A General Solution for Unstable Boronic Acids: Slow-Release Cross-Coupling from Air-Stable MIDA Boronates

Boronic acids can serve as excellent building blocks for the synthesis of a wide range of natural products, pharmaceuticals, and materials.¹ However, some of the potentially most useful boronic acids, including 2-heterocyclic,^2–4 vinyl,⁵ and cyclopropyl⁶ derivatives, are inherently unstable, which can significantly limit their benchtop storage and/or efficient cross-coupling. Many important surrogates have been developed, including trifluoroborate salts,^7–10 trialkoxy or trihydroxyborate salts,^11,12 diethanolamine adducts,¹³ sterically bulky boronic esters,¹⁴ and boroxines.¹⁵ However, none of these can provide air-stable and highly effective substitutes for all three of these challenging boronic acid classes. We herein report that N-methyliminodiacetic acid (MIDA) boronates^16,17 represent the first general solution to this problem by virtue of their uniform benchtop stability and remarkable capacity for in situ slow release of unstable boronic acids (Figure 1).

Figure 1.

2-Heterocyclic, vinyl, and cyclopropyl boronic acids are known to decompose on the benchtop under air via protodeboronation, oxidation, and/or polymerization.^2–6 In addition, these processes are thought to be accelerated in the presence of heat, base, and/or a Pd catalyst, causing the in situ decomposition of unstable boronic acids to compete with their cross-coupling.² This latter challenge is exacerbated in couplings with slower-reacting halides, such as unactivated aryl chlorides.² We hypothesized that both of these problems might be solved if we could achieve rate-controlled in situ hydrolysis of air-stable MIDA boronates, thereby promoting “slow release” of the corresponding unstable boronic acids from bench-stable building blocks.

The hydrolysis of MIDA boronates with aqueous NaOH is fast, typically requiring <10 min at 23 °C.¹⁷ In contrast, we discovered that K₃PO₄ in 5:1 dioxane/H₂O at 60 °C promotes the continuous release of boronic acids over ∼3 h.¹⁸ Remarkably, aryl, heteroaryl, alkenyl, and alkyl MIDA boronates all behave similarly. Moreover, this release rate can be adjusted from 24 h to 30 min by varying the temperature from 23 to 100 °C.¹⁸

Having verified this capacity for slow release, we systematically compared the benchtop stability and cross-coupling efficiency of freshly prepared boronic acids 1a–h^18,19 and the corresponding MIDA boronates 2a–h¹⁸ (Table 1). The benchtop instability of 2-heterocyclic, vinyl, and cyclopropyl boronic acids has frequently been discussed anecdotally,^2–15 yet there is very little quantitative data available. As shown in Table 1, we determined that boronic acids 1a–h all decompose significantly on the benchtop under air over the course of just 15 days (entries 1–8).¹⁸ In fact, with 2-furan, 2-pyrrole, 2-indole, vinyl, and cyclopropyl boronic acids, very little of the original material remains after this time. Alternatively, all of the MIDA boronates 2a–h are indefinitely air-stable, with no decomposition detectable by ¹H NMR even after g60 days on the benchtop under air.¹⁸

Table 1.

Benchtop Stability and Cross-Coupling Efficiency of Boronic Acids and the Corresponding MIDA Boronates

entry	R	% remaining after benchtop storage under aira		4	% isolation yield from cross-couplingc
		1 (15 days)	2 (60 days)		1	2
1		7	>95b		68	94
2		88	>95		50	92
3		80	>95		37	94
4		80	>95b		45	96
5		<5	>95		61	90
6		<5	>95		14	93
7d		5	>95b		79	98
8d		31	>95		95	96

^aFreshly prepared boronic acids 1 and MIDA boronates 2 were stored as solids on the benchtop under air for 15 and 60 days, respectively.¹⁸

^bStored for 107 days.

^cReaction conditions: 1.0 equiv of 3a (1 mmol), 1.0 equiv of 1 (freshly prepared, >95% pure) or 2, 5 mol % Pd(OAc)₂, 10 mol % SPhos, 7.5 equiv of K₃PO₄, 0.07 M in 5:1 dioxane/H₂O, 60 °C, 6 h.

^dCross-couplings were run at 100 °C.

We next tested the cross-coupling efficiency of freshly prepared boronic acids 1a–h¹⁹ with aryl chloride 3a using Pd(OAc)₂/SPhos^2,20 as the catalyst and K₃PO₄ as the base. Only very low to moderate yields (14–68%) were observed for the 2-heterocyclic derivatives 1a–f (entries 1–6), consistent with our observations that boronic acid decomposition kinetically competes with cross-coupling.^2,18 In stark contrast, all of the corresponding MIDA boronates 2a–f coupled under identical conditions with aryl chloride 3a in uniformly excellent yields (90–96%) using in each case only 1 equiv of MIDA boronate.²¹ In many cases, the improvement in yield using 2 versus 1 is striking [e.g., 92 vs 50% with 2-benzofuran (entry 2), 94 vs 37% with 2-thiophene (entry 3), and 93 vs 14% with 2-indole (entry 6)]. In addition, vinyl MIDA boronate (2g) was significantly more effective than freshly prepared 1g (entry 7).^5c,8

Consistent with our hypothesis that these increases in yield are attributable to in situ slow release of the corresponding boronic acids, no significant differences in yields were observed for 1a (64%) vs 2a (59%) under fast-release conditions, i.e., using aqueous NaOH as base.¹⁸ Moreover, the high yield observed for 2a under slow-release conditions was replicated via syringe-pump-mediated addition of freshly prepared 1a over the course of 3 h.¹⁸ It is noteworthy that cyclopropyl boronic acid (1h) prepared immediately prior to the reaction can be as effective as MIDA boronate 2h (entry 8), suggesting that benchtop decomposition of 1h may be in large part responsible for the challenges frequently encountered with this boronic acid.^6a

Encouraged by these results, we explored the scope of this slow-release method and found that even some of the most challenging aryl and heteroaryl chlorides can be efficiently coupled with MIDA boronates 2a–h (Table 2). For example, the highly deactivated (electron-rich and sterically hindered) compound 2,4-dimethoxychlorobenzene (3b) represents an exceptionally difficult cross-coupling partner for unstable 2-heterocyclic boronic acids. Nonetheless, just 1.2 equiv of the corresponding MIDA boronates promoted this coupling in generally excellent to outstanding yields (entries 1, 5, 9, 12, and 14). Because of the great importance of polyheterocyclic scaffolds in pharmaceuticals, similar cross-couplings with inexpensive and readily available heteroaryl chlorides would also be highly valuable. We explored this possibility with 3d–i and found that the 2-heterocyclic MIDA boronates are highly effective in such couplings (entries 3, 4, 6–8, 10, 11, 13, and 15). Even electronically deactivated heteroaryl chlorides such as 3f–h were coupled to 2b in good to excellent yields (entries 6–8).

Table 2.

Slow-Release Cross-Coupling of Air-Stable 2-Heterocyclic, Vinyl, and Cyclopropyl MIDA Boronates with Aryl and Heteroaryl Chlorides^a

entry	2	3	4	% isolated yield
1				99
2	2a			97
3	2a			99
4	2a			91
5		3b		94
6	2b			94
7b	2b			85
8b	2b			85
9		3b		98
10	2c	3d		99
11	2c			97
12c		3b		81
13c	2e	3d		98
14		3b		97
15	2f	3d		93
16d,e		3c		91
17d,e	2g	3i		87
18d,e	2g	3g		76
19d,e	2g	3d		96
20b,d,f		3c		79
21d	2h	3b		97

^aGeneral reaction conditions: 1 equiv of aryl halide (1 mmol), 1.2 equiv of MIDA boronate, 5 mol % Pd(OAc)₂, 10 mol % SPhos, 7.5 equiv of K₃PO₄, 0.07 M in 5:1 dioxane/H₂O, 60 °C, 6 h.

^bUsing 1.5 equiv of MIDA boronate.

^cUsing 0.5 mmol of aryl halide, 0.6 mmol of MIDA boronate (1.2 equiv)

^dAt 100 °C.

^eReaction time 2 h.

^fReaction time 24 h.

Vinylation of aryl and heteroaryl halides can provide styrene-like building blocks for a wide range of small molecules and materials.^5c Thus, the development of a highly effective, nontoxic, environmentally friendly, and air-stable vinyl metal species has long been an important goal.^5c Remarkably, vinyl MIDA boronate 2g embodies all of these favorable properties and efficiently coupled even with the highly deactivated aryl chloride 3c (entry 16) as well as a variety of heteroaryl chlorides (entries 17–19). Finally, 2h coupled with highly deactivated aryl chlorides 3c and 3b (entries 20–21).

As a final example, the 2-pyridyl subunit appears with remarkable frequency in biologically active small molecules. However, the corresponding boronic acid is notoriously unstable^4,22 and difficult to cross-couple,^4,7,11a,23 particularly with aryl chlorides.^11a Currently available surrogates either are not air-stable^11a,b,12 or cannot be isolated in chemically pure form.¹³

In contrast, 2-pyridyl MIDA boronate (2i) is isolable as a chemically pure and air-stable solid (X-ray structure shown in Table 3, ¹H NMR spectra showed no decomposition after 60 days on the benchtop under air). Consistent with a relatively lower rate of transmetalation for 2-pyridylboranes,^11a conditions like those used in Tables 1 and 2 were not generally effective for couplings with 2i. However, driven by the hypothesis that in-situ-generated 2-pyridyl boronic esters would be more stable than their boronic acid counterparts, we explored a variety of alcohol-containing solvent mixtures and found that DMF/IPA was advantageous. Moreover, it has been demonstrated that the addition of CuI^13,11b or CuCl^14c can promote cross-couplings with other 2-pyridylboranes. We therefore surveyed a series of copper salts and found that the inexpensive and nontoxic Cu(OAc)₂ was especially beneficial. As shown in Table 3, under these modified slow-release conditions, air-stable 2-pyridyl MIDA boronate (2i) can be cross-coupled with a variety of aryl and heteroaryl chlorides (entries 1–5). The capacity to effectively cross-couple two different 2-substituted heterocycles is a notable advantage of this methodology (Table 3, entries 3–5; also see Table 2, entries 4 and 11).

Table 3.

Slow-Release Cross-Coupling of Air-Stable 2-Pyridyl MIDA Boronate 2i with Aryl and Heteroaryl Chlorides^a

entry	3	4	% isolation yield
1			72
2			60
3			79
4			52
5			74

^aReaction conditions: 1.0 equiv of aryl halide 3 (1 mmol), 1.5 equiv of MIDA boronate 2i, 1.5 mol % Pd₂(dba)₃, 6 mol % XPhos, 50 mol % Cu(OAc)₂, 5 equiv of K₂CO₃, 0.1 M in 4:1 DMF/IPA, 100 °C, 4 h.

In summary, several highly advantageous features collectively make MIDA boronates an outstanding platform for the preparation and utilization of organoboranes in organic synthesis.¹⁷ These include reversibly attenuated reactivity toward anhydrous cross-coupling conditions, compatibility with a wide range of synthetic reagents, air stability, solubility in many common organic solvents, monomeric constitution, and compatibility with silica gel chromatography.¹⁷ We now report that MIDA boronates also possess the highly enabling capacity for in situ slow release of the corresponding unstable boronic acids. This remarkably general solution has transformed a wide range of unstable boronic acids into air-stable and highly effective cross-coupling partners, many of which are now commercially available.²⁴