Efficient reductive amination of HMF with well dispersed Pd nanoparticles immobilized in a porous MOF/polymer composite

Article abstract of DOI:10.1039/c9gc03140e

Aminated derivatives of 5-hydroxymethylfurfural (HMF) and furfural are critical intermediates for the pharmaceutical industry. The state-of-the-art catalysts currently used for these syntheses are mostly homogeneous in nature, motivating the design of rec

Full text of DOI:10.1039/c9gc03140e

Green Chemistry
View Article Online
View Journal
PAPER
Efficient reductive amination of HMF with well
dispersed Pd nanoparticles immobilized in a
porous MOF/polymer composite†
Cite this: DOI: 10.1039/c9gc03140e
a,b
Vikram V. Karve,a Daniel T. Sun,a Olga Trukhina,a Shuliang Yang,a Emad Oveisi,
a
Jeremy Luterbacher a and Wendy L. Queen
*
Aminated derivatives of 5-hydroxymethylfurfural (HMF) and furfural are critical intermediates for the
pharmaceutical industry. The state-of-the-art catalysts currently used for these syntheses are mostly
homogeneous in nature, motivating the design of recyclable, heterogeneous catalytic systems. As such,
the present study illustrates a new method for the design of metal–organic framework (MOF)/polymer
composites containing well-defined metal nanoparticles in a sustainable way. One such palladium func-
tionalized MOF/polymer composite is then employed in the reductive amination of HMF under mild con-
ditions. The catalyst shows excellent activity, including a high TON/TOF (h−1) of 604.8/302.4 and ∼94%
amine yield, which is maintained over a larger number of reaction cycles (up to 15) when compared to
several state-of-the-art materials, such as a commercial Pd/C (3 cycles). It is thought that the origin of the
improved catalyst recyclability is due to the added polymer, poly-para-phenylenediamine (PpPDA), which
helps to prevent the aggregation and leaching of the palladium nanoparticles. The synthetic approach is
further extended to design other potential catalysts with different metallic nanoparticles (NPs).
Received 6th September 2019,
Accepted 8th November 2019
DOI: 10.1039/c9gc03140e
rsc.li/greenchem
that is often found in molecules10–13 that are used to treat a
number of disorders (Fig. 1).13
Introduction
Fossil fuels are the primary source of many important chemi-
One of the first reports on the reductive amination of HMF
cals; however, their impending depletion and considerable describes a two-step, catalyst free conversion strategy wherein
impact on climate change,1 has spurred intense research the HMF is first condensed with a primary amine under an-
towards replacing hydrocarbons with renewable sources.2 hydrous conditions and then subsequently reduced by quanti-
Ligno-cellulosic biomass is one such resource that has tative amounts of NaBH4 producing 5-alkyl- and 5-arylamino-2-
attracted significant attention due to its high abundance.3 furfuryl alcohols.14 Subsequently, the selectivity and conver-
Further, it could be employed to produce a variety of platform sion of the same reaction was improved with the use of a
chemicals, such as 5-(hydroxymethyl)-furfural (HMF) and dichlorobis(2,9-dimethyl-1,10-phenanthroline) ruthenium(II)
furfural, that can potentially serve as secondary sources for the catalyst.15 Although this catalyst performs well for the conver-
production of fine organic chemicals.4 This prospect has
recently motivated much research devoted to studying the con-
version of biomass derived HMF and furfural into a variety of
value-added products.6–9 Interestingly, HMF and furfural are
important biomass derived intermediates which can be trans-
formed into a variety of aminated compounds that are relevant
to the pharmaceutical industry. This is due to the furanmetha-
namine moiety in the aminated derivative, a building block
aInstitute of Chemical Sciences and Engineering (ISIC), École Polytechnique Fédérale
de Lausanne (EPFL), CH-1051 Sion, Switzerland. E-mail: wendy.queen@epfl.ch
bInterdisciplinary Center for Electron Microscopy, École Polytechnique Fédérale de
Lausanne (EPFL), CH-1015 Lausanne, Switzerland
†Electronic supplementary information (ESI) available. See DOI: 10.1039/
Fig. 1 Examples of biomolecules showcasing the aminoalkyl furan
functionality (in red).5
c9gc03140e
This journal is © The Royal Society of Chemistry 2019
Green Chem.

View Article Online
Paper
Green Chemistry
sion, its homogeneous nature inhibits its recyclability. In this arises from a large number of connection points between the
context, recyclable, heterogeneous catalysts are preferred. As metal node and ligand (12-connected Zr-oxo cluster) and the
such, since this time, there have been numerous reports on strength of the coordination bond found between the carboxy-
supported nanoparticle (NP) catalysts for the reductive amin- late and Zr4+ ion. Given these favorable attributes, UiO-67 was
ation of HMF.16–22 Unfortunately, a common shortfall encoun- first synthesized according to a previously reported procedure
tered in these catalysts is their rapid loss of activity, a direct with slight modifications (see Experimental section). When
result of aggregation and leaching of the active NPs. The comparing the powder X-ray diffraction (PXRD) pattern
design of new catalysts able to resist NP aggregation and leach- obtained from the as-synthesized material and the simulated
ing will result in longer catalyst lifetimes, thereby making the diffraction pattern of UiO-67 (Fig. 2a), the resulting white
catalysis more economically and environmentally viable.
powder is confirmed to be a highly crystalline, phase pure
Among a plethora of different materials, the choice to sample of UiO-67. Further, N2 adsorption measurements,
employ MOFs as NP supports stems from their exceptionally carried out at 77 K, reveal a type I adsorption isotherm with two
high internal surface areas,23–25 which provide easy accessibil- steps that correspond to the two types of aforementioned cages
ity to active sites. Further, MOFs have a tunable pore (Fig. 2b) and a BET surface area of ∼2500 m2 g−1
.
environment26,27 that allows the easy introduction of reactive
Throughout the literature, the most common shortfall
functionalities meant to selectively facilitate a desired catalytic encountered in the reductive amination of HMF using hetero-
reaction.28,29 Recent research, demonstrates that the introduc- geneous catalysts is the aggregation and leaching of the active
tion of polymer guests into MOFs can improve both mechani- NPs, which results in a rapid loss of catalytic activity.16–22,46 In
cal stability and internal surface area.30–32 Further, separate previous works, we incorporated a variety of polymers into
studies have demonstrated that porous organic polymers, several MOF templates.47–49 While that work was focused on
having high densities of Lewis base functionality, can help to enhancing the separation properties of a given MOF, the high
inhibit the aggregation of metallic NPs.33,34 Considering these density of exposed Lewis basic sites on the polymer backbone,
aforementioned points, and the prominence of MOFs in the led us to employ a similar composite for catalysis. The idea
field of heterogeneous catalysis,35 we were inspired to develop was that the Lewis basic sites might help immobilize NPs
a new strategy to design MOF/polymer composites loaded with inside the MOF inhibiting aggregation/leaching during reac-
metallic NPs. For this, the MOF was first synthesized and then tions. As such, we developed a new, one-step in situ polymeriz-
in situ polymerization was carried out using palladium nitrate ation method to simultaneously generate both the polymeric
as a precursor. The MOF template provides extrinsic porosity species and the Pd NPs into the as-prepared UiO-67. After the
to the intrinsically non-porous polymer, poly para-phenylene modification, the resulting composite, denoted as UiO-67/
diamine (PpPDA), which contains a high-density of amines PpPDA/Pd, retains crystallinity (Fig. 2a) and a high BET surface
able to coordinate Pd. Further, the selected MOFs offer accessi- area of ∼1460 m2 g−1 (Fig. 2b). Pore size distribution analysis
ble acidic functionality like Zr4+, Fe3+ or hydroxyl groups on
their internal surface that can promote a dehydration reaction
and hence the formation of imines that can subsequently
undergo reduction to an amine by H2 adsorbed onto the Pd
NPs. We hypothesized that the short chain PpPDA polymers
inserted into the MOF structure would help to inhibit NP
aggregation inside the support. The resulting catalyst shows
good performance in the reductive amination of HMF with a
host of substrates containing primary or secondary amines
(with both electron withdrawing and donating functionalities)
under mild reaction conditions. We surmise that the polymer
promotes the appendage of aggregation-stable palladium
nanoparticles within the MOF crystallites, which we demon-
strate significantly improves stability of the composite when
compared to several other heterogeneous catalysts.
Results and discussion
Material design and characterization
Fig. 2 Characterization of UiO-67/PpPDA/Pd. (a) X-ray diffraction pat-
UiO-67, [Zr6(μ3-O)4(μ3-OH)4(BPDC)6],36–40 is the point of ingress
terns of the bare framework (red), the composite (blue) and the simu-
lated pattern for UiO-67 (black). (b) N2 adsorption isotherms at 77 K for
the bare framework (red), polymer modified framework (violet), polymer
and nanoparticle modified framework (blue) and the polymer-nano-
for various MOF-based catalysts owed to its high surface area,
∼2500 m2 g−1, the presence of large tetrahedral and octahedral
cages with diameters of 1.2 nm and 2.3 nm, respectively, and
particle composite (orange). X-ray photoelectron spectroscopy analysis
exceptional chemical and thermal stability.41–45 The stability for the (c) N 1s and the (d) Pd 3d regions for UiO-67/PpPDA/Pd.
Green Chem.
This journal is © The Royal Society of Chemistry 2019

View Article Online
Green Chemistry
Paper
indicates that the polymer and Pd likely fill the largest octa-
hedral pore (Fig. S1†). The polymer loading in the composite
was determined to be 10 wt% by combustion analysis
(Table T1, see ESI†), and X-ray photoelectron spectroscopy
(XPS) indicates the presence of both nitrogen and Pd in the
composite (Fig. 2d). The latter are present in both metallic
(Pd0) and ionized form (Pd2+) in a ratio of 46% to 54%, respect-
ively, based on the XPS spectrum. Further, the 3d peak corres-
ponding to the Pd2+ species are shifted to a higher binding
energy, 337.8 0.6 eV (Fig. 2d), than expected, 336.7 eV, which
is thought to be related to electronic interactions between the
palladium and the nitrogen rich polymer.50 We suspect that
the Pd2+ species are coordinated to nitrogen, forming a protec-
tive shell around the metallic NPs. Further, the N 1s XPS spec-
trum reveals peaks that correspond to secondary and tertiary
amines (Fig. 2c). These N species, which are not present in the
MOF or the monomer, confirm the formation of polymeric
units in the composite. MALDI-TOF-MS measurements,
(Fig. S2†) carried out on the digested MOF composite, reveals
oligomeric units composed of up to 8 monomers. These short
chain lengths indicate that the MOF pores could play a role in
limiting the polymer growth. This idea is further supported by
additional polymerization experiments carried out without the
MOF template, an act that expectedly results in polymers with
slightly higher molecular weights with no peaks evident for
low molecular weight oligomers (Fig. S3†). Attenuated total
reflectance infrared (ATR-IR) spectroscopy (Fig. S4†) addition-
ally provides qualitative evidence of polymerization. A new
peak at 732 cm−1 corresponds to the quinoid C–N–C
stretch;51,52 which is not expected in the monomer. Next,
microscopy was used to visually inspect the resulting compo-
site. Scanning electron microscope (SEM) images (Fig. S5†),
clearly indicate that the composite has the same crystal mor-
phology as observed for the parent UiO-67 structure. The crys-
tallites have an average size of ∼200 nm and relatively smooth
surfaces, with no signs of polymer on the external MOF
Fig. 3 BF-STEM image (a) of UiO-67/PpPDA/Pd along with the corres-
ponding EDX maps for zirconium (b), palladium (c) and nitrogen (d)
respectively. The scale bars in each image correspond to 200 nm.
reveals a uniform distribution of the nitrogen and Pd within
the MOF. While this does not rule out the possibility of having
a small portion of Pd NPs on the external surface of UiO-67,
these experiments do indicate that there is a significant
amount of the active species throughout the inside of the MOF
template. For comparison, when the polymerization is carried
out without the MOF present, the resulting size of the NPs
inside the bulk polymer is much larger, 20 to 40 nm (Fig. S9†),
supporting the idea that the MOF structure has a confinement
effect on the growth of the Pd NPs.
Reaction condition optimization
surface. In addition to this, transmission electron microscopy To study the catalytic activity of the newly designed MOF-
(BF-TEM) images show a fairly uniform distribution of Pd polymer composite, the initial aim was to optimize the reac-
nanoparticles throughout the MOF (Fig. 3a). While the sizes of tion conditions for UiO-67/PpPDA/Pd in the reductive amin-
the NPs in the composite range from 1 to 9 nm, ∼60% of the ation of HMF with aniline. It was expected that the reaction
NPs are below 4 nm (Fig. S6†), and hence of suitable size for would proceed through the adsorption of HMF near a Lewis
the inside of the defective MOF structure. It is important to acidic Zr site and subsequent formation of an imine inter-
note that the Zr-MOFs, such as UiO-67, often exhibit missing mediate would occur upon the removal of a water molecule.
linkers.53 These defects can lead to changes in the local pore Next, the imine intermediate would be reduced by the hydro-
size leading to voids that are larger than those expected from gen adsorbed on the Pd NPs present in close proximity to the
the crystal structure. Given this, titration experiments were Zr sites throughout the framework. To evaluate this, the
used to investigate the existence of such linker defects selected model reaction was first kept at a constant 50 °C and
(Fig. S7†). The experiments indicate that the Zr-node found in a 5 bar H2 pressure while the conversion of HMF and yields of
the MOF structure is indeed missing ∼1.8 of the 6 expected the intermediate imine and final amine products were moni-
linkers, a total of 30%. To confirm the presence of both the Pd tored as function of time (Fig. 4a). HMF conversion to the
nanoparticles and the polymer inside the MOF support, imine increases almost linearly up-to 30 minutes where the
energy dispersive X-ray (EDX) analysis was also carried out on HMF conversion is completed. The imine product, which is
the composite (Fig. 3b–d, Fig. S8†). The solid powder was first formed through a condensation reaction between the amine
embedded in a resin and sliced using a microtome. Each slice group of the aniline and the aldehyde of HMF, is likely pro-
had a thickness of around 60 nm. The EDX analysis, carried moted by Brønsted acidic defect sites found in UiO-67.53 It is
out in a scanning transmission electron microscope (STEM), noted that this intermediate is then totally converted to the
This journal is © The Royal Society of Chemistry 2019
Green Chem.

View Article Online
Paper
Green Chemistry
temperature were held constant at 2 hours and 50 °C, respect-
ively. At low hydrogen pressures of 1 to 4 bar, there is incom-
plete conversion from the imine to the final amine product. At
5 bar, the imine shows complete conversion, and hence this
pressure was selected as optimal. Finally, the conversion of
HMF and the reaction products were measured as a function
of temperature (Fig. 4c), with the H2 pressure and time held at
a constant 5 bar and 2 hours, respectively. Under these con-
ditions, complete conversion into the desired amine product
is indeed noted at 50 °C. Unless otherwise noted, the remain-
der of the reactions in this study were performed under the
optimized conditions: 5 bar H2 pressure, 2 hours, and 50 °C.
Under the selected reaction conditions UiO-67/PpPDA/Pd
exhibits a high performance for the model reaction with a
94.5% yield of the targeted amine product (Table 1, entry 5).
To prove that the catalysis was indeed heterogeneous in
nature, a hot extraction experiment was carried out. To do this,
the catalyst was removed from the reaction after 30 minutes
into the process (Fig. S10†). Afterwards, no further HMF or
imine conversion was observed indicating that the catalysis is
indeed heterogeneous. Another indication of the composites’
high catalytic activity is noted from Fig. 4c, which indicates
that appreciable amounts of the imine and amine products are
obtained even at room temperature. When the reaction time
was extended, total conversion to the targeted amine product
was observed after 10 hours at room temperature.
Evaluation of the catalytic performance under optimized
conditions
Firstly, it was necessary to demonstrate that the reductive
amination of HMF with aniline could not be achieved without
a catalyst or in the presence of the Pd-free UiO-67 framework
(Table 1, entries 1 and 2). Only UiO-67 revealed the formation
of the intermediate imine, due to the MOFs ability to promote
dehydration; however, the reaction showed no activity for the
targeted amine, even when the reaction time was prolonged to
10 hours (Table 1, entry 3). Next, to determine if the presence
of the polymer indeed can enhance catalyst performance, the
catalytic activity of UiO-67/PpPDA/Pd was compared to several
other Pd-containing catalysts without polymer. For this two
MOFs, namely UiO-67 and Fe-BTC, were prepared and sub-
sequently loaded with 0.40 and 0.47 wt% Pd NPs, respectively.
In addition to these, a state of the art, commercially available
carbon catalyst, containing 1 wt% Pd (Table 1, entries 4, 8 and
9) was purchased and tested. TEM images of UiO-67/Pd, Fe-
BTC/Pd and the commercial Pd/C reveal that the Pd NPs (see
methods) have sizes ranging from 4 to 9 nm (Fig. S11 and
S12†), a similar size and distribution as observed for UiO-67/
PpPDA/Pd. Next, PXRD was utilized to show that both UiO-67
and Fe-BTC frameworks remain crystalline after the addition
of Pd (Fig. S13†).
Fig. 4 Kinetic reaction profile as a function of (a) time, (b) hydrogen
pressure and (c) temperature for the condensation of HMF and aniline.
The black points indicate the conversion of HMF and the red and blue
points indicate the yields of the imine and the amine product respect-
ively. Reaction conditions: 10 mg UiO-67/PpPDA/Pd (0.4 wt% Pd
loading), 5-hydroxymethylfurfural (0.24 mmol), aniline (0.24 mmol),
5 mL ethanol.
Next, all materials were tested in the model reaction carried
out under the optimized conditions. After two-hours, the two
final amine product in 2 hours (Fig. 4a). These results indicate MOF/Pd samples and commercial Pd/C exhibit a high conver-
that the imine conversion to the amine is the rate limiting sion efficiency and high selectivity (Table 1, entries 4, 8 and 9)
step. Next, H2 pressure was varied (Fig. 4b), while time and as observed for UiO-67/PpPDA/Pd (Table 1, entry 5). This result
Green Chem.
This journal is © The Royal Society of Chemistry 2019

View Article Online
Green Chemistry
Paper
Table 1 Reaction of HMF with aniline under H2 using various catalysts using ethanol as the solvent
Entry
Catalyst
Pd loadinga (wt%)
Yieldb (%)
TON/TOFc (h−1
)
Cycles
1
2
3
4
5
6
7
8
9
No catalyst
—
—
—
0.4
0.4
—
0.62
0.47
1
<1%
<1%
<1%
93%
94.5%
<1%
<1%
95%
82%
—/—
—/—
—/—
595.2/297.6
604.8/302.4
—/—
—/—
608/304
524.8/262.4
—
—
—
3
15
—
UiO-67d
UiO-67e
UiO-67/Pdd
UiO-67/PpPDA/Pdd
UiO-67/PpPDAd
PpPDA/Pdd
Homogeneous
1
3
Fe-BTC/Pd
Commercial Pd/C
Reaction conditions: 5 bar H2, 50 °C, 10 mg catalyst (0.4 wt% Pd loading), 5-hydroxymethylfurfural (0.24 mmol), aniline (0.24 mmol), 5 mL
ethanol. a Pd loading determined by ICP-OES. b Determined by 1H NMR. c Turnover number (TON) and turnover frequency (TOF) calculated for
the optimized conditions for each substrate considering the total amount of Pd present in the catalyst as determined by ICP-OES. d Reaction time
of 2 h. e Reaction time of 10 h.
is expected, given that all materials have accessible Pd NPs
Given the significantly improved stability of the composite
having a similar size distribution (Fig. S6, S11 and S12†). catalyst when compared to heterogeneous catalysts used in
Despite similarities in the initial activity of the 4 catalysts, this study and others (Table T4, see ESI†), we carried out a
their differences can readily be seen with cycling. While more detailed investigation to assess the true stability of
UiO-67/Pd and Pd/C have improved longevity, they are found UiO-67/PpPDA/Pd. For this, the model reaction between HMF
to be catalytically inactive after 3 cycles (Table 1, entries 4 and and aniline was again used. This time, HMF consumption and
9, respectively). Further analysis via TEM indicates an increase rate of amine formation was monitored for each cycle at 40%
in NP size and
a decrease in number (Fig. S12†). conversion (Fig. 5). In between each cycle, the catalyst was
Unfortunately, for Fe-BTC/Pd, the catalyst is not able to tolerate thoroughly washed and then exposed to fresh reactants. For
the reaction conditions, and loses all activity after cycle 1 HMF consumption, the composite catalyst shows a steady
(Table 1, entry 8). TEM images taken afterwards indicate that decrease in the initial rate (Fig. 6a and Fig. S17†) between
there is a minimal amount of Pd remaining in the sample and cycle 1 and 15. From this, it was postulated that the decreasing
PXRD indicates complete degradation of the MOF structure HMF consumption could be due to the gradual degradation of
(Fig. S13†). Of the four tested catalysts, UiO-67/PpPDA/Pd, is the MOF support and/or catalyst loss upon filtration between
the best performing; it has high activity in the amination reac- each cycle. While we cannot rule out catalyst loss, PXRD pat-
tion up to cycle 15. While TEM images taken of the sample terns taken after every 4 cycles do indicate a gradual loss in the
after cycle 10 indicate minimal to no changes in the size or structural integrity of the MOF (Fig. S18†). Next, we assessed
quantity of the Pd nanoparticles (Fig. S14†), images taken after the rate of amine formation. We observe that the rate gradually
cycle 15 indicate that most of the Pd has leached out of the increases up to cycle 3, likely related to diffusion or catalyst
sample (Fig. S15, Table T2, see ESI†).
activation, and then appears to be steady up to cycle 10 within
In order to quantify Pd leaching, and hence better under- error of the experiment (Fig. 5b). The steady rate of amine pro-
stand catalyst failure, UiO-67/PpPDA/Pd and the three other duction up to this point indicates that there is preservation of
materials were soaked in ethanol at a temperature of 50 °C the Pd active sites, which is further supported by the afore-
under a 5-bar pressure of H2 for 2, 6, and 24 hours. Next, mentioned TEM images (Fig. S14†) and leaching tests carried
ICP-OES measurements were taken of the digested catalysts to out via ICP-OES (Table T3†). Unfortunately, over cycles 11 to
determine the amount of Pd loss (Table T3, see ESI†). For Fe- 15, there is an incomplete conversion to the desired amine,
BTC/Pd, there is a ∼77% Pd loss after 2 hours, and after which is accompanied by a nearly complete degradation of the
6 hours, the amount of Pd remaining in the sample is below MOF support (Fig. S18†). The latter lends to a noticeable leach-
the detectable limit. Both UiO-67/Pd and Pd/C have a minimal ing of the Pd active sites, as mentioned before, which is
Pd loss after 2 hours; however, after 6 hours, there is a ∼65% evident in TEM images taken of the composite (Fig. S15†).
and ∼76% loss, respectively. As expected, after 24 hours, the ICP-OES analysis reveals minimal Pd remaining in the sample
remaining Pd is again below the detectable limit. For UiO-67/ after cycle 15 (Table T2†) as determined by ICP-OES analysis. It
PpPDA/Pd, there is minimal Pd loss after 6 hours and only a should be considered that, despite the MOF degradation, we
∼15% loss after 24 hours. Further, MALDI-TOF-MS measure- see a nearly complete conversion of HMF at the end of the
ments taken of the ethanolic solutions indicate no apparent 2-hour reaction time up to cycle 10 (Fig. S19 and S20†). And,
leaching of polymer, even after 24 hours (Fig. S16†). When up to cycle 15, the HMF conversion decreases to only 90%.
considering that all MOF-based catalysts in this study have This data can be deceiving as our catalyst would look fully
similar Pd-loadings, the aforementioned results suggest that functional up to cycle 10, if only considering the HMF conver-
the N-containing polymer plays a positive role in stabilization sion and selectivity for the amine product over the reaction
of the Pd NPs.
time. However, when looking at the HMF consumption rate, it
This journal is © The Royal Society of Chemistry 2019
Green Chem.

View Article Online
Paper
Green Chemistry
Fig. 6 Comparison of the initial rate of HMF consumption as a function
of the para substitution Hammett co-efficient. Reaction conditions:
10 mg catalyst (0.4 wt% Pd loading), 5-hydroxymethylfurfural
(0.24 mmol), aniline (0.24 mmol), 5 mL ethanol, 5 bar H2, 50 °C,
10 minutes.
ate aggregation of the active Pd species. The latter demonstrates
the importance of the MOF in the composite.
Catalyst scope and substituent effect
After realizing the improved performance of UiO-67/PpPDA/Pd
compared to the other catalysts, it was necessary to test the
reductive amination of the HMF with a number of aromatic
and aliphatic amines. A few secondary amines and ammonia
were also tested to further expand the scope of the substrates.
The data (Table 2) clearly indicates that all examined substi-
tuted anilines with both electron withdrawing and electron
donating groups afford the target compounds. Secondary
amines (entries 12, 15 and 16) also afforded high conversion
Fig. 5 Kinetic study of the recyclability probed by plotting (a) the initial
rate of HMF consumption and (b) the rate of amine formation at each re-
cycling step. Each data-point has an error bar of 0.2 μM min−1 calculated
by triplicating the experiment; reaction conditions are HMF (0.24 mmol);
amine (0.24 mmol); UiO-67/PpPDA/Pd (10 mg, 0.4 wt% Pd loading); 5
bar H2; 50 °C, 5 mL ethanol, 10 minutes.
gives clearer signs of catalyst degradation (Fig. 6a) from the and selectivity towards the final amino substituted product.
start. As such, we believe that kinetic studies like the one pre- Moreover, even when glycine is used, which has two potential
sented in Fig. 5 can be a critical indicator of true long-term condensation sites, one on the amine and a second on the
catalyst cyclability.
hydroxyl group (entry 13, Table 2), condensation at the amino
While it is clear that UiO-67/PpPDA/Pd deactivates with center is favoured with >99% selectivity. Clearly the data shows
repeated cycling due to MOF degradation, the rate of de- that UiO-67/PpPDA/Pd is a versatile catalyst for the reductive
activation is significantly slowed when compared with other amination of HMF with a variety of substituted aromatic
catalysts tested in this work and several other heterogenous amines, as well as aliphatic amines, and secondary amines. To
catalysts reported by others (Fig. S21, Table T4, see ESI†). It is the best of our knowledge, the most efficient heterogeneous
thought that the enhanced cyclability is owed to the polymer, catalyst reported to date utilized 20 bar of CO at 60 °C over
which has high densities of Pd-complexing functionality. 2.5 h to catalyze the reductive amination using a titania sup-
Further, considering that the instability is directly related to the ported Au catalyst. On comparison, the catalyst developed in
MOF degradation, it implies that more stable supports, will this study has significant advantages, in terms of the turnover
lend to further extended catalyst lifetimes. It is likely that once number (604.8), turnover frequency (302.4 h−1) and cyclability
the MOF degrades, the polymers are no longer immobilized, as comparted to the Au catalyst and other catalysts reported
allowing aggregation and leaching of the Pd particles into solu- for reductive amination reactions. For comparison, a compre-
tion. It is also noted that for the bulk polymer containing Pd hensive list of previously reported catalysts is shown in
nanoparticles, referred to as PpPDA/Pd, the reaction appears to Table S20.† To better understand the substituent effects, the
be almost immediately homogeneous in nature (Table 1, entry reductive amination of HMF was also carried out with a series
7). The lack of rigidity of PpPDA polymer likely permits immedi- of substituted anilines having electron donating and electron
Green Chem.
This journal is © The Royal Society of Chemistry 2019

View Article Online
Green Chemistry
Paper
Table 2 Substrate scope using the UiO-67/PpPDA/Pd catalyst
No.
1
Substrate
Product
Time (h)
2
Yielda (%)
94.5
TON/TOFb (h−1
604.8/302.4
)
2
3
2
2
91
87
582.4/291.2
556.8/278.4
4
5
6
7
8
2
2
2
2
2
68
96
93
86
80
435.2/217.6
614.4/307.2
595.2/297.6
550.4/275.2
512/256
9
2
2
2
2
87
90
83
91
556.8/278.4
576/288
10
11
12
531.2/265.6
582.4/291.2
13
2
87
556.8/278.4
14
15
2
2
87
70
556.8/278.4
448/224
16
2
92
588.8/294.4
Unless otherwise mentioned, reaction conditions are HMF (0.24 mmol); amine (0.24 mmol); UiO-67/PpPDA/Pd (10 mg, 0.4 wt% Pd); 5 bar H2;
1
50 °C, 5 mL ethanol. a Determined by H NMR. b Turnover number (TON) and turnover frequency (TOF) calculated for the optimized conditions
for each substrate.
withdrawing substituents. For this study, each reaction was also probed using two substrates, including 4-toluidine and
carried out over 10 minutes, which corresponds to a conver- 2,4,6-trimethylaniline. The initial rate of HMF consumption
sion value of 40% in case of the model reaction. The Hammett was 5.1 0.2 μM min−1 and 4.6 0.2 μM min−1 for 4-toluidine
constants were used to empirically quantify and compare the and 2,4,6-methylaniline, respectively, indicating that the
effect of the substitution in the para position for aniline.54 As stearic effect of the methyl groups in the latter compound has
shown in Fig. 6, there is a linear decrease in the initial rate of a mild effect on the initial rate of HMF consumption. This is
HMF consumption with a decrease in electronic donating likely because of the stronger cumulative electron donating
character of the substituent. The effect of steric hindrance was effect from the larger number of methyl substituents in 2,4,6-
This journal is © The Royal Society of Chemistry 2019
Green Chem.

View Article Online
Paper
Green Chemistry
trimethylaniline; this can offset the stearic effect and hence the cavities of a host of different types of porous structures
help drive the reaction forward.
aimed at a wide array of potential hydrogenation processes.
Design template for making novel catalysts
While porous polymers and MOFs have been used as catalysts Experimental section
separately, to the best of our knowledge this is the first time a
MOF/polymer composite has been constructed for such an
Materials
5-Hydroxymethylfurfural (HMF, 99%) was supplied by
effort. The resulting composite catalyst, UiO-67/PpPDA/Pd,
demonstrates a significantly improved stability when com-
pared to the MOF (UiO-67/Pd) and polymer (PpPDA/Pd) tested
separately (Table 1, entries 4 and 7). This positive attribute
combined with the ease with which the catalyst building
blocks can be varied, strikes our interest to design such com-
posites for a wide array of interesting catalytic applications.
For instance, one can readily vary the MOF supports (or to
other porous frameworks), the stabilizing polymer additive, or
the catalytically active metal NPs for a number of targeted end
uses. Given this, we have also demonstrated that the active
metal NPs can indeed be varied. For this, several UiO-67/
PpPDA/M (M = Co, Ni and Cu) composite catalysts were syn-
thesized and characterized via powder XRD (Fig. S22†), TEM
(Fig. S23†) and N2 sorption measurements (Fig. S24†). The
testing of the utility of these catalysts for various hydrogen-
ation reactions is currently underway and will be the basis of
forthcoming reports.
Fluorochem. Zirconium chloride (ZrCl4, 99.5%) and 1,1′-biphe-
nyl-4,4′-dicarboxylic acid (BPDC, 98%) was obtained from
ABCR. Iron(III) chloride hexahydrate (97%) was bought from
Alfa Aesar and 1,3,5-benzenetricarboxylic acid (trimesic acid,
98%) was bought from ABCR GmbH. All other metal salts and
chemicals were bought from Sigma Aldrich and used without
further purification. Palladium nitrate (Pd(NO3)2, 99.8%
metals basis) was bought from Alfa Aesar. para-Phenylene-
diamine (pPDA, 99.5%) was obtained from Sigma-Aldrich.
Hydrochloric acid (HCl, 37%) was bought from Roth AG.
Commercial Pd/C (1 wt%) catalyst was acquired from ABCR
GmbH. The solvents, dimethylformamide (DMF, reagent
grade), ethanol (EtOH, reagent grade) and methanol (MeOH,
reagent grade) were bought from Roth AG and Reactolab SA
respectively. All chemicals were used without further
purification.
Characterization
Powder X-ray diffraction was performed on a Bruker D8
Discover system with a Cu Kα source (1.54056 Å) at 40 kV and
40 mA. The powder samples were loaded and flattened onto a
Conclusions
In conclusion, a highly stable MOF/polymer composite functio- zero-background silicon sample holder. The scanning range
nalized with Pd nanoparticles, denoted as UiO-67/PpPDA/Pd, was 1–60° over 2885 steps at a scanning rate of 2 steps per
has been developed and shows excellent catalytic activity for second. The simulated powder diffraction pattern of the MOFs
the reductive amination of HMF. In addition to the mild reac- was calculated using Mercury 3.6 crystallography software from
tion conditions utilized, the catalyst shows enhanced recycl- the .CIF files of the MOFs. Nitrogen adsorption measurements
ability, which is superior to others reported to date. The cata- were done on
a Micromeritics 3flex device. Samples
lyst is also highly active towards a wide variety of substrates (50–100 mg) were activated at 150 °C under vacuum for
(yields ranging from 80% to 96%) including aliphatic and aro- 12 hours. Nitrogen isotherms were obtained using incremental
matic amines with electron withdrawing or donating groups. exposure of ultra-high purity nitrogen in a liquid nitrogen
The polymer inclusion into the MOF promotes a relatively dewar at 77 K. The specific surface area was calculated by the
homogenous distribution of small Pd NPs and significantly BET method between a P/P0 range of 0.03–0.3. X-ray photo-
enhances the long-term stability of the catalyst, by a factor of electron spectroscopy measurements were carried out using a
5, when compared to other non-polymer containing MOF-NP Physical Instruments AG PHI VersaProbe II scanning XPS
composites and a commercially available Pd/C. We surmise microprobe. Experiments were performed with a beam size of
that this is due to the high densities of Lewis base functional- 100 μm using a monochromatic Al Kα X-ray source of 24.8 W
ity, which inhibit the aggregation of the Pd nanoparticles. Our power. The spherical capacitor analyzer was set at 45° take-off
data suggests that the eventual structural degradation of the angle with respect to the sample surface. The pass energy was
MOF with cycling coincides with an increase in the size distri- 46.95 eV yielding a full width at half maximum of 0.91 eV for
bution of the Pd NPs and also an eventual loss of this active the Ag 3d 5/2 peak. The data was analyzed using the CasaXPS
species from the composite; as such, further improving the software. Scanning Electron Microscopy (SEM) images were
stability of the MOF support, under the selected reaction con- taken using a FEI Teneo. Samples were sputter-coated with
ditions, could extend the catalyst lifetime significantly. Last, it iridium (∼7 nm thick) in a Quorum-Q150 to minimize electron
is noted that the oxidative polymerization of para-phenylene charging effects. SEM images were acquired using an in-
diamine (or other monomers) with metal nitrates is a general column (Trinity) detector at an accelerating voltage of 4.0 kV.
method, and hence can be used as a novel platform for the SEM-EDX data were acquired using an X-Flash silicon drift
design of many new heterogenous catalysts. This technique detector (Bruker) at an accelerating voltage of 10 kV.
could be used to introduce aggregation stable metal NPs into Transmission electron microscopy (TEM) images were taken
Green Chem.
This journal is © The Royal Society of Chemistry 2019

View Article Online
Green Chemistry
Paper
on a Tecno Spirit TEM device with an accelerating voltage of methanol repeatedly. Eventually, the UiO-67/Pd powder
200 kV. High-angle annular dark-field scanning transmission (0.4 wt% Pd by ICP-OES) was dried at room temperature in a
electron microscopy (HAADF-STEM) imaging and corres- vacuum oven at −30 inHg.
ponding STEM-EDX spectroscopy were performed on a FEI
Fe-BTC. 9.72 g of iron(III) chloride hexahydrate, 3.36 g of tri-
Titan Themis 60–300 operated in scanning mode at an acceler- mesic acid and 120 mL of distilled water were loaded in a
ating voltage of 200 kV. This microscope is equipped with a 180 mL Teflon autoclave. The reaction mixture was heated to
high brightness X-FEG gun and four silicon drift Super-X EDX 130 °C for 72 hours. After the reaction was cooled down to
detectors. Sample preparation for STEM measurements was room temperature the orange solid was filtered under vacuum
done using ultramicrotomy. A resin-embedded composite was and washed with copious amounts of water and methanol.
serially sectioned in ∼60 nm thick slices that were deposited The resulting powder was loaded into a double thickness
on a TEM grid with an ultrathin carbon support film. ATR-IR Whatman cellulose extraction thimble and underwent Soxhlet
spectra were collected using a PerkinElmer Frontier MIR/FIR purification with methanol for 24 hours. The sample was sub-
spectrometer equipped with a Quest ATR attachment. The sequently dried under vacuum overnight.
sample was pressed on a diamond window and the spectra
Fe-BTC/Pd. 100 mg of Fe-BTC were dispersed in 40 mL of
were recorded between 4000 and 400 cm−1 at a resolution of methanol by sonication for a period of 10 minutes. To this dis-
4 cm−1. The thermogravimetric analysis curve was obtained persion a solution of 1 mg of palladium nitrate in 0.68 mL of
using a TA Q-Series TGA Q500. Samples were loaded onto a MilliQ water was added in a dropwise fashion. The resulting
tared platinum pan. The balance flow rate was at 10 mL min−1 mixture was stirred at room temperature for 5.5 h.
with nitrogen and the sample flow rate was at 25 mL min−1 Subsequently, the powder was centrifuged and washed with
with air heating at a rate of 5 °C per minute. For polymer methanol repeatedly. Eventually, the Fe-BTC/Pd powder
characterization matrix-assisted laser desorption/ionization (0.47 wt% Pd by ICP-OES) was dried at room temperature in a
time of flight mass spectrometry (MALDI-TOF-MS) analysis vacuum oven at −30 inHg.
was performed using a Bruker Microflex. Composites were
UiO-67/PpPDA/Pd. 100 mg of activated UiO-67 were dis-
digested in a 5 mL aqua regia solution followed by treatment persed in 40 mL of methanol by sonication for a period of
in a microwave for 2 hours. A 10 mg ml−1 DHB (2,5-dihydroxy- 10 minutes. 10 mg of para-phenylenediamine were dissolved
benzoic acid) matrix consisting of H2O/acetonitrile/trifluoroa- in 1 mL of methanol and this solution was added to the MOF
cetic acid with a 49.9/50/0.1 ratio was mixed with the solution solution and stirred at room temperature for 1 hour. To this
of the digested composite, previously diluted 100 times, at a 1 solution, 10 mg of palladium nitrate in 2 mL of MilliQ water
to 1 ratio. 2 μL was drop casted onto a sample plate and was added and the solution was stirred for
1 day.
allowed to dry in air completely. The mode was set to linear Subsequently, the powder was filtered using a Buchner funnel
with a positive polarity and scanning range of 200–1500 m/z. and washed with ethanol (around 300 mL). After partial
Solution NMR spectra were recorded in deuterated solvents drying, the powder was further subjected to Soxhlet washings
(CDCl3) at room temperature on a BRUKER AVIII HD 400 with ethanol to remove any unreacted reagents or free poly-
spectrometer. The Bruker Topspin software package (version meric species. The powder was dried at room temperature in a
3.2) and Mestrenova NMR software (version 11.0.1) were used vacuum oven at −30 inHg to obtain the final composite
to measure and analyze the data respectively.
material (expected loading: 0.5 wt%; actual loading: 0.4 wt%
Pd by ICP-OES).
Catalyst preparation
UiO-67. UiO-67 was prepared according to the reported syn-
Catalytic performance
thetic procedure.36 ZrCl4 (2.41 g, 10.4 mmol) was sonicated in
a mixture of 180 mL of DMF and 18 mL of HCl (37 wt%) in a 1 The reductive amination of HMF was carried out in a stainless-
L cap-screw vessel for 15 min. BPDC (3.24 g, 13.4 mmol) and steel Parr autoclave (0.45 L), equipped with a stainless-steel
360 mL of DMF were added to the mixture and the reaction test-tube rack. HMF, the corresponding amine and the catalyst
mass was additionally sonicated for 15 minutes (BPDC does were transferred into Parr autoclave tubes in anhydrous
not dissolve completely) following by heating at 80 °C for ethanol. The tubes were purged thrice with gaseous hydrogen
24 hours. After the reaction was finished, the crude was under stirring (500 rpm). After purging, the autoclave was
washed with DMF (4 × 40 mL) and solvent exchanged with pressurized with 5 bars of hydrogen and stirred at 50 °C for
EtOH (3 × 40 mL, overnight). The sample was dried at RT the 2 hours to form the products. After the reaction, the
under vacuum for 6 hours, yielding 3.7 g of UiO-67 as white mother liquor was centrifuged, to remove the catalyst, and the
crystalline solid.
ethanolic supernatant was rid of ethanol under vacuum using
UiO-67/Pd. 100 mg of UiO-67 were dispersed in 40 mL of a rotary evaporator. The residue was re-dissolved in diethyl
methanol by sonication for a period of 10 minutes. To this dis- ether and dried using anhydrous magnesium sulphate. The
persion a solution of 1 mg of palladium nitrate in 0.68 mL of ether rich organic layer was evacuated under vacuum to yield
1
MilliQ water was added in a dropwise fashion. The resulting the final products. The product purity was analysed using H
mixture was stirred at room temperature for 5.5 h. NMR with deuterated chloroform as the solvent and TMS as
Subsequently, the powder was centrifuged and washed with the internal standard.
This journal is © The Royal Society of Chemistry 2019
Green Chem.

View Article Online
Paper
Green Chemistry
18 S. Nishimura, K. Mizhuhori and K. Ebitani, Res. Chem.
Intermed., 2016, 42, 19–30.
Conflicts of interest
There are no conflicts to declare.
19 J. J. Martinez, V. Nope, R. Eliana, H. Rojas, M. H. Brijaldo,
F. Passos and G. P. Romanelli, J. Mol. Catal. A: Chem., 2014,
392, 235–240.
20 M. Chatterjee, T. Ishikaza and H. Kawanami, Green Chem.,
2016, 18, 487–496.
Acknowledgements
This work was supported by the Swiss National Science 21 G. Hahn, P. Kunnas, N. de Jonge and R. Kempe, Nat.
Foundation (SNSF) under grant number Catal., 2019, 2, 71–77.
PYAPP2_160581. V. V. K. thanks Canton of Valais for financial 22 G. Liang, A. Wang, L. Li, G. Xu, N. Yan and T. Zhang,
support of his PhD Project titled “Biomass Conversion in Angew. Chem., Int. Ed., 2017, 56, 3050–3054.
Metal-Organic Frameworks”. O. T. acknowledges funding from 23 H. Furukawa, K. E. Cordova, M. O’Keeffe and O. M. Yaghi,
the European Union’s Horizon 2020 research and innovation Science, 2014, 341, 123044.
program under the Marie Skłodowska-Curie Grant No. 665667. 24 D. J. Tranchemontagne, J. L. Mendoza-Cortes, M. O’Keeffe
We acknowledge the Swiss-Norwegian Beam Line BM01 at
European Synchrotron Radiation Facility (ESRF) for the pro-
and O. M. Yaghi, Chem. Soc. Rev., 2009, 38(5),
1257–1283.
vided beam time. The authors humbly thank Mounir Mensi 25 I. M. Hönicke, I. Senkovska, V. Bon, I. A. Baburin,
and Mehrdad Asgari for their help with XPS and synchrotron
powder X-ray diffraction, respectively.
N. Bönisch, S. Raschke, J. D. Evans and S. Kaskel, Angew.
Chem., Int. Ed., 2018, 57(42), 13780–13783.
26 K. Koh, A. G. Wong-Foy and J. A. Matzger, J. Am. Chem.
Soc., 2009, 131(12), 4184–4185.
27 Z. Wang and S. M. Cohen, Chem. Soc. Rev., 2009, 38, 1315–
1329.
Notes and references
1 J. W. Tester, E. M. Drake, M. W. Golay, M. J. Driscoll and 28 A. Dhakshinamoorthy and H. Garcia, Chem. Soc. Rev., 2012,
W. A. Peters, Sustainable Energy: Choosing Among Options,
MIT Press, Cambridge, MA, 2005.
2 L. Shuai, M. T. Amiri and J. S. Luterbacher, Curr. Opin.
Green Sustainable Chem., 2016, 2, 59–63.
41, 5262–5284.
29 M. Meilikhov, K. Yusenko, D. Esken, S. Turner, G. van
Tendeloo and R. A. Fischer, Eur. J. Inorg. Chem., 2010,
3701–3714.
3 C. Li, X. Zhao, A. Wang, G. W. Huber and T. Zhang, Chem. 30 S. Yang, L. Peng, D. T. Sun, M. Asgari, E. Oveisi,
Rev., 2015, 115, 11559–11624.
4 R. J. van Putten, J. C. van der Waal, E. de Jong,
O. Trukhina, S. Bulut, A. Jamali and W. L. Queen, Chem.
Sci., 2019, 10, 4542–4549.
C. Rasrendra, H. Heeres and J. G. de Vries, Chem. Rev., 31 J. G. Nguyen and S. M. Cohen, J. Am. Chem. Soc., 2010,
2013, 113(3), 1499–1597. 132(13), 4560–4561.
5 B. J. Price, J. W. Clitherow and J. Bradshaw, U.S. Patent 32 L. Peng, S. Yang, S. Jawahery, S. M. Moosavi, A. J. Huckaba,
5750714A, 1995.
6 L. Hu, L. Lin, Z. Wu, S. Zhou and S. Liu, Renewable
Sustainable Energy Rev., 2017, 74, 230–257.
M. Asgari, E. Oveisi, M. K. Nazeeruddin, B. Smit and
W. L. Queen, J. Am. Chem. Soc., 2019, 141(31), 12397–
12405.
7 I. Yu and D. Tsang, Bioresour. Technol., 2017, 238, 716–732.
8 J. Chheda, Y. Roman-Leshkov and J. A. Dumesic, Green
Chem., 2007, 9, 342–350.
9 Y. Pagan-Torres, T. Wang, J. Gallo, B. Shanks and
J. A. Dumesic, ACS Catal., 2012, 2, 930–934.
10 D. L. Crookes, U.S. Patent 4521431A, 1980.
33 Y. Li, X. M. Hong, D. M. Collard and M. A. El-Sayed, Org.
Lett., 2000, 2, 2385–2387.
34 J. Huang, T. Jiang, B. Han, H. Gao, Y. Chang, G. Zhao and
W. Wu, Chem. Commun., 2003, 14, 1654–1655.
35 H. R. Moon, D.-W. Lim and M. P. Suh, Chem. Soc. Rev.,
2013, 42, 1807–1824.
11 B. J. Price, J. W. Clitherow and J. Bradshaw, U.S. Patent 36 M. J. Katz, Z. J. Brown, Y. J. Colon, P. W. Siu, K. A. Scheidt,
4279819A, 1976.
12 J. Cooper and D. V. Lee, U.S. Patent 4347191A, 1980.
R. Q. Snurr, J. T. Hupp and O. K. Farha, Chem. Commun.,
2013, 49, 9449–9451.
13 K. S. K. Murthy, G. Weeratunga, B. K. Radatus and 37 P. W. Siu, Z. J. Brown, O. K. Farha, J. T. Hupp and
K. P. S. Sidhu, U.S. Patent 5750714A, 1995. K. A. Scheidt, Chem. Commun., 2013, 49, 10920–10922.
14 A. Cukalovic and C. V. Stevens, Green Chem., 2010, 12, 38 G. L. Zhuang, J. Q. Bai, X. Zhou, Y. F. Gao, H. L. Huang,
1201–1206.
H. Q. Cui, X. Zhong, C. L. Zhong and J. G. Wang,
Eur. J. Chem., 2017, 1, 172–178.
39 S. Lin, A. K. Ravari, J. Zhu, P. M. Usov, M. Cai,
S. R. Ahrenholtz, Y. Pushkar and A. J. Morris,
ChemSusChem, 2018, 11, 464–471.
15 Z. Xu, P. Yan, W. Xu, S. Jia, Z. Xia, B. Chung and
Z. C. Zhang, RSC Adv., 2014, 4, 59083–59087.
16 M. Zhu, L. Tao, Q. Zhang, J. Dong, Y.-M. Liu, H.-Y. He and
Y. Cao, Green Chem., 2017, 19, 3880–3887.
17 T. Komanoya, T. Kinemura, Y. Kita, K. Kamta and M. Hara, 40 Y. L. Wei, Y. Li, Y. Q. Chen, Y. Dong, J. J. Dao, X. H. Yan
J. Am. Chem. Soc., 2017, 139, 11493–11499.
and Y. B. Dong, Inorg. Chem., 2018, 57(8), 4379–4386.
Green Chem.
This journal is © The Royal Society of Chemistry 2019

View Article Online
Green Chemistry
Paper
41 L. Chen, H. Chen, R. Luque and Y. Li, Chem. Sci., 2014, 5, 49 D. T. Sun, N. Gasilova, S. Yang, E. Oveisi and
3708–3714.
W. L. Queen, J. Am. Chem. Soc., 2018, 140, 16697–
16703.
50 R. Arrigo, M. E. Schuster, Z. Xie, Y. Yi, G. Wowsnick,
L. L. Sun, K. E. Hermann, M. Friedrich, P. Kast,
M. Havecker, A. Knop-Gericke and R. Schlögl, ACS Catal.,
2015, 5, 2740–2753.
51 S. Archana and R. J. Shanthi, Res. J. Chem. Sci., 2014, 4(2),
60–67.
52 S. Chavan, J. G. Vitillo, D. Gianolio, O. Zavorotynska,
B. Civalleri, S. Jakobsen, M. H. Nilsen, L. Valenzano,
C. Lamberti, K. P. Lillerud and S. Bordiga, Phys. Chem.
Chem. Phys., 2014, 14, 1614–1626.
42 L. Chen, B. Huang, X. Qiu, X. Wang, R. Luque and Y. Li,
Chem. Sci., 2016, 7, 228–233.
43 Y. Bu, F. Li, Y. Zhang, R. Liu, X. Luo and L. Xu, RSC Adv.,
2016, 6, 40560–40566.
44 H. Xu, Y. Li, X. Luo, Z. Xu and J. Ge, Chem. Commun., 2017,
53, 7953–7956.
45 Y. H. Hu, J. C. Wang, S. Yang, Y. A. Li and Y. B. Dong, Inorg.
Chem., 2017, 56(4), 8341–8347.
46 Z. Jin, D. Nackashi, W. Lu, C. Kittrell and J. M. Tour, Chem.
Mater., 2010, 22(20), 5695–5699.
47 D. T. Sun, L. Peng, W. Reeder, S. M. Moosavi, D. Tiana,
D. K. Britt, E. Oveisi and W. L. Queen, ACS Cent. Sci., 2018, 53 R. C. Klet, Y. Liu, T. C. Wang, J. T. Hupp and O. K. Farha,
4(3), 349–356. J. Mater. Chem. A, 2016, 4, 1479–1485.
48 L. Peng, S. Yang, D. T. Sun, M. Asgari and W. L. Queen, 54 K. C. Gross and P. G. Seybold, Int. J. Quantum Chem., 2000,
Chem. Commun., 2018, 54, 10602–10605.
80, 1107–1115.
This journal is © The Royal Society of Chemistry 2019
Green Chem.