Spectroscopy and magnetism of polymeric Ln(CCl₃COO)₃·2 H₂O and their heteronuclear Ln₂Cu(CCl₃COO)₈ ·6 H₂O analogues (Ln=Sm, Gd)

Abstract

In order to study the exchange interactions between f- and d-electron ions two series of compounds, heteronuclear Ln₂Cu(CCl₃COO)₈· 6 H₂O (1, Sm; 2, Gd) and polymeric Ln(CCl₃COO)₃·2 H₂O (3, Sm; 4, Gd) trichloroacetates, were synthesised. Polymeric trichloroacetates 3 and 4 are isomorphous with erbium chloroacetate, whose structure contains dimeric units linked by carboxyl bridges and water molecules into endless chains. Heteronuclear 1 and 2 are isomorphous with Nd₂Cu(CCl₃COO)₈·6 H₂O reported by us previously. Absorption spectra of single crystals of 1–4 for two orientations of the crystals down to 4 K were investigated. Magnetic susceptibility measurements in the range 300–1.8 K were carried out. Optical and magnetic properties are related to the X-ray data. Electron transition probabilities were calculated and the effect of coupled ion pairs in the polymeric structure on the optical and magnetic properties was analysed. Changes in magnetic properties in relation to the mononuclear systems are discussed.

---

In the last few years many studies have been devoted to polynuclear and heteronuclear dimeric and polymeric systems. Binucleating ligands may provide unusual structural features and/or magnetic, optical and catalytic properties and allow the preparation of sophisticated molecular optical and magnetic devices, etc.^1–9 For the same reasons, heteronuclear copper lanthanide compounds have been widely studied and also because of their potential applications in superconducting ceramics. In addition, it is well known that lanthanide(III) ions show peculiar physical and chemical properties that make them essential components in the preparation of many new materials. The main features of these materials can be influenced by structural parameters, the anisotropy of the ground state of 4f ions and by the nature and intensity of the exchange and other interactions involving rare earths and d-electron ions. In these systems two paramagnetic metal centres can interact either ferromagnetically or antiferromagnetically through the bridging group.

With the aim of shedding light on this field we synthesised two series of polynuclear (3,4) and heteronuclear (1,2) trichloroacetates. The correlation of their optical and magnetic properties to the X-ray data is the subject of this report.

Experimental

Polynuclear single crystals of Ln(CCl₃COO)₃·2H₂O (Ln=Sm, 3; Gd, 4) and their heteronuclear analogues Ln₂Cu(CCl₃COO)₈·6 H₂O (Ln=Sm, 1, Gd, 2) were synthesised according to the procedure described previously.^10,11 Samples 1–4 were checked by single crystal X-ray diffraction and the cell constants were determined. Polynuclear trichloroacetates 3 and 4 are isomorphous with the erbium complex.¹² Heteronuclear Cu:Ln 1 and 2 have a similar structure to that previously reported for the Cu:Nd crystal.¹⁰

Absorption spectra between 250 and 1600 nm for two orientations (a and b) of single crystals of compounds 1–4 were recorded in the 293–4 K temperature range using a Cary–Varian 5 spectrophotometer equipped with a helium (Oxford) cryostat. Intensities of the f–f and d–d transitions were calculated and converted into oscillator strength values and further applied for calculation of the Judd–Ofelt parameters (τ_λ). Raman spectra reported in ref. 13 were used in the vibronic components analysis. Magnetic susceptibility measurements were carried out with a SQUID magnetometer down to 1.7 K in both the low-field and high-field regimes. The data were corrected for diamagnetic contributions using Pascal's constants.

Results and discussion

X-Ray structure characteristics

As all the effects detected in the spectra (intensity, splitting, broadening) depend strongly on the structure of the system under investigation, essential X-ray structural data are given here. The crystal structure of Nd₂Cu(CCl₃COO)₈·6 H₂O reported by us earlier¹⁰ (and isomorphous with both crystals 1 and 2 investigated here) consists of heteronuclear chains composed of dimers of Nd(1)O₉ and Nd(2)O₈ polyhedra, linked through three bidentate and one chelating bridging (tridentate type) carboxylate groups of the trichloroacetate molecules. Thus, a noncentrosymmetric dimeric unit is formed, further connected by planar CuO₄ squares. Two water molecules (OW1, OW2) add to the strongly distorted octahedron and link together the Cu(II) and Nd(IIIIII) ions, as do the carboxylic bridges. In this manner endless chains are formed, as shown in Fig. 1. Intrachain M–M distances are as follows: Nd1–Nd2=4.38 Å; Cu1–Nd1=4.155 Å; Cu1–Nd2=4.137 Å. The interchain separations are 9.472 Å (Nd2–Nd1) and 8.680 Å (Cu1A–Cu2B). Magnetic properties are affected by the chain structure and may depend strongly on the dimensionality of polymeric systems.

Fig. 1 Part of the one-dimensional [Nd₂Cu(CCl₃COO)₈(H₂O)₄] chain in the crystal structure of Nd₂Cu(CCl₃COO)₈·6 H₂O.¹⁰

The structure of Ln(CCl₃COO)₃·2 H₂O is an example of a one-dimensional polymer, composed of two different centosymmetric dimeric subunits bridged by carboxyl units and water molecules.¹²

Spectroscopic results

Gd³⁺ and Cu:Gd systems.. The Gd³⁺ ion with the 4f⁷ configuration has an ⁸S_7/2 ground state and excited state multiplets ⁶P_J (J=7/2, 5/2, 3/2 in order of increasing energy), ⁶I_J (J=7/2, 17/2, 11/2, 15/2, 13/2) and ⁶D_J (J=9/2, 1/2, 7/2, 3/2, 5/2) with increasing energy. These excited states are located in the 32000–41000 cm⁻¹ range above the ground state.¹⁴

The absorption spectrum of a Gd(CCl₃COO)₃ ·2 H₂O single crystal in the range 270–314 nm is presented in Fig. 2. Rather broad, unsplit bands were detected and the splitting of these bands does not change with a decrease in temperature to 4 K. Only two transitions from the ⁸S_7/2 level to the ⁶P_J multiplet were detected (⁶P_7/2 and ⁶P_5/2, the weakest, ⁶P_3/2, was not observed). The ⁸S_7/2→⁶I_J transitions at higher energy also do not split cleanly. Such behaviour can indicate disorder in the structure or effects of the ion–ion interaction in the undiluted single crystal, leading to a broadening of the lines. This is so in our case, a fully concentrated single crystal of the gadolinium compound. Although the positions of the f–f absorption lines are weakly dependent on the material type, the absorption lineshape and the Stark splitting differ from material to material. A relatively small splitting is caused by the crystal field (CF) effect. The ⁸S_7/2 ground state has a very small Stark splitting¹⁵ (of the order of 0.1 cm⁻¹) so that the absorption spectra reflect directly the splitting of the excited states.^15–17

Fig. 2 Absorption spectra of a Gd(CCl₃COO)₃ ·2 H₂O crystal at 293 K (d=0.1355 cm; I: slit=0.1 nm, II: slit=0.2 nm) and 4.2 K (d=0.117 cm; slit=0.2 nm).

Comparison of the splitting of the ⁶P multiplets with those in other systems indicates a CF effect for Gd³⁺ in the discussed crystal 4 (7/2, 72 cm⁻¹; 5/2, 64 cm⁻¹), comparable to those reported for the GdAl₃(BO₃)₄ single crystal (7/2, 96 cm⁻¹; 5/2, 62 cm⁻¹; 3/2, 18 cm⁻¹)¹⁸ and larger than in Gd:LaCl₃ (7/2, 50 cm⁻¹; 5/2, 37 cm⁻¹; 3/2, 18 cm⁻¹).¹⁹ Intensity analysis of the f–f transitions in the observed spectral range is collected in Table 1, for one orientation of both the polynuclear and heteronuclear gadolinium trichloroacetate crystals. In the heteronuclear crystal, oscillator strength values indicate only a small increase in the intensities of the ⁸S_7/2→⁶P_7/2 , ⁶I_7/2 transitions in comparison to those in crystal 4. Since the splitting of the ⁸S_7/2 level is very small, the decrease in temperature does not reflect the CF effect. Thus, the observed subtle changes in the intensities reflect the vibronic coupling and cooperative effects.

Table 1 The oscillator strengths of the f–f (×10⁸) and d–d (×10⁶) transitions calculated from Gd, Cu:Gd and Cu trichloroacetate single crystal spectra at different temperatures

Gd(CCl3COO)3 ·2 H2O			Gd2Cu(CCl3COO)8·6 H2O
λ/nm	293 K	4.2 K	293 K
^8S7/2→^6P7/2	313–309	9.10	9.67	11.99
^8S7/2→^6P5/2	306.5–305	3.67	5.53	7.42
^8S7/2→^6I7/2	280–278	9.79	11.70	13.96
^8S7/2→^6I9/2,^6I17/2	278–275	69.55	71.31	77.49
^8S7/2→^6I11/2,^6I13/2, ^6I15/2	275–270	153.72	158.71	162.49
^8S7/2→^6D9/2	253.5–252	13.08
Cu(CCl3COO)2·3 H2O			Gd2Cu(CCl3COO)8·6 H2O
λ/nm	293 K		λ/nm	293 K		4.2 K
orientation	a	b	a	b	b
d–d Cu	500–1400	218.68	269.76	400–1300	407.77	456.66	410.28

---

A comparison of the f–f transition intensities in the single crystals under investigation with those reported by Carnall et al.¹⁴ in water solution (HClO₄) reveals slightly lower intensities in the single crystal spectra. We have calculated the oscillator strength values for two orientations of the crystals (axes a and b). It is necessary to calculate the third orientation of the crystal (axis c) to obtain the mean values, which can be compared to those in solution. Since the calculated changes in intensities for two orientations are not high they should not change much, even after correction for the third orientation.

The fitting procedure for the coordination polyhedra of the Ln³⁺ ions in the crystal structures leads to C_2v symmetry for the lanthanide centres, both for the polynuclear trichloroacetates and for one Nd ion position in their heteronuclear analogue. Hence, the symmetry of the metal ion in water solutions, C_3h,^14,20,21 is higher than in the trichloroacetate single crystals.

A comparison of the f–f transition probabilities of crystals 2 and 4 shows a smooth variation of the oscillator strengths. This confirms that the metal centres in both types of crystals have similar environments.

Let us now focus on another aspect of the spectroscopic results, the intensities of the d–d transition in the heteronuclear Gd:Cu system and the monomeric Cu(CCl₃COO)₂·3 H₂O crystal (Fig. 3). Both the intensities and energies of the d–d band are significantly different. Strong distortion of the octahedral symmetry is manifested by an increase of the d–d band intensity in the heteronuclear Gd:Cu crystal. Also, the energy of the split band of 2 at 4 K is shifted to higher values. This effect is accompanied by a shift of the charge transfer (CT) Cu²⁺ band to higher energies (≈280 nm). The energy of this band remains in resonance with the ⁶I multiplets of the Gd³⁺ ions.

Fig. 3 (I) Absorption spectra of a Gd₂Cu(CCl₃COO)₈·6 H₂O crystal at 293 K (orientations a and b; d=0.0475 cm; o: in paraffin oil) and at 4.2 K (orientation b; d=0.067 cm). (II) Absorption spectra of a Cu(CCl₃COO)₂·3 H₂O crystal at 293 K (orientations a and b; d=0.031 cm).

Sm³⁺ and Sm:Cu systems.. The absorption spectra of the two samarium compounds, Sm₂Cu(CCl₃COO)₈·6 H₂O (1) and Sm(CCl₃COO)₃ ·2 H₂O (3) are shown in Figs. 4–6. For the Sm³⁺ ion ground state, the ⁶H_5/2 multiplet is separated²² from the higher ⁶H_7/2 term by ≈880 cm⁻¹. In the Sm³⁺ aquaion spectrum this separation was evaluated as 1030 cm⁻¹ and seems to be comparable to that in our systems.^20,21 Such a comparison is possible because of the confrontation of the CF splitting of separated levels for the trichloroacetates and the aquaion spectrum. Experimental and theoretical studies of many of the 206 possible multiplet levels in the 4f⁶ configuration have been made by Magno and Dieke,²³ whose experimental results agree with those obtained by Gobrecht.²⁴ The absorption spectra (1000–2600 nm) have been interpreted as an overlap of the terms of the ⁶F state with the upper levels of the ⁶H ground state multiplet. Comparison of our single crystal spectra of the heteronuclear trichloroacetate 1 with data^20,23–25 at ca. 7000 cm⁻¹ reported earlier shows a structure of the lines for the ⁶H_5/2→⁶F_5/2 transition (as assigned in most of the absorption spectra of Sm³⁺ ions) that is too complex if we consider the two non-equivalent sites of the ions in the structure. At 4 K almost all transitions are composed of a sharp doublet structure, indicating two non-equivalent metal centres. However, in the region 7150–6850 cm⁻¹ very complex bands are observed as mentioned above, in which the number of components corresponds to that expected for the ⁶H_5/2→⁶H_15/2 transition. However, from the low values of the matrix elements of the unit tensor operator U^(λ) reported by Carnall et al. for this transition,²⁰ it is necessary to exclude this assignment. Thus, the complex structure of this band can only be explained by strong vibronic coupling with some vibrations observed in the IR and Raman spectra of lanthanide trichloroacetates, reported by us earlier.¹³ The assignment of vibronic components accompanying the ⁶H_5/2→⁶F_1/2 electronic hypersensitive transition in the high-energy range was made (see Fig. 5, spectrum b). These components are promoted by internal ligand modes and correspond to: (1) δ(COO) 361 cm⁻¹, (2) δ(COO) 433 cm⁻¹, (3) ω(H₂O) 488 cm⁻¹, (4) π(COO) 545 cm⁻¹, (5) δ(OCO) 620 cm⁻¹, (6) τ(OCO) 686 cm⁻¹ and (7) δ(H₂O) 757 cm⁻¹ energies calculated from the E₁=6355 cm⁻¹ and E₂=6329 cm⁻¹ zero-phonon lines in the single crystal spectra of Sm₂Cu(CCl₃COO)₈·6 H₂O at 4 K. The correct assignment of the levels in this region of the spectra and an explanation of this phenomenon will be made possible by a CF calculation, which will be the subject of future work. If the bands located at 6355 and 6329 cm⁻¹ correspond to the two components of the ⁶F_1/2 multiplet of the samarium centre (so that the other very weak components observed at 4 K in this region belong to the ⁶H_15/2 term) and if the adjacent band at 6620 cm⁻¹ corresponds to the ⁶F_3/2 transition, the number of components seems to confirm our assignment. The calculated intensities of the f–f and d–d transitions in the Sm:Cu and Sm trichloroacetates at room and helium temperatures along axis a are collected in Table 2. The results of the Judd–Ofelt²⁶ parameter calculation are also included. Estimation of the parameters was made with relatively good accuracy and relatively low estimated error. These parameters reproduce quite well the intensities of the Sm³⁺ f–f transitions in compound 1. In the 500–1000 nm spectral range the copper d–d band overlaps with the Sm³⁺ transitions. This wide band with no vibronic structure splits into two components at 4 K as a result of the Jahn–Teller (J–T) effect in the deformed octahedron of the Cu²⁺ environment. Although the energy of the split d–d bands corresponds quite well to that of similar components in the spectra of Nd:Cu and Pr:Cu trichloroacetates reported by us,^27,28 their intensities vary. They are highest for the samarium spectra (P_exp=665.71×10⁻⁶).

Fig. 4 Absorption spectrum of a Sm₂Cu(CCl₃COO)₈·6 H₂O crystal at 293 K. The insert shows the expanded spectrum between 320 and 540 nm.

Fig. 5 Absorption spectra of a Sm₂Cu(CCl₃COO)₈·6 H₂O crystal at 4.2 K. Spectra (a) and (b) show expanded regions of the full spectrum in (c).

Fig. 6 Absorption spectrum of a Sm(CCl₃COO)₃ ·2 H₂O crystal at 293 K.

Table 2 The oscillator strengths of the f–f (×10⁸) and d–d (×10⁶) transitions for Sm: Cu and Sm trichloroacetate crystals at different temperatures

Sm2Cu(CCl3COO)8·6 H2O		Sm(CCl3COO)3 ·2 H2O
λ/nm	293 K	4.2 K	293 K	4.2 K
a The τλ·10^9 parameters: τ2=6.80±3.20; τ4=13.12±1.02; τ6=8.94±0.95, for Sm2Cu(CCl3COO)8·6 H2O.
^6H5/2→^6F1/2, ^6H15/2	1786–1538	58.4	58.30	46.55	45.10
^6H5/2→^6F3/2	1538–1471	161.72	236.42	158.72	21.36
^6H5/2→^6F5/2	1471–1316	366.69	422.74	344.04	—
^6H5/2→^6F7/2	1316–1163	383.98	415.46	289.18	—
^6H5/2→^6F9/2	1136–1020	235.14	352.47	239.58	—
^6H5/2→^6F11/2	980–909	21.75	42.49	30.56	—
^6H5/2→^4G7/2	505–493	4.10	14.94	6.84	5.55
^6H5/2→^4I9/2, ^4M15/2, ^4I11/12	493–469	142.95	287.59	162.04	195.10
^6H5/2→^4I13/2	469–455	71.34	97.10	76.58	62.77
^6H5/2→^4F5/2	455–450	2.56	5.27	1.58	3.06
^6H5/2→^4M17/2, ^4G9/2, ^4I15/2	450–427	35.2	67.92	35.87	46.65
^6H5/2→^4L13/2	427–405	80.71	137.26	75.65	91.76
^6H5/2→^4F7/2, ^6P3/2, ^4K11/2, ^4M21/2	405–392	466.21	382.89	427.04	406.07
^6H5/2→^4L15/2, ^4G11/2	392–382	34.76	113.51	36.43	39.58
^6H5/2→^4D1/2, ^6P7/2, ^4L17/2, ^4K13/2, ^4F9/2	379–368	237.10	425.53	247.04	240.47
^6H5/2→^4D3/2, (^4D, ^6P)5/2	368–356	129.31	283.40	127.84	136.50
^6H5/2→^4H7/2	356–350	5.90	—	3.57	—
^6H5/2→^4K15/2, ^4H9/2, ^4D7/2,	350–336	130.03	346.30	175.14	176.22
(^4K, ^4L)17/2, ^4L19/2, ^4H11/2
d–d Cu	500–1200	665.71	667.28

---

Since the first excited state lies at relatively low energy, it can be partially populated at room temperature. We have estimated this population to be 1.3%. Therefore, the population of this term must affect both the spectroscopic and magnetic properties of the systems under investigation. In fact, with a decrease in temperature, the intensities of almost all transitions increase, similarly to the europium single crystal spectra. These results are in contrast to those for the other lanthanide trichloroacetates for which a decrease of intensities is observed at low temperatures.^11,13

Magnetic properties

Gd³⁺and Gd:Cu systems.. For most of the trivalent rare earth ions the ^2S+1L_J free ion ground state is well separated in energy from the first excited state, so that only this ground state is thermally populated at room and low temperatures. For mononuclear systems (in the free ion approximation) magnetic susceptibilities can be described by the relation²⁹

with T being the temperature, g the Zeeman factor and λ the spin-orbit coupling parameter. When considering the magnetic properties of lanthanide compounds one must remember the CF effect and the possible thermal population of the higher states for some ions (Sm and Eu).

In the polynuclear and heteronuclear systems ferromagnetic and antiferromagnetic ordering can be expected at low temperature. In fact, for 4f¹ through 4f⁶ configurations of Ln(III) ions the angular and spin momenta can lead to antiparallel ordering in the ^2S+1L_J free ion ground state. Parallel alignment of the Ln(III) and Cu(II) spin momenta can lead to an antiparallel alignment of the angular momenta and to an overall antiferromagnetic interaction. On the other hand, for 4f⁸ through 4f¹³ configurations angular and spin momenta are parallel in the ground state (J=L+S) and a parallel alignment of the Ln(III) and Cu(II) spin momenta may lead to parallel ordering of the magnetic moments, that is to a ferromagnetic interaction. In this case also, the CF effect complicates the situation.

In our earlier reports we have discussed the spectroscopic and magnetic data of two series of Ln and Ln:Cu trichloroacetates (Ln=Nd, Pr). Here we present results for the gadolinium and samarium polynuclear and heteronuclear analogues. In the first system, Gd(CCl₃COO)₃·2 H₂O (4), a simple model can be analysed. Since the Gd³⁺ ion has a ⁸S_7/2 ground state, which is almost unsplit (0.1 cm⁻¹) and is located ≈32000 cm⁻¹ below the first excited state, the magnetic moment is given by the spin momenta only. The second term of eqn. (1) is negligible and the magnetic susceptibility obeys the Curie law: χ_M·T=21Nβ²/k. However, our results are not consistent with the free ion approximation nor the reasoning and results reported for dimeric gadolinium systems by Kahn and coworkers.³⁰ This is understandable because, in contrast to the earlier data reported by Kahn and coworkers,^30,31 for both types of gadolinium chloroacetates, Gd₂Cu(CCl₃COO)₈·6 H₂O (2) and Gd(CCl₃COO)₃·2 H₂O (4), polymeric chains are formed with dimeric subunits, which are further linked to each other by the Cu ion, or in 4, by carboxyl groups and water molecules. They differ also from other heteronuclear systems investigated by Kahn et al.³²

χ_M and χ_M· T vs. T for the gadolinium systems 2 and 4 are plotted in Fig. 7. For the polynuclear system 4 χ_M ·T exhibits what looks like antiferromagnetic ordering at low temperature. At very low temperatures the dramatic decrease in χ_M·T could also be a result of a very small splitting of the ⁸S_7/2 multiplet at zero field or additional splitting of the ground state multiplet promoted by a Gd(III)–Gd(III) inter action. On the other hand, for the heteronuclear system Gd:Cu ferromagnetic ordering can be considered, but in this case also the relation of χ_M·T is not comparable to the previously reported results, most probably for the same reason as in the polymeric single crystals of compound 4 since the heteronuclear chain is composed also of Gd–Gd dimer units (see Fig. 1). Similar results and relations of χ_M·T for Cu:Gd heteronuclear compounds were reported by Benelli and co-workers.^33,34 in drastically different heteronuclear chelates.

Fig. 7 Experimental (■) magnetic susceptibility χ_M and (–■–) χ_M·T calculated per molecule in the complex plotted vs. temperature. (a) Gd (CCl₃COO)₃ ·2 H₂O, (b) Gd₂Cu(CCl₃COO)₈·6 H₂O.

Sm³⁺ and Sm:Cu systems.. The Sm³⁺ ion belongs to the first part of the lanthanide series and as we have demonstrated in the spectroscopic section, the ⁶H ground term is split by spin-orbit coupling into six levels. The first excited ⁶H_7/2 term (as was estimated by us) is separated by ca. 890–940 cm⁻¹ from the ⁶H_5/2 ground state and its population is 1.3% at 300 K. Similarly to Eu(III), E(J) and χ(J)T decrease with J. Therefore, for compound 1 χ_MT decreases as T decreases (the plot is nearly linear for T=50 to 300 K) and leads to a value of 0.36 cm³ mol⁻¹ K as T approaches 6.41 K and drops dramatically with further decrease of temperature to 1.7 K, giving a value of 0.035 cm³ mol⁻¹ K (Fig. 8). The plot of χ_Mvs. T shows that the magnetic susceptibility slowly increases below 160 K and then dramatically increases below 30 K to reach a value of 49.44×10⁻³ at 6.41 K, then decreases to 20.61×10⁻³ at 1.7 K (see also Table 3). This type of behaviour points to antiferromagnetic ordering at low temperatures. Such ordering of spins can be confirmed by the relation of 1/χ_Mvs. T. In the 300–40 K region it obeys the Curie–Weiss law with a Weiss constant of Θ=−20.04.

Fig. 8 Experimental (■) magnetic susceptibility χ (–■–) and χ·T calculated per molecule in the complex plotted vs. temperature. (a) Sm(CCl₃COO)₃ ·2 H₂O, (b) Sm₂Cu(CCl₃COO)₈·6 H₂O.

Table 3 Magnetic data for Sm₂Cu(CCl₃COO)₈ ·6 H₂O. C=0.705; Θ=−20.04

T/K	χM·10^3/cm^3 mol^−1	μeff/μB
1.73	20.61	0.53
2.75	21.57	0.68
3.50	22.85	0.80
4.25	25.66	0.93
5.44	43.60	1.38
6.41	49.44	1.59
7.97	44.89	1.69
9.09	40.20	1.71
10.50	35.22	1.72
20.06	19.90	1.78
52.94	8.68	1.91
100.19	5.05	2.01
160.31	3.58	2.14
220.31	2.89	2.26
300.27	2.42	2.41

---

Using the free ion approximation, Kahn and colleagues³¹ derived the relation of χ_M for the Sm³⁺ ion in monomeric systems, in which χ_M can be given by the following equation for the six states arising from the ⁶H ground state multiplet:

where χ(J) is given by eqn. (1) and E(J)=λ[J(J+1)−35/4]/2. Kahn et al. obtained excellent theory–experiment agreement. At low temperatures χ_MT reaches the value of 0.89 cm³ mol⁻¹ K predicted by theory. Our results for the heteronuclear 1 system show completely different behaviour for χ_M and χ_MT vs. temperature (Fig. 8, Table 3) in comparison to the monomeric systems reported by Kahn et al.³¹

EPR investigations (300–4 K) of the heteronuclear Ln:Cu trichloroacetates show Ln: Cu interactions larger than for Ln: Ln systems.^35,36 In addition, in the Pr:Cu trichloroacetates Cu:Cu interactions appear through the dimeric unit of the lanthanide ions. This was explained by a superexchange interaction involving the s and d orbitals of the Pr³⁺ ion.^35–37

Summary

Two types of single crystals, heteronuclear Sm₂Cu(CCl₃COO)₈·6H₂O (1) and Gd₂Cu(CCl₃COO)₈ ·6H₂O (2) and homonuclear Sm(CCl₃COO)₃·2H₂O (3) and Gd(CCl₃COO)₃·2H₂O (4) trichloroacetates were obtained. Their optical and magnetic properties were investigated in the 1.7–293 K range.

Their optical properties indicate marked differences in the structures of these two types of single crystals, with transformation of centro- to noncentrosymmetric lanthanide dimeric subunits.

The f–f transition probabilities were analysed and the Judd–Ofelt parameters calculated with a relatively small error.

Vibronic components were evaluated and assigned on the basis of IR spectra.

The magnetic moments of 1 and 3 decrease with temperature. This can be due to the crystal field effect and antiferromagnetic interactions of pairs of ions in compounds 1 and 3. Magnetic susceptibilities increase with decreasing temperature and ferromagnetic ordering seems to occur in compound 2.

References

1. O. Kahn, Struct. Bonding (Berlin), 1987, 68, 89.
2. G. M. Villacorta and S. J. Lippard, Inorg. Chem., 1987, 26, 3672.
3. J. M. Lehn, Angew. Chem., Int. Ed. Engl., 1988, 27, 89.
4. J. Legendziewicz, in Excited States of Transition Elements, World Scientific Publishing Co. Pte. Ltd., Singapore, 1992, pp. 149–179..
5. G. Blasse, Chem. Mater., 1989, 1, 294.
6. R. Reisfeld and C. Jørgensen, Laser and Excited States of Rare Earth, Springer, Berlin, 1977..
7. J. Legendziewicz, Acta Phys. Pol., 1996, 90, 127.
8. J. P. Costes, F. Dahan, A. Dupuis and J. P. Laurent, Inorg. Chem., 1977, 36, 3429.
9. J. P. Costes, F. Dahan, A. Dupuis and J. P. Laurent, Chem. Eur. J., 1998, 9, 1617.
10. I. Kutlu, G. Meyer, G. Oczko and J. Legendziewicz, Eur. J. Solid State Inorg. Chem., 1997, 34, 231.
11. J. Legendziewicz, G. Oczko and G. Meyer, Polyhedron, 1991, 10, 1921.
12. T. Imai and A. Ouchi, Bull. Chem. Soc. Jpn., 1987, 60, 408.
13. J. Legendziewicz, G. Oczko, W. Strek and J. Hanuza, Eur. J. Solid State Inorg. Chem., 1995, 32, 95.
14. W. T. Carnall, P. R. Fields and K. Rajnak, J. Chem. Phys., 1968, 49, 4443.
15. G. H. Dieke, Spectra and Energy Levels of Rare Earth Ions in Crystals, Interscience Publishers, New York, 1968, p. 250..
16. L. Couture, J. Lumin., 1979, 18/19, 891.
17. M. J. Weber, in Laser Spectroscopy of Solids, ed. W. M. Yen and P.M. Selzer, Springer-Verlag, Berlin, 1986, ch. 6. p. 189..
18. T. Tsuboi and S. Augusto, J. Phys. Conden. Matter, 1998, 10, 2787.
19. R. L. Schwiesow and H. M. Crosswhite, J. Opt. Soc. Am., 1969, 59, 592.
20. W. T. Carnall, P. R. Fields and K. Rajnak, J. Chem. Phys., 1968, 49, 4412 and 4424..
21. R. D. Peacock, Struct. Bonding (Berlin), 1975, 22, 83.
22. M. Malinowski, R. Wolski, Z. Srukacz, T. Łukasiewicz and N. Z. Łuczyński, J. Appl. Spectrosc., 1995, 62, 49.
23. M. S. Magno and G. H. Dieke, J. Chem. Phys., 1962, 37, 2354.
24. H. Gobrecht, Ann. Physik, 1938, 31, 755.
25. W. T. Carnall, D. M. Gruen and R. L. McBeth, J. Phys. Chem., 1962, 2159.
26. B. R. Judd, Phys. Rev., 1962, 127, 750.
27. G. Oczko, J. Legendziewicz, J. Mrozinski and G. Meyer, J. Alloys Compd., 1998, 275–277, 219.
28. J. Legendziewicz, M. Borzechowska, G. Oczko and J. Mrozinski, Spectrochim. Acta Part A, 1998, 54, 2197.
29. R. L. Carlin, Magnetochemistry, Springer Verlag, Berlin, 1986..
30. A. Panagiotopolus, T. F. Zafiropoulos, S. P. Perlepes, E. Bakalbasis, J. Masson–Ramade, O. Kahn, A. Terzis and C. Raptopoulu, Inorg. Chem., 1995, 34, 4918.
31. M. Andruh, E. Bakalbassis, O. Kahn, J. C. Trombe and P. Porcher, Inorg. Chem., 1993, 32, 1616.
32. O. Kahn, O. Guillou, R. L. Oushoorn, M. Drillon and P. Rabu, New J. Chem., 1995, 19, 655.
33. C. Benelli, A. Caneschi, D. Gatteschi and L. Pardi, Mater. Chem. Phys., 1992, 31, 17.
34. C. Benelii, T. Guerriero, S. Tanburini and P. A. Vigato, Mater. Chem. Phys., 1992, 31, 137.
35. Y. Yablokov, V. Voronkova, J. Legendziewicz, G. Oczko and M. Borzechowska, Appl. Magn. Reson., in press..
36. J. Legendziewicz, J. Alloys Compd., in press..
37. V. K. Voronkova, Y. V. Yablokov, J. Legendziewicz, G. Oczko, M. Borzechowska, Solid State Phys., 1999, 41, 41 (in Russian)..

---