Characterizing Defects in a UiO-AZB Metal-Organic Framework

Article abstract of DOI:10.1021/acs.inorgchem.7b01801

Exploring defect sites in metal-organic framework materials has quickly become an interesting topic of discussion in the literature. With reports of the enhancement of material properties with increasing defect sites, we were interested in probing the defect nature of UiO-AZB (UiO = University of Oslo, AZB = 4,4′-azobenzenedicarboxylate) nanoparticles. In this report, we investigate the use of acetic, formic, and benzoic acids as the modulators to prepare UiO-AZB. The results of ¹H NMR techniques and BET surface area analysis elucidate the extent of defects in our samples and are provided along with detailed discussions of the observed experimental trends. Interestingly, formic acid samples resulted in the most defected structure, reaching 36%. Additionally, for benzoic acid samples, with a 33% defect level, a drastic reduction in the accessible SA from 2682 m²/g to as low as 903 m²/g was observed, as the concentration of benzoic acid was increased. This was attributed to the creation of macropores in the individual crystallites and confirmed by average pore width analysis.

Full text of DOI:10.1021/acs.inorgchem.7b01801

Article
pubs.acs.org/IC
Cite This: Inorg. Chem. XXXX, XXX, XXX-XXX
Characterizing Defects in a UiO-AZB Metal−Organic Framework
Charity C. Epley, Madeline D. Love, and Amanda J. Morris*
Department of Chemistry and Macromolecules Innovation Institute, Virginia Tech, Blacksburg, Virginia 24061, United States
S
* Supporting Information
ABSTRACT: Exploring defect sites in metal−organic framework materials
has quickly become an interesting topic of discussion in the literature. With
reports of the enhancement of material properties with increasing defect sites,
we were interested in probing the defect nature of UiO-AZB (UiO =
University of Oslo, AZB = 4,4′-azobenzenedicarboxylate) nanoparticles. In
this report, we investigate the use of acetic, formic, and benzoic acids as the
1
modulators to prepare UiO-AZB. The results of H NMR techniques and
BET surface area analysis elucidate the extent of defects in our samples and
are provided along with detailed discussions of the observed experimental
trends. Interestingly, formic acid samples resulted in the most defected
structure, reaching 36%. Additionally, for benzoic acid samples, with a 33%
defect level, a drastic reduction in the accessible SA from 2682 m2/g to as low
as 903 m2/g was observed, as the concentration of benzoic acid was increased.
This was attributed to the creation of macropores in the individual crystallites and confirmed by average pore width analysis.
methods, mostly for UiO-66.17,21−26 Structural evidence
INTRODUCTION
■
quantifying missing linkers in UiO-66 came from a report by
Metal−organic framework (MOF) research has flourished over
the past several decades. MOFs are assemblies of metal clusters
Zhou et al.,25 where they used neutron diffraction of samples
synthesized with varying concentrations of acetic acid
linked through multidentate organic molecules and have been
modulator. They reported an average of 1 in every 12 linkers
proposed for a variety of applications from gas storage/
to be missing, bringing network connectivity to 11 instead of
separation1,2 to toxin sequestration3−5 and drug delivery.6−8
the 12 expected for an ideal UiO-66 sample. Farha10,27 and co-
Many structural motifs have been discovered and extensively
characterized over the years; however, the crystalline nature of
some MOFs renders them susceptible to bulk crystallographic
workers identified a maximum of 1.8 missing linkers for UiO-66
(Hf)10 using titration techniques to evaluate defects in several
UiO materials. Thermogravimetric analysis (TGA) is another
defects, typically not distinguished with general single crystal
technique reported to quantify defects in UiO-66,17,22 and
diffraction. While defective materials may intuitively be
considered undesirable, crystalline defects in MOFs have
using this method, it has been shown that different batches of
been shown to increase catalytic activity,9−11 enhance
UiO-66 MOF prepared by a single synthesis can result in
conductivity,12,13 and improve adsorption14−16 of the bulk
material. These material augmentations have led to increased
interest and debate regarding the essence of defects and the
materials with drastically different degrees of missing linkers
(up to 6 missing linkers per the 12 expected in an ideal UiO-
66).17
1H NMR techniques have also been used to elucidate the
best approaches of how to control and quantify them, especially
quantity of missing linker defects in UiO MOFs.21,22 Samples
concerning the University of Oslo (UiO) series of MOFs.
UiO materials are specifically attractive for many applications
due to the reports of exceptional chemical, thermal, and
are digested in a deuterated solvent, and the integration
corresponding to linker and modulator peaks present in the
spectra are compared to find the modulator: linker ratio
mechanical stability of UiO-66 (containing the 1,4-benzene-
dicarboxylate linker, BDC).17 The use of “modulators” in the
(discussed further in the results and discussion section). Using,
this technique, Lillerud et al.22 have reported ratios ranging
synthesis of UiO materials, to form octahedral Zr6O4(OH)4
secondary building units (SBUs), renders these materials
disposed to increased incidences of crystalline defects.
Modulators serve not only to form the SBUs but also to
competitively bind with the chosen linker and direct MOF
crystal growth. Therefore, in the final bulk material, the
from <0.1 to ∼0.8 for UiO-66 materials synthesized using acetic
acid and trifluoroacetic acid, respectively. A value of 0.8
corresponds to an impressive five missing linkers per cluster
and supports the assertion that UiO-66 can accommodate high
levels of missing linkers. They were also able to demonstrate an
increase in this ratio with increasing concentrations of
modulator may remain attached to the SBU, resulting in
missing linkers. Additionally, it has been proposed that OH−,
modulator, indicating that higher levels of modulator used,
H2O, and Cl− can also occupy missing linker sites.10,11,16,18−20
Several reports in the literature have attempted to describe
the quantity of defect sites in bulk materials using various
Received: July 17, 2017
© XXXX American Chemical Society
A
DOI: 10.1021/acs.inorgchem.7b01801
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
equivalents of modulator (x mol mod.: 1 mol Zr4+) to a glass vial. The
mixture was sonicated for ∼1 min and then placed into a 120 °C oven
for 24 h. The resultant powder was collected via centrifugation and
washed with DMF. All materials were then flushed of DMF solvent by
soaking in EtOH with fresh EtOH replacement every hour, for a total
5 h (5 EtOH soaking cycles). Materials were then evacuated in a
vacuum oven at 50 °C for 24 h. Evacuations at temperatures beyond
this resulted in loss of crystallinity.
Powder X-ray Diffraction (PXRD): For X-ray diffraction analysis, a
600 W Rigaku MiniFlex powder diffractometer operating with a Cu
(Kα = 0.15418 nm) radiation source was used, with a sweeping range
of 3°−50° in continuous scanning mode. Data were collected in 0.1°
increments at a scanning rate of 1°/min, and patterns were generated
with PDXL software.
Scanning Electron Microscopy (SEM): A Leo (Zeiss) 1550 field-
emission scanning electron microscope, equipped with an in-lens
detector and operating at 5.0 kV, was used to collect SEM images. NPs
were prepared for SEM imaging by first dispersing them in ethanol
(0.1 mg/mL) via sonication. The resulting solution was then deposited
via drop casting, with a short stem Pasteur pipet (1 drop), onto precut
5 mm × 5 mm glass slides. After room temperature drying, slides were
mounted onto an SEM sample peg using double sided Cu tape and the
sides of the slides were painted with conductive carbon paint. A
Cressington 208 High Resolution Sputter Coater with a Au/Pd target
(80/20) was used to deposit a conductive layer on the surface of the
slides in order to allow for imaging. Samples were coated for 60 s.
Thermogravimetric Analysis (TGA): A Q-series TGA from TA
Instruments was used to analyze the thermal stability of materials. A 10
mg amount of sample in a high temperature platinum pan was heated
under air from 25 °C up to 1000 °C with a heating rate of 10 °C/min.
Gas Sorption Isotherms: The N2 sorption isotherm measurements
were collected on a Quantachrome Autosorb-1 at 77 K. The samples
were placed in a 6 mm large bulb sample cell, which was degassed
under vacuum for 24 h at room temperature. The surface areas of the
materials were determined by fitting the adsorption data within the P/
P0 pressure range appropriate to the BET equation, ensuring that the
y-intercept of the linear BET plot was positive.
resulted in more missing linkers under their synthetic
conditions. For example, when formic acid concentrations
were increased from 6 to 100 mol equiv (mol modulator: mol
Zr4+), their integration ratios ranged from ∼0.1 to 0.35.
Although the point vacancy defects (missing linkers and
missing clusters) of UiO-66 have been well documented, there
are only a few reports addressing these defects for UiO
materials with longer linkers.10,28,29 In fact, reports suggest that
UiO-67, incorporating 1,4-biphenyldicarboxylate (BPDC), is
not as stable as UiO-66,30,31 and this is likely due to increases in
the defect nature of UiO-67 materials. It is also known that
minor changes in the synthetic procedure (i.e., changing the
concentration or identity of the modulator) can drastically alter
the properties of the final bulk material.17,25 Therefore, we were
interested in probing the defect nature of UiO-AZB (AZB =
4,4′-azobenzenedicarboxylate, Figure 1).32 We identified three
Figure 1. Representative diagram depicting the general synthetic
conditions used. The octahedral cage on the right was generated using
single crystal data from ref 30.
different modulators appropriate to produce highly crystalline
UiO-AZB materials, acetic (AA), formic (FA), and benzoic
acids (BA), and explored a range of concentrations for each
modulator, 30−70 molar equivalents (mol equiv) for AA and
FA and 10−30 mol equiv for BA. We provide the results of 1H
NMR analysis and N2 adsorption isotherms to elucidate the
extent of missing linkers in all samples generated using our
synthetic procedure.
1H Nuclear Magnetic Resonance: 1H NMR measurements were
made using an Agilent U4-DD2 400 MHz NMR with a 96 sample
robot. Samples were prepared for NMR by digesting ∼1.0 mg in 600
μL of 50 mM Na2HPO4 in D2O via sonication for 1 h. 50 μL of an 8
mM 3-(trimethylsilyl)propionic-2,2,3,3-d4 acid sodium salt (TMSP)
standard were added to each sample, and the sample was transferred to
an NMR tube for analysis.
Theoretical Surface Area Calculation: The geometric accessible
surface area calculations were performed using a nitrogen-sized
spherical probe that is rolled over the atoms of the structure,
according to the method described by Duren et al.35 The diameter of
the nitrogen molecule was set to 21/6 times the Lennard−Jones σ
parameter, with a σ value of 3.32 Å used, based on the TraPPE force
field.
EXPERIMENTAL DETAILS
■
Materials: The 4-nitrobenzoic acid (98%), sodium hydroxide (≥98%),
D-glucose (≥98%), ZrCl4 (anhydrous, 99.99%), and benzoic acid
(≥99.5%) were purchased from Sigma-Aldrich and used without
further purification. Glacial acetic acid (certified ACS), formic acid
(certified ACS), and dimethylformamide (DMF, Spectrophotometric
grade ≥99.8%) were used as received from Fisher Scientific.
Synthesis of 4,4′-azobenzenedicarboxylic acid (AZB):33 4-Nitro-
benzoic acid (13 g) and 250 mL of 5 M NaOH were heated to 50 °C
in a 500 mL round-bottom flask. An aqueous solution of 3.7 M D-
glucose (150 mL) was added slowly (30 mL/min) to the flask with
vigorous stirring. The reaction was allowed to stir for 10 min and then
cooled to room temperature. The solution was aerated for 12 h, and 5
M (200 mL) aqueous acetic acid was added to precipitate the AZB.
The orange solid was collected via vacuum filtration and dried in a 100
RESULTS AND DISCUSSION
■
Three different monocarboxylate modulators (AA, FA, and BA)
of varying concentrations were investigated for the synthesis of
UiO-AZB. The concentration ranges studied were 30−70 equiv
for AA and FA and 10−30 equiv for BA. Varying the
concentration of the modulator has been shown to control
particle sizes,30,36 where an increase in modulator concentration
typically results in larger particles. This is explained by faster
MOF nucleation at lower concentrations of modulator,
therefore, reducing crystal growth and resulting in smaller
particles. Indeed, upon increasing the modulator concentration,
particle sizes increased as evidenced by SEM (Figure 2, and
Figure S1). We originally attempted to synthesize UiO-AZB
without adding a modulator but were not able to obtain a
crystalline product, evidenced by broadened amorphous
features in the powder X-ray diffraction (PXRD) patterns of
1
°C oven for 18 h (9.6 g, ∼91%). H NMR (500 MHz, basic D2O, δ,
ppm) trans: 7.94, 7.92 (d, 2H, J = 10 Hz), 7.82, 7.80 (d, 2H, J = 10
Hz); cis: 7.68, 7.66 (d, 2H, J = 10 Hz), 6.91, 6.89 (d, 2H, J = 10 Hz),
(Predicted m/z = 270.2; M−H− = 269 m/z).
MOF Synthesis: UiO-AZB materials were made by a procedure
previously reported.32,34 AA (30−70 equiv), FA (30−70 equiv), and
BA (10−30 equiv) as the modulators in the synthesis were explored.
The general procedure consisted of adding 0.0233 g of ZrCl4 (0.1
mmol), 0.0270 g of H2AZB (0.1 mmol), 5 mL of DMF (20 mL when
benzoic acid was the modulator), 10 μL of H2O, and the appropriate
B
DOI: 10.1021/acs.inorgchem.7b01801
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
where IMOD is the summed integration of all the signals due to
the modulator, HMOD is the number of protons attributed to the
modulator in solution, HAZB is the number of protons
corresponding to AZB (16 due to 8 for each isomer), and
IAZB is the summed integration of all signals due to AZB. This
ratio can be used to quantify the extent of modulator that
remains bound to the structure after synthesis, washing, and
evacuation. For example, a ratio value of 1.0 would mean there
are equal concentrations of modulator and AZB linker,
signifying that 33% of the linkers are missing per cluster (or
4 of the expected 12).
1
To obtain samples for H NMR analysis, the MOFs were
digested in a 50 mM Na2HPO4 solution in D2O (see
Supporting Information for detailed digestion procedure).
1
Figure 3 shows example H NMR spectra for BA, AA, and
Figure 2. Particle sizes determined from SEM images of materials
synthesized using AA (top, scale bars are 200 nm), FA (middle, scale
bars are 1 μm), and BA (bottom, scale bars are 1 μm except for 10
equiv BA whose scale bar is 200 nm).
these syntheses (Figure S2). The PXRD patterns obtained for
all modulated samples matched single crystal data reported in
the literature30,34 (Figure S3), confirming that the modulators
used produced bulk materials isostructural with the reported
UiO-AZB.
With confirmation of crystalline samples, the focus became
1
quantification of the defect nature of the materials. H NMR
spectra and N2 adsorption isotherms were used in order to
elucidate the extent of missing linker defects in the UiO-AZB
materials. TGA analysis was ineffective to evaluate defects in
UiO-AZB due to a consistent and gradual loss in the TGA
curve below ∼440 °C (Figure S4). In the absence of a clear
plateau in this region, determination of defects had a large
degree of associated error. This specific complication with TGA
defect analysis was highlighted by Farha and co-workers.10
1H Nuclear Magnetic Resonance (1H NMR). UiO MOFs
are unstable in highly basic environments, and so, can be
1
digested in these conditions for H NMR analysis. This is a
powerful tool for the quantitative investigation of missing
linkers. Not only can it provide a precise concentration of
linkers in solution when a known amount of sample is digested,
it can also be used to deem whether missing linkers are replaced
Figure 3. (Top) Example spectrum of digested samples synthesized
with BA. (Middle) Example spectrum of digested samples synthesized
with AA. (Bottom) Example spectrum of digested samples synthesized
with FA.
1
by modulator species. In the H NMR spectra, the integration
of signals is proportional to the number of protons responsible
for the signal. When multiple species are present in solution
(e.g., linker and modulator) the integrations of signals assigned
to each species can be used to determine the relative amounts
FA samples. Signals due to the linker are found in all spectra.
Two doublets corresponding to the protons of trans-AZB are
found at 8.05 and 7.93 ppm, while the doublets due to protons
of the cis isomer are found at 7.77 and 7.02 ppm. Spectra of the
digested samples synthesized using BA displayed a doublet at
7.88 ppm, representative of the two protons located on BA
ortho to the carboxy group. There are also two triplets in the
BA sample spectra at 7.56 and 7.48 ppm, corresponding to
protons located meta and para to the carboxy group,
respectively. The spectra of digested samples synthesized
using AA display a singlet due to the three protons of the
acetyl group at 1.92 ppm. Spectra of the samples synthesized
using FA display the AZB peaks as well as the formate proton
peak at 8.46 ppm.
1
of the various components. Furthermore, the addition of a H
NMR standard allows for the determination of exact
concentrations of compounds in solution. This technique has
been utilized in the literature to investigate and quantify
missing linker defects for UiO-66.15,22
The technique reported by Lillerud22 et al. using 1H NMR to
investigate missing linkers was followed herein. By using the
relative integrations of each component, we are able to calculate
the modulator to AZB ratio (mod/AZB) using eq 1,
⎛
⎜
⎝
⎞
⎟
⎠
⎛
⎞
IMOD
HAZB
Mod
AZB
×
=
ratio
⎜
⎝
⎟
⎠
H
I
(1)
MOD
AZB
C
DOI: 10.1021/acs.inorgchem.7b01801
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Due to the acid/base hydrolysis of DMF, formic acid and
dimethylamine are formed at elevated temperatures and during
the sample digestion procedure. Therefore, a peak correspond-
ing to the formate proton is visible in all spectra (8.46 ppm).
Numerous washing cycles (up to 5) did not result in loss of the
formate peak, indicating either trapped solvent or additional
node interactions. To account for the presence of formate in
the BA and AA samples, we calculated the mod/AZB ratio
under two separate assumptions: (1) that the formate present is
binding to the structure (Figure 4A) such that its integration
experimental value of AZB, and by calculating the theoretical
value of AZB based on the amount of sample digested, we can
calculate a “% recovery” of AZB based on how much we
expected from an “ideal” material with no missing linkers
according to eqs 3 and 4. In eq 2, [AZB] is the experimental
concentration of AZB in solution, IAZB is the summed
integration from the signals attributed to both trans- and cis-
AZB, [S] is the concentration of the standard, and IS is the
integration of the signal from the standard (normalized to 9 in
the case of TMPS).
IAZB × [S]
[AZB] =
IS
(2)
In eq 3, mgUiOAZB is the mass of material digested, molAZB‑ideal
is the moles of AZB present in an ideal sample (6 mol AZB/1
mol UiOAZB), MWUiOAZB is the molecular weight of ideal
UiO-AZB = 2287.34 g/mol, V is the total volume of solution
analyzed (650 μL in our case), and TheoAZB is the theoretical
concentration of AZB expected for the sample assuming no
missing linkers. We can relate this % recovery to a % defects via
eq 5, which corrects the percent recovery for the difference in
molecular weight between an ideal sample and the defected
sample (derivation shown in the Supporting Information).
Figure 4. Modulator to AZB ratio as a function of modulator
concentration for samples synthesized using AA (black squares), FA
(red circles), and BA (blue triangles) as the modulator. Plot A assumes
formic acid present in the BA and AA samples is binding to the
structure, while plot B assumes FA is not binding.
(mgUiOAZB)(molAZB−ideal
)
= TheoAZB
(MWUiOAZB)(V)
(3)
(4)
(5)
[AZB]
TheoAZB
must be included in the consideration of the mod/AZB ratio or
(2) that it is not bound (Figure 4B) and is the result of trapped
FA and, therefore, not included in the calculation. From the
comparison of these two mathematical treatments, we see that
the assumption of bound DMF hydrolysis products results in
mod/AZB ratios that are slightly higher for the AA and BA
samples. In the case of FA samples, similar mathematical
treatments are not possible and, therefore, we assume that the
calculated FA/AZB ratio will be a slight overestimation when
compared to the other data in Figure 4B. Despite these
obstacles, the overall trends are clear.
× 100% = % recovery
MWUiOAZB−ideal
MWUiOAZB−defected
% recovery = (1 − % defects) ×
Again, for all of the modulators (excluding 10 equiv of BA),
there did not seem to be a distinct difference in the average %
defects when the concentration of the modulators was
increased (Figure 5A). Therefore, we averaged all the % defect
To assess the total modulator to AZB ratio for all samples,
the mod/AZB ratio was plotted as a function of modulator
concentration. First, there seems to be no significant increase in
the ratio with increasing concentrations of modulator for all
samples, except in the case of 10 equiv of BA, whose value was
lower at ∼0.3 compared to the remainder of the BA samples.
BA samples overall gave the highest ratios, up to ∼0.6 for
samples synthesized beyond 10 equiv of BA. A ratio of 0.6 (3
BA unit per 5 AZB units) corresponds to 23% of linkers
missing or 2.8 of the 12 expected per cluster. The samples
synthesized with AA display ratios at ∼0.2, indicating 1 unit of
modulator per 5 units of AZB or ∼9% of modulator remaining
(1 missing linker per cluster). The samples synthesized with FA
show ratios of ∼0.3, but again, we expected the FA ratios to be
artificially high. A ratio of 0.3 corresponds to 15% or roughly
1.8 missing linkers per cluster.
The issue of trapped vs bound modulator (including FA
formed via DMF decomposition) was circumvented by the
addition of a 3-(trimethylsilyl)propionic-2,2,3,3-d4 acid sodium
salt (TMPS) standard to the samples. This allows us to
calculate the AZB concentration in solution, independent of the
presence of varying modulators. Knowing the concentration of
the standard, the integration of standard signals and the
integration of AZB signals can be used to calculate the
concentration of AZB in solution, using eq 2. This gives an
Figure 5. (A) The % defects calculated from 1H NMR spectra
according to eqs 2−5) as a function of modulator equivalents for
samples synthesized using AA (black squares), FA (red circles), and
BA (blue triangles). (B) Average % defects of each modulator.
values obtained for each modulator regardless of the
concentration (Figure 5B). For BA, we omitted the % defects
from the samples made with 10 equiv. When comparing the
modulators to each other (Figure 5, B), it is clear that AA is on
average less defected, 29% defects, when compared to BA and
FA, 33% and 36% defected, respectively. The BA samples made
with 10 equiv resulted in ∼23% defects. Interestingly, the trend
is distinctively different from that found from the Mod/AZB
ratio and follows what would be expected from the acid
strength of the modulator, i.e. the pKa’s of the acids increase FA
D
DOI: 10.1021/acs.inorgchem.7b01801
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
(3.75) < BA (4.20) < AA (4.75). Such a trend was previously
observed by Lillerud in the study of defects in UiO-66 with AA,
FA, difluoroacetic acid, and trifluoroacetic acid.22 Given the
agreement of the % defect calculation with literature precedent,
the hypothesis that stronger acids result in more defected
structures is applied throughout the rest of the discussion.
The differences between the Mod/AZB ratio method and the
% defect calculation indicate a key finding in regards to defect
analysis in various MOF structures. Differentiating between
trapped modulator and bound modulator is critical to an
accurate picture of MOF defects. The difference in trend
between the two methods (AA < BA < FA; % defects vs AA <
FA < BA; Mod/AZB ratio) indicate that the benzoic acid/AZB
ratio overestimates defects, if it is assumed all benzoic acid
present in the 1H NMR spectrum is bound to nodes. Therefore,
to rectify the differences between the two values, trapped
benzoic acid must be present. Such a discrepancy is not
qualitatively seen with either AA or FA. However, this can be
expected, as AA and FA are liquid modulators and, thus, more
easily removed from the pores during traditional activation
procedures as compared to the solid modulator, benzoic acid.
This clearly provides support for the calculation of % defects in
comparison to a standard as opposed to the ratio method when
solid modulators are used in the synthetic procedure. It is also
important to note that neither method clearly identifies the
nature of all defects, which may include missing linkers replaced
by modulator, hydroxide, or water and/or missing nodes and
more importantly the orientation of defects to each other as
will be discussed below.
N2 Adsorption Isotherms. Gas sorption isotherms have
been extensively used throughout MOF literature as a
benchmark to provide insights into the porosity of frameworks.
N2 isotherms for all of the samples were measured, and BET
theory was applied to the appropriate pressure ranges governed
by the criteria outlined in the literature.35,37,38 Several reports
indicate that increases in the modulator concentration result in
an increase in gas uptake and thus the SA of UiO-66.22,25 This
experimental evidence supports the logical hypothesis that
increased modulator concentrations result in more defects that
open more surface area for gas sorption. Of the samples
investigated in this study, this trend was not observed. In the
case of UiO-66, there have only been two main types of defects
suggested and evidenced in the literature: missing linker defects
and missing clusters. In fact, drastic increases in SAs of UiO-66
as more modulator is added have been attributed mainly to
missing cluster defects, which is evidenced by broad peaks
below ∼5° in the PXRD pattern.22,28 Based on the PXRD
patterns for the materials studied, similar evidence of missing
clusters was not present. Given the larger unit cell or UiO-AZB,
it is possible that missing linker defects would appear in the
XRD pattern below 3° (the detection limit of our instrument).
That said, previous reports only address the “point-like”
vacancy defects characteristic of crystalline materials. During
crystal growth, other crystal defects are possible including line
and planar imperfections. For example, Siperstein24 and co-
workers show that decreasing CO2 uptake can be attributed to
pore blockage defects in Cu3(BTC)2 (BTC = 1,3,5-benzene-
tricarboxylate). These types of defects may arise from pockets
of trapped solvent or as a result of crystal growth around
synthetic byproducts (such as ZrO2 in the case of UiO
materials). The same MOF has also been shown to contain
plane slipping defects39 and crystal fractures40 that can result
from postsynthetic activation treatments. Furthermore, system-
atic experimental SA analyses, and how SA changes as more
modulator is added, have not been conducted for UiO materials
with longer linkers than UiO-66, which incorporates BDC.
All samples resulted in type II isotherms, and example
isotherms for AA and BA are shown in Figure 6 (A and B
Figure 6. (A) N2 isotherms for materials synthesized with 30 equiv
(black), 40 equiv (red), 50 equiv (blue), 60 equiv (orange), 70 equiv
(green) of AA. (B) N2 isotherms for materials synthesized with 10
equiv (black), 15 equiv (red), 20 equiv (blue), 25 equiv (orange), 30
equiv (green) of BA. (C) BET SA of materials synthesized with
varying concentrations of AA (black squares), FA (red circles), and BA
(blue triangles). (D) Pore width for AA (black squares), FA (red
circles), and BA (blue triangles) as a function of modulator
concentration.
respectively). Figure 6C displays the BET SA of all samples.
Overall, AA samples gave the highest SAs. Several trials of BET
analysis of AA samples were used to evaluate the batch variation
of AA over the concentration range. Given the standard
deviation of the surface area (∼10%), for all AA samples the SA
was the same (within error). Therefore, the averaged SA for AA
materials was 2687
218 m2/g. FA resulted in SAs that
increased from 2559 m2/g at 30 equiv of FA to 2769 m2/g with
40 equiv; however, above 40 equiv, the SA dropped as low as
1985 m2/g with 70 equiv of FA. When the lowest amount of
BA was used, we observed an SA of 2682 m2/g, which
consistently fell to 903 m2/g when the BA concentration was
increased from 10 to 30 equiv.
Type II isotherms are indicative of mesoporous (2−50 nm
pore widths) and macroporous (>50 nm) materials. In the low
pressure region of the isotherm, the formation of a monolayer
of N2 occurs on the adsorption sites at the framework “walls”.
At higher pressures, once the monolayer has formed,
adsorption becomes dominated by pore filling and the sharp
rise in the isotherm at relative pressures near unity indicates
unrestricted adsorption where saturation of the pores is taking
place.41 In the case of a sample with large mesopores or
macropores, the surface area decreases compared to a
microporous (<2 nm) sample of the same framework, because
the “walls” for gas sorption are absent in these large cavities
and, therefore, have less adsorption sites.42,43 The introduction
of macroporosity in our samples explains why we observe lower
surface areas than the theoretical SA of the fully intact structure,
which was found to be 3352 m2/g (see Supporting Information
for calculation parameters). Average pore width analysis for all
samples is displayed in Figure 6D.
E
DOI: 10.1021/acs.inorgchem.7b01801
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
For the FA samples, the pore width decreased as the
concentration of FA increases. If no other factors are playing a
role, the SA should therefore increase. However, we observe a
decrease in the SA above 40 equiv of FA, indicating that other
aspects of the defectivity should be considered other than the
1
pore width. From the H NMR analysis, it is clear that, in the
case of FA, the concentration of modulator does not affect the
mod/AZB ratio and, therefore, does not serve to introduce the
occurrence of more defects. However, we see that increasing
the FA concentration does affect the porosity of UiO-AZB,
evidenced by a decrease in the SA and pore width. This leads to
the conclusion that, at higher concentrations of FA, the total
void volume decreases due to increasing pore blockage defects.
In this case, pore blockage is not due to trapped solvent,
1
because the amount of solvent in the H NMR spectra are not
significantly different as the concentration of modulator is
increased. However, the pores could be blocked through ZrO2
production at higher concentrations of FA, resulting in crystal
growth around these contaminants. Alternatively, samples
could contain inseparable byproduct (ZrO2 or otherwise),
resulting in decreased SA due to an averaging effect. However,
SEM analysis does not indicate that an inseparable byproduct is
present.
Figure 7. Diagram of missing linkers on a UiO-AZB cluster. (Top)
Fully decorated cluster. (Bottom left) Four missing clusters oriented
90° from the others. (Bottom right) Four missing clusters that are all
coplanar. The colors represent: O (red), C (gray), N (blue), Zr
(green), and H (white).
For BA samples, the total uptake of N2 increased with
increased amounts of BA up to 25 equiv (dotted circle Figure
6B), which pointed to increased porosity. Average pore width
analysis indicated that when the concentration of BA is raised,
the pore width drastically increases up to ∼60 Å at 25 equiv.
Therefore, it is likely that the drop in the BET SA is a result of
the blockage of micropores and/or the formation of larger
meso-/macropores in the BA samples as the concentration of
synthesis remain attached to the Zr6 clusters, resulting in
various defect levels depending on modulator identity. Use of
AA as the modulator was found to result in ∼30% defected
structure. However, increased AA concentration had no effect
on the porosity of the samples. In the case of FA, with an ∼40%
defected structure, the average pore width and SA decrease as
the concentration of FA is raised. This suggests that increasing
the FA concentration results in either pore blockage defects or
more inseparable byproducts that contaminate the samples,
thus, lowering the SA. When BA is increased beyond 10 equiv,
the % defects reached 33%, and increased BA concentration
further results in larger meso- or macropores that serve to
decrease the SA.
Although more work is needed to investigate the chemical
nature of defects in MOFs, there are quick techniques that give
generalized information about the degree and structural nature
of defects in bulk samples. Several factors are responsible for
introducing defect sites into framework structures. Previous
studies have shown an increase in the SA of UiO-66 when the
concentration of modulator was raised due to more missing
cluster defects. However, the results presented here confirm
that defects are linker dependent and specific SA trends for a
single MOF should not be extended to other materials, even if
isostructural. Additionally, the results presented indicate that
different types of defects are observed within one MOF
structure depending on modulator identity. In fact, we show
that FA results in pore blockage defects, AA results in the
commonly observed missing linker point defects, and BA
results in clustered defects and macroporosity. Ultimately, it
appears that defect analysis should be conducted on each MOF
and each potential modulator. Only after this information is
gathered can the crystal properties desired (low defect material,
larger porosity) be designed into the material.
1
BA is raised. From H NMR analysis, it is clear that excess
benzoic acid is present in the structure. However, this does not
appear to change as a function of added modulator and,
therefore, cannot fully explain the decrease in surface area. In a
report by Thornton and co-workers,28 where they model
several possible defect scenarios and simulate the CO2 uptake
of the resulting “three-site model” (assumed to be three linked
clusters that possess the defect of interest), they point out that
two missing linkers on a UiO SBU cluster can be either linear
or not (i.e., adjacent or not). The orientation of missing linkers
that are replaced by modulator species becomes more complex
as the number of missing linkers per cluster is increased. For
example, imagine a fully decorated cluster (Figure 7, top). If
four missing linkers are placed the furthest apart from one
another (Figure 7, bottom left), then crystal growth can
proceed in all directions. However, if the four missing linkers
are all coplanar (Figure 7, bottom right), a whole face of the
cluster is restricted from crystal growth. This restriction allows
for nanoregions of crystal growth inhibition and results in the
formation of large vacant pockets in the final material. Due to
π−π stacking and known dimer formation of BA,44 increasing
the concentration of BA could result in more prevalent
coplanar vacancies and, therefore, increase the average pore
spaces in the bulk BA samples, resulting in lower SAs.
In the case of AA, the pore width showed a general decrease,
but the large standard deviations indicate that the values were
within error. When all AA samples were averaged, the pore
width was determined to be 30 12 nm.
CONCLUSIONS/OUTLOOK
UiO-AZB powders were studied in order to investigate the
defect nature of the MOFs. The modulators used in the
■
F
DOI: 10.1021/acs.inorgchem.7b01801
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Modulation as a Tool to Increase the Catalytic Activity of Metal−
Organic Frameworks: The Unique Case of UiO-66(Zr). J. Am. Chem.
Soc. 2013, 135, 11465−11468.
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.inorg-
chem.7b01801.
■
S
́
(12) Montoro, C.; Ocon, P.; Zamora, F.; Navarro, J. A. R. Metal−
Organic Frameworks Containing Missing-Linker Defects Leading to
High Hydroxide-Ion Conductivity. Chem. - Eur. J. 2016, 22, 1646−
1651.
(13) Jared, M.; Taylor, S. D. Ryuichi Ikeda, and Hiroshi Kitagawa
Defect Control to Enhance Proton Conductivity in a Metal−Organic
Framework. Chem. Mater. 2015, 27, 2286−2289.
All instrumentation used along with more detailed
experimental procedures are provided. Also, PXRD,
1
TGA, and a H NMR of an FA sample (PDF)
(14) Liang, W.; Coghlan, C. J.; Ragon, F.; Rubio-Martinez, M.;
D’Alessandro, D. M.; Babarao, R. Defect Engineering of UiO-66 for
CO2 and H2O Uptake - a Combined Experimental and Simulation
Study. Dalt. Trans. 2016, 45, 4496−4500.
(15) Wang, K.; Li, C.; Liang, Y.; Han, T.; Huang, H.; Yang, Q.; Liu,
D.; Zhong, C. Rational Construction of Defects in a Metal−Organic
Framework for Highly Efficient Adsorption and Separation of Dyes.
Chem. Eng. J. 2016, 289, 486−493.
(16) Ghosh, P.; Colon, Y. J.; Snurr, R. Q. Water Adsorption in UiO-
66: The Importance of Defects. Chem. Commun. 2014, 50, 11329−
11331.
(17) Valenzano, L.; Civalleri, B.; Chavan, S.; Bordiga, S.; Nilsen, M.
H.; Jakobsen, S.; Lillerud, K. P.; Lamberti, C. Disclosing the Complex
Structure of UiO-66 Metal Organic Framework: A Synergic
Combination of Experiment and Theory. Chem. Mater. 2011, 23,
1700−1718.
AUTHOR INFORMATION
Corresponding Author
*E-mail: ajmorris@vt.edu.
ORCID
Amanda J. Morris: 0000-0002-3512-0366
Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS
■
This material is based upon work supported by the National
Science Foundation under Grant No. 1551964. Also special
thanks to Krista Walton and Yang Jiao for help with theoretical
SA calculations.
(18) Katz, M. J.; Brown, Z. J.; Colon, Y. J.; Siu, P. W.; Scheidt, K. A.;
Snurr, R. Q.; Hupp, J. T.; Farha, O. K. A Facile Synthesis of UiO-66,
UiO-67 and Their Derivatives. Chem. Commun. 2013, 49, 9449−9451.
(19) Ling, S.; Slater, B. Dynamic Acidity in Defective Uio-66. Chem.
Sci. 2016, 7, 4706−4712.
REFERENCES
■
(1) McDonald, T. M.; Lee, W. R.; Mason, J. A.; Wiers, B. M.; Hong,
C. S.; Long, J. R. Capture of Carbon Dioxide from Air and Flue Gas in
the Alkylamine-Appended Metal−Organic Framework Mmen-
Mg2(DOBPDC). J. Am. Chem. Soc. 2012, 134, 7056−7065.
(2) Furukawa, H.; Cordova, K. E.; O’Keeffe, M.; Yaghi, O. M. The
Chemistry and Applications of Metal-Organic Frameworks. Science
2013, 341, 1230444.
(3) Bobbitt, N. S.; Mendonca, M. L.; Howarth, A. J.; Islamoglu, T.;
Hupp, J. T.; Farha, O. K.; Snurr, R. Q. Metal-Organic Frameworks for
the Removal of Toxic Industrial Chemicals and Chemical Warfare
Agents. Chem. Soc. Rev. 2017, 46, 3357.
(4) Saleem, H.; Rafique, U.; Davies, R. P. Investigations on Post-
Synthetically Modified UiO-66-NH2 for the Adsorptive Removal of
Heavy Metal Ions from Aqueous Solution. Microporous Mesoporous
Mater. 2016, 221, 238−244.
(5) Dias, E. M.; Petit, C. Towards the Use of Metal-Organic
Frameworks for Water Reuse: A Review of the Recent Advances in the
Field of Organic Pollutants Removal and Degradation and the Next
Steps in the Field. J. Mater. Chem. A 2015, 3, 22484−22506.
(6) He, C.; Liu, D.; Lin, W. Nanomedicine Applications of Hybrid
Nanomaterials Built from Metal−Ligand Coordination Bonds: Nano-
scale Metal−Organic Frameworks and Nanoscale Coordination
Polymers. Chem. Rev. 2015, 115, 11079−11108.
(7) Horcajada, P.; Gref, R.; Baati, T.; Allan, P. K.; Maurin, G.;
́
Couvreur, P.; Ferey, G.; Morris, R. E.; Serre, C. Metal−Organic
Frameworks in Biomedicine. Chem. Rev. 2012, 112, 1232−1268.
(8) Huxford, R. C.; Della Rocca, J.; Lin, W. Metal−Organic
Frameworks as Potential Drug Carriers. Curr. Opin. Chem. Biol.
2010, 14, 262−268.
(9) Vermoortele, F.; Ameloot, R.; Alaerts, L.; Matthessen, R.; Carlier,
B.; Fernandez, E. V. R.; Gascon, J.; Kapteijn, F.; De Vos, D. E. Tuning
the Catalytic Performance of Metal-Organic Frameworks in Fine
Chemistry by Active Site Engineering. J. Mater. Chem. 2012, 22,
10313−10321.
(10) Liu, Y.; Klet, R. C.; Hupp, J. T.; Farha, O. Probing the
Correlations between the Defects in Metal-Organic Frameworks and
Their Catalytic Activity by an Epoxide Ring-Opening Reaction. Chem.
Commun. 2016, 52, 7806−7809.
(11) Vermoortele, F.; Bueken, B.; Le Bars, G.; Van de Voorde, B.;
Vandichel, M.; Houthoofd, K.; Vimont, A.; Daturi, M.; Waroquier, M.;
Van Speybroeck, V.; Kirschhock, C.; De Vos, D. E. Synthesis
(20) Vandichel, M.; Hajek, J.; Vermoortele, F.; Waroquier, M.; De
Vos, D. E.; Van Speybroeck, V. Active Site Engineering in UiO-66
Type Metal-Organic Frameworks by Intentional Creation of Defects:
A Theoretical Rationalization. CrystEngComm 2015, 17, 395−406.
́
(21) Gutov, O. V.; Hevia, M. G.; Escudero-Adan, E. C.; Shafir, A.
Metal−Organic Framework (MOF) Defects under Control: Insights
into the Missing Linker Sites and Their Implication in the Reactivity of
Zirconium-Based Frameworks. Inorg. Chem. 2015, 54, 8396−8400.
(22) Shearer, G. C.; Chavan, S.; Bordiga, S.; Svelle, S.; Olsbye, U.;
Lillerud, K. P. Defect Engineering: Tuning the Porosity and
Composition of the Metal−Organic Framework UiO-66 Via
Modulated Synthesis. Chem. Mater. 2016, 28, 3749−3761.
(23) Lee, S. J.; Doussot, C.; Baux, A.; Liu, L.; Jameson, G. B.;
Richardson, C.; Pak, J. J.; Trousselet, F.; Coudert, F.-X.; Telfer, S. G.
Multicomponent Metal−Organic Frameworks as Defect-Tolerant
Materials. Chem. Mater. 2016, 28, 368−375.
(24) Al-Janabi, N.; Fan, X.; Siperstein, F. R. Assessment of Mof’s
Quality: Quantifying Defect Content in Crystalline Porous Materials. J.
Phys. Chem. Lett. 2016, 7, 1490−1494.
(25) Wu, H.; Chua, Y. S.; Krungleviciute, V.; Tyagi, M.; Chen, P.;
Yildirim, T.; Zhou, W. Unusual and Highly Tunable Missing-Linker
Defects in Zirconium Metal−Organic Framework UiO-66 and Their
Important Effects on Gas Adsorption. J. Am. Chem. Soc. 2013, 135,
10525−10532.
(26) 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. Cryst. Growth Des. 2016, 16, 2295−2301.
(27) Klet, R. C.; Liu, Y.; Wang, T. C.; Hupp, J. T.; Farha, O. K.
Evaluation of Bronsted Acidity and Proton Topology in Zr- and Hf-
Based Metal-Organic Frameworks Using Potentiometric Acid-Base
Titration. J. Mater. Chem. A 2016, 4, 1479−1485.
(28) Thornton, A. W.; Babarao, R.; Jain, A.; Trousselet, F.; Coudert,
F. X. Defects in Metal-Organic Frameworks: A Compromise between
Adsorption and Stability? Dalt. Trans. 2016, 45, 4352−4359.
(29) Cliffe, M. J.; Castillo-Martínez, E.; Wu, Y.; Lee, J.; Forse, A. C.;
Firth, F. C. N.; Moghadam, P. Z.; Fairen-Jimenez, D.; Gaultois, M. W.;
G
DOI: 10.1021/acs.inorgchem.7b01801
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Hill, J. A.; Magdysyuk, O. V.; Slater, B.; Goodwin, A. L.; Grey, C. P.
Metal−Organic Nanosheets Formed Via Defect-Mediated Trans-
formation of a Hafnium Metal−Organic Framework. J. Am. Chem. Soc.
2017, 139, 5397−5404.
(30) 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.
(31) DeCoste, J. B.; Peterson, G. W.; Jasuja, H.; Glover, T. G.;
Huang, Y.-g.; Walton, K. S. Stability and Degradation Mechanisms of
Metal-Organic Frameworks Containing the Zr6o4(Oh)4 Secondary
Building Unit. J. Mater. Chem. A 2013, 1, 5642−5650.
(32) Epley, C. C.; Roth, K. L.; Lin, S.; Ahrenholtz, S. R.; Grove, T. Z.;
Morris, A. J. Cargo Delivery on Demand from Photodegradable MOF
Nano-Cages. Dalt. Trans. 2017, 46, 4917−4922.
(33) Mary, L. J. F.; Kannan, P. Synthesis, Characterization and
Thermal Behaviour of Poly(Pyromellitimide-Ester)s and (Imide-
Urethane)s Containing Azobenzene Units. Eur. Polym. J. 1999, 35,
17−26.
(34) Schaate, A.; Duhnen, S.; Platz, G.; Lilienthal, S.; Schneider, A.
̈
M.; Behrens, P. A Novel Zr-Based Porous Coordination Polymer
Containing Azobenzenedicarboxylate as a Linker. Eur. J. Inorg. Chem.
2012, 2012, 790−796.
́
(35) Duren, T.; Millange, F.; Ferey, G.; Walton, K. S.; Snurr, R. Q.
̈
Calculating Geometric Surface Areas as a Characterization Tool for
Metal−Organic Frameworks. J. Phys. Chem. C 2007, 111, 15350−
15356.
(36) Tsuruoka, T.; Furukawa, S.; Takashima, Y.; Yoshida, K.; Isoda,
S.; Kitagawa, S. Nanoporous Nanorods Fabricated by Coordination
Modulation and Oriented Attachment Growth. Angew. Chem., Int. Ed.
2009, 48, 4739−4743.
́ ́
(37) Wang, T. C.; Bury, W.; Gomez-Gualdron, D. A.; Vermeulen, N.
A.; Mondloch, J. E.; Deria, P.; Zhang, K.; Moghadam, P. Z.; Sarjeant,
A. A.; Snurr, R. Q.; Stoddart, J. F.; Hupp, J. T.; Farha, O. K. Ultrahigh
Surface Area Zirconium MOFs and Insights into the Applicability of
the Bet Theory. J. Am. Chem. Soc. 2015, 137, 3585−3591.
(38) Walton, K. S.; Snurr, R. Q. Applicability of the Bet Method for
Determining Surface Areas of Microporous Metal−Organic Frame-
works. J. Am. Chem. Soc. 2007, 129, 8552−8556.
(39) Ameloot, R.; Vermoortele, F.; Hofkens, J.; De Schryver, F. C.;
De Vos, D. E.; Roeffaers, M. B. J. Three-Dimensional Visualization of
Defects Formed During the Synthesis of Metal−Organic Frameworks:
A Fluorescence Microscopy Study. Angew. Chem., Int. Ed. 2013, 52,
401−405.
(40) Shoaee, M.; Agger, J. R.; Anderson, M. W.; Attfield, M. P.
Crystal Form, Defects and Growth of the Metal Organic Framework
HKUST-1 Revealed by Atomic Force Microscopy. CrystEngComm
2008, 10, 646−648.
́ ́
(41) Gomez-Gualdron, D. A.; Moghadam, P. Z.; Hupp, J. T.; Farha,
O. K.; Snurr, R. Q. Application of Consistency Criteria to Calculate
Bet Areas of Micro- and Mesoporous Metal−Organic Frameworks. J.
Am. Chem. Soc. 2016, 138, 215−224.
(42) Du, H.; Bai, J.; Zuo, C.; Xin, Z.; Hu, J. A Hierarchical Supra-
Nanostructure of HKUST-1 Featuring Enhanced H2 Adsorption
Enthalpy and Higher Mesoporosity. CrystEngComm 2011, 13, 3314−
3316.
(43) Klimakow, M.; Klobes, P.; Rademann, K.; Emmerling, F.
Characterization of Mechanochemically Synthesized Mofs. Micro-
porous Mesoporous Mater. 2012, 154, 113−118.
(44) Yamaguchi, S.; Tominaga, K.; Saito, S. Intermolecular
Vibrational Mode of the Benzoic Acid Dimer in Solution Observed
by Terahertz Time-Domain Spectroscopy. Phys. Chem. Chem. Phys.
2011, 13, 14742−14749.
H
DOI: 10.1021/acs.inorgchem.7b01801
Inorg. Chem. XXXX, XXX, XXX−XXX