Generation of a Hetero Spin Complex from Iron(II) Iodide with Redox Active Acenaphthene-1,2-Diimine

The reaction of the redox active 1,2-bis[(2,6-diisopropylphenyl)imino]acenaphthene (dpp-BIAN) and iron(II) iodide in acetonitrile led to a new complex [(dpp-BIAN)FeIII2] (1). Molecular structure of 1 was determined by the single crystal X-ray diffraction analysis. The spin state of the iron cation in complex 1 at room temperature and the magnetic behavior of 1 in the temperature range of 2–300 K were studied using Mossbauer spectroscopy and magnetic susceptibility measurements, respectively. The neutral character of dpp-BIAN in 1 was confirmed by IR and UV spectroscopy. The electrochemistry of 1 was studied in solution and solid state using cyclic voltammetry. The generation of the radical anion form of the dpp-BIAN ligand upon reduction of 1 in a CH2Cl2 solution was monitored by EPR spectroscopy.


Introduction
The synthesis of molecular compounds with controlled physicochemical properties [1] is a great task in modern chemistry. For more than 20 years after the discovery of magnetic bistability in the Mn 12 OAc cluster [2], a large number of new complexes with magnetic behavior depending on temperature and applied magnetic field have been synthesized [3][4][5][6]. These compounds, which exhibit a slow relaxation of magnetization below a certain temperature, are called single molecule magnets (SMMs). Molecular materials based on SMMs can be considered as a part of future and even modern electronic devices (e.g., high-density information storage) [7][8][9][10][11].
Initially, the search for SMMs was focused on the synthesis of exchange-coupled clusters based on high-spin 3d metal ions or containing ions of 3d and 4f elements [12][13][14][15]. In 2010, Long and co-workers synthesized a series of mononuclear Fe(II) complexes exhibiting slow magnetic relaxation and named such compounds single-ion magnets (SIMs) [16]. Since this achievement, a large series of SIMs based on mononuclear complexes of 3d metal ions (Fe(I)/(II), Mn(II)/(III), Co(I)/(II), and Ni(I)/(II)) have been reported, and their magnetic properties have been thoroughly investigated [17][18][19][20][21]. Coordination environment, as well as the

Synthesis and Characterization
The dpp-BIAN was synthesized according to a known procedure [43]. Complex [(dpp-BIAN)Fe II I 2 ] (1) was synthesized from iron(II) iodide and dpp-BIAN in acetonitrile solution in a sealed evacuated ampoule (Scheme 1). Anhydrous FeI 2 is extremely air-sensitive compound, thus, it was synthesized and used in situ to prevent oxidation of Fe 2+ (commercial anhydrous FeI 2 undergoes oxidation even when stored in a glove box). The synthesis of iron diiodide in a Schlenk tube with a Teflon stopcock in CH 3 CN is a very simple procedure and takes about 30 min. Complex 1 was isolated from the mother liquor as brown crystals in an excellent yield of 96%. Although FeI 2 is air-sensitive, compound 1 is quite stable in the presence of oxygen and air moisture in solid. In the IR spectrum of compound 1 (see supplementary Figure S1), there is no absorption band associated with the dpp-BIAN radical anion (1510-1515 cm −1 [39,44]). The presence of the dpp-BIAN neutral form in complex 1 is indicated by the absorptions with maxima at 1627 and 1645 cm −1 , which are lower than for the C=N stretching vibra-Scheme 1. Synthesis of complex 1.
In the IR spectrum of compound 1 (see Supplementary Figure S1), there is no absorption band associated with the dpp-BIAN radical anion (1510-1515 cm −1 [39,44]). The presence of the dpp-BIAN neutral form in complex 1 is indicated by the absorptions with maxima at 1627 and 1645 cm −1 , which are lower than for the C=N stretching vibrations in free dpp-BIAN (1671, 1652 cm −1 ) [43] and are similar to the C=N stretching frequencies in [(dpp-BIAN)Fe II Br 2 ] (1643, 1614 cm −1 ) [32]. The band-shift to the lower wavenumbers indicates the coordination of two dpp-BIAN α-diimine nitrogen atoms.

Molecular Structure
Molecular structure for 1 was determined by single-crystal X-ray diffraction at 100 and 280 K ( Figure 1). Crystal data and refinement statistics for 1 are summarized in Table 1, while selected bond lengths and angles are presented in Table 2. In the IR spectrum of compound 1 (see supplementary Figure S1), there is no absorption band associated with the dpp-BIAN radical anion (1510-1515 cm −1 [39,44]). The presence of the dpp-BIAN neutral form in complex 1 is indicated by the absorptions with maxima at 1627 and 1645 cm −1 , which are lower than for the C=N stretching vibrations in free dpp-BIAN (1671, 1652 cm −1 ) [43] and are similar to the C=N stretching frequencies in [(dpp-BIAN)Fe II Br2] (1643, 1614 cm −1 ) [32]. The band-shift to the lower wavenumbers indicates the coordination of two dpp-BIAN α-diimine nitrogen atoms.

Molecular Structure
Molecular structure for 1 was determined by single-crystal X-ray diffraction at 100 and 280 K ( Figure 1). Crystal data and refinement statistics for 1 are summarized in Table 1, while selected bond lengths and angles are presented in Table 2.   [31] and slightly shorter than those in [(dpp-BIAN)Fe II Br2] (range: 2.131(6)-2.153(6) Å)) [32]. As a redox-active ligand, the dpp-BIAN    [31] and slightly shorter than those in [(dpp-BIAN)Fe II Br 2 ] (range: 2.131(6)-2.153(6) Å)) [32]. As a redox-active ligand, the dpp-BIAN can accept electrons, which should lead to a change in the α-diimine bond lengths [42,44]. In turn, an analysis of these bond lengths may suggest the ligand oxidation state. Inspection of the C−C and C−N bond lengths in the α-diimine fragment of  [45], confirms a double bond character for C-N bonds in the dpp-BIAN ligand. On the other hand, bond length distribution in non-innocent complexes is considered as a necessary but not sufficient criterion for determining the electronic state of a ligand [46,47]. The use of a set of complementary methods is a common procedure for investigating the electronic structure of such systems.
The crystal structures of 1 at 100 and 280 K are isostructural (Table 1), changes in bond lengths and angles are insignificant ( Table 2).

UV-VIS Spectra and Calculations
The diffuse reflectance spectra of free dpp-BIAN and complex 1 were recorded in the solid state (polycrystalline sample dispersed in the BaSO 4 matrix under argon atmosphere) and converted to the UV-VIS spectra ( Figure 2). Figure 2 shows that the UV-VIS spectrum of 1 has an intense absorption band in the range of 450-550 nm, the same as in dpp-BIAN, and a broad intense band (λ max = 740 nm) with a shoulder (at 660 nm) in the range of 600-850 nm. Two electron transitions contribute mainly to the latter band, and both of them are combinations (sum and difference) of promoting αand β-electrons from MOs localized on FeI 2 to π*-MOs of the dpp-BIAN ligand. Thus, this band is of the (M + X)LCT (metal-to-ligand and halide-to-ligand) type. mosphere) and converted to the UV-VIS spectra ( Figure 2). Figure 2 shows that the UV-VIS spectrum of 1 has an intense absorption band in the range of 450-550 nm, the same as in dpp-BIAN, and a broad intense band (λmax = 740 nm) with a shoulder (at 660 nm) in the range of 600-850 nm. Two electron transitions contribute mainly to the latter band, and both of them are combinations (sum and difference) of promoting α-and β-electrons from MOs localized on FeI2 to π*-MOs of the dpp-BIAN ligand. Thus, this band is of the (M + X)LCT (metal-to-ligand and halide-to-ligand) type.

Solid State Magnetic Properties and Calculations
Before investigating the magnetic properties, it is very important to establish the oxidation state of Fe ion and the stability of complex 1 under ambient conditions. The purity of complex 1 has been determined by powder X-ray diffraction ( Figure S2, SI), its chemical stability, and the Fe(II) oxidation state have been additionally confirmed by recording only Fe 2+ lines in the Mössbauer spectrum ( Figure S3, SI).
To understand magnetic properties of polycrystalline samples of 1 on a molecular level, we performed also high-level ab initio calculations (at both the relativistic CASSCF/NEVPT2 and non-relativistic CASCI levels) of the electronic structure and properties of complex 1 (for details, see SI). It was established that complex 1 has a quintet ground state (S = 2) with negligible contribution (<1%) of the configurations with occupation of the dpp-BIAN virtual MOs. The next four excited states are also quintet states, which have substantially higher energy ( Figure S4a, SI) and are not occupied in the investigated temperature range. The zero-field splitting (ZFS) of the ground state was also predicted with a negative D value (from −7.8 to −16.7 cm −1 at different levels)

Solid State Magnetic Properties and Calculations
Before investigating the magnetic properties, it is very important to establish the oxidation state of Fe ion and the stability of complex 1 under ambient conditions. The purity of complex 1 has been determined by powder X-ray diffraction ( Figure S2, SI), its chemical stability, and the Fe(II) oxidation state have been additionally confirmed by recording only Fe 2+ lines in the Mössbauer spectrum ( Figure S3, SI).
To understand magnetic properties of polycrystalline samples of 1 on a molecular level, we performed also high-level ab initio calculations (at both the relativistic CASSCF/NEVPT2 and non-relativistic CASCI levels) of the electronic structure and properties of complex 1 (for details, see SI). It was established that complex 1 has a quintet ground state (S = 2) with negligible contribution (<1%) of the configurations with occupation of the dpp-BIAN virtual MOs. The next four excited states are also quintet states, which have substantially higher energy ( Figure S4a, SI) and are not occupied in the investigated temperature range. The zero-field splitting (ZFS) of the ground state was also predicted with a negative D value (from −7.8 to −16.7 cm −1 at different levels) and a moderate E/D ratio. At different levels of theory, the g iso value is predicted to be in the range of 2.13-2.20 (Table S2, SI).
Thus, the results of Mössbauer and diffuse reflectance spectroscopies, as well as quantum chemical calculations, unambiguously prove the Fe(II) oxidation state and the absence of electron transfer between Fe(II) and dpp-BIAN during complexation, leading to the formation of [(dpp-BIAN)Fe II I 2 ] complex. Since the value of D for 1 is negative, single-ion magnetic (SIM) behavior is possible for this complex. Therefore, the magnetic behavior of this complex on direct and alternating current was studied.
The dc magnetic behavior of 1 was investigated under 5000 Oe dc magnetic field in the temperature range of 2-300 K. The experimental temperature dependences of the molar magnetic susceptibility (χ m ) in the form of χ m vs. T and χ m T vs. T are shown in Figure 3. As can be seen, the experimental χ m T values in the range of 200-300 K are close to the theoretical ones for the corresponding non-interacting complexes with S = 2 and g iso = 2.15. This is due to the fact that the magnetic behavior of complex 1 is provided only by the contribution of Fe 2+ , since all ligands are diamagnetic. Figure 3 also displays the theoretical temperature dependences for non-interacting complexes 1, calculated using the results of high-level quantum chemical calculations. For both the experimental and theoretical dependences, the χ m T values remain practically constant upon cooling from 300 to 100 K and close to each other. However, with a subsequent temperature lowering, the growth of the experimental χ m T values was observed with a maximum of 3.77 cm 3 mol −1 K at 16 K followed by a sharp decrease to χ m T = 1.32 cm 3 mol −1 K at 2 K. On the contrary, the theoretical χ m T value decreases monotonically with decreasing temperature from 100 to 2 K, reaching a value 1.83 cm 3 mol −1 K at 2 K. The monotonic decrease in the theoretical dependence of χ m T vs. T is associated with the depopulation of the magnetic sublevels of the ground quintet state of Fe(II) (Figure S4b, SI). The presence of the maximum on the experimental curve can be explained by the reorientation of crystals of 1 under the 5000 Oe dc-field. Comparison with the theoretical curve suggests that the sharp decrease in the experimental χ m T at the low temperatures is mainly due to the depopulation of the magnetic sublevels of Fe II ions.
to the formation of [(dpp-BIAN)Fe II I2] complex. Since the value of D for 1 is negative, single-ion magnetic (SIM) behavior is possible for this complex. Therefore, the magnetic behavior of this complex on direct and alternating current was studied.
The dc magnetic behavior of 1 was investigated under 5000 Oe dc magnetic field in the temperature range of 2-300 K. The experimental temperature dependences of the molar magnetic susceptibility (χm) in the form of χm vs. T and χmT vs. T are shown in Figure 3. As can be seen, the experimental χmT values in the range of 200-300 K are close to the theoretical ones for the corresponding non-interacting complexes with S = 2 and giso = 2.15. This is due to the fact that the magnetic behavior of complex 1 is provided only by the contribution of Fe 2+ , since all ligands are diamagnetic. Figure 3 also displays the theoretical temperature dependences for non-interacting complexes 1, calculated using the results of high-level quantum chemical calculations. For both the experimental and theoretical dependences, the χmT values remain practically constant upon cooling from 300 to 100 K and close to each other. However, with a subsequent temperature lowering, the growth of the experimental χmT values was observed with a maximum of 3.77 cm 3 mol −1 K at 16 K followed by a sharp decrease to χmT = 1.32 cm 3 mol −1 K at 2 K. On the contrary, the theoretical χmT value decreases monotonically with decreasing temperature from 100 to 2 K, reaching a value 1.83 cm 3 mol −1 K at 2 K. The monotonic decrease in the theoretical dependence of χmT vs. T is associated with the depopulation of the magnetic sublevels of the ground quintet state of Fe(II) (Figure S4b, SI). The presence of the maximum on the experimental curve can be explained by the reorientation of crystals of 1 under the 5000 Oe dc-field. Comparison with the theoretical curve suggests that the sharp decrease in the experimental χmT at the low temperatures is mainly due to the depopulation of the magnetic sublevels of Fe II ions. In order to investigate single ion molecule magnet behavior, we performed magnetic measurements of [(dpp-BIAN)Fe II I2] in alternating current (ac) magnetic field, but there was no evidence of slow relaxation of the magnetization. In order to investigate single ion molecule magnet behavior, we performed magnetic measurements of [(dpp-BIAN)Fe II I 2 ] in alternating current (ac) magnetic field, but there was no evidence of slow relaxation of the magnetization.

Cyclic Voltammetry and EPR Spectroscopy
It is well-known that dpp-BIAN can be reversibly reduced by alkali metals; it can accept stepwise from one to four electrons [42]; dpp-BIAN can exist in radical-anion or dianion form conjugated with a metal center (Ln or d-elements) [39,44,[48][49][50][51]. However, electrochemical properties of the BIAN-complexes are not sufficiently investigated. Redox properties of several BIAN derivatives [52,53], as well as those of various α-diimine complexes of iron [30,54] have been reported. In order to enrich the level of discussion, we have studied the electrochemical behavior of the new complex 1 both in the solid state and in solution.
The initial powder of 1 has an EPR signal, which is a broad line with parameters: g = 2.10 and ∆H = 900 G (Figure 4). Unfortunately, this information itself is not enough to unequivocally assign this spectrum to definite paramagnetic center. However, we can definitely say that the spectrum is associated with paramagnetic iron ions, and not with organic radicals in the composition of the sample. In contrast to solid state (Figure 4), solutions of complex 1 in acetonitrile and CH 2 Cl 2 are EPR silent. This is not just a decrease in the signal intensity due to a multiple decrease in the concentration of the complex in solution as compared to its content in the powder, but it is indeed the disappearance of the EPR signal. The EPR silence of 1 in solution is associated with a rather large ZFS in the paramagnetic ground state of this complex. Thus, we can propose that the EPR spectrum in Figure 4 belongs not to the paramagnetic Fe ions of individual complexes 1, but to the system of interacting paramagnetic Fe ions in the solid state. On the whole, this assumption is consistent with the unusual magnetic properties of 1.
dox properties of several BIAN derivatives [52,53], as well as those of various α-diimine complexes of iron [30,54] have been reported. In order to enrich the level of discussion, we have studied the electrochemical behavior of the new complex 1 both in the solid state and in solution.
The initial powder of 1 has an EPR signal, which is a broad line with parameters: g = 2.10 and ∆H = 900 G (Figure 4). Unfortunately, this information itself is not enough to unequivocally assign this spectrum to definite paramagnetic center. However, we can definitely say that the spectrum is associated with paramagnetic iron ions, and not with organic radicals in the composition of the sample. In contrast to solid state (Figure 4), solutions of complex 1 in acetonitrile and CH2Cl2 are EPR silent. This is not just a decrease in the signal intensity due to a multiple decrease in the concentration of the complex in solution as compared to its content in the powder, but it is indeed the disappearance of the EPR signal. The EPR silence of 1 in solution is associated with a rather large ZFS in the paramagnetic ground state of this complex. Thus, we can propose that the EPR spectrum in Figure 4 belongs not to the paramagnetic Fe ions of individual complexes 1, but to the system of interacting paramagnetic Fe ions in the solid state. On the whole, this assumption is consistent with the unusual magnetic properties of 1. The voltammetric study of [(dpp-BIAN)Fe II I2] was carried out in two solvents (acetonitrile and CH2Cl2) as well as in the solid state [55] (Figure 5 and Table 3). The electrochemical behavior of 1 does not differ from that for the previously described chloride analogue [(dpp-BIAN)Fe II Cl2] [54,56] and, thus, can be explained by Scheme 2. It should be noted that the electrochemical behavior of 1 in the solid state and in the methylene chloride solution is the same, which indicates the sufficient stability of the complex in this solvent. The electrochemical behavior in the acetonitrile solution is different when the complex is dissolved in an inert gas atmosphere, a new characteristic reoxidation peak appears on the voltammogram ( 1 Epa = −0.21 V). The voltammetric study of [(dpp-BIAN)Fe II I 2 ] was carried out in two solvents (acetonitrile and CH 2 Cl 2 ) as well as in the solid state [55] (Figure 5 and Table 3). The electrochemical behavior of 1 does not differ from that for the previously described chloride analogue [(dpp-BIAN)Fe II Cl 2 ] [54,56] and, thus, can be explained by Scheme 2. It should be noted that the electrochemical behavior of 1 in the solid state and in the methylene chloride solution is the same, which indicates the sufficient stability of the complex in this solvent. The electrochemical behavior in the acetonitrile solution is different when the complex is dissolved in an inert gas atmosphere, a new characteristic reoxidation peak appears on the voltammogram ( 1 E p a = −0.21 V).  Using the EPR spectroelectrochemistry for the CH2Cl2 solution of 1, the EPR signal was recorded at a potential of −1.65 V (Figure 6). The value of the g-factor (g = 2.0027) and splitting on two nitrogen nuclei (2:aN = 4.8 G) indicates that we are dealing with the dpp-BIAN radical. Noteworthy is the presence of additional splitting at 4 protons (4:aH =    Using the EPR spectroelectrochemistry for the CH2Cl2 solution of 1, the EPR signal was recorded at a potential of −1.65 V (Figure 6). The value of the g-factor (g = 2.0027) and splitting on two nitrogen nuclei (2:aN = 4.8 G) indicates that we are dealing with the dpp-BIAN radical. Noteworthy is the presence of additional splitting at 4 protons (4:aH = Scheme 2. Cathodic transformation of complex 1. Using the EPR spectroelectrochemistry for the CH 2 Cl 2 solution of 1, the EPR signal was recorded at a potential of −1.65 V (Figure 6). The value of the g-factor (g = 2.0027) and splitting on two nitrogen nuclei (2:a N = 4.8 G) indicates that we are dealing with the dpp-BIAN radical. Noteworthy is the presence of additional splitting at 4 protons (4:a H = 4.7 G), which is not observed for the radical anion and radical trianion of the free dpp-BIAN. Thus, it is most likely that the reduction of the product of 1 in solution leads to the paramagnetic hetero-spin complex [(dpp-BIAN) •− Fe I ].
Molecules 2021, 26, x FOR PEER REVIEW 9 of 14 4.7 G), which is not observed for the radical anion and radical trianion of the free dpp-BIAN. Thus, it is most likely that the reduction of the product of 1 in solution leads to the paramagnetic hetero-spin complex [(dpp-BIAN) •− Fe I ].

General Remarks
All the synthetic procedures were carried out under a vacuum using glass ampules or under an argon atmosphere with standard Schlenk techniques. Acetonitrile was dried with P2O5, kept on the activated molecular sieves (3 Å), toluene was refluxed over sodium with benzophenone-both solvents were distilled under vacuum prior to use. The dpp-BIAN was synthesized by the condensation of acenaphthenequinone with 2,6diisopropylaniline (both from Aldrich) under reflux in acetonitrile with a catalytic amount of acetic acid. Iron (II) iodide was synthesized every time in situ from carbonyl iron and iodine (both from Aldrich) in acetonitrile under heating in a Schlenk ampule with a Teflon stopcock. The yield of the product was calculated from the amount of the dpp-BIAN (0.50 g, 1.0 mmol) used in the syntheses. IR spectra were recorded on a Perkin Elmer Spectrum 65 spectrophotometer equipped with a Quest ATR Accessory (Specac) by the attenuated total reflectance (ATR) in the range of 400-4000 cm −1 . Elemental analysis was performed on a Euro EA-3000 (Euro Vektor) automated C,H,N,S-analyzer. The powder patterns were measured using Bruker D8 Advance (λ(CuKα) = 1.5418 Å, Ni filter, Bragg-Brentano geometry) diffractometer. The obtained patterns were Rietveld refined using TOPAS 5 software. The diffuse reflectance data for a mixture of powder of 1 (or free dpp-BIAN) with BaSO4 were converted into the electronic absorption spectra applying a Kubelka-Munk (K-M) function [57].
The X-ray diffraction data sets for crystal of 1·CH3CN were collected using Bruker SMART APEX II diffractometer equipped with a CCD detector (Mo-Kα, λ= 0.71073 Å, graphite monochromator) [58]. Semiempirical absorption correction for 2 was applied [59]. The structure was solved by direct methods and refined by the full-matrix least squares with anisotropic displacement parameters using the SHELX-2014 program package [60]. The hydrogen atoms of the ligands were positioned geometrically and refined using the riding model.

General Remarks
All the synthetic procedures were carried out under a vacuum using glass ampules or under an argon atmosphere with standard Schlenk techniques. Acetonitrile was dried with P 2 O 5 , kept on the activated molecular sieves (3 Å), toluene was refluxed over sodium with benzophenone-both solvents were distilled under vacuum prior to use. The dpp-BIAN was synthesized by the condensation of acenaphthenequinone with 2,6-diisopropylaniline (both from Aldrich) under reflux in acetonitrile with a catalytic amount of acetic acid. Iron (II) iodide was synthesized every time in situ from carbonyl iron and iodine (both from Aldrich) in acetonitrile under heating in a Schlenk ampule with a Teflon stopcock. The yield of the product was calculated from the amount of the dpp-BIAN (0.50 g, 1.0 mmol) used in the syntheses. IR spectra were recorded on a Perkin Elmer Spectrum 65 spectrophotometer equipped with a Quest ATR Accessory (Specac) by the attenuated total reflectance (ATR) in the range of 400-4000 cm −1 . Elemental analysis was performed on a Euro EA-3000 (Euro Vektor) automated C,H,N,S-analyzer. The powder patterns were measured using Bruker D8 Advance (λ(CuKα) = 1.5418 Å, Ni filter, Bragg-Brentano geometry) diffractometer. The obtained patterns were Rietveld refined using TOPAS 5 software. The diffuse reflectance data for a mixture of powder of 1 (or free dpp-BIAN) with BaSO 4 were converted into the electronic absorption spectra applying a Kubelka-Munk (K-M) function [57].
The X-ray diffraction data sets for crystal of 1·CH 3 CN were collected using Bruker SMART APEX II diffractometer equipped with a CCD detector (Mo-K α , λ= 0.71073 Å, graphite monochromator) [58]. Semiempirical absorption correction for 2 was applied [59]. The structure was solved by direct methods and refined by the full-matrix least squares with anisotropic displacement parameters using the SHELX-2014 program package [60]. The hydrogen atoms of the ligands were positioned geometrically and refined using the riding model. The structure parameters were deposited with the Cambridge Structural Database (CCDC Nos. 2077891 (1, 100 K) and 2077892 (1, 280 K); deposit@ccdc.cam.ac.uk or http://www.ccdc.cam.ac.uk/data_request/cif (accessed on 16 April 2021)).
Voltammograms of the [(dpp-BIAN)Fe II I 2 ] complex were recorded in dry acetonitrile and CH 2 Cl 2 . A recrystallized n-Bu 4 NBF 4 sample was placed in a Schlenk ampoule, it was evacuated and filled with argon. For this, 15 mL of acetonitrile was added through the septum. Then, 5 mL of the resulting solution was taken with a syringe to fill the electrochemical bridge. A portion of the complex was transferred under argon flow into a Schlenk vessel with n-Bu 4 NBF 4 solution and dissolved under the influence of ultrasound. The resulting solution in argon flow was transferred by syringe into an electrochemical cell filled with argon.
DC-Magnetic susceptibility measurements were performed using a Quantum Design PPMS-9 susceptometer, under 5000 Oe magnetic field in the 2-300 K temperature range on polycrystalline sample sealed in polyethylene bag. The paramagnetic components of the magnetic susceptibility χ were determined taking into account both the diamagnetic contribution evaluated from Pascal's constants and the contributions of the sample holder.

Synthesis of [(dpp-BIAN)Fe II I 2 ]·CH 3 CN (1·CH 3 CN)
To a colorless solution of iron(II) iodide which was prepared in situ from iodine (0.254 g, 1.0 mmol) and an excess of iron powder in acetonitrile (25 mL) was added dpp-BIAN (0.500 g, 1.0 mmol). The reaction mixture was heated at 110 • C for 1 h in a sealed ampule. Colorless solution became dark-brown. Then, the reaction mixture slowly cooled (by 10 • C in an hour) to RT. A product started to crystallize at 50 • C. Big brown cubic crystals were isolated (0.82 g, 96%) after 24 h. Anal. Calcd for C 38

Details of Quantum Chemical Calculations
The positions and oscillator strengths of electronic transitions in the UV-VIS spectrum of 1 were calculated using time-dependent (TD) DFT [61] at the TD-UB2PLYP/def2-TZVP level [62].
The state-averaged (SA) CASSCF/NEVPT2 procedure [63,64] was used for the calculation of the wave functions of spin-free states. The scalar relativistic effects were taken into account using standard second-order Douglas−Kroll−Hess (DKH2) procedure [65,66]. The segmented all-electron relativistically contracted version of Ahlrichs polarized basis set def2-TZVP was used [67]. The active space for most of calculations (6, 10) consisted of six electrons distributed on five 3d-orbitals and five double-shell d-orbitals [68]. The fine structure of ground and excited states was computed with the Breit-Pauli approximation and the mean field approximation (SOMF) [69]. Five quintet and 45 triplet states of Fe II were included. Spin Hamiltonian (SH) parameters and g tensors have been computed with an effective Hamiltonian (QDPT) [70]. All calculations were performed with the ORCA 4.1.1 computational package [71].
The CASSCF/NEVPT2 wavefunctions of complex 1 were also used to calculate temperature dependence of its molar magnetic susceptibility. In these calculations, both the SOC and Zeeman effects were taken into account. Magnetic susceptibility of each magnetic sublevel was computed numerically as the second derivative of the energy with respect to the magnetic field. The molar magnetic susceptibility is computed as averaged according to Boltzmann statistics. In order to reproduce powder measurement, the averaging over sphere for directions of magnetic field has been performed. Described treatment, realized in the ORCA package [71], takes into account mixing of states due to both the spin−orbit coupling and magnetic field up to infinite order of perturbation theory.

Conclusions
Here, we reported on a novel molecular iodide iron(II) complex with neutral 1,2bis[(2,6-diisopropylphenyl)imino]acenaphthene, [(dpp-BIAN)Fe II I 2 ]. It is known that dpp-BIAN can exist in a radical-anion or dianionic form conjugated with a metal center, which were previously isolated [50]. In this paper, based on the data of single crystal X-ray diffraction, we proved the absence of redox isomerism in the solid state and the preservation of an almost identical crystal structure of 1 in the temperature range of 100-300 K. The EPR spectrum of solid sample of 1 is associated with paramagnetic Fe ions, and not with organic radicals in the composition of the sample. The oxidation state of the Fe(II) ion in 1 was confirmed by Mossbauer spectroscopy, magnetic susceptibility measurements, and high-level quantum chemical calculations. The neutral form of dpp-BIAN is confirmed by the features of the IR and UV-VIS spectra. Thus, according to comprehensive experimental studies and ab initio calculations, solid sample of 1 consists of an iron(II) ion and a neutral form of a non-innocent ligand.
Cyclic voltammetry of 1 demonstrated that the transition of the dpp-BIAN ligand to the anion-radical form occurs in two steps: (1) reduction of the iron(II) ion into iron(I) ion and (2) reduction of the ligand. The electrochemical behavior of complex 1 in the solid state and in methylene chloride is the same. The product of two-electron reduction in a CH 2 Cl 2 solution (E p = −1.65 V vs. Fc/Fc+) was detected by EPR method and assigned to the paramagnetic hetero-spin complex [(dpp-BIAN) •− Fe I ].
Supplementary Materials: The following are available online. Figure S1: IR spectrum of compound; Figure S2: Theoretical and experimental powder patterns of 1, Figure S3: Mossbauer spectrum of compound 1, Figure S4: the energies of the lowest spin-multiplets (S = 2) of complex 1 (a) and energies of magnetic sublevels of the ground quintet state, Figure S5: d-orbitals of Fe(II) in upper row and LUMO and LUMO+1 of dpp-BIAN ligand in 1, Table S1: Mossbauer parameters of iron in compound 1, Table S2: The fine structure and SH-parameters for the quintet ground states of complexes 1.