The Synthesis, Crystal Structure, and Magnetic Properties of Mono-Scorpionate Eu(III) Complexes

Abstract

Three Eu³⁺ complexes containing a neutral tripodal ligand possessing a predictable coordination mode have been obtained and studied. The trispyrazolylmethane complexes have an aqua ligand in the coordination sphere, forming both the mononuclear species [Eu(HCPz₃)H₂O(NO₃)₃] and the dimer [Eu(HCPz₃)H₂O(CF₃SO₃)₃]₂, having a Chinese lantern structure, whereas the use of the methylated tripod leads to the water-free complex, [Eu(HC(Pz^Me₂)₃)(NO₃)₃]. A qualitative analysis of the magnetic susceptibility of polycrystalline samples demonstrated that the magnetic properties can be described using a simple Van Vleck formula with spin–orbit coupling parameter (λ = 383 ÷ 406 cm⁻¹) close to the values for free Eu³⁺ ions. The stereochemical analysis of the coordination environment of [Ln(HC(Pz^Me₂)₃)(NO₃)₃] has shown that these complexes can be used as diamagnetic model systems to obtain information on the crystal field effects in the paramagnetic monoradical complexes, [LnRad(NO₃)₃], since both types of compounds have the same type of coordination polyhedron (symmetry point group D_3h) and very close Ln–donor atom distances.

1. Introduction

The design of molecular materials with newly developed functionalities holds the potential for technological advances. Materials based on molecules that switch under the action of various stimuli will form the basis for the creation of multifunctional devices that combine optical, magnetic, and other electronic properties, as well as extreme miniaturization. For these purposes, understanding what the fundamental properties of molecular coordination compounds depend on, and how they can be controlled, is very important for the creation of new functional molecular materials.

The unique electronic and physical properties of rare-earth ions are used for various applications, including the development of magnetic and luminescent materials. In this direction, the main attention is paid to the design of luminescent single-molecule magnets (SMMs), which requires the development of rational approaches to the search for compounds with strong magnetic anisotropy and enhanced photoluminescence [1,2].

The main candidates for the role of SMMs are lanthanide (Ln) complexes, in which the metal ion is in the trivalent state and the redox activity is usually limited to the ligand. However, lanthanide-based redox activity is also possible. Among the lanthanides, europium possesses the most reachable divalent oxidation state, Eu²⁺, because of its half-filled 4f⁷ electronic configuration [3,4]. For this reason, a vast majority of the magnetically studied Eu³⁺ molecular complexes contain some amount of Eu²⁺ (S = 7/2) impurities, which can be easily detected through low-temperature magnetic susceptibility measurements [5,6]. To date, there is abundant information on the photophysical properties of Eu³⁺ molecular complexes, which are quite well studied spectroscopically. At the same time, because the ground-state multiplet ⁷F₀ of Eu³⁺ ions is non-magnetic (S = L = 3; J = S − L = 0), researchers often leave the magnetic properties of europium luminescent compounds unattended. However, owing to the excited multiplets that are located closely in energy to the ground-state one, Eu³⁺ ions exhibit a considerable magnetic activity associated with Van Vleck paramagnetism at low temperatures and the thermal population of the excited states that is already manifested at moderate temperatures, with the effective magnetic moment reaching a value of ~3.2 µ_B at room temperature. In its turn, the low redox potential of the Eu²⁺/Eu³⁺ pair [7] in combination with non-innocent organic ligands can lead to interesting materials both photo- and magnetically switchable, including SMMs [8,9].

Tuning the electronic properties of a compound is the most common challenge for materials scientists. This can be achieved by varying the ligand environment of a central atom, namely the polyhedron geometry and the ligand field strength. The latter is mainly determined by nature of the donor atoms, whereas the former depends not only on the ligand geometry but also on its coordination manner. Organic tripod molecules with donor atoms on each of the three legs are most promising for the chemical construction of various complexes with a predictable geometry. Moreover, the latter can be easily adjusted using methods of synthetic organic chemistry.

Previously, we have obtained the monoradical [LnRad(NO₃)₃] (Ln³⁺ = Gd, Tb, Dy, Tm [10] and biradical [Ln(Rad)₂(OTf)₃] (Ln³⁺ = Eu and Dy) [11] complexes, where Rad is a tripodal nitroxyl radical (Scheme 1a), 4,4-dimethyl-2,2-bis(pyridin-2-yl)-1,3-oxazolidine-3-oxyl. Upon comparing the magnetic susceptibility (χ_MT) measured for the biradical Eu complex (1.78 and 0.348 emuK/mol at 300 and 2 K, respectively) [11] with the values known for the [Eu(radical)₂(anion)₃] compounds [12,13,14,15,16,17,18], we noticed a remarkable dispersion of χ_MT in the latter, varying from 1.89 to 3.09 emuK/mol at room temperature, and from 0.035 to 0.42 emuK/mol at low temperatures. Correspondingly, the theoretical estimates of the magnetic parameters for Eu³⁺ biradical complexes that are reported in the literature resulted in a surprisingly large scatter in the values of the spin–orbit coupling (SOC) parameter, λ. In turn, in the case of monoradical [Eu(radical)(anion)₃] complexes, there is no reasonable explanation for the deviation of their magnetic susceptibility, measured at low temperatures (χ_MT = 0.192–0.252 emuK/mol) [19,20,21], from the value of 0.375 emuK/mol, naturally expected for radicals with S = 1/2 non-interacting with the Eu³⁺ ion’s singlet state. The observed discrepancies point to an insufficient understanding of the interaction between the Eu³⁺ ions with strong SOC and paramagnetic ligands.

In order to clarify the nature of Eu³⁺–Rad interaction one can compare, for example, the complexes with paramagnetic and diamagnetic ligands, [EuRad(NO₃)₃] and [EuL_dia(NO₃)₃]. The latter is a model complex containing a diamagnetic tripodal ligand with a coordination polyhedron geometry that is close to that of the radical complex. Such a strategy of diamagnetic substitution was used for obtaining information on the crystal field (CF) effects in paramagnetic Ln ions [15,22,23]. For instance, for two homologous series of coordination compounds, [LnNN^trz₂(NO₃)₃] and [Ln(Nitrone)₂(NO₃)₃] [15], and [Ln(HBPz₃)₂SQ] and [Ln(HBPz₃)₂Trp] [22,23], where NN^trz is 2-(4′,5′-dimethyl-1′,2′,3′-triazolyl)-4,4,5,5-tetramethyl-4,5-dihydro-1H-imidazolyl-1-oxy-3-oxide, and SQ is 3,5-di-tert-butylsemiquinonato, respectively, while the diamagnetic homologue ligands, Nitrone and Trp, are 3-N-tert-butylnitrone-4,5-dimethyltriazole and tropolonate, correspondingly. In the present case of Eu complexes, this experimental approach should help to elucidate the nature of coupling between the radical and Eu³⁺ ion by subtracting the CF splitting effect.

In this paper, we have synthesized and magnetically studied three Eu³⁺ complexes comprising diamagnetic tripodal pyrazolyl-based ligands (see Scheme 1b,c) to determine the reference values of the magnetic susceptibility for such Eu(III) complexes, and to establish a ground for further studies of the Eu³⁺–Rad interactions in complexes with a similar geometry of the coordination polyhedron.

2. Short Theoretical Background

Trivalent europium ions (Eu³⁺) exhibit intense red photoluminescence upon irradiation with ultra violet light [24]. This luminescence is observed not only for Eu³⁺ ions doped into crystalline host matrices or glasses, but also for europium(III) complexes with organic ligands [25]. Such a ligand can drastically enhance the photoluminescence quantum yield by acting as an “antenna” to absorb the excitation light and to transfer the energy to the Eu³⁺ ion, thereby bringing about the population of excited Eu³⁺ energy levels, including the emitting one [2]. Not only its red luminescence, but also the narrow transitions in the luminescence spectra are typical features of the Eu³⁺ ion [26], and these spectroscopic properties have been known since the earliest history of the chemical element europium [27]. Many Eu³⁺ compounds show an intense photoluminescence due to the ⁵D₀ → ⁷F_J transitions (J = 0–6) from the ⁵D₀ excited state to the J levels of the ground term ⁷F. The fine structure and the relative intensities of the transitions in the luminescence spectra of Eu³⁺ can be used to probe the local surroundings of the metal center. Spectroscopic experimental data provide information on the point group symmetry of the Eu³⁺ site, as well as on the polyhedron geometry [26], which is especially useful for solution studies. Other techniques can also be used for the determination of the CF energy levels inside the 4f shell, including two-photon absorption, Zeeman spectroscopy, and magnetic circular dichroism [24].

To a first approximation, a lanthanide central atom in a molecular coordination compound behaves as an Ln³⁺ free ion, for which the paramagnetic behavior is due to its unpaired 4f electrons. As these electrons are shielded by the outer electronic closed shell (2p), the Coulomb electron–electron interaction and spin–orbit coupling between 4f electrons are much stronger than the ligand field, and hence, the former plays a decisive role in the electronic level structure and magnetic properties. Depending on the sign of the spin–orbit interaction, the total magnetic moment (J) of a Ln³⁺ ion is formed as J = L ± S, where L is the angular momentum and S is the spin momentum. For the majority of Ln³⁺ ions, the ^2S+lL_J ground-state (GS) multiplet is well-isolated in energy from the first excited one. Then, only the GS is thermally populated in the temperature range of 0–300 K. In the free ion approximation, the molar magnetic susceptibility for a mononuclear species is given in Equation (1) [28]:

$${\chi }_J=\frac{N{g}_J^2{\beta }^{2}J\left(J+1\right)}{3kT}+\frac{2N{\beta }^2\left(g_J-1\right)}{3\lambda }$$

(1)

where λ is the spin–orbit coupling parameter and g_J is equal to

$$g_J=\frac{3}{2}+\frac{\left[S\left(S+1\right)- L\left(L+1\right)\right]}{2J\left(J+1\right)}$$

(2)

For J = 0, there is obviously no first-order Zeeman splitting. However, the application of a magnetic field can result in second-order splitting and the appearance of a net magnetic moment. The last part in (1) is a temperature-independent contribution arising from the coupling between the ground and excited states. For Ln molecular compounds, the χ_MT versus T curve often differs slightly from what is predicted by (1). This deviation occurs for two reasons: the crystal field partially removing the 2J + 1 degeneracy of the GS in zero-field, and the thermal population of the Ln³⁺ excited states. The latter is most pronounced for tri-positive Sm and Eu, which possess excited states that are located closely in energy to the ground one. In the case of Eu³⁺, its ⁷F ground term is splitted by SOC into seven multiplets (⁷F₀, ⁷F₁, ⁷F₂, ⁷F₃, ⁷F₄, ⁷F₅, ⁷F₆) with energies of E(J) = λJ(J + 1)/2, where the energy of the ⁷F₀ ground state is taken as zero. Since λ is small enough for the excited states to be thermally populated, the magnetic susceptibility may be written as

$${\chi }_M=\frac{{{\displaystyle \sum }}_{J=0}^6\left(J+1\right){\chi }_J{e}^{-\frac{\mathsf{\lambda }J\left(J+1\right)}{2kT}}}{{{\displaystyle \sum }}_{J=0}^6\left(2J+1\right){\chi }_J{e}^{-\frac{\mathsf{\lambda }J\left(J+1\right)}{2kT}}}$$

(3)

where χ_J is given in Equation (1). In the case of Eu³⁺, all of the g_J are equal to 3/2, except g₀, which is equal to 2 + L(=2 + S) = 5, i.e., (χ_MT)_Eu can be expanded as [29]

$$\frac{\frac{N{\beta }^2}{3kx}\left[24+\left(\frac{27}{2}x-\frac{3}{2}\right)e^{-x}+\left(\frac{135}{2}x-\frac{5}{2}\right)e^{-3x}+\left(189x-\frac{7}{2}\right)e^{-6x}+\left(405x-\frac{9}{2}\right)e^{-10x}+\left(\frac{1485}{2}x-\frac{11}{2}\right)e^{-15x}+\left(\frac{2457}{2}x-\frac{13}{2}\right)e^{-21x}\right]}{\left[1+3e^{-x}+5e^{-3x}+7e^{-6x}+9e^{-10x}+11e^{-15x}+13e^{-21x}\right]}$$

(4)

where x = λ/kT.

At moderate temperatures, owing to the value of λ~350 cm⁻¹ [28], only the first three lower states (⁷F₀, ⁷F₁, ⁷F₂), with energies of 0, λ, and 3λ, can be considerably populated; since the ⁷F₀ ground state is formally nonmagnetic, the lower limit of χT is zero. However, the low-temperature limit of the magnetic susceptibility, χ_LT, is non-zero due to the temperature-independent contribution arising from the coupling between the ground and excited states. According to Equation (4), which acquires a very simple form in the T = 0 limit, the χ_LT value is directly related to λ as

χ_LT = 8Nβ²/λ = 2.085/λ (λ in cm⁻¹)

(5)

Although the aforementioned description has been proven to capture the main qualitative features of the magnetic susceptibility of europium(III)-based compounds, its performance turns out to be not as good when it comes to the precise determination of λ and quantitative comparison with spectroscopic results. The problem stems from the CF splitting of the excited J ≠ 0 multiplets, which is not taken into account in Equation (4) [29,30,31]. Given that the ⁷F₁ multiplet is the closest one in energy to the ground level, its splitting into three Stark sublevels plays the most important role. For example, when one considers the low-temperature magnetic susceptibility and allows for the splitting of the first excited multiplet, Equation (5) turns into

χ_LT = (8Nβ²/3)(1/E₁ + 1/E₂ + 1/E₃)

(6)

where E₁, E₂, E₃ are the energies of the ⁷F₁ multiplet sublevels. Apparently, when the multiplet splitting is negligible, Equations (5) and (6) are equivalent, as 1/E₁ = 1/E₂ = 1/E₃ = 1/λ. But, in the opposite case of strong splitting, the average (1/E₁ + 1/E₂ + 1/E₃)/3 becomes significantly larger than 1/λ, resulting in the magnetic susceptibility in the low-temperature plateau region to be enhanced compared to the values predicted by Equation (4) [31].

It is worth noting that as long as a simple model is used, which implies that the ⁷F term splitting into multiplets is caused solely by the spin–orbit coupling, λ should be considered as an “effective” SOC parameter. Such an effective λ absorbs, along with the true SOC, other effects related to the fine level structure due to the LF. In fact, it is the effective parameter, λ, that is usually found by empirical analyses of magnetic or spectroscopic data obtained for Eu(III) complexes.

3. Results and Discussion

3.1. Synthetic Aspects

As can be seen in Scheme 1, the paramagnetic tripod (a) and its diamagnetic relatives (b,c) have a similar organic molecule organization. Formally, Rad can be classified as a methane derivative, like the other two tripodal ligands, since, for all three molecules, the bridgehead atom is an sp³ carbon. In addition, despite the difference in the donor atoms on the “legs” of the paramagnetic tripod (oxygen atom from the nitroxyl group instead of pyrazole nitrogen), Rad also forms three six-membered chelate cycles when coordinated.

Upon the coordination of the tris(pyrazol-1-yl)methanes (neutral tripods), anionic ligands are required to compensate for the positive charge of the central atom [32,33], in contrast to their negatively charged congeners, tris(pyrazol-1-yl)borates (TPBs), which form the neutral complexes Ln^III(TPB)₃ [34,35,36,37,38] and Ln^II(TPB)₂ [39,40,41].

The interaction of Eu(NO₃)₃·5H₂O with a tripodal ligand in a 1:1 ratio in an acetonitrile solution leads to the formation of an aqua complex, [Eu(HCPz₃)(NO₃)₃H₂O]·2MeCN (1), when reacting with unsubstituted trispyrazolylmethane tripod, (HCPz₃), and the water-free nitrate compound, [Eu(HC(Pz^Me₂)₃)(NO₃)₃]·MeCN (2), upon the coordination of the methylated congener—tris(3,5-dimethyl-1-pyrazolyl)methane. The absence of an aqua–ligand in 2 is most likely due to the steric hindrance of the Pz^Me₂–tripod. It is worth noting that, for the chloride analogue, [Ln(HC(Pz^Me₂)₃)Cl₃]·2MeCN, no entry of water molecules into the coordination sphere was also detected upon the reaction in acetonitrile under normal conditions [42], whereas the interaction of hydrated salt, EuCl₃·6H₂O, with HCPz₃ in tetrahydrofuran (THF) results in the six-aqua complex, [EuHC(Pz)₃(H₂O)₆]Cl₃ [25].

The seven-coordinate mono-tripod complexes, [Ln(HC(Pz^Me₂)₃)(OTf)₃THF], where Ln = Y, Ho, Dy were formed in a dry and inert tetrahydrofuran medium [32]. In open air, the reaction in acetonitrile media of Eu(OTf)₃ and HCPz₃ in a 1:1 ratio results in the dimerized entity [Eu(HCPz₃)H₂O(OTf)₃]₂ (3), in which the Eu ions are also seven-coordinate. The crystals of 3 contain the acetonitrile and ether solvate molecules, which escape easily during storage. Interestingly, upon using unsubstituted anionic hydro-tris(pyrazolyl)borate (HBPz₃) in THF under an inert atmosphere, the dimeric species [Ln(HBPz₃)₂(OTf)₃]₂ were also obtained for Y, Eu, in which the two triflates are bridging. In contrast, the molecular complexes of Gd and Yb are monomeric THF adducts, [Ln(HBPz₃)₂(OTf)₃] [43].

3.2. Description of the Crystal Structures of 1–3

A summary of the crystallographic data and the structure refinement is given in Table S1 (Supplementary Materials). As established through an X-ray diffraction experiment, the mononuclear compounds crystallize in the following space groups: P–1, I2/c, and P2₁/n for 1, 1a, and 2, respectively, as well as the binuclear species 3 in P2₁/c. In all compounds, each europium ion is capped with one neutral tripodal ligand. Among the structurally studied heteroleptic complexes, only 1a does not have solvate molecules; 1 differs from 1a by the presence of solvate acetonitrile molecules in its crystal structure. In both compounds, the europium center has a coordination number of ten, comprising three nitrogens of trispyrazolylmethane, six oxygens of bidentate nitrate anions, and one aqua–oxygen; see Figure 1.

The Eu–N distances of 1 are in the range of 2.5810(12) to 2.5932(12) Å, which are somewhat larger than those for nitrates (2.4846(10)–2.5687(11) Å), the length of the Eu–O_aqua bond of 2.4191(11) Å being the shortest one (

Table 1. The coordination polyhedron bond and Eu–Eu distances (Å) in the studied compounds.

	1	1a	2	3
			Eu1 1	Eu2 1
Eu–N	2.5810 (12)	2.5678 (16)	2.5402 (19)	Eu1–N1	2.520 (8)	Eu2–N7	2.533 (8)
	2.5926 (12)	2.5180 (17)	2.5592 (19)	Eu1–N3	2.537 (8)	Eu2–N9	2.542 (9)
	2.5932 (12)	2.6072 (16)	2.5158 (18)	Eu1–N5	2.552 (8)	Eu2–N11	2.552 (9)
mean	2.5889	2.5643	2.5384		5.36		5.42
Eu–Oanion	2.5687 (11)	2.4909 (14)	2.4440 (17)	Eu1–O1	2.378 (6)	Eu2–O11	2.382 (7)
	2.4846 (10)	2.7172 (16)	2.4883 (17)	Eu1–O3i	2.404 (6)	Eu2–O14	2.520 (8)
	2.5153 (11)	2.4288 (13)	2.4718 (17)	Eu1–O4	2.376 (6)	Eu2–O17	2.537 (8)
	2.4655 (11)	2.4983 (15)	2.4465 (17)	Eu1–O7	2.334 (6)	Eu2–O19 ii	2.552 (8)
	2.5234 (11)	2.5306 (15)	2.4674 (17)	mean	2.373	mean	2.498
	2.5005 (11)	2.5199 (15)	2.4514 (17)
mean	2.5097	2.5310	2.4616
Eu–Oaqua	2.4191 (11)	2.4237 (15)		Eu1–O10	2.382 (7)	Eu2–O20	2.360 (8)
Eu–Eu	6.151	6.646	8.931		6.137		6.073

¹ Symmetry codes for compound 3: (i) −x + 3, −y + 2, −z + 2; (ii) −x + 2, −y + 2, −z + 2.

). The same tendency is observed for other compounds. However, compound 2 has the most compact coordination environment and contains a solvated acetonitrile molecule (Figure 2a). According to the ten-coordinate polyhedrons analysis using the SHAPE program [44], the geometry of the lanthanide site in 1 and 1a can be best described as a sphenocorona (C_2v), while for the nine-coordinate environment of 2, a spherical tricapped trigonal prism (D_3h) is more acceptable (Table S2, Supplementary Materials).

Among all studied complexes, the most intriguing is the molecular structure of 3, in which there are two independent Eu³⁺ sites with a coordination number of eight (Figure S1), which is composed of three tripod nitrogens, two μ₂-oxygens of bridging triflates, two oxygens from monodentate anions, and one aqua–oxygen. A duplication of the corresponding independent parts via an inversion center gives two different dimeric molecules, one of which is shown in Figure 2b. In addition to the two bidentate triflates, each binuclear entity is supported by two water–triflate hydrogen bonds, which are in a plane perpendicular to that in which the bridging anions are located, thus forming a Chinese lantern structure. The stereochemical analysis indicates that the coordination polyhedron for the two Eu sites corresponds well to a square antiprism (D_4d) (Figure S2, Table S2, Supplementary Materials). Nevertheless, despite the similar polyhedron geometry, the two Eu sites differ quite significantly in the Eu–ligand bond lengths, which are, on average, shorter for Eu1 (Table 1).

A crystal packing analysis shows that the shortest Eu–Eu distances are in the range of 6.073–8.931 Å (Table 1). These distances have an intermolecular character only for 2, in which no hydrogen bonds were detected, while 3 is a binuclear species, and the compounds 1 and 1a are dimerized due to the formation of a double hydrogen bond (Figure S3, Supplementary Materials). The lengths of the shortest intermolecular contacts between a tripod and anionic ligand, CH^L…O_anion, are 3.281, 3.245, 3.334, and 3.325 Å for 1, 1a, 2, and 3, respectively.

3.3. Magnetic Behavior of The Complexes

The experimental dc magnetic susceptibility data for complexes 1–3, measured in the temperature range from 1.77 to 300 K at the applied magnetic field H = 1 kOe, are shown as χ_MT vs. T plots in Figure 3 (where χ_M is the molar magnetic susceptibility corrected for the diamagnetic contribution). As expected, due to the thermal depopulation of the excited levels with non-zero J values, χ_MT gradually decreases with cooling and tends towards a value that is very close to zero as the temperature approaches absolute zero. The χ_MT values at low and room temperature for all complexes are summarized in

Table 2. Magnetic parameters for europium(III) molecular compounds.

Compound	χMT (emu·K/mol)		λ (cm−1) a	R b	Reference
	2 K	300 K
Free Eu3+	0	1.280	≈380		[26,45,46]
1	0.0113	1.228	406	1.02 × 10−4	this work
2	0.0114	1.251	396	4.10 × 10−5	this work
3/Eu	0.0120	1.255	383	3.26 × 10−4	this work
[Eu2(L1)6(H2O)4]n/Eu	0.025	1.325	343/360 c		[47]
[Eu2(tzf)3(H2O)6]n/Eu	0.020	1.388	344	2.7 × 10−4	[48]
[Eu2L26Pen2(H2O)2]/Eu			404/379 d	3.1 × 10−4	[49]
[EuL3(H2O)2SO4]n			336	4.28 × 10−4	[50]
[EuL43(CH3OH)3]n			360	9.65 × 10−5	[51]
[Eu2(Hpcpa)3(H2O)5]n	sap	1.2	356/351 c	2.2 × 10−5	[52]
[Eu(HLCl)4(dmso)]−		1.37	348	4.1 × 10−5	[53]
[Eu(HLF)4(dmso)]−		1.37	336	2.1 × 10−5	[53]
[EuL52(H2O)4](ClO4)3]			362/370 c	7.0 × 10−4	[29]
[Eu(trippBAP)3]			363		[6]
[EuL6PhCOO](ClO4)2	0.015	1.30	370/410 d		[54]
[Eu(Nitrone)2(NO3)3]		1.25	371.5		[15]
[Eu(HBPz3)2Trp]	0.037	1.48			[22]
[Eu(hfac)3NN e]	0.192	1.946	332	2.0 × 10−3	[19]
Eu(OEP)2−(OEP)−]	0.208	1.575			[20]
[EuIII(AP2−)(ISQ)−]	0.252	1.834			[21]
[Eu(HBPz3)2SQ]	0.28	1.59			[22]
[Eu(hfac)3NN2]SMMTb	0.035	2.24	259		[12]
[Eu(hfac)3PhOMeNN2]	0.11	1.89			[13]
[Eu(hfac)3NN2]	0.12	1.98	309		[14]
[EuNNtrz2(NO3)3]	0.24	1.93			[15]
[Eu(hfac)3NN2]	0.419	3.09			[16]
[Eu(hfac)3NN2]		2.38	175		[17]
[Eu(hfac)3NN2]		2.12	331		[18]
[EuRad2(OTf)3]	0.348	1.78			[11]
[Eu(hfac)3NNEt]n	0.89	1.9			[55]
[Eu(hfac)3NN]n		1.88	267	1.47 × 10−4	[56]
[Eu2(acac)4PhONN2]	0.66	3.47	326	2.67 × 10−5	[57]
[NN4CuL2(Eu(hfac)3)3]	0.38	5.78		1.8  × 10−3	[58]

^a Spin–orbit coupling parameter; ^b agreement factor; ^c found from the optic data at 77 K; ^d found from the optic data at RT; ^e NN—nitronyl nitroxide radical ligand.

. Below ≈100 K, χ_MT varies almost linearly with T (Figure 3), implying that χ_M is nearly temperature-independent at low T. Indeed, presented in Figure S4, the χ_M(T) dependences exhibit a smooth growth upon cooling and tend to saturate, and form a plateau below ~100 K. A small increase in χ_M at the lowest temperatures is due to the inevitable presence of an Eu²⁺ admixture (at a level of 0.01–0.017 at.%) or other rare-earth impurity ions possessing a paramagnetic ground state. The subtraction of this impurity contribution gives the intrinsic temperature-independent magnetic susceptibility of Eu³⁺ ions.

The temperature dependences of χ_MT in Figure 3 can be analyzed using the Van Vleck approximation (4), and assuming a simplified electronic structure of a Eu³⁺ ion composed of seven multiplets with energies of E(J) = λJ(J + 1)/2 [29]. The least squares fitting using Equation (4) with effective λ as the only adjustable parameter (dashed blue lines in Figure 3) results in the λ values of ≈383–406 cm⁻¹ for 1–3 (Table 2)—values typical for Eu³⁺ molecular compounds. A common feature of all these fits is that they underestimate the χ_MT values in the low-temperature region, where χ_M exhibits a plateau. As was discussed in Section 2, this feature is quite expected because Equation (4) does not take into account the zero-field splitting of the excited multiplets with J ≠ 0. The latter inevitably reduces the energy difference between the ground-state J = 0 level and the lowest-energy sublevel of the J = 1 multiplet, E₁ [29,30,31]. In the case of strong zero-field splitting, E₁ may be significantly smaller than λ, thereby providing the low-T magnetic susceptibility to be much larger than χ_LT = 8Nβ²/λ, predicted using Equation (4) [31].

In the entire available temperature range, the field dependences of the magnetization, M(H), demonstrate an almost perfect linear behavior (Figure 4). However, the magnetization value is 2–3 orders of magnitude lower than that of other rare-earth ions with a J ≠ 0 ground state. Even at the lowest temperature, T = 1.77 K, the magnetization at H = 10 kOe reaches only ≈0.01 μ_B per Eu³⁺ ion. These features are common to many other europium(III) complexes with diamagnetic ligands presented in Table 2.

3.4. Spin–Orbit Coupling Parameter Analysis

As can be seen from Table 2, the values of the coefficients of agreement (R) indicate the validity of the chosen theoretical model for all compounds studied by us. However, for the nitrate-containing complexes, 1 and 2, this agreement is better compared to 3. This is most likely due to the significant difference in the splitting of Eu³⁺ excited levels (especially of the ⁷F₁ state) under the action of the ligand field. Indeed, despite approximately the same set of donor atoms—nitrogens from a tripodal ligand and oxygens from the anion ligands—the latter, in compound 3, have a pronounced electron-withdrawing character, whereas nitrate ligands possess this property to a lesser extent, which is confirmed by the difference in Eu–O bond lengths (Table 1). In addition, despite the presence of two different europium sites in 3, the symmetry of their local environment (D_4d) is higher than that of the other two: C_2v for 1 and D_3h for 2 (Table S2, Supplementary Materials). Based on the above, it is not surprising that fitting the experimental curve with the model with only one variable, λ, leads, in the case of compound 3, to a less accurate result compared to 1 and 2. Moreover, for a more precise theoretical description, it might be necessary to take into account the interaction between Eu ions by introducing into the model the coupling constant, j_Eu-Eu. However, without additional experimental data such as the luminescence spectra, the use of five independent variables would lead to overparametrization.

According to the agreement coefficients, R, the fitting quality achieved is sufficient to conclude that the spin–orbit coupling parameters, λ, in 1 and 2 are different (by ~10 cm⁻¹). Indeed, despite the same type of tripod and anionic ligands, the coordination environment of the Eu site is significantly different for 1 and 2, due to the presence of one water molecule in the former. Therefore, 1 has a lower symmetry of the coordination polyhedron compared to 2, which leads to a smaller splitting of the excited levels in the crystal field. This fact is consistent with the higher value of λ: 406 for 1 vs. 396 cm⁻¹ for 2.

4. Experimental Section

4.1. Materials and Physical Measurements

Ln(CF₃SO)₃ salts (Thermo Fisher GmbH, Kandel, Germany) were purchased from Alfa Aesar. Eu(NO₃)₃·6H₂O was prepared upon the dissolution of the corresponding Eu₂O₃ in diluted HNO₃ at 50 °C, followed by crystallization during the slow evaporation of the reaction mixture. Solvents of the reagent grade (EKOS-1, Moscow, Russia) were distilled prior to use. The complexes were synthesized under aerobic conditions. Elemental (C, H, N, S) analyses were carried out using standard methods with a Euro-Vector 3000 analyzer (Eurovector, Redavalle, Italy).

Magnetic measurements were performed using a MPMS-XL SQUID magnetometer (Quantum Design, San Diego, CA, USA) in the temperature range of 1.77–300 K at magnetic fields of up to 10 kOe. In order to determine the paramagnetic component of the molar magnetic susceptibility, χ_MT(T), the temperature-independent diamagnetic contribution, χ_d, and a possible magnetization of ferromagnetic micro-impurities, χ_FM(T), were evaluated and subtracted from the measured values of the total molar susceptibility, χ = M/H. The diamagnetic intrinsic contributions of the compounds, χ_d, were estimated using the Pascal’s constants [59]. Fourier transform infrared (FTIR) spectra were measured in KBr pellets with a Perkin–Elmer System 2000 FTIR spectrometer (Perkin Elmer, Waltham, MA, USA) in the 4000–400 cm⁻¹ range.

4.2. Single-Crystal XRD Experiment and Data Refinement Details

Single-crystal XRD data for crystals of the compounds were collected at 150 K with a Bruker D8 Venture diffractometer (Bruker, Karlsruhe, Germany) (0.5° ω- and φ-scans, fixed-χ three circle goniometer, CMOS PHOTON III detector, IμS 3.0 micro-focus source, focusing Montel mirrors, λ = 0.71073 Å MoK_α radiation, N₂-flow thermostat). Data reduction was performed routinely via APEX 3 suite [60]. The crystal structures were solved using the Bruker SHELXTL (version 2018/3) [61] and were refined using ShelXL [62] programs assisted by Olex2 GUI (Version 1.3.0) [63]. Atomic displacements for non-hydrogen atoms were refined in harmonic anisotropic approximation, with the exception for the partially occupied diethyl ether molecule in 3. Hydrogen atoms were located geometrically, with the exception of those in water molecules in 1 and 1a, which were refined freely with the restraints on the O–H bond. All hydrogens were placed in geometrically idealized positions, and were refined in a riding model. The crystals of compound 3 tend to form twins; as a result, the quality of the structure is lower than that of the others. The structures of the complexes were deposited to the Cambridge Crystallographic Data Centre (CCDC) as a supplementary publication, No. 2285187-2285190.

4.3. Synthesis of the Complexes

[Eu(HCPz₃)(NO₃)₃H₂O]·2MeCN (1): A solution of trispyrazolylmethane (64.5 mg, 0.3 mmol) in acetonitrile (0.5 mL) was added dropwise to a stirred warm solution of Eu(NO₃)₃(H₂O)₅ (128 mg, 0.3 mmol) in acetonitrile (0.5 mL). The next day, the crystalline, colorless precipitate was filtered, washed with a small amount of acetonitrile and Et₂O, and air-dried. Yield: 158 mg (81%) (652.32 g/mol). The storage of the compound resulted in unsolvated 1. Anal. calcd. (%) for C₁₀H₁₂EuN₉O₁₀: C, 21.02; H, 2.12; N, 22.07. Found: C, 21.1; H, 2.0; N, 22.1. IR (KBr): ν (cm⁻¹) 3423 (m), 3115 (m), 3111 (m), 2505 m, 2427 (sh), 1507 (s), 1384 (s), 1317 (s), 1285 (s), 1230 (m), 1148 (s), 1186 (s), 1148 (s), 1136 (s), 1119 (m), 1042 (m), 1028 (m), 1009 (m), 981 (m), 963 (m), 904 (m), 876 (m), 730 (s), 723 (s), 672 (s), 664 (s), 651 (m).

[Eu(HCPz₃)(NO₃)₃H₂O] (1a): The mother liquor and flushing liquids that remain after the synthesis of 1 were combined and evaporated to dryness. The white powder that formed (30 mg) was dissolved in hot dry acetonitrile (2 mL). After filtration, the filtrate was left in a refrigerator at a temperature of 4 C for one week. The colorless crystals were filtered, rinsed with a small amount of cold acetonitrile, and air-dried. Yield: 17 mg (570.22 g/mol); Anal. calcd. (%) for C₁₀H₁₂EuN₉O₁₀: C, 21.02; H, 2.12; N, 22.07. Found: C, 21.2; H, 2.1; N, 22.0. I. IR (KBr) is similar to that of 1.

[Eu(HC(Pz^Me₂)₃)(NO₃)₃]·MeCN (2): To a stirred and hot solution of the metal salt of Eu(NO₃)₃(H₂O)₅) (84 mg, 0.2 mmol) in acetonitrile (1.5 mL) solid tris(3,5-dimethyl-1-pyrazolyl)methane (60 mg, 0.2 mmol) was gradually added under heating. The solution was filtered and kept at room temperature overnight, and then the resulting colorless crystalline solid was sucked off, washed twice with cold acetonitrile, and air-dried. Yield: 112 mg (84%) (677.41 g/mol). The storage of the compound resulted in unsolvated 2. Anal. calcd. (%) for C₁₆H₂₂EuN₉O₉: C, 30.14; H, 3.48; N, 19.65. Found C, 30.2; H, 3.3; N, 19.65. IR (KBr) ν (cm⁻¹): 487 (m); 636 (w); 707 (s); 746 (s); 802 (m); 811 (s); 822 (w); 836 (m); 859 (s); 864 (s); 906 (s); 983 (m); 1024 (s); 1042 (s); 1109 (m); 1153 (w); 1169 (w); 1276 (s); 1304 (s); 1417 (m); 1515 (s); 1530 (m); 1570 (s); 1732 (w); 1771 (w); 1977 (w); 2043 (w); 2268 (m); 2292 (w); 2533 (w).

[Eu(HCPz₃)H₂O(CF₃SO₃)₃]₂·2MeCN·0.73Et₂O (3): A solution of trispyrazolylmethane (54 mg, 0.25 mmol) in acetonitrile (1 mL) was added dropwise to a stirred solution of Eu(CF₃SO₃)₃·(150 mg, 0.25 mmol) in acetonitrile (2 mL). The reaction mixture was halved via heating (65 C) and slowly cooled to room temperature, and then left undisturbed overnight in a closed vessel with a few drops of Et₂O. The crystalline, colorless solid was filtered, washed with a small amount of acetonitrile and Et₂O, and air-dried. Yield: 209 mg (45%) C_32.9H_37.24Eu₂F₁₈N₁₄O_20.73S₆ (1798.66 g/mol). The storage of the compound resulted in unsolvated 3. Anal. calcd. (%) for C₂₆H₂₄Eu₂F₁₈S₆N₁₂O₂₀ (1662.8 g/mol): Anal. calcd. (%): C, 18.78, H, 1.45, N, 10.11; S, 11.7; found: C, 18.6, H, 1.6, N, 10.0; S, 11.5. IR (KBr) ν (cm⁻¹): 3140 (w); 3130 (w); 1578 (m); 1549 (m); 1527 (m); 1409 (m); 1292 (s); 1200 (s); 1166 (s); 1110 (m); 1094 (m); 1029 (m); 960 (m); 920 (m); 889 (m); 857 (m); 824 (m); 810 (m); 755 (s); 722 (m); 638 (s); 597 (m); 574 (m); 517 (m); 410 (w); 389 (w); 351 (w).

5. Conclusions and Perspectives

In this work, europium(III) complexes involving one neutral tripodal ligand with a predictable mode of coordination have been synthesized and structurally studied. In the case of unsubstituted trispyrazolylmethane and hydrated europium nitrate, the complex [Eu(HCPz₃)(NO₃)₃H₂O] has been obtained upon crystallization, of which, from acetonitrile, two different solid phases can be formed, both with (1) and without (1a) solvated molecules in the crystal. In the case of europium(III) trifluoromethylsulfonate, the dimeric coordination compound [Eu(HCPz₃)H₂O(CF₃SO₃)₃]₂ (3) is formed, in which the two anionic ligands are bidentate, and two of the four monodentate triflates form short hydrogen bonds with the aqua ligand of the neighboring Eu center, thus forming a Chinese lantern structure. The replacement of tripod HCPz₃ by its methylated analogue HC(Pz^Me₂)₃ leads to the formation of a nine-coordination complex, [Eu(HC(Pz^Me₂)₃)(NO₃)₃] (2), which does not include an aqua ligand, unlike the complexes of the unsubstituted congener.

For compounds 1–3, dc magnetic measurements were performed in the range of 2 ÷ 300 K at H = 1000 and 10,000 Oe. Based on these data, and using the theoretical modeling of the temperature dependences of magnetic susceptibility, the spin–orbit coupling parameter, λ, was obtained, which varies in the range of 383 ÷ 406 cm⁻¹, only slightly differing from that for the free Eu³⁺ ions.

In the absence of analytical models accounting for strong spin–orbit coupling in Eu³⁺ complexes with paramagnetic ligands, alternative approaches are needed to determine the nature and magnitude of the exchange coupling between paramagnetic centers in such complicated systems. The earlier proposed solution, based on subtraction from the experimental dependence of the magnetic susceptibility, χ_M(T), for a series of lanthanide complexes with radicals [15,23], the same dependence of the model compound of the corresponding Ln with a diamagnetic ligand possessing a similar coordination polyhedron geometry turned out to be particularly effective for [Ln(HBPz₃)₂SQ] [23].

Based on the stereochemical analysis of the coordination environment in compound 2, it was stated that the complexes [Ln(HC(Pz^Me₂)₃)(NO₃)₃] [42] can be used as a model system for the monoradical complexes [LnRad(NO₃)₃], since both types of compounds have the same type of coordination polyhedron (symmetry point group D_3h) and very close Ln–donor atom distances. It is quite possible that the neutral ligand HC(Pz^Me₂)₃ can also be used for the synthesis of bis-tripodal complexes, [Ln(HC(Pz^Me₂)₃)₂](anion)₃, as model systems for biradical compounds with sterically hindered paramagnetic tripods, [Ln(Rad)^Me₂](anion)₃, for which the corresponding ligands are under preparation [64]. As surprising as it may be, the homoleptic Ln³⁺ complexes involving two balky diamagnetic tripods of type HC(Pz^Me₂)₃ in their composition are still unknown. Consequently, developing methods to synthesize them is a particular challenge.