Controlling spin transition in one-dimensional coordination polymers through polymorphism.

A series of polymer 1 microcrystals with several different morphologies have been systematically synthesized by controlling experimental parameters, namely concentration of reactants, temperature, solvent nature, and the use of surfactants, and their valence tautomerism (VT) has been studied by combined electron microscopy, X-ray diffraction data, and magnetization. Our results indicate that all of them can be grouped exclusively into two different crystalline phases, or a mixture of them, that critically determine the VT process, independent of the morphology and/or dimensions of the crystals. Moreover, a difference in the critical temperature of both phases by more than 50 K allows us to regulate VT. These results head the use of valence tautomeric 1D polymers in devices where strict control and reproducibility of the switching behavior at different length scales and integration procedures is highly required.


INTRODUCTION
The influence of polymorphism on spin transition coordination polymers is used to tune the critical interconversion transition temperatures by establishing novel methodologies for the morphology-selective fabrication of coordination polymer microcrystals with reproducible crystalline phases and properties.
Crystal engineering of one-dimensional coordination polymers (1D-CP) with welldefined functionalities has been shown as an efficient route to produce materials of The assembly of coordination polymers at the micro-/nanoscale (CPPs) with welldefined crystal symmetries and morphologies represents an excellent scenario for these studies. CPPs allow for a proper structure-morphology-functionality correlation(9) and, therefore, enable the rational design of new generations of coordination polymers targeting specific desired properties. For instance, the influence of the precursors,(10) nature of the metal nodes, ancillary ligand, and counterions,(11) and use of surfactants(12) or aggregation phenomena(13) on the final morphology/properties of CPPs have been stated. However, even though some examples of spin transition 1D-CP at the micro-/nanoscale have already been reported, (14) no detailed studies to determine morphology and crystalline phase influence on the resulting spin transition was reported. (15) To fill this gap, we have envisioned a systematic study of several different microsized structures of the same coordination polymer. An excellent model for these studies is the one-dimensional coordination polymer 1 (see Scheme 1), bearing a cobalt metal ion and catechol ligands connected by the 4,4′-bipyridine ligand. In a previous communication polymer 1 was already obtained as amorphous nanoparticles and exhibited VT behavior with striking and unpredictable differences with respect to its crystalline counterpart.(18) Such differential behavior is not new. VT switching takes place through an intramolecular electron between the metal ion and a redox-active ligand in response to temperature modifications.(17) Although of intramolecular origin, the matrix nature has already been shown to condition such equilibria (18)  Herein several different batches of polymer 1 have been obtained by precipitation (sometimes spontaneous) from a poor solvent. Control over the morphology and size of the resulting precipitate is finely tuned by controlling experimental parameters, namely concentration, temperature, solvent nature, and even the use of surfactants.
A detailed study by combined electron microscopy, X-ray diffraction data, and magnetization studies have allowed us for the first time to extract conclusive information about the phonon matrix influence on the VT equilibria.

Experimental Section
All operations were carried out open to air. Chemicals were obtained from different suppliers and used without further purification. The stirring time was kept constant during the different experiments to facilitate the methodological study; all of the experiments shown in this work have been done in a semistatic range (2 min of stirring at 800 rpm) to ensure the mixing of the reagents and the mixture was subsequently allowed to stand. Finally, the resulting CPPs were characterized by FT-IR spectroscopy, X-ray diffraction, scanning electron microscopy (SEM), thermal gravimetric analysis (TGA), and magnetic measurements.

Synthesis of 1: [CoIII(4,4′-bipy)(3,5-dbsq)(3,5-dbcat)] (α-and β-phase)
In general, the complexes was prepared by procedures similar to those in our previous report, changing some experimental conditions.(19) In this work close to 25 different reaction conditions were performed to synthesize materials with the same chemical composition but containing different crystalline phases (α/β) depending on two main factors: reagent concentration and temperature. In all cases the ratio among cobalt salt, linker ligands, and catechol ligands was 1:1:2, respectively. In general, the stabilization of the α-phase at room temperature was achieved when the concentration (referred to metallic salt) was lower than 0.01 M: 4,4′-bipyridyl (20 mg, 0.125 mmol) and 3,5-di-tert-butylcatechol (55 mg, 0.25 mmol) were first dissolved in

Physical Measurements
Scanning electron microscopy (SEM) images were collected on a scanning electron microscope (HITACHI S-570) at acceleration voltages of 10-15 kV. Aluminum was used as support. Magnetic susceptibilities of samples were measured in the temperature range 5-400 K with a Quantum Design MPMS XL SQUID magnetometer operating at a magnetic field strength of 0.1 T. Infrared (IR) spectra were performed on a FTIR Tensor 27 infrared spectrophotometer (Bruker) equipped with a Bruker Golden Gate diamond ATR (attenuated total reflection) cell. Thermal stabilities of the compounds were measured as a function of temperature and time using a simultaneous thermogravimetric analysis-differential scanning calorimetry/differential thermal analysis (TGA-DSC/DTA) system, NETZSCH-STA 449 F1 Jupiter, under nitrogen and air atmosphere and at ambient preasure. The sensitivity of the balance is 0.07 μg, and the furnace can operate from room temperature to 1400 °C. Powder X-ray diffraction spectra were recorded at room temperature on a PANalytical X'Pert PRO MRD (high-resolution texture goniometer) working in reflection mode. The diffractometer was equipped with a Co Kα radiation source (λ = 1.7903 Å). Ultraviolet-visible and reflectance spectra were obtained using an Agilent Cary 5000 spectrophotometer equipped with an external diffuse reflectance integrating sphere and a temperature-controlled cuvette.

Results and Discussion
In a typical reaction, an aqueous solution of Co(CH3COO)2·4H2O is added to an ethanol solution containing 1 equiv of 4,4′-bipyridyl and 2 equiv of 3,5-di-tertbutylcatechol at 25 °C, resulting in the production of a blue crystalline precipitate within a few minutes, identified after filtration as polymer 1. The powder diffraction pattern of a sample obtained by using a 0.01 M concentration of Co(CH3COO)2·4H2O (while keeping constant the molar ratios of the two ligands) nicely fits with that simulated from the single-crystal X-ray data (monoclinic space group C2/c) previously reported (see S2 in the Supporting Information), (16)     with an arbitrary color code: α-phase (blue), α+β-phase (violet), β-phase (orange).
As far as the crystal phase is concerned, we only obtain one of the two phases, either the α-phase or β-phase, or a combination of both. No appearance of peaks associated with new phases was detected. At 0 °C the α-phase is predominant, while at 70 °C the predominant phase is the β-phase; at an intermediate temperature of 25 °C the phase is strongly dependent on the reagent concentrations (see Figure   3).

Use of Surfactants
The surfactants used (in all cases below 5%) were the polymeric polyvinylpyrrolidone (PVP) and the anionic surfactant sodium dodecylbenzenesulfonate (SDS). As can be seen in Figure   Finally, the use of surfactants does not modify the crystalline phase of the resulting material, it being dictated by the initial concentration of the cobalt salt. Moreover, it was confirmed that the surfactants do not take part in the final material composition.

Solvent Effects
In addition to EtOH, DMF and THF were also used, due to the good solubility of the ligands and their miscibility with water. SEM images of the aforementioned structures are shown in Figure 5. Interestingly, all the microcrystals obtained in THF consist of superstructures hierarchically formed by very thin needles that range from urchinlike to unstructured samples, depending on the concentration or temperature used. A more broad range of morphologies is found by using DMF. Depending on both the initial cobalt salt concentration and/or the temperature, several different morphologies ranging from urchinlike to plates or flowerlike supersturctures are obtained. As far as the crystal phase is concerned, all the microcrystals obtained in THF yield the β-phase independent of the concentration or temperature used. In contrast, the crystal phase of the samples obtained in DMF changes depending on the initial metal salt concentration and/or temperature, in a manner similar to that observed for the structures obtained in ethanol. Color code: α-phase (blue), α+β-phase (violet), β-phase (orange).

Magnetic Measurements
Magnetic susceptibility data were measured for all the different morphologies in the 5-380 K temperature range. All of them can be grouped into two different magnetic behaviors (or a combination of them), depending on the crystalline phase and independent of the morphology. Representative examples of magnetization curves found for the α-phase and β-phase and their mixtures are shown in Figure 6a. In the low-temperature range of 5-250 K only the ls-[CoIII(4,4′-bipy)(3,5-dbsq) (3,5-dbcat)] (S = 1/2) tautomer is observed in all cases. As the temperature is increased, ligand-to-metal intramolecular electron transfer is activated as one of the ligands evolves from the catecholate form (3,5-dbcat) to the semiquinone radical ligand (3,5dbsq Figure 6a).
Finally, we also monitored changes in the unit cell (C2/c) parameters upon increasing the temperature from 223 to 423 K, where the VT has been shown to take place (calculation details are presented in S4 in the Supporting Information). As can be seen in Figure 6b, an increase of the Co-N bond length by ca. 0.2 Å (0.36 Å if an increase of ionic radius from ls-CoIII to hs-CoII is included) is observed, involving a significant expansion along the b and c axes:(22) i.e., the polymeric chain propagation vector. These changes imply an increase of the total cell volume of 218 Å3 along the temperature range studied, though the most abrupt volume variation (171 Å3) is observed between room temperature and 423 K. This is the range where the largest changes in the effective magnetic moment are found. These structural changes are reversible over several cycles, as previously observed by variabletemperature magnetization data.

CONCLUSIONS
By controlling experimental reaction parameters, several microparticles of polymer 1 have been obtained and studied by combined X-ray diffraction and SEM studies (a detailed list of all the different reactions and morphologies obtained is given in S1 in the Supporting Information). Interestingly, all of them can be exclusively grouped into two crystalline phases or a mixture of them. Moreover, the VT transition is conditioned by the crystal symmetry of the phase independent of the crystal morphology and/or dimensions. The use of low reactant concentrations at room temperature (or below) favors the α-phase, where the transition starts at 275 K with a maximum value of μ eff = 5.3 μ B reached at 380 K. A 4-fold increase of the concentration above room temperature ensures the formation of the β-phase, which exhibits a T c shift by more than 50 K. Therefore, the possibility of controlling the fabrication of microparticles with specific crystal symmetry allow us for the first time to design VT microparticles with reproducible switching and controlled VT temperatures, a major problem in the design of future molecular devices based on these materials.

SUPPORTING INFORMATION
Text, figures, and tables giving a summary of the experimental conditions and the corresponding morphologies and crystalline phase of the [CoIII(4,4′-bipy)(3,5dbsq) (3,5-dbcat)] samples obtained, X-ray data for 1, thermogravimetric analysis of the α-and β-phase of polymer 1, and cell parameters extracted from XRD data at different temperatures.