Photolabile protecting groups in metal-organic frameworks: Preventing interpenetration and masking functional groups

Article abstract of DOI:10.1039/c1cc12884a

Photolabile groups can be incorporated into metal-organic frameworks (MOFs) and then quantitatively cleaved following MOF formation. Here, a 2-nitrobenzyl ether prevents lattice interpenetration (catenation) in a cubic MOF derived from zinc(ii). Subsequen

Full text of DOI:10.1039/c1cc12884a

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
ChemComm
Cite this: Chem. Commun., 2012, 48, 1574–1576
www.rsc.org/chemcomm
COMMUNICATION
Photolabile protecting groups in metal–organic frameworks: preventing
interpenetration and masking functional groupswz
Rajesh K. Deshpande,a Geoffrey I. N. Waterhouse,b Geoffrey B. Jamesona and
Shane G. Telfer*a
Received 17th May 2011, Accepted 7th July 2011
DOI: 10.1039/c1cc12884a
Photolabile groups can be incorporated into metal–organic
frameworks (MOFs) and then quantitatively cleaved following
MOF formation. Here, a 2-nitrobenzyl ether prevents lattice
interpenetration (catenation) in a cubic MOF derived from
zinc(II). Subsequent photolysis unmasks a hydroxyl group, and
produces an open MOF that cannot be synthesized directly.
(UMCM = University of Michigan Crystalline Material) they
were able to mask hydroxyl groups during framework synthesis
then reveal them in a post synthetic photolysis step.6
Several other pioneering studies have established that
MOFs can be addressed and transformed by light. For
example, carbonyl ligands of Cr(CO)3 units appended to the
phenyl struts of MOF-5 can be labilized by visible light,7 and
cycloaddition reactions of unsaturated ligands have been
explored.8 In very recent work, an azide group was converted
into a nitrene within a coordination polymer using UV light9
and long-lived radicals have been generated within a series of
cadmium(II) frameworks upon irradiation.10
The post-synthetic modification of crystalline metal–organic
frameworks (MOFs) has emerged as a versatile and powerful
method for tuning the functional attributes of these materials.1 We
are interested in the reverse of this process, i.e., the controlled post-
synthetic cleavage of specific framework components. In the first
example of this concept, we recently demonstrated how a bulky
tert-butoxycarbonyl (Boc) group appended to functionalized
ligands could be decomposed into volatile fragments by simple
heating.2,3 The thermolysis reaction served to deprotect amino
functional groups, which had hindered MOF formation when left
exposed during the crystal growth step. Further, the steric bulk of
the Boc group suppresses interpenetration, and its cleavage
produces an open framework with significant void volumes.
This methodology builds on a significant body of work on
thermolabile protecting groups.4 We reasoned that the scope of
our approach may be expanded by introducing photolabile
protecting groups5 into MOFs. This would serve to diversify
the functional groups that can be surreptitiously incorporated
into MOFs, allowing for a wider range of reaction conditions to
be used for the synthesis of protected MOFs (specifically higher
temperatures and low pH), and provide an alternative physical
means by which deprotection of the frameworks can be induced.
While this manuscript was in preparation, Cohen et al. published
a superb demonstration of the viability of this concept. By
incorporating 2-nitrobenzylether groups in UMCM-type MOFs
The design of a framework material for trialing our methodo-
logy was influenced by several considerations. First, it was
recognized that some metal/ligand combinations may lead to
undesirable photochemical reactions that may degrade the
integrity of the framework. For example, carboxylate complexes
of metals such as copper(II),11 iron(III)12 and cobalt(III)13 are
known to decarboxylate upon irradiation into their LMCT bands.
Second, a relatively large protecting group is expected to prevent
interpenetration and thus encourage the formation of a frame-
work with a network of wide pores (after photolysis) suitable for
applications in storage and catalysis. Third, the ideal protective
group is one that can be released with a high quantum yield at a
wavelength where the remainder of the structure is transparent.
With these considerations in mind, we designed ligand H21
and envisaged that the synthesis and photolytic transformation
of the MOF [Zn4O(1)3] could be achieved as outlined in
Scheme 1. The side-arm 1 features an hydroxyl group that is
protected by a large 2-nitrobenzyl moiety. 2-Nitrobenzyl ethers
are well established photolabile protecting groups that are
cleaved with high efficiency upon irradiation around 350 nm to
liberate the parent alcohol.5,14 Since MOFs comprising zinc(II)
and arylcarboxylate ligands transmit a large fraction of light of
this wavelength, the light will be predominantly absorbed by the
nitrobenzyl chromophore. 2-Nitrosobenzaldehyde is generated
as a side product in this deprotection step.
a MacDiarmid Institute for Advanced Materials and Nanotechnology,
Institute of Fundamental Sciences, Palmerston North, New Zealand.
E-mail: s.telfer@massey.ac.nz; Tel: +64 27 3566655
b School of Chemical Sciences, University of Auckland, Auckland,
New Zealand
The synthesis of H21 was achieved in three steps starting from
dimethyl 2-aminobiphenyl-4,40-dicarboxylate. As expected on
the basis of the pioneering work of Yaghi’s group15 and more
recent reports from Burrows et al.,16 and our own group,2,3 the
solvothermal reaction of H21 with Zn(NO3)2 in N,N-diethyl-
formamide (DEF) produced well-faceted colorless cubic crystals.
w This article is part of the ChemComm ‘Emerging Investigators 2012’
themed issue.
z Electronic supplementary information (ESI) available: Full details of
ligand and MOF synthesis and characterization, photolysis reactions,
TGA, crystallography, and gas sorption experiments. CCDC 826054
and 826055. For ESI and crystallographic data in CIF or other
electronic format see DOI: 10.1039/c1cc12884a
c
1574 Chem. Commun., 2012, 48, 1574–1576
This journal is The Royal Society of Chemistry 2012

View Article Online
Scheme 1 Conversion of ligand H21 to a noninterpenetrated cubic
metal–organic framework, [Zn4O(1)3], followed by the photolytic
deprotection of the hydroxyl group to generate noninterpenetrated
[Zn4O(2)3]. MOF synthesis using ligand H22 or H23 generates the
interpenetrated frameworks b-[Zn4O(2)3] and b-[Zn4O(3)3].
Fig. 1 (a and b) The structure of [Zn4O(1)3] as determined by single
crystal X-ray crystallography. For clarity only three nitrobenzyl
groups are shown at representative positions in CPK format in (b);
(c) the proposed structure of [Zn4O(2)3], produced by the photolysis of
X-Ray crystallographyz revealed that these crystals comprised
a cubic framework of Zn4O nodes linked by bridging ligands
of 1 (Fig. 1a and b). The overall stoichiometry of the frame-
1
[Zn4O(1)3], as deduced from X-ray diffraction and H NMR spectro-
scopy; (d) the structure of interpenetrated b-[Zn4O(3)3], which serves
as a model for b-[Zn4O(2)3], as determined by X-ray crystallography.
The methoxy groups are shown in CPK format and the carbon and
zinc atoms of the two lattices are coloured differently for clarity.
%
work is [Zn4O(1)3], and it belongs to the space group P43m
with a unit cell length of 17.211(2) A.
The cubic topology of [Zn4O(1)3] is consistent with other
MOFs derived from biphenyl-4,40-dicarboxylic acid ligands.
The framework exhibits an open, non-interpenetrated structure,
as observed for a related ligand appended with a bulky tert-
butylcarbamate group,2 but in contrast to similar ligands that
feature small functional groups, which typically yield inter-
penetrated structures.15,16 It appears that the nitrobenzyl
group of ligand 1 presents sufficient steric bulk to discourage
the formation of interpenetrated frameworks.
The nitrobenzyl group could not be located in the Fourier
difference map, due to the statistical disorder over four sites
coupled with likely conformational disorder at each of these sites.
1
However, H NMR spectroscopy on digested crystals indicated
that this group remained intact in the MOF and analysis of the
residual electron density within the framework voids was con-
sistent with its presence. The atoms of this unit were placed in
calculated positions to complete the refinement. The phase purity
of [Zn4O(1)3] was established by powder XRD; the observed
diffraction pattern (Fig. 2b) closely matched that predicted from
the single-crystal structure (Fig. 2a).
Fig. 2 Powder X-ray diffractograms (Cu Ka radiation) of (a) the
pattern predicted from the single-crystal structure of [Zn4O(1)3],
(b) [Zn4O(1)3], (c) [Zn4O(2)3] produced by photolysis of [Zn4O(1)3],
(d) b-[Zn4O(2)3], and (e) b-[Zn4O(3)3].
To effect the single-crystal-to-single-crystal photolytic
conversion of [Zn4O(1)3] to [Zn4O(2)3], crystals of [Zn4O(1)3]
suspended in DMF or THF were irradiated with 355 nm laser
light with stirring. Reaction progress was monitored by digest-
directly observed due to decomposition via secondary photolysis
reactions.17 The powder XRD diffractogram of the photolyzed
material (Fig. 2c) is very similar to that of [Zn4O(1)3], indicating
that the skeletal structure of the framework is unchanged by
photolysis and that long-range order is maintained. By combining
the NMR and PXRD results, a structural model of [Zn4O(2)3] can
be developed, as depicted in Fig. 1c. In these experiments, the
mechanical agitation of a suspension of crystals of [Zn4O(1)3]
promotes the photolysis reaction by grinding the material into
1
ing the MOF in DMSO-d6/DCl and recording the H NMR
spectrum (Fig. S1z). We were delighted to observe that, after
around two hours, a remarkably clean conversion to [Zn4O(2)3]
had taken place in a single-crystal-to-single-crystal fashion. The
expected photolysis by-product, 2-nitrosobenzaldehyde, is not
c
This journal is The Royal Society of Chemistry 2012
Chem. Commun., 2012, 48, 1574–1576 1575

View Article Online
microcrystallites. When similar photochemical reactions were
performed on large single crystals of [Zn4O(1)3], it was impossible
to drive the photochemical reaction to beyond 50% completion
(as indicated by 1H NMR spectroscopy on digested samples).
Although the external form of the crystals was maintained, they
became noticeably opaque upon irradiation. Retardation of the
photolysis reaction in this case may be attributed to the difficulty
of photons penetrating to the core of the crystals as a consequence
of (i) scattering from defects and (ii) absorption from competing
chromophores (given the density of chromophores in the crystal,
even weakly absorbing transitions may prevent the transmission
of a significant fraction of the photons). This is a clear drawback
of photochemical deprotection as compared to thermolytic
deprotection. Single-crystal XRD experiments on the partially
photolyzed crystals produced sharp diffraction spots to a
resolution beyond 0.90 A, however a full structural determination
was not pursued (see ESIz for details).
porous, are produced. This methodology for the surreptitious
incorporation of functional groups that line the pores of open
MOFs is potentially quite general in terms of the framework
topology, the ligand backbone, and the functional group itself.
It may find applications in the synthesis of MOFs that cannot
be prepared directly, for example because the ligands bear
functional groups that cannot withstand the solvothermal
reaction conditions or that hinder MOF growth by coordinating
to the metal ion. This opens up new perspectives on tailoring
the chemical space within MOFs to optimize interactions with
incoming guest molecules, particularly for applications in gas
sorption and catalysis. We are actively exploring these
research avenues employing both photolabile and thermo-
labile protecting groups.
We are very appreciative of the funding received from
MacDiarmid Institute that supported this work. Dr Mark
Waterland is gratefully acknowledged for generously providing
access to a laser.
Following activation by supercritical CO2,18 N2 gas
sorption experiments gave an average BET surface area of
131 m2 gꢀ1 (Fig. S4z), which is lower than predicted on the
basis of X-ray crystallography,19 indicating a degree of pore
collapse upon desolvation, but on par with previous studies.3
It is noteworthy that the direct solvothermal reaction of H22
with Zn(NO3)2 in DEF produces a crystalline material,
b-[Zn4O(2)3]. Single-crystal XRD of b-[Zn4O(2)3] showed a
clear set of diffraction spots out to a resolution of around
1.30 A (Fig. S3z). Although satisfactorily indexing to a reason-
able unit cell (see ESIz for a full discussion) was impossible,
the data are consistent with the presence of a pair of
non-commensurate interpenetrating lattices i.e., b-[Zn4O(2)3]
is an interpenetrated analogue of the phase produced by the
photolysis of [Zn4O(1)3].
Notes and references
1 (a) S. M. Cohen, Chem. Sci., 2010, 1, 32–36; (b) Z. Wang and
S. M. Cohen, Chem. Soc. Rev., 2009, 38, 1315; (c) Z. Wang and
S. M. Cohen, J. Am. Chem. Soc., 2007, 129, 12368–12369.
2 R. K. Deshpande, J. L. Minnaar and S. G. Telfer, Angew. Chem.,
Int. Ed., 2010, 47, 4598–4602.
3 D. J. Lun, G. I. N. Waterhouse and S. G. Telfer, J. Am. Chem.
Soc., 2011, 133, 5806–5809.
4 P. G. M. Wuts and T. W. Greene, Greene’s Protective Groups in
Organic Synthesis, Wiley, Hoboken, 4th edn, 2007.
5 C. G. Bochet, J. Chem. Soc., Perkin Trans., 2002, 1, 125–142.
6 K. K. Tanabe, C. A. Allen and S. M. Cohen, Angew. Chem., Int.
Ed., 2010, 49, 9730–9733.
7 S. S. Kaye and J. R. Long, J. Am. Chem. Soc., 2008, 130, 806–807.
8 (a) M. H. Mir, L. L. Koh, G. K. Tan and J. J. Vittal, Angew.
Chem., Int. Ed., 2010, 49, 390–393; (b) D. Liu, Z.-G. Ren, H.-X. Li,
J.-P. Lang, N.-Y. Li and B. F. Abrahams, Angew. Chem., Int. Ed.,
2010, 49, 4767–4770; (c) J. F. Eubank, V. C. Kravtsov and
M. Eddaoudi, J. Am. Chem. Soc., 2007, 129, 5820–5821.
9 H. Sato, R. Matsuda, K. Sugimoto, M. Takata and S. Kitagawa,
Nat. Mater., 2010, 9, 661–666.
To gather complementary evidence that b-[Zn4O(2)3] is an
interpenetrated phase, we determined the structure of
b-[Zn4O(3)3], which is derived from 2-methoxybiphenyl-4,40-
dicarboxylic acid, H23 (Scheme 1).20 This structure, which
%
belongs to the tetragonal space group P421m, comprises a
10 Q.-K. Liu, J.-P. Ma and Y.-B. Dong, J. Am. Chem. Soc., 2010, 132,
7005–7017.
commensurate doubly-interpenetrated pair of cubic lattices
built up from struts of 3 and Zn4O nodes (Fig. 1d). It is
noteworthy that, to date, all known non-interpenetrated
zinc(II) MOFs based on biphenyl-4,40-dicarboxylic acid
ligands belong to cubic space groups,2,15,21 while inter-
penetrated variants belong to non-cubic space groups.15,16,21
Thermal gravimetric analyses (TGA) were conducted on the
MOFs (Fig. S2z). For [Zn4O(1)3], fragmentation of the nitro-
benzyl groups in the region 300–350 1C6 is followed by
framework decomposition beyond B400 1C. The onset of
decomposition occurs at a similar temperature for [Zn4O(2)3],
while b-[Zn4O(2)3] and [Zn4O(3)3] exhibit slightly greater thermal
stability. This is probably a consequence of the mutual bracing of
one lattice by the other in these interpenetrated materials.
In summary, these results demonstrate that photolabile
groups can be introduced to MOF ligands to (i) prevent
framework interpenetration, and (ii) mask hydroxy functional
groups during MOF synthesis. Point (ii) complements the
results obtained by Cohen et al.,6 while point (i) adds a new
dimension to the photolabile protecting group strategy: in the
absence of the photolabile group, it is shown that inter-
penetrated frameworks, which are inherently less open and
11 P. Natarajan and G. Ferraudi, Inorg. Chem., 1981, 20, 3708–3712.
12 H. B. Abrahamson, A. B. Rezvani and J. G. Brushmiller, Inorg.
Chim. Acta, 1994, 226, 117–127.
13 (a) R. M. Hartshorn and S. G. Telfer, J. Chem. Soc., Dalton Trans.,
1999, 3217–3224; (b) R. M. Hartshorn and S. G. Telfer, J. Chem.
Soc., Dalton Trans., 1999, 3565–3571.
14 (a) S. Muralidharan and J. M. Nerbonne, J. Photochem.
Photobiol., B, 1995, 27, 123–137; (b) Y. Zhao, Q. Zheng,
K. Dakin, K. Xu, M. L. Martinez and Wen- Hong, J. Am. Chem.
Soc., 2004, 126, 4653–4663.
15 M. Eddaoudi, J. Kim, N. Rosi, D. Vodak, J. Wachter, M. O0Keeffe
and O. M. Yaghi, Science, 2002, 295, 469–472.
16 (a) A. D. Burrows, C. G. Frost, M. F. Mahon and C. Richardson,
Chem. Commun., 2009, 4218–4220; (b) A. D. Burrows, C. Frost,
M. F. Mahon and C. Richardson, Angew. Chem., Int. Ed., 2008,
47, 8482–8486.
17 (a) H. Tomioka, N. Ichikawa and K. Komatsu, J. Am. Chem. Soc.,
1992, 114, 8045–8053; (b) Y. V. Il’ichev, M. A. Schworer and
¨
J. Wirz, J. Am. Chem. Soc., 2004, 126, 4581–4595.
18 A. P. Nelson, O. K. Farha, K. L. Mulfort and J. T. Hupp, J. Am.
Chem. Soc., 2009, 131, 458–460.
19 K. S. Walton and R. Q. Snurr, J. Am. Chem. Soc., 2007, 129,
8552–8556.
20 This compound has been reported (ref. 16b) but the crystal
structure could not be fully determined.
21 (a) U.S. Pat., 6,930,193 B2, 2005; (b) T.-H. Park, K. Koh,
A. G. Wong-Foy and A. J. Matzger, Cryst. Growth Des., 2011, ASAP.
c
1576 Chem. Commun., 2012, 48, 1574–1576
This journal is The Royal Society of Chemistry 2012