Seth M.
Cohen
*
Department of Chemistry and Biochemistry, University of California, San Diego, 9500 Gilman Drive, La Jolla, California, USA. E-mail: scohen@ucsd.edu; Fax: +1-858-822-5598; Tel: +1-858-822-5596
First published on 12th May 2010
The postsynthetic modification (PSM) of metal–organic frameworks (MOFs) has grown substantially in the past few years. In this minireview, recent progress in the area of PSM is highlighted, with an emphasis on several recent advancements. The scope and limitations of PSM are described and future prospects are discussed.
Seth M. Cohen | Seth M. Cohen obtained a Bachelor of Science degree in Chemistry and a Bachelor of Arts degree in Political Science at Stanford University. He obtained his PhD at the University of California, Berkeley in inorganic chemistry with Prof. Kenneth N. Raymond. He performed postdoctoral research in the laboratory of Prof. Stephen J. Lippard at the Massachusetts Institute of Technology. In 2001, he joined the University of California, San Diego in the Department of Chemistry and Biochemistry, where he is currently an Associate Professor. His research interests are primarily in the areas of metal–organic frameworks and medicinal bioinorganic chemistry. |
Fig. 1 Top: illustration comparing PSM modification of MOFs with conventional incorporation of functional groups embedded in the ligand precursor. The MOF depicted is IRMOF-3 (isoreticular metal–organic framework), which is a cubic lattice comprised of NH2-BDC ligands (represented in stick form) and Zn4O clusters (zinc ions represented by blue tetrahedra). Bottom: scheme of a simple PSM reaction on IRMOF-3. The MOF is represented as a line drawing with only one modification site shown for clarity. |
1. The MOF must be sufficiently porous to allow access of all required reagents to the interior of the lattice (unless only surface modification is desired, vide infra).
2. The MOF must possess an available functional group that can undergo a chemical transformation.
3. The MOF must be stable to the reaction conditions (e.g. solvent, caustic reagents, temperature).
4. The MOF must be stable to any by-products produced by the reaction conditions (e.g. acids, radicals).
To date, PSM is limited by these parameters and so both the choice of MOF (e.g. suitable porosity, chemical stability) and choice of reaction (e.g. suitable solvents, chemical by-products) will govern the scope of transformations that can be realized by PSM.
Fig. 2 List of organic PSM reactions that have been reported on MOFs. |
To date, the most clear example showing the difference between a solution and a MOF-based chemical reaction has been reported by Jones and Bauer.9 In their report, a cubic MOF constructed of trans-4,4′-stilbene dicarboxylate (SDC) having the formula Zn4O(SDC)3 (Fig. 3) was prepared and treated with Br2 at room temperature for 48 h. Under these conditions, the SDC ligands within the framework were converted to the dibromide adduct (Fig. 2, entry 6). Most importantly, the sole product of the MOF reaction with Br2 was the meso-stereoisomer. This is in striking contrast to the homogenous reaction of the ligand with Br2 in CH2Cl2, which resulted in a 4:1 ratio of meso- to rac-isomers. The stereoselectivity of the MOF reaction is rationalized as follows: the bromonium intermediate is accessible for back-side attack by bromide, but carbon–carbon bond rotation is inhibited because the ligand is immobilized as part of the MOF framework (Fig. 3).9 The restricted bond rotation allows for only the formation of the meso-product. This study definitively shows that PSM reactions within the MOF lattice can influence the stereochemical outcome of a reaction. These findings suggest that MOFs may become interesting platforms for generating new reaction outcomes and obtaining a better understanding of reaction mechanisms. Indeed, another recent report has shown that an unstable organic intermediate could be isolated and structurally characterized within the matrix of a MOF;10 a commentary on this fascinating piece of work can be found elsewhere.11
Fig. 3 The stereochemical outcome of a bromination reaction is different with PSM on a MOF (top) than with the same ligand in solution (bottom). Restricted bond rotation within the MOF leads to a single stereoisomer. |
Hupp and Nguyen et al. have reported a clever solution to this problem that allows for PSM to be applied in such a way that the surface and interior of the MOF structure can be distinguished.12 In their study, a MOF was constructed from two different ligands (Fig. 4); one of these ligands contained a trimethylsilyl-protected (TMS) acetylene group suitable for modification. The acetylene group can undergo ‘click’ chemistry with an azide upon removal of the TMS group (Fig. 2, entry 7) with a fluoride source. In order to achieve selective functionalization, the MOF crystals were first soaked in chloroform and then exposed to an aqueous solution of KF. The aqueous solution will not permeate the interior of the chloroform-saturated MOF due to the low partition coefficient of KF in chloroform, resulting in deprotection of only the surface TMS-acetylene groups. The MOF is then subjected to ‘click’ conditions and only the surface-exposed, deprotected acetylene groups are available to undergo the reaction. Subsequent deprotection of the interior TMS-acetylene groups with TEAF in THF, allowed for transformation of the interior acetylene groups with a different azide, giving a MOF with a ‘core–shell’ modification pattern.12 This study clearly showed, for the first time, control over the spatial distribution of PSM. The availability of such core–shell MOFs will likely have a substantial impact on the application of MOFs in many areas, including their use in biotechnology and biomedicine.
Fig. 4 Selective functionalization of the exterior (red) and interior (black) of a MOF crystal. The ligands used to prepare this MOF are shown in the upper left. |
MIL-101(Fe) (MIL = Material Institut Lavoisier),18,19 comprised of Fe(III) clusters and a mixture of 1,4-benzenedicarboxylate and NH2-BDC ligands, was formulated into NMOFs for use as biocompatible particles. Specifically, the MIL-101(Fe) material was selected for its very high porosity and low toxicity. The NH2-BDC ligand of MIL-101(Fe) was used so that dye (imaging) or drug (therapeutic) molecules could be attached to the NMOFs via PSM. Alkylation of the NH2-BDC ligands (Fig. 2, entry 3) with an organic fluorophore (Br-BODIPY) was used to integrate an imaging component. Similarly, acylation (Fig. 2, entry 1) with a cisplatin prodrug (ESCP = c,c,t-[PtCl2(NH3)2(OEt)(O2CCH2CH2CO2H)]) was used to provide a therapeutic payload (Fig. 5). The fluorescence of the BODIPY dye was quenched by the Fe(III) centers of the NMOF. The quenching allowed for degradation of the NMOF components to be monitored by emergence of BODIPY emission upon release from the particle. The NMOFs showed a t1/2 of ∼2.5 h in PBS buffer at 37 °C; however, encapsulation of the particles in silica (Fig. 5) increased this t1/2 to ∼16 h.17 The dye-loaded particles were shown to be taken up by HT-29 human colon adenocarcinoma cells. Treatment of the same cell line with the ESCP-NMOFs resulted in cytotoxicity quite comparable to known platinum drugs.17 Overall, this study showed the potential of NMOFs in biomedicine and how PSM will play a central role in their development.
Fig. 5 Preparation of fluorophore- and drug-loaded NMOFs using PSM. Combining the two approaches could lead to a complete ‘theranostic’ NMOF. Encapsulation in silica is represented by the yellow spheres. |
Bioorthogonal chemistry | Postsynthetic modification | |
---|---|---|
Solvent limitations | Water, buffer, biological media | Non-caustic solvents |
Reaction conditions | Mild, no toxic by-products | Mild, no degrading (strongly acidic) by-products |
Introduction of chemical ‘handle’ | Reaction must occur with native functional groups or a synthetic ‘handle’ must be readily introduced by chemical biology | Reaction must proceed with MOF ‘as is’ or otherwise a ‘handle’ compatible with solvothermal synthesis must be provided |
Reaction selectivity | Reagents must be highly selective to avoid reactivity with native functional groups | Reagents must be selective for groups installed on the MOF framework |
In essence, the challenges facing bioorthogonal chemistry parallel those in the PSM of MOFs. While the chemical biologist tries to modify a select biomolecule without denaturing the substrate or killing the cell, the MOF chemist tries to modify the framework without degrading the porous, crystalline structure of the MOF. However, the chemical biologist is largely confined by the limits of the biological medium, while the MOF chemist has the opportunity to broaden the scope of chemical reactions to be used for PSM with the ongoing development of more chemically robust and porous MOFs (e.g. MILs, ZIFs).19,21 The opportunities and significance of PSM on MOFs are only just now being revealed, and certainly many exciting discoveries will be reported in the near future.
The author would like to thank the University of California, San Diego, the National Science Foundation (CHE-0546531; instrumentation grants CHE-9709183, CHE-0116662, and CHE-0741968, MOF synthesis), and the Department of Energy (DE-FG02-08ER46519, gas sorption by MOFs) for support of research projects related to the PSM of MOFs.
This journal is © The Royal Society of Chemistry 2010 |