Functionalization of Metal-Organic Frameworks To Achieve Controllable Wettability

Article abstract of DOI:10.1021/acs.inorgchem.7b00373

The overall versatility of a material can be immensely expanded by the ability to controllably tune its hydrophobicity. Herein we took advantage of steric bias to demonstrate that tricarboxylate metal-organic frameworks (MOFs) can undergo covalent postsyn

Full text of DOI:10.1021/acs.inorgchem.7b00373

Article
pubs.acs.org/IC
Functionalization of Metal−Organic Frameworks To Achieve
Controllable Wettability
Heather N. Rubin† and Melissa M. Reynolds*,†,‡,§
‡
§
†Department of Chemistry, School of Biomedical Engineering, and Department of Chemical & Biological Engineering, Colorado
State University, Fort Collins, Colorado 80523, United States
S
* Supporting Information
ABSTRACT: The overall versatility of a material can be immensely
expanded by the ability to controllably tune its hydrophobicity. Herein
we took advantage of steric bias to demonstrate that tricarboxylate
metal−organic frameworks (MOFs) can undergo covalent post-
synthetic modification to confer various degrees of hydrophobicity.
MOF copper 2-aminobenzene-1,3,5-tricarboxylate was modified with
varying-length aliphatic carbon chains. Unmodified Cu3(NH2BTC)2
degrades in minutes upon contact with water, whereas modification as
low as 14% results in powders that show significantly enhanced
hydrophobic character with contact angles up to 147°. The modified
material is capable of withstanding direct contact with water for 30
min with no visual evidence of altered surface characteristics. A linear
relationship was observed between the length of the tethered chain
and the water contact angle. These results reveal a predictable method for achieving a range of desirable sorption rates and highly
controllable hydrophobic character. This work thereby expands the possibilities of rationally modifying MOFs for a plethora of
target-specific applications.
applications, while tunable degradation rates would hold
potential for drug carriers with tunable drug release.11,12
Postsynthetic modification (PSM) provides a method of
obtaining new physical and chemical characteristics with
preexisting MOFs via late-stage functionalization of the material
including the hydrophobicity.13−17 While changing the overall
wettability of material from hydrophilic to hydrophobic is
valuable, control over the degree of hydrophobicity holds even
more promise for expanding the overall usefulness of a
material.18,19 Tricarboxylate ligands have an increased level of
steric hindrance, which may change the rate of PSM and enable
fine-tuning of the amount of functionalization. These systems
have yet to be thoroughly investigated with a covalent
postsynthetic functionalization strategy. This is partially
attributable to concerns that steric hindrance may impede
access to the functional handle intended for modification.
Moreover, tricarboxylate ligands with useful functional handles
(such as primary amines) have limited availability. Thus, the
present study was conducted to determine (a) if NH2H3BTC
can be accessed with a scalable, cost-effective synthesis to
provide a tricarboxylate MOF for PSM, (b) if MOFs with
sterically hindered constituents can undergo covalent PSM with
long-chained and branched hydrocarbons, and (c) if such PSM
can expand functionality of the material by providing an
isostructural system with improved hydrophobicity.
INTRODUCTION
■
The ability to predictably tune the properties of a material
expands the overall usefulness, leading to enhanced function-
ality for target-specific applications. Hybrid materials are
composed of two or more constituents and inherently possess
a high degree of tunability attributable to the ability to
manipulate each individual component.1 Metal−organic frame-
works (MOFs) are a distinct class of organic−inorganic hybrid
nanomaterials with intricate porous networks. The unique
chemical environments at the pore sites have led to the
investigation of MOFs for applications such as gas adsorption,
catalysis, drug delivery, electronics, and luminescence.2−4
Wettability is a characteristic that attracts significant attention
in the practical applicability of materials, including MOFs.5−7
Engineering the wettability of a material presents a rational, yet
complex solution to combat failure related to water absorptivity
in biomedical, energy, environmental, and industrial systems.
The ability to carefully and predictably tune MOF wettability
would enhance potential applications and provide a material
with controllable desired characteristics.
Currently counted among the most-studied frameworks,
copper benzene-1,3,5-tricarboxylate (Cu3(BTC)2) has shown a
range of notable applications; however, water instability
presents a significant obstacle for practical use.8−10 A method
to control interaction of a MOF with water would permit this
instability to be circumvented or exploited for a specific
application. For example, rendering the material more stable in
aqueous environments would significantly impact long-term
Received: February 9, 2017
© XXXX American Chemical Society
A
DOI: 10.1021/acs.inorgchem.7b00373
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
beam side during the XRD studies. The instrument alignment was
tested using the NIST 1976b SRM. A typical scan rate was 0.1 s/step
with a step size of 0.02°. Infrared (IR) data were acquired on a Nicolet
6700 Fourier transform infrared (FTIR) spectrometer with all samples
prepared as pressed potassium bromide (KBr) pellets. In short, ∼5−10
mg of gently ground MOF was added to a clean 10 mL beaker
followed by ∼200 mg of FTIR-grade potassium bromide. After being
thoroughly mixed, samples were pressed to obtain sample containing
KBr pellets.
We report a practical 2-step synthetic route to 2-amino-
benzene-1,3,5-tricarboxylic acid for tricoordinate MOFs. The
ensuing synthesis of the MOF Cu3(NH2BTC)2 produced a
highly uniform, crystalline green powder that was subsequently
functionalized with various anhydrides. This yields a series of
modified frameworks having mixed ligands. On the basis of
equivalents of anhydride added, the percentage of modification
is varied. Remarkable reliability and reproducibility to obtain
mixed-ligand ratios were observed, while drastic changes in the
hydrophobic character of the material were simultaneously
achieved even at low conversion quantities. The wettability of
the modified MOFs was carefully evaluated for each sample
with a method outlined herein that allows for comparison of
water sorption rates into MOF powders. With 7% conversion
using valeric anhydride, water sorption rates decreased from 8.8
to 0.46 μL/s. Furthermore, this type of modification resulted in
Cu3(NH2BTC)2 derivatives demonstrating hydrophobic to
nearly superhydrophobic water contact angles (120−150°).
As a whole, this work presents a new approach to access
NH2H3BTC, a reliable covalent PSM with long-chained and
branched hydrocarbons; notably, PSM results in altered
physical and chemical properties that provide an isostructural
system with tunable hydrophobicity and enhanced stability to
water. For the first time, the type and percentage of
modification is observed to result in significant variations to
the wettability of the material in this 3D MOF system.
Synthetic Methods. N-Mesitylacetamide was synthesized in two
steps beginning with 2,4,6-trimethylaniline via modifications to a
known literature procedure.21 Under argon atmosphere, 2,4,6-
trimethylaniline (10.5 mL, 75 mmol, 1 equiv) was added to a flame-
dried 500 mL round-bottom flask equipped with a Teflon stir bar.
Dichloromethane (100 mL) was added while stirring, and the flask was
cooled to 0 °C. Acetyl chloride (5.64 mL, 79 mmol, 1.05 equiv) was
added dropwise via addition funnel, followed by triethylamine (11.01
mL, 79 mmol, 1.05 equiv). The reaction was monitored by thin-layer
chromatography (TLC). When little to no starting material remained
(∼2 h), the reaction mixture was filtered. The solid was suspended in
water for 30 min, collected by filtration, and dried overnight in vacuo
to yield 9.5 g, 72%. Crude material was carried on without further
purification. 1H NMR (CDCl3, 400 MHz, δ): 6.93, s, 2H; 2.29, s, 3H;
2.25, s, 6H; 1.73, s, 3H. 13C NMR (CDCl3, 101 MHz, δ): 169.0, 137.1,
136.3, 135.3, 129.2, 128.9, 23.18, 20.90, 19.75, 18.33. MS ESI-APCI:
found, M + 1 = 178.12. IR (KBr pellet) ν (cm−1): 3147, 3097, 2914,
2852, 2764, 2534, 2447, 1896, 1716, 1676, 1483, 1414, 1225, 766.
2-Aminobenzene-1,3,5-tricarboxylic acid was synthesized from
crude N-mesitylacetamide, in a one-pot oxidation of the methyl
substituents2 and removal of the acetyl group. N-mesitylacetamide (5.0
g, 28.2 mmol, 1 equiv) was added to a 500 mL round-bottom flask
equipped with a Teflon stir bar and condenser, followed by 165 mL of
deionized H2O. NaOH was added (0.55 g, 14 mmol, 0.5 equiv) slowly
in ∼0.1 g portions over 5 min. KMnO4 was added last (34 g, 215
mmol, 7.8 equiv) in ∼5 g portions over 2 h at room temperature.
Once the KMnO4 addition was complete, the reaction was stirred for 1
h at room temperature and then heated to reflux at 85 °C for 72 h.
The resultant brown slurry was filtered through filter paper to remove
MnO2 and rinsed with 200 mL of hot water. The clear filtrate was
acidified with 20 mL of concentrated HCl while stirring to a pH of 1−
2. The mixture was refluxed overnight at 100 °C. Upon cooling to
room temperature, the product crystallized in solution as a white solid.
The solid was filtered and rinsed with 200 mL of ice-cold water,
yielding 3 g, 40%, of the product. 1H NMR (DMSO-d6, 400 MHz, δ):
13.27 (s, 3H); 8.56 (s, 2H); 8.55 (s, 2H). 13C NMR (DMSO-d6, 101
MHz, δ): 169.1, 166.6, 155.6, 138.7, 115.1, 112.0. MS ESI-APCI:
found, M + 1 = 226.03. IR (KBr pellet) ν (cm−1): 3418, 3291, 3074,
1716, 1678, 1446, 1570, 1292, 1252, 1114, 1084, 883, 813, 694.
Synthesis of MOF Cu3(NH2BTC)2. The MOF was synthesized
following modification to a known literature procedure.10 In brief, 2-
aminobenzene-1,3,5-tricarboxylic acid (510 mg, 2.65 mmol, 1 equiv)
was added to a 100 mL Pyrex glass container followed by Cu(NO3)2·
3H2O (1.03 g, 4.26 mmol, 1.61 equiv). Five mL of H2O and 40 mL of
DMA were added by syringe, and the solution was mixed via
sonication for 10 min. The homogeneous solution was then heated to
85 °C for 42 h and slowly cooled to room temperature. The resulting
green/teal crystalline MOF material was collected by filtration and
dried in air overnight. The MOF was then loaded into a cellulose
thimble and was treated by Soxhlet extraction with ethanol for 20 h to
exchange the solvent. Lastly, the MOF was thermally activated in a
vacuum oven at 120 °C for 20 h, yielding a dark-green powder.
Postsynthetic Modification of MOF Cu3(NH2BTC)2. For each
modification, Cu3(NH2BTC)2 (75 mg, 0.2 mmol relative to the
amine) was added to a 2-dram vial, followed by 4 mL of CHCl3, and
soaked overnight. After 12 h each vial was dosed with 1, 10, 20, or 40
equiv of the desired anhydride. The heterogeneous mixtures were
mixed via vortex for 3 s at 3000 rpm. Vials were then heated to 45 °C
for 24 h in a heating block. Upon cooling, the vials were centrifuged at
3000 rpm for 10 min. Solvent was decanted with a pipet, and 4 mL of
clean CHCl3 was added. The vials were vortexed once again and
EXPERIMENTAL SECTION
■
Materials. Syntheses sensitive to air and moisture were performed
using standard Schlenk techniques under an atmosphere of inert
nitrogen or argon. Anhydrous dimethylacetamide (DMA) and
anhydrous methanol (MeOH) were purchased through Sigma-Aldrich.
Anhydrous chloroform (CHCl3) was obtained from Fisher Scientific.
Cu3(BTC)2 was purchased as Basolite C300 from Sigma-Aldrich.
Additional solvents and reagents were obtained from TCI, Aldrich, and
Alfa Aesar and used without further purification. The synthesis of the
MOF herein is previously described in the literature.18 A Millipore
purification system set at 18 MΩ was used for all water-submersion
experiments.
Characterization. All NMR experiments were performed using an
Agilent (Varian) 400MR equipped with automated tuning and a 7600
sample changer at room temperature. All NMR spectra were analyzed
using MestReNova NMR software. To verify and quantify
modification to each ligand after the above PSM procedure, all
samples were digested in acid to analyze the individual components.
For NMR analysis, ∼5 mg of modified MOF was added to an NMR
tube followed by 0.8 mL of dimethylsulfoxide (DMSO) and 3 drops of
DCl. NMR tubes were vortexed briefly and sonicated for 1 min,
whereby the solution took on a fluorescent green hue. The material
was allowed to settle for 10 min, and 1H NMR spectra were acquired.
For mass spectrometry (MS) data, ∼1 mg of each modified sample
was added to a 2 mL Agilent vial followed by 1 mL of MeOH and 1
drop of HCl. The vials were shaken by hand until all particulate was
fully dissolved (∼5 min), and mass spectra were acquired using an
Agilent 6224 Accurate Mass TOF LC/MS with electrospray and
multimode (ESI/APCI) sources in negative or positive ion mode by
direct injection in methanol. For each sample, [M − H]− or [M + H]+
was detected for both the unmodified and modified ligands. Thermal
stability was evaluated using thermal gravimetric analysis (TGA)
acquired with a TGA Q500 V20.13 Build 39 instrument with a ramp
rate of 20 °C/min under an inert flow of nitrogen from 25 to 800 °C,
and analyzed with TA Universal Analysis. Sample mass was ∼1−5 mg.
Powder X-ray diffraction (PXRD) patterns were obtained using a
Bruker D8 Discover DaVinci Powder X-ray diffractometer with Cu Kα
radiation operated at 40 kV and 40 mA. A 0.6 mm divergent slit was
placed on the primary beam side, and a high-resolution energy-
dispersive LYNXEYE-XE-T detector was placed on the diffracted
B
DOI: 10.1021/acs.inorgchem.7b00373
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
a
Table 1. Percent Modification of Cu3(NH2BTC)2 with Anhydrides
% conversion
entry
MOF
1 equiv
10 equiv
20 equiv
40 equiv
1
2
3
4
5
6
7
Cu3(NH-AMiPr-BTC)2
Cu3(NH-AMiBu-BTC)2
Cu3(NH-AM4-BTC)2
Cu3(NH-AM5-BTC)2
Cu3(NH-AM6-BTC)2
Cu3(NH-AM7-BTC)2
Cu3(NH-AM10-BTC)2
6
11
14
7
1
0
1
4
2
1
1
10
28
31
20
20
18
10
6
1
1
6
5
2
1
11
37
43
25
27
22
11
2
0
0
8
7
2
1
20
43
50
32
32
26
13
6
0
3
3
1
4
1
13
9
6
a
All values are the average and standard deviation of three independent experiments.
a
Scheme 1. Synthesis of 2-Aminobenzene-1,3,5-tricarboxylic Acid (NH2H3BTC)
a
Reagents: (a) 1 (1.0 equiv), CH3COCl (1.05 equiv), TEA (1.05 equiv), DCM, 0 → 25 °C, 2 h, 72% yield; (b) 2 (1.0 equiv), KMnO4 (7.8 equiv), 1
M aq NaOH (0.5 equiv), H2O, 25 → 85°, 72 h; then 12 M aq HCl, pH = 1, reflux, 12 h, 40% yield; 30% overall yield. Abbreviations: TEA =
triethylamine; DCM = dichloromethane.
centrifuged, and the solution was decanted. This was repeated 2 times.
Once more CHCl3 was added and soaked overnight. After 12 h, the
vials were centrifuged and the solvent was decanted. The MOFs were
dried in open atmosphere for 12 h followed by 12 h in the oven at 100
°C. To remove residual chloroform, the material can be rinsed with
methanol as described above; however, rinsing with methanol bears no
impact on the hydrophobic effects observed, nor does it remove any
additional byproduct observed. All modifications were performed three
times using three independent Cu3(NH2BTC)2 samples. Various
batches of the MOF were used for each modification to confirm that
the percent conversions were independent of the different batches of
MOF used. The values given represent the average and standard
deviation of three independent trials. The percent conversions were
calculated by comparing the ratio of the aromatic chemical shifts
values given represent the average and standard deviation of three
independent trials.
Structural Stability in Water Studies. To test the stability of the
MOF in water, ∼10 mg of Cu3(NH2BTC)2, or modified MOF, was
added to a 20 mL glass vial. Slowly, 5 mL of Millipore water was
poured into each vial. Vials sat undisturbed for 30 min, after which
Kimwipes were used to absorb the water. The material was dried for 1
h in open atmosphere, followed by 1 h in the oven at 100 °C, and
immediately analyzed with PXRD and scanning electron microscope
(SEM) imaging. SEM imaging was performed using a JEOL JSM-
6500F microscope prior to and after water exposure. An accelerating
voltage of 15.0 kV and a working distance of 10.0 mm were used. All
samples were placed under vacuum and coated with 10 nm gold prior
to imaging. Five to six representative images were taken at three
different magnifications for each sample.
1
associated with the unmodified and modified ligands in the H NMR
spectra of each independent acid-digested material.
Wettability Studies. Contact angle goniometry (CAG) was
performed to analyze the interaction of the materials with water
Data Analysis. All PSMs were performed three times using three
independent Cu3(NH2BTC)2 samples. Various batches of the MOF
were used for each modification to confirm that the percent
conversions were independent of the different batches of MOF
used. The percent conversions were calculated by comparing the ratio
of the aromatic chemical shifts associated with the unmodified and
using a Kruss DSA30 goniometer with ultrapure water. Approximately
̈
10 mg of each MOF powder was gently ground and added to a glass
slide. Using another glass slide, the material was carefully pressed to 2
mm in height. Videos were collected at 25 frames per second for 10 s.
All water contact angles (WCAs) were determined with a 6−8 μL
droplet of water, whereby contact angles were calculated at 1 s after
deposition. For less hydrophobic materials, measuring a static WCA
was not possible. Therefore, water sorption rates were determined
using 1 μL droplets of water. The time at which the water droplet
made contact with the material was subtracted from the time at which
the entire water droplet went into the material. This value in
milliseconds was used to determine the rate that 1 μL took to sorb
into the material and is reported as the sorption rate in μL/s. The
1
modified ligands in the H NMR spectra of each independent acid−
digested material. The data reported in Table 1 represent the average
of the three modified samples and the standard deviation associated
with those averages. In the CAG experimentation, the standard
deviation shown is associated with the average of the three separate
rates acquired for the sample. When contact angles could be
determined, 6−8 μL droplets were used. The contact angles reported
are the average of three independent droplet contact angles after 1 s of
water deposition, and the standard deviations reported are the
respective calculated standard deviations. Contact angles were plotted
versus the number of carbons in the PSM amide modification. A linear
C
DOI: 10.1021/acs.inorgchem.7b00373
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
fit was plotted with an equation of y = 5.7x + 89.5, R2 = 0.97. Each data
point represents the average of three trials. Error bars represent the
standard deviation of 3 separate water droplets. By inserting a desired
number of carbons in the carbon chain length for x, one can solve for
the observed contact angle, y. Alternatively, if a desired contact angle is
inserted for y, an estimated number of carbons needed in the amide
chain can be deduced by solving for x.
successive rinsing and drying steps afforded the modified
MOFs. Little to no effect on conversion was achieved when
increasing equivalents beyond 40, increasing reaction time, or
increasing the temperature of the reaction. The masses [M −
H]− of both the unmodified and modified linkers were
abundant in the digested mass spectrum of each modification,
which confirmed mixed-ligand functionalization. The appear-
1
ance and intensity of a chemical shift within the H NMR
RESULTS AND DISCUSSION
■
spectra of the digested MOF samples at 8.39 ppm, indicative of
the aromatic protons on the modified amino tricarboxylate
ligands, were identified (Figure 1). The unmodified ligand
Synthesis and Characterization. Having direct and
efficient access to a benzene-1,3,5-tricarboxylate ligand
containing a functional group for late-stage modification
enables the possibility to modify isostructural MOFs.
Tricarboxylate ligands with useful functional handles (such as
primary amines) have limited availability. An amine functional
handle is particularly attractive from a PSM perspective because
of its susceptibility to an assortment of potential chemical
modifications.14,22 The only existing preparation of the amine-
functionalized tricarboxylate ligand 2-aminobenzene-1,3,5-
tricarboxylic acid (NH2H3BTC) involves an expensive synthesis
with 2-fluoromesitylene ($45.70/g AA) achieved in low to
moderate yields.20,23 Thus, initial efforts were focused on
devising a more efficient synthetic route. Protection and
deprotection steps are minimized in our 2-step approach
followed by an HCl workup to 3 in 30% overall yield on a gram
scale (Scheme 1). Using 2,4,6-trimethylaniline 1 decreased the
cost of synthesis more than 120-fold as compared to previous
reports ($0.36/g AA).23
Figure 1. 1H NMR spectra after acid digestion with DCl in DMSO-d6
of modified Cu3(NH2BTC)2 with valeric anhydride to confirm
synthesis of Cu3(NH-AM5-BTC)2. All samples were dried at 100 °C
for 12 h and digested in 0.8 mL of DMSO with 2 drops of DCl. The
unmodified ligand peaks are denoted with a black square and appear
slightly downfield of the modified ligands denoted with red circles.
The amine was first protected as an amide using acetyl
chloride and subsequently oxidized with KMnO4 over 3 days.
After filtering to remove unwanted oxidized manganese
products, the desired product was simultaneously deprotected
and protonated in a hydrochloric acid workup. Upon cooling,
the product crystallized as a pure white powder and was
isolated by filtration. This approach was found to be a
remarkably straightforward method of accessing this com-
pound. NH2H3BTC was then used for the synthesis of the
MOF Cu3(NH2BTC)2 under hydrothermal conditions in
DMF.20 The resultant green crystalline material was rinsed
thoroughly and solvent exchanged with ethanol. Thermal
activation of the material results in a much darker green color
(almost black). Overall this route to NH2H3BTC requires the
use of minimal purification steps, reduces the number of
protection and deprotection steps, and is accomplished with
significant economical improvements, yielding an attractive
efficient synthesis of the ligand and the MOF Cu3(NH2BTC)2.
This approach enables practical exploration of Cu3(NH2BTC)2
for postsynthetic modification including amidation, alkylation,
diazonium salt formation, enamine formation, and carbamate
formation, among others.
Postsynthetic Modification. PSM of Cu3(BTC)2 is
limited to dative modification and modification of the
carboxylate moiety.24−26 There is one example of covalent
PSM with the isostructural system Cu3(NH2BTC)2 that
showed Cu3(NH2BTC)2 can successfully undergo modification
with acetic anhydride, while maintaining porosity.20 Thus, the
present study aimed to modify Cu3(NH2BTC)2 with a variety
of anhydrides containing inherent hydrophobic moieties to
confer greater hydrophobicity and altered wettability.
1
The appearance and intensity of a chemical shift within the H NMR
spectra of the digested MOF samples at 8.39 ppm, indicative of the
aromatic protons on the modified amino tricarboxylate ligands, were
identified. Aromatic region (8−9 ppm) was used to quantify percent
conversion to the desired amides. As the equivalents of anhydride
increase, so does the percentage of modification, denoted by the peak
in the aromatic region with the red circle.
peaks are denoted with a black square and appear slightly
downfield of the modified ligands denoted with red circles. The
two peaks in the aromatic region were used to quantify the
conversion ratio of modified to unmodified ligands after PSM.
Reproducible control of a range of percent conversions for
each different anhydride used was accomplished (Table 1). As
the carbon chain of the PSM amide increases in length for
linear aliphatic chains, a decrease in the maximum percent
conversion is observed. Overall, this trend is consistent with
reports of PSM with NH2BDC; however, these reports indicate
that amide chains with 8 or less carbons led to near complete
conversion (>90%). In our study with NH2BTC, chains as low
as 4 carbons in length result in limited maximum conversions of
50%. Modifications with branched chain functionalities do not
seem to fit this trend. We hypothesize that these observations
may be due to steric hindrance, preventing access and diffusion
of the anhydrides used for these reactions to the intended
modification sites.
Complete conversion of Cu3(NH2BTC)2 with anhydrides
containing carbon lengths of 4 or more does not appear to be
possible; however, this provides a unique opportunity to
control mixed-ligand ratios within this framework and may be
attributable to the decrease in disorder of the BTC system due
to the increased substitution of the ligand. This is not often the
case with mixed-ligand syntheses, wherein the ratio of mixed
In each PSM procedure, Cu3(NH2BTC)2 was soaked in
chloroform overnight and subsequently dosed with 1, 10, 20, or
40 equiv of the desired anhydride with respect to the amount of
free amine (Table 1). After 24 h, solvent was removed and
D
DOI: 10.1021/acs.inorgchem.7b00373
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
ligands added to form the desired MOF does not predictably
impact the percent incorporation of each ligand present in the
resultant MOF formulation.27 The ability to reliably access
mixed-ligand frameworks yields materials that have multiple
distinctive pore environments that exhibit different chemical
properties all within the same material.28 This type of control
could permit multiselective gas adsorption (among other
multifunctional applications), due to the ability to tune a
percentage of the ligands, and by default affect the properties of
a desired percentage of pore sites within the same material.29
Conversion limitations may also be advantageous for the
retention of MOF crystallinity. Using acetic anhydride (2
carbons in amide), up to 92% conversion can be achieved;
however, after 73% conversion, substantial loss of crystallinity is
observed.20 In this study, when using butyric anhydride with a
4-carbon chain, the maximum conversion is limited to 50% and
the same degradation concerns are not observed. A close
examination of the 1H NMR spectra obtained from the digested
modified MOFs shows that minor peaks are present, consistent
with carboxylic acid byproduct generated during the formation
of the amide. The presence of this byproduct may suggest that
it is simply trapped within the MOF. Alternatively, the
byproduct may be coordinating to the copper metal ions,
replacing the labile ethanol ligands.30 This provides insight into
a plausible mechanism for modification-related degradation.
Nevertheless, even at maximum levels of modification, thermal
gravimetric analysis indicates that the modified materials in this
study maintain thermal stability as compared with the
unmodified material (Supporting Information, Figure S14).
Additionally, PXRD reveals crystallinity is retained after each
modification with 40 equiv of anhydride to amine as evidenced
by patterns having matching 2θ reflections (Figure 2). Thus, we
the material, leading to improved water stability. Contact angle
goniometry (CAG) has previously been used with PSM to
show hydrophobic and super hydrophobic properties of zinc
MOFs.16 We expanded upon this technique to use CAG to not
only observe contact angles achieved after maximum
modification but also compare the sorptive rates of the
materials at varying amounts of conversion.
The material was gently ground and pressed using a glass
slide to nearly 2 mm thick onto a separate glass slide. Water
sorption rates were calculated using 1 μL droplets, and the total
time necessary to completely sorb 1 μL of water into the
powder was used to determine a relative sorption rate for the
resultant material after minimum and maximum conversions.
As samples increased in hydrophobic character, water droplets
of 1 μL would not deposit on the surface of the material; for
these samples, 6−8 μL water droplets were dispensed and
contact angles were calculated 1 s after the droplet made
contact with the MOF material. Table 2 shows the results of
the CAG experimentation with three averaged independent
water droplet experiments.
The hydrophobicity of the material closely correlates to the
modification and overall percent of modification.
Cu3(NH2BTC)2 exhibits a similar sorption rate as
Cu3(BTC)2 at 8.8 μL/s (Table 2). All of the modified material
has slower sorption rates than the unmodified material,
demonstrating that all of the modifications made increase the
hydrophobic character of the MOF, albeit to different extents.
For instance, Cu3(NH-AMiBu-BTC)2 with 11% conversion
results in nearly half the sorption rate of the unmodified
material, sorbing water at a rate of 4.1 1.6 μL/s. Additionally,
nearly all materials exhibit greater hydrophobicity with slower
sorption rates or contact angles achieved using 40 equiv of
anhydride versus 1 equiv. Materials displaying contact angles
from 90°−150° are considered hydrophobic.31 Maximum
percent modifications with 5 or more carbons in the alkyl
chain result in contact angles of 95°−147°, rendering these new
mixed-ligand systems hydrophobic (Table 2, Figure 3C and D).
Upon increasing modification to nearly 32% with valeric
anhydride, sorption rates are significantly decreased so that,
after 1 s, a water contact angle of 118
4° is achieved.
Unexpectedly, as little as 6% modification with decanoic
anhydride led to a contact angle of 66 6°. At 13% conversion
an impressive hydrophobic effect was achieved where the
material exhibited an observed contact angle of 147 6°.
Over the course of these experiments, several intriguing
trends unique to this system were identified. Increased
hydrophobicity and slower sorption rates are achieved with
straight-chain versus branched moieties. This holds true for
both 4- and 5-carbon modifications, such as valeric versus
isovaleric and butyric versus isobutyric anhydrides. The
modified samples displayed distinctive hydrophobic properties,
whereby the MOF powder spread over the entire surface of the
water droplet during CAG experiments (Figure 3C). This may
indicate a change in adhesion properties or electrostatic charge
in the modified materials. These observations may be a
consequence of changes in mechanical interlocking, interparti-
culate forces of cohesion, liquid bridging or moisture-induced
particle adhesion, or reduced particle stiffness.32−35 These
properties are not observed with the unmodified materials,
further supporting the capability of late-stage functionalization
to tune the chemical properties of the material. Super-
hydrophobic materials are characterized by contact angles
>150°; accordingly, this suggests that modifying
Figure 2. Powder X-ray diffraction patterns for unmodified
Cu3(NH2BTC)2 (black) and all modified samples with 40 equiv of
anhydride, showing little to no loss of crystallinity after modification.
All samples were dried at 100 °C, gently ground, and pressed onto a
zero diffraction silicon plate.
have developed a covalent PSM approach with Cu3(NH2BTC)2
that allows reliable access to mixed-ligand frameworks having
conversions ranging from 6−50% with a retention of structural
integrity.
Enhanced Wettability. We anticipated that modifying
Cu3(NH2BTC)2 would increase the overall hydrophobicity of
E
DOI: 10.1021/acs.inorgchem.7b00373
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
a,b
Table 2. Effect of PSM on Wettability
entry
MOF
% conversion
0
sorption (μL/s)
8.8
contact angle
1
2
Cu3(NH2BTC)2
0
n/a
Cu3(NH-AMiPr-BTC)2
Cu3(NH-AMiBu-BTC)2
Cu3(NH-AM4-BTC)2
Cu3(NH-AM5-BTC)2
Cu3(NH-AM6-BTC)2
Cu3(NH-AM7-BTC)2
Cu3(NH-AM10-BTC)2
6
1
1.8 0.1
1.5 0.3
4.1 1.6
n/a
n/a
20
11
43
14
50
7
6
0
0
1
3
n/a
3
4
5
6
7
8
n/a
94.9 1.3
n/a
7.8 1.9
1.02 0.4
0.46 0.1
n/a
n/a
4
n/a
32
13
32
9
3
2
1
118.2 0.9
n/a
0.63 0.2
n/a
126
n/a
127
66
5
6
6
1
1.21 0.2
n/a
26
6
4
1
n/a
6
13
1
n/a
147
a
Measurements are the average of three independent experiments. Percent conversions shown for each modification are for modifications using 1
b
equiv and 40 equiv of anhydride to free amine, and the average and standard deviation of the three trials are reported. For samples in which the
water droplet was completely taken into the material after 1 s of deposition, a WCA could not be determined; therefore, sorption rates are reported
for these samples.
Figure 3. (A) PSM of the MOF with various anhydrides. (B) 3D image showing the MOF and potential sites for modification (carbon, black;
hydrogen, white; oxygen, red; nitrogen, blue; and copper, green). (C) CAG images of water on the surface of the MOF showing tunable wettability
with different modifications and percentages. (D) Water contact angle increases as the number of carbons in the linear chain increase. Values and
standard deviations shown are the average of three independent experiments.
Cu3(NH2BTC)2 can yield near superhydrophobic powders.
These data indicate that modifications have a relatively
significant impact on the hydrophobicity of this material,
resulting in varied sorption rates.
When the material is modified at maximum conversion, a
linear relationship is observed where the contact angle increases
in proportion to the number of carbons in the amide (Figure
3D). This allows for deliberate modification to access the
desired level of hydrophobic character. This observation may
suggest either the entire surface or a particular pore is modified
on the MOF, leading to predictable contact angles. Using the
linear fit, one or two more carbons to the alkyl chain may
produce superhydrophobic material with calculated contact
angles of ∼152−158°. This is the first report of using late-stage
functionalization with a metal−organic framework that allows
for control of the hydrophobic properties of the material.
Efforts to manipulate materials to attain tunable degradation
and release is especially appealing for the delivery of
therapeutics, among other biomedical applications.36
Enhanced Stability to Water. Because this is the first
report of synthetic functionalization of a copper MOF to
control hydrophobicity, we probed the limitations of water
F
DOI: 10.1021/acs.inorgchem.7b00373
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Figure 4. Results after water submersion. (A) (left to right) Cu3(NH2BTC)2, Cu3(NH-AMiBu-BTC)2, and Cu3(NH-AM10-BTC)2 in vials with 5
mL of Millipore water. (B) Powder X-ray diffraction patterns of Cu3(NH-AM10-BTC)2 before and after 30 min of submersion in water. SEM images
of Cu3(NH2BTC)2 (C) before and (d) after water submersion; SEM images of Cu3(NH-AMiBu-BTC)2 (E) before and (F) after water submersion.
SEM images of Cu3(NH-AM10-BTC)2 (G) before and (H) after water submersion. All images shown are representative of 4,000× (small) and
20 000× (large) magnifications; the black bar represents 1 μm.
stability with the modified MOFs. Cu3(NH2BTC)2 and
Cu3(NH-AMiBu-BTC)2 and Cu3(NH-AM10-BTC)2 with
maximum conversions were added to three separate 20 mL
glass vials, and 5 mL of water was added to each for 30 min.
The modified materials did not remain at the bottom of the vial
and were dispersed around and at the surface of the water,
demonstrating enhanced hydrophobic character (Figure 4A).
This was especially apparent with Cu3(NH-AM10-BTC)2
(Figure 4A, far right). Upon removal of the water and adequate
drying, the powder was analyzed. The PXRD pattern of
Cu3(NH-AM10-BTC)2 showed that a crystalline material was
present after water submersion (Figure 4B). Conversely,
Cu3(NH2BTC)2 no longer exhibited well-resolved peaks in
the diffraction pattern after water submersion. These results
demonstrate that modifying the MOF with 14% decanoic
anhydride renders this material significantly more stable in
aqueous environments.
Cu3(BTC)2 is known to rapidly degrade in water-rich
environments.10 As such, we would expect similar degradation
observations with unmodified Cu3(NH2BTC)2. We evaluated
this by analyzing the MOF’s surface characteristics before and
after water submersion. SEM imaging shows that
Cu3(NH2BTC)2 is comparable to Cu3(BTC)2 with regard to
surface deterioration upon water submersion (Figure 4C and
D). Prior to water submersion, the surface of Cu3(NH2BTC)2
appears smooth and uniform. After submersion there is a
significant change to the morphology of the MOF; pores are
expanded and the original octahedral crystals are nearly
impossible to distinguish. Cu3(NH-AMiBu-BTC)2 still appears
to undergo minor surface alteration, exhibiting a more roughly
textured morphology with breaks and cracks, albeit much less
severe than in the unmodified material (Figure 4E and F).
Cu3(NH-AM10-BTC)2, on the other hand, showed little to no
impact to the surface characteristics after being submerged in
water (Figure 4G and H). PXRD patterns of Cu3(NH-AM10-
BTC)2 and Cu3(NH2BTC)2 on the material after water
submersion verified that the modified material retained
crystallinity after water submersion (Supporting Information,
Figure S23). The water submersion results indicate that tunable
degradation rates are achievable by modifying Cu3(NH2BTC)2.
These results suggest that the stability of the material correlates
with the relative amount of hydrophobicity observed. This
technique can be used with other modifications made herein to
assess stability to water after modification. Taken together,
these data showcase the elevated hydrophobic character of the
modified material and its immunity to moisture -amaging
effects present with Cu3(BTC)2.
CONCLUSIONS
■
In summary, we have determined that modifying a sterically
hindered MOF like Cu3(NH2BTC)2 serves as an excellent
approach to improve functionality of these materials. In this
study, we designed a new synthesis of 2-aminobenzene-1,3,5-
tricarboxylic acid affording a rapid, inexpensive route to MOF
Cu3(NH2BTC)2. This accessibility allowed for late-stage
functionalization with a series of anhydrides containing greasy
aliphatic hydrocarbon chains. Importantly, controlled access to
new mixed-ligand copper MOFs that are isostructural to
Cu3(NH2BTC)2 can now be obtained without affecting the
overall stability of the system. Lastly, these modifications result
in tunable changes to the chemical and physical properties of
the system, specifically herein leading to control over the
stability of the MOF to water, hydrophobic properties, and
wettability. The results of this study now expand the methods
to covalently PSM MOFs to include copper MOFs and also
open the possibilities to PSM other BTC ligand-containing
MOFs. This work foreshadows new capabilities for the same
finely tuned hybrid material to be employed in many diverse
applications, including biological, gas storage, and electronics,
previously impeded by instability to water.
G
DOI: 10.1021/acs.inorgchem.7b00373
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
(11) Sun, K.; Li, L.; Yu, X. L.; Liu, L.; Meng, Q.; Wang, F.; Zhang, R.
J. Colloid Interface Sci. 2017, 486, 128.
ASSOCIATED CONTENT
■
S
* Supporting Information
(12) DeCoste, J. B.; Peterson, G. W.; Schindler, B. J.; Killops, K. L.;
Browe, M. A.; Mahle, J. J. The effect of water adsorption on the
structure of the carboxylate containing metal−organic frameworks Cu-
BTC, Mg-MOF-74, and UiO-66. J. Mater. Chem. A 2013, 1, 11922−
11932.
(13) Burrows, A. D.; Frost, C. G.; Mahon, M. F.; Richardson, C.
Post-Synthetic Modification of Tagged Metal−Organic Frameworks.
Angew. Chem., Int. Ed. 2008, 47, 8482−8486.
(14) Garibay, S. J.; Wang, Z.; Tanabe, K. K.; Cohen, S. M.
Postsynthetic modification: a versatile approach toward multifunc-
tional metal-organic frameworks. Inorg. Chem. 2009, 48, 7341−7349.
(15) Tanabe, K. K.; Wang, Z.; Cohen, S. M. Systematic
Functionalization of a Metal-Organic Framework via a Postsynthetic
Modification Approach. J. Am. Chem. Soc. 2008, 130, 8508−8517.
(16) Nguyen, J. G.; Cohen, S. M. Moisture-Resistant and
Superhydrophobic Metal-Organic Frameworks Obtained via Post-
synthetic Modification. J. Am. Chem. Soc. 2010, 132, 4560−4561.
(17) Wittmann, T.; Siegel, R.; Reimer, N.; Milius, W.; Stock, N.;
Senker, J. Enhancing the water stability of Al-MIL-101-NH2 via
postsynthetic modification. Chem. - Eur. J. 2015, 21, 314−323.
(18) Rodriguez-Hermida, S.; Tsang, M. Y.; Vignatti, C.; Stylianou, K.
C.; Guillerm, V.; Perez-Carvajal, J.; Teixidor, F.; Vinas, C.;
Choquesillo-Lazarte, D.; Verdugo-Escamilla, C.; Peral, I.; Juanhuix,
J.; Verdaguer, A.; Imaz, I.; Maspoch, D.; Giner Planas, J. Switchable
Surface Hydrophobicity−Hydrophilicity of a Metal−Organic Frame-
work. Angew. Chem. 2016, 128, 16283−16287.
(19) Sun, Q.; He, H.; Gao, W.-Y.; Aguila, B.; Wojtas, L.; Dai, Z.; Li,
J.; Chen, Y.-S.; Xiao, F.-S.; Ma, S. Imparting amphiphobicity on single-
crystalline porous materials. Nat. Commun. 2016, 7, 13300.
(20) Peikert, K.; Hoffmann, F.; Froba, M. Amino substituted
Cu3(btc)2: a new metal-organic framework with a versatile
functionality. Chem. Commun. 2012, 48, 11196−11198.
(21) Cai, Y.; Kulkarni, A.; Huang, Y.-G.; Sholl, D.; Walton, K.
Control of Metal−Organic Framework Crystal Topology by Ligand
Functionalization: Functionalized HKUST-1 Derivatives. Cryst.
Growth Des. 2014, 14, 6122−6128.
(22) Tanabe, K. K.; Cohen, S. M. Postsynthetic modification of
metal-organic frameworks–a progress report. Chem. Soc. Rev. 2011, 40,
498−519.
(23) Data accessed on March 20, 2016, at www.alfa.com and
reported as price per gram. All values are based on the lowest quantity
available.
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.inorg-
chem.7b00373.
All NMR spectra of the organic compounds; NMR and
MS spectra of the acid-digested material after PSM; TGA
showing thermal stability before and after modification;
SEM images of the material; additional CAG images;
pXRD of the modified and unmodified materials; IR
spectroscopy of the ligand and MOF materials; and 3D
drawings of the MOF (PDF)
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: melissa.reynolds@colostate.edu. Tel.: +1 970 491
3775.
ORCID
Melissa M. Reynolds: 0000-0002-1836-7324
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We acknowledge financial support from the National Science
Foundation (NSF/DMR) Career Grant CBET-1351909. We
thank Jesus Tapia for help acquiring mass spectrometry data,
Dr. Morgan Hawker and Michelle Mann for help with the
contact angle goniometer, and Dr. Patrick McCurdy for
assistance using the scanning electron microscope.
REFERENCES
■
(1) Wang, C.; Liu, D.; Lin, W. Metal−Organic Frameworks as A
Tunable Platform for Designing Functional Molecular Materials. J. Am.
Chem. Soc. 2013, 135, 13222−13234.
(2) Parola, S.; Julian-Lopez, B.; Carlos, L. D.; Sanchez, C. Optical
́ ́
Properties of Hybrid Organic-Inorganic Materials and their
Applications. Adv. Funct. Mater. 2016, 26, 6506.
(3) Li, J.-R.; Kuppler, R. J.; Zhou, H.-C. Selective gas adsorption and
separation in metal-organic frameworks. Chem. Soc. Rev. 2009, 38,
1477−1504.
(4) Furukawa, H.; Cordova, K. E.; O’Keeffe, M.; Yaghi, O. M. The
Chemistry and Applications of Metal-Organic Frameworks. Science
2013, 341, 1230444.
(5) Li, B.; Wen, H.-M.; Cui, Y.; Zhou, W.; Qian, G.; Chen, B.
Emerging Multifunctional Metal−Organic Framework Materials. Adv.
Mater. 2016, 28, 8819−8860.
(6) Guo, Z.-J.; Yu, J.; Zhang, Y.-Z.; Zhang, J.; Chen, Y.; Wu, Y.; Xie,
L.-H.; Li, J.-R. Water-Stable In(III)-Based Metal−Organic Frameworks
with Rod-Shaped Secondary Building Units: Single-Crystal to Single-
Crystal Transformation and Selective Sorption of C2H2 over CO2 and
CH4. Inorg. Chem. 2017, 56, 2188.
(24) Chui, S. S.-Y.; Lo, S. M.-F.; Charmant, J. P. H.; Orpen, A. G.;
Williams, I. D. A Chemically Functionalizable Nanoporous Material
[Cu3(TMA)2(H2O)3]n. Science 1999, 283, 1148−1150.
(25) Al-Janabi, N.; Deng, H.; Borges, J.; Liu, X.; Garforth, A.;
Siperstein, F. R.; Fan, X. A Facile Post-Synthetic Modification Method
to Improve Hydrothermal Stability and CO2 Selectivity of CuBTC
Metal−Organic Framework. Ind. Eng. Chem. Res. 2016, 55, 7941.
(26) Mohideen, M. I. H.; Xiao, B.; Wheatley, P. S.; McKinlay, A. C.;
Li, Y.; Slawin, A. M. Z.; Aldous, D. W.; Cessford, N. F.; Duren, T.;
̈
Zhao, X.; Gill, R.; Thomas, K. M.; Griffin, J. M.; Ashbrook, S. E.;
Morris, R. E. Protecting group and switchable pore-discriminating
adsorption properties of a hydrophilic−hydrophobic metal−organic
framework. Nat. Chem. 2011, 3, 304−310.
(7) Jeong, H.; Kim, J. Electrodeposition of Nanoflake Pd Structures:
Structure-Dependent Wettability and SERS Activity. ACS Appl. Mater.
Interfaces 2015, 7, 7129−7135.
(8) Sezginel, K.; Keskin, S.; Uzun, A. Tuning the Gas Separation
Performance of CuBTC by Ionic Liquid Incorporation. Langmuir
2016, 32, 1139−1147.
(27) Yin, Z.; Zhou, Y.-L.; Zeng, M.-H.; Kurmoo, M. The concept of
mixed organic ligands in metal−organic frameworks: design, tuning
and functions. Dalton Trans. 2015, 44, 5258−5275.
(28) Chen, D.-M.; Xu, N.; Qiu, X.-H.; Cheng, P. Functionalization of
Metal−Organic Framework via Mixed-Ligand Strategy for Selective
CO2 Sorption at Ambient Conditions. Cryst. Growth Des. 2015, 15,
961.
(29) Burrows, A. D. Mixed-component metal-organic frameworks
(MC-MOFs): enhancing functionality through solid solution for-
mation and surface modifications. CrystEngComm 2011, 13, 3623−
3642.
(9) Burtch, N. C.; Jasuja, H.; Walton, K. S. Water stability and
adsorption in metal-organic frameworks. Chem. Rev. 2014, 114,
10575−10612.
(10) Horcajada, P.; Serre, C.; Vallet-Regí, M.; Sebban, M.; Taulelle,
F.; Fer
́
ey, G. Metal−Organic Frameworks as Efficient Materials for
Drug Delivery. Angew. Chem., Int. Ed. 2006, 45, 5974.
H
DOI: 10.1021/acs.inorgchem.7b00373
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
(30) Bordiga, S.; Regli, L.; Bonino, F.; Groppo, E.; Lamberti, C.;
Xiao, B.; Wheatley, P. S.; Morris, R. E.; Zecchina, A. Adsorption
properties of HKUST-1 toward hydrogen and other small molecules
monitored by IR. Phys. Chem. Chem. Phys. 2007, 9, 2676−2685.
(31) Hawker, M. J.; Pegalajar-Jurado, A.; Fisher, E. R. Innovative
Applications of Surface Wettability Measurements for Plasma-
Modified Three-Dimensional Porous Polymeric Materials: A Review.
Plasma Processes Polym. 2015, 12, 846−863.
(32) Lumay, G.; Boschini, F.; Traina, K.; Bontempi, S.; Remy, J. C.;
Cloots, R.; Vandewalle, N. Measuring the flowing properties of
powders and grains. Powder Technol. 2012, 224, 19−27.
(33) Gregory, J. Particles in water properties and processes; IWA Pub
Taylor & Francis: London, 2006.
(34) Sandler, N.; Reiche, K.; Heinamaki, J.; Yliruusi, J. Effect of
̈
̈
Moisture on Powder Flow Properties of Theophylline. Pharmaceutics
2010, 2, 275.
(35) Armstrong, B.; Krockbank, K.; Clayton, J. Understand the
Effects of Moisture on Powder Behavior. AIChE J. 2014, 110, 25−30.
(36) Sankaranarayanan, J.; Mahmoud, E. A.; Kim, G.; Morachis, J.
M.; Almutairi, A. Multiresponse Strategies to Modulate Burst
Degradation and Release from Nanoparticles. ACS Nano 2010, 4,
5930−5936.
I
DOI: 10.1021/acs.inorgchem.7b00373
Inorg. Chem. XXXX, XXX, XXX−XXX