Quesada Vilar, Manuel
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).
3
Influence of substituted alkyl-based
spacers on the formation of bistetrazole
Fe(II) spin-transition polymers
Abstract:
Two new bistetrazole-based ligands, btzmp (1,2-bis(tetrazol-1-yl)-2-methylpropane) and btzpol (1,3-bis(tetrazol-1-yl)-2-propanol) generate unusual polymeric species when reacted with an iron(II) salt. In both cases, the triply-bridged 1D polymer, typically observed with certain bistetrazole ligands, is not attained. Instead, one of the ligands is perpendicular to the direction of the main chain, yielding a 2D network in the case of btzpol ([FeII(btzpol)1.8(btzpol-OBF3)1.2](BF4)0.8), and a 1D polymer exhibiting monocoordinated ligands for btzmp ([FeII(µ-btzmp)2(btzmp)2](ClO4)2). The influence of these peculiar spatial arrangements of the coordinated ligands on the spin-transition properties of the corresponding compounds is discussed. In addition, the LIESST properties of [FeII(btzpol)1.8(btzpol- OBF3)1.2](BF4)0.8 are studied, and the relaxation process of its metastable state is analyzed.
Parts of this chapter have appeared in the literature: Manuel Quesada, Ferry Prins, Olivier Roubeau, Patrick Gamez, Simon J. Teat, Petra J. van Koningsbruggen, Jaap G. Haasnoot, Jan Reedijk., Inorg. Chim. Acta (2007), doi:10.1016/j.ica.2006.12.010
3.1 Introduction.
The search for materials with predefined and tunable properties has for long been appealing for scientists. In the 1970’s, Schmidt noticed that the physical properties of a crystalline solid are both dependent on the properties of the individual molecular components, and on the distribution of these molecules in the crystal lattice.1 This search for a control over the “overall structure” was then further explored by Wells with the “node and spacer”
approach, defining the inorganic networks in terms of their topology,2 and was brought to the field of coordination polymers by Robson.3-5 The porosity of these coordination polymers was primarily exploited in the field of inclusion chemistry, motivated by applications in storage media and in catalysis.6 Different types of linkers and coordination geometries at the metal centers are nowadays used to rationally achieve predetermined structural properties such as the size and shape of the cavities.7 The increasing interest and development of the field8 has brought up a more precise definition of this discipline, for which the term “reticular chemistry” or “reticular synthesis” has been introduced.9 This use of rigid molecular building blocks connected with rigid linkers is emphasised to stress the difference with related areas of research, such as supramolecular chemistry. New challenges within this field involve the design of reticular networks, which now exhibit one or more functional properties.9 These hybrid inorganic-organic materials have already opened promising routes towards the preparation of active catalysts, enantiomeric separation, the synthesis of bistable systems, or in solid-state synthesis.10-16, 17
The incorporation of the spin-transition property in reticular networks has also been investigated.12, 18 The targeted search for rigid polymeric structures that would bring stability to the material and enhance the desired cooperativity has been a subject of investigation, but the development of this type of compounds is still limited.19-21 4-Substituted 1,2,4-triazole- based ligands have extensively been used to coordinate FeII centers and produce 1D polymers displaying cooperative SCO behaviours.22-30 4,4'-bis-1,2,4-triazole (btr) yielded 2D ([Fe(btr)2(NCX)2], X = S,31, 32 or Se33) and 3D ([Fe(btr)3](ClO4)2) networks34 exhibiting spin- crossover properties, in which the bistriazole acts as a rigid linker between the iron centres.
Iron(II) spin-crossover materials having rigid 1D, 2D or 3D structures have also been synthesised with cyanometallate-based linkers.35-38
Bistetrazoles are also well-known ligands in this field and generate polymeric species with a wide variety of geometries and conformations.39, 40 The size and conformation of the spacer affects the dimensionality of the resulting polymer. Short spacers such as btzp (1,2- bis(tetrazol-1-yl)propane) and endi (1,2-bis(tetrazol-1-yl)ethane) lead to 1D polymers,39, 41 while btzb (1,4-bis(tetrazol-1-yl)butane) and longer spacers give rise to 3D networks.42, 43 Endi and btzp possess the same spacer length (same number of carbons between the two azole rings), but btzp holds one methyl group attached to one of the carbon atoms of the ethane bridge. Due to the columnar disposition of the 1D polymer obtained with btzp, the methyl groups of different chains lie close to each other (sterical interaction).39 In an attempt to increase this beneficial interaction (cooperativity), two strategies are followed which are discussed in two different parts in the present chapter. In Part I, the new ligand btzmp (1,2- bis(tetrazol-1-yl)-2-methylpropane, Figure 3.1), which represents the next logical step in the
endi–btzp series, is presented. Indeed, two methyl substituents are now included in the ethane bridge (Figure 3.1 c). The effect of this additional methyl substituent on the structural arrangement of the complex [Fe(µ-btzmp)2(btzmp)2](ClO4)2 (1) is studied by means of single crystal X-ray diffraction. Magnetic susceptibility measurements and 57Fe Mössbauer spectroscopy are used to investigate the spin-transition properties of the material. In Part II, the ligand btzpol (1,3-bis(tetrazol-1-yl)-2-propanol) (Figure 3.1 d) is reported, which contains a three-carbon atom spacer and an alcohol function. The hydroxyl group is thus expected to be involved in hydrogen bonding interactions within the coordination network. The solid-state structure, spin-transition properties, LIESST behaviour and relaxation processes of the HS metastable state of the resulting complex, [FeII(btzpol)1.8(btzpol-OBF3)1.2](BF4)0.8
(H2O)0.8(CH3CN) (2), are presented. The unusual, partial reaction of the alcohol group of the ligand with BF4– anions, and its consequences on the structural and/or physical properties of the ensuing material are also explained.a
Figure 3.1. Short-spacer ligands used to prepare the coordination compounds discussed in this chapter: a) endi = 1,2-bis(tetrazol-1-yl)ethane; b) btzp = 1,2-bis(tetrazol-1-yl)propane; c) btzmp =1,2-bis(tetrazol-1-yl)-2-methylpropane; d) btzpol = 1,3-bis(tetrazol-1-yl)-2-propanol.
Part I: [Fe(µ-btzmp)2(btzmp)2](ClO4)2
3.2 Crystal Structure of the free btzmp ligand.
The ligand btzmp crystallises in methanol as colourless blocks (space group P21/c). The solid-state conformation of btzmp is depicted in Figure 3.2, together with a Newman Projection and the crystal packing. The methyl substituents favour a gauche conformation. By contrast, a trans conformation is observed for the free endi ligand,41 which does not have hindering methyl groups (see Figures 3.1 a and c). In addition, the bulkiness of the methyls tilts the tetrazole planes with an angle of 57.28 º with respect to each other. These sterical constraints result in the particular crystal packing observed in Figure 3.3, in which the atoms H4 and H1 of the tetrazole rings are involved in hydrogen bonds (H4···N4 = 2.381 Å and H1···N8 = 2.543 Å) with tetrazole units of adjacent molecules. Contrary to the tetrazole rings of the free endi ligand,41 the heteroaromatic rings of btzmp are not π–π stacked.
a See chapter 2 for details on the experimental procedures and recipes for all compounds presented in this
a) b)
Figure 3.2. a) View of the Newman Projection along the C2–C3 bond (top) and view of the solid-state conformation (bottom) of btzmp. b) Crystal packing of the ligand btzmp along the a axis.
Table 3.1. Crystallographic data for complex 1. When the parameters depend on the temperature, they are entered in the order 200 K and 100 K.
[Fe(µ-btzmp)2(btzmp)2](ClO4)2
Formula C24H40Cl2FeN32O8
FW/g mol-1 1031.63 Crystal system triclinic Space group P-1
a/Å 8.5157(15); 8.3039(15) b/ Å 11.1223(18);10.9503(18)
c/ Å 12.432(2);12.311(2) α/º 79.24(2);79.38(2) β/º 83.69(3);78.96(2)
γ/º 73.65(4);74.62(4) V/Å3 1102.8(4);1053.8(4)
Z 2
ρcalcd/g cm-3 1.5534(6);1.6256(6) T/K 200(K); 100(K) Crystal shape hexagon
Colour white
3.3 Crystal Structure of [Fe(µ-btzmp)2(btzmp)2](ClO4)2 (1)
Compound 1 crystallises in the P-1 space group at 100 K and 200 K (Table 3.1). The iron ion is located on a crystallographic inversion centre. The unit cell is composed of an FeII ion, two perchlorate anions and four btzmp ligands. One btzmp is acting as a bridging ligand while the other is monocoordinated to an iron ion. At 200 K, the perchlorate anion is disordered in two positions (0.542(4):0.458(4)), while at 100 K no disorder is observed. The coordination sphere is constituted of 6 tetrazole rings coordinated through their ND atom,b forming a slightly distorted octahedral environment (see Figure 3.3). As expected,44 the coordination sphere of the HS state is more distorted than the LS one (Table 3.2). To evaluate the octahedral distortion,44 the parameter Σ is considered.c Thus, the geometry of the HS state of 1 shows the largest octahedral distortion so far observed for tetrazole-based compounds (23.64º, see Chapter 5, Table 5.1). In its LS state, 1 exhibits a much smaller octahedral distortion (15.6º). The Fe–N distances at 200 K are 2.207(4) ((Fe–N18), 2.181(4) (Fe–N1) and 2.171(3) Å (Fe–N8), falling in the expected range for a HS FeII centre (see Table 3.2).45 It is noticeable that the Fe–N distances for monocoordinated ligands are approximately 0.02 Å longer, compared to those of bridging tetrazole ligands. At 100 K, the Fe–N distances decrease by a factor of 8-9 %. The observed values of 2.002(3) (Fe-N18), 2.001(3) (Fe-N1) and 1.986(3) Å (Fe-N8) are thus normal for LS FeII centres (see Table 3.2).45 The change in the metal-to- ligand bond length affects the angles of the coordinated tetrazole rings. A considerable modification of the [tetrazole-centroid]–ND–Fe angle is occurring upon the spin transition.
This angle characterises the tilting of the whole tetrazole ring with respect to the FeII metal ion. The increase of this angle in the low-spin state thus suggests that the tetrazole rings tend to open-up in order to compensate the decrease of the Fe–N distances (see Figure 3.4).
Figure 3.3. Labeled ORTEP representation of [Fe(µ-btzmp)2(btzmp)2](ClO4)2 (1).
b N stands for “donor atom”, and represents the atom coordinated to the metal ion.
a) b)
Figure 3.4. [Tetrazole-centroid]–ND–Fe angle for the bridging btzmp ligands of compound 1 (a) in the LS (170.36º) and (b) HS state (163.82º). Non-bridging btzmp ligands and counterions have been omitted for clarity.
Surprisingly, only two of the btzmp ligands extend the structure along the a axis, while the other two only act as monodentate ligands. This coordination mode contrasts with those observed for other bistetrazole-based 1D polymers, in which the iron(II) centres are triply bridged, and present no monocoordinated azole ligands.46 The presence of a double bridge, instead of the typical triple bridge, affects the intrapolymeric distance (namely the Fe–Fe separation in the polymeric chain), which is 8.516 Å at 200 K. This distance is about 1 Å longer those found for the related [Fe(endi)3](BF4)2 and [Fe(btzp)3](ClO4)2 1D polymers (see Discussion). The change in spin state results in a decrease of 0.212 Å of the Fe···Fe intrapolymeric distance. Additionally, these monocoordinated ligands have an effect on the crystal packing of [Fe(µ-btzmp)2(btzmp)2](ClO4)2, which does not show the typical hexagonal motif usually observed for bistetrazole-based 1D polymers (Chapter 5, section 5.3).40 In 1, each polymeric chain is surrounded by six other chains with the anions occupying the ensuing voids (see Figure 3.5). The interpolymeric distances are 11.122 Å and 12.432 Å in the HS state and 10.950 Å and 12.311 Å in the LS state. No solvent molecules are present in the crystal lattice, which is a characteristic feature of such 1D bistetrazole spin-transition materials (see Chapter 5).
3.4 Magnetic Properties
Compound 1 crystallises as colourless small single crystals. For [tetrazole-FeII] materials, white/colourless crystals generally indicate a HS state at room temperature. Indeed, as shown in Figure 3.5, at 300 K χmT = 3.11 cm3 mol-1 K-1, implying that the compound is mainly in its HS state. At T½ = 133 K, the LS state is abruptly populated. The transition takes place for 98% (χmT = 0.073 cm3 mol-1 K-1) of the FeII centres, 2% remaining HS through the whole temperature range. A drastic colour change, from colourless to purple, accompanies the HS→LS transition. As the temperature is increased, a hysteresis loop with a width of about 3 K (T1/2↑= 136 K and T1/2↓= 133 K) is detected, as observed in the magnetic susceptibility data (Figure 3.5 b). The process shows no fatigue, and the hysteresis is reproducible over several
cooling/heating cycles (5 cycles). In case of less crystalline batches of [Fe(btzmp)2µ- (btzmp)2](ClO4)2, the hysteresis loop is not observed.
Table 3.2. Selected interatomic distances [Å] and angles [º] for (1). a = -1+x,y,z; c = -x,-y,2- z; d = 1-x,-y,2-z.
a) b)
Figure 3.5. a) View along the a axis of the crystal packing of [Fe(µ-btzmp)2(btzmp)2](ClO4)2. b) Plot of χmT vs. T for a crystalline sample of [Fe(btzmp)2µ-(btzmp)2](ClO4)2 (1), recorded both in the cooling mode (black squares) and in the heating mode (white circles).
[Fe(µ-btzmp)2(btzmp)2](ClO4)2 200 K 100 K Fe1–N1 2.1814(35) 2.0005(29) Fe1–N18 2.2068(36) 2.0021(25)
Fe1–N8a 2.1707(34) 1.9864(30) N1–Fe1–N18 92.65(14) 91.44(12)
N1–Fe1–N8a 87.04(13) 88.12(12) N1–Fe1–N1c 180.00 180.00 N1–Fe1–N18c 87.35(14) 88.56(12) N1–Fe1–N8d 92.96(13) 91.88(12) N18–Fe1–N8a 89.70(14) 90.58(12) N18–Fe1–N1c 87.35(14) 88.56(12) N18–Fe1–N18c 180.00 180.00 N18–Fe1–N8d 90.30(14) 89.42(12) N8a–Fe1–N1c 92.96(13) 91.88(13) N8a–Fe1–N18c 90.30(14) 89.42(12) N8a–Fe1–N8d 180.00(19) 180.00(16) N1c–Fe1–N18c 92.65(14) 91.44(12) N1c–Fe1–N8d 87.04(13) 88.12(13) N18c–Fe1–N8d 89.70(14) 90.58(12)
3.5 57Mössbauer Spectroscopy
The spin transition has been investigated by Mössbauer spectroscopy, only in the heating mode. In Figure 3.6, four significant Mössbauer spectra are depicted. From room temperature to 250 K (Figure 3.6 a), the Mössbauer data does not significantly change. A main signal is observed with quadrupole splitting (1.8–2.0 mm/s) and isomer shift (0.94–0.97 mm/s) values typical for high spin FeII centres (Table 3.3).47 In addition, a smaller HS signal corresponding to less than 10% of the total iron(II) centres is present. This signal corresponds to an impurity, and does not vary in intensity throughout the whole temperature range. As the temperature is lowered down to 140 K, no significant changes are observed for both signals. At 135 K, a new signal with a quadrupole splitting of 0.203(1) mm/s and an isomer shift, δ, of 0.443(2) mm/s is detected, which is typical for LS FeII ions (Table 3.3).47 The decrease of intensity of the HS signal with concomitant increase of the LS signal (see HS/LS percentages, Figure 3.6) can be assigned to the HS→LS spin transition of the Fe centres. The temperature of the transition, i.e T½, determined by Mössbauer spectroscopy is comparable to the one obtained by magnetic susceptibility measurements. From 135 to 4.2 K, no further significant changes are noted, indicating that the spin-crossover process is complete within a range of 5 K around T½. Indeed, the steepness of the transition is clearly evidenced in Figure 3.6 a, bottom.
Unfortunately, no cooling experiments have been performed, precluding the possibility to investigate the occurrence of a hysteresis.
Table 3.3. Mössbauer Spectral Hyperfine Parameters for complex 1
LS component HS component HS residual
Temperature
[K] δ[mm/s] ∆[mm/s] δ[mm/s] ∆[mm/s] HS fract. ∆[mm/s] δ[mm/s]
4.2 0.45(14) 0.214(97) 1.10(61) 3.3(12)
25 0.448(1) 0.214(1) 1.108(6) 3.301(12) 50 0.445(1) 0.213(1) 1.106 (7) 3.306(14) 75 0.446 (2) 0.210(1) 1.099(8) 3.300(16) 100 0.445 (2) 0.207(1) 1.092(10) 3.275(21) 125 0.443 (2) 0.204(2) 1.085(13) 3.263(27) 130 0.441(1) 0.203(2) 1.041(34) 2.649(68) 1.48(82) 1.082(18) 3.267(36) 135 0.443 (2) 0.203(1) 1.026(11) 2.529(22) 5.06(82) 1.097(16) 3.237(32) 145 1.031(74) 2.48(15) 96(74) 1.1(12) 3.3(24) 175 1.016(80) 2.34(17) 95(83) 1.1(16) 3.3(32) 225 0.99(10) 2.09(20) 95(95) 1.0(20) 3.2(39) 250 0.97(11) 1.98(22) 90(120) 1.1(34) 3.2(68)
290 0.96(12) 1.86(24) 95 1.0(26) 3.2(53)
294 0.94(15) 1.78(29 96 1.0(41) 3.1(82)
a) b)
Figure 3.6. a) From top to bottom; plots of quadrupole splitting ∆, isomer shift δ, and high- spin fraction vs. temperature (obtained from Mössbauer). b) Selected 57Fe Mössbauer spectra obtained at 250, 145, 135 and 4.2 K. The LS, HS and HS impurity signals are depicted in dark grey, light grey and black respectively.
Figure 3.7. UV-vis-NIR spectra for compound 1 at room temperature (dotted line) and at liquid nitrogen temperature (full line).
3.6 UV-vis-NIR
The reflectance UV-vis spectra for [Fe(µ-btzmp)2(btzmp)2](ClO4)2 show a dependence on the temperature. The spectrum at room temperature (Figure 3.7) exhibits a main band centred at 260 nm with a shoulder at 360 nm, corresponding to metal-to-ligand charge transfer. A second band at 850 nm can be assigned to the 5T2→5E transition, characteristic of FeII HS compounds. The spectrum at 77 K (Figure 3.7) illustrates a shift of the metal-to-ligand charge transfer band to higher wavelengths (λ = 316 nm), and the appearance of two new bands at 383 nm and 551 nm. These two absorption bands correspond to the 1A1→1T2 and the
1A1→1T1transitions, respectively, and are distinctive of LS FeII compounds.45 The absorption band at 383 nm seems to be already present at room temperature, as a shoulder of the 260 nm band. At 77 K, the band ascribed to the 5T2→5E transition in FeII HS systems is significantly less intense; the occurrence of this absorption at this temperature is most likely due to HS impurities, already observed by Mössbauer spectroscopy. Unfortunately, the irradiation with green light (λ = 559 nm) in the SQUID cavity at 10 K, did not result in a detectable population of the HS metastable state (LIESST effect).
3.7 Discussion
The btzmp ligand completes the endi–btzp series (Figure 3.1). Indeed, these ligands all possess a C2 bridge, substituted differently. The reaction of these ligands with different FeII salts results in 1D polymers. For [Fe(btzp)3](ClO4)2, the syn conformation of the ligand has been suggested to be the origin for its dimensionality.39 This spatial arrangement, in which the tetrazole rings are almost eclipsed (Newman Projection), is also observed for [Fe(endi)3](BF4)2. In contrast, the solid-state structure of [Fe(µ-btzmp)2(btzmp)2](ClO4)2
are pointing towards different directions, which is reflected in a longer ND···ND’ distance (endi, 5.107 Å; btzp, 5.038 Å and btzmp, 5.687 Å). However, the dimensionality, i.e. 1D linear chain, of the compound is not affected by the different ligand conformation. Interestingly, the disposition adopted by the bridging btzmp ligands in compound 1 is also observed in the crystal structure of the free ligand. Apparently, it is not the coordination to the metal ions that determines the conformation of the ligand, as observed for endi. The second methyl substituent most likely forces the ligand to adopt a special spatial arrangement which affects the disposition of the tetrazole rings.
As stated above, the dimensionality of the compound is not influenced by the specific conformation of the btzmp ligand. However, it appears that the steric hindrance, owing to the additional methyl group, determines the number of ligands bridging the FeII centres. Indeed, for both coordination compounds [Fe(btzp)3](ClO4)2 and [Fe(endi)3](BF4)2, the FeII centres are triply bridged by their corresponding ligands, while in [Fe(µ-btzmp)2(btzmp)2](ClO4)2, the iron ions are doubly bridged. This steric hindrance is again evidenced by the fact that, contrary to endi and btzp which present disorder when coordinated, btzmp shows only one stable position of its bridging mode. The two btzmp molecules acting as monodentate ligands consequently modify the crystal packing, which differs notably from those of [Fe(btzp)3](ClO4)2 and [Fe(endi)3](BF4)2 (Chapter 5, Figures 5.5). The different solid-state packing is reflected in longer interpolymeric distances and shorter intrapolymeric distances for compound 1 (Chapter 5, section 5.4).
The structural modifications occurring during the spin transition are distinct for each compound. Compound 1 shows the largest alterations of the metal to ligand distances, while [Fe(btzp)3](ClO4)2 has the smallest variations of the bond lengths. Surprisingly, for the btzp derivative, the change in bond length does not structurally modify the ligand. In the case of endi, a variation of the torsion angle is observed, which amounts to 7º. This different response of the ligands to the decrease of the Fe–N distances does not appear to affect the spin- crossover properties of the corresponding materials, as they both show gradual HS→LS transitions at similar temperatures. In compound 1, the torsion angle does not significantly vary (a minor variation of about 1º is observed). However, a structural change is noted for the ligand btzmp upon the spin crossover. The [ttz-centroid]–ND–Fe angle diverges by almost 6°
from the HS to the LS forms. In 1D bistetrazole-based SCO coordination polymers, this angle usually is larger for the LS compound (see Chapter 5, section 5.3). This increase suggests that the tetrazole rings “open up” to counterbalance the shortening of the Fe–N distances. The adjustment of the [ttz-centroid]–ND–Fe angle is especially large for compound 1, while it is almost insignificant for the other two complexes. The increase of the [ttz-centroid]–ND–Fe angle has a drastic effect on the disorder of the perchlorate anions. At high temperatures, the ClO4– ions are disordered and involved in weak anion–π interactions with the tetrazole rings.
Upon the HS→LS transition, the tetrazole rings “open up”, i.e. [ttz-centroid]–ND–Fe angle increases, which result in closer contacts with the perchlorate anions (no shift of the anion is detected). Accordingly, the anion–π interactions strengthen (shorter anion···centroid separations), with concomitant loss of disorder.
The magnetic susceptibility measurements reveal that the spin-transition phenomenon detected in 1 presents a hysteresis loop of 3 K. The HS↔LS transition is much steeper than the ones observed for [Fe(btzp)3](ClO4)2 and [Fe(endi)3](BF4)2. Abrupt transitions are commonly assigned either to highly cooperative systems,45 or to compounds which present a crystallographic phase transition.48, 49 The main structural differences between the SCO coordination polymers obtained from the ligands endi, btzp, and btzmp are found in their respective crystal packing. The steric bulk of the non-bridging btzmp ligands present in compound 1 generates a different packing of the polymer chains, possibly giving rise to a more rigid and compact framework. This increased rigidity would induce a higher cooperativity, and thus the occurrence of a hysteresis loop. The thermal order/disorder transition observed with the perchlorate anions may also cause the abrupt spin-state change. It appears that the spin transition is responsible for the order/disorder transition of the anions in the lattice. This affirmation, although speculative, is based on the X-ray structural data. The aforementioned variation of the [ttz-centroid]–ND–Fe angle during the spin transition observed for all 1D SCO materials, in this case strengthens the anion–π interactions and favours just one position of the anion in the crystal lattice. Nevertheless, both structural transformations, i.e. adjustment of the tetrazole plane–iron angle with the consecutive ordering of the ClO4– ions, are most likely involved in the abruptness of the transition. The transition temperature T½ determined by Mössbauer spectroscopy is comparable to the one obtained by magnetic susceptibility measurements. 57Fe Mössbauer measurements also reveal the presence of HS iron(II) impurities, whose amount is higher than the remaining HS fraction observed by magnetic susceptibility experiments. These impurities are also detected by UV- vis spectroscopy at low temperature (absorption band at λ = 850 nm).
3.8 Conclusions Part I
The incorporation of an additional methyl substituent on the backbone of the spacer separating the tetrazole rings, i.e. the ligand btzmp, significantly affects the solid-state arrangement of the ligand in comparison with the original ligands endi and btzp. As a result, the iron(II) polymer chains obtained from this btzmp ligand exhibit a notably distinct crystal packing, compared to the related materials prepared from the less hindered ligands endi and btzp. Indeed, the presence of two methyl substituents in btzmp most likely decreases the flexibility of the ligand, thus affecting its coordinating properties. Accordingly, in contrast to the triple-bridged [Fe(btzp)3](ClO4)2 and [Fe(endi)3](BF4)2 systems, [Fe(µ- btzmp)2(btzmp)2](ClO4)2 is a 1D polymer generated via the bridging of the FeII centres by only two btzmp ligands. The coordination sphere of the metal ions is then completed by two monodentate btzmp ligands. The presence of these non-bridging monodentate btzmp induces an unprecedented crystal packing in the well-known series of 1D bistetrazole-based compounds. The structural changes occurring during the transition are also affected by the presence of the two methyl groups. The steric bulk probably prevents the modification of the torsion angle. However, the [ttz-centroid]–ND–Fe angle is drastically altered. This common structural feature observed in bistetrazole compounds is most likely responsible for the thermal order/disorder transition of the ClO4– counterions observed in [Fe(µ-
btzmp)2(btzmp)2](ClO4)2. The combined structural variations, namely the adjustment of the [ttz-centroid]–ND–Fe angle to counteract the variations of the Fe–N separations, and the thermal order/disorder of the ClO4– anions, appear to produce an abrupt spin transition, with occurrence of a hysteresis loop. A simple increase in rigidity of the overall network is not discarded.
Part II: [FeII(btzpol)1.8(btzpol-OBF3)1.2](BF4)0.8(H2O)(CH3CN)
3.9 X-ray Structure of [FeII(btzpol)1.8(btzpol-OBF3)1.2](BF4)0.8(H2O)(CH3CN) (2)
[FeII(btzpol)1.8(btzpol-OBF3)1.2](BF4)0.8·(H2O)(CH3CN) crystallises in the space group P21/m, with Z=2. The asymmetric unit consists of a half of an FeII on a special position, a btzpol ligand, a slightly modified ligand btzpol-OBF3 (see scheme 3.1)in which the alcohol moiety has reacted with a BF4– anion, and a ligand that presents an occupational disorder between OH/OBF3 in a 1:4 ratio.
Scheme 3.1. Schematic representation of the reaction of btzpol with a BF4– anion.
Table 3.4 includes the most relevant crystallographic parameters. A view of the FeII coordination sphere is depicted in Figure 3.8. The metal centre is hexacoordinated by six N tetrazole donors belonging to six different ligands. Two of these six bistetrazole ligands are covalently bound to a BF3 moiety through the alcoholic function. An almost perfect Oh
symmetry is found around the FeII atom, with N–Fe–N angles ranging from 88.88(12)º to 91.12(12)º (see Table 3.4). The slight distortion observed is not unusual for FeII tetrazole compounds in the high-spin state. The Fe–N distances (2.169(3) to 2.182(3) Å) are in the expected range for FeII in the HS state.45 Along the b axis, the Fe–N distances are of 2.169(3) Å and 2.175(3) Å, slightly shorter than those along the c axis, but still in the range of Fe–N distances for a high-spin FeII compound.
This iron coordination entity consists of a building unit whose self-assembly generates a two-dimensional network, in which the FeII ions are connected through single 1,3- bis(tetrazol-1-yl)-2-propanol (btzpol) bridges in the b direction and by means of double ligand bridges in the c direction (Figure 3.8). Along the latter direction, one of the two ligands is a btzpol-OBF3 entity. The other ligand presents an occupational disorder,39, 50 with an OBF3/OH occupancy ratio of 1:4. This particular arrangement explains the decimal numbers present in the molecular formula. The representation shown in Figure 3.8 characterises the major component of the disordered structure. BF4– anions are interacting by means of C–H···F forces with the hydrogen atoms of the ligands along the c direction (Table 3.5).
N N N N N
HO N
N N
+
BF4-
N N N N N
O N
N
N B H
F F F F
N N N N N
O N
N N -F3B
+ HF -
Table 3.4. Crystallographic data for 2.
[Fe(btzpol)1.8(btzpol-OBF3)1.2](BF4)0.8(H2O)(CH3CN) Formula C17H28.20B2F6.80FeN25O4.20
FW/g mol-1 856.72 Crystal system monoclinic Space group P21/m λ (Å) 0.6894 a/Å 7.1402(4) b/Å 24.0050(15) c/Å 10.6885(7)
α/º 90
β/º 90.9(2)
γ/º 90
V/Å3 1831.77(19)
Z 2
ρ(g/cm3) calculated 1.553
T/K 150
Crystal Shape block
Colour Colourless R(F) 0.0719
Rw(F2) 0.2132
The bridging ligand in the b direction (Figure 3.9) has a Gauche-Gauche conformation (GG), when viewed along C12···C13 and C13···C14 axes, respectively, with an N11···N11g separation of 8.433(4) Å. The Fe1–N11 bond distance is 2.182(3) Å (see Table 3.6). Since the btzpol ligands are bridging the iron centres (Fe1···Fe1c; 12.003(1) Å) in a zigzag fashion, the alcohol functions are alternating in opposite directions. These alcohol moieties do not show any type of intermolecular contact with other adjacent molecules present in the lattice.
Table 3.5. C–H···F interactions for 2.
Atoms Distance (Å) C1–H1···F6 2.27 C2–H2B···F4 2.52 C4–H4A···F3 2.52 Intra 1 C5–H5···F3 2.33
Intra 1 C11–H11···F2 2.34
a) b)
Figure 3.8. a) Labeled ORTEP representation of the iron coordination sphere of [FeII(btzpol)1.8(btzpol-OBF3)1.2](BF4)0.8(H2O)(CH3CN). Hydrogen atoms, solvent molecules and counterions are not shown for clarity. b) Representation of the major component of the disordered structure of 2along the a axis. Hydrogen atoms, solvent molecules and counterions are not shown for clarity.
The view along the b axis shows another remarkable feature of the coordination polymer (Figure 3.9 b). The two btzpol ligands bridging the iron centres (Fe1···Fe1a;
10.6885(7) Å) have a TG’ conformation,51 i.e. a Trans-Gauche conformation, if viewed along the C2···C3 and C3···C4 bonds, respectively, with a N1···N8 distance of 7.647(5) Å. The fact that two ligands bridge the iron centres results in a significant decrease of the distance of about 1.3 Å between the metal ions, as compared to the Fe–Fe separation observed along the c axis. However, the conformation of the ligand may also play a role in determining these distances, which will be discussed below. The alcohol function of one of the two bridging ligands has reacted with a BF4– anion (scheme 3.1). The resulting O–BF3 group interacts with one of the tetrazole rings of the same ligand through anion-π interactions (C5···F2 3.067(8) Å). A water molecule is sitting in the cavity dislocated in four positions (occupancy factor of 0.2 for each position). The 2D layers have a thickness (d, C4···C2d) of about 7.409(7) Å, produced by the two bridging ligands in this direction (see Figure 3.9). No interactions between layers are observed, as solvent molecules or counterions are not found between the polymer sheets. Both solvent molecules and counterions tend to be involved in hydrogen bonding networks that increase the rigidity of a system. The presence of only Van der Waals interactions makes the interaction between layers weak, probably one of the reasons behind the weak cooperative transition (see section 3.10).
Table 3.6. Selected bond lengths, intra-atomic separations (Å) and angles (º) for 2 at 150 K.
Symmetry operations: b = x, y, 1+ z; d = –x, 1-y, 1–z; e = –x, 1–y, 2–z; g = x, 3/2– y, z
a) b)
Figure 3.9. Representation of 2 along the c axis (a), andalong the b axis (b). Hydrogen atoms are not shown. d is defined as (d, C4···C2d), see text.
[Fe(btzpol)1.8(btzpol-OBF3)1.2](BF4)0.8 (2)
Fe–Fe (double ligand bridge) 10.689(1) Fe–Fe (single ligand bridge) 12.003(1)
N1–N8 7.647(5) N11–N11g 8.433(4) Fe1–N1 2.175(3) Fe1–N11 2.182(3) Fe1–N8d 2.169(3) N1–Fe1–N1e 180.00(0)
N1–Fe1–N8b 89.62(12)
N1–Fe1–N8d 90.38(12)
N1–Fe1–N11 90.78(13)
N1–Fe1–N11e 89.22(13)
N8d–Fe1–N8b 180.0(17)
N8d–Fe1–N11 91.12(12)
N8d–Fe1–N11e 88.88(12)
N11e–Fe1–N11 180.00(0)
3.10 Magnetic Studies
Compound 2 easily looses some of the lattice solvent molecules resulting in [(Compound 2)-(H2O)-0.5(CH3CN)] (2a). In all batches crystallised, the loss of solvent molecules led to the same composition.d All measurements here presented were carried out with the desolvated compound 2a. Magnetic data for a polycrystalline sample of 2a have been recorded in the temperature range 6–300 K. The χmT vs T plot is depicted in Figure 3.10 (white circles), where χm is the molar magnetic susceptibility per iron center and T the temperature. At 300 K, the χmT value is 3.4 cm3 mol–1 K, which is close to that expected for a spin-only HS iron center (3 cm3 mol–1 K , g = 2). Upon cooling, the χmT value remains constant down to a temperature of 200 K, where a gradual transition sets in, expanding over a temperature range of 130 K, after which it reaches a χmT value of 0.9 cm3 mol–1 K at 50 K. At very low temperatures a slight decrease in the χmT value is observed, which can only be assigned to the zero-field splitting of the remaining high-spin iron centers. The transition temperature, T½, is 112 K. This value is a rather low transition temperature, although not unusual since low T½
have also been observed for bistetrazole-based iron complexes,39 and especially for mononuclear tetrazole-based compounds.45, 52-55
3.11 LIESST and Relaxation Experiments
UV-Vis-NIR spectroscopic measurements at liquid nitrogen temperature (Figure 3.10 b) of 2a indicated that the 1A1→1T1 absorption band for LS FeII is situated at 555 nm; therefore a cut-off filter with transmittance below 559 nm was used. Excitation with green light at 6 K within the squid cavity created the so-called high-spin metastable state, which was followed by the increase in magnetic response (Figure 3.10 a). Times of excitation pointed at a low efficiency of the light-induced process, as more than three hours of continuous irradiation were needed to reach saturation (a maximum population of the metastable HS state). The estimation of the exact percentage of metastable HS state formed is hampered by the fact that the χmT value at this temperature shows a large contribution from the zero-field splitting of the HS FeII species. Consequently, the increase in χmT value on irradiation yields a χmT signal which holds the contributions from the newly created metastable HS form, as well as its own zero-field splitting. A χmT value of 2.2 cm3 mol–1 K was obtained at 6 K, indicating by comparison with the high temperature χmT data, that a nearly complete population of the metastable state was achieved. The skinning effect, due to the amount of sample used, and the inherent technical problems of irradiation experiments can account for the incompleteness of the population.
After having yielded the maximal population of the metastable HS state, i.e. a χmT value of 2.2 cm3 mol–1 K, the temperature was increased at a rate of 0.3 K min–1 with the lamp switched off. The initial increase of χmT reaching a value of 2.6 cm3 mol–1 K seems to confirm that the low χmT value obtained during excitation is partly due to the presence, at this temperature, of a zero-field splitting effect. This phenomenon is observed until it is masked at
d See chapter 2 for details on the experimental procedures and recipes for all compounds presented in this
20 K by the relaxation of the metastable state, placing the T(LIESST) value at 46 K (inset Figure 3.10 a).56 A similar behaviour is observed for the complex [Fe(nditz)3](ClO4)2 bearing alkyl-based spacers.57
a) b)
Figure 3.10. a) Plot of experimental χmT vs T for 2a(white circles). Temperature dependence of the product χmT after LIESST, in the range 6–300 K (black and white hexagons). The inset shows the derivative of the product χmT. UV-Vis-NIR measurement for 2a at room temperature (dashed lined) and at liquid N2 temperature (full line).
Figure 3.11. HS→LS relaxation curves after LIESST (2a) at six different temperatures.
Normalised as: χmT (t) - χmT (LT) / χmT (Initial)- χmT (LT). χmT (LT) = low temperature value obtained from the normal spin-crossover curve; χmT (Initial) = value after population of the metastable high-spin state.
Relaxation kinetics of the HS→LS conversion for the compound [FeII(btzpol)1.8(btzpol- OBF3)1.2](BF4)0.80.5(CH3CN) (2a) was studied at different temperatures: 6, 12, 24, 36, 42 and 50 K. In all cases, the compound was excited at 6 K and then rapidly set to the corresponding subsequent temperature. The χmT value was normalised to γHS = 1.58 In the case of the highest temperature, a part of the iron centers may have relaxed before the measurement was started, although this should not alter the value for the high-spin to low-spin relaxation rates (KHL).
The χmT value reflecting the decay in the percentage of metastable HS form was then measured every 30 seconds. Similarly shaped relaxation curves are obtained for the 12, 24, 36 and 42 K, temperatures at which total relaxation is not reached even after several hours (Figure 3.11). By contrast, at 50 K the compound relaxes in a few hours. This result indicates that the compound is already in the thermally activated region. In this region, the compound may indeed populate higher vibrational states from which a more effective relaxation to the LS state occurs.
3.12 Discussion
[FeII(btzpol)1.8(btzpol-OBF3)1.2](BF4)0.8·(H2O)0.8(CH3CN) (2) is a 2D SCO polymer based on a bistetrazole ligand. Bistetrazoles are widely used in reticular synthesis combined with different metals. Copper, nickel and zinc complexes have shown different topologies, and 1D, 2D or 3D structures have been reported.59-61 The use of FeII as the metal ion not only generates unique supramolecular networks, but may also produce spin-transition compounds.
1-Substituted tetrazoles provide the proper crystal field splitting to form spin-transition materials. In this respect, all bistetrazole-based spin-transition compounds previously reported do present 1D or 3D networks.40 The present coordination polymer is thus the first 2D bistetrazole-based material exhibiting such magnetic properties. Compound 2 looses solvent molecules to form compound 2a. This change in composition is probably due to the open structure of compound 2 and to the lack of strong interactions of the lattice solvent molecules present.
The synthesis of complexes with BF4– counterions may yield products where the instability of the anion is observed.62, 63 Stable hydrogen-bonding interactions between the fluorine atoms and protons of donor ligands64 can indeed result in the formation of HF molecules. The ensuing fluorine atoms may then act as ligands and form unexpected complexes.65 In the present case, the btzpol ligand, through its alcohol functions, provides protons creating an R–O–H·(Fn)BF4-n nucleophile (Scheme 3.1).64 The elimination of HF molecules gives rise to the formation of stable R–O–BF3 entities. This phenomenon which has been reported before,64, 66 generates a slightly modified ligand which bridges the iron centers through the N1 atom of the tetrazole rings. In this case, it has been proven through repetitive synthesis of the material that the composition and physical behaviour (see below) are reproducible, indicating that the formation of the OBF3 modified ligand always takes place under these experimental conditions.
In bis(4-pyridyl)alkanes, the length of the spacer and the conformation seem to play an important role in determining the dimensionality of the coordination compound.67 In
bis(polyazole) derivatives, such as bistetrazoles, an additional factor is present, namely the free rotation of the azole ring around the ring-spacer bond.51 This free rotation influences the ND···ND’ distance (coordinated nitrogens of the bis(tetrazole)) and thus will also affect the important structural features of the resulting coordination architectures, such as dimensionality, interpenetration, and size of the cavity. Ultimately, this free rotation will affect the properties of the SCO as well.
Recently, copper(II) and zinc(II) coordination polymers prepared with the 1,3- bis(tetrazol-2-yl)propane ligand were reported.51 These studies demonstrate that the metal–
metal distance in alkyl-bis(polyazole)-based polymers only depends on the spacer length and not on its conformation. In fact, the difference in distances owing to the distinct conformations is compensated by the tilting of the N(ring)–C(spacer) bond, resulting in an ND···ND’ distance difference of 0.188 Å between the two disparate conformers (which is considered as negligible). In the present study, two types of conformation are observed for the ligands, i.e. a TG’ conformation along the b axis (double ligand bridge) and a GG conformation along the c axis (single ligand bridge). The disparity between the ND···ND’ and M···M distances of the two ligands is 0.786 Å and 1.314 Å, respectively. The reaction of the ligand with the BF4– anion to form the altered ligand btzpol–OBF3 appears to be responsible for these significant differences in distances between the two ligands. The presence of the more bulky OBF3 moiety stabilises a shorter conformation of the ligand through anion-π interactions (C5···F2 3.067(8) Å) between the fluoride atoms and the electron-deficient tetrazole rings. Moreover, these anion-π interactions annihilate the rotational freedom of the tetrazole around the N(ring)–C(spacer) bond, which is believed to be responsible for the compensation of the difference in ND···ND’ distance between different conformers.51
The presence of the linked OBF3 moieties is apparently responsible for the formation of the 2D framework obtained. Bistetrazole ligands bearing a short linear alkyl spacer (2 carbon spacer) lead to the formation of 1D coordination polymers,41 like [Fe(endi)3](BF4)2
(endi = 1,2-bis(tetrazol-1-yl)ethane) or [Cu(endi)3](ClO4)2. In the present complex, a third btzpol–OBF3 ligand or a btzpol ligand obviously cannot bind to the metal ion along the same direction, due to the steric constraints brought in by the R–OBF3 group. As a result, the binding of the third ligand, i.e. btzpol, can only extend the coordination structure along another direction, resulting in this peculiar 2D layered architecture.
Tetrazole-based ligands are known to center the spin transition at relatively low temperatures. The current transition at 112 K, accompanied by a pronounced thermochromic effect, is therefore not surprising. In a recent systematic study on a series of FeII compounds based on bistetrazole ligands bridged by alkyl chains as spacers, a linear relation between the number of carbons and the T½ has been found; moreover a dependence on the parity of the chain has also been observed.57 Although the present compound does not exhibit a three- dimensional network, the T½ value follows this trend. This appears to be reasonable as the dimension of the polymer formed has not been found to affect the T½ of the transition, and thus T½ seems only to depend on the nature of the ligand itself.57 Nevertheless, this case represents one of the lowest transition temperatures reported for Fe(II) until now.52, 53 In the present study, the relatively high percentage of iron centers that do not undergo the transition
is reproducible over different samples and independent of the cooling rate. It is hence neither related to impurities, nor to trapping in the HS state. Structural defects within the solid-state framework therefore constitute a possible explanation of the incompleteness of the transition.
Another reasonable possibility is that the structural changes associated with the spin transition, e.g. the contraction of the crystal lattice, restricts the number of iron(II) ions capable of switching from the HS to the LS state. This can explain that the remaining HS value at low temperatures is always similar. Both explanations can account for this effect.
The cooperativity in spin-crossover materials is a desired property as it leads to steeper transitions and, in some cases, to hysteresis effects. This cooperative behaviour results from the structural changes associated with the transition, which are elastically communicated throughout the lattice. The gradual transition observed for 2a, indicates that the flexibility of the 2D network is hampering the propagation of the structural changes intrinsic to the transition.21, 46 Normally, it is through rigid linkers,68 or via an efficient crystal packing42, 43 that the rigidity is achieved in coordination polymers. In the present investigation, the btzpol ligand has an alkyl-based spacer which confers flexibility to the ligand, and thus acts as a shock absorber, as in the case of [Fe(btzp)3](ClO4)2.39 At the same time, in [FeII(btzpol)1.8(btzpol-OBF3)1.2](BF4)0.8(H2O)0.8(CH3CN) the BF4– anions, that are found in the cavities, do not occupy completely the voids. This allows the ligand to perfectly adapt its conformation in order to counterbalance the molecular distortions created by the SCO.
Unfortunately, the use of other counterions that would maybe fill up entirely the space in the cavities, namely Cl–, ClO4– and PF6–, has not yielded spin-transition materials. Furthermore, intermolecular interactions between layers are not observed, which makes the structure less rigid.
The features of the transition indicate that the metastable high-spin state can be observed at low temperatures. The low transition temperature and the lack of cooperativity enhance the stability of this metastable state.69 Indeed, this is observed when the compound is excited with green light during long excitation times. These long excitation times needed can be explained by a skinning effect and/or a low intensity light emission. The slight incompleteness of the population of the metastable HS state and the presence of zero-field splitting for HS FeII when exciting at 6 K, lead to a χmT value which is lower than that expected for an FeII center. Once the light is turned off and the temperature is slowly increased (see Figure 3.10), the relaxation sets in gradually. This behaviour indicates that part of the iron centers relax earlier and faster than the bulk majority. Consequently, a thermally activated process cannot account for this initial decay of the metastable HS form. Actually, this effect can be attributed to the lack of homogeneity, most likely as a result of the occupational disorder of the OH/OBF3 moiety in one of the ligands. This disorder gives rise to slightly different FeII SCO chromophores with slightly disparate transition temperatures, and hence different relaxation rates.16 This difference matches with the fact that before the measurement has started, the relaxation rates show a fast decay to the low-spin state of a small fraction of the iron centers (normalised to 1).58
A relatively high T(LIESST)56 of 46 K is found experimentally with a warming rate of 0.3 K min–1 for 2a. Based on Letard’s study,70 for monodentate compounds, the T(LIESST)
and the T½ relate as T(LIESST) = T0–0.3T½ (1). The present compound [FeII(btzpol)1.8(btzpol- OBF3)1.2](BF4)0.80.5(CH3CN), lies close to the T0 = 100 K line (monodentate ligands). A value of 66 K is obtained for T(LIESST) when using equation (1); the difference with the experimental value can be due to the fact that this compilation does not consider any polynuclear material.70 The study of the relaxation rates for bistetrazole-based compounds had not been performed yet and thus was of particular interest. Single exponential relaxation curves are found for all temperatures studied (Figure 3.11). The relaxation data were fitted to an exponential decay equation to obtain the relaxation constant values (kHL) (Table 3.7), which are shown in Figure 3.12 as an Arrhenius plot. The curve obtained exhibits two distinct parts, corresponding to the two different relaxation regions. At higher temperatures, the temperature dependence of the relaxation constant follows a simple Arrhenius behaviour (Equation 3.1). This is the so-called thermally activated region where the relaxation can be considered as occurring from excited vibrational states. At low temperatures, a deviation from the simple Arrhenius behaviour is observed as the relaxation constant does not depend on the temperature. This behaviour is characteristic of a low-temperature tunneling process.69, 71 The curvature results from the overlap of both regions.
kHL = A×exp(–Ea/KBT) (2.1)
Table 3.7. HS to LS relaxation rates after quantitative LIESST effect performed at 6 K.
Considering the high-temperature region (from 36 to 50 K), an Arrhenius fit (S2) results in a very low activation energy value (Ea). As the energy difference between the two lowest vibrational levels of the HS state and the LS state is directly dependent on the T½,72, 73 such a low transition temperature (112 K) indicates that both states are close in energy. Thus, one would expect a higher activation energy for this compound. The large deviation from linearity observed in the temperature range 36–50 K evidences the presence of tunneling, which justifies the low Ea value obtained.73 Indeed the low value obtained for the pre- exponential factor A confirms a dominant tunneling effect.36 Relaxation curves recorded at higher temperatures are required to obtain a more accurate value.
Temperature (K) KHL(10–4 s–1)
50 3.57 42 5.37 36 3.2 24 1.4 12 0.58 6 0.725
Figure 3.12. Arrhenius plot for the relaxation constants of 2a. The inset shows the linear fit for the points T = 50 K, 42 K and 36 K.
3.13 Conclusions Part II
The reaction of bistetrazole-based ligands with FeII usually leads to the formation of 1D or 3D spin-transition polymers, depending on the length and/or the conformation of the spacer involved. The present ligand 1,3-bis(tetrazol-1-yl)-2-propanol btzpol, has demonstrated for the first time, that a 2D spin-transition polymer can be obtained with bistetrazole-based ligands. Two btzpol ligands bind along the same axis forming a 1D polymer, as for the previously reported cases with short spacers (2-carboned links), but the unexpected reaction of the btzpol ligand with BF4– anions has forced the third bridging ligand to accommodate along a different axis, forming a 2D polymer. The distortion on the bistetrazole conformation created by the R–OBF3 moiety and its implicit bulkiness explain the singular disposition adopted by the compound. [FeII(btzpol)1.8(btzpol-OBF3)1.2](BF4)0.80.5(CH3CN) shows spin- crossover properties induced both by temperature and light. This first study of the relaxation behaviour of the metastable HS state carried out for this type of bidimensional coordination polymer proves that it follows a nearly temperature-independent rate at low temperatures (<
36 K) and a thermally activated behaviour at higher temperatures (> 36 K). There is an obvious overlap between these two limits, which results in low kinetic values for the temperature activated behaviour. The low degree of cooperativity in the SCO characteristics of the polymer can be explained by the presence of voids in the solid-state structure, which permits the flexible btzpol (btzpol–OBF3) to absorb the structural changes associated with the spin transition. Accordingly, the use of alcohol moieties to favour beneficial intermolecular interactions that would bring an extra rigidity to the material (cooperativity) has been surprisingly disrupted by the peculiar reaction with the BF4– counterions. The resulting R–
OBF3 moiety has been dominating the overall structure and thus the properties of the material.
3.14 General Conclusions
The compounds [FeII(btzpol)1.8(btzpol-OBF3)1.2](BF4)0.8(H2O)(CH3CN) and [FeII(µ- btzmp)2(btzmp)2](ClO4)2, although described separately, present similar structural and functional features. Actually, in both cases, the substituents on the spacer hinder the formation of the expected triply-bridged 1D polymers. Instead, doubly-bridged polymers are achieved, where the third ligand is pointing perpendicular to the axis of the polymeric chain. In the case of [FeII(µ-btzmp)2(btzmp)2](ClO4)2, the third bistetrazole molecule acts as a monodentate ligand, binding through only one of its tetrazole rings. This particular binding mode results in a 1D polymer with a different crystal packing, compared to the usual one found in all reported bistetrazole-based coordination networks. In the case of [FeII(btzpol)1.8(btzpol- OBF3)1.2](BF4)0.8(H2O)(CH3CN), the third ligand bridges the chain in a new direction, forming a 2D spin transition framework. The role of steric effects and/or rigidity of the ligand on the structural arrangement of the ensuing hybrid organic-inorganic materials (HOIMs) herein reported needs to be further investigated via slight modifications of the ligands.
3.15 References
1. Schmidt, G. M., Pure Appl. Chem. 1971, 27, 647-678.
2. Wells, A. F., J. Chem. Educ. 1977, 54, 273-276.
3. Abrahams, B. F.; Hoskins, B. F.; Liu, J. P.; Robson, R., J. Am. Chem. Soc. 1991, 113, 3045-3051.
4. Batten, S. R.; Hoskins, B. F.; Moubaraki, B.; Murray, K. S.; Robson, R., J. Chem. Soc.
Dalton Trans. 1999, 17, 2977-2986.
5. Hoskins, B. F.; Robson, R., J. Am. Chem. Soc. 1990, 112, 1546-1554.
6. Fujita, M.; Kwon, Y. J.; Washizu, S.; Ogura, K., J. Am. Chem. Soc. 1994, 116, 1151- 1152.
7. Yaghi, O. M.; O'Keeffe, M.; Kanatzidis, M., J. Solid State Chem. 2000, 152, 1-2.
8. Moulton, B.; Zaworotko, M. J., Chem. Rev. 2001, 101, 1629-1658.
9. Rosi, N. L.; Eckert, J.; Eddaoudi, M.; Vodak, D. T.; Kim, J.; O'Keeffe, M.; Yaghi, O.
M., Science 2003, 300, 1127-1129.
10. Chae, H. K.; Siberio-Perez, D. Y.; Kim, J.; Go, Y.; Eddaoudi, M.; Matzger, A. J.;
O'Keeffe, M.; Yaghi, O. M., Nature 2004, 427, 523-527.
11. Eddaoudi, M.; Kim, J.; Rosi, N.; Vodak, D.; Wachter, J.; O'Keefe, M.; Yaghi, O. M., Science 2002, 295, 469-472.
12. Halder, G. J.; Kepert, C. J.; Moubaraki, B.; Murray, K. S.; Cashion, J. D., Science 2002, 298, 1762-1765.
13. van Zutphen, S.; Robillard, M. S.; van der Marel, G. A.; Overkleeft, H. S.; den Dulk, H.; Brouwer, J.; Reedijk, J., Chem. Commun. 2003, 634-635.
14. Ohmori, O.; Fujita, M., Chem. Commun. 2004, 1586-1587.
15. Ohmori, O.; Kawano, M.; Fujita, M., J. Am. Chem. Soc. 2004, 126, 16292-16293.
16. Kahn, O.; Martinez, C. J., Science 1998, 279, 44-48.
17. Seo, J. S.; Whang, D.; Lee, H.; Jun, S. I.; Oh, J.; Jeon, Y. J.; Kim, K., Nature 2000, 404, 982-986.
18. Real, J. A.; Andrés, E.; Muñoz, M. C.; Julve, M.; Granier, T.; Bousseksou, A.; Varret, F., Science 1995, 268, 265-267.
19. Real, J. A.; Gaspar, A. B.; Niel, V.; Muñoz, M. C., Coord. Chem. Rev. 2003, 236, 121- 141.
20. Gütlich, P.; Goodwin, H. A., Spin crossover - An overall perspective. In Spin
Crossover in Transition Metal Compounds I, Topics in Current Chemistry, Gütlich, P.;
Goodwin, H. A., Eds. Springer: 2004; Vol. 233, pp 1-47.
21. Garcia, Y.; Niel, V.; Muñoz, M. C.; Real, J. A., Spin crossover in 1D, 2D and 3D polymeric Fe(II) networks. In Spin Crossover in Transition Metal Compounds I, Topics in Current Chemistry, Gütlich, P.; Goodwin, H. A., Eds. Springer: 2004; Vol.
233, pp 229-257.
22. Jay, C.; Grolière, F.; Kahn, O.; Kröber, J., Mol. Cryst. Liq. Cryst. Sci. Technol. Sect.
A-Mol. Cryst. Liq. Cryst. 1993, 234, 255-262.
23. Haasnoot, J. G., Coord. Chem. Rev. 2000, 200, 131-185.
24. 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.
25. Codjovi, E.; Sommier, L.; Kahn, O.; Jay, C., New J. Chem. 1996, 20, 503-505.
26. Garcia, Y.; van Koningsbruggen, P. J.; Codjovi, E.; Lapouyade, R.; Kahn, O.;
Rabardel, L., J. Mater. Chem. 1997, 7, 857-858.
27. Kahn, O.; Codjovi, E., Philos. Trans. R. Soc. Lond. Ser. A-Math. Phys. Eng. Sci. 1996, 354, 359-379.
