1-s2.0-s0010854509001271-main
DESCRIPTION
MOFTRANSCRIPT
-
Coordination Chemistry Reviews 253 (2009) 30423066
Contents lists available at ScienceDirect
Coordination Chemistry Reviews
journa l homepage: www.e lsev ier .com/ locate /ccr
Review
Potenti
Ryan J. KMark D. Ya Department ob Department o
Contents
1. Introd2. Synth
2.1.2.2.
3. Hydro3.1.3.2.3.3.
4. Select4.1.4.2.
4.3.4.4.
5. Cataly5.1.5.2.5.3.5.4.5.5.5.6.5.7.
6. Magn6.1.6.2.
6.3.7. Lumin
7.1.7.2.
7.3.
Correspon
0010-8545/$ doi:10.1016/j.cal applications of metal-organic frameworks
uppler a, Daren J. Timmons b, Qian-Rong Fang a, Jian-Rong Li a, Trevor A. Makal a,oung a, Daqiang Yuan a, Dan Zhao a, Wenjuan Zhuang a, Hong-Cai Zhou a,
f Chemistry, Texas A&M University, PO Box 30012, College Station, TX 77842-3012, United Statesf Chemistry, Virginia Military Institute, 303 Science Hall, Lexington, VA 24450, United States
uction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3043esis and structure of MOFs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3043Synthetic routes to metal-organic frameworks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3043Structural highlights of metal-organic frameworks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3043
gen and methane storage in MOFs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3045Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3045Hydrogen storage in MOFs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3045Methane storage in MOFs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3048
ive gas adsorption in MOFs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3048Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3048Selective gas adsorption in MOFs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30494.2.1. Selective gas adsorption of O2 over N2 based on size-exclusion in rigid MOFs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30494.2.2. Selective gas adsorption based on size-exclusion in dynamic MOFs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30494.2.3. Selective gas adsorption based on adsorbatesurface interactions in rigid MOFs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30494.2.4. Selective gas adsorption based on adsorbatesurface interactions in dynamic MOFs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3049Useful gas separations in MOFs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3049MAMS in selective gas adsorption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3050
sis in MOFS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3050Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3050Catalysis in MOFs with active metal sites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3050Catalysis in MOFs doped with metal catalysts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3050Catalysis in post-synthesized MOFs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3051Selective catalysis in MOFs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3051Catalysis in chiral MOFs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3053Future research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3053
etic properties of MOFs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3053Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3053Magnetic properties of MOFs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30546.2.1. Ferromagnetic properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30546.2.2. Antiferromagnetic properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30556.2.3. Ferrimagnetic properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30566.2.4. Frustration and canting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3057Spin-crossover and induced magnetic change in MOFs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3058
escence and sensors in MOFs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3059Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3059Luminescence in MOFs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30597.2.1. Luminescence in MOFs based on metal centers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30597.2.2. Luminescence in MOFs based on organic ligands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30607.2.3. Luminescence based on guest molecules in MOFs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3060
Sensors in MOFs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30607.3.1. Sensors for selective ion monitoring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30607.3.2. Sensors for the presence and/or types of guest/solvent molecules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3061
ding author. Tel.: +1 979 845 4034.
see front matter 2009 Elsevier B.V. All rights reserved.cr.2009.05.019
-
R.J. Kuppler et al. / Coordination Chemistry Reviews 253 (2009) 30423066 3043
7.3.3. Sensors for stress-induced chemical detection. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30617.3.4. Sensors for anisotropic photoluminescence probes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3062
8. Drug8.1.8.2.
8.3.9. Concl
AppeRefer
a r t i c l
Article history:Received 16 JaAccepted 25 MAvailable onlin
Keywords:Metal-organicStructurePotential applApplicationsPorous materi
1. Introdu
Porous mbased gas/vstorage anddimensionabeen eithermon organiprepared byareas and hstructures.many uses,cation of w
Inorgani(e.g. zeolitetemplate wframeworkquence, remframework.sity, as theSi and chalcuseful in se
In orderand inorganorganic framand orderethat MOFsporous coothe same gture merelywho constrformed inmetal ionscinating strgreat deal onumber ofremarkableration, sizeand uoresexplored [8storage and delivery in MOFs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3062Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3062Drug-delivery methods. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30628.2.1. Inorganic drug-delivery materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30638.2.2. MOF drug-delivery materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3063Future work in drug delivery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3064
usion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3064ndix A. Supplementary data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3064ences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3064
e i n f o
nuary 2009ay 2009e 28 June 2009
frameworks
ications
als
a b s t r a c t
Metal-organic frameworks have received much attention in recent years especially as newly developedporous materials. As such, they possess a wide array of potential applications including materials for gasstorage, gas/vapor separation, catalysis, luminescence, and drug delivery. In the review, these potentialapplications of metal-organic frameworks are examined and an outlook will be proposed.
2009 Elsevier B.V. All rights reserved.
ction
aterials are very useful in gas storage, adsorption-apor separation, shape/size-selective catalysis, drugdelivery, and as templates in the preparation of low-
l materials [16]. Traditionally, porous materials haveorganic or inorganic materials. Perhaps the most com-c porous material is activated carbon. These are usually
pyrolysis of carbon-rich materials, have high surfaceigh adsorption capacities, yet do not possess ordered
Despite this lack of order, porous carbon materials haveincluding the separation and storage of gases, the puri-ater, and solvent removal and recovery [7].c porous frameworks possess highly ordered structuress). Syntheses often require an inorganic or organicith strong interactions forming between the inorganicand the template during the synthesis. As a conse-oval of the template can result in collapse of the
Inorganic frameworks also suffer from a lack of diver-variation of elements used seldom deviates from Al,ogens. Nevertheless, inorganic frameworks have been
paration and catalysis applications [3].to take advantage of the properties of both organicic porous materials, porous hybrids, known as metal-eworks (MOFs), can be generated that are both stable
d and possess high surface areas. It should be notedgo by many names (porous coordination networks,
rdination polymers, etc.) yet all refer to similar if noteneral type of materials. The difference in nomencla-
reects the type of framework and the researchersucted it. MOFs are essentially coordination polymersthe most elementary sense by connecting togetherwith polytopic organic linkers often resulting in fas-uctural topologies. These materials have attracted af attention in the past decade, and the increase in thepapers published in this area during recent years is(Fig. 1). Applications in gas storage, gas/vapor sepa-
-, shape-, and enantio-selective catalysis, luminescentcent materials, and drug storage and delivery have been10].
2. Synthesis and structure of MOFs
2.1. Synthetic routes to metal-organic frameworks
MOFs are typically synthesized by combining organic ligandsand metal salts in solvothermal reactions at relatively low tem-peratures (below 300 C). The characteristics of the ligand (bondangles, ligand length, bulkiness, chirality, etc.) play a crucial role indictating what the resultant framework will be. Additionally, thetendency of metal ions to adopt certain geometries also inuencesthe structure of the MOF. The reactants are mixed in high boiling,polar solvents such as water, dialkyl formamides, dimethyl sulfox-ide or acetonitrile. The most important parameters of solvothermalMOF synthesis are temperature, the concentrations of metal saltand ligand (which can be varied across a large range), the extent ofsolubility of the reactants in the solvent, and the pH of the solution.Although experience often dictates the best conditions for growingthese crystalline frameworks, experimentation and trial-and-errormethods are still often necessary. Reviews which describe the syn-thesis and characterization of MOFs have been previously published[1116].
In addition to this standard method, several other syntheticmethodologies are described in the literature including the mix-ture of non-miscible solvents [17], an electrochemical route [18],and a high-throughput approach [19]. One of the most promisingalternatives is microwave irradiation which allows access to a widerange of temperatures and can be used to shorten crystallizationtimes while controlling face morphology and particle size distribu-tion [12,20,21]. A serious limitation of this approach is the generallack of formation of crystals large enough to obtain good structuraldata.
2.2. Structural highlights of metal-organic frameworks
When considering the structure of MOFs, it is helpful to recog-nize the secondary building units (SBUs), which dictate the naltopology of a framework. While the organic linkers are also impor-tant SBUs, their structure seldom changes during MOF assembly.The following discussion will focus on metal-cluster-based SBUs,which result from the initial bonding between the metal ions and
-
3044 R.J. Kuppler et al. / Coordination Chemistry Reviews 253 (2009) 30423066
Fig. 1. Number of publications on MOFs over the past decade.
bridging ligands [16]. Fig. 2a and b shows SBUs with a trigonal and asquare planar arrangement of the metal atoms, respectively. Fig. 2cfeatures a tetrahedron of metal atoms surrounding a central oxoanion, and a dimetal paddlewheel SBU is shown in Fig. 2d. In eachcase, the edges between two metal ions are bridged by the coordi-nating atoms of the ligand and result in control of the orientationof the linker.
Recently, the geometry of the SBU has been proven to be depen-dent on not only the structure of the ligand and type of metalutilized, but also the metal to ligand ratio, the solvent, and thesource of anions to balance the charge of the metal ion [22]. Severalpublicationdepth [16,2
Pores armaterials) u
Fig. 2. Structusquare planar,
Fig. 3. Lookinof PCN-9 [28].reader is refer
general, larchemistrybetween 20than 500
less tlecustor
entsctingby gpor
ust bcur ae. Pe
and is more difcult to achieve in mesoporous MOFs than inorous analogues.ough MOFs can be constructed with ligands designed to gen-rge pores, frameworks will often interpenetrate one anotherimize packing efciency [23]. In such cases, the pores sizesatly reduced, but this may be benecial for some applica-s discussed the topic of SBU formation and structure in326].e the void spaces formed within MOFs (or any porous
pon the removal of guest molecules (Fig. 3) [27]. In
poresgas mofor gassuremsubtraaccess
Thethat mcan occollapsintactmicrop
Altherate lato maxare greral representations of several SBUs, including (a) trigonal planar, (b)(c) tetrahedral, and (d) tetragonal paddlewheel.
tions. Indeeformed andin H2 storag
Followinmers, maythe framewknown as ppresence oformation oto rst formfunctional gthat are detions requirpore geomeimportant fenhance thof the MOFframeworkg through the window created by the ligands into the pore (red sphere)(For interpretation of the references to color in this gure legend, thered to the web version of the article.)
ge pores are advantageous for conducting hostguestsuch as catalysis, therefore mesoporous (openings
and 500 ) or even macroporous (openings greater) materials are attractive. Microporous materials havehan 20 which result in strong interactions betweenles and the pore walls making them good candidatesage and gas separation applications. In all cases, mea-of these openings are done from atom to atom whilethe van der Waals radii to give the space available for
uest molecules.es of MOFs are usually occupied by solvent molecules
e removed for most applications. Structural collapsend, in general, the larger the pore, the more likely thermanent porosity results when the framework remainsd interpenetrated frameworks have been intentionallyfound to lead to improved performance, for example,e [29].g the synthesis, MOFs, like other coordination poly-participate in further chemical reactions to decorateorks with molecules or functional groups in what isost-synthetic modication (PSM) [30]. Sometimes the
f a certain functional group on a ligand prevents thef the targeted MOF. In this situation, it is necessarya MOF with the desired topology, and then add the
roup to the framework. This may be applied to MOFssigned for catalysis and gas storage, as these applica-e functional groups to modify the surface property andtry. It is important to keep in mind that the two mostactors in PSM are making sure that the reagent used toe functionality is small enough to t inside the cavityand that the reaction conditions will not destroy the
. If the reagent is too small to enter the cavity or the
-
R.J. Kuppler et al. / Coordination Chemistry Reviews 253 (2009) 30423066 3045
framework is destroyed by the reaction, the modication will beuseless.
3. Hydroge
3.1. Introdu
A tank cstored at a madsorbent.can be avoimethod. Maadsorbentslites [6]. MOdue to theirneed to redtive technoforefront ofstate-of-the
3.2. Hydrog
Hydrogemetric heat[31], and thmakes hydrhydrogen epressure wwhich requthe density(700 kg/msity presenfuel. In ordeDepartmenage targetsand 2015 (9
Currentpressure tanPure tank-bissues [33].chemical borials, leadinkinetics, rev[34]. Physis(mainly vangen and themanageabledata from pat a cryogeinsignican
In 2003,age study [3their hydroon this topithis sectionserve as an
Hydrogewell establhydrogen uface area ofincreasing tand the ads
Comparesurface areity. One of
which has a BET surface area of 4500 m2/g and an excess gravi-metric hydrogen uptake of 7.5 wt% at 70 bar, 77 K [42,43]. However,when the temperature was raised to ambient temperature, thehydrogen uptake dropped signicantly, as is the general case. The
ticaltionmper5 barpy cload
orptiof 4
adsing tant.wa
is toerlapn hyt tha
ve kin difinterpara
etation,ach ozes fodisc
and c(isored thgenire [5t wh
res. Amorper
ree IRa pa
(Figd thad 13imetroge
howsK/50
50 bad thais st
. 6).he atgs.
adve M
y the[41
higheeat o], verrepa
red thave
wt%/50 bn and methane storage in MOFs
ction
harged with a porous adsorbent enables a gas to beuch lower pressure than an identical tank without an
Thus, high pressure tanks and multi-stage compressorsded providing a safer and more economical gas storageny gas storage studies have been conducted on poroussuch as activated carbon, carbon nanotubes, and zeo-Fs have received growing attention as such adsorbentstunable pore geometries and exible frameworks. The
uce global reliance on fossil fuels by the use of alterna-logies has pushed hydrogen and methane gases to the
gas storage applications. This section will review the-art study of hydrogen and methane storage in MOFs.
en storage in MOFs
n is an ideal energy carrier. It almost triples the gravi-of combustion of gasoline (120 MJ/kg vs. 44.5 MJ/kg)e main byproduct after energy release is water. Thisogen a leading candidate for on-board fuel. However,xists in a gaseous state at ambient temperature andith a density of 0.08 kg/m3. Even in its liquid state,ires pressurizing at a very low temperature (20.27 K),can only reach 70.8 kg/m3, one tenth of that of gasoline3) [31]. This extremely low volumetric storage den-ts a hurdle for the practical usage of hydrogen as ar to guide the research into hydrogen storage, the U.S.
t of Energy (DOE) set gravimetric and volumetric stor-for on-board hydrogen storage for 2010 (6 wt%, 45 g/L)wt%, 81 g/L) [32].hydrogen storage techniques involve the use of highks, cryogenic tanks, chemisorption, and physisorption.
ased hydrogen storage suffers from safety and economicThe chemisorption approach allows the formation ofnds between adsorbed hydrogen and the storage mate-g to greater hydrogen storage density. However, theersibility and heat management still remain a problem
orption, on the other hand, is based on weak interactionsder Waals interactions) between the adsorbed hydro-
adsorbent, leading to fast kinetics, full reversibility, andheat during hydrogen fueling. However, the promisinghysisorption-based hydrogen storage are all obtained
nic state (normally 77 K), and the adsorption becomest at ambient temperature.Rosi et al. reported the rst MOF-based hydrogen stor-5]. Subsequently, about 150 MOFs have been tested for
gen uptake capacity [13], and several reviews focusingc have appeared [13,3641]. In order to avoid repetition,of the review will cover the most current literature andupdate to Ref. [13].n storage in MOFs is based on physisorption. It has beenished that under physisorption mode, the saturationptake at 77 K has a positive correlation with the sur-the materials [13,38,40,42]. This is not surprising sincehe surface area enhances the contact between hydrogenorbent resulting in an increased hydrogen uptake.d to other porous materials, some MOFs have higher
as and subsequently higher hydrogen uptake capac-the benchmarks is provided by the study on MOF-177,
theoreadsorpent teand 1.enthalwholephysisrangeas theincreasimportstorage
OneMOFstial ovbetweesupporeffectineutrostrongis comrated mCatenatrate epore siever, aThe grIRMOFproposat cryoperaturange apressuadsorbent temthe theratedPCN-6showearea anin gravon hydture sat 298298 K/showelinkers6 (Figfrom tloadin
Onein somated bvacuummuchteric h[52,53et al. pcompaUMCsvs. 2.07at 77 Kstudy from Bhatia and Myers demonstrated that anenthalpy change of 15.1 kJ/mol is needed for ambi-ature storage of hydrogen and delivery between 30of pressure [44]. More importantly, the adsorption
hange should be kept at the same level during theing range. However, in current MOF-based hydrogenon, the adsorption enthalpy change is only within the10 kJ/mol [13], and the enthalpy drops dramatically
orption amount increases. It has become clear thathe interaction between hydrogen and MOFs is the moststep towards practical usage of MOF-based hydrogen
y to increase the interaction between hydrogen andtailor the pore size in the MOF to maximize the poten-
of the walls thereby allowing enhanced interactiondrogen and MOFs. Theoretical and experimental resultst the optimal pore size is around 6 , about twice the
netic diameter of the hydrogen molecule [45]. A recentfraction study on rare earth MOFs indicated that theaction between hydrogen and the optimized pore wallsble to the interaction between hydrogen and unsatu-l centers, a point which will be discussed later [46].when two or more identical frameworks interpene-
ther, could be used to generate MOFs with appropriater enhanced hydrogen uptake [36,4750]. There is, how-
repancy between the theory and experimental results.anonical Monte Carlo (GCMC) simulations done on theeticular metal-organic framework) series by Ryan et al.at catenation can improve hydrogen uptake in IRMOFs
c temperatures and low pressures, but not at room tem-1]. Based on these calculations, there is a temperatureich catenation is benecial for hydrogen uptake at lowt higher pressures, however, the non-catenated IRMOFs
e hydrogen due to a greater free volume (Fig. 4). At ambi-ature, catenation does not improve hydrogen uptake for
MOFs studied. Using a template, Ma et al. have gen-ir of catenated and non-catenated MOFs, PCN-6 and
. 5) [49]. The low pressure and 77 K gas sorption resultst catenation yields a 41% increase in Langmuir surface3% enhancement in volumetric hydrogen uptake (29%ric). A further investigation of the effect of catenationn uptake at higher pressure and ambient tempera-promising results: 6.7 wt% at 77 K/50 bar (0.92 wt%
bar) for PCN-6 vs. 4.0 wt% at 77 K/50 bar (0.40 wt% atr) for PCN-6 [29]. Inelastic neutron scattering studiest the interaction between hydrogen and the organic
ronger in catenated PCN-6 than in non-catenated PCN-This may stem from a greater number of interactionsoms in the organic ligands, especially at high hydrogen
antage of using MOFs over other porous materials is thatOFs, unsaturated metal centers (UMCs) can be gener-
removal of the coordinated solvent molecules under]. The interaction between hydrogen and the UMCs isr than that with pure carbon materials, and the isos-
f adsorption can sometimes go as high as 1213 kJ/moly close to the projected optimum 15.1 kJ/mol [44]. Leered isostructural MOFs with and without UMCs, andheir hydrogen sorption capacities [54]. MOFs containinghigher hydrogen uptake both at low pressure (2.87 wt%at 77 K/1 atm) and high pressure (5.22 wt% vs. 3.70 wt%ar) than MOFs with saturated metal centers. The higher
-
3046 R.J. Kuppler et al. / Coordination Chemistry Reviews 253 (2009) 30423066
Fig. 4. Simulamission from R
hydrogen ution of hydsupported b(11.60 kJ/mo
It has being units coadsorbed g[55]. By linkdlewheel mto generatenally aligne[56]. The r(PCN-12) hone (PCN-1UMCs-alignUMCs will
Fig. 5. (a) Catefrom Ref. [49].
xcess hydrogen sorption isotherms of PCN-6 and PCN-6 at 77 K (red) andlack): circles, PCN-6; squares, PCN-6; solid symbols, adsorption; open sym-orption. Reproduced with permission from Ref. [29]. (For interpretation ofences to color in this gure legend, the reader is referred to the web versionticle.)
ide the void and higher afnity for the subsequent adsorbeden.oretical studies revealed that the binding energy betweenen and transition metals can be tuned from about 10 tool1 by using different transition metals in the MOF sys-7]. This was partially conrmed by a recent experiment ina series of isostructural MOFs were prepared from a vari-Fig. 6. E298 K (bbols, desthe referof the ar
lap inshydrog
Thehydrog50 kJ mtem [5whichted hydrogen adsorption isotherms on IRMOFs. Reproduced with per-ef. [51].
ptake capacity was attributed to the stronger interac-rogen molecules with the UMCs of the MOF, and wasy the higher zero-coverage isosteric heat of adsorptionl vs. 7.24 kJ/mol).
en suggested that MOFs with cage-like polyhedral build-uld be useful in the storage of small molecules, becauseuests may remain kinetically trapped inside the cagesing two isophthalate moieties with the dicopper pad-otif, Wang et al. developed the close-packing strategyMOFs with polyhedral cage structures and UMCs ratio-d to interact directly with the guest inside the void
esult was encouraging: the MOF with aligned UMCsas much higher hydrogen uptake than the misaligned2) (3.05 wt% vs. 2.40 wt%, 77 K/1 atm, Fig. 7). In theed-MOF, the initially adsorbed hydrogen around thepoint into the void, leading to higher potential over-
nated PCN-6. (b) Non-catenated PCN-6 . Reproduced with permission Fig. 7. The symorphs: PCN-nthesis, UMCs alignment, and hydrogen uptake of two MOF poly-12 and PCN-12 . Adapted from Ref. [56].
-
R.J. Kuppler et al. / Coordination Chemistry Reviews 253 (2009) 30423066 3047
Fig. 8. Reaction of Zn4O(BDC)3 with Cr(CO)6 to generate Zn4O[(BDC)Cr(CO)3]3 (1),followed by photolysis under N2 or H2 to afford Zn4O[(BDC)Cr(CO)2(N2)]3 (3) andZn4O[(BDC)Cr(CO)2(H2)]3 (4). Reproduced with permission from Ref. [65].
ety of metal ions (Mg, Mn, Co, Ni, Zn) [58]. The hydrogen sorptiondata showeheat of adsoest (12.9 kthe IrvingWprevious cathe UMCs aVitillo et al.different naTheir resulthydrogen aalmost comarea and pogen storageof UMCs is
Many thMOFs with[6164]. Thinteractionsthis idealis very harddevelopedinto MOFs
very limited, and the gravimetric hydrogen uptake capacity wasimpaired due to the added mass. Mulfort and Hupp reported thedoping of a catenated MOF with lithium ions by chemical reduc-tion of the ligand with lithium metal [66]. The lithium-doped MOFexhibited a 75% increase in gravimetric hydrogen uptake (1.63 wt%vs. 0.93 wt% at 77 K/1 atm) and a higher heat of adsorption through-out the entire loading range (65 kJ/mol vs. 5.63.3 kJ/mol). Theextreme enhancement of hydrogen binding (60 H2 molecules peradded Li+) was attributed to the direct H2/cation binding, as wellas other factors such as framework displacement and/or enhancedstrut polarizability.
A further study of the alkali metal cation effects on hydro-gen uptake and binding in catenated MOFs was carried out usingLi+, Na+, and K+ [67]. The 77 K/1 atm gravimetric hydrogen uptakeincrease followed as K+ > Na+ > Li+, while the isosteric heat ofadsorption adopted the opposite trend (Fig. 9). Two observationscast doubt on the hydrogenmetal binding mechanism: (1) anextraordinarily large number of hydrogen molecules adsorbed perdopant cation; and (2) a much smaller isosteric heat of adsorptioncompared to the calculated result. More importantly, further dopingwith K+ led to decreased hydrogen uptake. After careful examina-tion of the nitrogen isotherms, the alkali metal cation effects wereattributed to a molecular-adsorbate-driven displacement of inter-woven networks, as well as the possible positioning of the dopant
tweerogened tadsoemetakea noetricisplaginalreaseere we excllove
ragef dialp ofit hyhe instrat
beddtion
Fig. 9. Left: 77from Ref. [67].d that within this series the zinc MOF has the lowestrption (8.5 kJ/mol) while the nickel one has the high-J/mol), and the order of the heat of adsorption matches
illiams sequence [59] reasonably well. This conrmedlculations indicating that the major interaction betweennd hydrogen molecules is Coulombic attractions [60].studied the role of UMCs in hydrogen storage in MOFs oftures and with different accessibilities of the UMCs [53].s indicated that UMCs can increase the binding betweennd MOFs, but their effect in the hydrogen uptake ispletely hidden in the high pressure range where surfacere volume play primary roles. The efciency of hydro-by physisorption should increase as the surface density
increased.eoretical calculations support the idea that doping
metal ions could enhance the hydrogen uptake capacityis enhancement is proposed to originate from the strongbetween hydrogen and the doped metal ions. However,metal ion geometry adopted in the calculated studiesto achieve experimentally. Even though Kaye and Long
a photosensitive doping method to embed chromium(Fig. 8) [65], additional open coordination sites were
ions beto hydincreasheat ofdisplacgen upwithingravimMOF dthe orithe incally, thand th
Spigen stosplit othe hethe spl[70]. Tdemonal. emdeposiK hydrogen isotherm for neutral MOF-1 and metal ion doped-MOF 1 M; Right: isosteric hyn the frameworks where they are not readily accessible. More recently it was reported that the metal doping
he hydrogen uptake without a signicant change of therption [68]. In order to rule out the effect of framework
nt in evaluating the doping ions impact on the hydro-, Yang et al. prepared a Li+-doped MOF by ion exchangen-catenated MOF [69]. Surprisingly, a 25% increase inhydrogen uptake was observed while the Li+-doped
yed a lower isosteric heat for hydrogen adsorption thanMOF. The increase in hydrogen uptake was attributed toof BET surface area in the Li+-doped MOF, and, addition-as no apparent interaction between adsorbed hydrogenhanged Li+.r techniques have been used to further advance hydro-
in MOFs. Hydrogen storage by spillover involves thetomic hydrogen molecules into hydrogen atoms with
heavy transition metal (e.g. Pd) and the diffusion ofdrogen atoms onto the supportive framework (e.g. MOF)itial results of MOF-based hydrogen storage by spillovered promising capacity and reversibility [71]. Proch eted Pt into MOF-177 via metal-organic chemical vaporand tested the hydrogen uptake capacity at ambient
drogen heat of adsorption for 1 and 1 M. Reproduced with permission
-
3048 R.J. Kuppler et al. / Coordination Chemistry Reviews 253 (2009) 30423066
temperature [72]. The rst cycle showed an encouraging 2.5 wt%hydrogen uptake, but dropped to 0.5 wt% in consecutive cycles.The uptake loss was attributed to the formation of metal hydrides,which are not desorbed at ambient temperature.
Ionic MOFs are another proposed approach towards practicalhydrogen storage in which the binding between hydrogen andMOFs can be enhanced by the attractive electrostatic interactions[7375]. A calculation was performed by Kuc et al. in which thehydrogen molecule was placed in a ctitious linear Na+H2Cl
complex, where Na+ and Cl were placed on opposite sides ofhydrogen at a distance of 3.2 from the hydrogen center [76]. Thedipole component at the center hydrogen molecule was calculatedto be 0.73 (0.97) Debye using the NBO (Mulliken) charges; however,it is extremely difcult to introduce such a large charge separationin real MOF systems.
Despite the rapid advancement in MOF research, the DOE hydro-gen storage target has yet to be reached with a pure MOF system.However, due to novel properties and well-dened structures,MOFs are still attracting much attention. For example, supercrit-ical processing in the MOF activation step permits the opening ofa gate of even higher surface area [77]. As reported from the highpressure hydrogen storage study done on IRMOF-1, the upper-limithydrogen density based on the physisorption method would hardlygo beyond the density of liquid hydrogen (70.8 g/L) [78]. It was fur-ther concluded that the development of a new method would benecessary to reach DOEs 2015 volumetric hydrogen storage target(81 g/L).
3.3. Methan
NaturalThe main cmixture of eide [79]. Mewith gasolinlack of effeabout 72% ocryogenic coffer aboutoperates atnatural gasadsorbed ondensity forthat of CNG
Unlike for hydrogen, the heat of adsorption for methane (about20 kJ/mol) is already within the ideal scope for practical usage. DOEhas set a methane storage target: 180 v/v at ambient temperatureand pressure no more than 35 bar [80]. Some of the carbon materialshave already reached this target [79], but they have limited packingdensity. Thus, the focus has been on increasing the surface area ofthe porous sorbent.
In 1997, Kondo et al. reported the rst methane sorption studyusing MOFs [81]. Table 1 summarizes the surface area, porosity andmethane uptake data for selected MOFs. The breakthrough resultobtained by Ma et al. showed that the methane uptake in a MOFcan exceed the DOE target [82].
One point of concern is that uptake data calculations are basedon the MOFs crystallographic density, which is higher than thepacking density due to the void generated by particle packing [83].More methane uptake data needs to be calculated from the MOFsreal packing density from MOFs in order to evaluate the potentialof MOFs in methane storage.
4. Selective gas adsorption in MOFs
4.1. Introduction
With the ever-increasing demand for cheaper and more environ-mentally friendly industrial applications, new means for selective
sorprial ms wetion-s, moel [8stryer, almai
) theties aadsourin
ectivlar, mpe onabled se
].
Table 1Surface area, p
Materiala (g cm
Co2(4,4-bipy)Cu2(pzdc)2(pyCu2(pzdc)2(4,4Cu2(pzdc)2(piaCuSiF6(4,4-bipZn4O(R6-bdc)3MIL-53(Al) Al(MIL-53(Cr) Cr(PCN-14 Cu2(adPCN-11 Cu2(sbHKUST-1 Cu3(Zn2(bdc)2dabcMIL-101 Cr3FO
a 4,4-bpy = -(pyri3,6-dicarboxyl iyl)di-sbtc = trans-sti abicy
b Calculatede storage in MOFs
gas (NG) is another good candidate for on-board fuel.omponent of NG is methane (>95%), while the rest is athane, other hydrocarbons, nitrogen, and carbon diox-
thane has a comparable gravimetric heat of combustione (50.0 MJ/kg vs. 44.5 MJ/kg) [31], but it suffers from the
ctive storage. Liqueed natural gas (LNG) can providef the volumetric energy density of gasoline yet requiresonditions (112 K). Compressed natural gas (CNG) can26% of the volumetric energy density of gasoline but
pressures about 200 bar for compression [79]. Adsorbed(ANG) provides another storage method in which NG is
a porous adsorbent, and the volumetric storage energyANG (at 500 psig) has been reported to be up to 80% of(at 3000 psig) [79].
gas adindustgenic aadsorpzeolitesilica gin induHowevon twoand (2properof theature dfor selparticunew tyand tution anRef. [2
orosity, and methane uptake data for selected MOFs.
Surface area (m2 g1) Pore volume (cm3 g1) Densityb
BET Langmuir
3(NO3)4 1.36z) 1.75-bipy))y)2 0.86IRMOF-6 2630 0.60 0.65
OH)(bdc) 1100 1590 0.59 0.98OH)(bdc) 1100 1500 0.56 1.04ip) 1753 0.87 0.83tc) 1931 2442 0.91 0.75
btc)2 1502 2214 0.76 0.88o 1448 2104 0.75 0.87(bdc)3 2693 4492 1.303 0.31
4,4-bipyridine, pzdc = pyrazine-2,3-dicarboxylate, pyz = pyrazine, pia = Nate, dbc = benzenedicarboxylate, adip = 5,5-(9,10-anthracenedlbene-3,3 ,5,5-tetracarboxylate, btc = benzene-1,3,5-tricarboxylate, dabco = 1,4-diazfrom single crystal structure without guest molecule and labile ligand binding.tion and separation are being examined. Currently,ethods for selective gas adsorption rely heavily on cryo-ll as membrane- and adsorption-based techniques. Inbased separation, commonly used adsorbents includelecular sieves, carbon nanotubes, aluminosilicates, and
994]. A review of the individual properties and usagesof these materials is beyond the scope of this review.l materials for selective gas adsorption are chosen basedn criteria: (1) the adsorption capacity of the adsorbent;selectivity of the adsorbent for an adsorbate [2]. Thesere dictated by the chemical composition and structure
rbent, as well as the equilibrium pressure and temper-g the adsorption. MOFs are very promising candidatese gas adsorption, which can lead to gas separation. In
esh-adjustable molecular sieve (MAMS) [95] is such af MOFs that stands out due to its remarkable selectivity
properties. A complete review of selective gas adsorp-paration in metal-organic frameworks can be found in
3) Ambient temperatureCH4 uptake
Hads (kJ mol1) Ref.
wt% v/v
3.6 (30.4 bar) 71 [81]1.3 (31.4 bar) 32 [84]3.9 (31.4 bar) [84]4.4 (31.4 bar) [84]9.4 (36.5 bar) 124 [85]
14.7 (36.5 bar) 155 [86]10.2 (35 bar) 155 17 [87]10.2 (35 bar) 165 17 [87]16.0 (35 bar) 220 30 [82]14.1 (35 bar) 171 14.6 [88]15.7(150 bar) 228 [83]14.3 (75 bar) 202 [83]14.2 (125 bar) 72 [83]
din-4-yl)isonicotinamide, R6-bdc = 1,2-dihydrocyclobutabenzene-isophthalate, abtc = azobenzene-3,3 ,5,5-tetracarboxylate,clo[2.2.2]octane.
-
R.J. Kuppler et al. / Coordination Chemistry Reviews 253 (2009) 30423066 3049
Fig. 10. HydroPCN-13 at 77 K
4.2. Selectiv
The printion is achisize-exclusichemical anthe adsorbaof the framtwo effectseratively.
4.2.1. Selectsize-exclusio
Selectivehas been s13 (Zn4O(Hchannel ofThe largeras opposedpletely preO2 to passMg3(ndc)3,demonstratapproximatH2O)(C24H1connected btimes as mu
4.2.2. Selecdynamic MO
The caseclassied aior. Dynaminot only onditions (poand/or areNeverthelesmainly on sNi2(cyclametrated netenough for
4.2.3. Selecinteractions
Adsorbain the selecwalls of the
nation sites, leading to high afnity for O2. This MOF can adsorb14 mmol of O2 but only 3 mmol of N2 [100].
Selective gas adsorption based on adsorbatesurfacetions in dynamic MOFsadsorbentguest interactions in dynamic MOFs have also
strated adsorption selectivity of O2 over N2 [99,101103].ding on the interaction between the adsorbate and the adsor-ore expansion or constriction also occurs. This is known ase-opening process [2]. Characteristic gate-opening selectivetion of O2 over N2 has been observed in Cd(bpndc)(4,4-bpy),ch approximately 150 mL/g of O2 was adsorbed with almostuptake [103]. This signicant difference in gas uptakes wasted to the interaction of the O2 molecules with the frame-
which is stronger than that of N2. Similar results have beenwith Cu(dhbc)2(4,4-bpy) and Cu(bdc)(4,4-bpy)0.5 but withtake of O2 [101,102].
seful gas separations in MOFs
separation of alkane isomers from natural gas is of primaryn for08 mraphary fs bys to
betnalles thentlyaid iin
tionOF desed s). Thlicatentlyd CHe COhigh
demh a Meparions,gen, oxygen, nitrogen, and carbon monoxide adsorption isotherms of. Reproduced with permission from Ref. [96].
e gas adsorption in MOFs
cipal mechanisms based on which selective gas adsorp-eved in MOFs are adsorbatesurface interactions andon (molecular sieving effect). The former involves thed/or physical interaction between the adsorbent and
te while the latter depends on the dimension and shapeework pores. It is important to keep in mind that theare capable of working independently as well as coop-
ive gas adsorption of O2 over N2 based onn in rigid MOFs
gas adsorption based on size-exclusion principleuccessfully demonstrated with several MOFs. PCN-2O)3(C16H8O4)32DMF) contains a square hydrophobic3.5 3.5 and has a pore volume of 0.3 cm3/g [96].size of the N2 molecule (kinetic diameter of 3.64 )to the O2 molecule (kinetic diameter of 3.46 ) com-
vents N2 from entering the aperture while allowingthrough the pore as shown in Fig. 10. Likewise,
with its pore opening of approximately 3.463.64 ,es almost no N2 uptake at 77 K and 880 Torr whereasely 3.5 mmol/g of O2 is adsorbed [97]. PCN-17 (Yb4(4-2N3O6)8/3(SO4)23H2O10DMSO), with its large cagesy small apertures, is able to selectively adsorb over tench O2 as N2 [98].
tive gas adsorption based on size-exclusion inFss in the foregoing discussion involve MOFs that are
s rigid and do not display any obvious exible behav-
4.2.4.interac
ThedemonDepenbent, pthe gatadsorpin whino N2attribuwork,shownless up
4.3. U
TheconcerMOF-5matognecessalkanealkaneactionsAdditiomixtur
Recals toMOF thseparaThis Mposses(Fig. 11for app
RecCO2 anBecaushave amentsthrougcient sconditc, or exible MOFs exist and their applications dependthe size of the aperture, but also on adsorption con-
re sizes vary with pressure/temperature adjustmentcontrolled by hostguest induced gate opening) [2].s, the adsorption selectivity of O2 over N2, based
ize-exclusion in dynamic MOFs, has also been observed.)2(mtb) can selectively adsorb O2 because its interpen-work results in narrow channels that are only largeO2 molecules to pass through [99].
tive gas adsorption based on adsorbatesurfacein rigid MOFs
tesurface interactions also play a very important roletive gas adsorption of O2 over N2. For example, the pore
rigid MOF Cu(bdt) contain unsaturated metal coordi-Fig. 11. The seCu3(btc)2 MOFindustrial applications. It has been demonstrated thatay be utilized to separate such alkanes using gas chro-
ic techniques [104]. The 3D pillared layer provides theramework to selectively separate linear and branchedadsorbing the linear ones while allowing the branchedpass because of the difference in van der Waals inter-ween the isomers and the frameworks pores [105].y, this MOF proved effective in separating natural gasrough gas chromatography., MOFs have been incorporated into thin-lm materi-
n gas separation. The copper net supported Cu3(btc)2lm developed by Guo et al. demonstrated the successfulof H2 from H2/N2, H2/CO2, and H2/CH4 mixtures [106].monstrated excellent permeation selectivity for H2 andeparation factors much higher than traditional zeolitese recyclability of this MOF further enhances its potentialion in H2 separation and purication., Couck et al. demonstrated the successful separation of4 gases using amino-functionalized MIL-53(Al) [107].
2 possesses a quadrupole moment, the CO2 moleculesafnity for the amino groups. Breakthrough experi-
onstrated that the weakly adsorbed CH4 is able to passOF-packed column while CO2 is adsorbed. Thus, ef-
ation of the two gases can be performed at ambientas shown in Fig. 12.
paration factors for various H2 gas mixtures by copper net supportedthin lm. Reproduced with permission from Ref. [106].
-
3050 R.J. Kuppler et al. / Coordination Chemistry Reviews 253 (2009) 30423066
Fig. 12. Separation of CO2 and CH4 gas mixtures using amino-MIL-53(Al). Repro-duced with permission from Ref. [107].
4.4. MAMS in selective gas adsorption
One of the most promising materials for selective gas adsorp-tion is the Mare very usdifference ipore sizes iadjusting thbeen reporhydrophilicthrough a schambers athrough thenels and theas the gate tin temperabetween 60where fromon MAMS-1long as the tIndustry-reand have dand N2 at 7at 113 K; CHover iso-C4HCu(etz) also[108,109].
Isostructpared withdisplay temLike MAMSmesh size, w2.9 to 4.6 temperaturtert-butyl gallowing laThis selecti
lene, butane, and, theoretically, any gas that has a kinetic diameterbetween 2.9 and 4.6 .
5. Catalysis in MOFS
5.1. Introduction
As porous materials, MOFs may prove to be very useful incatalysis. Theoretically, the pores of MOFs can be tailored in asystematic way allowing optimization for specic catalytic applica-tions. Besides the high metal content of MOFs, one of their greatestadvantages is that the active sites are rarely different because ofthe highly crystalline nature of the material. Although catalysis isone of the most promising applications of such materials, only afew examples have been reported to date. In these MOFs, size- andshape-selective catalytic applications depend on porosity and thepresence of catalytically active transition-metal centers.
talys
ete meoledhis M
o CO2drupeolitedaraayersann
(OH)Thisto b
hydeundses in
talys
entlyof I
icallyct onatilewas
s tranmetpre
inioOF)b).OF-based mesh-adjustable molecular sieve [95]. MAMSeful when attempting to separate gases in which then size is very small. Unlike in more rigid systems, then MAMS can be tuned across a wide array by simplye temperature (Fig. 13). Such a material has recently
ted by Ma et al. as MAMS-1, a layered structure withchannels and hydrophobic chambers interconnected
ize-adjustable gate [95]. In MAMS-1, the hydrophobicre the site of gas storage, yet the gas must rst passhydrophilic channels. Between the hydrophilic chan-hydrophobic chambers lie two bbdc ligands that serve
hat can be opened or closed with an increase or decreaseture, respectively. By simply varying the temperatureand 300 K, the gate size of MAMS-1 can be adjusted any-2.9 to 5.0 . Thus, fractional adsorption may be attainedfor separation of virtually any multi-component gas so
arget molecules kinetic diameter falls within this range.levant experiments have been carried out on MAMS-1emonstrated selective adsorption of H2 over CO2, O2,7 K; O2 over CO and N2 at 87 K; N2 over CO2 and CH44 over C2H4 at 143 K; C2H4 over C3H6 at 196 K; C3H610 at 241 K [95]. In addition, the MOFs Mn(HCOO)2 andexhibit temperature-dependent gate-opening behavior
ural MAMSs (MAMS-2, MAMS-3, and MAMS-4, pre-metal centers of Zn, Co, and Cu, respectively) alsoperature-dependant molecular sieving effects [110].
-1, a linear relationship exists between temperature andith the new MAMSs mesh size existing anywhere from
when the temperature is adjusted from 77 to 273 K. Ase increases, the van der Waals interactions between theroups of the ligands of the framework decrease, thusrger molecules to pass through the frameworks gate.vity provided selective adsorption of O2, N2, CO, ethy-
5.2. Ca
Zoudiscretimidaz[111]. Tof CO tby quaNiY z
Gnthick lThe chin [Inacid)).provenof aldecompoysis do
5.3. Ca
Reclizationcatalytits effea verswhichvariousulatedin the4-pyridrho-ZM(Fig. 15Fig. 13. The structure of MAMS-1 and the pore opening dependence on temperatis in MOFs with active metal sites
al. have prepared a 3D functional MOF by using atalorganic cubic building block [Ni8L12]20 (H3L = 4,5-icarboxylic acid) bridged by alkali-metal ions (Na+)
OF exhibited stable catalytic activity for the oxidation. Fig. 14b shows the rate of CO oxidation, RQMS, obtainedole mass spectroscopy in comparison with those forand nickel oxide (NiO).et al. obtained a new In(III) MOF composed ofcontaining square-shaped channels (Fig. 14a) [112].
els are empty in In(OH)L and lled with pyridineL]xPy (L = 4,4-(hexauoroisopropylidene)bis(benzoic
microporous, thermally stable compound has beene an efcient heterogeneous catalyst for acetalizations (Fig. 14b). The difference in catalytic activity betweenwith empty or lled channels demonstrates that catal-
deed take place inside the pores.
is in MOFs doped with metal catalysts
, Eddaoudi and co-workers have demonstrated the uti-n-HImDC-based rho-ZMOF (Fig. 15a) as a host for large
active molecules, specically metalloporphyrins, andthe enhancement of catalytic activity [113]. To produceplatform, they encapsulated the free-base porphyrinreadily metallated and post-synthetically modied bysition metal ions to produce a wide range of encap-alloporphyrins. Hydrocarbon oxidation was performedsence of Mn-RTMPyP (5,10,15,20-tetrakis(1-methyl-)porphyrin tetra(p-toluenesulfonate encapsulated into assess catalytic activity for cyclohexane oxidation
ure. Reproduced with permission from Ref. [2].
-
R.J. Kuppler et al. / Coordination Chemistry Reviews 253 (2009) 30423066 3051
Fig. 14. (a) Polyhedral representation of the structure. (b) Acetalization catalyzed by [In(OH)L]0.5Py (white columns) and [In(OH)L] (black columns). Reproduced withpermission from Ref. [112].
The embmetal-organet al. (Fig[Ru(cod)(coinside the cnary resultscatalytic ap
5.4. Catalys
Hwangcoordinativterephthalawith electrothe therma[115]. Thisalize the unethylenediaing agentsremarkablytionally, palactivities duto couple al
Recentlymodicatioasp = L-aspaprotonated
enerCl inand
lectiv
gawnatio-btap3)24Thes (Fiion ons oonitrwere
subble rts an
can4C
comubic210
rve anosire de
Fig. 15. (a) Crycatalytic oxidaedding of Ru nanoparticles in the otherwise unchangedic framework MOF-5 has been investigated by Schrder
. 16) [114]. After the inclusion of Ru to formt)]3.5@MOF-5, hydrogenolysis formed Ru nanoparticlesavities and lead to the material Ru@MOF-5. Prelimi-for alcohol oxidation with Ru@MOF-5 revealed limited
plications of the water-sensitive MOF-5 host material.
is in post-synthesized MOFs
et al. found that the presence of chromium(III)ely unsaturated metal sites (CUSs) in chromium(III)te MIL-101 can provide an intrinsic chelating propertyn-rich functional groups that leads to the formation of
lly stable amine species grafted to the surface (Fig. 17)feature offers a powerful way to selectively function-saturated sites in MIL-101. It was demonstrated thatmine and diethylenetriamine can be used as new graft-to produce the amine-grafted MIL-101, which exhibithigh activities in the Knoevenagel condensation. Addi-ladium loaded APS-MIL-101 and ED-MIL-101 have highring the Heck reaction (393 K), a powerful reaction usedkenes with organic moieties.
Ingleson et al. reported the rational, post-syntheticn of a series of MOFs, M(asp)L0.5 (M = Ni2+ or Cu2+,rtate or D-aspartate, L = bipy or bpe) (Fig. 18) [116]. Theframeworks needed to conduct acid-catalyzed reactions
were gwith Hcrucial
5.5. Se
Kitacoordi([Cd(4Cd(NO[117].surfaceactivatreactiomalonester)a goodnegligireactan
TheMn3[(Mal. Theing a carea ofthat sethe cyathe mostal structure of rho-ZMOF (left) and schematic presentation of [H2TMPyP]4+ porphyrin rintion using Mn-RTMPyP as a catalyst at 65 C. Reproduced with permission from Ref. [113ated by simply treating the porous, homochiral MOFsdiethyl ether. Containment within the MOF pore wall isno homogeneous analogues are known.
e catalysis in MOFs
a and co-workers successfully synthesized a 3D porousn polymer (PCP) functionalized with amide groups,a)2(NO3)2]6H2O2DMF)n, from the reaction betweenH2O and a three connector-type amide ligand (4-btapa)amide groups are ordered uniformly on the channelg. 19a) and facilitate the selective accommodation andf guests within the channels. Knoevenagel condensationf benzaldehyde with active methylene compounds (e.g.ile, ethyl cyanoacetate, and cyano-acetic acid tert-butylconducted (Fig. 19b). While malononitrile proved to be
strate (98% conversion), the other substrates producedesults implicating a relationship between the size of thed the pore window of the host.
talytic activity of the sodalite-type compoundl)3(BTT)8(CH3OH)10]2 has been investigated by Long etpound is a thermally stable microporous solid exhibit-network of 7 and 10 pores that affords a BET surface0 m2 g1 [118]. The surface contains exposed Mn2+ ionss Lewis acids (Fig. 20). Indeed, this compound catalyzeslylation of aromatic aldehydes and ketones, as well asmanding Mukaiyama-aldol reaction. Moreover, in eachg enclosed in rho-ZMOF R-cage (right, drawn to scale). (b) Cyclohexane].
-
3052 R.J. Kuppler et al. / Coordination Chemistry Reviews 253 (2009) 30423066
Fig. 16. Model of the inclusion compound [Ru(cod)(cot)]3.5@MOF-5. Reproduced with permission from Ref. [114].
Fig. 17. Site-selective functionalization of MIL-101 with unsaturated metal sites: (A) perspective view of the mesoporous cage of MIL-101 with hexagonal windows; (B, C)evolution of coordinatively unsaturated sites from chromium trimers in mesoporous cages of MIL-101 after vacuum treatment at 423 K for 12 h; (D) surface functionalization ofthe dehydrated MIL-101 through selective grafting of amine molecules onto coordinatively unsaturated sites; (E) selective encapsulation of noble metals in the amine-graftedMIL-101 via a three-step process. Adapted from Ref. [115].
-
R.J. Kuppler et al. / Coordination Chemistry Reviews 253 (2009) 30423066 3053
Fig. 18. Left, C Right,displaying the from Rlegend, the rea
Fig. 19. (a) CryReproduced w
case, a prondimensions
5.6. Catalys
In 2000ral metal-o(referred toure metal-oThe presenprovides POalytic transPOST-1 (Figas isobutantransesterirate under o
Wu an[Cd3L4(NO31,1-binaphinto a mixdihydroxy-1room tempeu(asp) 2D layer dashed red lines emphasise JahnTeller elongated CuO bonds.1D porous channels between chiral Cu(asp) layers. Reproduced with permissionder is referred to the web version of the article.)stal structure to form another type of zigzag channels with dimensions of 3.3 3.6 . (b)ith permission from Ref. [117].
ounced size-selectivity effect consistent with the poreis observed.
is in chiral MOFs
, Seo et al. [119] reported the synthesis of a homochi-rganic porous material, [Zn3(3-O)(L)6]2H3O12H2Oas D-POST-1, L = D-tartaric acid), which uses enantiop-rganic clusters as secondary building blocks (Fig. 21a).
ce of the pyridyl groups exposed in the channels alsoST-1 with unique opportunities in catalysis. The cat-
esterication of ethanol occurred in 77% yield with. 21b). The transesterication with bulkier alcohols suchol, neopentanol and 3,3,3-triphenyl-1-propanol usingcation occurred at a much slower or even negligibletherwise identical reaction conditions (Fig. 21c).
d Lin synthesized a homochiral porous MOF,)6]7MeOH5H2O (L = (R)-6,6-dichloro-2,2-dihydroxy-
thyl-4,4-bipyridine), by slow diffusion of diethyl etherture of Cd(NO3)24H2O and (R)-6,6-dichloro-2,2-,1-binaphthyl-4,4-bipyridine in DMF/CHCl3/MeOH atrature (Fig. 22) [120]. Treatment of this compound with
excess Ti(Oof diethylzalcohols upgenerated bFor examplcatalyzed tconversion
5.7. Future
Future rmetal centethese can ework is neeheterogene
6. Magnet
6.1. Introdu
Magnetsnumber ofcross-sectional view down the a axis (solvent removed for clarity)ef. [116]. (For interpretation of the references to color in this gureKnoevenagel condensation reaction of benzaldehyde with substrates.
iPr)4 in toluene led to an active catalyst for the additioninc to aromatic aldehydes to afford chiral secondaryon hydrolytic workup. Chiral secondary alcohols werey this MOF in very high yields and enantioselectivities.
e, the addition of diethylzinc to 1-naphthaldehyde waso afford (R)-1-(1-naphthyl)-propanol with completeand 90.0% ee.
research
esearch should be dedicated to elucidate whether thers, the ligands, particle size, or some combination of
ngender MOFs with unusual catalytic properties. Moreded to determine if MOFs can compete with well-knownous industrial catalysts.
ic properties of MOFs
ction
are very important materials with an ever-increasinguses. Thus, an important goal of the research of mag-
-
3054 R.J. Kuppler et al. / Coordination Chemistry Reviews 253 (2009) 30423066
Fig. 20. A porsites exposed wis ve-coordinis 3.420(8) . R
netic materwell as explother usefuferromagnemetallic sys
between the paramagnetic metal ions or organic radicals throughdiamagnetic bridging entities. Therefore, their magnetic behaviorsdepend on the intrinsic nature of both the metal and the organicligand as well as the particular level of organization created bythe metalligand coordination interaction. As a result, in pursu-ing the magnetism of MOFs, the ligand design is crucial both toorganize the paramagnetic metal ions in a desired topology and toefciently transmit exchange interactions between the metal ionsin a controlled manner. The formation of a bulk material with non-zero spin requires a framework that allows for parallel coupling ofthe spins of neighboring paramagnetic spin carriers or antiparal-lel coupling of unequal spins. Canted spin orientations may alsoresult. It should be pointed out that there is always a tendency forantiparallel coupling of spins because the state of low-spin multi-plicity is often more stable than the state of high-spin multiplicity[122].
Magnetic studies of MOFs are embedded in the area of molecu-lar magnets and the design of low-dimensional magnetic materials,magnetic sensors, and multifunctional materials. Indeed, closed-shell organic ligands that are typically used in MOFs mostly give riseto only weak magnetic interactions. In order to achieve a strong cou-pling between the metal centers, short oxo, cyano, or azido bridges
deds aretsides pr
tic p
Fig. 21. (a) Threactions withare neepoundfall ouof MOFmagnetion of the crystal structure showing the two different types of MnII
ithin its three-dimensional pore system of 10 wide channels. Site Iate, while site II is only two-coordinate; the separation between themeproduced with permission from Ref. [118].
ials is the improvement of the properties of magnets asoring new functions, in particular in combination withl phenomena [121]. The magnetic properties such astism, antiferromagnetism, and ferrimagnetism of poly-tems derive from the cooperative exchange interactions
engineeringwith adjusterties of Msection wilinvestigatio
6.2. Magne
6.2.1. FerroFerroma
pling of theproperties ostill limited{[Cu2(pic)3by sh baThese chaingroups to gFig. 23a. Thsheets. Magence of wetted with
e hexagonal framework with large pores that is formed with the trinuclear secondaryethanol, and (c) with different alcohols. Reproduced with permission from Ref. [119].[123125]. Alternatively, polymeric metal cyanide com-frequently encountered in magnetic investigations butthe scope of this review. Furthermore, the porosity
ovides additional interesting phenomena in regards toroperties. The use of chemical coordination or crystal
techniques allows for the systematic design of MOFsable magnetic properties. Because the magnetic prop-OFs have been recently reviewed [122,126128], thisl give only a brief enumeration of recent results in thens of magnetic MOFs.
tic properties of MOFs
magnetic propertiesgnetism requires a structure that allows for parallel cou-spins. There are numerous reports about the magneticf MOFs; however, literature on ferromagnetic MOFs is
. Biswas et al. [129] reported a magnetic MOF material,(H2O)]ClO4}n (pic = 2-picolinate), which is constructedckbone chains through syn-anti carboxylate groups.s are linked to one another by syn-anti carboxylate
ive rise to a rectangular grid-like 2D net as shown ine ClO4 anions are located between these 2D cationicnetic susceptibility measurements revealed the pres-
ak ferromagnetic coupling for this MOF that has beena suitable model (Fig. 23b).
building units. (b) Catalytic activity of POST-1 in transesterication
-
R.J. Kuppler et al. / Coordination Chemistry Reviews 253 (2009) 30423066 3055
Fig. 22. (a) The 2D square grid in the crystal structure and (b) space-lling model as viewed down the c axis showing the chiral 1D channels of 13.5 13.5 in dimensions.Reproduced with permission from Ref. [120].
Fig. 23. (a) 2Dtemperature (
Anothermetal compAwaga [130in zigzag chby two shoand the O a(Fig. 24a). Tcoupling asTTTA and adue to the w
Fig. 24. (a) A vdependence olayer [1 1 0] of the CuII-picolinate-bridged structure in {[Cu2(pic)3(H2O)]ClO4}n . (b) ThermT) for {[Cu2(pic)3(H2O)]ClO4}n . Reproduced with permission from Ref. [129].
interesting example is TTTACu(hfac)2, an adduct of alex and an organic free radical, reported by Fujita and
]. In this structure TTTA ligands bridge Cu2+ ions to resultains along the b axis, which are assembled together
rt contacts between the S atom on the dithiazolyl ringtoms of Cu(hfac)2 to form a 2D supramolecular layerhe magnetic measurements showed a ferromagnetic
cribed to the coordination bond between Cu(hfac)2 andweak antiferromagnetic coupling at low temperatureeak interaction between Cu(hfac)2 and TTTA and/or the
interchain ipreted in teantiferroma
6.2.2. AntifMany M
antiparallelradical entiwork, Mn2(pillared by
iew of the structure of TTTACu(hfac)2. The CF3 groups of Cu(hfac)2 are omitted for clarity. Tf pT for TTTACu(hfac)2. Reproduced with permission from Ref. [130].mal variation of the product of the molar magnetic susceptibility and
nteractions. Thus, the magnetic properties were inter-rms of a ferromagnetic dimer with a weak inter-dimergnetic interaction.
erromagnetic propertiesOFs have antiferromagnetic properties due to thecoupling of the spin metal ions or the organic free
ties. Recently, Jia et al. [131] reported a 3D Mn2+ frame-tzc)2(bpea), in which 2D layers with 3-tzc bridges arethe bpea spacers to generate a 3D structure with a 3,4-
he interchain contacts are shown by the broken lines. (b) Temperature
-
3056 R.J. Kuppler et al. / Coordination Chemistry Reviews 253 (2009) 30423066
Fig. 25. The e ds, (b)self-catenation
connected sThe magnetbetween M
Yu etior of aCo3(TATB)2of a trimetthe clusterMOF. In theclusters anda propelle
achaned th
Fig. 26. (a) 3DReproduced wxtended structure of Mn2(tzc)2(bpea). (a) A 2D layer with tzc as 3-bridging liganhighlighted by bold lines. Reproduced with permission from Ref. [131].
elf-penetrating topology of (482 103)(482) (Fig. 25).ic measurements revealed antiferromagnetic coupling
n2+ ions in this framework.
to formchiralindicatal. [132] also characterized the magnetic behav-non-interpenetrating chiral porous cobalt MOF,
(H2O)22DMA3H2O. The framework is composedallic hourglass cluster linked by organic ligands withbeing responsible for the magnetic properties of theframework, every ligand pair is bound to three Co2+
every cluster connects with three ligand pairs to formr. The clusters are extended by the TATB ligand pairs
measured tcoupling ex
6.2.3. FerrimIn MOFs
typically leresulting inally, the fer
framework structure of Co3(TATB)2(H2O)2. (b) Susceptibility as function of temperatuith permission from Ref. [132].the 2D topology, (c) the 3D structure, and (d) the 3D topology with
non-interpenetrated (10, 3)-a framework containingnels (Fig. 26a). Magnetic susceptibility measurementsat this MOF exhibited paramagnetic properties within
emperature range 2300 K, though antiferromagneticisted between neighboring Co2+ atoms (Fig. 26b).
agnetic properties, the exchange coupling among paramagnetic centersads to ferromagnetic or antiferromagnetic behavior,
very few instances of ferromagnetic behavior. Usu-rimagnetic systems contain two kinds of spin carriers
re (inset shows the 1/ vs. T curve along with a theoretical tting).
-
R.J. Kuppler et al. / Coordination Chemistry Reviews 253 (2009) 30423066 3057
Fig. 27. (a) 3D framework structure of Cu3(L)2(VO3)4. (b) mT vs. T plot at different elds and the zero eld cooled (ZFC) and eld cooled (FC) susceptibility plot for Cu3(L)2(VO3)4.Reproduced with permission from Ref. [133].
that alternate regularly and interact antiferromagnetically [121]. Incontrast, homospin ferrimagnetic systems contain only one typeof spin carrier that, in spite of the antiferromagnetic coupling,leads to a non-cancellation of spins as a result of a partic-ular topological arrangement of the spins. Recently, Li et al.[133] reported a 3D homospin ferrimagnet, Cu3(L)2(VO3)4 (L= 5-(pyrimidin-2-yl)tetrazolate) constructed from (VO3)n chains link-ing ([5-(pyrimidin-2-yl)tetrazolate-(Cu2+)1.5]2+)n layers (Fig. 27a).In the structure, there exist two crystallographically independentCu2+ centers that are responsible for the unusual magnetic property.Magnetic measurements revealed this complex to be a ferrimagnetwith TN = 10 K (Fig. 27b). This framework can be seen as two latticeswith an antiferromagnetic coupling between them that results inan uncompensated spin S = 1/2 per every three Cu2+ atoms. At lowtemperatures the uncompensated spins align parallel to each other,resulting in the ferromagnetic long-range order.
6.2.4. FrustFrustrate
because th
press or signicantly reduce long-range magnetic ordering andmay lead to unusual ground-state behaviors. The magnetic frustra-tions are usually observed in triangular or tetrahedral plaquettemagnetic lattices [134136]. Gao et al. [137] recently reported aMn2+ complex, Mn[Mn3(3-F)(bta)3(H2O)6]2 which has a 2D frus-trated lattice with alternating triangular motifs and mononuclearcenters. Magnetic studies showed this compound to behave as alow-temperature antiferromagnet (Fig. 28). The stoichiometry andconnecting mode of the mono- and tri-nuclear Mn2+ moieties in the2D lattice led to an uncompensated spin moment, which caused thefrustration.
In contrast to ferrimagnetism, residual spin resulting from per-turbation of antiparallel or parallel coupling may induce spincanting. Asymmetric exchange interactions and single-ion mag-netic anisotropy are the origins of this magnetic behavior [138].Several MOFs have presented spin canted antiferromagnetic prop-
Li etzoled (utcrys
Fig. 28. Left: v3,6-connected(b) eld-depenReproduced wration and cantingd magnetic materials have attracted much attention
e competing interactions in these materials can sup-
erties.thiadia(10,3)-in theiews of the structure of Mn[Mn3(3-F)(bta)3(H2O)6]2, showing (a) coordination environet of the layer. Right: (a) temperature dependence of the susceptibility of Mn[Mn3(3dent magnetization, (c) schematic illustration of the 2D lattice, (d) a segment of the la
ith permission from Ref. [137].al. [139] reported a CoII MOF, Co2(L)2(H2O)2 (L= 2,1,3--4,5-dicarboxylate), which has a non-interpenetratedp) topological network structure. It is interesting thattallographic asymmetric unit there exist two uniquenments of the metal ion and the ligand, (b) the 2D layer, and (c) the-F)(bta)3(H2O)6]2. Inset: FC (20 and 1000 Oe) and ZFC (20 Oe) plots,ttice illustrating the frustration and the triangular sublattice of Mn.
-
3058 R.J. Kuppler et al. / Coordination Chemistry Reviews 253 (2009) 30423066
Fig. 29. Left: view of the 3D framework structure of Co2(L)2(H2O)2 (un-coordinated carboxylate O atoms and water molecules were omitted for clarity). Right: (a) mT vs. Tplot for Co2(L)2(H2O)2 (insert: plot of the reduced magnetization at 2 K), (b) hysteresis loop, (c) mT vs. T plots at different elds (from 1000 to 50 G) in the low-temperatureregion, (d) plot of the ac susceptibility. Reproduced with permission from Ref. [139].
but chemically similar Co2+ ions, which were linked by ligandsto form a 3D framework (Fig. 29, left). These two crystallograph-ically different CoII ions were determined to be responsible forthe magnetic behaviors. Magnetic measurements showed that thiscomplex takes on unexpected long-range magnetic ordering asshown in Fig. 29, right. Due to the shape of the mT curve, ferromag-netic coupling between the Co2+ ions does not seem likely at hightemperatures. There are different magnetic pathways Co1Co2,Co1Co1, and Co2Co2 that have to be antiferromagnetic unlessthe syn-anti carboxylate coordinate mode is ferromagnetic yetvery small in magnitude. Thus, the origin of this magnetic ordercan be attributed to a weak magnetic ordering, the so-calledcanting.
6.3. Spin-crossover and induced magnetic change in MOFs
Among efforts to introduce specic functions to MOFs, theincorporation of electronic switching centers is very importantin the development of new advanced functional materials formolecular-scale switching devices. Spin-crossover centers repre-sent a convenient, addressable molecular-scale magnetic switchin which d4d7 transition metals can change between the high-spin and low-spin state in response to variation in temperature,pressure, or light irradiation [140]. Recently, Halder et al. [141]reported a nanoporous MOF, Fe2(azpy)4(NCS)4(guest) which hasan interpenetrated 2D layer to 3D framework structure contain-ing electronic switching Fe2+ centers and displays reversible uptake
Fig. 30. (a) Thsample at diffspin-crossovercrossover withRef. [141].e structures of Fe2(azpy)4(NCS)4(EtOH) at 150 and 375 K. (b) The temperature-dependenerent stages of guest desorption and resorption, showing 50% spin-crossover behavior b
for the fully desorbed phase. The ethanol and methanol-loaded phases undergo a single sa plateau at 120 K. The inset shows the effect of partial and complete removal of methant magnetic moment of Fe2(azpy)4(NCS)4(guest), recorded on a singleetween 50 and 150 K for the fully loaded phases and an absence oftep spin-crossover, whereas the 1-propanol adduct shows a two-step
ol from Fe2(azpy)4(NCS)4(MeOH). Reproduced with permission from
-
R.J. Kuppler et al. / Coordination Chemistry Reviews 253 (2009) 30423066 3059
Fig. 31. (a) Structure of Cu3(PTMTC)2(py)6. (b) Reversible magnetic behavior of the amorphous and evacuated phase in contact with ethanol liquid, as observed by plottingT as a functio duced
and releaseinterpenetrin turn causcenters. Aswithin thethe sorbedorbed phashas shown tfunctionalitnatures to ptomagneticFe(NCS)2(bp[142].
Solvent-in a nanopowhich wasical (PTMTCThis MOF hcomposed oweak awork strucinterestingtive solventvolume chapropertiescould be us
ine
trodu
potenterebini
s, as w
mine
uminthan
dueties.thesoratiy [Baine-se apniceparasingn of temperature T at a eld of 1000 Oe. Inset: similar behavior at 10,000 Oe. Repro
of guest molecules (Fig. 30a). The exibility from theated framework allows guest uptake and release, whiches substantial changes in the local geometry at the Fe2+
a result, the presence of Fe2+ spin-crossover centersframework lattice led to the switching of this material:phases undergoing half-spin crossovers, and the des-e showing no switching property (Fig. 30b). This workhat MOF materials can incorporate an additional level ofy, such as spin-transition associated with their porousroduce new, useful properties. Furthermore, the pho-properties of a similar porous Fe2+ spin-crossover MOF,ed)23EtOH has been investigated by the same group
induced magnetic properties have also been observedrous MOF structure, Cu3(PTMTC)2(py)6(EtOH)2(H2O),constructed by using a persistent organic free rad-) functionalized with three carboxylic groups [143].as a 2D layer structure extending along the ab planef a honeycomb structure. These layers stack by usingnd van der Waals interactions to form an open frame-
ture with hexagonal pores (Fig. 31a). It is even morethat this material shows a reversible and highly selec--induced shrinkingbreathing process involving large
7. Lum
7.1. In
Themuch iby comligand
7.2. Lu
7.2.1. LLan
thesesproperto synincorpnamelbipyridstepwithe sigthe prdecreanges (2535%) that strongly inuence its magnetic(Fig. 31b). Thus, this magnetic sponge-like behaviored as a new route towards magnetic solvent sensors.
work was pAdditionallthe develop
Fig. 32. Lanthanide metal complex and structure of [Ba2(H2O)4[LnL3(H2O)2](H2O)nCwith permission from Ref. [143].
scence and sensors in MOFs
ction
ntial use of MOFs as luminescent materials has spurredst in the area [144,145]. These materials can be prepared
ng the luminescent metal ions or clusters and organicell as special guest molecules [146].
scence in MOFs
escence in MOFs based on metal centerside metal ions have been widely used in MOF syn-
to their coordination diversity and luminescentChandler et al. [147] reported a stepwise approach
izing a MOF with photophysical properties throughon of lanthanide metal complexes in the framework,2(H2O)4[LnL3(H2O)2](H2O)nCl] (L = 4,4-disulfo-2,2-
N,N-dioxide, Ln = Sm, Eu, Gd, Tb, Dy) (Fig. 32). Aproach was used and allowed for an examination of
ance of the ratio of metal to organic building units intion of a MOF containing lanthanide metal ions. Bythe ligand to metal ratio, a less dense, porous frame-
repared that maintained its luminescent characteristics.y, metalloligands were referred to as building units forment of lanthanide containing MOFs for use as sensing
l] . Reproduced with permission from Ref. [147].
-
3060 R.J. Kuppler et al. / Coordination Chemistry Reviews 253 (2009) 30423066
Fig. 33. Structure and luminescence properties of Zn3L3(DMF)2 (left) and Zn4OL3 (right). Reproduced with permission from Ref. [150].
devices [148]. Proper selection of the linker is necessary to ade-quately shield the metal to prevent quenching of its luminescentproperties.
de Lill etnescent prothe mixed swith an incrhas been atlinkers and
7.2.2. LumiTwo lum
on trans-4,4two differeobtained inresulted frodemonstratIn both cascoordinatiolifetimes fostilbene.
Studies hers, such asa ligand, hslightly moleading to t
7.2.3. LumiHuang
[Cd3L6](BF4in which thand reintro
work (Fig. 34a). Heating in air at different temperatures (180, 200,225, 250 C) for 1 day generated a series of dehydrated products.One of the dehydrated complexes was rehydrated when exposed
vape comed/renteraes an UVole
tifun
nsor
MOr shas [14
ensoentlyb(BTpropthan) in m, Cl
sultsorateincohe pexhi
oriden et[Eu(
Fig. 34. (a) 3Dand yellow); (those partiallyal. [149] reported in 2007 the synthesis and photolumi-perties of a Eu MOF and a Eu/Tb mixed system MOF. Inystem, both Eu and Tb emissions were observed alongease in Eu emission intensity relative to the Eu MOF. Thistributed to the Eu being sensitized by both the organicthe Tb.
nescence in MOFs based on organic ligandsinescent stilbene-based MOFs were prepared based-stilbene dicarboxylic acid (LH2) and zinc nitrate in
nt solvents. A 2D network structure Zn3L3(DMF)2 wasDMF, while a 3D porous framework structure Zn4OL3m DEF (Fig. 33) [150]. The optical properties of bothe that the LH2 organic ligand serves as the chromophore.es, the rigidity of the stilbene ligand increases uponn to the metal center, resulting in increased emissionr the MOF crystals as compared to solutions of trans-
ave also been conducted with luminescent organic link-H2hpbb as reported by Gndara et al. in 2007 [151]. Aspbb emits a blue-white color under UV light, which is
died with the addition of different lanthanide metals,he possible uses of these MOFs as diodes.
nescence based on guest molecules in MOFset al. [152] reported a 3D porous MOF,
)2(SiF6)(OH)213.5H2O (L = 2,6-di(4-triazolyl)pyridine),e guest species in the open channels can be removed
duced reversibly without destroying the porous frame-
to H2Oof thesremovweak iprovidbetweeguest mof mul
7.3. Se
Thesize- odevice
7.3.1. SRec
MOF, TcenceG = meTb(BTC(X = F
The reincorpthe F
lying tanionsthe u
CheMOFs,framework of [Cd3L6](BF4)2(SiF6)(OH)213.5H2O with the guest water molecules and anib) Luminescence spectra of complexes with increasing dehydration, in the solid state atrehydrated, in the solid state at RT. Reproduced with permission from Ref. [152].or for 1 or 2 days. The solid-state luminescence spectraplexes (Fig. 34b and c) revealed that guest molecules
hydrated from the MOF undoubtedly inuence thections between a ligand and a metal center. The workconvenient and effective route for tuning emissionsand visible wavelengths by controlling the number of
cules and may be useful for the design and fabricationctional luminescent materials.
s in MOFs
Fs that possess luminescent properties together withpe-selective sorption properties can be used as sensing6].
rs for selective ion monitoring, Chen et al. [153] reported a prototype luminescentC)G (MOF-76, G = guest solvent) (Fig. 35a). The lumines-erties of the anion incorporated Tb(BTC)G (MOF-76b,ol) microcrystalline solids were studied by immersing
ethanolic solutions of different concentrations of NaX, and Br) and Na2X (X = CO32 and SO42) (Fig. 35b).
show that the luminescence intensity of the aniond MOF-76b is signicantly increased, particularly forrporated MOF (Fig. 35c). The special properties under-otential of MOF-76 for the recognition and sensing ofbit a high-sensitivity sensing function with respect toanion.
al. [154] recently reported the synthesis of luminescentpdc)1.5(dmf)](DMF)0.5(H2O)0.5 (1) (pdc = pyridine-3,5-ons in the cavities (SiF62 , polyhedron, blue; BF4 , space lling, greenroom temperature; (c) Luminescence spectra of complexes including
-
R.J. Kuppler et al. / Coordination Chemistry Reviews 253 (2009) 30423066 3061
Fig. 35. (a) Crystal X-ray structure of MOF-76b containing NaF with the model of uoride (green) at the center of the channel; (b) 5D4 7F5 transition intensities of MOF-76bactivated in different types of 102 M NaX and Na2X methanol solution (excited and monitored at 353 and 548 nm, respectively); (c) Excitation (dotted) and PL spectra (solid)of MOF-76b solid activated in different concentrations of NaF methanol solution (excited and monitored at 353 and 548 nm, respectively). Reproduced with permission fromRef. [153].
Fig. 36. (a) Cr ardsactivated in DM and 61b incorporati m Ref
dicarboxyla(Fig. 36a). Tied as the pintroducedidentity of tnescence cametals haviions, such a
7.3.2. Sensomolecules
Chen etEu(BTC) (wethanol, aceexhibit diffecence inten
nsingscen
ng ul
Fig. 37. (a) Str(b) DMF and (referred to theystal X-ray structure of 1 with immobilized Lewis basic pyridyl sites oriented towF solutions of Cu(NO3)2 at different concentrations (excited and monitored at 321
ng different metal ions, activated in 10 mm DMF solutions of M(NO3)x . Adapted fro
te), with Lewis basic pyridyl sites for metal ion sensinghe luminescence characteristics of this MOF were stud-yridyl sites were coordinated to additional metal ions
and selumine
Usi
in a DMF solution (Fig. 36b). It was found that thehe additional metal ion is very signicant to the lumi-pability of the complex, with alkali and alkaline earthng little effect on the luminescence while other metals Cu2+, cause signicant quenching (Fig. 36c).
rs for the presence and/or types of guest/solvent
al. [155] reported a rare earth microporous MOF ofith open Eu3+ metal sites) (Fig. 38a). Within the MOF,tone, dimethyl formamide, and other small moleculesrent enhancing and quenching effects on the lumines-
sity (Fig. 37b and c). This MOF is suitable for the binding
tals of a uoand quantittonitrile solRemarkableing) were oacetonitrilesuggests ahighly sens
7.3.3. SensoBesides
be used foAllendorf e
ucture of Eu(BTC) MOF viewed along the c axis, and accessible Eu3+ sites are shown by tc) acetone solvent, respectively (excited at 285 nm). Adapted from Ref. [155]. (For interpweb version of the article.)pore centers; (b) The excitation () and PL spectra (- - -) of solid 1b18 nm, respectively); (c) Comparison of the luminescence intensity of. [154].
of small molecules by using the specic properties oft open Eu3+ sites.trasonic methods, Qiu et al. [156] obtained nanocrys-
rescent microporous MOF, Zn3(BTC)212H2O (Fig. 38a),atively analyzed the signals of organoamines in an ace-ution using uorescence spectrophotometric titrations.
changes of emission intensity (uorescence quench-bserved when the volume of ethylamine added intowas changed (Fig. 38b). This uorescence quenching
high sensitivity to ethylamine and may lead to someitive sensors for organoamines.
rs for stress-induced chemical detectionsensors for anions and guest molecules, MOFs can alsor stress-induced chemical detection. Recent work byt al. [157] demonstrated that the energy of molecular
he pink arrows. The PL spectra in the presence of various content ofretation of the references to color in this gure legend, the reader is
-
3062 R.J. Kuppler et al. / Coordination Chemistry Reviews 253 (2009) 30423066
Fig. 38. (a) Molecular-packing diagram of Zn3(BTC)212H2O viewed along the c axis; (b) Variation of emission spectra with the volume of ethylamine in 2 mL of acetonitrile(excited at 327 nm). Reproduced with permission from Ref. [156].
adsorption can be converted to mechanical energy to create a highlyresponsive, reversible, and selective sensor by integrating a thinlm of MOFsensor respbut yields nthe use of sthe structur
7.3.4. SensoHarbuza