2,5-Bridged 1-Carba-arachno-pentaborane(10) Derivatives – Intermediates on the Way to Pentaalkyl-1,5-dicarba-closo-pentaboranes(5)

2.1 1-Carba-arachno-pentaborane(10) derivative 1

Aiming at a better viable synthetic route to 1, similar to the synthesis of 5 [9], we had at first to solve the problem how to obtain diethyl(ethynyl)borane, Et₂B–C≡C–H. By contrast to the 1-propynyl, Et₂B–C≡C–Me, or the vinyl derivative, Et₂B–CH=CH₂, diethyl(ethynyl)borane Et₂B–C≡C–H was not accessible by reaction of Et₂B–Cl with ethynyltin compounds, such as Bu₃Sn–C≡C–H, Bu₂Sn(C≡C–H)₂ or Oct₃Sn–C≡C–H. Although the organotin chlorides were formed in each case, the boron containing materials were only red-brown polymers. We decided to generate Et₂B–C≡C–H in situ in the hydride bath via ethynyl/H exchange, hoping that the exchange reaction was preferred over 1,1-carboboration [14–16]. For this purpose a cooled (–78 °C) solution of Oct₃Sn–C≡C–H in pentane was added to the cold (–78 °C) hydride bath, followed by addition of an equimolar amount of Et₂B–CH=CH₂. The reaction mixture was allowed to reach room temperature. After removal of the readily volatile materials pentane, ethlyldiboranes(6) and BEt₃, fractional distillation gave compound 1(exo) (b. p. 65 –75 °C/10^–3 Torr) in 22 % yield, contaminated with a small amount of the endo-isomer. The residue consisted of Oct₃Sn–H (δ¹H(Sn–H)[¹J(¹¹⁹Sn,¹H)] = 4.8 [1620 Hz]) and polymeric materials, containing tin and boron, which could not be distilled. At present we cannot offer a conclusive explanation which factors govern the formation of the endo or the exo isomer. These results (Scheme 3) agree with the behaviour of diethyl(1-propynyl)borane, Et₂B–C≡C–Me, present in the hydride bath [9, 10]. Like (Et₂B)₃C–Et, the intermediacy of (Et₂B)₃C–Me (Scheme 3) was just assumed, whereas (Et₂B)₂CH₂CH₃ [9] can be prepared independently by stoichiometric 1,2-hydroboration.

Scheme 3:

Alternative synthesis of 1-carba-arachno-pentaborane(10) derivative 1.

The properties of 1 prepared by following Scheme 3 are identical to those found for the “old” sample of Köster’s experiments [8], when the crude material had been purified. In particular boiling point, mass spectra, ¹¹B (Figs. 1 and 2) and ¹³C NMR data (Fig. 2) are conclusive. The structure of a single molecule of 1 was optimised at the B3LYP/6-311+G(d,p) level of theory [17–21], and the NMR parameters were calculated [22–25] at the same level (Fig. 2).

Fig. 1:

160.5 MHz ¹¹B{¹H} NMR spectrum of 1 (23 °C; 5 mg in 0.5 mL of C₆D₆). Small signals may belong to the endo-isomer.

Fig. 2:

Optimized geometry and experimental and calculated [RB3LYP/6-311+G(d,p) level] ¹¹B and ¹³C NMR parameters of 1(exo).

Figure 2 and data in Table 1 show that the experimental NMR data are fairly well reproduced by the calculations, which means that the optimised single-molecule structures are reliable. ¹H/¹H NOE difference experiments served for distinguishing between the endo- and exo-isomer. Coupling constants ¹J(¹³C,¹¹B) of 1 were not resolved owing to rapid ¹¹B quadrupolar relaxation. However, the broadening of ¹³C NMR signals of carbon atoms linked to boron, typical of partially relaxed scalar ¹³C-¹¹B spin-spin coupling [26], revealed some information. Selective heteronuclear decoupling experiments ¹³C{¹¹B,¹H} enabled to assign the B(2,5)-Et and B(3,4)-Et groups. Moreover, decoupling of ¹¹B(2,5) sharpened the μ-¹³C NMR signal and caused only slight narrowing of the broad ¹³C(1) NMR signal. The calculated values ¹J((μ-¹³C,¹¹B(2,5)) are only slightly smaller than for trialkylboranes [26]. On the other hand, decoupling of ¹¹B(3.4) gave rise to a sharpened ¹³C(1) NMR signal, in agreement with the calculated coupling constants ¹J(¹³C(1),¹¹B(3,4)) >> ¹J(¹³C(1)¹¹B(2,5)). Expectedly, the calculated values for ¹J(¹¹B,¹¹B) are small, typical for B–H–B bridges. Therefore, cross peaks for ¹¹B-¹¹B coupling were absent in 2D ¹¹B/¹¹B COSY spectra [27].

2.2 Thermal rearrangement of the 1-carba-arachno-pentaborane(10) derivatives 1, 4 and 5

When the compounds 1, 4 and 5 were heated in benzene to 80 °C for 1 h, there were no appreciable changes in their NMR spectra. However, heating to >100 °C without a solvent caused formation of a large amount of polymeric materials along with volatile ethyldiboranes(6) and some (<30%) pentaalkyl-1,5-dicarba-closo-pentaboranes 7a (from 1) and 7b (from 4 and 5), both isolated by fractional distillation (Scheme 4). The properties of 7a [28] and 7b [13], such as boiling points, mass spectra and NMR spectroscopic data [29] correspond to reported data. Apparently, decomposition of 1, 4 and 5 had occurred, leading back to the respective 1,1,1-tris(diethylboryl)alkane, which condenses by Et₂B–H-catalysis to the closo carbaboranes 7a and 7b, respectively. We note that Me₂SiH₂ (from 4) was not eliminated.

Scheme 4:

Thermal decomposition of 1, 4 and 5.

4.1 2,3,4,5-Tetraethyl-1-methyl-2,5-μ-(ethane-1′,1′-diyl)-1-carba-arachno-pentaborane(10) (1)

An equimolar mixture of diethyl(vinyl)borane (0.96 g: 10 mmol) and ethynyl(trioctyl)stannane (4.83 g, 10 mmol) was dissolved in 5 mL of pentane, cooled to –78 °C and added slowly within 45 m to the cold (–78 °C) hydride bath. Then, the stirred reaction mixture was warmed to room temperature and the readily volatile materials were removed in a vacuum. Fractional distillation (oil bath <100 °C) of the high boiling residue gave the arachno carbaborane 1 (b. p. 65–75 °C/10^–3 Torr; 0.48 g, 22%) as a colorless oil, identified by NMR as the exo-isomer. – EI-MS (70 eV) for C₁₂H₃₀B₄ (217.6): m/z = 218 (18) [M]⁺, 148 (53) [M–70]⁺, 119 (100) [M–99]⁺. – ¹H NMR (500.1 MHz, C₆D₆, 23 °C). δ¹H = 1.53 (br, 1H, B3-H-B4), 0.11 (br, 2H, B2-B3, B4-H-B5), 0.82 (m, 1H μ-CH), 0.70 (d, 3H, μ-C-Me), 1.10 (s, 3H, C1-Me), 062–0.95 (m, 20 H, B-Et). – ¹¹B NMR (160.5 MHz, C₆D₆, 23 °C: δ¹¹B = see Figs. 1 and 2. – ¹³C NMR (125.8 MHz, C₆D₆, 23 °C): δ¹³C = see Fig. 2. Redistillation of Köster’s “old” sample [8] (0.8 g) containing presumably 1 gave polymers and 0.05 g of 1 along with traces of 7a.

4.2 Thermal degradation of 1, 1,2,3,4,5-pentaethyl-2,5-μ-(1′-dimethylsilylpropane-1′,1′-diyl)-1-carba-arachno-pentaboranes(10) (4) [7], and 1,2,3,4,5-pentaethyl(ethane-1′,1′-diyl)1-carba-arachno-pentaboranes(10) (5) [9]

The compounds were heated without a solvent in a vacuum (10^–2 Torr) for 1 h to 150 °C and all volatile materials were collected in a trap cooled with liquid N₂. Fractional distillation gave ethyldiboranes(6) and BEt₃, followed by the air-stable closo-carbaboranes 7a, b. p. 56–58 °C/11 Torr (from 1), and 7b, b. p. 35–40 °C/1 Torr, (from 4 or 5). Polymeric materials (ca. 70–75%) were left and could not be distilled.