; Amaratunga, G.; Lee, S. Y.; Choi, B. L.; Kuk, Y.;
Kim, J. M.; Kim, K. Full-colour quantum dot displays fabricated by transfer
printing. Nature Photon 2011, 5, 176–182.
205
163. Cai, X.; Xie, Z.; Li, D.; Kassymova, M.; Zang, S.-Q.; Jiang, H.-L. Nano-
sized metal-organic frameworks: Synthesis and applications. Coordination
Chemistry Reviews 2020, 417, 213366.
164. Gao, R.; Zhang, G.; Ru, X.; Xu, C.; Li, M.; Lin, R.; Wang, Z. Morphology
control of metal-organic frameworks by Co-competitive coordination strategy
for low-temperature selective catalytic reduction of NO with NH3. Journal of
Solid State Chemistry 2021, 297, 122031.
165. Bonnett, B. L.; Smith, E. D.; De La Garza, M.; Cai, M.; Haag, J. V.;
Serrano, J. M.; Cornell, H. D.; Gibbons, B.; Martin, S. M.; Morris, A. J.
PCN-222 Metal–Organic Framework Nanoparticles with Tunable Pore Size
for Nanocomposite Reverse Osmosis Membranes. ACS Appl. Mater. Interfaces
2020, 12, 15765–15773.
166. Mandal, T. N.; Karmakar, A.; Sharma, S.; Ghosh, S. K. Metal-Organic
Frameworks (MOFs) as Functional Supramolecular Architectures for Anion
Recognition and Sensing. Chem. Rec. 2018, 18, 154–164.
167. Qian, Q.; Asinger, P. A.; Lee, M. J.; Han, G.; Mizrahi Rodriguez, K.; Lin, S.;
Benedetti, F. M.; Wu, A. X.; Chi, W. S.; Smith, Z. P. MOF-Based Membranes
for Gas Separations. Chem. Rev. 2020, acs.chemrev.0c00119.
168. Maka, V. K.; Mukhopadhyay, A.; Jindal, S.; Moorthy, J. N. Redox-
Reversible 2D Metal–Organic Framework Nanosheets (MONs) Based on the
Hydroquinone/Quinone Couple. Chem. Eur. J. 2019, 25, 3835–3842.
169. D’Alessandro, D. M. Exploiting redox activity in metal–organic frameworks:
concepts, trends and perspectives. Chem. Commun. 2016, 52, 8957–8971.
206
170. Cravillon, J.; Schröder, C. A.; Bux, H.; Rothkirch, A.; Caro, J.; Wiebcke, M.
Formate modulated solvothermal synthesis of ZIF-8 investigated using
time-resolved in situ X-ray diffraction and scanning electron microscopy.
CrystEngComm 2012, 14, 492–498.
171. Guo, C.; Zhang, Y.; Guo, Y.; Zhang, L.; Zhang, Y.; Wang, J. A general and
efficient approach for tuning the crystal morphology of classical MOFs. Chem.
Commun. 2018, 54, 252–255.
172. Manuel, D.; Mayoral, A.; Isabel, D.; Sa, M. Nanoscaled M - MOF-74 Materials
Prepared at Room Temperature. Cryst. Growth Des. 2014, 14, 2479–2487.
173. Xu, X. Controllable synthesis of ultra-small metal-organic framework
nanocrystals composed of copper(II) carboxylates. Nanoscale 2016, 8, 16725–
16732.
174. Zheng, W.; Hao, X.; Zhao, L.; Sun, W. Controllable Preparation of Nanoscale
Metal-Organic Frameworks by Ionic Liquid Microemulsions. Ind. Eng. Chem.
Res. 2017, 56, 5899–5905.
175. Shimpi, J. R.; Sidhaye, D. S.; Prasad, B. L. V. Digestive Ripening: A Fine
Chemical Machining Process on the Nanoscale. Langmuir 2017, 33, 9491–
9507.
176. Van Vleet, M. J.; Weng, T.; Li, X.; Schmidt, J. In Situ, Time-Resolved, and
Mechanistic Studies of Metal–Organic Framework Nucleation and Growth.
Chemical Reviews 2018, 118, 3681–3721.
207
177. Pan, Y.; Liu, Y.; Zeng, G.; Zhao, L.; Lai, Z. Rapid synthesis of zeolitic
imidazolate framework-8 (ZIF-8) nanocrystals in an aqueous system. Chem.
Commun. 2011, 47, 2071.
178. Cheng, C.; Zhang, R.; Wang, J.; Zhang, Y.; Wen, C.; Tan, Y.; Yang, M.
An ultrasensitive and selective fluorescent nanosensor based on porphyrinic
metal–organic framework nanoparticles for Cu2+ detection. Analyst 2020, 145,
797–804.
179. Razavi, S. A. A.; Masoomi, M. Y.; Morsali, A. Stimuli-Responsive
Metal–Organic Framework (MOF) with Chemo-Switchable Properties for
Colorimetric Detection of CHCl3. Chemistry – A European Journal 2017,
23, 12559–12564.
180. Wang, J.; Fan, Y.; Lee, H.-w.; Yi, C.; Cheng, C.; Zhao, X.; Yang, M.
Ultrasmall Metal–Organic Framework Zn-MOF-74 Nanodots: Size-Controlled
Synthesis and Application for Highly Selective Colorimetric Sensing of
Iron(III) in Aqueous Solution. ACS Appl. Nano Mater. 2018, 1, 3747–3753.
181. Sun, L.; Hendon, C. H.; Park, S. S.; Tulchinsky, Y.; Wan, R.; Wang, F.;
Walsh, A.; Dincă, M. Is iron unique in promoting electrical conductivity in
MOFs? Chem. Sci. 2017, 8, 4450–4457.
182. Norris, D. J.; Bawendi, M. G. Measurement and assignment of the size-
dependent optical spectrum in CdSe quantum dots. Phys. Rev. B 1996, 53,
16338–16346.
208
183. Siddiqui, S. A.; Domanov, O.; Schafler, E.; Vejpravova, J.; Shiozawa, H.
Synthesis and size-dependent spin crossover of coordination polymer
[Fe(Htrz)2(trz)](BF4). J. Mater. Chem. C 2021, 9, 1077–1084.
184. Hauser, A.; Guetlich, P.; Spiering, H. High-spin / low-spin relaxation
kinetics and cooperative effects in the hexakis(1-propyltetrazole)iron
bis(tetrafluoroborate) and [Zn1 − xFex(ptz)6](BF4)2 (ptz = 1-propyltetrazole)
spin-crossover systems. 1986, 25, 4245–4248.
185. Hauser, A. Intersystem crossing in the [Fe(ptz)6](BF4)2 spin crossover system
(ptz=1-propyltetrazole). The Journal of Chemical Physics 1991, 94, 2741–
2748.
186. Rajh, T.; Chen, L. X.; Lukas, K.; Liu, T.; Thurnauer, M. C.; Tiede, D. M.
Surface Restructuring of Nanoparticles: An Efficient Route for LigandMetal
Oxide Crosstalk. J. Phys. Chem. B 2002, 106, 10543–10552.
187. Dai, Q.; Wang, Y.; Li, X.; Zhang, Y.; Pellegrino, D. J.; Zhao, M.; Zou, B.;
Seo, J.; Wang, Y.; Yu, W. W. Size-Dependent Composition and Molar
Extinction Coefficient of PbSe Semiconductor Nanocrystals. ACS Nano 2009,
3, 1518–1524.
188. Liu, X.; Atwater, M.; Wang, J.; Huo, Q. Extinction coefficient of gold
nanoparticles with different sizes and different capping ligands. Colloids and
Surfaces B: Biointerfaces 2007, 58, 3–7.
189. Moreels, I.; Lambert, K.; De Muynck, D.; Vanhaecke, F.; Poelman, D.;
Martins, J. C.; Allan, G.; Hens, Z. Composition and Size-Dependent
Extinction Coefficient of Colloidal PbSe Quantum Dots. 2007, 19, 6101–6106.
209
190. Zhang, Y.; Wang, X.; Ma, M.; Fu, D.; Gu, N.; Lu, Z.; Xu, J.; Xu, L.; Chen, K.
Size dependence of second-order optical nonlinearity of CdS nanoparticles
studied by hyper-Rayleigh scattering. Journal of Colloid and Interface Science
2003, 266, 377–381.
191. Mayerhofer, T. G.; Hofer, S.; Popp, J. Deviations from Beer’s law on the
microscale – nonadditivity of absorption cross sections. Phys Chem Chem
Phys 2019, 21, 9793–9801.
192. Kornyshev, A. A.; Spohr, E.; Vorotyntsev, M. A. In Encyclopedia of
Electrochemistry ; Bard, A. J., Ed.; Wiley-VCH, 2007; p PAGES.
193. Reviakine, I.; Johannsmann, D.; Richter, R. P. Hearing What You Cannot See
and Visualizing What You Hear: Interpreting Quartz Crystal Microbalance
Data from Solvated Interfaces. Anal. Chem. 2011, 83, 8838–8848.
194. Celis-Salazar, P. J.; Cai, M.; Cucinell, C. A.; Ahrenholtz, S. R.; Epley, C. C.;
Usov, P. M.; Morris, A. J. Independent Quantification of Electron and Ion
Diffusion in Metallocene-Doped Metal–Organic Frameworks Thin Films. J.
Am. Chem. Soc. 2019, 141, 11947–11953.
195. Huang, Z.; Xie, N.; Zhang, M.; Xu, B. Nonpyrolyzed FeN Coordination-
Based Iron Triazolate Framework: An Efficient and Stable Electrocatalyst
for Oxygen Reduction Reaction. ChemSusChem 2019, 12, 200–207.
196. Okada, K.; Sawai, S.; Ikigaki, K.; Tokudome, Y.; Falcaro, P.; Takahashi, M.
Electrochemical sensing and catalysis using Cu3(BTC)2 coating electrodes
from Cu(OH)2 films. CrystEngComm 2017, 19, 4194–4200.
210
197. Kobayashi, H.; Tokita, Y. Modeling of hole transport across grain boundaries
in organic semiconductors for mesoscale simulations. 2015, 8, 051602.
198. Vladimirov, I.; Kühn, M.; Geßner, T.; May, F.; Weitz, R. T. Energy barriers
at grain boundaries dominate charge carrier transport in an electron-
conductive organic semiconductor. Sci Rep 2018, 8, 14868.
199. Horowitz, G.; Hajlaoui, M. E. Mobility in Polycrystalline Oligothiophene
Field-Effect Transistors Dependent on Grain Size. Advanced Materials 2000,
12, 1046–1050.
200. Chmiola, J.; Yushin, G.; Gogotsi, Y.; Portet, C.; Simon, P.; Taberna, P. L.
Anomalous Increase in Carbon Capacitance at Pore Sizes Less Than 1
Nanometer. 2006.
201. Boyd, S.; Ganeshan, K.; Tsai, W.-Y.; Wu, T.; Saeed, S.; Jiang, D.-e.;
Balke, N.; van Duin, A. C. T.; Augustyn, V. Effects of interlayer confinement
and hydration on capacitive charge storage in birnessite. 2021, 20, 1689–1694.
202. Cai, M.; Loague, Q.; Morris, A. J. Design Rules for Efficient Charge Transfer
in Metal–Organic Framework Films: The Pore Size Effect. J. Phys. Chem.
Lett. 2020-02-06, 11, 702–709.
203. Castner, A. T.; Su, H.; Svensson Grape, E.; Inge, A. K.; Johnson, B. A.;
Ahlquist, M. S. G.; Ott, S. Microscopic Insights into Cation-Coupled Electron
Hopping Transport in a Metal–Organic Framework. 2022, 144, 5910–5920.
204. Marshall, C. R.; Dvorak, J. P.; Twight, L. P.; Chen, L.; Kadota, K.;
Andreeva, A. B.; Overland, A. E.; Ericson, T.; Cozzolino, A. F.; Brozek, C. K.
211
Size-Dependent Properties of Solution-Processable Conductive MOF
Nanocrystals. 2022, 144, 5784–5794.
205. Saji, T.; Yamada, T.; Aoyagui, S. Electron-transfer rate constants for redox
systems of Fe(III)/Fe(II) complexes with 2,2-bipyridine and/or cyanide ion as
measured by the galvanostatic double pulse method. 1975, 61, 147–153.
206. Bondarenko, A. S.; Stephens, I. E. L.; Hansen, H. A.; Pérez-Alonso, F. J.;
Tripkovic, V.; Johansson, T. P.; Rossmeisl, J.; Nørskov, J. K.; Chorkendorff, I.
The Pt(111)/Electrolyte Interface under Oxygen Reduction Reaction
Conditions: An Electrochemical Impedance Spectroscopy Study. 2011, 27,
2058–2066.
207. Patwardhan, S.; Schatz, G. C. Theoretical Investigation of Charge Transfer
in Metal Organic Frameworks for Electrochemical Device Applications. 119,
24238–24247.
208. Sheberla, D.; Bachman, J. C.; Elias, J. S.; Sun, C.-J.; Shao-Horn, Y.;
Dincă, M. Conductive MOF electrodes for stable supercapacitors with high
areal capacitance. Nature Mater 2017-02, 16, 220–224.
209. Mendecki, L.; Mirica, K. A. Conductive Metal–Organic Frameworks as Ion-to-
Electron Transducers in Potentiometric Sensors. ACS Appl. Mater. Interfaces
2018-06-06, 10, 19248–19257.
210. Qi, Y.; He, C.-T.; Lin, J.; Lin, S.; Liu, J.; Huang, J.; Xue, W.; Yu, G.;
Chao, H.-Y.; Tong, Y.; Qiao, Z. Mild metal-organic-gel route for synthesis of
stable sub-5-nm metal-organic framework nanocrystals. Nano Research 2017,
10, 3621–3628.
212
211. Wang, X. G.; Cheng, Q.; Yu, Y.; Zhang, X. Z. Controlled Nucleation and
Controlled Growth for Size Predicable Synthesis of Nanoscale Metal–Organic
Frameworks (MOFs): A General and Scalable Approach. Angew. Chem. Int.
Ed. 2018, 57, 7836–7840.
212. Bachman, J. E.; Smith, Z. P.; Li, T.; Xu, T.; Long, J. R. Enhanced ethylene
separation and plasticization resistance in polymer membranes incorporating
metal-organic framework nanocrystals. Nat. Mater. 2016, 15, 845–849.
213. Abánades Lázaro, I.; Haddad, S.; Rodrigo-Muñoz, J. M.; Orellana-Tavra, C.;
Del Pozo, V.; Fairen-Jimenez, D.; Forgan, R. S. Mechanistic Investigation
into the Selective Anticancer Cytotoxicity and Immune System Response of
Surface-Functionalized, Dichloroacetate-Loaded, UiO-66 Nanoparticles. ACS
Appl. Mater. Interfaces 2018, 10, 5255–5268.
214. Chalati, T.; Horcajada, P.; Gref, R.; Couvreur, P.; Serre, C. Optimisation of
the synthesis of MOF nanoparticles made of flexible porous iron fumerate
MIL-88A. J. Mater. Chem. 2011, 21, 2220–2227.
215. Grall, R.; Hidalgo, T.; Delic, J.; Garcia-Marquez, A.; Chevillard, S.;
Horcajada, P. In vitro biocompatibility of mesoporous metal (III; Fe, Al, Cr)
trimesate MOF nanocarriers. J. Mater. Chem. B 2015, 3, 8279–8292.
216. Saouma, C. T.; Richard, S.; Smolders, S.; Delley, M. F.; Ameloot, R.;
Vermoortele, F.; De Vos, D. E.; Mayer, J. M. Bulk-to-Surface Proton-Coupled
Electron Transfer Reactivity of the Metal-Organic Framework MIL-125. J.
Am. Chem. Soc. 2018, 140, 16184–16189.
213
217. Harvey, S. P.; Hupp, J. T.; Farha, O. K. Nanosizing a Metal Organic
Framework Enzyme Carrier for Accelerating Nerve Agent Hydrolysis. ACS
Nano 2016, 10, 9174–9182.
218. Schaate, A.; Roy, P.; Godt, A.; Lippke, J.; Waltz, F.; Wiebcke, M.;
Behrens, P. Modulated Synthesis of Zr-Based Metal–Organic Frameworks:
From Nano to Single Crystals. Chem. Eur. J. 2011, 17, 6643–6651.
219. Cravillon, J.; Nayuk, R.; Springer, S.; Feldhoff, A.; Huber, K.; Wiebcke, M.
Controlling zeolitic imidazolate framework nano- and microcrystal formation:
Insight into crystal growth by time-resolved in situ static light scattering.
Chem. Mater. 2011, 23, 2130–2141.
220. Lim, I. H.; Schrader, W.; Schüth, F. Insights into the molecular assembly of
zeolitic imidazolate frameworks by ESI-MS. Chem. Mater. 2015, 27, 3088–
3095.
221. Chouhan, A.; Pilet, G.; Daniele, S.; Pandey, A. Shape Controllable
Preparation of Submicronic Cadmium Tetrazole-Based Metal–Organic
Frameworks via Solvothermal or Microwave-Assisted Methods and Their
Photocatalytic Studies. Chinese J. Chem. 2017, 35, 209–216.
222. Enŕıquez, J.; Manquian, C.; Chi-Duran, I.; Herrera, F.; Singh, D. P.
Controlled Growth of the Noncentrosymmetric Zn(3-ptz)2 and Zn(OH)(3-ptz)
Metal-Organic Frameworks. ACS Omega 2019, 4, 7411–7419.
223. Liu, Y.-y.; Decadt, R.; Bogaerts, T.; Hemelsoet, K.; Kaczmarek, A. M.;
Poelman, D.; Waroquier, M.; Speybroeck, V. V.; Deun, R. V.; Voort, P.
V. D. Bipyridine-Based Nanosized Metal Organic Framework with Tunable
214
Luminescence by a Postmodification with Eu (III): An Experimental and
Theoretical Study. J. Phys. Chem. C 2013, 117, 11302–11310.
224. Qi, Z.; Yang, J.; Kang, Y.; Guo, F.; Sun, W. Facile water-stability evaluation
of metal–organic frameworks and the property of selective removal of dyes
from aqueous solution. Dalton Trans. 2016, 45, 8753–8759.
225. Klein, N.; Senkovska, I.; Baburin, I. A.; Grünker, R.; Stoeck, U.;
Schlichtenmayer, M.; Streppel, B.; Mueller, U.; Leoni, S.; Hirscher, M.;
Kaskel, S. Route to a family of robust, non-interpenetrated metal-organic
frameworks with pto-like topology. Chem. Eur. J. 2011, 17, 13007–13016.
226. Zhou, Q.; Yang, F.; Xin, B.; Zeng, G.; Zhou, X.; Liu, K.; Ma, D.; Li, G.;
Shi, Z.; Feng, S. Reversible switching of slow magnetic relaxation in a classic
lanthanide metal-organic framework system. Chem. Commun. 2013, 49, 8244–
8246.
227. Cai, X.; Liu, B.; Pang, M.; Lin, J. Interfacially synthesized Fe-soc-MOF
nanoparticles combined with ICG for photothermal/photodynamic therapy.
Dalton Trans. 2018, 47, 16329–16336.
228. Cai, X.; Xie, Z.; Pang, M.; Lin, J. Controllable Synthesis of Highly Uniform
Nanosized HKUST-1 Crystals by Liquid–Solid–Solution Method. Cryst.
Growth Des. 2019, 19, 556–561.
229. Xiang, S.; Zhou, W.; Gallegos, J. M.; Liu, Y.; Chen, B. Exceptionally high
acetylene uptake in a microporous metal-organic framework with open metal
sites. J. Am. Chem. Soc. 2009, 131, 12415–12419.
215
230. Wang, F.; Guo, H.; Chai, Y.; Li, Y.; Liu, C. The controlled regulation of
morphology and size of HKUST-1 by “coordination modulation method”.
Microporous Mesoporous Mater. 2013, 173, 181–188.
231. Zhuang, J. L.; Ceglarek, D.; Pethuraj, S.; Terfort, A. Rapid room-temperature
synthesis of metal-organic framework HKUST-1 crystals in bulk and as
oriented and patterned thin films. Adv. Funct. Mater. 2011, 21, 1442–1447.
232. Diring, S.; Furukawa, S.; Takashima, Y.; Tsuruoka, T.; Kitagawa, S.
Controlled multiscale synthesis of porous coordination polymer in nano/micro
regimes. Chem. Mater. 2010, 22, 4531–4538.
233. Xin, C.; Zhan, H.; Huang, X.; Li, H.; Zhao, N.; Xiao, F.; Wei, W.; Sun, Y.
Effect of various alkaline agents on the size and morphology of nano-sized
HKUST-1 for CO2 adsorption. RSC Advances 2015, 5, 27901–27911.
234. Abdelhameed, R. M.; Carlos, L. D.; Silva, A. M.; Rocha, J. Near-infrared
emitters based on post-synthetic modified Ln3+-IRMOF-3. Chem. Commun.
2013, 49, 5019–5021.
235. Biswas, S.; Grzywa, M.; Nayek, H. P.; Dehnen, S.; Senkovska, I.;
Kaskel, S.; Volkmer, D. A cubic coordination framework constructed from
benzobistriazolate ligands and zinc ions having selective gas sorption
properties. Dalton Trans. 2009, 9226, 6487–6495.
236. Denysenko, D.; Grzywa, M.; Tonigold, M.; Streppel, B.; Krkljus, I.;
Hirscher, M.; Mugnaioli, E.; Kolb, U.; Hanss, J.; Volkmer, D. Elucidating
gating effects for hydrogen sorption in MFU-4-type triazolate-based metal-
216
organic frameworks featuring different pore sizes. Chem. Eur. J. 2011, 17,
1837–1848.
237. Volkringer, C.; Popov, D.; Loiseau, T.; Férey, G.; Burghammer, M.; Riekel, C.;
Haouas, M.; Taulelle, F. Synthesis, single-crystal X-ray microdiffraction, and
NMR characterizations of the giant pore metal-organic framework aluminum
trimesate MIL-100. Chem. Mater. 2009, 21, 5695–5697.
238. Qiu, M.; Chen, C.; Li, W. Rapid controllable synthesis of Al-MIL-96 and its
adsorption of nitrogenous VOCs. Catalysis Today 2015, 258, 132–138.
239. Férey, G.; Serre, C.; Mellot-Draznieks, C.; Millange, F.; Surblé, S.; Dutour, J.;
Margiolaki, I. A hybrid solid with giant pores prepared by a combination of
targeted chemistry, simulation, and powder diffraction. Angew. Chem. Int. Ed.
2004, 43, 6296–6301.
240. Seo, Y. K.; Yoon, J. W.; Lee, J. S.; Lee, U. H.; Hwang, Y. K.; Jun, C. H.;
Horcajada, P.; Serre, C.; Chang, J. S. Large scale fluorine-free synthesis of
hierarchically porous iron(III) trimesate MIL-100(Fe) with a zeolite MTN
topology. Microporous Mesoporous Mater. 2012, 157, 137–145.
241. Suresh, M.; David Raju, B.; Rama Rao, K. S.; Raveendranath Reddy, K.;
Kantam, M. L.; Srinvasu, P. Metal organic framework MIL-101(Cr) for
dehydration reactions. J. of Chem. Sci. 2014, 126, 527–532.
242. Yang, L.; Zhao, T.; Boldog, I.; Janiak, C.; Yang, X.; Li, Q.; Zhou, Y.; Xia, Y.;
Lai, D.; Liu, Y. Benzoic acid as a selector–modulator in the synthesis of MIL-
88B(Cr) and nano-MIL-101(Cr). Dalton Trans. 2019, 48, 989–996.
217
243. Wang, J.; Wan, J.; Ma, Y.; Wang, Y.; Pu, M.; Guan, Z. Metal-organic
frameworks MIL-88A with suitable synthesis conditions and optimal dosage
for effective catalytic degradation of Orange G through persulfate activation.
RSC Advances 2016, 6, 112502–112511.
244. Shih, Y. H.; Lo, S. H.; Yang, N. S.; Singco, B.; Cheng, Y. J.; Wu, C. Y.;
Chang, I. H.; Huang, H. Y.; Lin, C. H. Trypsin-immobilized metal-organic
framework as a biocatalyst in proteomics analysis. ChemPlusChem 2012, 77,
982–986.
245. Ma, M.; Bétard, A.; Weber, I.; Al-Hokbany, N. S.; Fischer, R. A.; Metzler-
Nolte, N. Iron-based metal-organic frameworks MIL-88B and NH2-MIL-88B:
High quality microwave synthesis and solvent-induced lattice ”breathing”.
Cryst. Growth. Des. 2013, 13, 2286–2291.
246. Kim, S.; Kim, J.; Kim, H.; Cho, H.; Ahn, W. Adsorption / catalytic properties
of MIL-125 and NH2-MIL-125. Catalysis Today 2013, 204, 85–93.
247. Liu, Y.; Zhao, X. Impact of moderative ligand hydrolysis on morphology
evolution and the morphology-dependent breathing effect performance of MIL-
53(Al). CrystEngComm 2018, 20, 2102–2111.
248. Xiao, Y.; Han, T.; Xiao, G.; Ying, Y.; Huang, H.; Yang, Q.; Liu, D.; Zhong, C.
Highly selective adsorption and separation of aniline/phenol from aqueous
solutions by microporous MIL-53(Al): A combined experimental and
computational study. Langmuir 2014, 30, 12229–12235.
249. Han, S.; Lah, M. S. Simple and Efficient Regeneration of MOF-5 and HKUST-
1 via Acid-Base Treatment. Cryst. Growth. Des. 2015, 15, 5568–5572.
218
250. Choi, J. S.; Son, W. J.; Kim, J.; Ahn, W. S. Metal-organic framework MOF-
5 prepared by microwave heating: Factors to be considered. Microporous
Mesoporous Mater. 2008, 116, 727–731.
251. Burgaz, E.; Erciyes, A.; Andac, M.; Andac, O. Synthesis and characterization
of nano-sized metal organic framework-5 (MOF-5) by using consecutive
combination of ultrasound and microwave irradiation methods. Inorganica
Chim. Acta 2019, 485, 118–124.
252. Ma, M.; Zacher, D.; Zhang, X.; Fischer, R. A.; Metzler-Nolte, N. A Method
for the Preparation of Highly Porous, Nanosized Crystals of Isoreticular
MetalOrganic Frameworks. Cryst. Growth Des. 2011, 11, 185–189.
253. Huang, L.; Wang, H.; Chen, J.; Wang, Z.; Sun, J.; Zhao, D.; Yan, Y.
Synthesis, morphology control, and properties of porous metal–organic
coordination polymers. Microporous Mesoporous Mater. 2003, 58, 105–114.
254. Islamoglu, T.; Otake, K. I.; Li, P.; Buru, C. T.; Peters, A. W.; Akpinar, I.;
Garibay, S. J.; Farha, O. K. Revisiting the structural homogeneity of NU-
1000, a Zr-based metal-organic framework. CrystEngComm 2018, 20, 5913–
5918.
255. Pankajakshan, A.; Sinha, M.; Ojha, A. A.; Mandal, S. Water-Stable Nanoscale
Zirconium-Based Metal–Organic Frameworks for the Effective Removal of
Glyphosate from Aqueous Media. ACS Omega 2018, 3, 7832–7839.
256. Zhang, G. Y.; Zhuang, Y. H.; Shan, D.; Su, G. F.; Cosnier, S.; Zhang, X. J.
Zirconium-based porphyrinic metal-organic framework (PCN-222):
219
Enhanced photoelectrochemical response and its application for label-free
phosphoprotein detection. Anal. Chem. 2016, 88, 11207–11212.
257. Decker, G. E.; Stillman, Z.; Attia, L.; Fromen, C. A.; Bloch, E. D. Controlling
Size, Defectiveness, and Fluorescence in Nanoparticle UiO-66 through Water
and Ligand Modulation. Chem. Mater. 2019, 31, 4831–4839.
258. Cavka, J. H.; Jakobsen, S.; Olsbye, U.; Guillou, N.; Lamberti, C.; Bordiga, S.;
Lillerud, K. P. A New Zirconium Inorganic Building Brick Forming Metal
Organic Frameworks with Exceptional Stability. J. Am. Chem. Soc. 2008,
130, 13850–13851.
259. Zhao, Y.; Zhang, Q.; Li, Y.; Zhang, R.; Lu, G. Large-Scale Synthesis of
Monodisperse UiO-66 Crystals with Tunable Sizes and Missing Linker Defects
via Acid/Base Co-Modulation. ACS Appl. Mater. Interfaces 2017, 9, 15079–
15085.
260. Lozano, L. A.; Faroldi, B. M. C. Efficient solvothermal synthesis of highly
porous UiO-66 nanocrystals in dimethylformamide-free media. J. Mater. Sci.
2018, 1862–1873.
261. Han, Y.; Liu, M.; Li, K.; Sun, Q.; Song, C.; Zhang, G.; Zhang, Z.; Guo, X.
Cu2O Mediated Synthesis of MetalOrganic Framework UiO-66 in Nanometer
Scale. Cryst. Growth Des. 2017, 17, 685–692.
262. Cao, J.; Yang, Z.; Xiong, W.; Zhou, Y.; Peng, Y.; Li, X. One-step synthesis
of Co-doped UiO-66 nanoparticle with enhanced removal efficiency of
tetracycline: Simultaneous adsorption and photocatalysis. Chem. Eng. J.
2018, 353, 126–137.
220
263. Park, T. H.; Cychosz, K. A.; Wong-Foy, A. G.; Dailly, A.; Matzger, A. J.
Gas and liquid phase adsorption in isostructural Cu3[biaryltricarboxylate]2
microporous coordination polymers. Chem. Commun. 2011, 47, 1452–1454.
264. Lan, X.; Huang, N.; Wang, J.; Wang, T. A general and facile strategy
for precisely controlling the crystal size of monodispersed metal-organic
frameworks: Via separating the nucleation and growth. Chem. Commun.
2018, 54, 584–587.
265. Zarȩba, J. K.; Nyk, M.; Samoć, M. Co/ZIF-8 Heterometallic Nanoparticles:
Control of Nanocrystal Size and Properties by a Mixed-Metal Approach.
Cryst. Growth Des. 2016, 16, 6419–6425.
266. Tu, M.; Wiktor, C.; Ro, C.; Fischer, R. A. Rapid room temperature syntheses
of zeolitic-imidazolate framework (ZIF) nanocrystals. Chem. Commun. 2014,
50, 13258–13260.
267. Orsi, A.; Price, D. J.; Kahr, J.; Pillai, R. S.; Sneddon, S.; Cao, S.; Benoit, V.;
L ozińska, M. M.; Cordes, D. B.; Slawin, A. M.; Llewellyn, P. L.; Casely, I.;
Ashbrook, S. E.; Maurin, G.; Wright, P. A. Porous zinc and cobalt 2-
nitroimidazolate frameworks with six-membered ring windows and a layered
cobalt 2-nitroimidazolate polymorph. CrystEngComm 2017, 19, 1377–1388.
268. Zhao, P.; Lampronti, G. I.; Lloyd, G. O.; Wharmby, M. T.; Facq, S.;
Cheetham, A. K.; Redfern, S. A. Phase transitions in zeolitic imidazolate
framework 7: The importance of framework flexibility and guest-induced
instability. Chem. Mater. 2014, 26, 1767–1769.
221
269. Lively, R. P.; Dose, M. E.; Thompson, J. A.; McCool, B. A.; Chance, R. R.;
Koros, W. J. Ethanol and water adsorption in methanol-derived ZIF-71.
Chem. Commun. 2011, 47, 8667–8669.
270. Cravillon, J.; Münzer, S.; Lohmeier, S.; Feldhoff, A.; Huber, K. Rapid Room-
Temperature Synthesis and Characterization of Nanocrystals of a Prototypical
Zeolitic Imidazolate Framework- Supporting Information. Chem. Mater. 2009,
21, 1–21.
271. Park, K. S.; Ni, Z.; Cote, A. P.; Choi, J. Y.; Huang, R.; Uribe-Romo, F. J.;
Chae, H. K.; O’Keeffe, M.; Yaghi, O. M. ZIFs - first synthesis. P. Natl. Acad.
Sci. 2006, 103, 10186–10191.
272. Colombo, V.; Montoro, C.; Maspero, A.; Palmisano, G.; Masciocchi, N.;
Galli, S.; Barea, E.; Navarro, J. A. Tuning the adsorption properties of
isoreticular pyrazolate-based metal-organic frameworks through ligand
modification. J. Am. Chem. Soc. 2012, 134, 12830–12843.
273. He, T.; Xu, X.; Ni, B.; Wang, H.; Long, Y.; Hu, W.; Wang, X. Fast and
scalable synthesis of uniform zirconium-, hafnium-based metal–organic
framework nanocrystals. Nanoscale 2017, 19209–19215.
274. Wang, F.; Guo, H.; Chai, Y.; Li, Y.; Liu, C. The controlled regulation of
morphology and size of HKUST-1 by “coordination modulation method”.
Microporous Mesoporous Mater. 2013, 173, 181–188.
275. Bagherzadeh, E.; Zebarjad, S. M.; Hosseini, H. R. M. Morphology
modification of the iron fumarate MIL-88A metal–organic framework using
222
formic acid and acetic acid as modulators. Eur. J. Inorg. Chem. 2018, 2018,
1909–1915.
276. Bellido, E.; Hidalgo, T.; Lozano, M. V.; Guillevic, M.; Simón-Vázquez, R.;
Santander-Ortega, M. J.; González-Fernández, Á.; Serre, C.; Alonso, M. J.;
Horcajada, P. Heparin-Engineered Mesoporous Iron Metal-Organic Framework
Nanoparticles: Toward Stealth Drug Nanocarriers. Adv. Healthc. Mater. 2015,
4, 1246–1257.
277. Xiang, L.; Sheng, L.; Wang, C.; Zhang, L.; Pan, Y.; Li, Y. Amino-
Functionalized ZIF-7 Nanocrystals: Improved Intrinsic Separation Ability
and Interfacial Compatibility in Mixed-Matrix Membranes for CO2/CH4
Separation. Adv. Mater. 2017, 29, 1–8.
278. Xing, T.; Lou, Y.; Bao, Q.; Chen, J. Surfactant-assisted synthesis of ZIF-8
nanocrystals in aqueous solution via microwave irradiation. CrystEngComm
2014, 16, 8994–9000.
279. Springer, S.; Satalov, A.; Lippke, J.; Wiebcke, M. Nanocrystals and
nanomaterials of isoreticular zeolitic imidazolate frameworks. Microporous
Mesoporous Mater. 2015, 216, 161–170.
280. Toby, B. H.; Von Dreele, R. B. GSAS-II: the genesis of a modern open-
source all purpose crystallography software package. 46, 544–549, eprint:
https://onlinelibrary.wiley.com/doi/pdf/10.1107/S0021889813003531.
281. Vazquez-Molina, D. A.; Mohammad-Pour, G. S.; Lee, C.; Logan, M. W.;
Duan, X.; Harper, J. K.; Uribe-Romo, F. J. Mechanically Shaped Two-
223
Dimensional Covalent Organic Frameworks Reveal Crystallographic Alignment
and Fast Li-Ion Conductivity. 138, 9767–9770.
224