Hydroboration with Pyridine Borane at Room Temperature

Hydroboration has been an essential reaction in synthetic organic chemistry since Brown’s discovery that borane etherates are reactive at room temperature.^1,2 Diverse hydroborating agents including THF borane, dimethyl sulfide borane, 9-BBN, and thexylborane are readily available and offer many options for selective hydroboration.¹ However, each has limitations as well as advantages and all are air-sensitive. The far more stable pyridine borane (py·BH₃) has also been considered as a hydroborating agent,^3a,b but heating to 75-100 °C is required for dissociation to free borane, a prerequisite for π-complexation of the olefin and eventual hydroboration. Hindered amine boranes dissociate more readily and react at lower temperatures, but they are air-sensitive.^3c The remaining challenge is to obtain high reactivity without compromising reagent stability and practicality.

We have explored the possibility of activating Py·BH3 by replacing one of the hydrides with a good leaving group (Scheme 1; Py·BH₂X(1) with X ) I, Br, OTf, NTf2). If this approach is used, the strength of the B-N bond would no longer be problematic, provided that departure of the new leaving group X leads to hydroboration. This might occur by some process equivalent to SN2-like displacement of X to form the olefin π-complex 2 (path A) or an SN1-like heterolysis via 5 (path B), followed by 4-center addition of B-H(3) to give 4. A third possibility is dissociation of 1 to BH₂X(6, path C), conventional hydroboration, and complexation with pyridine to afford 4. Prior studies show that intramolecular hydroborations using activated, unsaturated amine and phosphine boranes are consistent with internal versions of paths A or B.⁴ We now report that a similar hydroboration pathway is also viable as an intermolecular process.

Scheme 1.

Several amine boranes and activation methods were compared to see if intermolecular hydroboration according to Scheme 1 is possible. The best results were achieved when commercially available pyridine borane (Py·BH3) was activated with 50 mol % of I2 in dichloromethane to generate Py·BH₂I(1; rapid hydrogen evolution).⁵ Addition of β-methylstyrene followed by oxidative workup gave alcohol products (92%; 15:1 ratio, 7/8; entry 1, Table 1). This improved selectivity, compared to the 5:1 ratio using BH3·THF,⁶ suggests that activation produces a unique hydroborating agent and does not simply release BH3. Activation of Py·BH3 with bromine gave higher selectivity, but a much slower reaction (entry 2), while TfOH and HNTf2 (entries 3 and 4) induced faster but less selective hydroborations.

Table 1.

Hydroboration of β-methylstyrene with L·BH₂X

entry	L·BH3a	activation	time (h)	7:8	yield (%)
1	Py·BH3	I2b	2	15:1	92
2	Py·BH3	Br2b	12	>20:1	10d
3	Py·BH3	TfOHc	2	10:1	72
4	Py·BH3	HNTf2c	2	10:1	90
5	Lut·BH3	I2b	2	2.4:1	13d
6	Me2S·BH2Ie		2	3.5:1	62

^a1:1 ratio, L·BH₃/alkene, room temperature.

^b50 mol %.

^c100 mol %.

^dReaction quenched prior to completion.

^ePreformed (ref ⁷).

Next we compared the reagent 1 (X ) I) with Lut·BH₂I (Table 1, entry 5; from lutidene borane + I2) and the known Me₂S·BH₂I (entry 6).⁷ Different hydroboration regioselectivity was found in each case, and unique ¹¹B NMR signals were observed prior to the addition of alkene (Py·BH₂I, δ -28.5 ppm; Lut·BH₂I, δ -34.5 ppm; Me₂S·BH₂I, δ -20.5 ppm). The NMR data do not exclude the presence of BH₂I in equilibrium with L·BH₂I in one or more cases, but the regioselectivity results (entries 1, 5, and 6) prove that dissociation (as in path C) cannot be the only reaction pathway.

The hydroboration of 1-Ph-1-propyne (9) with BH3·THF is reported to give a 3:1 ratio of 10/11, while sia₂BH, thexylBH₂, catecholborane, and Br₂BH·SMe₂ afford mostly 11.⁸ In contrast, Py·BH₂I produces a striking 15:1 selectivity favoring 10 (Table 2, entry 1), an effect that is amplified for p-CF₃Ph-1-propyne and reversed for the p-MeOPh analogue (entries 2 and 3). Related trends are reported for styrene hydroboration.⁹ Lut·BH₂I reacts nonselectively (entry 4), but Me₂S·BH₂I gives 10 with only traces of 11 (entry 5). Other alkynes (entries 8 and 9) are hydroborated with low regioselectivity, similar to the results with BH3·THF.⁸

Table 2.

Alkyne Hydroboration

entry	L	alkyne	R	R′	10:11	yield (%)
1	Py	9a	Ph	CH3	15:1	64
2	Py	9b	pCF3Ph	CH3	>20:1	NA
3	Py	9c	pMeOPh	CH3	1:2	NA
4	Lut	9a	Ph	CH3	1.2:1	51
5	Me2S	9a	Ph	CH3	30:1	46
6	Py	9d	Ph	C2H5	10:1	66
7	Py	9e	C3H7	C3H7		63
8	Py	9f	CH3	C5H11	1.5:1	61
9	Py	9g	CH3	cC6H11	3:1	64

The simplest interpretation of the pyridine and lutidine borane results (Tables 1, 2) is that the ligand (L ) Py or Lut) remains attached to boron in the product-determining step for each reaction(Table 1, entries 1 and 5; Table 2, entries 1 and 4). However, the data require only that the Py·BH₂I reagent follows a pathway different from path C (Scheme 1), assuming that the reaction of Me₂S·BH₂I involves dissociation to free BH₂I.

Rate-determining dissociation of 1 (X ) I) to 5 (path B) is ruled out because the rate of methylstyrene hydroboration with Py·BH₂I increases with alkene concentration (qualitatively, first order in alkene). The strong counterion dependence for hydroboration regiochemistry (Table 1) also argues against formal dissociation in an SN1-like mechanism, but neither the rate nor the regiochemistry data can rule out pathways where the conversion from 5 to 3 is rate-limiting if species analogous to tight ion pairs are involved. Path A (Scheme 1) is the simplest rationale that is consistent with facile hydroboration from Py·BH₂I at room temperature. By way of analogy, Ryschkewitsch et al. have reported that Py·BH₂I reacts readily with nitrogen nucleophiles, resulting in iodide displacement in an SN2-like process.^5b Of course, the alkene is a much weaker nucleophile, and thus it would be premature to conclude that it can be sufficiently reactive to trigger the simplest version of path A. Furthermore, tight ion pair versions of path B cannot be ruled out, and other mechanistic variants remain to be evaluated.

Good functional group compatibility was observed with the Py·BH₂I reagent (Table 3). Hydroboration of 12 followed by oxidative workup gave >95% primary alcohols 13 (NMR assay). Complete conversion of ester, amide, and amine substrates 12d-g required 2 equiv of Py·BH₂I, but no reduction of these functional groups was observed within 2 h at room temperature. On the other hand, reduction of ketones and carboxylic acids (12,R = C(O)Me or CO²H) was fast compared to hydroboration of the alkene.

Table 3.

Functional Group Compatibility

entry	alkene	R	yield (%)
1	12a	n-C6H13	98
2	12b	OBn	83
3	12c	OTBS	83
4	12d	OBz	84a,b
5	12e	NBn2	74a
6	12f	NHBn	80a
7	12g	NHBz	89a,b

^a2:1:1 py·BH₃/l₂/alkene; 2 h at room temperature; NaOOH/MeOH.

^bOxidative workup: NaBO₃·H₂O, THF/H₂O.

Monoalkyl boronic acid derivatives cannot be generated directly from unhindered alkenes using BH3·THF because the initially formed monoalkylborane is more reactive in hydroboration than is the parent BH3.¹⁰ However, the Py·BH₂I method forms the 1:1 adducts considerably faster than 2:1 adducts, as might be expected according to path A (Scheme 1). Thus, hydroboration of 1-dodecene 12a was monitored after quenching in MeOH using positive ion detection ESMS. Strong signals for the 1:1 adducts 14 (Z ) MeO, Py) were observed, together with a weak signal for the 2:1 adduct 15 (Chart 1). Subsequent treatment with KHF₂¹¹ allowed assay in the negative ion detection mode. A strong signal for 16 was observed, but 17 was not detected after precipitation from acetonitrile. Preparative experiments were performed from alkenes 18 to afford the corresponding potassium alkyltrifluoroborates 19 in 59-84% yield (Table 4). In all cases, ESMS with negative ion detection revealed the presence of 1:1 adducts, but not 2:1 adducts. On the other hand, use of excess alkene allowed the ESMS detection of a substantial signal for 17.

Table 4.

Preparation of Potassium Alkyltrifluoroborates 19

entry	alkene	R	R′	R″	yield (%)
1	18a	Ph	H	H	84
2	18b	C4Hg	H	H	76
3	18c	H	-C4H8-		82
4	18d	Ph	CH3	H	61
5	18e	Ph	-C4H8-		59

Molander has shown that alkyltrifluoroboratesalts are attractive reagents for Suzuki coupling applications,¹²but preparation of these salts required the use of catecholborane or BBr₂H·SMe₂. The Py·BH₂I hydroboration is a simple alternative that cleanly affords the 1:1 adducts 19 and provides a high-yielding and convenient route to useful organoborane substrates.

In conclusion, we have presented evidence for an unusual hydroboration mechanism involving leaving group displacement from activated pyridine boranes 1. Hydroboration with Py·BH₂Iis easily controlled to give the monoadducts and does not require handling sensitive trivalent boranes.

Chart 1.