pubs.acs.org/crystal Published on Web 10/23/2009 r 2009 American Chemical Society

DOI: 10.1021/cg900985w

2009, Vol. 95024–5034

Crystallization Behavior of Coordination Polymers. 1. Kinetic and

Thermodynamic Features of 1,3-Bis(4-pyridyl)propane/MCl2 Systems

Lucia Carlucci,† Gianfranco Ciani,† Juan Manuel Garcıa-Ruiz,§ Massimo Moret,*,‡

Davide M. Proserpio,† and Silvia Rizzato*,†,#

†Dipartimento di Chimica Strutturale e Stereochimica Inorganica and #Facolt�a di Farmacia,Universit�a degli Studi di Milano, Via G. Venezian 21, I-20133 Milano, Italy, §Laboratorio de EstudiosCristalogr�aficos, IACT (CSIC-Universidad de Granada), Avda. del Conocimiento s/n, P.T. Ciencias dela Salud, 18100 Armilla, Spain, and ‡Dipartimento di Scienza dei Materiali, Universit�a degli Studi diMilano Bicocca, Via R. Cozzi 53, I-20125 Milano, Italy

Received August 18, 2009; Revised Manuscript Received October 1, 2009

ABSTRACT: A series of one-, two-, and three-dimensional (1D, 2D, and 3D) coordination polymers has been crystallized fromsolution as well as in gelled media, by controlled reaction of the flexible ligand 1,3-bis(4-pyridyl)propane withMCl2 salts (M=Mn, Fe, Co, Ni, Cu, Cd, Zn). Attention was paid to the role of kinetic and thermodynamic control upon selection of the finalproducts that can be obtained with different optimized crystallization techniques. It is shown that the correct use ofcrystallization techniques allows tuning of the metal to ligand ratio and absolute concentrations and, hence, control of thecrystallization of metal-organic frameworks, including their solvent-mediated phase transitions.

Introduction

Recent years have witnessed a surge of interest in supra-molecular chemistry,1 and, in particular, there is hope forfuture technological applications2 of coordination polymers,a specific class of supramolecular objects also knownasmetal-organic frameworks (hereafterMOFs).3,4 The intriguing vari-ety of architectures and topologies5 exhibited by MOFs areexplored in crystal engineering,6 as they can reveal interestingproperties for opto-electronic7 and magnetic8 functional ma-terials, molecular sieving,9 ion exchange,10 catalysis11-13 aswell as stable and highly porous scaffolds14 for sorption ofvolatile organic compounds and gas storage.15-18Attempts tobuild novel and useful coordination networks rely uponalmost all metal cations available in the periodic table ofelements, that is, transition, post-transition, lanthanide,19-21

and actinide21 metals. The counterpart of ligands shows aneven wider variety of choices being based on multifunctionalorganic or inorganic moieties, displaying diversified localstereochemistries supposedly able to force the self-assemblingof MOFs to follow a pre-established structural motif aswished by crystal engineers.22-24 Many structurally relevantMOFs which do have promising properties for key technolo-gical applications and high commercial value25 have beensynthesized starting from nitrogen-based multifunctionalligands.3,16c,d,26 Among them, pyridine-based ligands differ-ing in length, flexibility, and conformational freedom pro-duced a variety of supramolecular networks with differentmetal cations.12

While the structural characterization of MOFs by X-raydiffraction is well established, there is no comparable under-standing of their solution chemistry, under ambient or solvo-thermal conditions.27,28 Particularly unclear are themolecularrecognition and assembling steps which eventually lead to acoordination polymer. This failure partly stems from thecomplexity of chemical equilibria involving these polymers

which fragment upon solubilization, and whose high mole-cular weight frameworks arise frommostly unknown solutionspecies,29 apart froma fewnoteworthyoligomeric entities.30,31

Thus, no accurate model for the network formation fromsolution is presently available; chemical synthesis usually isperformed following a buy-and-mix strategy. However, it isalso largely acknowledged that the final outcome of reactantsmixing canbe driven by subtle kinetical and thermodynamicaldetails of the chemical processes. As several authors noticed,supramolecular inorganic coordination frameworks representa dynamic challenge for structural chemists.32,33 Starting fromthe mother solutions, the overall supramolecular synthesis ofMOFs includes molecular recognition and self-assembly ofsolution species through nucleation of crystalline phases,growth of mature crystals, and possibly phase transitionsaccording to the relative kinetic prevalence and thermody-namic stability. Therefore, solution coordination chemistry isdeeply entangled with nucleation and growth mechanisms ofMOFs crystals to determine the final products and theirproperties, through a complicated interplay between kineticsand thermodynamics.

We have previously synthesized coordination polymerswith reversible uptake/release of guest solvent molecules, forexample, compounds [Cu5(bpp)8(SO4)4(EtOH)(H2O)5](SO4) 3EtOH 3 25.5H2O

34 (bpp=1,3-bis(4-pyridyl)propane; EtOH=ethanol) and [Cu(bipy)2(CF3SO3)2] 3 2CH2Cl2 3H2O,

35 (bipy=4,40-bipyridyl); both exhibit interestingmicroporous behaviorarising fromnovel 3Darchitectures associatedwith cavities ortunnels, respectively, fromwhich guestmolecules canbe easilyand reversibly removed due to some degree of flexibility36 ofthe metal-organic framework. Moreover, the self-assemblyof polymeric networks from different Ag(I) salts and theflexible ligand 1,3-bis(4-pyridyl)propane (bpp) has been sys-tematically investigated in order to obtain basic informationuseful for the crystal engineering of coordination frames uponvariation of the counterions.37

We have also studied the crystallization of inorganic poly-mers based on the flexible nitrogen ligand bpp and metal(II)chlorides, trying to improve understanding of the self-assembly

*Corresponding authors. (M.M.) Phone: þ39 0264485218; fax: þ390264485400; e-mail: massimo.moret@mater.unimib.it; (S.R.) phone:þ390250314442; fax: þ39 0250314454; e-mail: silvia.rizzato@unimi.it.

Article Crystal Growth & Design, Vol. 9, No. 12, 2009 5025

processes of coordination networks. Reaction of aqueousCuCl2 with bpp ligand under different conditions afforded afamily of coordination polymers with one-, two-, and three-dimensional (1D, 2D, and 3D) polymeric network dimension-alities. A preliminary account on these systems has alreadybeen published38mainly dealingwith the structural features oftheCu(II)/bpp systems but also gathering preliminary data onhow different metal cations assemble with bpp to buildcoordination networks.

The present study aims at a better characterization of theformation of coordination polymers, that is, crystal nuclea-tion, crystal growth, and kinetical or thermodynamical selec-tion of the products, factors which for MOFs are intimatelyconnected with self-assembly and building of the polymericspecies. We choose a simple but flexible ligand such as 1,3-bis(4-pyridyl)propane reacted with the chlorides of a series ofdivalent cations (Mn2þ, Fe2þ, Co2þ, Ni2þ, Cu2þ, Cd2þ, andZn2þ) aiming at forcing the chloride anion in the innercoordination sphere to obtain neutral polymers, thus remov-ing the role of the counterion as a possible templating agentduring crystallization ofMOFs.37 The present paper expandsthe case study of the CuCl2/bpp system38 and improvesdescription of the subtle interplay between experimental con-ditions under whichMOFs based on theMCl2/bpp couple aresynthesized, including effects of concentration, metal/ligandratio, nature of solvent, and kinetic or thermodynamic com-petition between different solid phases.

Experimental Procedures

Materials and Methods. All reagents (metal salts: CdCl2 3 2H2O,CuCl2 3 2H2O,CoCl2 3 6H2O,FeCl2 3 4H2O,MnCl2 3 4H2O,NiCl2 3 6H2O,ZnCl2; ligand 1,3-bis(4-pyridyl)propane) and solvents were com-mercially available high-purity materials (from Aldrich Chemicalsand Pharmacia LKB Biotechnology), and used as supplied withoutfurther purification apart from the bpp ligandwhichwas purified bysublimation. Deionized Milli-Q water was used for electroconduc-tance measurements using a Hanna HI255 conductimeter with avessel thermostatted at 25 �C ( 0.05 �C. The Job plot absorbancedata points were collected with a Hewlett-Packard 8453 UV/visspectrophotometer by using CuCl2 and bpp solutions both 10.7 or20.0 mM. Optical micrographs were taken with a polarizing SZX12Olympus stereomicroscope equipped with a digital camera forsingle shot or time sequence images acquisition. X-ray powderdiffraction spectra were collected on a Philips PW 3050 vertical-scan diffractometer. For the chemical synthesis, no crystal yieldsare reported due to unavoidable difficulties during crystal recoveryfrom gels and the very long times required to reach chemicalequilibrium. A selection of yields is reported in Table 2S inSupporting Information for precipitation of microcrystalline pow-ders at high concentrations.

Synthesis of [M(bpp)3Cl2] 3 2H2O (M=Cd (1Cd), Co (1Co), Fe(1Fe), Ni (1Ni)). Compounds 1Cd and 1Co were crystallized by theslow evaporation method from a mixture of reagents using water/acetone for 1Cd (CdCl2 3 2H2O: 28.3 mg, 0.129 mmol in 10 mL ofwater added to bpp: 51.5 mg, 0.260 mmol in 10 mL of acetone) andwater for 1Co (CoCl2 3 6H2O: 32.4 mg, 0.136 mmol dissolved in10mL of water added to bpp: 53.4 mg, 0.269mmol in 10mL of water)as solvent. In the case of 1Cd a white precipitate formed immedi-ately. Both reaction mixtures were filtered and left to evaporate atambient temperature until crystals were obtained, colorless for 1Cdand pink for 1Co.

1Fe has been prepared by the liquid-liquid diffusion techniqueusing aqueous solutions of reagents and a 1:2 molar ratio(FeCl2 3 4H2O: 15.6 mg, 0.078 mmol dissolved in 3.0 mL of waterlayered on bpp: 31.2 mg, 0.157 mmol dissolved in 3.0 mL of water).Because of the oxidation of Fe(II) a brownpowder precipitates fromthe solution; however, in a few days formation of aggregates ofyellow crystals of 1Fe has been observed.

1Ni: 1.00 g of solid PEO (poly (ethylene oxide), average MW∼ 2000000) was added to 20mLof a 0.060Msolution of bpp inwater.After a few minutes of stirring, a stable suspension was obtained.Part of this solution was poured into a test tube (10 mm i.d. and115 mm long) approximately to half height. In the same way 20 mLof a 0.060M solution of NiCl2 3 6H2O was gelled with 0.50 g of PEOand then poured above the bpp suspension until the top of the tubewas reached. The tube was closed and the mixture was left to standat room temperature to crystallize. Crystals appeared within a fewdays as clusters of green plates.

Synthesis of [M(bpp)2Cl2] 3 solv (M=Cd (3Cd), Co (3Co), Fe (3Fe),Mn (3Mn), Ni (3Ni) solv = MeOH, EtOH, H2O). 3Cd: crystal-lization was accomplished using single gel diffusion technique. Thecrystals were grown in a single glass tube of length 115 mm anddiameter 10 mm. A 1.0% (w/v) agarose gel was obtained by addingagarose powder (Tg=42 �C) to 4.5 mL of an aqueous solution ofCdCl2 3 2H2O (32.0 mg, 0.146 mmol) under continuous stirring andthen heating the mixture in a water bath at a temperature above theagarose melting point until a clear solution was obtained. The hotsol was then poured in the test tube and cooled. After setting of thegel, an ethanolic solution of bpp (19.1 mg, 0.406 mmol) was care-fully layered over it. In a fewweeks, white well-formed bymiramidalcrystals of 3Cd were formed at the gel-solution interface and in thesupernatant ethanolic solution.

3Co: 0.75 g of a bead-formed dextran gel, Sephadex LH-20(Pharmacia LKB Biotechnology), was added to 5 mL of aCoCl2 3 6H2O methanolic solution (34.7 mg, 0.146 mmol). The gelwas allowed to swell for several hours into a test tube (10 mm i.d.and 115 mm long). The excess of solution was removed beforelayering on top of the packed beds an equal volumeof a bpp solutionat requiredmolarity (0.030M inmethanol). The tubewas closed andallowed to stand at room temperature. Pink well-formed crystals,with characteristic bipyramidal shape, grew in a few days togetherwith blue crystals of species 6Co both within the body of the gel andin the supernatant solution.

3Mn and 3Fe: Both compounds were crystallized by carefullylayering 3 mL of a methanolic solution of the metal salt (MnCl2 34H2O: 42.3 mg, 0.214 mmol; FeCl2 3 4H2O: 9.1 mg, 0.0458 mmol)over a solution of bpp in methanol (82.1 mg, 0.414 mmol, 3 mL) for3Mn and dichloromethane (17.6 g, 0.0887 mmol, 3 mL) for 3Fe atroom temperature. In both cases, the mixture was left to stand atambient temperature for one day to yield typical bipyramidalcrystals (colorless for 3Mn and yellow for 3Fe).

3Ni: solid PEO (0.40 g, average MW ∼ 2 000 000) was dissolvedby stirring in 20 mL of a dichloromethane solution containingbpp (80.6 mg, 0.406 mmol) to give a gel-like high viscosity fluid.Approximately 3 mL of the gel material, characterized by the pre-sence of gas bubbles, were transferred into a test tube (10 mm i.d.and 115 mm long). The bubbles were removed by letting the samplestand for several hours. Then, an equal volume of NiCl2 3 6H2O(11.7 mg, 0.0492 mmol) methanolic solution was layered above thegel. The tube was closed and the mixture was left to stand at roomtemperature to crystallize. Many small green crystals, with a tetra-gonal bipyramidal shape, were formed at the gel-solution interfacein a few days.

Synthesis of 5Cu.This compoundwasobtained as a pure crystallinepowder by a counter-diffusion method using isopropanol as solvent,suitable concentrations of reactants, and a 1:1 metal-ligand molarratio (bpp solution: 24.6 mg, 0.124 mmol in 3.0 mL, layered over aCuCl2 3 2H2O solution: 20.9 mg, 0.123 mmol in 3.0 mL)

Synthesis of [Co(bpp)Cl2] (6Co). Single crystals of 6Co wereobtained by using a modification of the gel-acupuncture method.39

A 1.0% (w/v) agarose gel was prepared and poured into a test tube(13 mm i.d. and 160 mm long) until a layer of 10 mm in height wasobtained. A 100 mm long glass tube, open at both ends, with a2.0 mm inner diameter, was immediately inserted into the biggertube and fixed with one side of the tube into the sol-gel. Aftercompletion of gelling of the agarose sol, 1.5 mL of a methanolicsolution of bpp (11.9 mg, 0.060 mmol) and 6.0 mL of a solution ofCoCl2 3 6H2O (47.6 mg, 0.200 mmol) in the same solvent werepoured respectively in the inner and in the outer tube, and storedat room temperature. Well-formed prismatic blue crystals wereformed in the bpp solution in a few weeks.

5026 Crystal Growth & Design, Vol. 9, No. 12, 2009 Carlucci et al.

Synthesis of [Zn(bpp)Cl2] (6Zn). Crystals of polymeric Zn-bppwere grownby liquid-liquid diffusionmethod using bothwater andmethanolic solutions of reagents and 1:2 or 1:1 metal-ligand molarratio (for example: ZnCl2 3.4 mg, 0.025 mmol, 2.5 mL of H2O/bpp5.0 mg, 0.025 mmol, 2.5 mL of H2O; ZnCl2 3.6 mg, 0.026 mmol,3.0 mL of H2O/bpp 5.1 mg, 0.026 mmol, 3 mL of MeOH).

Synthesis of [Cu(bpp)1.5Cl2] 3 (THF) 3 0.75(H2O) (7Cu). Crystal-lization of compound 7Cu was accomplished by layering on a solu-tion of bpp in tetrahydrofuran (THF) (4.6 mg, 0.030 mmol, 4 mL)an isopropanolic solution of CuCl2 3 2H2O (11.9 mg, 0.060 mmol,4mL). Themixture was left to stand at ambient temperature for oneday to yield small prismatic light-blue crystals. The nature of thebulk material as pure 7Cu was confirmed by X-ray powder diffrac-tion data compared with the pattern calculated from the single-crystal X-ray structure.

Determination of Phase Diagrams. The batch technique40 hasbeen used to perform a screening of crystallization conditions and tobuild the phase diagram as a function of metal salts and ligandconcentrations. Crystallization trials were rapidly set up in astandard 24-well plate made of polystyrene (aqueous systems) orpolypropylene (methanolic systems). Each well was air-tightlysealed with a 22 mm diameter glass coverslip and vacuum grease.A set of solutions with different concentrations of metal salts andbpp ligand were prepared in deionized water or methanol. Crystal-lization trials were performed by mixing the appropriate amount ofreagent solutions and pure solvent so that a final volume of 2.0 mLfor each fixed concentration of the components was obtained. Theconcentrations of metal salt and bpp ligand were systematicallyvaried along the rows and the columns of the plate. The sampleswere monitored, by using an optical microscope coupled to a CCDcamera, until equilibrium conditions were reached.

An extensive study was carried out only for copper and cobaltchlorides in water andmethanol solvent systems. For all othermetalcations (Mn2þ, Fe2þ, Ni2þ, Cd2þ, Zn2þ), the nature and purity ofthe products formed by precipitation from water or methanol werechecked by XRPD analysis by comparing the experimental patternswith the simulated ones from the single crystal structures. Micro-crystalline powder samples were prepared following a standardprocedure: typically, a water (or methanolic) solution of bpp ofappropriate concentrationwas added to an aqueous (ormethanolic)solution of the metal salt while stirring. In water a 1:3 metal-ligandmolar ratio was used while in methanol the reagents were mixed in1:1 and 1:2 ratio according to the stoichiometry of the two cobaltspecies isolated from this solvent. Precipitate was formed and thereaction mixture was left to react for ∼2 h. The precipitate wasfiltered through a Buchner funnel, washed with small portions ofsolvent and dried in air. A summary of the results and further detailsabout crystallization conditions for each plate and powder samplesare given as Supporting Information.

Gel Double Diffusion Experiments. Two similar crystallizationexperiments, based on the double diffusion technique, were carriedout by usingU-shaped tubes41 (5mm i.d. and 130mm long with twoidentical arms of 30 mm length). The tubes were filled up withagarose gel of high strength prepared at 1% (w/v) and then solutionsof the reagents (∼0.5 mL) of the appropriate concentrations werelayered above the gel on opposite arms of the tubes. The twocrystallization trials differ both for the solvent system and theconcentration of reactants. In the first experiment (Figure 6) equi-molar (0.128 M) aqueous solutions of bpp and copper salt (CuCl2 32H2O) were employed while in the second one (Figure 7) methanoland lower concentrations (0.099 M) were used.

Phase Transformations. The crystal transformations among dif-ferent MOFs were monitored by means of an optical microscopecoupled to a CCD camera connected to a computer equipped with aframe-grabber. Time-lapse acquisition for automatic archiving ofimage series was performed at a suitable frequency.

1Df 2D Transformation.A crystal of the 1D copper species 1Cuwas placed in a small cell to which 0.2 mL of a 0.1MCuCl2 aqueoussolution was added. The transformation took place at 25 �C inabout 4 h (Figure 3).

1Df 3D Transformation. A crystal of 1Cu was placed in a smallcell to which pure ethanol was added. The transformation tookplace at 25 �C in about 5 min.

2Df 1D Transformation. Crystals of the 2D copper species 2Cuwere spontaneously nucleated in a small cell by preparing anaqueous solution 4 mM and 12 mM with respect to CuCl2 andbpp, respectively. Subsequent appearance of 1Cu crystals startswithin 15-30min; equilibrium is reached at 25 �C in about 3-5 dayswith complete comsumption of 2Cu crystals (see Figure 1S, Sup-porting Information).

2Df 3D Transformation. A crystal of 2Cu was placed in a smallcell to which pure ethanol was added. The transformation tookplace at 25 �C in about 50 min.

Electroconductance Measurements. Solubilities of compounds1Co, 1Ni, 1Cu, and 1Cd have been determined at 25.0 ( 0.1 �Cusing a thermostatic bath. Metal(II) chloride and bpp ligand solu-tions were mixed in a 1:3 M/L ratio and molar concentrationsufficient to give a microcrystalline precipitate (1Cd) or crystals(1Co, 1Ni, 1Cu) of the corresponding 1D MOF species. Chemicalequilibrium was reached after 3 weeks. Solubilities were determinedby comparison of electroconductance value of the saturated solu-tion after diluting 1:2 with deionized water against a set of standardsolutions of the pertinent 1D species.

Single Crystal X-ray Structure Characterization. Crystal data forall the compounds are listed in Table 1S, Supporting Information.The data collections were performed with Mo KR (λ=0.71073 A)on a Bruker APEX II CCD area-detector diffractometer for 1Co,1Ni, 3Cd, 3Co, 3Fe, 3Mn, 3Ni, 6Co, and 7Cu and on an Enraf-Nonius CAD4 for 1Cd, by the ω-scan method. An empiricalabsorption correction was applied42 for the structures collectedwith the CCD detector, while the ψ-scan43 method was usedfor 1Cd. The structures were solved by direct methods (SIR9744

or SHELX-S97)45 and refined by full-matrix least-squares onF2 (SHELXL-97)46 with WINGX interface.47 Anisotropic thermalparameters were commonly assigned to all the nondisordered non-hydrogen atoms except for the partially occupied atoms of theclathrate solvent molecules in some of the structures. In compounds1Cd, 1Co, and 1Ni one of two independent ligands was founddisordered and refined isotropically using two models with occu-pancies of 52 and 48% (1Cd), 50 and 50% (1Co), 59 and 41% (1Ni).In the same structures the water molecules were refined anisotropi-cally but with half occupancy. The hydroxylH atom associated withthe ethanol molecule in 3Mn was located on a difference Fouriermap and refined freely, withUiso(H)=1.2Ueq(O), while all the otherones were placed in geometrically calculated positions and there-after refined using a riding model with the correct occupancy fordisorderedmodels. All the structure diagramswere performed usingthe TOPOS48 SCHAKAL9949 programs.

Results and Discussion

As the nature of coordination polymers can bemodified bythe conformation of the chosen ligands, coordination geome-try preferred by the metal, counterions, solvent system, andmetal-to-ligand ratio, in the following sections we discuss thephenomenology and the interplay among some of the chemi-cal parameters relevant to the selection of the crystallinespecies obtained aftermixing ofmetal(II) chlorides and ligandbpp.50

The CuCl2/bpp/Water System. We already reported38 onthe possibility to obtain a whole family of coordinationpolymers based on the couple CuCl2/bpp in water, withnew interesting 1D, 2D, and 3D topologies (Figure 1) de-pending on the choice of several chemical parameters. TheCuCl2/bpp system subsequently afforded another new spe-cies when the reaction was conducted in THF solutions (seenext section). In ref 38, we also briefly explored the network-ing properties of the chlorides of other M2þ cations withsuitable ionic radius and coordination stereochemistry thatin principle can afford species isostructural to those obtainedfrom CuCl2 (see Table 1).

By working with aqueous CuCl2, a 1D species ([Cu(bpp)3-Cl2] 3 2H2O, 1Cu), consisting of interdigitated zigzag chains

Article Crystal Growth & Design, Vol. 9, No. 12, 2009 5027

containing dangling monocoordinated bpp ligands, and a2D coordination polymer ([Cu(bpp)2Cl]Cl 3 1.5H2O, 2Cu),that now can be better defined as a 3D array formed frominclined polycatenation of 2D square layers in the diago-nal-diagonal mode (density of catenation DOC= (2/2))5a

have been obtained using different Cu/bpp ratios and

concentrations. Independently from us another group pre-pared a 1D MnCl2/bpp coordination polymer isostructuralwith our Cu(II) 1D system.51 On the contrary, the presenceof alcohols or other organic solvents (acetone, dichloro-methane) leads to the formation of a 3D 4-fold interpene-trated diamondoid network ([Cu(bpp)2Cl2] 3 2.75H2O, 3Cu)(see next section). Curiously, while in these systems phenom-ena of polymorphism and supramolecular isomerism areoften encountered, this seems to be not the case here; indeed,compounds 2Cu and 3Cu have very similar compositiondiffering only in the amount of solvated water molecules,but the chloride ions play a different role in these two speciesso that it is hard to consider them supramolecular isomers.

During initial exploration of the behavior of aqueoussolutions of CuCl2 and bpp in a 1:2 ratiowe obtained crystalsof the 1D phase 1Cu by slow solvent evaporation at RT.While trying to improve size and quality of 1Cu crystals byvarying crystallization conditions, that is, M/L ratio, abso-lute concentrations, and crystal growth technique (see later),we observed the formation of a new crystallineMOF (species2Cu), easily recognized due to its deep blue color anddifferent crystal morphology compared to 1Cu. However,upon standing for a few days in the presence of the mothersolution the deep blue crystals of 2Cu disappeared leavingonly pale blue crystals of 1Cu. Therefore, under the selectedconditions, 1Cu was the thermodynamically stable product(showing indeed a lower solubility) whose formation ispreceded by the nucleation of crystals of 2Cu, the kineticallyfavored species. This sequence of events follows Ostwald’srule of stages52 which allows the possibility that differentcrystalline species (e.g., polymorphs, or different hydrates orsolvates) can nucleate sequentially starting from the leaststable one that in turn transforms into a thermodynamicallymore stable species and so on until only the most stablespecies survive owing to its lowest solubility. Supersatura-tion conditions that provoke at first the nucleation of 2Cufollowed by the more stable 1Cu (but see later for theopposite 1Cu to 2Cu transformation) are mirrored in thelarge number of small single crystals of 2Cu while crystals of1Cu are generally heavily twinned rosettes. Nucleation of2Cu occurs within minutes after mixing of reactants (e.g.,CuCl2 4.0 mM and bpp 12.0 mM), while the subsequentsolvent-mediated transformation into 1Cu requires severaldays to go to completion as monitored by means of timelapse optical microscopy (Figure 1S, Supporting Information).

Table 1. Summary of Crystallographic Data for Coordination Polymers Obtained from Different MCl2/bpp/Solvent Systems

compound MOF a (A) b (A) c (A) β (deg) crystal system sp group

1Cda 1D 16.516(4) 26.779(6) 17.705(3) orthorhombic Ibca

1Coa 1D 16.580(3) 27.108(3) 17.122(6) orthorhombic Ibca

1Cub 1D 17.191(1) 16.242(1) 26.859(2) 91.59(1) monoclinic I2/a

1Fea 1D 16.517(2) 26.922(4) 17.221(2) orthorhombic Ibca1Mna 1D 16.571(3) 26.888(5) 17.525(3) orthorhombic Ibca1Nia 1D 16.484(3) 26.901(3) 16.973(6) orthorhombic Ibca2Cu

b 2D 16.314(1) 18.211(1) 36.247(3) orthorhombic P2121213Cd

a 3D 17.536(1) 17.536(1) 42.966(1) tetragonal I41/a3Co

a 3D 17.187(1) 17.187(1) 42.299(3) tetragonal I41/a3Cub 3D 17.221(1) 17.221(1) 40.965(2) tetragonal I41/a3Fea 3D 17.081(2) 17.081(2) 42.554(5) tetragonal I41/a3Mnc 3D 17.201(2) 17.201(2) 42.697(4) tetragonal I41/a3Ni

a 3D 16.991(4) 16.991(4) 41.947(9) tetragonal I41/a6Co

a 1D 5.189(1) 12.980(1) 10.493(1) 93.589(2) monoclinic P21/m6Zn

a,d 1D 5.2254(4) 12.9371(9) 10.5425(6) 94.247(3) monoclinic P21/m7Cua 2D 22.062(9) 16.984(8) 15.647(7) 118.183(6) monoclinic C2/c

aThis work. bRef 38. cRef 51. dRef 54.

Figure 1. Coordination polymers obtained with the CuCl2/bpp/water system showing new interesting 1D, 2D, and 3D topologies.

5028 Crystal Growth & Design, Vol. 9, No. 12, 2009 Carlucci et al.

This behavior is highly reproducible and has been checkedseveral times. In the context of MOFs chemistry severalother systems displayed crystal-to-crystal transformations.53

After this accidental discovery, we pursued a more ra-tional approach by studying the influence of the Cu2þ/bpp

ratio and absolute concentration values on kinetics and finalequilibrium composition. Considering that solid 1Cu and2Cu have Cu2þ/bpp ratios respectively of 1:3 and 1:2, asystematic exploration of Cu2þ/bpp ratio and total concen-tration of reactants has been undertaken with the batchapproach.40 Results for aqueous CuCl2/bpp system aresummarized in Figure 2 which extends up to ca. 15 mM forboth CuCl2 and bpp (solubility of bpp being of ca. 60 mM inwater at RT). The equilibrium phase diagram clearly showsdominance of the 1D polymer over the 2D species startingfrom a bpp/CuCl2 ratio of∼2, while the absolute concentra-tion value has a negligible influence over stability regions ofthe two species. The phase diagram also shows the presenceof the microcrystalline species Cu4Cl2(OH)6 (4) identified bymeans of XRPD. This copper(II) hydroxochloride is pro-duced when the Lewis basicity of bpp is overcome by itsBroensted-Lowry basicity at low ligand concentrations.From the point of view of kinetics, increasing the bpp/CuCl2ratio reduces the nucleation time for species 1Cu to an extentfor which the preliminary nucleation and appearance ofcrystals of 2Cu becomes negligible or not discernible underthe microscope at the highest magnification. Low ligand-to-metal ratios favor formation of 2Cu at the expense of 1Cu.Hence, when proper conditions are used, the 1T 2 reversibletransformation can be driven to completion at will so that weare able to selectively nucleate and obtain pure crystal

Figure 3. Selected frames from a video microscopy record of the solvent mediated 1Cuf 2Cu transformation induced by a proper increase ofthe CuCl2 concentration (time lapse between subsequent images is 15 min; crystal size is ca. 2 mm).

Figure 2. Equilibrium phase diagram for aqueous CuCl2/bpp sys-tem with indication of the stability regions for 1D polymer 1Cu, 2Dpolymer 2Cu, andCu4Cl2(OH)6 (4) species as a function of [L]þ [M]total concentration and [L]/[M] ratio (M=Cu2þ, L=bpp).

Article Crystal Growth & Design, Vol. 9, No. 12, 2009 5029

samples of both species through equilibration. Indeed, basedon our knowledge of chemical equilibria of aqueous CuCl2/bpp system, it has been possible to drive to completion inabout 4 h also the solvent-mediated 1Cuf 2Cu transforma-tion as shown in Figure 3.

Since compounds 1Cu and 2Cu are characterized bydifferent Cu/bpp stoichiometries the 1Cuf 2Cu transforma-tion produces the release of the excess of bpp as a ghostlyhalo surrounding the transforming crystals (last frames inFigure 3). Clearly, the unidentified solution species leadingto 1Cu or 2Cu depending on chemical conditions can bemanipulated tomove equilibrium bymass action law towardthe desired product, independently from the kinetic featuresof the system previously described. A hint about the natureof the solution species that self-assemble to give 1Cu or 2Cu(the nucleation step in term of crystal growth mechanisms)has been collected by constructing a Job plot55 with CuCl2and bpp solutions 10.7 and 20.0 mM. Unfortunately, stabi-lity constants for complexes between the bpp ligand andcopper(II) are not very high and the Job plot of Figure 4shows a significant rounding of absorbance vsmolar fractionascending and descending arms which hampers a clear-cutanalysis of data. Indication is that at total [Cu2þ] þ [bpp]=10.7mM the average formula of solution species lies between

a 1:2 and 1:3M:L ratio, while at 20.0mMtotal concentrationtheM:L ratio slightly shifts toward 1:3, but basically withoutstatistical significance. This trend is however in accordancewith the decreasing nucleation time required for 1Cu whenincreasing the bpp/Cu ratio.

Demonstration that with proper knowledge of thermo-dynamics and kinetics of ametal-ligand system it is possibleto control the outcome of MOFs synthesis is a relevantmessage to all those involved with this kind of solid-statepreparation; in fact, the selected pair of ligand andmetal saltcan afford several different species in a perfectly controllablemanner when the system is carefully studied (Figure 5).

The previous results could be further refined in terms ofcapability to selectively obtain the desired product by ex-ploring the parameters relevant to nucleation and growth ofcrystals of the different species. By crystallizing in gels56

within U-shaped tubes and choosing experimental condi-tions on the basis of the phase diagram (Figure 2), it waspossible to afford in the same crystallization tube the threespecies observed in water, that is, Cu4Cl2(OH)6 (4), 2Cu, and1Cu in order of increasing bpp to Cu ratio. On the right armof the tube in Figure 6 was put an aqueous solution of CuCl2while on the left reservoir an aqueous solution of bpp wascharged; the horizontal branch of the tube contained a 1%w/w agarose gel. Counter diffusion leads to concentrationgradients of the two reactants which enabled us to obtainthe aforementioned species in a single experiment accordingto the phase diagram, that is, 4 at very low, 2Cu at inter-mediate, and 1Cu at high bpp/Cu2þ ratios, respectively(Figure 6). This experimental setup is inter alia very usefulto tailor reactants concentration in order to obtain goodsingle crystals avoiding excessive nucleation, spherulitic,dendritic or twinned crystals, all representative of a veryhigh supersaturation. By properly selecting concentrations,it has also been possible to drive the system toward produc-tion of pure samples of 2Cu, without 1Cu or 4 as byproducts.

Figure 5. Scheme illustrating the reversible solvent mediated chemical transformation 1Cu T 2Cu.

Figure 4. Job plot for the CuCl2/bpp aqueous system with indica-tion of the average composition of solution species at χCu ∼0.28,that is, intermediate between the stoichiometry of 1Cu and 2Cu.

Figure 6. Crystallization in aU-shaped tube by counterdiffusion ofbpp (left arm) and CuCl2 (right arm) showing the concomitantpresence of species Cu4Cl2(OH)6 (4, microcrystalline), 2Cu and 1Cuin order of increasing bpp to Cu ratio from right to left. Thehorizontal part of the U-shaped tube is filled with a gel of highstrength agarose to avoid convective mixing of the reactants.

5030 Crystal Growth & Design, Vol. 9, No. 12, 2009 Carlucci et al.

The same experimental setup can be used to produce mor-phodromes, that is, a detailed study of dependence of crystalmorphology on supersaturation or ratio between reactingspecies, a feature that can be of relevance for technologicalapplications of MOFs. It must be mentioned that morphol-ogy of crystals can become a delicate issue when planningpractical applications of crystalline materials, for example,for catalysis or gas sorption. These results, in view of the factthat no general and reliable models for the formation ofcoordination networks from solution are presently available,can be a useful guide for careful exploration and optimiza-tion of conditions for synthesis of crystalline coordinationpolymers. An illustrative example of the severe difficultieswhen trying to reproducibly synthesize and study the arche-typal structureMOF-5 are discussed in a recent paper.27Alsotemperature plays a fundamental role on determining kine-tics and thermodynamics of coordination polymer synthesisand we briefly discussed it in ref 38. Recently, temperaturehas been exploited to select the final product and polymertopology through variation of conformation of a flexibleligand.57

The CuCl2/bpp/Organic Solvent Systems. Other crystal-lization trials involved use of organic solvents (methanol,ethanol, isopropanol, acetone, dichloromethane) for dissol-ving the ligand bpp and copper chloride. Presence of anorganic solvent directs the self-assembling process towardthe 3D species 3Cu, [Cu(bpp)2Cl2] 3 2.75H2O, that does notcontain any solvated organic molecule within the cavities ofthe 3D metal framework.38 Therefore, it must be assumedthat the organic solvent, without entering the coordinationnetwork, modifies the nature of the unknown solutionspecies and/or their self-organization which eventually as-semble into the interpenetrated structure of 3Cu. The role ofmethanol is well illustrated in a refined crystallization ex-periment with aqueous agarose gel which, thanks to compo-sition gradients, in particular those involving methanol,produced both species 3Cu and 2Cu on sides of the growthtube where bpp concentration was higher or lower, respec-tively (Figure 7). Hence, beyond the presence of the organicsolvent also the bpp/Cu2þ ratio comes again into play ondetermining which species can nucleate and develop intomature crystals.

Aiming at a better control of chemical equilibria involvingour polymeric coordination species, we tried to transform1Cu and 2Cu, obtained in the absence of organic solvents,into 3Cu by selecting the proper conditions. Indeed, crystalsof 1Cu and 2Cuwere unstable in the presence of pure organicsolvents such as methanol, ethanol, or methylene chloride.As an example, upon immersing a crystal of 1Cu in ethanolwe could observe its quick conversion into microcrystallineaggregates pseudomorphic of the starting crystal and whosecomposition was showed by XRPD to be that of 3Cu

(Figure 8). Analogously, 2Cu can be converted into 3Cu

with the same procedure but with a process that takes longertimes to reach equilibrium. Therefore, as highlighted inScheme 1, we could devise procedures to reversibly convert1Cu into 2Cu and vice versa, together with conversion of 1Cuand 2Cu into 3Cu.

Parallel to formation of 3Cu, crystallization from organicsolvents produced also a yellow sparingly soluble, and al-most amorphous, powder (5Cu), represented by the yellowzones in the phase diagram and appearing at low total con-centrations and bpp/Cu ratios. The formation of compound5Cu has been observed in several preparations involving

organic solvents such as methanol, isopropanol, or amylacetate, but it has been possible isolate it in a crystalline pureform only by the method described above. In any case, theanalysis of crystallization trials performed using a wideselection of conditions and configurations has confirmedthat a low ligand-to-metal molar ratio favors unknownspecies 5Cu as evidenced by the phase diagram. The ele-mental analysis of the yellow phase58 indicated that thestoichiometric composition for 5Cu corresponds to a 1:1combination of ligand and metal ion. These results and thesimilarities between the copper diagram in methanol and thecobalt one led us to hypothesize that the structures of yellow5Cu and the blue 1D-polymer 6Co, containing cobalt ionstetrahedrically coordinated, could be closely related. Indeed,there are many examples of tetrahedral Cu(II) complexescontaining N-donor ligands, some of which, but not all,show a yellow color. Unfortunately, other derivatives, aspyridinium salts of copper halide,59 could precipitate underthe experimental conditions used, sowe cannot be sure aboutthe polymeric nature of the yellow phase.

We also built the equilibriumphase diagram for theCuCl2/bpp/MeOH system with a batch procedure (Figure 9). Devia-tions from the correct 3:1 bpp/Cu stoichiometry of 3Cu

produce significant variations of the crystal morphology,but the combined effects upon supersaturation hamperinterpretation of morphodromes (Figure 9 right). The fea-ture common to all growth conditions is that, on average,species 3Cu exhibits a fast growing [001] direction, alongwhich the 4-fold interpenetrated networks stack on top ofeach other. This feature has also been observed in otherdiamantoid networks based on nitriles.

When using THF to dissolve ligand bpp the new species[Cu(bpp)1.5Cl2] 3 (THF) 3 0.75H2O (7Cu) has been obtained,exemplifying the different role for this solvent, that can besupposed to be an active templating agent. Compound 7Cu

consists of a two-dimensionally extended hexagonal array(Cu 3 3 3Cu 13.349 and 13.489 A) with open pores that accom-modate thenoncoordinatingTHFandwatermolecules stackedalong the [201] direction with an ABAB sequence. The copperatoms exhibit a distorted square pyramidal co-ordinationgeometry with the basal plane formed by two nitrogens ofpyridyl groups and two chloride atoms arranged in transconfiguration, while another pyridine nitrogen occupies theapical position at a rather long distance 2.27(1) A. The twoindependent bpp ligands display both a trans-trans (TT)conformation (N-to-N of 9.644 and 9.859 A). A single layerof 7Cu is shown in Figure 2S in the Supporting Information.

Figure 7. Crystallization by counterdiffusion in aU-shaped tube ofmethanolic solutions of bpp (diffusing from the left arm) and CuCl2(diffusing from the right arm). Note the concomitant presence ofmicrocrystalline Cu4Cl2(OH)6 (4), 2Cu and 3Cu in order of incre-asing bpp/Cu ratio from right to left. The horizontal part of theU-shaped tube is filled with a gel of high strength agarose to avoidconvective mixing of the reactants.

Article Crystal Growth & Design, Vol. 9, No. 12, 2009 5031

The MCl2/bpp/Water (M=Mn, Fe, Co, Ni, Cd) Systems.

As preliminarly reported in ref 38 substitution of copper(II)chloride with other bivalent cations allowed us to isolatecoordination polymers isostructural to 1Cu (Table 1), whileno analogues of 2Cu have been recovered. This differentbehavior has to be related to the octahedral coordination ofcopper in 1Cu, a local stereochemistry easily available alsofor Mn, Fe, Co, Ni and Cd 2þ ions, at variance with thesquare pyramidal pentacoordination of copper in 2Cuwhichis found less frequently for the other 2þ cations.60 Differences

in ionic radii induced only minor structural effects on the 1Dpolymers of Co(II), Ni(II), and Cd(II) even though theircrystal structures are described in different crystal systems(Table 1). Electroconductance measurements of saturatedsolutions and comparison between crystallization trials in-dicate that for 1D polymers of formula [M(bpp)3Cl2] 3 2H2O1Cu, 1Cd, 1Co, 1Ni, 1Mn, and 1Fe the solubility order isMn>Fe>Co>Ni>Cu>Cd. Values for the solubilityat 25 �C for 1Co, 1Ni, 1Cu, and 1Cd are 2.81, 1.66, 0.73, and0.51 mM, respectively.

Further insights about similarities and differences betweencopper and the other divalent cations were gathered from theexperimental determination with the batch technique of thephase diagrams for MCl2/bbp systems in water. As anexample, concentrations had to be increased up to 50 mMfor both bpp and Co2þ, about three times higher than thatused for the less soluble copper polymer. A thorough analy-sis of batch experiments for the 1D polymer 1Co shows,compared to copper, a shrinking of the nucleation zonemoved at higher concentrations. From a practical point ofview, the increased solubility of all 1D networks with theexception of Cd, is reflected in more difficulties to obtaincrystals on moving from copper to cobalt and nickel, thelatter one being the most soluble system. Crystals of thederivative 1Ni could be obtained only from isothermalevaporation of solutions containing Ni2þ cation and bppligand because there is no possibility to reach the criticalconcentration necessary for nucleation by simply mixingcation and ligand. On the contrary, trials in high densityhydrogels evidenced for the cadmium derivative 1Cd a prompt

Figure 8. Video recording of the 1Cu f 3Cu transformation in ethanol by optical microscopy with polarized light. Time interval betweenframes is 20 s. Development of opacity tracks conversion of a single crystal of 1Cu into microdomains of crystalline 3Cu.

Scheme 1. Chemical Reactions Relative to the Feasible

Interconversions among Species 1Cu-3Cu

5032 Crystal Growth & Design, Vol. 9, No. 12, 2009 Carlucci et al.

crystallization due its very low solubility. In fact, crystallizationof 1Cdmust be carefully controlled in order to avoid dendriticor spherulitic growth owing to the very high supersaturationreachable in the mother solution. A high growth rate can beeasily obtained producing several centimeters long dendriticcrystals in a few minutes after contact of the bpp and CdCl2solutions.

TheMCl2/bpp/Organic Solvent (M=Mn, Fe, Co, Ni, Cd)Systems. Similarities among Cu andMn, Fe, Co, Ni, and Cd2þ cations were further investigated by using an alyphaticalcohol (methanol, ethanol, isopropanol) as solvent for bothMCl2 and bpp or alternatively an alcohol for dissolving themetal salt and dichloromethane to dissolve bpp. In this way,it has been possible to obtain crystalline coordination poly-mers akin to 3Cu for all metal cations studied. Solubilityranking for these 3D metal-organic networks is analogousto that discussed for the 1D polymers crystallizing fromwater as confirmed by the reaction yields obtained frommethanol at high concentrations of the reagents and atmetal/bpp molar ratio of 1:2. The formation of pure pro-ducts in all cases (with exception of Cd) was checked byXRPD analysis, by comparing the experimental patternswith the simulated ones from the single crystal structures.These results evidence once more the necessity to explore forany solvent system chosen as reaction media a wide range ofconcentrations and reagents molar ratio. In fact, data col-lected for the phase diagram of the CoCl2/bpp/methanolsystem with the batch technique evidenced a shift towardhigher concentrations with respect to the case of copper,together with a higher bpp/M2þ molar ratio. Analysis of thecrystallization batch (Figure 10) revealed also the presence ofthe new species 6Co characterized by blue crystals appearingjust beneath the metastable zone of the 3D polymer 3Co.Also the prismatic morphology of species 6Co is clearlydifferent from that typically bipyramidal (with or withoutthe (100) pinacoid) of the 3D polymer 3Cu but also for Ni,Cd,Mn, andFe. The crystal structure of 6Co has the formula[Co(bpp)Cl2] and is isomorphous to the published 1D poly-mer [Zn(bpp)Cl2]

54 obtained via hydrothermal synthesisusing a mixture of ZnCl2 in H2O and bpp ligand in ethanol.This species, which is probably related to compound 7Cu

discussed in the case of copper, consists of festoon chains ofCo atoms interconnected by bpp ligands in TT conforma-tion. The distorted tetrahedral coordination geometry ofcobalt cations are completed by two chloride anions. To

confirm the primary role of the tetrahedral coordinationaround the metal ion in the assembly of Zn and Co 1Dpolymers, we have reprepared the Zn-bpp compound 6Zn, inboth powder and single crystal forms, following a moreconventional way of crystallization (i.e., using the liquid-liquid diffusion method) than the hydrothermal reactionused by Yao and co-workers.54 We have found that thebetter solvent system to prepare this compound in a suitablecrystalline form for X-ray diffraction measurements is waterand a combination of water and methanol.

Conclusion

We discussed kinetical and thermodynamical features ofthe crystallization behavior ofMOFs based onMCl2 salts andthe ligand 1,3-bis(4-pyridyl)propane. Our systematic ap-proach, using different crystallization techniques, to exploreconditions enabling the synthesis of new MOF systems is ofgeneral application to a variety of coordination polymers. Inparticular, the self-assembly of coordinationnetworks and thephase transformations here described between species withdifferent metal to ligand ratios and/or clathrated solvents are

Figure 9. (left) Equilibrium phase diagram for CuCl2/bpp/MeOH with indication of the stability regions for 3D polymer 3Cu (blue area) and5Cu (yellow area) species as a function of [L]þ [M] total concentration and [L]/[M] ratio (M=Cu2þ, L=bpp). (right)Morphodrome showingdependence of crystal morphology and habit upon metal and ligand concentration.

Figure 10. Equilibrium phase diagram for methanolic CoCl2/bppsystem with indication of the stability regions for 1D polymer 6Co(blue area), species 3Co (pink area), and an intermediate region ofcoexistence, as a function of [L] þ [M] total concentration and[L]/[M] ratio (M=Co2þ, L=bpp).

Article Crystal Growth & Design, Vol. 9, No. 12, 2009 5033

intimately related tokinetic and thermodynamic factors activeduring the crystallization processes. We showed that MOFspecies which are kinetically favored can be substituted by thethermodynamical product and conversion among differentspecies canbeperformedafter proper chemical knowledgehasbeen gained.

The presentwork, while not resolving the very nature of theself-assembling processes leading to polymeric coordinationnetworks, shows, however, that it is possible to finely tune andcontrol the synthetic route toward a desired MOF. Thecapability to explore extensively the physicochemical vari-ables of the metal-ligand solution equilibria is a majorimprovement over a simple buy-and-mix approach.

Acknowledgment. M.M. and S.R. thank FondazioneCARIPLO for financial support. J.M.G.-R. acknowledgesMICINN project “Factorıa de Crystalizaci�on, Consolider-Ingenio-2010”.

Supporting Information Available: X-ray crystallographic infor-mation files (CIF) for compounds 1Cd, 1Co, 1Fe, 1Ni, 3Cd, 3Co,3Fe, 3Mn, 3Ni, 6Co, 7Cu, Figure 1S showing crystallization of 2Cufollowed by its transformation into 1Cu, Figure 2S with details ofthe structure of 7Cu, Figures 3S and 4S showing batch crystal-lization results for 1,3-bis(4-pyridyl)propane þ CuCl2 3 2H2O inmethanol and CoCl2 3 6H2O in water, respectively, Table 1S withcrystal data and Table 2S with a summary of microcrystallinepowder samples prepared with bpp and Mn, Fe, Ni, Cd, Znchlorides in water and methanol. This information is available freeof charge via the Internet at http://pubs.acs.org.

References

(1) Lehn, J.-M.Supramolecular Chemistry - Concepts andPerspectives;VCH: Weinheim, 1995.

(2) (a) Janiak, C. Dalton Trans. 2003, 2781–2804. (b) Mueller, U.;Schubert, M.; Teich, F.; Puetter, H.; Schierle-Arndt, K.; Pastr�e, J.J. Mater. Chem. 2006, 16, 626–636.

(3) Zaworotko, M. J. Cryst. Growth Des. 2007, 7, 4–9.(4) (a) Kitagawa, S.; Uemura, K. Chem. Soc. Rev. 2005, 34, 109–119.

(b) Kitagawa, S.; Kitaura, R.; Noro, S. Angew. Chem., Int. Ed. Engl.2004, 43, 2334–2375. (c) Matsuda, R.; Kitaura, R.; Kitagawa, S.;Kubota, Y.; Kobayashi, T. C.; Horike, S.; Takata, M. J. Am. Chem.Soc. 2004, 126, 14063–14070.

(5) (a) Carlucci, L.; Ciani, G.; Proserpio, D. M. Coord. Chem. Rev.2003, 246, 247–289. (b) Barnett, S. A.; Champness, N. R. Coord.Chem. Rev. 2003, 246, 145–168.

(6) Brammer, L. Chem. Soc. Rev. 2004, 33, 476–489.(7) Evans, O. R.; Lin, W. Acc. Chem. Res. 2002, 35, 511–522.(8) (a) Kahn, O. Acc. Chem. Res. 2000, 33, 647–657. (b) Carballo, R.;

Covelo, B.; El Fallah, M. S.; Ribas, J.; V�azquez-L�opez, E. M. Cryst.Growth Des. 2007, 7, 1069–1077.

(9) Ma, S.; Sun, D.; Wang, X.-S.; Zhou, H.-C. Angew. Chem., Int. Ed.Engl. 2007, 46, 2458–2462.

(10) Khlobystov, A. N.; Champness, N. R.; Roberts, C. J.; Tendler,S. J. B.; Thompson, C.; Schr€oder,M.CrystEngComm 2002, 4, 426–431.

(11) Fujita,M.; Kwon, Y. J.; Washizu, S.; Ogura, K. J. Am. Chem. Soc.1994, 116, 1151–1152.

(12) Robin, A. Y.; Fromm, K. M. Coord. Chem. Rev. 2006, 250, 2127–2157.

(13) Wu, C.-D.; Lin, W. Angew. Chem., Int. Ed. Engl. 2007, 46, 1075–1078.

(14) (a) Rosseinsky, M. J. Microporous Mesoporous Mater. 2004, 73,15–30. (b) Fletcher, A. J.; Thomas, K. M.; Rosseinsky, M. J. J. SolidState Chem. 2005, 178, 2491–2510.

(15) (a) Llewellyn, P. L.; Bourrelly, S.; Serre, C.; Filinchuk, Y.; F�erey,G. Angew. Chem., Int. Ed. Engl. 2006, 45, 7751–7754. (b) Serre, C.;Mellot-Draznieks, C.; Surbl�e, S.; Audebrand, N.; Filinchuk, Y.; F�erey,G. Science 2007, 315, 1828–1831.

(16) (a) Kubota, Y.; Takata, M.; Matsuda, R.; Kitaura, R.; Kitagawa,S.; Kato, K.; Sakata, M.; Kobayashi, T. C.Angew. Chem., Int. Ed.Engl. 2004, 43, 2–5. (b)Matsuda, R.; Kitaura, R.; Kitagawa, S.; Kubota,

Y.; Kobayashi, T. C.; Horike, S.; Takata, M. J. Am. Chem. Soc. 2004,126, 14063–14070. (c) Kondo, A.; Noguchi, H.; Ohnishi, S.; Kajiro, H.;Tohdoh, A.; Hattori, Y.; Xu, W.-C.; Tanaka, H.; Kanoh, H.; Kaneko, K.NanoLett. 2006, 6, 2581–2584. (d) Yildirim, T.; Hartman,M. R.Phys.Rev. Lett. 2005, 95, 215504. (e) Ueda, T.; Kurokawa, K.; Eguchi, T.;Kachi-Terajima, C.; Takamizawa, S. J. Phys. Chem. C 2007, 111,1524–1534. (f) Dinca,M.; Han,W. S.; Liu, Y.; Dailly, A.; Brown, C.M.;Long, J. R. Angew. Chem., Int. Ed. Engl. 2007, 46, 1419–1422.(g) Kondo, A.; Noguchi, H.; Carlucci, L.; Proserpio, D. M.; Ciani, G.;Kajiro, H.; Ohba, T.; Kanoh, H.; Kaneko, K. J. Am. Chem. Soc. 2007,129, 12362–12363.

(17) Jia, J.; Lin, X.; Wilson, C.; Blake, A. J.; Champness, N. R.;Hubberstey, P.; Walker, G.; Cussen, E. J.; Schr€oder, M. Chem.Commun. 2007, 840–842.

(18) (a) Wu, H.; Zhou, W.; Yildirim, T. J. Am. Chem. Soc. 2007, 129,5314–5315. (b) Lin, X.; Jia, J.; Hubberstey, P.; Schr€oder,M.; Champness,N. R.CrystEngComm 2007, 9, 438–448. (c) Collins, D. J.; Zhou, H.-C.J. Mater. Chem. 2007, 17, 3154–3160.

(19) Hill, R. J.; Long,D.-L.; Hubberstey, P.; Schr€oder,M.; Champness,N. R. J. Solid State Chem. 2005, 178, 2414–2419.

(20) Daiguebonne, C.; Deluzet, A.; Camara, M.; Boubekeur, K.;Audebrand, N.; Gerault, Y.; Baux, C.; Guillou, O. Cryst. GrowthDes. 2003, 3, 1015–1020.

(21) Cahill, C. L.; de Lill, D. T.; Frisch,M.CrystEngComm 2007, 9, 15–26.

(22) Soldatov, D . V.; Enright, G. D.; Ripmeester, J. A. Cryst. GrowthDes. 2004, 4, 1185–1194.

(23) Blake, A. J.; Champness, N. R.; Hubberstey, P.; Li, W.-S.;Withersby, M. A.; Schr€oder, M. Coord. Chem. Rev. 1999, 183,117–138.

(24) Biradha, K.; Sarkar, M.; Rajput, L. Chem. Commun. 2006, 4169–4179.

(25) Rowsell, J. L. C.; Yaghi, O. M. Angew. Chem., Int. Ed. Engl. 2005,44, 4670–4679.

(26) (a) Khlobystov, A. N.; Blake, A. J.; Champness, N. R.; Lemenovskii,D. A.; Majouga, A. G.; Zyk, N. V.; Schr€oder, M.Coord. Chem. Rev.2001, 222, 155–192. (b) Barnett, S. A.; Champness, N. R. Coord. Chem.Rev. 2003, 246, 145–168. (c) Th�ebault, F.; Barnett, S. A.; Blake, A. J.;Wilson, C.; Champness, N. R.; Schr€oder, M. Inorg. Chem. 2006, 45,6179–6187.

(27) Hafizovic, J.; Bjoergen, M.; Olsbye, U.; Dietzel, P. D. C.; Bordiga,S.; Prestipino, C.; Lamberti, C.; Lillerud, K. P. J. Am. Chem. Soc.2007, 129, 3612–3620.

(28) Livage, C.; Egger, C.; Ferey, G. Chem. Mater. 2001, 13, 410–414.(29) Caveat should be noted to avoid the misuse of solution spectro-

scopy in the rationalization of structural features of the extendedsolid-state species: IR and NMR solution data, indeed, can onlygive information about the fragments of the dissolution process.

(30) Levin, M. D.; Stang, P. J. J. Am. Chem. Soc. 2000, 122, 7428–7429.(31) Wheaton, C. A.; Jennings, M. C.; Puddephatt, R. J. J. Am. Chem.

Soc. 2006, 128, 15370–15371.(32) May, L. J.; Shimizu, G. K. H. Z. Kristallogr. 2005, 220, 364–372.(33) Kawano, M.; Haneda, T.; Hashizume, D.; Izumi, F.; Fujita, M.

Angew. Chem., Int. Ed. Engl. 2008, 47, 1269–1271.(34) Carlucci, L.; Ciani, G.; Moret, M.; Proserpio, D. M.; Rizzato, S.

Angew. Chem., Int. Ed. Engl. 2000, 39, 1506–1510.(35) Carlucci, L.; Cozzi, N.; Ciani, G.; Moret, M.; Proserpio, D. M.;

Rizzato, S. Chem. Commun. 2002, 1354–1355.(36) (a) Du, M.; Zhang, Z.-H.; Wang, X.-G.; Tang, L.-F.; Zhao, X.-J.

CrystEngComm 2008, 10, 1855–1865. (b) Ren, P.; Liu, M.-L.; Zhang,J.; Shi, W.; Cheng, P.; Liao, D.-Z.; Yan, S.-P. Dalton Trans. 2008,4711–4713. (c) Lan, Y.-Q.; Li, S.-L.; Fu, Y.-M.; Xu, Y.-H.; Li, L.; Su,Z.-M.; Fu, Q. Dalton Trans. 2008, 6796–6807. (d) Wen, Q. Y.; Chen,J. Q.; Fang, H. C.; Chen, L.; Zhou, Z. Y.; Hu, Z. J.; Cai, Y. P.Z. Anorg.Allg. Chem. 2008, 634, 567–571. (e) Yu,W.; Zhang, T. L.; Zhang, J. G.;Yang, L.; Wang, S. Z.; Wu, R. F. Z. Anorg. Allg. Chem. 2008, 634,754–757. (f) Sun, C. Y.; Li, X. X.; Wang, S. F.; Li, W. J.; Zheng, X. J.Z. Anorg. Allg. Chem. 2008, 634, 950–955. (g) Yang, P. P.; Li, B.;Wang, Y. H.; Gu, W.; Liu, X.Z. Anorg. Allg. Chem. 2008, 634, 1210–1214. (h) Habib, H. A.; Hoffmann, A.; H€oppe, H. A.; Steinfeld, G.;Janiak, C. Inorg. Chem. 2009, 48, 2166–2180. (i) Yang, J.; Ma, J.-F.;Liu, Y.-Y.; Batten, S. R.CrystEngComm 2009, 11, 151–159. (j) Habib,H. A.; Hoffmann, A.; H€oppe, H. A.; Janiak, C. Dalton Trans. 2009,1742–1751.

(37) Carlucci, L.; Ciani, G.; Proserpio, D. M.; Rizzato, S. CrystEng-Comm 2002, 4, 121–129.

(38) Carlucci, L.; Ciani, G.; Moret, M.; Proserpio, D. M.; Rizzato, S.Chem. Mater. 2002, 14, 12–16.

5034 Crystal Growth & Design, Vol. 9, No. 12, 2009 Carlucci et al.

(39) (a) Garcia-Ruiz, J. M.; Moreno, A. Acta Crystallogr. D 1994, 50,484–490. (b) Garcia-Ruiz, J. M.; Moreno, A. J. Cryst. Growth 1997,178, 393–401. (c) Garcia-Ruiz, J. M.; Moreno, A.; Ot�alora, F.; Rondon,D.; Viedma, C.; Zauscher, F. J. Chem. Educ. 1998, 75, 442-446.

(40) (a) Robert, M. C.; Garcıa-Ruiz, J. M.; Vidal, O.; Ot�alora, F. InDucruix, A.; Gieg�e, R., Eds.; Crystallization of Nucleic Acids andProteins:APracticalApproach; IRLPress at OxfordUniversity Press:Oxford, 1999. (b) García-Ruiz, J. M. Key Eng. Mater. 1991, 58, 87–106.

(41) http://www.trianatech.com(42) Sheldrick, G. M. SADABS, Program for Area Detector Absorption

and Other Corrections; University of G€ottingen: Germany, 2001.(43) North, A. C. T.; Phillips, D. C.;Mathews, F. S.Acta Crystallogr, A

1968, 24, 351–359.(44) Altomare,A.; Burla,M.C.; Camalli,M.; Cascarano,G.;Giacovazzo,

C.; Guagliardi, A.;Moliterni, A. G.; Polidori, G.; Spagna, R. J. Appl.Crystallogr. 1999, 32, 115–119.

(45) Sheldrick, G. M. SHELXS-97, Program for Crystal StructureSolution; University of G€ottingen: Germany, 1997.

(46) Sheldrick, G. M. SHELXL-97, Program for Crystal StructureRefinement; University of G€ottingen: Germany, 1997.

(47) Farrugia, L. J. J. Appl. Crystallogr. 1999, 32, 837–838.(48) (a) Blatov, V. A.; Shevchenko, A. P.; Serezhkin, V. N. J. Appl.

Crystallogr. 2000, 33, 1193. (b) Blatov, V. A. Crystallogr. Rev. 2004,10, 249–318.

(49) Keller, E. SCHAKAL99; University of Freiburg: Germany, 1999.(50) Melikhov, I. V.; Berdonosova, D. G. Theor. Found. Chem. Eng.

2004, 38, 255–273.(51) Ghosh, A. K.; Ghoshal, D.; Lu, T.-H.; Mostafa, G.; Chaudhuri,

N. R. Cryst. Growth Des. 2004, 4, 851–857.(52) Ostwald, W. Z. Phys. Chem. 1897, 22, 289–330.(53) (a) Khlobystov, A. N.; Champness, N. R.; Roberts, C. J.; Tendler,

S. J. B.; Thomson, C.; Schr€oder, M. CrystEngComm 2002, 4, 426–

431. (b) Thomson, C.; Champness, N. R.; Khlobystov, A. N.; Roberts,C. J.; Schr€oder, M.; Tendler, S. J. B.; Wilkinson, M. J. J. Microscopy2004, 214, 261–271. (c) Xue, D.-X.; Zhang, W.-X.; Chen, X.-M.;Wang, H.-Z. Chem. Commun. 2008, 1551–1553. (d) Habib, H. A.;Sanchiz, J.; Janiak, C.Dalton Trans. 2008, 1734–1744. (e) Hu, S.; He,K.-H.; Zeng, M.-H.; Zou, H.-H.; Jiang, Y.-M. Inorg. Chem. 2008, 47,5218–5224. (f) Ye, J.; Liu, Y.; Zhao, Y.; Mu, X.; Zhang, P.; Wang, Y.CrystEngComm 2008, 10, 598–604. (g) Bataille, T.; Costantino, F.;Lorenzo-Luis, P.; Midollini, S.; Orlandini, A. Inorg. Chim. Acta 2008,361, 9–15. (h) Aslani, A.; Morsali, A. Chem. Commun. 2008, 3402–3404. (i) Aslani, A.; Morsali, A.; Zeller, M.Dalton Trans. 2008, 5173–5177.

(54) Chen, Y.-B.; Kang, Y.; Qin, Y.-Y.; Li, Z.-J.; Cheng, J.-K.; Hu,R. F.;Wen, Y.-H.; Yao, Y.-G.Acta Crystallogr. C 2004, 60, m168–m169.

(55) Hill, Z. D.; MacCarthy, P. J. Chem. Educ. 1986, 63, 162-167.(56) (a) Henisch, H. K. Crystal Growth in Gels; The Pennsylvania State

University Press: University Park, PA, 1970. (b) Henisch, H. K.;Garcia-Ruiz, J. M. J. Cryst. Growth 1986, 75, 195–202. (c) Henisch,H. K.; Garcia-Ruiz, J. M. J. Cryst. Growth 1986, 75, 203–211.

(57) Dong, Y.-B.; Jiang, Y.-Y.; Li, J.; Ma, J.-P.; Liu, F.-L.; Tang, B.;Huang, R.-Q.; Batten, S. R. J. Am. Chem. Soc. 2007, 129, 4520–4521.

(58) Elemental analyses were carried out at the Microanalytical Labo-ratory of the University of Milano. Calcd. for [Cu(bpp)Cl2]:C13H14CuN4Cl2: C, 46.93; H, 4.24; N, 8.42%. Found: C, 42.27;H, 4.25; N, 7.45%.

(59) (a) Rice, C. R.; Onions, S.; Vidal, N.; Wallis, J. D.; Senna, M. C.;Pilkington, M.; Stoeckli-Evans, H. Eur. J. Inorg. Chem. 2002, 8,1985–1997. (b) Willett, R. D.; Awwadi, F.; Butcher, R.; Haddad, S.;Twamley, B. Cryst. Growth Des. 2003, 3, 301–311.

(60) Cotton, F. A.; Wilkinson, G.; Murillo, C. A.; Bochmann, M.Advanced Inorganic Chemistry; Wiley & Sons: New York, 1999.