Synthesis of Ferrocenyl-Substituted Organochalcogenyldichlorogermanes

Abstract

Reaction of the isolable ferrocenyldichlorogermyl anion, Fc*GeCl₂⁻ (Fc* = 2,5-bis(3,5-di-t-butylphenyl)-1-ferrocenyl), with the isolable chalcogenenyl halides resulted in the formation of the corresponding organochalcogenyldichlorogermanes that were structurally characterized. Thus, it was demonstrated the use of sterically demanding ferrocenyl groups allowed isolating stable crystalline organochalcogenyldichlorogermanes.

1. Introduction

Germanium chalcogenides are interesting chemical species for optoelectronic modules due to the appropriate combination between the electron-accepting element Ge and an electron-donating element, such as Se or Te, because the size and energy levels of frontier orbitals should be close to each other (4p and 4p/5p) [1,2]. Therefore, organogermanium species that bear a chalcogen (Ch) moiety should be able to serve as building blocks for organic–inorganic hybrid materials that contain a Ge–Ch bond. Given that the Ge–Ch bond is redox-active, metallocenyl-substituted germanium chalcogenides could be promising prospective building blocks for such Ge–Ch hybrid materials. Although the appropriate molecular design for such building blocks should be Mc–GeX₂–ChR (Mc = metallocenyl; R = organic substituent; X = leaving group), it is generally difficult to isolate such species on account of the lability due to facile hydrolysis of the Ge–Ch and Ge–X bonds. In addition, a conceivable synthetic strategy such as the nucleophilic substitution of the RCh moiety toward Mc-GeCl₃, would most likely not be selective, i.e., two- and three-fold substitution could easily occur (Scheme 1). In this paper, we report a solution to this problem by using a method that is based on kinetic control using a bulky ferrocenyl group. We have already prepared sterically demanding ferrocenyl groups [3,4,5,6] that are able to stabilize anionic species that bear a halogen group due to multi-hydrogen bonding [7]. The use of the sterically demanding ferrocenyl group Fc* (2,5-bis(3,5-di-t-butylphenyl)-1-ferrocenyl) enabled us to isolate the dichlorogermyl anion [Fc*GeCl₂]⁻, which was identified as a chlorogermylenoid [7]. Subsequently, we speculated that the germylenoid could not only work as an electrophile but also as a nucleophile towards chalcogens, even in the presence of two halogen atoms. Herein, we demonstrate the reactions of a stable germylenoid with selenenyl and tellurenyl chlorides, which affords stable dichlorochalcogenagermanes that bear a bulky ferrocenyl group. These dichlorochalcogenagermanes represent promising prospective building blocks for organogermanium chalcogenides.

2. Results and Discussions

To isolate stable chalcogenenyl halides, it is necessary to introduce sterically demanding substituents on the Ch atom [8,9,10]. We have decided to use the 9-triptycyl (Trp) group as a steric protection group, as the simple synthesis of the corresponding dichalcogenides (1a,b) has already been reported [11,12,13]. The treatment of an ether solution of TrpSeSeTrp (1a) [11] with SO₂Cl₂ at room temperature afforded the corresponding selenenyl chloride, TrpSeCl (2a), as a stable crystalline compound (Scheme 2). In a similar fashion, TrpTeCl (2b) was obtained from the reaction of TrpTeTeTrp (1b) [12,13] with SO₂Cl₂. The molecular structures of TrpSeCl (2a) and TrpTeCl (2b) were unambiguously determined by X-ray diffraction (XRD) analyses, which delivered C–Ch–Cl angles of 99.93(6)° (Ch = Se) and 96.38(9)° (Ch = Te), as well as Ch–Cl bond lengths of 2.1860(7) Å (Ch = Se) and 2.348(1) Å (Ch = Te). These structural parameters are similar to those of previously reported stable selenenyl- and tellurenyl-chlorides [8,9,10], indicating negligible electronic perturbations from the Trp group toward the Ch–Cl moieties. The packing structures of 2a and 2b suggest that these compounds are monomeric in the crystalline state (Figure 1). As only 2a contains one molecule of benzene per unit cell, the packing structures and space groups of 2a and 2b are different. While 2b exhibits head-to-tail-type interactions, 2a shows head-to-head-type interactions, albeit that the intramolecular interactions should be negligible due to the long intramolecular Ch···Ch, Ch···Cl, and Cl···Cl distances. In addition, the ⁷⁷Se (907 ppm) and ¹²⁵Te NMR chemical shifts (1756 ppm) are consistent with those of previously reported monomeric chalcogenenylhalides [8,9,10].

The sterically hindered germylenoid Fc*GeCl₂Li (4) was prepared according to literature procedures [5] from the reaction of the isolable lithioferrocene dimer 3 with GeCl₂·(dioxane). Subsequently, 4 was treated with the isolated chalcogenenylchlorides 2a or 2b in toluene at room temperature. The NMR spectra of the crude reaction mixtures suggested the predominant formation of the expected products (5a,b) together with small amounts of the by-product Fc*₂GeCl₂ (6) [14] in both cases (5:6 = 8:1 for Ch = Se; 5:6 = 21:1 for Ch = Te). The purification processes, including GPC separation and recrystallization from hexane, afforded the stable chalcogenyldichlorogermanes 5a and 5b in 41% and 58% isolated yields, respectively, together with the corresponding bis(ferrocenyl)dichlorogermane in both cases (31% for Ch = Se; 17% for Ch = Te) (Scheme 3). Although the formation mechanism for 6 cannot be explained unequivocally at present, the oxidation of germylenoid 4 by chalcogenenylchloride 2 could initiate the unexpected formation of 6.

The molecular structures of chalcogenyldichlorogermanes 5a and 5b were determined by single-crystal XRD analyses (Figure 2). Unexpectedly, their geometries are different, i.e., the Ch moiety in 5a (Ch = Se) is spatially removed from the Fe atom or located “outside of the ferrocenyl unit” (Form A), while that of 5b (Ch = Te) is close to the Fe atom or oriented “toward the ferrocenyl unit” (Form B). In both cases, the two energy minima, i.e., Form A and B, were identified by theoretical calculations at the M062x/6-311G(3d) (Ge, Cl, Fe)/ 6-31G(d) (C, H)/ SDD (Se, Te) level of theory [15]. The optimized structures of 5a-Form A and 5b-Form B were in good agreement with those experimentally obtained from the XRD analyses (

Table 1. Selected structural parameters for 5a-Form A and 5b-Form B (observed: XRD analysis) together with the corresponding theoretical (calculated) values for 5a-Form A, 5a-Form B, 5b-Form A, and 5b-Form B that were optimized at the M062x/6-311G(3d) (Ge, Cl, Fe)/ 6-31G(d) (C, H)/ SDD (Se, Te) level of theory.

Distance/Å	5a (Ch = Se) Form A Observed	5a (Ch = Se) Form A Calculated	5a (Ch = Se) Form B Calculated	5b (Ch = Te) Form A Calculated	5b (Ch = Te) Form B Calculated	5b (Ch = Te) Form B Observed
C(Cp)–Ge	1.926(6)	1.9289	1.9290	1.9308	1.9309	1.927(8)
Ge–Ch	2.3474(9)	2.3870	2.3904	2.5878	2.5869	2.5489(8)
Ge–Cl1	2.209(2)	2.1873	2.2018	2.1979	2.2081	2.212(2)
Ge–Cl2	2.164(2)	2.1577	2.1561	2.1609	2.1624	2.164(2)
Fe···Ge	3.696(1)	3.6587	3.5460	3.7326	3.5252	3.505(2)
Fe···Cl1	4.194(2)	4.0287	5.3396	4.2108	5.3303	5.305(3)
Fe···Cl2	5.176(2)	5.1614	4.3337	5.1220	4.2763	4.317(2)
Fe···Ch	5.229(1)	5.3217	4.6049	5.6007	4.7724	4.761(1)
Angles/°
Cl1–Ge–Cl2	104.06(7)	105.85	105.68	102.93	105.82	104.51(8)
Cl1–Ge–Ch	110.59(5)	103.47	108.33	108.95	108.17	106.62(6)
Cl2–Ge–Ch	110.92(5)	110.60	110.15	110.48	109.03	110.32(6)

). Although the thermodynamic energies of Form A and Form B are similar in both cases, 5a-Form A and 5b-Form B are more stable than their imaginary forms, i.e., 5a-Form B (+0.33 kcal/mol) and 5b-Form A (+1.38 kcal/mol), which supports the experimental results. At present, however, we do not have a reasonable explanation regarding the energy differences between Form A and Form B.

The experimentally observed and theoretically optimized structural parameters are summarized in Table 1. In both cases, i.e., Form A and Form B, one of the two chlorine atoms (Cl1) is vertically oriented toward the Cp plane of the ferrocenyl unit. In both cases, the Ge–Cl1 bonds are slightly longer than the Ge–Cl2 bonds, indicating an orbital interaction between the σ*(Ge–Cl1) orbital and lone pairs on the Cl2 and Ch (Se or Te) atoms. Indeed, the NBO (Natural Bond Orbitals) calculations [16] suggested effective π(Cp) → σ*(Ge–Cl1), LP(Cl2) → σ*(Ge–Cl1), and LP(Ch) → σ*(Ge–Cl1) interactions, all of which would result in an elongation of the Ge–Cl1 bond. These results indicate that the Ge–Ch and Ge–Cl2 bonds would be maintained, even after the functionalization of the Ge–Cl1 moiety, given that the Ge–Ch and Ge–Cl2 bonds should strengthen rather than weaken the Ge–Cl1 bond. Thus, 5a and 5b should be suitable as potential building blocks for ferrocenyl-substituted germanium chalcogenides.

3. Materials and Methods

3.1. General Information

All manipulations were carried out under an argon atmosphere using either Schlenk-line or glovebox techniques. Solvents were purified using the Ultimate Solvent System (Glass Contour Company, CA, USA) [17]. ¹H, ¹³C, ⁷⁷Se, and ¹²⁵Te NMR spectra were measured on JEOL 300 or 400 MHz spectrometers (JEOL, Tokyo, Japan). Signals arising from residual C₆D₅H (7.15 ppm) in C₆D₆ and CHCl₃ (7.25 ppm) in CDCl₃ were used as an internal standard for the ¹H NMR spectra, while signals of C₆D₆ (128.0 ppm) and CDCl₃ (77.0 ppm) where used to reference the ¹³C NMR spectra. PhSeSePh (460 ppm) and PhTeTePh (450 ppm) were used as external standards for the ⁷⁷Se and ¹²⁵Te NMR spectra. High-resolution mass spectra (HRMS) were measured on a Bruker micrOTOF focus-Kci mass spectrometer (DART) (Bruker Japan K.K. Daltonics Division, Kanagawa, Japan) or a JEOL JMS-700 spectrometer (FAB) (JEOL, Tokyo, Japan). All melting points were determined on a Büchi Melting Point Apparatus M-565 (Büchi Japan, Tokyo, Japan) and are uncorrected. Elemental analyses were carried out at the Microanalytical Laboratory, Institute for Chemical Research, Kyoto University. Dichalcogenides 1a and 1b [11,12,13] as well as chlorogermylenoid 4 [7] were prepared according to literature procedures.

3.2. Experimental Details

3.2.1. Synthesis of Selenenylchloride 2a

A suspension of TrpSeSeTrp (1a, 665 mg, 1.00 mmol) in CH₂Cl₂ (10 mL) was treated with SO₂Cl₂ (135 mg, 1.00 mmol) at room temperature for 1 h. After the removal of all volatiles, the residue was recrystallized from CH₂Cl₂ at room temperature to give 2a as orange crystals in 51% yield (376 mg, 1.02 mmol). Data for 2a: orange crystals, m.p. = 212.2 °C (dec.); ¹H NMR (400 MHz, CDCl₃, 298 K): δ (ppm) 5.41 (s, 1H), 7.00–7.10 (m, 6H), 7.40–7.42 (m, 3H), 7.50–7.52 (m, 3H); ¹³C NMR (100 MHz, CDCl₃, 298 K): δ (ppm) 54.1 (d), 64.3 (s), 123.5 (d), 123.7 (d), 125.3 (d), 126.1 (d), 143.7 (s), 145.4 (d); ⁷⁷Se NMR (76 MHz, CDCl₃, 298 K): δ (ppm) 907; Anal. Calcd. for C₂₀H₁₃ClSe: C, 65.32; H, 3.56. Found: C, 65.01; H, 3.73. MS (DART-TOF, positive mode): m/z Calcd. for C₂₀H₁₃³⁵Cl⁸⁰Se 367.9871 ([M]⁺), found 367.9886 ([M]⁺).

3.2.2. Synthesis of Tellurenylchloride 2b

A suspension of TrpTeTeTrp 1b (762 mg, 1.00 mmol) in CH₂Cl₂ (10 mL) was treated with SO₂Cl₂ (135 mg, 1.00 mmol) at room temperature for 1 h. After the removal of all volatiles, the residue was recrystallized from CH₂Cl₂/benzene at room temperature to give 2b as blue crystals in 57% yield (416 mg, 1.13 mmol). Data for 2b: blue crystals, m.p. = 228.1 °C (dec.); ¹H NMR (400 MHz, CDCl₃, 298 K): δ (ppm) 5.44 (s, 1H), 7.00–7.10 (m, 6H), 7.32–7.35 (m, 3H), 7.40–7.43 (m, 3H); ¹³C NMR (100 MHz, CDCl₃, 298 K): δ (ppm) 54.5 (d), 57.3 (s), 123.8 (d), 125.6 (d), 126.0 (d), 126.3 (d), 145.4 (s), 145.8 (d); ¹²⁵Te NMR (125 MHz, CDCl₃, 298 K): δ (ppm) 1756; Anal. Calcd. for C₂₀H₁₃ClTe: C, 57.69; H, 3.15. Found: C, 57.46; H, 3.14. MS (DART-TOF, positive mode): m/z Calcd. for C₂₀H₁₃³⁵Cl¹³⁰Te 417.9768 ([M]⁺), found 417.9733 ([M]⁺).

3.2.3. Reaction of Chlorogermylenoid 4 with Selenenylchloride 2a

A toluene solution (3 mL) of chlorogermylenoid 4 (123.5 mg, 0.173 mmol) was treated with TrpSeCl (2a, 64.8 mg, 0.176 mmol) at room temperature. After stirring the reaction mixture for 3 h, the solvent was removed under reduced pressure. The residue was extracted into toluene and filtered before the solvent was removed from the filtrate under reduced pressure. The residue was purified by high performance liquid chromatography (HPLC) (eluent: toluene) and recrystallization from hexane to give 5a as the main product in 41% yield (73.7 mg, 0.0710 mmol), and 6 (34.0 mg, 0.0268 mmol, 31%). Data for 5a: orange crystals, m.p. 231 °C (dec.); ¹H NMR (300 MHz, C₆D₆, r.t.): δ (ppm) 1.40 (s, 36H), 4.47 (s, 5H), 4.72 (s, 2H), 4.96 (s, 1H), 6.67–6.72 (m, 3H), 6.76–6.79 (m, 3H), 6.99 (d, 3H, J = 7.1 Hz), 7.55 (t, 2H, J = 1.7 Hz), 7.76 (d, 3H, J = 7.4 Hz), 7.91 (d, 4H, J = 1.7 Hz); ¹³C NMR (75 MHz, C₆D₆, 298 K): δ (ppm) 31.77 (q), 35.22 (s), 54.52 (d), 66.94 (s), 71.86 (d), 73.06 (d), 83.69 (s), 96.62 (s), 122.30 (d), 123.20 (d), 124.91 (d), 125.28 (d), 125.89 (d), 126.13 (d), 136.92 (s), 145.07 (s), 146.01 (s), 151.02 (s); ⁷⁷Se NMR (57 MHz, C₆D₆, 298 K): δ (ppm) 168; MS (DART-TOF, positive mode): m/z Calcd. for C₅₈H₆₃³⁷Cl₂⁵⁷Fe⁷²Ge⁷⁶Se 1038.2068 ([M + H]⁺), found 1038.2084 ([M + H]⁺); Anal. Calcd. for C₅₈H₆₂Cl₂FeGeSe:  C, 67.15; H, 6.02. Found:  C, 66.90; H, 6.16.

3.2.4. Reaction of Chlorogermylenoid 4 with Tellurenylchloride 2b

A toluene solution (3 mL) of chlorogermylenoid 4 (109.9 mg, 0.154 mmol) was treated with TrpTeCl (2b, 65.0 mg, 0.156 mmol) at room temperature. After stirring the reaction mixture for 3 h, the solvent was removed under reduced pressure. The residue was extracted into toluene and filtered before the solvent was removed from the filtrate under reduced pressure. The residue was purified by HPLC (eluent: toluene) and recrystallized from hexane to give 5b as the main product (58% yield, 95.1 mg, 0.0874 mmol) together with 6 (17%, 16.0 mg, 0.0126 mmol). Data for 5b: orange crystals, m.p. 151 °C (dec.); ¹H NMR (300 MHz, C₆D₆, r.t.): δ (ppm) 1.42 (s, 36H), 4.62 (s, 5H), 4.63 (s, 2H), 5.03 (s, 1H), 6.67–6.72 (m, 3H), 6.79–6.83 (m, 3H), 7.01 (d, 3H, J = 7.1 Hz), 7.56 (t, 2H, J = 1.7 Hz), 7.90 (d, 3H, J = 7.3 Hz), 7.97 (d, 4H, J = 1.7 Hz); ¹³C NMR (75 MHz, C₆D₆, 298 K): δ (ppm) 31.71 (q), 35.21 (s), 54.94 (d), 60.81 (s), 72.83 (d), 72.97 (d), 78.16 (s), 96.94 (s), 122.44 (d), 123.24 (d), 125.03 (d), 126.11 (d), 128.51 (d), 129.28 (d), 136.21 (s), 145.04 (s), 147.01 (s), 151.09 (s); ¹²⁵Te NMR (94 MHz, C₆D₆, 298 K): δ (ppm) 244; MS (FAB): m/z calcd. for C₅₈H₆₂³⁷Cl₂⁵⁸Fe⁷⁴Ge¹²²Te 1086.1849 ([M]⁺), found 1086.1853 ([M]⁺); Anal. Calcd. for [C₅₈H₆₂Cl₂FeGeTe + C₆H₁₄]:  C, 65.57; H, 6.53. Found:  C, 65.32; H, 6.66.

3.3. Computational Methods

The level of theory and the basis sets used for the structural optimization are given in the main text. Frequency calculations confirmed minimum energies for all optimized structures. All calculations were carried out using the Gaussian 09 program package [15].

3.4. X-ray Crystallographic Analyses

Single crystals of [2a·benzene], 2b, 5a, and 5b were obtained upon recrystallizations from benzene ([2a·benzene]) or hexane (2b, 5a, and 5b). Intensity data for [2a·benzene], 5a, and 5b were collected on a RIGAKU Saturn70 CCD system (RIGAKU, Tokyo, Japan) with VariMax Mo Optics using Mo Kα radiation (λ = 0.71073 Å), while those for 2b were collected at the BL40XU beam line at Spring-8 (JASRI, projects 2017A1647, 2017B1179, 2018A1167, and 2018A1405) on a Rigaku Saturn 724 CCD system (RIGAKU, Tokyo, Japan) using synchrotron radiation (λ = 0.7823 Å). Crystal data are shown in the references. The structures were solved by direct methods (SHELXT-2014 [18]) and refined by a full-matrix least square method on F² for all reflections (SHELXL-2014 [19]). All hydrogen atoms were placed using AFIX instructions, while all other atoms were refined anisotropically. Supplementary crystallographic data were deposited at the Cambridge Crystallographic Data Centre (CCDC; under reference numbers 1846194–1846197 for [2a·benzene], 2b, 5a, and 5b, respectively) and can be obtained free of charge via www.ccdc.cam.ac.uk/data_request.cif. Crystal data; [2a·benzene] (C₂₆H₁₉ClSe): M = 445.82, λ = 0.71073 Å, T = –170 °C, orthorhombic, Pbca (no. 61), a = 10.3039(3) Å, b = 18.5923(5) Å, c = 21.1023(6) Å, V = 4042.6(2) ų, Z = 8, D_calc = 1.465 g cm⁻³, μ = 1.998 mm⁻¹, 2θ_max = 51.0°, measd./unique refls. = 33309/3943 (R_int = 0.0495), param = 253, GOF = 1.035, R₁ = 0.0279/0.0411 [I > 2 σ(I)/all data], wR₂ = 0.0570/0.0615 [I > 2σ(I)/all data], largest diff. peak and hole 0.341 and −0.279 e·Å⁻³ (CCDC-1846194); 2b (C₂₀H₁₃ClTe): M = 416.35, λ = 0.7823 Å, T = −180 °C, monoclinic, P2₁/c (no. 14), a = 15.3100(3) Å, b = 8.0496(1) Å, c = 13.8261(3) Å, β = 115.690(3)°, V = 1535.49(6) ų, Z = 4, D_calc = 1.801 g·cm⁻³, μ = 2.690 mm⁻¹, 2θ_max = 56.0°, measd./unique refls. = 20016/4065 (R_int = 0.0676), param = 218, GOF = 1.084, R₁ = 0.0483/0.0517 [I > 2σ(I)/all data], wR₂ = 0.1270/0.1284 [I > 2 σ(I)/all data], largest diff. peak and hole 2.241 and −0.948 e·Å⁻³ (CCDC-1846195); 5a (C₅₈H₆₂Cl₂FeGeSe): M = 1037.37, λ = 0.71073 Å, T = −170 °C, triclinic, P−1 (no. 2), a = 9.3412(3) Å, b = 13.2993(7) Å, c = 21.3355(14) Å, α = 93.977(4)°, β = 101.963(4)°, γ = 103.821(2)°, V = 2498.1(2) ų, Z = 2, D_calc = 1.379 g cm⁻³, μ = 1.765 mm⁻¹, 2θ_max = 53.0°, measd./unique refls. = 48637/10245 (R_int = 0.1265), param = 596, GOF = 1.136, R₁ = 0.0736/0.1313 [I > 2σ(I)/all data], wR₂ = 0.1431/0.1712 [I > 2σ(I)/all data], largest diff. peak and hole 0.831 and –0.594 e·Å⁻³ (CCDC-1846196); 5b (C₅₈H₆₂Cl₂FeGeTe): M = 1086.01, λ = 0.71073 Å, T = −170 °C, triclinic, P–1 (no. 2), a = 12.6102(6) Å, b = 15.3206(4) Å, c = 15.9034(7) Å, α = 66.692(2)°, β = 68.112(2)°, γ = 71.139(3)°, V = 2562.03(19) ų, Z = 2, D_calc = 1.408 g cm⁻³, μ = 1.568 mm⁻¹, 2θ_max = 53.0°, measd./unique refls. = 29752/10195 (R_int = 0.0891), param = 632, GOF = 1.167, R₁ = 0.0680/0.1112 [I > 2σ(I)/all data], wR₂ = 0.1261/0.1487 [I > 2σ(I)/all data], largest diff. peak and hole 0.900 and –0.851 e·Å⁻³ (CCDC-1846197).

4. Conclusions

Chalcogenyldichlorogermanes 5a and 5b were successfully synthesized from the reaction between an isolable ferrocenyl-substituted chlorogermylenoid and a sterically demanding ferrocenyl group (Fc*). Ferrocenylchlorogermylenoid 4 is an appropriate precursor for the targeted ferrocenyl-substituted chalcogenyldichlorogermanes via nucleophilic reactions towards the sterically hindered chalcogenenyl chlorides. Thus, reactions of a halogermylenoid with a chalcogenenyl chloride represent an effective synthetic route to chalcogenyldichlorogermanes. Theoretical calculations showed that the Ge–Ch bonds in these chalcogenyldichlorogermanes are strengthened due to LP(Ch) → σ*(Ge–Cl) interactions, suggesting promising potential for such chalcogenyldichlorogermanes as building blocks for organochalcogenylgermanes that bear a redox-active ferrocenyl moiety.