Citation
Quesada Vilar, M. (2007, March 29). Spin-transition frameworks based on bistetrazole and triazine ligands. Retrieved from https://hdl.handle.net/1887/11463
Version: Corrected Publisher’s Version
License: Licence agreement concerning inclusion of doctoral thesis in the Institutional Repository of the University of Leiden
Downloaded from: https://hdl.handle.net/1887/11463
Note: To cite this publication please use the final published version (if applicable).
4
Counterion effect on the spin-transition
properties of the [Fe(btzx) 3 ] 2+ cation
(btzx = m-xylylene(bis)tetrazole)
Abstract
The spin-transition properties of the system [Fe(btzx)3]X2 (X = PF6– (3), CF3SO3– (4) and ClO4– (5))depend on the nature of X. The degree of completion and the transition temperature are both drastically influenced by the counterion used. Surprisingly, the cooperative nature of the transition is not affected by the choice of counterion. The X-ray crystal structures of compounds 3 and 4 are presented, and the magnetic susceptibility measurements of all three compounds are compared. Compound 4 is further studied by Mössbauer spectroscopy and its LIESST properties are presented, while compound 3 has been analysed by means of DSC.
4.1 Introduction
The rational design of highly efficient and specific functional materials is now a main area of research, as the so-called top-down approach is reaching its limits of miniaturisation.1 The bottom-up approach represents an exciting alternative, based on the idea that the miniaturisation limit of an electronic function is the molecule. Spin-crossover (SCO) systems fall in this category, and can work as molecular switches or memory storage devices.2, 3 These materials are able to pass from a low spin (LS) to a high spin (HS) state by means of an external stimulus.4 This entropy driven phenomenon5 is sometimes accompanied by a thermocromic effect, which makes these systems valuable for optical-storage devices.
The spin-transition phenomenon is strictly molecular, but its characteristics are dependent on the overall crystal lattice. The narrow range of ligand field strength for which a spin-transition may be observed makes these systems sensitive, not only to the first coordination sphere, but also to the second coordination sphere. Entities such as counterions or lattice-solvent molecules may have a drastic effect on the spin-transition properties of the material.4 Bistetrazole-based polymers belong to this family of compounds showing spin- transition properties.6 Gradual to steep temperature dependent magnetic responses have been reported with bis-tetrazole compounds.7 The variation of the size of the spacer linking the two tetrazole rings has resulted in compounds with different dimensionality.8, 9 The size, conformation and flexibility of the spacer all seem to have an important effect on the magnetic behaviour of the resulting materials. Weinberg and co-workers recently published a study on the relation between the size of the alkyl spacer and the T1/2 and optical properties.10 An influence of the size of the spacer on the thermo-magnetic and magneto-optical properties is expected, but a relation with the parity (number of carbons) of the spacer has also been found. Further systematic studies are necessary to understand the role played by each chemical piece constituting the metal-organic network.
In this respect, no study has yet been undertaken to understand in detail the influence of the counterion on the spin-transition behaviour within this family of compounds. This lack is mainly due to the synthetic problems that arise when preparing coordination compounds from these ligands with different FeII salts, and to the subsequent lack of occurrence of SCO.
The role of the counterion is to template the crystallisation process. Most importantly, the size of the counterion has an effect on the quality of the crystallisation.10 In addition, the exchange of counterions very often results in an annihilation of the SCO properties. Consequently, its influence on the transition is very difficult to investigate. So far, the only report mentioning the preparation of FeII spin-transition compounds with two different counterions involves the ligand btzb (btzb, 1,4-bis(tetrazole-1-yl)butane).8, 11
The compounds described in the present chapter are obtained from a new ligand, bearing two tetrazoles rings bridged by a spacer, namely m-xylene (Chapter 2). All previously reported bistetrazole ligands are based on aliphatic-chain linkers. The synthesis, characterisation and magnetic properties of a new series of compounds containing the cation [Fe(btzx)3]2+ (btzx = m-xylylene(bis)tetrazole) and various anions, i.e. PF6– (3), CF3SO3– (4), and ClO4– (5) are reported below. The crystal structures of the PF6– and CF3SO3– derivatives
are described. Calorimetric measurements and LIESST properties are also included. The syntheses of both ligands and complexes have been described in detail in chapter 2.
4.1 Crystal Structure descriptions.
[Fe(btzx)3](PF6)2·MeOH (3)
The space group for compound 3 at 200 K and at 100 K is P63/m (Table 4.2). The PF6–
anions of [Fe(btzx)3](PF6)2·MeOH are not disordered, while the methanol is crystallographic disordered between two positions. This disorder of the solvent molecule is observed at both temperatures. The FeII metal ion is surrounded by six crystallographically related tetrazole
rings which coordinate through the ND atom,
a forming an almost perfect octahedral environment (see Table 4.1). At 200 K, all Fe-N4 distances amount 2.160(4) Å, in the expected range for an FeII HS centre (Table 4.1).4 This bond length is close to those of published bistetrazole-based compounds,8, 12, 13 and slightly shorter than Fe–Ntz bonds observed for mononuclear tetrazole-based compounds.14 At 100 K, the Fe-N4 distance is 2.001(3) Å, which corresponds to a 7% decrease of the bond length upon the transition.
Figure 4.1. View of the X-ray crystal structure along the b axis, showing the polymeric nature of [Fe(btzx)3](PF6)2·MeOH. The disordered methanol molecules are represented in the space- filling mode.
The coordination sphere for the HS centres characterises a slightly distorted octahedron (set of angles N4–Fe1–N4; 90.48 and 89.52 º), while the LS iron(II) centres exhibit a nearly regular octahedral geometry (set of angles N4-Fe1-N4; 90.13 and 89.87). Based on the distortion parameter Σ for all bistetrazole ST polymers so far reported, compound 3 represents the one with the smallest octahedral distortion in both, the HS and the LS states (see Chapter
a ND stands for “donor atom”, and represents the atom coordinated to the metal ion.
5 for details).b The bent conformation of the ligands results in the formation of a 1D polymer, extending along the c axis. The FeII centres are bridged by three ligands which adopt the same conformation, forming cages in which the disordered methanol molecules (no interactions) are trapped. The Fe···Fe separation along the polymer chains at 200 K is 11.397 Å, and is 11.205 Å at 100 K, while the interpolymeric Fe···Fe distance is 10.679 Å and 10.543 Å, at these respective temperatures. The most important structural changes caused by the spin transition are found along the c axis (see Table 4.2). Surprisingly, there is no significant modification of the structural arrangement in the ligand.
Table 4.1. Selected interatomic distances [Å] and angles [º] for [Fe(btzx)3](PF6)2·CH3OH.
Symmetry operations: i = -y, x-y, z; k = -x + y,- x, z; m = -x, -y, -z; o = y, -x + y, -z; q = x - y, x, -z.
[Fe(btzx)3](PF6)2·CH3OH 200 K 100 K Fe1–N4 (i, k, m, o, q) 2.160(4) 2.0011(29) N4–Fe1–N4i 90.48(17) 90.13(11) N4–Fe1–N4k 90.48(18) 90.13(11) N4–Fe1–N4m 180.00 180.00 N4–Fe1–N4o 89.52(17) 89.87(11) N4–Fe1–N4q 89.52(18) 89.87(11) N4i–Fe1–N4k 90.5(2) 90.13(13) N4i–Fe1–N4m 89.52(17) 89.87(11) N4i–Fe1–N4o 180.0(3) 180.00(16) N4i–Fe1–N4q 89.5(2) 89.87(13) N4k–Fe1–N4m 89.52(18) 89.87(11) N4k–Fe1–N4o 89.5(2) 89.87(13) N4k–Fe1–N4q 180.0(3) 180.00(20) N4m–Fe1–N4o 90.48(17) 90.13(11) N4m–Fe1–N4q 90.48(18) 90.13(11) N4o–Fe1–N4q 90.5(2) 90.13(13)
As for other bistetrazole-based 1D polymers,9, 15 the particular packing of the alkyl chains generates cavities in which the counterions are located (see Figure 4.2). Each polymer chain is separated from its six neighbours by six PF6– (see Figure 4.2 b). Each PF6– anion is shared by three polymers, resulting in the expected Fe/PF6– 1 to 2 ratio. The distance separating the counterions from the metal centres is 5.166 Å for the HS state, and decreases by 0.068 Å upon changing to the LS state (Table 4.4). All the anions are lying in the plane formed by the FeII centres, and each of them interacts via anion–π contacts with three different chains (Figure 4.2). No solvent molecules are present between the polymeric chains in the crystal lattice.
However, solvent molecules are trapped inside the cages formed by three ditopic ligands (see Figure 4.3 a). No intermolecular interactions between the polymer chains are observed.
Unexpectedly, the benzyl-based linkers do not show π-π interactions. The solid-state structure is most likely imposed by the coordination of the ligands to the metal ions.
b Σ symbolises the sum of the deviations from 90º of the 12 cis N–Fe–N angles (see Chapter 5 for more detail)
Table 4.2. Crystallographic data for complex 3 and complex 4. When the parameters depend on the temperature, they are entered in the order 200 K and 100 K (3), and 170 K and 100 K (4).
a) b)
Figure 4.2. a) Representation of the iron coordination sphere of [Fe(btzx)3](PF6)2·MeOH showing the anion–π interactions (distance F1···centroid of the tetrazole ring = 3.137 Å). b) View of the X-ray crystal structure along the c axis.
[Fe(btzx)3](CF3SO3)2·CH3CN (4)
In comparison with [Fe(btzx)3](PF6)2·CH3OH (3), [Fe(btzx)3](CF3SO3)2·CH3CN exhibits a more intricate structure, most likely due to the asymmetric triflate anions.
Nevertheless, a comparable overall picture is maintained; the compound is still constituted of 1D polymeric chains which extend along the c-axis. The crystal turned out to be a merohedral
[Fe(btzx)3](PF6)2·CH3OH (3) [Fe(btzx)3](CF3SO3)2·CH3CN (4) Formula C31H34F12FeN24OP2 C102H99F18Fe3N75O18S6
FW/g mol-1 1104.61 3365.53
Crystal system Hexagonal Trigonal
Space group P63/m P–3
a/Å 10.679(2); 10.543(2) 18.6954(10);18.5942(12) b/ Å 10.679(2);10.543(2) 18.6954(10);18.5942(12)
c/ Å 22.793(6);22.409(6) 23.293(2);23.1106(15) α/º 90;90 90;90
β/º 90;90 90;90
γ/º 120;120 120;120 V/Å3 2251.1(8);2157.2(8) 7050.6(8);6919.9(8)
Z 2 2
ρcalcd/g cm-3 1.6297(6);1.7006(6) 1.5853(2);1.6152(2)
T/K 200(2); 100(K) 170(2); 100(K)
Crystal shape Hexagon Hexagon
Colour White White
twin, with a two-fold rotation axis parallel to the c-axis as the twin operation. The compound crystallises in the P-3 space group and, possesses two unique chains: (i) chain2 lies on a crystallographic –3 axis (spectroscopic symmetry S6) and contains ions labelled as Fe3 and Fe4 (Figure 4.3); (ii) chain1 lies at a crystallographic 3 axis (spectroscopic C3) and contains the Fe1 and Fe2 centres (Figure 4.3). The two-fold axis of the twin operation is parallel to the 3 and –3 axis.
a) b)
Figure 4.3. a) View of the 1D chains present in [Fe(btzx)3](CF3SO3)2·CH3CN (counterions are omitted for clarity), along the a axis. b) View of the different anion-surroundings for each of the four independent FeII centres present in [Fe(btzx)3](CF3SO3)2·CH3CN, all depicted independently and viewed along the c axis.
Chain1: Fe1 and Fe2 centres.
Chain1 is formed by two independent iron centres, namely Fe1 and Fe2, which are both coordinated by six tetrazole rings and are alternatingly connected by btzx ligands. At 170 K, the Fe–N bond lengths for Fe1 are 2.208(4) Å (Fe1–N18) and 2.178(4) Å (Fe1–N1), and those for Fe2 are 2.192(4) Å (Fe2–N8) and 2.196(4) Å (Fe2–N11). These values are all in the range expected for HS FeII centres (see Table 4.3).4 At 100 K, the Fe–N distances for Fe2 amount to 2.061(3) Å (Fe2–N8) and 2.059 (3) Å (Fe2–N11), corresponding to a decrease of 6%. These distances are in the range for LS FeII centres,4 and thus their shortening reflects the HS→LS transition occurring in Fe2 (see below). In contrast, Fe1 shows practically no variation of the
Fe–N coordination bond lengths at 100 K, indicating that this metal centre remains HS throughout the entire temperature range. The distortion of the octahedral geometry can be estimated with the parameter Σ. Both the Fe1 and the Fe2 ions are closer to the ideal octahedron (Σ = 0º) at 100 K than at 170 K (see Table 4.4). At 170 K, the octahedral coordination sphere is significantly more distorted for Fe2, as compared to Fe1. At lower temperatures, the distortion in both centres is comparable. These Σ values contrast with those of the previous related 1D polymeric compound, i.e. [Fe(btzx)3](PF6)2·CH3OH (3), which shows Σ values closer to 0 (reflecting a less important distorsion) both in the HS and in the LS sate (see Table 4.4).
a)
b)
Figure 4.4. View along the b axis of the two crystallographically independent chains, showing the orientation of the trapped solvent molecules. Chain1 (b) is shown below and chain2 (a) above (pictures taken separately).
The metal centres are enclosed by six triflate counterions, each shared by two other metal centres, thus resulting in the expected 2:1 ratio (triflate/Fe). The spatial arrangement of these counterions differs depending on the metal centre. As seen in Figure 4.3 b, two different types of arrangements for the counterions around Fe1 are observed. Both types have their CF3
group close to the metal ion, with Fe–F distances of 4.384 Å (F21–Fe1) and 5.135 Å (F23–
Fe1). These counterions are distributed alternatingly around the metal centre (see Figure 4.3 b). The counterions surrounding the Fe2 ions are organised in a different way. As for Fe1, there are two different spatial dispositions for the counterions around Fe2, which are alternatingly distributed. In this case, the two distinct types of triflate are clearly differentiated as one has its CF3 group closer to the metal centre, while for the other triflate, the –SO3
function is adjacent to the Fe2. The corresponding distances are 4.905 Å for Fe2–O11SO3, and 4.896 Å for Fe2–F13CF3 (Table 4.4). When the temperature decreases, only the Fe1 centre experiences a significant variation of the metal–counterion distances, with a shortening of the F21–Fe1 and F23–Fe1 distances by more than 0.5 Å. The peculiar crystal packing of 1D bistetrazole-based polymers is responsible for this proximity of the counterion to the metal centre (see Chapter 5 for details). These close contacts give rise to strong anion–tetrazole
interactions (Figure 4.5), observed for the anions closest to Fe1 and Fe2 (Table 4.4). The deviation of the anion–tetrazole (centroid) distance between the structure measured at 170 K, and the one measured at 100 K is only significant for the anion closest to Fe1.
Table 4.3. Selected interatomic distances [Å] and angles [º] for [Fe(btzx)3](CF3SO3)2·CH3CN.
Symmetry operations: a = x, y, –1 + z; c = 1–y, x–y, –1 + z; d = 1–y, x–y, z; e = 1–y, x–y, 1+z; f = 1–x + y, 1–x, –1 + z; g = 1– x + y, 1–x, z; i = –y, x–y, z; j = –x+y, –x, z; k = –x, –y, – z; l = –x, –y, 1–z; m = y, –x+y, –z; n = y, –x+y, 1–z; o = x–y, x, –z; p = x–y, x, 1–z.
[Fe(btzx)3](CF3SO3)2·
CH3CN (Chain1) 170 K 100 K [Fe(btzx)3](CF3SO3)2·
CH3CN (Chain2) 170 K 100 K Fe1–N1 2.1782(37) 2.1806(30) N21–Fe3–N21i 90.16(16) 90.27(13) Fe1–N18 2.2082(37) 2.2010(31) N21–Fe3–N21j 90.16(15) 90.27(11) Fe2–N8 2.1920(38) 2.0611(30) N21–Fe3–N21k 180.00 180.00 Fe2–N11a 2.1962(37) 2.0585(31) N21–Fe3–N21m 89.84(16) 89.73(13) Fe3–N21 2.1777(38) 2.1818(29) N21–Fe3–N21o 89.84(15) 89.73(11) Fe4–N28 2.1854(38) 2.0386(31) N21i–Fe3–N21j 90.16(18) 90.27(14) N1–Fe1–N18 178.18(14) 178.33(12) N21i–Fe3–N21k 89.84(16) 89.73(13) N1–Fe1–N1d 89.82(16) 89.74(13) N21i–Fe3–N21m 180.00(23) 180.00(19) N1–Fe1–N18d 89.78(15) 89.66(12) N21i–Fe3–N21o 89.84(18) 89.73(14) N1–Fe1–N1g 89.82(15) 89.74(12) N21j–Fe3–N21k 89.84(15) 89.73(11) N1–Fe1–N18g 91.96(15) 91.82(12) N21j–Fe3–N21m 89.84(18) 89.73(14) N18–Fe1–N1d 91.96(15) 91.82(12) N21j–Fe3–N21o 180.00(25) 180.00(20) N18–Fe1–N18d 88.45(15) 88.79(13) N21k–Fe3–N21m 90.16(16) 90.27(13) N18–Fe1–N1g 89.78(15) 89.66(12) N21k–Fe3–N21o 90.16(15) 90.27(11) N18–Fe1–N18g 88.45(15) 88.79(12) N21m–Fe3–N21o 90.16(18) 90.27(14) N1d–Fe1–N18d 178.18(17) 178.33(14) N28–Fe(4)–N28i 90.97(16) 91.01(14) N1d–Fe1–N1g 89.82(16) 89.74(13) N28–Fe4–N28j 90.97(15) 91.01(13) N1d–Fe1–N18g 89.78(16) 89.66(13) N28–Fe4–N28l 180.00 180.00 N18d–Fe1–N1g 91.96(16) 91.82(13) N28–Fe4–N28n 89.03(16) 88.99(14) N18d–Fe1–N18g 88.45(16) 88.79(13) N28–Fe4–N28p 89.03(15) 88.99(13) N1g–Fe1–N18g 178.18(13) 178.33(11) N28i–Fe4–N28j 90.97(18) 91.01(16) N8–Fe2–N11a 177.46(15) 178.23(14) N28i–Fe4–N28l 89.03(16) 88.99(14) N8–Fe2–N11c 88.24(15) 88.92(13) N28i–Fe4–N28n 180.00(23) 180.00(20) N8–Fe2–N8d 89.54(16) 89.40(14) N28i–Fe4–N28p 89.03(18) 88.99(16) N8–Fe2–N11f 91.67(15) 91.12(13) N28j–Fe4–N28l 89.03(15) 88.99(13) N8–Fe2–N8g 89.54(16) 89.40(14) N28j–Fe4–N28n 89.03(18) 88.99(16) N11a–Fe2–N11c 90.60(15) 90.58(14) N28j–Fe4–N28p 180.00(25) 180.00(21) N11a–Fe2–N8d 91.67(15) 91.12(13) N28l–Fe4–N28n 90.97(16) 91.01(14) N11a–Fe2–N11f 90.60(15) 90.58(13) N28l–Fe4–N28p 90.97(15) 91.01(13) N11a–Fe2–N8g 88.24(15) 88.92(13) N28n–Fe4–N28p 90.97(18) 91.01(16) N11c–Fe2–N8d 177.46(17) 178.23(15)
N11c–Fe2–N11f 90.60(16) 90.58(14) N11c–Fe2–N8g 91.67(16) 91.12(14) N8d–Fe2–N11f 88.24(16) 88.92(14) N8d–Fe2–N8g 89.54(16) 89.40(14) N11f–Fe2–N8g 177.46(13) 178.23(12)
Similarly to the hexafluorophosphate analogue (see above), [Fe(btzx)3](CF3SO3)2·CH3CN possesses solvent molecules trapped in the cages formed by the three bridging btzx ligands.
Each cavity contains one solvent molecule, and all these acetonitrile molecules are orientated in the same direction (head-to-tail, namely with the nitrogen atom alternatetingly pointing to Fe1 and Fe2 (see Figure 4.4). These acetonitrile molecules do show weak solvent-π interactions with the coordinated tetrazole rings, and are therefore not disordered in the cage, contrary to the methanol molecules in [Fe(btzx)3](PF6)2·CH3OH. TGA analyses for both compounds show that these solvent molecules are tightly trapped in the cages, as they cannot be released without the decomposition of the material.
Chain2: Fe3 and Fe4.
The chain2 is also constituted of two independent FeII centres, i.e. Fe3 and Fe4, which are alternatively connected by three btzx ligands (Figure 4.3 a). Similarly to chain1, the coordination sphere for both Fe3 and Fe4 is formed by six tetrazoles rings, but all tetrazoles are here crystallographically related. As a result, chain2 has a higher symmetry than chain1.
The metal–to–ligand bond lengths for Fe3 (Fe3–N21) and Fe4 (Fe4–N28) at 170 K are 2.178 (4) and 2.185(4) Å, respectively (Table 4.3). These bonds are slightly shorter than those observed for chain1, but they are still within the range for HS FeII systems.4 At 100 K, the metal–to–ligand bond distances decrease (the values observed are 6.7% lower) for Fe4, indicating a HS→LS transition (see Table 4.3).4 As for the Fe1 ions found in chain1, the metal–to–ligand bond lengths remains practically unchanged for Fe3 (Fe–Ntz = 2.182 Å), illustrative of a HS centre. At 170 K, the octahedral distortion for Fe4 is comparable to that observed for Fe1 and Fe2. Fe3 instead shows an almost perfect regular octahedral geometry (see Table 4.4), close to that observed for [Fe(btzx)3](PF6)2·CH3OH. Surprisingly, the low temperature structure (LS state) reveals slightly more distorted coordination spheres for both Fe3 and Fe4. Indeed, more symmetric, or less distorted, coordination spheres are expected for LS, which is the reverse for HS FeII centres.16
In chain2, six counterions “surround” each FeII centre, and each anion is shared by two adjacent metal ions. The counterions are equally arranged around Fe3 and Fe4, in contrast to the two different spatial arrangements observed for Fe1 and Fe2 (see Figure 4.3). The Fe3–
F11CF3 distance is 4.852 Å and the Fe4–O23SO3 distance is 5.369 Å. F11 also “interacts” with Fe2, while the O23 atom also interacts with Fe1; accordingly, the iron pairs Fe1/Fe4 and Fe2/Fe3 are sharing the same counterions (Figure 4.5). The distances do not vary significantly with the temperature for both metal centres forming chain2. This suggests that the variation observed for Fe1 (chain1) is due to a rearrangement of the spatial orientation of the counterion, and not to a displacement. Anion–tetrazole interactions are only observed for Fe3, with an anion···tetrazole (centroid) distance of 3.005 Å (Figure 4.5). Remarkably, Fe4 does not show this type of interactions which may be related to the solvent disposition in the cavities (see below).
As for chain1, chain2 also contains acetonitrile molecules in its cages formed by the three bridging btzx ligands. However, the acetonitriles are here arranged in a head-to-head
manner (see Figure 4.4). Consequently, the acetonitrile molecules are always pointing with their N41 atoms towards the Fe4 ions, interacting with the coordinated tetrazoles by means of solvent-π interactions. The N41–centroid distance is 3.275 Å, which is longer than those observed in the other chain. Fe4 possesses two acetonitrile molecules very close to the first coordination sphere. The resulting increased electron density at the metal centre may explain the lack of anion-π interactions for Fe4.
a) b)
Figure 4.5. a) Anion–π interactions occurring in chain1 and chain2. All tetrazoles involved in anion-π interactions are marked with its centroid. b) View of the packing along the c axis for [Fe(btzx)3](CF3SO3)2·CH3CN.
Table 4.4. HS/LS anion–to–metal distances, anion−π separations, Σ (distortion parameter reflecting the deformation of the octahedral coordination environment) and cone angles for [Fe(btzx)3](PF6)2·CH3OH (3), and for the four independent FeII centres of [Fe(btzx)3](CF3SO3)2·CH3CN (4).
Anion-metal distance
(Å)
Ttz-anion interaction.
(Å)
Σº Cone Angle b (º)
[Fe(btzx)3](PF6)2·CH3OH HS 5.166, 5.146 3.137 5.82 85.67 [Fe(btzx)3](PF6)2·CH3OH LS 5.098, 5.088 3.089 1.56 85.88 [Fe(btzx)3](CF3SO3)2·CH3CN(170 K) a
Fe1 HS 5.135, 4.384 2.998, 3.385 11.73 81.58/86.3 Fe2 HS 4.905, 4.896 2.968, 3.214 21.47 83.65/86.16
Fe3 HS 4.852 3.086 1.92 83.13
Fe4 HS 5.369 none 11.64 87.39
[Fe(btzx)3](CF3SO3)2·CH3CN(100 K) a
Fe1 HS 5.063, 4.319 2.944, 3.32 10.89 81.55/86.23 Fe2 LS 4.867, 4.898 2.961 3.190 10.14 84.41/87.12
Fe3 HS 4.815 3.005 3.24 82.77
Fe4 LS 5.343 none 12.12 88.43
a [Fe(btzx)3](CF3SO3)2 has four crystallographically independent centres which are shown separately.
b The cone angle is the angle formed by the centroids of two tetrazoles of the same side of the plain formed by the iron(II) atoms.
Crystal Packing.
The crystal packing of [Fe(btzx)3](CF3SO3)2·CH3CN is similar to those previously observed for other bistetrazole-based linear polymers (see Chapter 5). In Figure 4.5, a chain1:chain2 ratio of 2:1 is clearly evidenced. Chain2 is surrounded by six units of chain1, and each chain1 is shared by three units of chain2. The asymmetry of the chains and the spatial position of the counterions lead to a less symmetrical crystal packing, by comparison with the one of compound 3. The different tilting of the tetrazole rings, and the disparate spatial positions of the counterions are clearly visible in Figure 4.5 (b). The alignment of the polymeric chains generates planes with the iron(II) ions and the triflate counterions, which are separated by the cages formed by the btzx ligands (Figure 4.6).
a) b)
Figure 4.6. a) View along the a axis of the layers formed by the polymeric chains of compound 4 (only the carbon and the sulfur atoms of the counterions are shown for clarity).
b) View of the position of the sulfur atoms in the two planes. Plane1 (six sulfur atoms adopt a chair-like conformation, while the carbon atoms are almost all in the plane) is shown above and plane2 (six sulfur atoms are all in-plane, and the six carbon atoms are in a chair-like disposition ) is shown below. Figures are taken from the structure measured at 170 K.
The plane1 incorporates the Fe1 and the Fe4 atoms, while plane2 contains the Fe2 and the Fe3 atoms (Figure 4.6 a). These two planes present some differences, especially in the position of the counterion. As shown in Figure 4.6 b, in plane1, six sulfur atoms adopt a chair- like conformation, while the carbon atoms are almost all in the plane. In contrast, in plane2, the six sulfur atoms are all in-plane, and the six carbon atoms are in a chair-like disposition.
The atoms-to-plane distances are much longer for the sulfur atoms in plane1 than for the
carbon atoms in plane2, suggesting that plane2 is more compact than plane1. This difference is reflected by the cone-anglec between the metal centres forming the different planes. For Fe1 and Fe4, these angles are closer to the ideal value of 90º than for Fe2 and Fe3, indicating that the tetrazole-coordination angles are clearly influenced by the anions disposition in the plane.
Interestingly, these different dispositions of the counterions are not reflected in the distances between the metal centres of the two planes (see Table 4.4).
The solid-state structure of the complex 4 at 100 K presents an unusual crystal packing (Figure 4.7). As mentioned above, only one of the two crystallographically independent FeII centres of each chain undergoes the spin-transition. This particular situation results in the first example of a 1D chain exhibiting the unusual [HS-LS] pattern, typically observed for dinuclear entities.17 The […HS-LS-HS…] alternation observed in the chains does not produce full HS or LS planes. In fact, the spin-transition (ST) centres of chain1 share the same plane (plane2) with the non-ST centres of chain2, and vice versa (plane1). This arrangement in fact produces the peculiar packing depicted in Figure 4.7. Thus in plane1, each ST (Fe4) centre is surrounded by six non-ST (Fe1) centres, whereas in plane2, the non-ST (Fe3) ion is surrounded by six ST (Fe2) centres.
Figure 4.7. View of the HS-LS pattern observed for compound 4 at 100 K (LS = dark, HS = grey).
[Fe(btzx)3](ClO4)2·CH3CN (5)
The reaction conditions previously described were used to prepare [Fe(btzx)3](ClO4)2·CH3CN. In contrast to compounds [Fe(btzx)3](CF3SO3)2·CH3CN and
c The cone angle is the angle formed by the centroids of two tetrazoles of the same side of the plain formed by the iron(II) atoms.
[Fe(btzx)3](PF6)2·CH3OH, suitable single crystals of complex [Fe(btzx)3](ClO4)2·CH3CN could not be obtained, and thus the solid-state structure is not available. The IR and Elemental analyses suggest that [Fe(btzx)3](ClO4)2·CH3CN must be structurally related to the previous two compounds. Considering that all the structural asymmetry observed for [Fe(btzx)3](CF3SO3)2·CH3CN is due to the triflate anions, the structure of complex [Fe(btzx)3](ClO4)2·CH3CN is expected to be comparable to the one of complex [Fe(btzx)3](PF6)2·CH3OH. Unfortunately, no accurate structural information could be obtained from XRPD analysis, due to the poor crystallinity of the material. Nevertheless, the main structural features observed for the previous compounds most likely are also present in [Fe(btzx)3](ClO4)2·CH3CN. Thus, it is assumed that compound [Fe(btzx)3](ClO4)2·CH3CN is constituted of 1D polymeric chains which most likely pack in a similar manner as for compound [Fe(btzx)3](PF6)2·CH3OH.
a) b)
Figure 4.8. a) Plots of χmT vs. T for compounds 3 (PF6, full circles), 4 (CF3SO3, empty circles) and 5 (ClO4, squares). b) Plot of χmT vs. T for compound 5, showing the dependence on the sample preparation and the cooling rate. Crystalline sample (empty circles), amorphous sample (prepared by fast precipitation) measured upon cooling (full circles) at 10 K/min, and measured upon heating (full squares).
4.2 Bulk Magnetic Properties.
Magnetic susceptibility measurements were performed for the samples 3-5, in the temperature range 6–300 K, under an applied magnetic field of 1000 G (see Figure 4.8 (a)).
[Fe(btzx)3](PF6)2·CH3OH(3) shows a gradual spin transition centred at T1/2 = 160 K. At room temperature, χmT = 3.5 cm3 mol K–1. This value remains constant until 210 K, where the transition gradually sets in, extending over more than 100 K. χmT then remains constant at a value of 0.5 cm3 mol K–1 and finally decreases at very low temperatures, due to the zero-field splitting of remaining HS FeII centres. The residual high-spin fraction at low temperatures corresponds to 14% of the FeII atoms. [Fe(btzx)3](CF3SO3)2(CH3CN) (4) shows a similar type
of behaviour. The transition is centred at T1/2 = 110 K, and is less complete, compared to compound 3. The χmT value at room temperature is 3.5 cm3 mol K–1. At 146 K χmT starts to gradually decrease until 80 K, where a χmT value of 2.2 cm3 mol K–1 is reached. In this case, the transition corresponds to 40% of the iron(II) centres, indicating a 60% high-spin fraction.
For compound [Fe(btzx)3](ClO4)2·CH3CN (5), the transition is centred at T1/2 = 130 K, involving only 20% of the iron centres. At very low temperatures, χmT decreases as a result of the zero-field splitting of the remaining HS FeII ions. The completeness of the transitions for compound 5 is dependent on both, the cooling rate, and the quality of the sample.
The magnetic behaviours of compounds 3 and 4 are in agreement with their structural features. For compound 3, the remnant HS fraction at low temperatures can be assigned to defects in the structure, like the extremities of the polymers. In triazole-based compounds, the remnant HS fraction at low temperatures is used to estimate the length of the polymeric 1D chains.18 For instance, in compound [Fe(btzx)3](PF6)2·CH3OH, the FeII centres at the end of the polymeric chains are most likely coordinated by solvent molecules or non-bridging btzx ligands. As these ‘end-centres’ remain HS, and taking into account that coordinated btzx ligands would create a favourable crystal field to produce ST FeII centres (like the ‘in-chain’
ST centres), it is therefore expected that methanol molecules are completing the coordination spheres of the terminal FeII ions. For compound 4, the incompleteness of the transition is also in agreement with its structural features. As previously mentioned, only two out of four crystallographically independent FeII centres undergo the spin transition, while the other two remain HS. The transition temperature is 40 K lower than the one observed for 3, but the slope of the curve remains practically the same. Hence, the cooperativity seems not to be affected by the change of counterion.
Compound [Fe(btzx)3](ClO4)2·CH3CN (5) exhibits an incomplete transition with only 20% of the FeII changing spin state. As mentioned earlier, this value is to a certain extent dependent on the cooling rate,19 and on the sample preparation.20 Samples which have been obtained by rapid precipitation usually exhibit lower percentages of transition centres (full squares), while more crystalline samples (white circles) show slightly higher values (Figure 4.8 b). The cooling rate also influences this value, although its effect is smaller, compared to the sample preparation (amorphous vs. crystalline). Accordingly, if the temperature is kept at 120 K for about 22 hours, only a negligible decrease of the χmT value is detected. Hence, most of the incompleteness observed for compound 5 is independent of the sample preparation and/or the cooling rate. The incompleteness is apparently caused by an inherent structural impossibility for the compound to achieve a regular network in which all metallic centres are ST. A structural change occurring during the transition may also justify this incompleteness.
4.3 Light Induced Spin State Trapping (LIESST)
LIESST measurements have been performed on compounds 3 and 4. Compound 3 is not excited to the metastable HS state upon irradiation at 530 nm at 10 K. In contrast, compound 4 can be trapped at low temperatures in its HS metastable state. Indeed, the irradiation of 4 with a wavelength of 530 nm, at 10 K, for 12 hours results in an increase of the amount of
paramagnetic species. As seen in Figure 4.10, the photo-saturation, which is the equilibrium point between the relaxation and the excitation, is reached after 3 hours of irradiation, and no further increase is observed during the remaining 9 hours. The value reached after 3 hours suggests that the excitation to the metastable state is quantitative. Once the saturation is reached, the temperature is brought down to 6 K, and a diminution of χmT is observed due to the zero-field splitting effect of HS FeII centres. The temperature is then raised at a rate of 0.3 K min–1. At 19 K, χmT reaches its maximum value, i.e. 3.41 cm3 mol K-1, and then starts to decrease, owing to the relaxation of the HS metastable state back to the LS state. The critical T(LIESST) temperature is obtained through the plotting the derivative of χmT vs. T (inset Figure 4.10). As the temperature continues to increase, χmT follows the same line as the one observed during the thermal spin-transition.
a) b)
Figure 4.10. a) χmT vs. time plot showing the excitation of the LS state to the HS metastable state for compound 4. b) Plot of experimental χmT vs. T (squares) and temperature dependence of the product χmT after LIESST (circles), in the 6–300 K range. The inset shows the derivative of the product χmT with a T(LIESST) = 54 K.
4.4 DSC (Differential Scanning Calorimetry) Measurements.
Tetrazole-based spin-transition materials have low transition temperatures due to the crystal field splitting created by the ligands. Compounds 3–5 follow this trend, and their transition temperatures are close to 100 K. Due to these low transition temperatures and to equipment limitations, a DSC measurement was only carried out for compound 3. The heat capacity technique using a Quantum design PPMS was applied for all three compounds, but no peak was detected with the samples measured.
A crystalline sample of compound 3 has been used to determine the molar heat capacity, Cp, between 120 and 300 K, both in the cooling and heating modes (0.08 Ks-1). The spin- transition temperature found by DSC (position of the peak) differs substantially from the one
established by magnetic susceptibility. Although DSC is a dynamic measurement, the disparity observed between the two methods is too pronounced. However, broad peaks, as obtained for compound 3, are observed for gradual spin-transitions, and the peak cannot be assigned to any other structural phenomenon. Thus, the broad peak centred at 185 K (heating and cooling) is assigned to the spin-transition of compound 3.
To determine the excess heat capacity value (∆Cp) corresponding to the spin-crossover phenomenon, an estimation of a normal heat capacity curve with the high and low temperature data is required, which is then subtracted from the total heat capacity curve (see Figure 4.11). The entropy and enthalpy variations caused by the spin transition are then obtained by integrating the peak with respect to lnT and T, respectively. The ∆H = 1.91 K J mol–1 and ∆S = 17.47 J mol–1 K–1 values thus calculated for compound 3 are even lower than those reported in the literature for gradual transitions.21 The determined entropy value is superior to the value expected from the change in spin manifold of an FeII complex (R·ln 5 = 13.4 J mol–1 K–1). This excess arises from changes in internal vibrations caused by structural modifications occurring during the spin transition.5, 22 A larger vibrational entropy should be expected even for such a gradual transition.
Figure 4.11. Excess heat capacity due to the spin transition of[Fe(btzx)3](PF6)2·CH3OHin the cooling mode at scan rate of 0.08 Ks-1.
Calorimetric data for spin-transition compounds were reported for the first time by Sorai and Seki for the compound [Fe(phen)2(SCN)2].5 The large enthalpic and entropic values obtained were related to the cooperative nature of the spin transition. Consecutive calorimetric analyses for other spin-transition systems, both gradual and cooperative, corroborated this initial assumption.22 For compound 3, the values obtained are slightly lower than the range expected for gradual spin-transition systems. In these systems, the estimation of the normal heat capacity curve is intricate, as a result of the broadness of the peaks. Moreover, for compound 3, the transition temperature is close to the temperature limit of the DSC technique.
Consequently, the low temperature normal heat capacity curve is subjected to potential errors.
Both ∆H and ∆S are extremely sensitive to the calculation of the normal heat capacity curve, thus explaining the abnormal values obtained.
4.5 Discussion.
4.5.1 The case of [Fe(btzx)3](CF3SO3)2·(CH3CN) (4)
The four crystallographically independent iron(II) centres of compound [Fe(btzx)3](CF3SO3)2·(CH3CN) (4) possess very similar coordination environments, but disparate magnetic behaviours. Therefore, subtle differences in the coordination spheres of the metallic centres or, minor but key structural variations due to the crystal packing, may play a crucial role in the overall magnetic behaviour.
Magnetic susceptibility measurements reveal that for compound 4 the spin crossover is incomplete, with only half of the iron(II) centres undergoing the transition. The crystal structure analysis shows that the structure is formed by two crystallographically independent chains, namely chain1 and chain2, present in the crystal lattice in a 2:1 ratio. Each chain includes two independent FeII centres, namely Fe1 and Fe2 for chain1, and Fe3 and Fe4 for chain2. In each chain, only one of the two iron centres undergoes the spin transition (ST) (Fe2 and Fe4), while the other stays HS throughout the whole temperature range (Fe1 and Fe3).
The alternate distribution of the ST and HS centres results in the first 1D polymeric chain showing the typical [HS-LS] pattern observed for dinuclear species.23, 24 The major difference between [Fe(btzx)3](CF3SO3)2·(CH3CN) (4) and all bistetrazole-based 1D ST polymers, including [Fe(btzx)3](PF6)2·CH3OH (3), is the comparatively lower symmetry of its solid-state structure. The asymmetric nature of the CF3SO3– anions (in contrast to BF4– or PF6–), associated with their spatial disposition in the crystal lattice, are responsible for the overall asymmetry of the network. Indeed, many particular structural features of 4 can be related to its counterion. The existence of two distinct planes (plane1 containing the Fe1 and Fe4 ions, and plane2 the Fe2 and Fe3 centres) is obviously caused by the different disposition adopted by the counterions in each of these planes (Figure 4.6). The position of the triflates in the lattice also influences their association with the tetrazole moieties via anion-π interactions, and vice versa. As a result of these triflate···tetrazole contacts, the coordination angle of the tetrazole rings (cone angle) is affected. The proximity of the counterion to the metal centre sterically and electronically influences the ligand-field strength of the btzx ligands, and thus the magnetic properties of the correspondingly coordinated iron (see cone angles).
Furthermore, the orientation of the solvent molecules in the cavities is most likely determined by the counterion interactions with the tetrazole rings, resulting in the polarisation of the host cages. All these structural characteristics, induced by the triflate anions, apparently lead to a unique magnetic behaviour.
Dinuclear FeII complexes are known to be the most common molecular systems that exhibit the [HS-LS] pattern, as the one observed in the present polynuclear compound.23-25 Usually, this type of materials presents a two-step transition, due to the stabilisation of the intermediate [HS-LS] domain.26,27 Interestingly, the first two-step spin-transition phenomenon had been observed in a mononuclear entity, i.e. [Fe(pic)3]Cl2·EtOH (pic, 2-picolylamine).28
Subsequent studies have proven that this system forms 1D supramolecular chains through intermolecular interactions that stabilise the HS-LS pattern.29 Similarly, in compound 4, a HS- LS domain can be detected; however, in this case the metallic centres are part of a polymer generated by means of coordination bonds. It has been proposed that, in order to observe this supramolecular phenomenon, the enthalpic energy of the [HS-LS] pairs should be lower than half the sum of the enthalpic energies of the HS and LS states.26, 27 In addition, intermolecular interactions are required to stabilise the [HS-LS] domain formation.26 In compound 4, the intramolecular (intrachain) interactions that stabilise this […HS-LS-HS…] motif, apparently arise from the btzx-anion “pairs” which create sterical strain within the polymer chains. This structural constraint is reflected in the [tetrazole-centroid]–ND–Fe angles which are exceptionally small in comparison with other bistetrazole-based 1D polymers (see Chapter 5, section 5.3). In the case of the ST compound [Fe4L4](ClO4)8 (L = 4,6-bis(2’,2’’-bipyridyl-6’- yl)-2-phenyl-pyrimidine),30 the multistep character of the transition is assigned to the intramolecular cooperativity, or short range interactions, between the iron centres.
Comparatively, for compound 4, the ligand btzx is less rigid, but the close packing of the chains confers an extra rigidity to the overall network. 1D bistetrazole-based polymers usually experience a tilt of the coordinated tetrazole rings upon the transition ([tetrazole-centroid]–
ND–Fe angles, Chapter 5 section 5.3). In the present case, this change of angle is hindered for Fe1 and Fe3 due to a rigid packing, which generates intramolecular strain within the ligand, and/or intramolecular communication. The intermolecular interactions can only come from the triflate-tetrazole interactions occurring in the two planes formed by the pairs of FeII centres. In contrast to compound 4, [Fe(btzx)3](PF6)2 does not show a two-step spin crossover.
This absence of an intermediate step during the transition can only be attributed to the presence of only one crystallographically independent centre, or to weak intramolecular cooperativity. The intrachain strain in compound 4, ascribed to the close anion-to-metal contacts, is less strong in compound 3, due to the longer anion–metal separations. On the other hand in 3, the longer anion–metal separation results in a less distorted octahedron, and consequently in a different crystal field, which is reflected by a higher T½.31
4.5.2 The effect of the counterion
As mentioned above, the effect of the counterion on the magnetic properties of bistetrazole-based spin-transition materials had not yet been investigated. Indeed, the only cation reported8, 11 with two different anions, namely ClO4– and PF6–, is [Fe(btzb)3]2+. It is known that the crystallinity of the cationic structure of bistetrazole-based compounds is dependent on the anion used.10 The role of the anion in this type of materials is to act as a template for the generation of the cationic polymeric network. Thus, the size of the anion is crucial to obtain suitable single crystals of the material, as observed for [Fe(btzb)3]2+. Surprisingly, crystalline compounds of [Fe(btzx)3]2+with three different counterions have been obtained which all exhibit spin-transition properties.
An obvious effect of the counterion on the spin-transition properties concerns the remaining fraction of HS species at low temperatures, which is 14%, 50% and 80% for
[Fe(btzx)3](PF6)2(MeOH), [Fe(btzx)3](CF3SO3)2(CH3CN) and [Fe(btzx)3](ClO4)2(CH3CN), respectively. For [Fe(btzb)3](ClO4)2, the high fraction of iron centres remaining HS at low temperatures has been assigned to a phase transition.11 In the case of the [Fe(btzx)3]2+ cation, a plausible explanation is an irregular packing of the network, due to the inappropriate size of the anion, as mentioned in the magnetic susceptibility section. The structural changes associated with the spin transition may also restrict the number of FeII ions undergoing the transition, as proposed for [FeII(btzpol)1.8(btzpol-OBF3)1.2](BF4)0.8 in Chapter 3. In certain cases, extremely rigid networks apparently prohibits the occurrence of the transition.31 The spin-transition slopes for compounds containing the cation [Fe(btzx)3]2+, seem to indicate that this latter explanation is not plausible (weak cooperativity), unless the intramolecular cooperativity is compensating the inherent intermolecular interactions of its rigid network.30
The nature of the counterion also affects the temperature of the spin transition (T½).
Triazole-based materials have been intensively studied, and the influence of the different chemical “pieces” constituting the compound on the spin-transition properties has been thoroughly investigated.32-37 In the case of spherical counterions, the smallest ion leads to the highest transition temperature.38 The greater electrostatic pressure generated on the cation by the smaller anions, stabilises the LS state over the HS state.38 In the present case, the smaller perchlorate anion (compared to PF6–) has shifted the transition to lower temperatures stabilising the HS state, thus acting in the opposite sense. This behaviour is similar to that observed by Long et al. and Tuchagues and co-workers,{Long, 2004 #41; Yamada, 2003
#392} in which steric interactions involving groups close to the metal centre stabilise the HS state (shift T½ towards lower temperatures). In some cases, these steric constraints may even annihilate the spin crossover.31, 39 Bistetrazole-based 1D polymers usually exhibit comparable crystal packings in which the counterions are situated around the FeII centres. The size and shape of the counterion will determine the anion-to-metal distance. Thus, for a certain cation, the use of smaller anions will result in shorter anion-to-metal distances, and will sterically limit the ‘tilting freedom’ of the tetrazole rings. As a result, sterically hindered tetrazole rings give rise to low transition temperatures. Ultimately, these steric effects may lead to the disappearance of the spin transition. This is observed for compound 4, where the non-ST centres Fe1 and Fe3 exhibit shorter anion-to-metal distances, compared to the ST centres Fe2 and Fe4. Similarly, one can conceive that the high degree of remnant HS fraction present in compound 5 is caused by the proximity of the anion to those metal centres. This tendency can be related to the trend previously observed for the Σ values, where high Σ are associated with weaker crystal fields and therefore stabilisation of the HS state.40 In other SCO materials long Fe–N distances have also been a sign of HS stable states (lower transition temperatures).39 In the case of [Fe(btzx)3](CF3SO3)2(CH3CN), neither the Fe–N distances of the different metal centres, nor their Σ have any relation with the energy of the HS state. Based on the structural data available in the literature on bistetrazole spin-transition compounds, a correlation between the anion-to-metal distance and the transition temperature is suggested (Figure 4.12).
As mentioned earlier, most of the structural features of this compound 4 are related to the position of the anions in the crystal lattice (two planes observed, cone angle, [centroid- tetrazole]–ND–Fe angle, etc.). Long anion-to-metal distances are associated to small changes in the distortion of the octahedron during the spin transition (Figure 4.12).
Figure 4.12. a) T½ (temperature at which 50% of the molecular complex is in the HS state) vs.
HS anion-to-metal distance. b) Octahedral distortion vs. HS anion-to-metal distance. See Table 4.5 for number/compound assignment.
Interestingly, the cooperative nature of the spin transition has not been modified by the change of counterion. Indeed, both the triflate and the hexafluorophosphate derivatives show similar slopes. The slightly less steep slope for the perchlorate compound is probably caused by the lower crystallinity of the material, which is known to affect the properties of the transition.4, 20 This insignificant effect of the anions suggest that the cooperative nature of this cationic network arises from the intramolecular communication along the polymeric chains.
Based on the solid-state dilution studies that proved that the cooperativity is influenced by the number of ST centres in the material,41 the alternating distribution of the ST centres along the chains most likely affects the cooperativity within the system. Therefore, a more cooperative behaviour is expected for [Fe(btzx)3](PF6)2(CH3OH), for which all FeII centres are ST, than for [Fe(btzx)3](CF3SO3)2(CH3CN), which displays an alternating distribution of HS and ST centres. This is obviously not the case, as revealed by the similar transition slopes obtained from magnetic susceptibility measurements for the two compounds. As proposed earlier, [Fe(btzx)3](PF6)2(CH3OH) may suffer from an internal constraint in the form of short-range intramolecular interactions which counterbalance the long range cooperative interactions (see Chapter 9 conclusions).30
Table 4.5. Transition temperatures (T½), references for bistetrazole-based spin-transition compounds and numbers assigned to each compound (Figure 4.12).
Compounds T½ (K) Reference Number for
Compound
[Fe(endi)3](BF4)2 140 9 1
[Fe(btzp)3](ClO4)2 130 13 2
[Fe(btzmp)2µ-(btzmp)2](ClO4)2 135 Chapter 3 3
[Fe(btzb)3](PF6)2 173 8 4
[Fe(btzx)3](PF6)2(CH3OH) 150 Chapter 4 5
[Fe(btzx)3](CF3SO3)2(CH3CN) 110 Chapter 4 6
[Fe(btzx)3](PF6)2(CH3OH) and [Fe(btzx)3](CF3SO3)2·(CH3CN) do differ in their LIESST properties. Only for [Fe(btzx)3](CF3SO3)2·(CH3CN) the irradiation leads to the trapping of the high-spin metastable state. The χmT value for the LIESST-generated HS metastable state of [Fe(btzx)3](CF3SO3)2 corresponds to the excitation of two iron(II) centres, namely Fe2 and Fe4. This metastable state relaxes back to the LS states with a T(LIESST) of 54 K. The stability of this metastable state depends on the cooperativity, as well as on the transition temperature.4, 42 High T½ are associated with lower energy barriers and faster relaxation to the LS state. For compound 3, the T½ value is low enough and thus cannot explain the absence of a HS metastable state. In a recent work the distortion of the coordination sphere, ∆Θ,d has been related to the T(LIESST) temperature.43 High ∆Θ values result in higher T(LIESST) temperatures. The ∆Σ values corresponding to complexes 3 and 4 allow to compare the respective octahedral distortions.16 In this case, the octahedral distortion for compound 3 is much smaller than observed for the atoms Fe2 and Fe4 of compound 4.
These distinct geometric distortions may be responsible for the different photomagnetic behaviours.
4.6 Conclusions.
The series of compounds [Fe(btzx)3](PF6)2(CH3OH), [Fe(btzx)3](CF3SO3)2(CH3CN) and [Fe(btzx)3](ClO4)2(CH3CN) has shown that the template-role of the counterions is determining the spin-transition properties. Short anion-to-metal distances lead to significant distortions of the octahedral coordination environment of the iron(II) centres, resulting in lower temperatures for the spin transition. For the [Fe(btzx)3]2+ cation several structural features, i.e. ∆Σ, cone angle and [tetrazole-centroid]–ND–Fe angle, reflect the steric hindrance around the FeII centre. These features are all associated with the size and shape of the counterions, as well as their position in the lattice. The HS fraction remaining at low temperatures is also linked to the distortion of the octahedron. Apparently, for a certain degree of distortion the FeII ion remains in HS state through the whole temperature range. This is observed for the triflate derivative, whose two HS centres show with the highest distortion of the octahedron. The cooperative nature of the cationic species barely depends on the nature of the anion used, suggesting that the nature of the cooperativity is intrapolymeric. This observation is in contradiction with the fact that the alternating distribution of the ST centres in the 1D chains (…HS-LS-HS…) of the triflate derivative does not affect the cooperative nature of the ([Fe(btzx)3]2+)-basedmaterial. [Fe(btzx)3](PF6)2 and [Fe(btzx)3](CF3SO3)2 also differ in their LIESST properties. Only the HS metastable state of the complex [Fe(btzx)3](CF3SO3)2 can be trapped under light irradiation. The origin of this disparity is not clearly identified, but the greater distortion of the octahedral environment around the FeII centres of compound [Fe(btzx)3](CF3SO3)2 appears to be critical. Further structural analysis, including optical and pressure dependent studies, are necessary for a better understanding of this cationic system.
d θ corresponds to the sum of the deviations from 60º of the 24 possible θ angles.
4.7 References
1. Létard, J. F.; Guionneau, P.; Goux-Capes, L., Towards spin crossover applications. In Spin Crossover in Transition Metal Compounds III, Topics in Current Chemistry, Gütlich, P.; Goodwin, H. A., Eds. Springer: 2004; Vol. 235, pp 221-249.
2. Kahn, O.; Kröber, J.; Jay, C., Adv. Mater. 1992, 4, 718-728.
3. Kahn, O.; Martinez, C. J., Science 1998, 279, 44-48.
4. Gütlich, P.; Hauser, A.; Spiering, H., Angew. Chem. Int. Ed. Engl. 1994, 33, 2024.
5. Sorai, M.; Seki, S., J. Phys. Chem. Solids 1974, 35, 555-570.
6. van Koningsbruggen, P. J., Special classes of iron(II) azole spin crossover compounds.
In Spin Crossover in Transition Metal Compounds I, Topics in Current Chemistry, Gütlich, P.; Goodwin, H. A., Eds. Springer: 2004; Vol. 233, pp 123-149.
7. van Koningsbruggen, P. J.; Grunert, M.; Weinberger, P., Mon. Chem. 2003, 134, 183- 198.
8. Grunert, C. M.; Schweifer, J.; Weinberger, P.; Linert, W.; Mereiter, K.; Hilscher, G.;
Müller, M.; Wiesinger, G.; van Koningsbruggen, P. J., Inorg. Chem. 2004, 43, 155- 165.
9. Schweifer, J.; Weinberger, P.; Mereiter, K.; Boca, M.; Reichl, C.; Wiesinger, G.;
Hilscher, G.; van Koningsbruggen, P. J.; Kooijman, H.; Grunert, M.; Linert, W., Inorg. Chim. Acta 2002, 339, 297-306.
10. Absmeier, A.; Bartel, M.; Carbonera, C.; Jameson, G. N. L.; Weinberger, P.; Caneschi, A.; Mereiter, K.; Létard, J. F.; Linert, W., Chem.-Eur. J. 2006, 12, 2235-2243.
11. van Koningsbruggen, P. J.; Garcia, Y.; Kooijman, H.; Spek, A. L.; Haasnoot, J. G.;
Kahn, O.; Linarès, J.; Codjovi, E.; Varret, F., J. Chem. Soc. Dalton Trans. 2001, 466- 471.
12. Bronisz, R., Inorg. Chim. Acta 2004, 357, 396-404.
13. van Koningsbruggen, P. J.; Garcia, Y.; Kahn, O.; Fournès, L.; Kooijman, H.; Spek, A.
L.; Haasnoot, J. G.; Moscovici, J.; Provost, K.; Michalowicz, A.; Renz, F.; Gütlich, P., Inorg. Chem. 2000, 39, 1891-1900.
14. Alvarez, S., J. Am. Chem. Soc. 2003, 125, 6795-6802.
15. van Koningsbruggen, P. J.; Garcia, Y.; Bravic, G.; Chasseau, D.; Kahn, O., Inorg.
Chim. Acta 2001, 326, 101-105.
16. Guionneau, P.; Marchivie, M.; Bravic, G.; Létard, J. F.; Chasseau, D., Structural aspects of spin crossover. Example of the (FeLn)-L-II(NCS)(2) complexes. In Spin Crossover in Transition Metal Compounds II, Gütlich, P.; Goodwin, H. A., Eds. 2004;
Vol. 234, pp 97-128.
17. Gaspar, A. B.; Muñoz, M. C.; Real, J. A., J. Mater. Chem. 2006, 16, 2522-2533.
18. Roubeau, O. Solid State Properties of Triazole-Based Fe(II) Materials. PhD thesis, Leiden University, Leiden, 2002.
19. Yamada, M.; Hagiwara, H.; Torigoe, H.; Matsumoto, N.; Kojima, M.; Dahan, F.;
Tuchagues, J. P.; Re, N.; Iijima, S., Chem.-Eur. J. 2006, 12, 4536-4549.
20. Haddad, M. S.; Federer, W. D.; Lynch, M. W.; Hendrickson, D. N., Inorg. Chem.
1981, 20, 131-139.
21. Hauser, A., J. Chem. Phys. 1991, 94, 2741-2748.
22. Sorai, M., Heat capacity studies of spin crossover systems. In Spin Crossover in Transition Metal Compounds III, Topics in Current Chemistry, Gütlich, P.; Goodwin, H. A., Eds. Springer: 2004; Vol. 235, pp 153-170.
23. Real, J. A.; Bolvin, H.; Bousseksou, A.; Dworkin, A.; Kahn, O.; Varret, F.;
Zarembowitch, J., J. Am. Chem. Soc. 1992, 114, 4650-4658.
24. Real, J. A.; Gaspar, A. B.; Muñoz, M. C.; Gütlich, P.; Ksenofontov, V.; Spiering, H., Bipyrimidine-bridged dinuclear iron(II) spin crossover compounds. In Spin Crossover in Transition Metal Compounds I, Topics in Current Chemistry, Gütlich, P.; Goodwin, H. A., Eds. Springer: 2004; Vol. 233, pp 167-193.
25. Ortega-Villar, N.; Thompson, A. L.; Muñoz, M. C.; Ugalde-Saldivar, V. M.; Goeta, A.
E.; Moreno-Esparza, R.; Real, J. A., Chem.-Eur. J. 2005, 11, 5721-5734.
26. Ksenofontov, V.; Gaspar, A. B.; Niel, V.; Reiman, S.; Real, J. A.; Gütlich, P., Chem.- Eur. J. 2004, 10, 1291-1298.
27. Zein, S.; Borshch, S. A., J. Am. Chem. Soc. 2005, 127, 16197-16201.
28. Sorai, M.; Ensling, J.; Gütlich, P., Chem. Phys. 1976, 18, 199-209.
29. Chernyshov, D.; Hostettler, M.; Tornroos, K. W.; Burgi, H. B., Angew. Chem. Int. Ed.
2003, 42, 3825-3830.
30. Breuning, E.; Ruben, M.; Lehn, J. M.; Renz, F.; Garcia, Y.; Ksenofontov, V.; Gutlich, P.; Wegelius, E.; Rissanen, K., Angew. Chem. Int. Ed. 2000, 39, 2504-2507.
31. Reger, D. L.; Gardinier, J. R.; Smith, M. D.; Shahin, A. M.; Long, G. J.; Rebbouh, L.;
Grandjean, F., Inorg. Chem. 2005, 44, 1852-1866.
32. Garcia, Y.; van Koningsbruggen, P. J.; Lapouyade, R.; Fournès, L.; Rabardel, L.;
Kahn, O.; Ksenofontov, V.; Levchenko, G.; Gütlich, P., Chem. Mater. 1998, 10, 2426- 2433.
33. Haasnoot, J. G., Coord. Chem. Rev. 2000, 200, 131-185.
34. Jay, C.; Grolière, F.; Kahn, O.; Kröber, J., Mol. Cryst. Liq. Cryst. Sci. Technol. Sect.
A-Mol. Cryst. Liq. Cryst. 1993, 234, 255-262.
35. Kahn, O.; Codjovi, E., Philos. Trans. R. Soc. Lond. Ser. A-Math. Phys. Eng. Sci. 1996, 354, 359-379.
36. Lavrenova, L. G.; Ikorskii, V. N.; Varnek, V. A.; Oglezneva, I. M.; Larionov, S. V., J.
Struct. Chem. 1993, 34, 960-965.
37. Lavrenova, L. G.; Yudina, N. G.; Ikorskii, V. N.; Varnek, V. A.; Oglezneva, I. M.;
Larionov, S. V., Polyhedron 1995, 14, 1333-1337.
38. Roubeau, O.; Gomez, J. M. A.; Balskus, E.; Kolnaar, J. J. A.; Haasnoot, J. G.; Reedijk, J., New J. Chem. 2001, 25, 144-150.
39. Long, G. J.; Grandjean, F.; Reger, D. L., Spin crossover in pyrazolylborate and pyrazolylmethane complexes. In Spin Crossover in Transition Metal Compounds I, Topics in Current Chemistry, Gütlich, P.; Goodwin, H. A., Eds. 2004; Vol. 233, pp 91- 122.
40. Guionneau, P.; Brigouleix, C.; Barrans, Y.; Goeta, A. E.; Letard, J. F.; Howard, J. A.
K.; Gaultier, J.; Chasseau, D., C. R. Acad. Sci. Ser. IIc 2001, 4, 161-171.
41. Hauser, A.; Gütlich, P.; Spiering, H., Inorg. Chem. 1986, 25, 4245-4248.
42. Hauser, A., Light-induced spin crossover and the high-spin -> low-spin relaxation. In Spin Crossover in Transition Metal Compounds II, Topics in Current Chemistry, Gütlich, P.; Goodwin, H. A., Eds. Springer: 2004; Vol. 234, pp 155-198.
43. Marchivie, M.; Guionneau, P.; Létard, J. F.; Chasseau, D., Acta Crystallogr., Sect. B:
Struct. Sci. 2005, 61, 25-28.
