Reactivity of a series of triaryl borates, B(OAr x ) 3 , in hydroboration catalysis †

In this paper, we compare the reactivity of a series of triaryl borates B(OAr x ) 3 as catalysts for the hydroboration of alkenes and alkynes. It was observed that commercially available B(OPh) 3 performed the poorest, whereas catalysts with o -F atoms appeared to perform much better

Organoboron compounds are versatile building blocks in organic synthesis as the high chemical reactivity of the boryl moiety allows for their multiple derivatisations, especially in Suzuki-Miyaura coupling reactions, giving access to numerous natural products and complex organic molecules. 1Therefore, novel approaches for the preparation of these reagents are highly sought after.Hydroboration is one of the simplest methods for the synthesis of a wide array of organoboranes.Typically, these reactions are promoted by precious transition metal complexes based on rhodium, ruthenium, palladium, platinum and others; however, an increasing focus on the application of cheaper and more Earth-abundant alternatives such as first-row transition metals 2 and main group elements 3 has recently been observed.In particular, boron Lewis acids have sparked growing attention (Scheme 1).For example, Hoshi described that dicyclohexylborane and 9-borabicyclo (3.3.1)nonane(9-BBN) can catalyse regioselective cis-hydroboration of alkynes with HBcat (catecholborane) at ambient temperature. 4Thomas reported that simple, commercially available borane adducts, H 3 B•THF and H 3 B•SMe 2 , can be used as effective catalysts for the hydroboration of alkynes and alkenes with HBpin ( pinacolborane), 5 and Okuda demonstrated that alkali metal hydridotriphenylborate complexes [(Me 6 TREN)M][HBPh 3 ] (Me 6 TREN = tris{2-(dimethylamino) ethyl}amine) can serve as efficient catalysts for the hydroboration of a broad range of substrates with carbonyl groups. 6One area that has gained particular attention is where fluorinated aryl borates such as tris( pentafluorophenyl)borane [B(C 6 F 5 ) 3 ] 7 or Piers' borane [HB(C 6 F 5 ) 2 ] are used as pre-catalysts. 8We and Oestreich have later developed the use of other borane catalysts including tris[3,5-bis(trifluoromethyl)phenyl]borane, 9 tris (2,4,6-trifluorophenyl)borane, 10 and tris(3,4,5-trifluorophenyl) borane 11 as effective catalysts for a range of hydroboration reactions.In these cases, the catalytic activity was generally found to be higher than that of the archetypical B(C 6 F 5 ) 3 catalyst.
Earlier this year, we reported the synthesis of a range of fluorinated triaryl borates [B(OAr F ) 3 ] with varying Lewis acidity, prepared by reacting various fluorophenols with BCl 3 . 12This concept stems from Britovsek's findings that the introduction of an O-atom spacer between the boron atom in B(C 6 F 5 ) 3 and the C 6 F 5 -aryl ring increases the Lewis acidity of the borate product B(OC 6 F 5 ) 3 relative to B(C 6 F 5 ) 3 . 13In this project, we were interested in investigating the relative reactivity of these new borates in hydroboration catalysis.
The Lewis acidity of the two new boranes was then determined by both experimental and computational methods.Using the Gutmann-Beckett (GB) Lewis acidity test, 14 the 31 P NMR chemical shifts (δ ppm) of the Et 3 PvO → B(OAr x ) 3 adducts were 78.9 and 78.7 ppm for 1l and 1m, respectively.The change in the 31 P NMR chemical shift between the free phosphine oxide (δ = 52.5 ppm) and the adduct (Δδ) was determined to be 26.4 (1l) and 26.2 ppm (1m).Computational studies 15 at the M06-2X+D3(0)/def2-QZVPP level of theory gave fluoride ion affinities (FIAs) of 336 (1l) and 400 (1m) kJ mol −1 , and hydride ion affinities (HIA) of 317 (1l) and 381 (1m) kJ mol −1 .Finally, the Lewis acidity was also determined using the global electrophilicity index (GEI), 16 which gave values of 1.31 (1l) and 1.39 (1m) eV.In comparison to previously reported borates, 1l has the lowest Lewis acidity of the series when considering Gutmann-Beckett, HIA and FIA values, for example in comparison to the weakly Lewis acidic borate, B(OPh) 3 (GB: Δδ = 23.0 ppm, FIA: 350 kJ mol −1 , and HIA: 323 kJ mol −1 ).On the other hand, 1m shows higher Lewis acidity than previously reported ortho-substituted borates from Gutmann-Beckett, FIA and HIA [i.e., B(O(2-FC 6 H 4 )) 3 has GB: Δδ = 23.8ppm, FIA: 351 kJ mol −1 , HIA: 339 kJ mol −1 ], and comparable Lewis acidity to the most Lewis acidic borates, B(O(3,4,5-F 3 C 6 H 2 )) 3 and B(OC 6 F 5 ) 3 . 12The GEI Lewis acidity metric is intrinsic, and considers the HOMO-LUMO gap rather than the coordination to an external probe and, from this, both 1l and 1m fall within the range of the reported values for F-substituted borates (GEI: 0.88-1.45eV). 12This suggests that the larger chlorine substituents at both the ortho-positions in 1l hinder adduct formation over intrinsic factors significantly more than the previously explored fluorine substituents.However, higher Lewis acidity is observed for 1m with an ortho-CF 3 group, which has comparable size to chlorine, from both intrinsic and extrinsic metrics, suggesting that the electronic effects of the electronwithdrawing CF 3 group outweigh steric effects.

Dalton Transactions Communication
under four different sets of reaction conditions: (a) HBpin (1.2 eq.), catalyst loading (10 mol%), temperature (50 °C), time (24 h); (b) HBpin (2 eq.), catalyst loading (2 mol%), temperature (80 °C), time (48 h); (c) HBpin (2 eq.), catalyst loading (5 mol%), temperature (80 °C), time (48 h); and (d) HBpin (2 eq.), catalyst loading (10 mol%), temperature (80 °C), time (48 h) (Table 1).Using these conditions, we found that the first set of conditions (a) gave the poorest overall results with the catalysts giving yields from 29% (1i) to 67% (1b).Upon changing the HBpin equivalents to 2, increasing the reaction temperature to 80 °C, and the time to 48 h (conditions (c)), the yields predictably increased for all catalysts (range: 69% (1g) to 96% (1e)) whilst keeping the catalyst loading the same.However, when we reduced the catalyst loading to 2 mol% or 5 mol% while keeping the other conditions the same (conditions (b) and (c), respectively), the yields expectedly decreased to 52% (1i)-75% (1f ), but were not as low as those under the initial set of conditions (a) with the exception of catalysts 1a and 1g.While very good yields are obtained for conditions (d), we chose to use conditions (b) as the standard conditions for examining the catalysts in the hydroboration of other substrates as these allowed for better differentiation between the different activities of the catalysts.With these conditions in hand, we investigated different catalysts in the hydroboration of other substrates to compare their catalytic activity (Table 2).Initially, two further alkenes were trialed including the electron-deficient 4-fluorostyrene (2b) and electron-rich 4-methoxystyrene (2c).For all catalysts, the yields increased when using the electron-deficient substrate 2b (range: 58% (1a) to 81% (1m) versus 52% (1i) to 75% (1f )).Conversely, the yields were generally lower with the more electron-rich substrate 2c than those with 2b (range: 44% (1e) to 81% (1m)).These results are interesting as other Lewis acidic boranes such as tris[3,5-bis(trifluoromethyl)phenyl] borane were previously found to give trace products with substrate 2c. 9ollowing this, we investigated the catalysts for the hydroboration of alkynes including electron-neutral phenyl acetylene (4a), electron-deficient 1-ethynyl-4-fluorobenzene (4b), and electron-rich 4-ethynylanisole (4c).For all substrates, the catalysts mostly performed better with the alkyne substrates (4) than the alkene substrates (2).When looking at the catalysts, the least Lewis acidic borate B(OPh) 3 (1a) performed the poorest.The other borates, however, showed little trend between their Lewis acidity and their yield for the reaction.It was noticed, that the catalysts that had ortho-F atoms includ-  as the active catalytic species when using tris[3,5-bis(trifluoromethyl) phenyl]borane. 9We similarly investigated the stoichiometric reactions between 1b and HBpin and found full conversion of both reagents to form 2 new species as revealed in the 1 H NMR spectrum (see the ESI †).We hypothesise that the formation of a catalytically active hydroborate species is stabilised by ortho-fluorine substituents on the borates, and this accounts for their higher activity.Importantly, a control experiment with TMEDA suggested no involvement of B 2 H 6 , which may catalyse the reaction through "hidden boron catalysis" (see the ESI †). 17 Using one of the best catalysts, B(O(2,3,5,6-F 4 C 6 H)) 3 (1j), we investigated a small scope of aliphatic, aromatic, and internal and terminal alkene or alkyne substrates (Scheme 2).Styrene derivatives (2a,b) worked well but a lower yield was observed with a p-OMe substituent (2c) generating 3c in, 48% yield.
In conclusion, we have reported the reactivity of a series of triaryl borates B(OAr x ) 3 as catalysts for the hydroboration of alkenes and alkynes.The catalysts tested included commercially available B(OPh) 3 , previously reported fluorinated triaryl borates, and two new borates with varying ortho-substituents

Communication
Dalton Transactions B(OAr x ) 3 (Ar X = 2,6-Cl 2 C 6 H 3 and 2-(CF 3 )C 6 H 4 ).Although all catalysts were active in the reaction, it was observed that there was no obvious trend between their Lewis acidity and the reaction yield.However, commercially available B(OPh) 3 performed the poorest, whereas catalysts with o-F atoms appeared to perform much better.One of the more active catalysts, B(O(2,3,5,6-F 4 C 6 H)) 3 , was then trialed with a range of aliphatic, aromatic, and internal and terminal alkenes or alkynes.

Table 2
Hydroboration reactions of alkenes and alkynes using borate catalysts a Yields are isolated.Colour scale from dark red to dark blue indicating low to high yields.