Organoboranes and tetraorganoborates studied by ¹¹B and ¹³C NMR spectroscopy and DFT calculations

1 Introduction

The considerable importance of organoboron compounds in synthesis [1–3] requires careful characterization by reliable and efficient analytical techniques. In this context, ¹H NMR spectroscopy is helpful, although many organyl groups do not give simple first order spectra. In spite of the quadrupolar nature of ¹¹B (spin I = 3/2), ¹¹B NMR spectra are extremely useful, since in many cases the chemical shifts δ¹¹B are diagnostic for the surroundings of the boron atom [4]. Typically, ¹³C NMR spectroscopic measurements [5] often provide conclusive evidence for the solution-state molecular structures. However, even nowadays, with modern powerful NMR spectrometers, one finds frequently the message that the ¹³C NMR signal of a carbon atom linked to boron was not observed. Indeed, this signal can be rather broad because of partially relaxed scalar ¹³C-¹¹B spin-spin coupling, and careful measurements are necessary, in particular if ¹³C nuclear spin relaxation tends to be slow. One can distinguish three cases, depending on ¹¹B nuclear spin relaxation: (i) long relaxation time T^Q(¹¹B), typical for highly symmetrical surroundings (or with a small electric field gradient) of the boron atom as in borates, e.g. in [BPh₄]^–, for which all ¹³C–¹¹B spin–spin couplings over one-, two-, three- and even four bonds are perfectly resolved [6]; (ii) moderately fast ¹¹B nuclear spin relaxation, e.g. in trigonal boranes of low molecular weight, which allows sometimes to observe residual splitting of the ¹³C NMR signal [7–9] (see also Fig. 1); (iii) rather fast ¹¹B nuclear spin relaxation, observed e.g. for most arylboranes, leading to more or less broadened ¹³C NMR signals for boron-bonded carbon atoms (see examples for ¹³C NMR spectra in [10, 11]).

In the latter cases (iii), the coupling constants ¹J(¹³C,¹¹B) can be evaluated using eq. (1) [12] with T^Q(¹¹B) = [πh_1/2(¹¹B)]^–1, since scalar relaxation of the second kind (T₂^SC(¹³C) [13]) is by far the major source of broadening of these ¹³C(B-C) NMR signals. In the absence of chemical exchange and when the line width exceeds 10–15 Hz, other contributions to the line width Δν_b(¹³C) can be safely neglected, an advantage of the magnetically diluted ¹³C nucleus. The data ¹J(¹³C,¹¹B) thus obtained can be compared with calculated data based on quantum-chemical calculations of coupling constants [14, 15], to be checked by comparison with directly measured data ¹J(¹³C,¹¹B) for the cases (i) and (ii). The numerous examples depicted in Table 1 indicate that both evaluation of ¹J(¹³C,¹¹B) from line widths as well as quantum-chemical calculations are fairly reliable.

Fig. 1:

Part of the 50.3 MHz ¹³C{¹H} NMR spectrum of methylamino-dimethylborane (24 in Table 1; 5 % in C₆D₆ at 27 °C) showing the ¹³C(BMe) NMR signals (hindered rotation about the B–N bond) split by partially relaxed ¹³C–¹¹B spin–spin coupling (the distance in Hz between the inner lines represents the coupling constant).

Table 1

¹¹B and ¹³C NMR parameters of some organoboranes and -borates.

Compounda	δ11B	δ13C	nJ(13C,11B) (n) in Hz calcd.b [exp.]c
1 BMe3	86.0	14.8	49.1 (1) [52.0]
2 [BMe4]–	–20.7	13.3	41.0 (1) [39.6]
3 BEt3	86.5	19.8 (BCH2), 8.5	49.1 (1) [50.0], –1.2 (2)
4 [BEt4]–	–17.5	17.6 (BCH2), 11.7	42.1 (1) [40.7], –1.0 (2) [1.1]
5 BnPr3	86.0	31.1 (BCH2), 18.1, 17.6	49.0 (1) [50.0], –1.0 (2), 2.2 (3)
6 [BnPr4]–	–17.8	33.6 (BCH2), 21.6, 18.8	41.1 (1) [41.0], –0.8 (2) [1.0], 4.8 (3) [4.7]
7 BPh3	68.0	143.3 (br, i), 138.4 (o), 127.3 (m), 131.1 (p)	64.9 (1) [66.0], 2.9 (2), 3.4 (3), –0.5 (4)
8 [BPh4]– d	–8.0	164.8 (i), 128.8 (o), 129.8 (m), 125.8 (p),	51.1 (1) [49.4], 1.8 (2) [1.5], 2.9 (3) [2.7], –0.5 (4) [0.5]
9	44.3	147.0 (br, 2), 129.9 (3), 112.3 (4), 132.1 (5), 35.2 (NMe)	76.6 (1) [75.0], 4.9 (2), 3.5 (3), 2.8 (4)
10	–18.8	162.7 (2), 116.9 (3), 106.7 (4), 119.5 (5), 32.5 (NMe)	58.7 (1) [57.5], 3.4 (2) [3.5], 2.7 (3) [2.8], 2.0 (4) [2.0]
11	35.0	166.7 (br, 2), 132.5 (3), 111.4 (4), 148.9 (5)	84.6 (1) [83.0], 7.2 (2), 2.4 (3), 3.8 (4)
12	–19.9	185.6 (2), 110.7 (3), 106.8 (4), 137.9 (5)	63.2 (1) [64.0], 5.6 (2) [5.8], 2.1 (3) [2.0], 2.4 (4) [2.5]
13	47.3	144.2 (br, 2), 141.5 (3), 128.9 (4), 136.4 (5)	76.4 (1) [77.0], 4.1 (2), 4.3 (3), 2.5 (4)
14	–14.1	179.9 (2), 127.4 (3), 122.3 (4), 127.0 (5)	57.5 (1) [58.0], 5.8 (2) [6.0], 2.9 (3) [3.0], 2.5 (4) [2.5]
15	57.0	145.0 (br, 2), 140.0 (3), 129.0 (4),136.1 (5), 12.3 (br, BMe)	73.9 (1) [71.5], 4.1 (2), 4.1 (3), 22.6 (4)Me: 53.6 (1) [54.0]
16	49.0	148.7 (br, 2), 145.1 (3), 128.0 (4), 136.7 (5) Ph: 146.1 (br, i), 136.8 (o), 129.7 (m), 130.7 (p)	75.1 (1) [73.0], 4.1 (2), 4.2 (3), 2.6 (4) Ph: 66.9 (1) [65.0], 2.9 (2), 3.5 (3), –0.6 (4)
17	35.2	142 (br, 2), 129.8 (3), 127.8 (4), 135.2 (5) NEt2: 44.5, 16.2	79.9 (1) [78.0], 3.3 (2), 4.5 (3), 2.1 (4)
18	48.5	139.1 (br, 2), 143.0 (3), 128.6 (4), 140.2 (5)	107.7 (1) [110.0], 5.4 (2), 5.7 (3), 3.8 (4)
19 [B(C≡C–Ph)4]– d	–31.0	102.8 (B–C≡), 94.1 (≡C–Ph)	76.7 (1) [70.0], 15.7 (2) [14.0]
20e	62.1	139.5 (br, 9a), 153.4 (4a), 133.0 (2), 129.3 (3), 120.7 (4)	76.20 (1) [75.0], 8.2 (2), 3.9 (2), 3.8 (3), 3.1 (3), –0.3 (4)
21 EtBCl2	63.4	28.1, 8.3	71.8 (1) [74.0], –1.0 (2)
22 PhBCl2	55.0	134.0 (br, i), 126.8 (o), 136.8 (m), 134.9 (p)	86.6 (1) [88.0], 4.5 (2), 6.0 (3),–0.6 (4)
23 Me2B–NH2	47.1	5.6	55.6 (1) [59.0]
24 Me2B–N(H)Me	45.7	1.5, 6.5, 29.4 (NMe)	57.0 (1) [56.0], 56.0 (1) [56.0]
25 Et2B–NH2	47.3	11.2 (BCH2), 8.7	55.7 (1) [57.0], –1.2 (2)
26 [Et3B–NH2]– d	–13.5	21.5 (BCH2), 10.1	44.0 (1) [43.0], –1.9 (2)

^aBoranes measured in CDCl₃ solutions, except 23–25 (in C₆D₆), and borates as lithium salts in [D₈]THF; ^bcalculated for optimized geometries at the B3LYP/6-311+G(d,p) level of theory; ^cdirect measurements from ¹³C{¹H} NMR spectra for 1, 3, 21, 23–25 (± 2 Hz); and all borates (± 1 Hz); for all other boranes, the experimental coupling constants ¹J(¹³C,¹¹B) were evaluated (± 3 Hz) from line widths measured from ¹³C{¹H} and ¹¹B{¹H} NMR spectra; ^dsodium salt; ^eRef. [10].

(1)Δνb(13C) = (4/3) π 3/2(3/2 + 1) [J(13C,11B)]2 [TQ(11B)] (1)

Expectedly, all calculated coupling constants ⁿJ(¹³C,¹¹B) for n = 1 possess a positive sign [7, 9, 16]. For n = 2, calculations show negative or positive signs depending on the nature of the intervening carbon atom. In the cases of [BPh₄]^–(8) and the heteroaryborates 10, 12, 14, the calculated positive sign of ⁿJ(¹³C,¹¹B) (n = 2, 3, 4) can be confirmed by appropriate selective heteronuclear ¹H{¹¹B} experiments [17], observing ¹³C satellite signals in the ¹H NMR spectra (Fig. 2).

Fig. 2:

Two representative examples of the determination of coupling signs ⁿJ(¹³C,¹¹B) relative to ¹J(¹³C,¹H) (known as > 0) by selective low-power irradiation of ¹¹B transitions at low or high frequency, sharpening the respective low- or high-frequency ¹³C satellite signal in the ¹H NMR spectrum. ¹H and ¹¹B are the “active” nuclei and ¹³C (neither irradiated nor observed) is the "passive" nucleus.

2 Experimental

Most compounds (Table 1) were commercially available or obtained following literature procedures [1]. Heteroarylboranes were accessible as reported [18, 19]. Heteroarylborates were prepared as lithium salts in situ, as described previously [20], extracted with ether, dried and dissolved in [D₈]THF. NMR measurements: Bruker WP200, ARX 250 and DRX 500 spectrometers in 5 mm (o.d.) tubes at 23 or 27 °C. Chemical shifts are given in ppm relative to BF₃-OEt₂ (δ¹¹B = 0 ppm for Ξ(¹¹B) = 32.083971 MHz) and SiMe₄ [δ¹³C(CDCl₃) = 77.0, δ¹³C([D₈]THF) = 67.4, 25.2 ppm]. Typically, 1000 to 4000 transients were used to record ¹³C{¹H} NMR spectra with a S/N ratio sufficient for line width measurements of broadened ¹³C(B-C) NMR signals. The decoupler power for selective ¹H{¹¹B} experiments (Bruker DRX 500) was carefully calibrated by observing the desired effects for the central signals of the respective multiplets before measuring the ¹³C satellite signals in the ¹H NMR spectra of the tetraorganoborates. In the case of ⁴J(¹³C,¹¹B) for [BPh₄]^– (8), the calculated negative sign of the coupling constant could not be confirmed, since there was no differential sharpening of the ¹³C satellites upon selective ¹¹B irradiation. Presumably the coupling constant ⁵J(¹¹B,¹H) is too small (calcd. 0.25 Hz),

All quantum-chemical calculations were carried out using the Gaussian 09 program package [21].