Magnetic , Photophysical and Thermal Properties of Complexes of Iron ( II ) with Structurally Different Schiff Bases

Fe(II) complexes made up of N-donor ligands are widely investigated as spin crossover (SCO) materials. This is because the valence electronic configuration of Fe(II) (3d) may be reversibly switched between low spin (LS, t2g) and high spin (HS, t2geg) state by external stimuli, such as temperature, pressure and light irradiation. More recently, Fe(II) complexes are attracting attention as potential photosensitizers in dye-sensitized solar cells (DSSC), to replace the more expensive and toxic Ru(II) complexes. For these purposes, ligands derived from Schiff bases are ideally suited as they are easy to prepare, structurally versatile and form stable complexes with most transition metal ions. This paper presents the syntheses of three structurally different Schiff bases, L1, H2L2 and H2L3 (Fig. 1) and the corresponding Fe(II) complexes. Ligand L1 was a neutral N3-donor Schiff base, H2L2 was a multi-N donor conjugated Schiff base and H2L3 has two N,O-donors separated by an eight-carbon aliphatic chain. The main objective was to show the effect of different donor atoms of the ligands on the structures, magnetic, photophysical and thermal properties of Fe(II) complexes formed. [Fe(L1)2](BF4)2 (1) was formed in a one-pot reaction involving Fe(BF4)2·6H2O and L1, while [Fe2(OOC(CH2)14CH3)2(L2)(H2O)2] (2) and [Fe2(OOC(CH2)14 CH3)2(L3)(H2O)4]·21⁄2H2O (3) were formed in step-wise reactions involving [Fe(OOC(CH2)14CH3)2(EtOH)] with H2L2 Magnetic, Photophysical and Thermal Properties of Complexes of Iron(II) with Structurally Different Schiff Bases

and H2L3, respectively.A common feature of these complexes was the presence of linear 16-carbon alkyl chains, introduced in order to lower their melting temperatures 20 and to induce mesomorphism (s) 21,22 .Asian Journal of Chemistry; Vol. 27, No. 7 (2015), 2359-2364 EXPERIMENTAL All chemicals were analaR reagents and used as received.The elemental analyses (CHN) were carried out on a Thermo Finnigan Flash EA 1112.The 1 H NMR spectra were recorded in CDCl 3 and DMSO on a JEOL FT-NMR lambda 400 MHz spectrometer.The FTIR spectra were recorded for neat samples from 4000 to 450 cm -1 on a Perkin-Elmer Frontier FTIR spectrometer equipped with a diamond attenuated total reflectance attachment.Magnetic susceptibility at room temperature was determined on a Sherwood automagnetic susceptibility balance by the Gouy method.The instrument was calibrated using Hg[Co(NCS) 4 ] and the molar susceptibility value was corrected for the diamagnetism of the constituent atoms using Pascal's constants.Variable temperature magnetic susceptibility was measured on a Quantum Design MPMS XL EverCool SQUID magnetometer at Kinki University, Higashiosaka-shi, Osaka, Japan.About 10 mg of the sample was placed inside a gelatine capsule and inserted halfway inside a drinking straw to a depth of about 10 cm from the top.The straw was then inserted into the instrument.The measurements were recorded at 1 Tesla (10,000 Gauss) in the temperature range of 300-4 K.The raw data was analysed using Microsoft Excel and IGOR Pro.The UV-visible spectrum was recorded in CHCl 3 and DMSO from 1200 to 400 nm on a Shimadzu UV-visible-NIR 3600 spectrophotometer.Thermogravimetric analysis (TGA) was done on a Perkin-Elmer Pyris Diamond TG/DTA thermal instrument under N 2 at a flow rate of 10 cm 3 min -1 .The temperature range was 50-900 °C and the scan rate was 20 °C min -1 .Differential scanning calorimetry (DSC) was done on a Mettler Toledo DSC 822 calorimeter in the temperature range 25 °C to a maximum 250 °C under N 2 at a flow rate of 20 cm 3 min -1 and a scan rate of 10 °C min -1 .The scans were recorded during one heatingand-cooling cycle for all complexes and the onset temperatures were quoted for all peaks observed.Polarising optical microscopy (POM) was carried out on an olympus polarizing microscope equipped with a Mettler Toledo FP90 central processor and a Linkam THMS 600 hot stage.The sample was heated in an oven at 60 °C for a few days prior to the analysis.The heating and cooling rates were 10 and 3 °C min -1 , respectively and the magnification was 50x.
Its IR spectrum showed two strong peaks at 2916 and 2850 cm -1 for νasymCH2 and νsymCH2, respectively, a weak peak at 1636 cm -1 for νC=N, a strong peak at 1024 cm -1 for νB-F of noncoordinated BF4 -ion 23 and a medium peak at 519 cm -1 for νFe-N.From these, it may be inferred that L1 was coordinated to Fe(II) ion as a neutral N3-donor ligand (Fig. 2).
HO Scheme-II: Equations for the preparation of compounds 2 and 3: The IR spectrum of compound 2 showed two broad peaks centred at 3333 and 3138 cm -1 for H-bonded H2O molecules, two strong peaks at 2918 and 2850 cm -1 for νasymCH2 and νsymCH2, respectively, a strong peak at 1618 cm -1 for ν(C=N), a strong peak at 1530 cm -1 for νasymCOO, a strong peak at 1393 cm -1 for symCOO, a weak peak at 578 cm -1 for ν(Fe-N) and a weak peak at 491 cm -1 for ν(Fe-O) 24 .For comparison, the IR spectrum of H2L2 showed a weak peak at 3295 cm -1 for ν(N-H) and a strong peak at 1615 cm -1 for ν(C=N).Hence, it may be inferred that in the formation of compound 2, H2L2 were doubly deprotonated at pyrrole N-H, the iminyl nitrogen atoms were not involved in the bonding and the binding mode of CH3(CH2)14COO -ion to Fe(II) atoms was bidentate chelating (∆ = 137 cm -1 ) 25 (Fig. 2).
The IR spectrum of compound 3 showed a weak peak at 3396 cm -1 for water, two strong peaks at 2917 and 2849 cm -1 for νasymCH2 and νsymCH2, respectively, a weak peak at 1612 cm -1 for νC=N, a strong peak at 1574 cm -1 for νasymCOO, a strong peak at 1446 cm -1 for νsymCOO, a medium peak at 592 cm -1 for νFe-N and a medium peak at 461 cm -1 for ν(Fe-O) 24 .For comparison, the IR spectrum of H2L3 showed a weak peak at 3406 cm -1 for ν(O-H), two strong peaks at 2919 cm -1 and 2851 cm -1 for νasymCH2 and νsymCH2, respectively and a strong peak at 1633 cm -1 for ν(C=N).Hence, it may be inferred that in the formation of compound 3, the two phenolic O-H group of H2L3 was deprotonated and the phenolic oxygen and iminyl nitrogen were coordinated to Fe(II) atom.In addition, the ∆ value for CH3(CH2)14COO -ligand was 128 cm -1 , suggesting a chelating binding mode to both Fe(II) centres 25 (Fig. 2).Magnetic and photophysical properties: Complex 1 was diamagnetic at room temperature as the value of its mass susceptibility (χg), determined by the Gouy method, was negative.Its electronic absorption spectrum in chloroform shows two strong intraligand bands at 473 nm (εmax = 6476 M -1 cm -1 ) and 576 nm (εmax = 7100 M -1 cm -1 ), a strong singlet metal-to-ligand charge transfer band ( 1 MLCT) at 596 nm (εmax = 8710 M -1 cm -1 ) assigned to t2g → π* electronic transition 9 and a weaker d-d band at 721 nm (εmax = 470 M -1 cm -1 ) assigned to 1 A1g → 1 T1g electronic transition 26 .Hence, compound 1 was a low-spin iron(II) complex.From this, it may be inferred that L1 was a strong field ligand.From the electronic spectral data, the optical band gap (Eg) for compound 1, calculated using the equation: Eg = hc/λ, where c = velocity of light, h = Planck constant and λ (absorption edge of CT band) = 688 nm) 27 , was 1.8 eV.
In contrast, compound 2 was paramagnetic at room temperature.The value of χMT (χM = molar magnetic susceptibility and T = temperature in K), as determined by the Gouy method, was 6.2 cm 3 K mol -1 at 298 K.The expected value for a dimeric high spin Fe(II) octahedral complex 28 is 6 cm 3 K mol -1 .From these, it may be inferred that both Fe(II) atoms in the complex were high spin, with negligible electronic communication between the two Fe(II) centres.Thus, L2 2-ion was a weak field ligand.The electronic absorption spectrum for the complex in DMSO shows a shoulder at 488 nm (εmax = 3478 M -1 cm -1 ) assigned to the intraligand electronic transition and a weak d-d band at 773 nm (εmax = 489 M -1 cm -1 ) assigned to 5 T2g → 5 Eg electronic transition 29,30 .The Eg value, similarly calculated as for compound 1 using λ = 630 nm, was 1.9 eV.
For compound 3, the value of χMT, as determined by the Gouy method, was 4.17 cm 3 K mol -1 at 300 K. Since low spin Fe(II) is diamagnetic (t2g 6 ) and the expected χMT value for a complex with two high spin Fe(II) atoms and g = 1.3 (see below) 29 is 2.54 cm 3 K mol -1 , it may be inferred that compound 3 comprised of two high spin Fe(II) atoms at this temperature 31 .Additionally for compound 3, its temperature-dependence magnetic susceptibility was measured using a SQUID magnetometer in the temperature range 300-2 K.The experimental curve of χMT versus T (Fig. 3) shows a good fit with the theoretical curve constructed from the formula for a symmetrical dinuclear complex given below 28 and inserting the values of g = 1.3 and J (the isotropic interaction parameter) = -28.4cm -1 into the formulae: From Fig. 3, it is noted that the χMT values decreased gradually from 4.15 cm 3 K mol -1 at 294 K to 2.03 cm 3 K mol -1 at 8.8 K and then more rapidly to about 1.78 cm 3 K mol -1 at 4 K as a result of zero-field splitting.It is also noted that the g value for compound 3 was significantly lower than the theoretical value (2.0023), suggesting a highly distorted octahedral environment, expected for high spin Fe(II) atoms due to the weak Fe-L bonds.The low J value (-28.4 cm -1 ) indicates a weak antiferromagnetic interaction between the two Fe(II) centres, postulated to occur through H-bonds between the coordinated and lattice H2O 32 .The results suggest a normal but incomplete SCO behaviour in this temperature range.The electronic absorption spectrum of compound 3 in chloroform shows two overlapping bands at 468 nm (εmax = 114.3M -1 cm -1 ) and 510 nm (εmax = 63.5 M -1 cm -1 ) assigned to intraligand electronic transitions.The Eg value was 2.3 eV (λonset = 534 nm).
Thermal properties: The thermal data (TGA and DSC) for complexes 1-3 are shown in Table-1.The TGA traces are shown in Fig. 4, while the DSC scans are shown in Fig. 5.The TGA trace for compound 1 (Fig. 4a) shows a gradual weight loss totaling 90.2 % on heating from 260 °C to about 640 °C.This may be due to the decomposition of BF4 -ion and L1 ligand (expected 93.3 %).The amount of residue above this temperature was 9.8 %.For compound 2, the TGA trace (Fig. 4b) shows a rapid weight loss of 90.4 % on heating from about 205 to 620 °C.This may be due to evaporation of coordinated H2O and decomposition of CH3(CH2)14COO -and L2 ligands (expected 88.8 %).The amount of residue above this temperature was 9.6 %.For compound 3, the TGA trace (Fig. 4c) shows an initial weight loss of 4.1 % in the temperature range 100-250 °C, assigned to evaporation of lattice H2O (expected 3.5 %).On further heating to about 660 °C, it suffered a total weight loss of 78 %, assigned to loss of coordinated H2O and decomposition of CH3(CH2)14COO -(expected 85.1 %).The amount of residue above this temperature was 17.9 %.For phase transitions, the DSC of complex 1 (Fig. 5a) shows two endotherms on heating at 62.6 °C (∆H = +16.0kJ mol -1 ) assigned to crystal-to-crystal transition and at 89.6 °C (∆H = +94.1 kJ mol -1 ) assigned to crystal-to-mesophase transition.On cooling from 150 °C, there was an exotherm at 44.5 °C (∆H = -54.1 kJ mol -1 ) assigned to mesophase-to-crystal transition.The sample, sandwiched between two glass covers, was then heated on a hot stage and the changes observed under a polarized optical microscope (POM) were as follows: it melted at about 89 °C and then cleared to an isotropic liquid at 193 °C.On cooling from the isotropic liquid phase, batonetts of smectic A developed at 133 °C, which transformed into the smectic A phase at 126 °C (Fig. 6b).Hence, the complex has liquid crystal properties of a calamitic mesogen.Similar mesophase was found for [Co(C16-terpy)2](BF4)2 compound 33 .
For compound 3, its DSC (Fig. 5c) shows two overlapping endotherms on heating at 77 °C (∆Hcombined = -48 kJ mol -1 ), assigned to crystal-to-crystal and crystal-to-isotropic liquid transitions (the latter was observed at 104 °C under POM).Its DSC shows no peaks on cooling from 180 °C to room temperature and POM did not show any optical textures.Hence, it may be concluded that 3 was not mesogenic.temperature with weak antiferromagnetic interaction and incomplete SCO at low temperature, large optical bandgap (2.3 eV), but no liquid crystal properties.Hence, these complexes are potential SCO materials, but only compound 1 is a potential dye-sensitized solar cell material.