Tetrahedral boronates as basic catalysts in the aldol reaction

Tobias Müller; Kristina Djanashvili; Joop A. Peters; Isabel W.C.E. Arends; Ulf Hanefeld

doi:10.1515/znb-2015-0029

Article Publicly Available

Tetrahedral boronates as basic catalysts in the aldol reaction

Tobias Müller , Kristina Djanashvili , Joop A. Peters , Isabel W.C.E. Arends and Ulf Hanefeld

Published/Copyright: June 24, 2015

Published by

Become an author with De Gruyter Brill

Author Information Explore this Subject

From the journal Zeitschrift für Naturforschung B Volume 70 Issue 8

Abstract

β-Hydroxyketones are versatile building blocks in organic synthesis, which can be conveniently synthesized from ketones and aldehydes by aldol reactions. Unfortunately, these reactions often suffer from dehydration of the initially formed β-hydroxyketones. Previously, tetrahedral 3,5-difluorophenylborate was shown to be an efficient and selective catalyst for this reaction. The present investigation concerns the catalytic performance of phenyl borates with different substitution patterns in the aldol reaction. It appears that the dehydration reaction can be suppressed by selecting substituents and substituent positions with reduced electron withdrawing effects on the borate function. Optimal suppression of the dehydration of β-hydroxyketones was obtained for compounds corresponding to phenylboronic acids with a pK_a > 7. The reactions between benzaldehyde and butanone or 3-pentanone did not show diastereoselectivity, which suggests that the catalysts merely act as bases rather than as templates for the transition state of the aldol reaction. Sterically more demanding ketones were not converted.

Keywords: aldol condensation; aldol reaction; boron; catalysis; dehydration

1 Introduction

The aldol reaction was discovered 150 years ago and originally was performed with stoichiometric amounts of sodium to deprotonate the enol in the activation step of the reaction [1]. Later, other inorganic bases such as LiOH, NaOH, Na₂CO₃, and Ca(OH)₂ were applied in stoichiometric amounts for this purpose [2–5]. Unfortunately, the base used for the initial deprotonation of the ketone is often also active in an undesired consecutive reaction, dehydration of the formed β-hydroxyketone. Suppression of this elimination reaction is highly desired as β-hydroxyketones can be used as building blocks for diols, amino alcohols, and polyketides [6–11].

A way to increase the chemo-selectivity towards the β-hydroxyketone was introduced by Mukaiyama et al., applying stoichiometric amounts of silyl enol ether as donor in the aldol reaction to form β-hydroxyketone [12–14]. A further improvement was introduced by Mori et al., who applied diarylborinic acid in the Mukaiyama reaction [15]. The advantage is a higher stereoselectivity of the reaction than that obtained before. This can be explained by a shorter bond length of the B–O bond as compared to metal oxygen bonds, resulting in a tighter six-membered ring transition state of the addition reaction [16, 17]. The boron-mediated reaction by Mori with silyl enol ethers was performed in water at 30 °C.

More recently, activation of the ketone by the highly reactive trimethylsilyl (TMS) chloride was avoided by generating the boron enolate in situ [18]. However, with this procedure, high selectivities towards aldols could only be obtained when the ketone had an α-hydroxygroup.

In a recent communication, we presented a borate salt, as a basic catalyst for the aldol reaction, which showed high selectivity towards the aldol product with a wide range of aldehydes. This catalyst was soluble in organic solvents, which allowed us to perform the reaction in the substrate (acetone) as solvent [19]. No further co-solvent was needed, which is desirable for an environmentally friendly process. The good solubility of the catalyst was achieved by attaching electron withdrawing groups to the aromatic ring. Furthermore, the electron withdrawing groups stabilized the negatively charged borate function.

Here, we report on an investigation of the influence of electron withdrawing groups on the performance of these borate salts as catalysts in the aldol reactions. Fine tuning of β-hydroxyketone selectivity was obtained in this way. Additionally, different ketones were investigated to support the previously proposed mechanism with the catalyst acting as a base rather than as a template.

2 Results and discussion

The catalysts were prepared in two steps. First, the substituted phenylboronic acid was esterified with isopropanol in toluene using a Dean–Stark trap to remove the produced water.

Then, the desired catalysts were obtained by reacting the resulting diesters with an equimolar amount of sodium isopropanolate (see Scheme 1). An attempt to prepare the 2,6-difluoro and 2,4,6-trifluoro substituted borate salts failed due to cleavage of the B–C bond during the reaction with sodium isopropanolate. This is most likely caused by an increase of steric strain and electrostatic repulsion during the conversion of the boron atom from planar trigonal to tetrahedral [20, 21].

Scheme 1:

Synthesis of the catalysts. The pK_a values of the starting boronic acids (1) are given in brackets [22, 23].

The catalysts were tested in the aldol reaction of benzaldehyde (4) and acetone (5) at 30 °C (see Scheme 2). High conversions were obtained rapidly after mixing the reactants. With 10 mol% of catalyst, a maximum conversion to the desired β-hydroxyketone (6) was obtained within 10 min. In addition, small amounts of side products (mainly 7) were obtained due to elimination and oligomerization reactions.

Scheme 2:

The aldol reaction catalyzed by borate salts 3a–f.

Figure 1a and b display the course of the reaction as a function of time (0–90 min) for the reaction using 10 or 20 mol% of catalyst 3, respectively. After fast conversion of benzaldehyde (4) to the aldol product 6 during the first few minutes, a relatively slow decrease of the concentration of 6 occurs. Analysis of the reaction mixture showed that this is caused by elimination of water from 6 to give 7 and minor amounts of its oligomerization products. The large difference in reaction rates between the aldol reaction and the subsequent elimination allows termination of the reaction when the yield of aldol is maximal (Table 1).

Fig. 1:

Amount of aldol (6) (in % of initial benzaldehyde 4 concentration) as a function of time during the reaction of 2 mmol of 4 in 20 mmol of acetone at 30 °C in the presence of 10 mol% of 3 (a) and 20 mol% of 3 (b); 3a ▴, 3b •, 3c ■, 3d ◂, 3e ▾, and 3f ▸.

Table 1

Conversion and selectivity for the boron catalyzed aldol reactions between benzaldehyde (2 mmol) and acetone (20 mmol), at 30 °C after 10 min.^a

Entry	Catalyst	Amount of catalyst (mol%)	Conversion (%)	Aldol 6 (%)^b
1	3a	10	93	87
2	3b	10	90	80
3	3c	10	94	73
4	3d	10	93	91
5	3e	10	95	62
6	3f	10	94	77
7	3a	20	96	62
8	3b	20	95	77
9	3c	20	95	72
10	3d	20	96	64
11	3e	20	97	60
12	3f	20	96	56
13	3f	5	42	40
14	3f	2	25	23

^aAs determined by ¹H NMR spectroscopy.

^bThe remainder is mainly elimination product 7 in addition to minor amounts of products derived from sequential oligomerization and elimination reactions.

The final slopes of the curves in Fig. 1 suggest that the relative rates of the elimination reaction for the various catalysts increase in the order 3a < 3b < 3c < 3d < 3e < 3f. This order is the same as that of the pK_a values of the corresponding phenylboronic acids (1) in water (see Scheme 1) [22, 23]. The pK_a values of the boronic acids are determined by the charge density at the boron atom, and hence it can be concluded that the activities in the elimination reaction with catalysts 3a–f depend on the charge density of the boron atoms. Those with the highest positive charge density (the catalysts derived from the boronic acids with the lowest pK_a) show the lowest elimination rates, whereas the rates of the aldol reaction are very high for all catalysts investigated.

The initial rates of the reactions with 10 or 20 mol% of catalyst were too high to enable accurate kinetics. Neither reducing the catalyst concentration (Table 1, entries 13 and 14) nor cooling the reaction mixture enabled evaluation of accurate kinetics. A comparison of the benzaldehyde (4) conversion at various catalyst concentrations (see Table 1) and that of the final slopes of the aldol concentration as a function of time (Fig. 1) suggest that both the rates of the aldol reaction and the subsequent elimination reaction are approximately first order in catalyst, which underlines the importance of the catalyst in the rate determining step of these reactions.

The effect of variation of the ratio 4:5 was investigated in reactions with 10 mol% of 3b under the conditions given in Table 1. An increase of this ratio from 1:5, 1:10 to 1:20 led to an increase of the aldol reaction rate as reflected in an increase of the conversion from 55 % to 95 %, while the selectivity for aldol 6 remained about 80 %. All phenomena observed are consistent with the previously suggested mechanism (see Scheme 3) [19].

Scheme 3:

The mechanism of aldol reaction catalyzed by the synthesized borate salts 3a–f.

Catalyst 3 deprotonates acetone to form the boron-bound enolate. Then reaction with the aldehyde results in β-hydroxyketone (6). The subsequent dehydration is not catalyzed by any of the catalysts 3 due to their steric bulk. This is in contrast to sterically less demanding bases such as NaOiPr [19].

Increasing the positive charge density on the boron atoms through the substituents at the phenyl groups of the catalysts will make them more electron withdrawing and hence the neighboring O atoms will have higher negative charge densities. Consequently, these catalysts will be stronger Brønsted bases leading to enhancement of the reaction rate of the deprotonation of acetone. Accordingly, catalysts 3 corresponding to boronic acids (1) with the lowest pK_a values show the slowest elimination reactions.

To evaluate the regio- and stereoselectivity, catalysts 3a–c and 3e were also applied in the aldol reactions of benzaldehyde (4) and 3-pentanone (8a) or butanone (8b) at 30 °C (Scheme 4). In this study 20 mol% of boron catalyst was used, and again very fast conversion of 4 was observed during the first few minutes of the reaction, after which the reaction slowed down dramatically. The maximum amount of β-hydroxyketone was obtained already after 10 min (Table 2).

Scheme 4:

The boron-catalyzed aldol reaction of benzaldehyde and 3-pentanone (8a) or butanone (8b).

Table 2

Conversion and selectivity for the boron catalyzed aldol reactions between benzaldehyde (2 mmol) and 3-pentanone (8a) or butanone (8b, 20 mmol) in the presence of 20 mol% of catalyst, at 30 °C after 10 min.^a

Entry	Catalyst	Conversion (%)	9 (%)	10 (%)	11 (%)	12 (%)
8a
1	NaOiPr	91	33	56	–^b	–^b
2	3a	92	25	41	–	–
3	3b	91	26	47	–	–
4	3c	94	29	57	–	–
5	3e	94	24	48	–	–
6	3e^c	26	10	15	–	–
8b
7	NaOiPr	93	10	16	25	41
8	3a	92	13	17	32	7
9	3b	75	12	12	36	–^b
10	3c	97	7	8	21	37
11	3e	95	7	10	22	37

^aAs determined by ¹H NMR spectroscopy; GC-MS suggested that the remainder consists of oligomerization products.

^bNot observed.

^cAfter 30 min at 0 °C.

For 3-pentanone (8a), the selectivity towards the rac-erythro- and rac-threo-β-hydroxyketones (9a, 10a) was high, and no dehydration was observed. In contrast, one of the aldols formed from the reaction of butanone with benzaldehyde, 11 dehydrated to give 12 (Scheme 4). Probably, steric hindrance by the methyl group prevents the elimination reaction of 9 and 10. With NaOiPr as the basic catalyst, the behavior was similar. These reactions appeared to be prone to oligomerization of the aldol product after 90 min, as observed by GCMS analysis. This oligomerization could be suppressed by decreasing the reaction temperature to 0 °C. However, under those conditions, the reaction stopped after 10 min with < 30 % conversion of benzaldehyde (Table 2, entry 6). At low temperatures, the aldol reaction products probably coordinate relatively strongly to the catalyst, thereby blocking it for further reaction. (The termination of the reaction at 0 °C after about 10 min cannot be due to hydrolysis of the catalyst by water formed during the dehydration reaction, since that would lead to precipitation of the catalyst. In all cases the reaction mixture remained homogeneous. Additionally, it had earlier been shown that water does not inhibit the reaction [19].)

All reactions with butanone and 3-pentanone showed no diastereoselectivity, neither with NaOiPr nor with 3a,b,c, and 3e as the catalyst (Table 2, entries 1–6 and 7–11). The ratio of 9 and 10 is about the same in all cases. Therefore, it is not very likely that catalysts 3 act as a template for a six-membered transition state in the reaction between the enolate of the ketone (5, 8a/b) and benzaldehyde (4) [24–26]. Compounds 3 rather are basic catalysts for the deprotonation of the ketone in this case.

When more rigid cyclohexanone (8c) was employed instead of 3-pentanone 8a, only traces of the aldol products were observed (Scheme 5). While 3a displayed sufficient activity to catalyze the conversion of acetone, 8a and 8b, the boron catalyst is too weak as a base to deprotonate the cyclic ketone 8c. This might be due to an increased ring strain when deprotonating 8c. Also no dehydration occurred and product 13 was not detected. NaOiPr as a much stronger base catalyzed the reaction; formation of an aldol product and subsequent dehydration took place (Table 3). As expected for NaOiPr, no diastereoselectivity was observed, and 9c and 10c were formed in equal amounts. Due to the higher activation energy for the deprotonation, this catalyst displayed slightly better selectivity and only 4 % of elimination product 13c was detected.

Scheme 5:

The boron catalyzed aldol reaction of benzaldehyde and cyclohexanone (8c).

Table 3

Conversion and selectivity for the boron catalyzed aldol reaction between benzaldehyde and cyclohexanone (8c) in the presence of 20 mol% of catalyst, at 30 °C after 10 min.

Entry	Catalyst	Conversion (%)	9c + 10c (%)	13 (%)
1^a	NaOiPr	72	34 + 34	4
2^b	3a	∼2	Traces	–

^a2 mmol of benzaldehyde in 20 mmol of cyclohexanone with 20 mol% of NaOiPr; conversion was determined by ¹H NMR spectroscopy.

^b1 mmol of benzaldehyde in 10 mmol of perdeuterated cyclohexanone with 20 mol% of 3a; the reaction was followed in situ by ¹H NMR spectroscopy.

3 Conclusion

Triisopropyl esters of fluoro-substituted phenylboronic acids are efficient basic catalysts for the aldol reactions of unhindered starting materials. They are highly stable in the substrate, which can also function as solvent. Manipulation of the charge density of the borate group through the fluoro-substituents can be used to optimize the selectivity towards the desired β-hydroxyketone. This effect is mainly due to suppression of the undesired subsequent elimination reaction of the β-hydroxyketone. The highest selectivity towards the aldol product was obtained with borates corresponding to fluoro-substituted phenylboronic acids with a pK_a > 7. As the catalyst acts as a mild base rather than as template, no diastereoselectivity was observed; also more demanding ketones were not converted due to this weak basicity.

4 Experimental section

4.1 Materials

Experiments were performed with commercially available chemicals. Dry toluene and dry 2-propanol were used to synthesize the catalysts. All phenylboronic esters were synthesized by following the procedure for diisopropyl 3,5-difluorophenylboronate or sodium triisopropoxy 3,5-difluorophenylborate (see below). Molecular sieves (3 Å) were purchased from Metrohm. Silica gel (Fluka, particle size 0.06–0.2 nm) was used for the isolation of the products. For the separation of the products, ethyl acetate/petroleum ether (gradient) was used as eluent. Benzaldehyde was freshly distilled. All chemicals were purchased from Sigma-Aldrich (Schnelldorf, Germany) unless otherwise stated.

4.2 Analytical techniques

¹H, ¹¹B, and ¹³C NMR spectra were recorded at 400, 128, and 100 MHz, respectively, with a Bruker Avance-400 spectrometer, and at 300 (¹H) or 75 MHz (¹³C) with a Varian Inova-300 spectrometer. The chemical shifts in ¹¹B NMR spectra were referenced to a 0.1 m boric acid solution in D₂O at 0 ppm. All other resonances in the ¹H and ¹³C spectra were referenced to the residual solvent peak and are reported in ppm with respect to TMS. Boron-substituted aromatic carbon nuclei were not observed in some cases due to severe line broadening. NMR data of aldol products were compared with literature to assign structures. Thin layer chromatography (TLC) was performed with silica gel 60 F254 (Merck) and analyzed at 254 and 365 nm. If needed, the product was visualized with iodine on the TLC plate. Mass spectrometry was performed with a Shimadzu QP-2010S GCMS spectrometer using GCMS solution software for data processing. Satisfactory elemental analyses or high-resolution mass spectra could not be obtained probably due to the boron and fluorine content of the compounds. Similar problems have been reported previously [27–29].

4.3 Synthesis of the catalyst

All catalysts were prepared according to the procedure described for 2b and 3b.

4.3.1 Diisopropyl 3,5-difluorophenylboronate (2b)

3,5-Difluorophenylboronic acid (4.74 g, 30 mmol) and 2-propanol (12 g, 15 mL, 200 mmol) were dissolved in 15 mL of toluene. The reaction mixture was stirred for 24 h under reflux under a Dean–Stark trap filled with molecular sieve 3 Å (11 g), 2-propanol (5 mL), and toluene (5 mL). After that the solvent was evaporated under reduced pressure, and the product was distilled under reduced pressure. Pure 2b (4.1 g) was obtained as a colorless liquid (16.9 mmol, 56 %). B.p. 50 °C/6.5 × 10^–2 mbar. – ¹H NMR (400 MHz, CDCl₃): δ = 1.25 (d, 12H, 4 × CH₃, J= 6.5 Hz), 4.58 (sept, 2H, 2 × CHOB, J= 6.5 Hz), 6.81 (m, 1H, H_Ar), 7.06 (m, 2H, 2 × H_Ar). – ¹¹B NMR (128 MHz, CDCl₃): δ = 7.6. ¹³C NMR (100 MHz, CDCl₃): δ = 24.6, 66.7, 104.7 (t, J= 25.2 Hz), 115.1 (m), 138.5 (C–B), 163.0 (dd, J₁= 249.7 Hz, J₂= 11.2 Hz).

4.3.2 Sodium triisopropoxy (3,5-difluorophenyl)borate (3b)

The reaction was carried out under a nitrogen atmosphere. Sodium (345 mg, 15 mmol) was dissolved in 60 mL of 2-propanol. The solution was stirred for 1 h under reflux and then cooled to 30 °C. Then a solution of 3.63 g (15 mmol) of diisopropyl 3,5-difluorophenylboronate (2b) in 8 mL of 2-propanol was added dropwise over 30 min. The mixture was stirred for 16 h at 30 °C, after which the solvent was removed under vacuum to give 4.78 g of white powder (14.8 mmol; yield: 98 %). – ¹H NMR (300 MHz, [D₈]2-propanol): δ = 1.11 (d, 18H, 6 × CH₃, J= 6 Hz), 3.9 (sept, 3H, CHOB, J= 6 Hz), 6.54–7.04 (m, 3H, 3 × H_Ar). – ¹¹B NMR (128 MHz, [D₈]2-propanol): δ = –15.5. – ¹³C NMR (75 MHz, [D₈]2-propanol): δ = 25.3, 63.6, 99.3 (m), 115.3 (m), 162.6 (dd, J= 10.3 Hz; J= 245.8 Hz), 162.9 (dd, J₁= 246.5 Hz; J₂= 10.4 Hz).

4.4 NMR data of the other phenylboronic esters

4.4.1 Diisopropyl 3,5-bis-(trifluoromethyl)phenylboronate (2a)

¹H NMR (400 MHz, CDCl₃): δ = 1.27 (d, 12H, 4 × CH₃, J= 6.1 Hz), 4.59 (sept, 2H, 2 × CHOB, J= 6.1 Hz), 7.89 (m, 1H, H_Ar), 8.00 (m, 2H, 2 × H_Ar). – ¹¹B NMR (128 MHz, CDCl₃): δ = 7.5. ¹³C NMR (100 MHz, CDCl₃): δ = 24.70, 67.10, 123.19 (quin, J = 3.8 Hz), 123.2 (q, J= 274.0 Hz), 130.90 (q, J= 33.0 Hz), 133.0, 137.3 (C–B).

4.4.2 Sodium triisopropoxy 3,5-bis-(trifluoromethyl)phenyl-borate (3a)

¹H NMR (300 MHz, CD₃OD): δ = 1.15 (d, 18H, 6 × CH₃, J= 6.2 Hz), 3.92 (sept, 3H, CHOB, J= 6.2 Hz), 7.56 (s, 1H, H_Ar), 8.03 (s, 2H, 2 × H_Ar). – ¹¹B NMR (128 MHz, [D₈]2-propanol): δ = –15.51. – ¹³C NMR (75 MHz, CD₃OD): δ = 25.3, 64.7, 119.3 (quin, J= 3.89 Hz), 124.3, 127.9, 130.0 (q, J= 117.68 Hz), 131.54, 134.1.

4.4.3 Diisopropyl 2,4-difluorophenylboronate (2c)

¹H NMR (300 MHz, CDCl₃): δ = 1.22 (d, 12H, 4 × CH₃, J= 6.0 Hz), 4.44 (sep, 2H, 2 × CHOB, J= 6.0 Hz), 6.70–6.80 (m, 1H, H_Ar), 6.84–6.94 (m, 1H, H_Ar), 7.30–7.42 (m, 1H, H_Ar). – ¹¹B NMR (128 MHz, CDCl₃): δ = 8.0. ¹³C NMR (75 MHz, CDCl₃): δ = 24.6, 67.2, 103.51 (dd, J₁= 12.2 Hz; J₂= 28.8 Hz), 111.5 (dd, J₁= 20.3, Hz; J₂= 3.2 Hz), 134.9 (dd, J₁= 7.6 Hz; J₂= 9.5 Hz), 162.9 (dd, J₁= 67.34 Hz; J₂= 11. 7 Hz), 166.2 (dd, J₁= 11.6 Hz; J₂= 11.5 Hz).

4.4.4 Sodium triisopropoxy (2,4-difluorophenyl)borate (3c)

¹H NMR (300 MHz, CD₃OD): δ = 1.15 (d, 18H, 6 × CH₃, J= 6.2 Hz), 3.93 (sept, 3H, CHOB, J= 6.2 Hz), 6.51–6.59 (m, 1H, H_Ar), 6.65–6.72 (m, 1H, H_Ar), 7.44–7.51 (m, 1H, H_Ar). – ¹¹B NMR (128 MHz, [D₈]2-propanol): δ = –15.8. – ¹³C NMR (75 MHz, [D₄]MeOH): δ = 25.3, 64.7, 102.5 (dd, J₁= 16.9 Hz; J₂= 33.0 Hz), 110.0 (dd, J₁= 18.3 Hz; J= 2.9 Hz), 131.7 (C–B), 138.0 (dd, J₁= 15.8 Hz; J₂= 8.6 Hz), 163.1 (dd, J₁= 241.5, Hz; J₂= 13.7 Hz), 165.8 (dd, J₁= 238.6 Hz; J₂= 10.5 Hz).

4.4.5 Diisopropyl 3-fluorophenylboronate (2d)

¹H NMR (300 MHz, CDCl₃): δ = 1.29 (d, 12H, 4 × CH₃, J= 6.1 Hz), 4.64 (sept, 2H, 2 × CHOB, J= 6.1 Hz), 7.06–7.14 (m, 1H, H_Ar), 7.29–7.38 (m, 3H, H_Ar). ¹¹B NMR (128 MHz, CDCl₃): δ = 7.3. – ¹³C NMR (75 MHz, CDCl₃): δ = 24.7, 66.5; 116.2 (d, J= 21.0 Hz), 119.5 (d, J= 19.1 Hz), 128.4 (d, J= 3.0 Hz), 129.6 (d, J= 7.2 Hz), 136.4 (C–B), 162.7 (d, J= 244.9 Hz).

4.4.6 Sodium triisopropoxy (3-fluorophenyl)borate (3d)

¹H NMR (300 MHz, CD₃OD): δ = 1.15 (d, 18H, 6 × CH₃, J= 6.2 Hz), 3.93 (sept, 3H, CHOB, J= 6.2 Hz), 6.67–6.73 (m, 1H, H_Ar), 7.08–7.19 (m, 2H, H_Ar), 7.27 (d, 1H, H_Ar,J= 7.1 Hz). – ¹¹B NMR (128 MHz, [D₈]2-propanol): δ = –15.1.

¹³C NMR (75 MHz, CD₃OD): δ = 25.3, 64.7; 112.1 (d, J= 21.2 Hz), 119.8 (d, J= 16.2 Hz), 128.6 (d, J= 7.5 Hz), 129.6 (d, J= 2.3 Hz), 155.9 (C–B), 164.1 (d, J= 240.7 Hz).

4.4.7 Diisopropyl 2-fluorophenylboronate (2e)

¹H NMR (400 MHz, CDCl₃): δ = 1.22 (d, 12H, 4 × CH₃, J= 6.1 Hz), 4.44 (sept, 2H, 2 × CHOB, J= 6.1 Hz), 7.01 (t, 1H, H_Ar,J= 7.3 Hz), 7.14 (t, 1H, H_Ar,J= 8.5 Hz), 7.35 (m, 2H, 2 × H_Ar). – ¹¹B NMR (128 MHz, CDCl₃): δ = 8.3. ¹³C NMR (100 MHz, CDCl₃): δ = 24.6, 67.1, 114.9 (d, J= 24.1 Hz), 121.9 (C–B), 124.6 (d, J= 2.8 Hz), 130.8 (d, J= 8.2 Hz), 133.7 (d, J= 10.0 Hz), 164.4 (d, J= 239.7 Hz).

4.4.8 Sodium triisopropoxy (2-fluorophenyl)boronate (3e)

¹H NMR (400 MHz, [D₈]2-propanol): δ = 1.16 (d, 18H, 6 × CH₃, J= 6.2 Hz), 3.94 (sept, 3H, CHOB, J= 6.2 Hz), 6.62–6.78 (m, 1H, H_Ar), 6.88–7.06 (m, 2H, H_Ar); 7.61–7.82 (m, 1H, H_Ar).

¹¹B NMR (128 MHz, [D₈]2-propanol): δ = –15.7. – ¹³C NMR (100 MHz, CDCl₃): δ = 25.2, 64.7; 112.0 (d, J= 28.3 Hz), 119.77 (d, J= 21.6 Hz), 128.6 (d, J= 9.0 Hz), 129.6 (d, J= 3.1 Hz), 155.9, 164.1 (d, J= 320.9 Hz).

4.4.9 Diisopropyl 4-fluorophenylboronate (2f)

¹H NMR (300 MHz, CDCl₃): δ = 1.25 (d, 12H, 4 × CH₃, J= 6.1 Hz), 4.60 (sept, 2H, 2 × CHOB, J= 6.1 Hz), 7.05 (t, 2H, H_Ar,J= 9.1 Hz), 7.58 (t, 2H, H_Ar,J= 7.29. Hz). – ¹¹B NMR (128 MHz, CDCl₃): δ = 8.2. ¹³C NMR (75 MHz, CDCl₃): δ = 24.8, 66.4, 114.9 (d, J= 19.8 Hz), 129.3 (C–B), 135.2 (d, J= 7.5 Hz), 163.9 (d, J= 246.6 Hz).

4.4.10 Sodium triisopropoxy (4-fluorophenyl)borate (3f)

¹H NMR (400 MHz, CD₃OD): δ = 1.15 (d, 18H, 6 × CH₃, J= 6.2 Hz), 3.92 (sept, 3H, CHOB, J= 6.2 Hz), 7.05 (m, 2H, H_Ar,J= 9.0 Hz), 7.46 (dd, 2H, H_Ar,J₁ = 6.15, J₂= 8.5 Hz). – ¹¹B NMR (128 MHz, [D₈]2-propanol): δ = –14.9. – ¹³C NMR (100 MHz, MeOD): δ = 25.3, 64.7, 113.5 (d, J= 18.4 Hz), 135.5 (d, J= 6.3 Hz), 162.9 (d, J= 237.7 Hz).

4.4.11 Diisopropyl 2,6-difluorophenylboronate

¹H NMR (300 MHz, CDCl₃): δ = 1.21 (d, 12H, 4 × CH₃, J= 6.0 Hz), 4.38 (sept, 2H, 2 × CHOB, J= 6.0 Hz), 6.82 (m, 2H, 2 × H_Ar), 7.28 (m, 1H, H_Ar). – ¹¹B NMR (128 MHz, CDCl₃): δ = 5.0. – ¹³C NMR (75 MHz, CDCl₃): δ = 24.6, 67.7, 111.0 (dd, J₁= 25.7 Hz; J₂= 1.7 Hz), 131.2 (t, J= 9.8 Hz), 164.5 (dd, J₁= 243.2 Hz; J= 14.7 Hz).

4.4.12 Diisopropyl 2,4,6-trifluorophenylboronate

¹H NMR (300 MHz, CDCl₃): δ = 1.21 (d, 12H, 4 × CH₃, J= 6.2 Hz), 4.38 (sept, 2H, 2 × CHOB, J= 6.2 Hz), 6.60 (dd, 2H, H_Ar, J= 6.8 Hz, J= 9.0 Hz). – ¹¹B NMR (96 MHz, CDCl₃): δ = 7.0. ¹³C NMR (75 MHz, CDCl₃): δ = 24.5, 67.8, 100.0 (ddd, J= 3.2 Hz; J= 25.1 Hz; J= 28.6 Hz), 106.6 (C–B), 163.9 (td, J₁= 249.6 Hz, J₂= 15.7 Hz), 164.7 (ddd, J₁= 243.3 Hz; J₂= 18.0 Hz; J₃= 16.9 Hz).

4.5 General procedure for the aldol reaction (with acetone)

Aldol reactions were performed in an NMR tube with 5 mm diameter at 30 °C. The reaction was monitored by ¹H NMR spectroscopy. The amount of sodium borate (3a–f) desired for the particular experiment, 1 mmol of benzaldehyde, and 68 mg (0.4 mmol) of 1,3,5-trimethoxy benzene, which was used as an internal standard, were dissolved in 640 mg (10 mmol) of [D₆]acetone.

4.6 General procedure for the aldol reaction (with butanone, 3-pentanone)

The aldol reactions were performed in a 10 mL one-necked flask (dried in the oven overnight, flushed with nitrogen). Benzaldehyde (212 mg, 2 mmol, 1 eq) and 1,3,5-trimethoxybenzene (100 mg, 0.6 mmol, 0.3 eq, internal standard for quantitative measurements by integration of signals) were dissolved in butanone or 3-pentanone (20 mmol, 10 eq). An aliquot of 25 μL was taken from the reaction mixture to prepare the NMR samples in CDCl₃. A catalyst (borate salt, 0.4 mmol, 20 mol%) was added to the reaction mixture and the conversion was followed by ¹H NMR spectroscopy at 30 °C.

4.7 General procedure for the aldol reaction (with cyclohexanone)

With sodium isopropoxide:

The aldol reaction was performed in a 10 mL one-necked flask (dried in the oven overnight, flushed with nitrogen). Benzaldehyde (212 mg, 2 mmol, 1 eq) and 1,3,5-trimethoxybenzene (67 mg, 0.4 mmol, 0.2 eq, internal standard for quantitative measurements by integration of signals) were dissolved in cyclohexanone (20 mmol, 10 eq). A 25 μL sample was taken from the reaction mixture to prepare the ¹H NMR samples in CDCl₃. NaOiPr (33 mg, 0.4 mmol, 20 mol%) was added to the reaction mixture and the conversion was followed by ¹H NMR spectroscopy at 30 °C. The products were compared with literature known compounds [30].

With sodium borate 3a:

The aldol reaction was performed in a 5 mm ø NMR tube at 30 °C. The reaction was monitored by ¹H NMR spectroscopy. Eighty-five milligrams (0.2 mmol) of sodium borate (3a), 106 mg (1 mmol) of benzaldehyde, and 68 mg (0.4 mmol) of 1,3,5-trimethoxybenzene, which was used as an internal standard, were dissolved in 1.08 g (10 mmol) of [D₁₀]cyclohexanone.

4.8 Isolation of the aldol product 6

Borate salt (0.4 mmol 3b) and 2 mmol of benzaldehyde were dissolved in 1.16 g (20 mmol) of acetone. The reaction mixture was stirred for 20 min at 30 °C. The excess of acetone was evaporated and the residue was dissolved in 3 mL of water. The product was extracted with 3 × 5 mL of ethyl acetate. The organic layers were combined, dried with MgSO₄, and filtered, and the solvent was evaporated. The crude product was purified by column chromatography to yield 34 % of 4-hydroxy-4-phenylbutan-2-one (6). – ¹H NMR (400 MHz, CDCl₃): δ = 2.18 (s, 3H, CH₃), 2.77–2.92 (m, 2H, CH₂), 3.30 (s, 1H, OH), 5.16 (m, 1H, CHCOH), 7.28–7.36 (m, 5H, H_Ar). – ¹³C NMR (100 MHz, CDCl₃): δ = 30.8, 52.1, 69.9, 125.7, 127.7, 128.6, 142.9, 209.1.

4.9 Isolation of the aldol products 9a and 10a

Boronate salt (0.2 mmol 3b) and 212 mg (2 mmol) of benzaldehyde were dissolved in 1.72 g (20 mmol) of 3-pentanone (8a). The reaction mixture was stirred for 20 min at 30 °C. The excess of 3-pentanone was evaporated and the residue was dissolved in 3 mL of water. The product was extracted with 3 × 5 mL of ethyl acetate. The organic layers were combined, dried over MgSO₄, and filtered, and the solvent was evaporated to give 489 mg of crude product, which was purified by column chromatography using a column of 10 g silica gel (0.060–0.200 mm, pore diameter: 6 nm) The column was eluted subsequently with 9:1 petroleum ether-ethyl acetate 9:1, 8:2, and 7:3. The fractions containing the aldol products were combined and the solvents were evaporated to give 50 mg of colorless oil consisting of a syn-1-hydroxy-2-methyl-1-phenylpentan-3-one (0.26 mmol, 13 % yield) and anti-1-hydroxy-2-methyl-1-phenylpentan-3-one (0.52 mmol, 26 % yield).

4.9.1 Syn-1-hydroxy-2-methyl-1-phenyl-pentan-3-one (9a)

¹H NMR (300 MHz, CDCl₃): δ=0.97 (t, J = 7.2 Hz, 3H, CH₃), 1.01 (d, J = 7.3 Hz, 3H, CH₃), 2.28–2.52 (m, 2H, CH₂), 2.93 (dq, J₁ = 7.2 Hz, J₂ = 4.2 Hz, 1H, CHCH₃), 5.05 (J = 4.2 Hz, 1H, CHOH), 7.27–7.37 (m, 5H, H_Ar). – ¹³C NMR (100 MHz, CDCl₃): δ = 7.61, 10.7, 35.5, 52.4, 73.4, 126.0, 127.5, 128.0, 142.0, 216.4.

4.9.2 Anti-1-hydroxy-2-methyl-1-phenyl-pentan- 3-one (10a)

¹H NMR (300 MHz, CDCl₃): δ=0.93 (d, J = 7.2 Hz, 3H, CH₃), 1.01 (t, J = 7.3 Hz, 3H, CH₃), 2.28–2.52 (m, 2H, CH₂), 2.93 (dq, J₁ = 7.2 Hz, J₂ = 8.3 Hz, 1H, CHCH₃), 4.75 (J = 8.2 Hz, 1H, CHOH), 7.27–7.37 (m, 5H, H_Ar). – ¹³C NMR (100 MHz, CDCl₃): δ = 7.55, 14.6, 36.6, 52.8, 76.8, 126.1, 128.4, 128.6, 142.3, 216.2.

4.10 Isolation of the aldol products 9b, 10b, and 11

Boronate salt (0.2 mmol) and 212 mg (2 mmol) of benzaldehyde were dissolved in 1.44 g (20 mmol) of butanone (8b). The reaction mixture was stirred for 20 min at 30 °C. Butanone was evaporated and the residue was dissolved in 3 mL of water. The product was extracted with 3 × 5 mL of ethyl acetate. The organic layers were combined, dried with MgSO₄, and filtered, and the solvent was evaporated to give 320 mg of the crude product. For purification by column chromatography, 20 g of silica gel (0.060–0.200 mm, pore diameter: 6 nm) was used, along with an 8:2 petroleum ether:ethyl acetate eluent. From two combined fractions containing the aldol products, the solvents were evaporated. The residues were two colorless oils, one of these being anti-4-hydroxy-3-methyl-4-phenylbutan-2-one in a yield of 28.6 mg (0.16 mmol, 8.0 %) and the other being a mixture of 1-hydroxy-1-phenylpentan-3-one (34.7 mg, 0.20 mmol, 9.7 %) and syn-4-hydroxy-3-methyl-4-phenylbutan-2-one (11.6 mg, 0.07 mmol, 3.2 %).

4.10.1 Syn-4-hydroxy-3-methyl-4-phenylbutan-2-one (9b)

¹H NMR (400 MHz, CDCl₃): δ = 1.02–1.06 (d, J = 7.4 Hz, 3H, CH₃), 2.14 (s, 3H, CH₃), 2.82 (dd, J₁ = 3.8 Hz, J₂ = 7.4 Hz, 1H, CH), 3.08 (s, br, 1H, OH), 5.12 (d, J = 4.2 Hz, 1H, CHOH), 7.30–7.36 (m, 5H, H_Ar). – ¹³C NMR (100 MHz, CDCl₃): δ = 10.2, 29.4, 53.2, 73.0, 125.6, 125.9, 128.5, 141.8, 213.5.

4.10.2 Anti-4-hydroxy-3-methyl-4-phenylbutan- 2-one (10b)

¹H NMR (400 MHz, CDCl₃): δ = 0.93 (d, 3H, CH₃), 2.21 (s, 3H, CH₃), 2.84 (s, br, 1H, OH), 2.92 (dd, 1H, CHCH₃), 4.74 (d, 1H, CHOH), 7.31–7.35 (m, 5H, H_Ar). – ¹³C NMR (100 MHz, CDCl₃): δ = 14.1, 30.0, 53.7, 76.5, 126.6, 128.0, 128.5, 141.9, 213.4.

4.10.3 1-Hydroxy-1-phenylpentan-3-one (11)

¹H NMR (400 MHz, CDCl₃): δ = 1.06 (t, J = 7.4 Hz, 3H, CH₃), 2.45 (q, J = 7.3 Hz, 2H, CH₂), 2.82 (dd, J₁ = 3.7 Hz, J₂ = 8.7 Hz, 2H, CH₂), 3.36 (s, br, 1H, OH), 5.16 (dd, J₁ = 8.8 Hz, J₂ = 4.0 Hz, 1H, CHOH), 7.30–7.36 (m, 5H, H_Ar). – ¹³C NMR (100 MHz, CDCl₃): δ = 7.50, 36.8, 50.7, 70.0, 125.9, 128.5, 142.9, 211.9.

Corresponding author: Ulf Hanefeld, Gebouw voor Scheikunde, Afdeling Biotechnologie, Technische Universiteit Delft, Julianalaan 136, 2628 BL Delft, The Netherlands, Fax: +31-15-278-1415, E-mail: u.hanefeld@tudelft.nl

Acknowledgments

This research has been performed within the framework of the CatchBio program. The authors gratefully acknowledge the support of the Smart Mix Program of the Dutch Ministry of Economic Affairs and the Dutch Ministry of Education, Culture and Science. The authors are thankful to L. Panella (DSM), P. Alsters (DSM), J. G. de Vries (DSM), B. Kaptein (DSM), and G. Kemperman (MSD) for fruitful discussions. The authors are also thankful to Wuyuan Zhang (TU Delft – NMR) and Guzman Torrelo Villa (TU Delft – Chemicals) for their support.

References

[1] J. Podlech, Angew. Chem. Int. Ed. 2010, 49, 6490.Search in Google Scholar

[2] P. S. Gradeff, U. S. Patent 3840601, 1974.Search in Google Scholar

[3] P. W. D. Mitchell, U.S. Patent 4874900, 1989.Search in Google Scholar

[4] K. Weissermel, H. J. Arpe, Industrial Organic Chemistry, 4th ed., Wiley-VCH, Weinheim, 2003.10.1002/9783527619191Search in Google Scholar

[5] Z. Zhang, Y. W. Dong, G. W. Wang, Chem. Lett.2003, 32, 966.10.1246/cl.2003.966Search in Google Scholar

[6] C. Hertweck, Angew. Chem. Int. Ed.2009, 48, 4688.10.1002/anie.200806121Search in Google Scholar

[7] P. Clapes, W.-D. Fessner, G. A. Sprenger, A. K. Samland, Curr. Opin. Chem. Biol.2010, 14, 154.10.1016/j.cbpa.2009.11.029Search in Google Scholar

[8] R. Mestres, Green Chem. 2004, 6, 583.10.1039/b409143bSearch in Google Scholar

[9] S. K. Karmee, U. Hanefeld, ChemSusChem2011, 4, 1118.10.1002/cssc.201100083Search in Google Scholar

[10] A. M. Frey, S. K. Karmee, K. P. de Jong, J. H. Bitter, U. Hanefeld, ChemCatChem2013, 5, 594.10.1002/cctc.201200282Search in Google Scholar

[11] H. Zheng, D. G. Hall, Aldrichimica Acta.2014, 47, 41.Search in Google Scholar

[12] T. Mukaiyama, M. Hayashi, Chem. Lett.1974, 3, 15.10.1246/cl.1974.15Search in Google Scholar

[13] T. Mukaiyama, T. Izawa, K. Saigo, Chem. Lett. 1974, 3, 323.10.1246/cl.1974.323Search in Google Scholar

[14] T. Mukaiyama, K. Narasaka, K. Banno, Chem. Lett. 1973, 2, 1011.10.1246/cl.1973.1011Search in Google Scholar

[15] Y. Mori, J. Kobayashi, K. Manabe, S. Kobayashi, Tetrahedron2002, 58, 8263.10.1016/S0040-4020(02)00976-6Search in Google Scholar

[16] D. A. Evans, J. Bartroli, T. L. Shih, J. Am. Chem. Soc. 1981, 103, 2127.10.1021/ja00398a058Search in Google Scholar

[17] D. A. Evans, J. V. Nelson, E. Vogel, T. R. Taber, J. Am. Chem. Soc. 1981, 103, 3099.10.1021/ja00401a031Search in Google Scholar

[18] K. Aelvoet, A. S. Batsanov, A. J. Blatch, C. Grosjean, L. G. F. Patrick, C. A. Smethurst, A. Whiting, Angew. Chem. Int. Ed. 2008, 47, 768.10.1002/anie.200704293Search in Google Scholar

[19] T. Müller, K. Djanashvili, I. W. C. E. Arends, J. A. Peters, U. Hanefeld, Chem. Commun. 2013, 49, 361.10.1039/C2CC37047FSearch in Google Scholar

[20] S. W. J. Benson, Chem. Educ.1965, 42, 502.10.1021/ed042p502Search in Google Scholar

[21] T. L. Cottrell, The Strengths of Chemical Bonds, 2nd ed., Butterworth, London, 1958.Search in Google Scholar

[22] J. Yan, G. Springsteen, S. Deeter, B. Wang, Tetrahedron2004, 60, 11205.10.1016/j.tet.2004.08.051Search in Google Scholar

[23] Y. Yamamoto, T. Matsumura, N. Takao, H. Yamagishi, M. Takahashi, S. Iwatsuki, K. Ishihara, Inorg. Chim. Acta2005, 358, 3355.10.1016/j.ica.2005.05.026Search in Google Scholar

[24] D. A. Evans, E. Vogel, J. V. Nelson, J. Am. Chem. Soc. 1979, 101, 6120.10.1021/ja00514a045Search in Google Scholar

[25] T. Nitabaru, N. Kumagai, M. Shibasaki, Tetrahedron Lett. 2008, 49, 272.10.1016/j.tetlet.2007.11.055Search in Google Scholar

[26] T. Nitabaru, A. Nojiri, M. Kobayashi, N. Kumagai, M. Shibasaki, J. Am. Chem. Soc. 2009, 131, 13860.10.1021/ja905885zSearch in Google Scholar

[27] V. P. Fadeeva, V. D. Tikhova, O. N. Nikulicheva, J. Anal. Chem. 2008, 63, 1094.10.1134/S1061934808110142Search in Google Scholar

[28] R. Perry, Organo-fluorine compounds, in Houben-Weyl, Methods of Organic Chemistry, Vol. E 10a (Eds.: K. H. Büchel, J. Falbe, H. Hagemann, M. Hanack, D. Klamann, R. Kreher, H. Kropf, M. Regitz, E. Schaumann), Thieme, Stuttgart, 1999, chapter 4a, p. 27.Search in Google Scholar

[29] K. K. Shen in Non-Halogenated Flame Retardant Handbook (Eds.: A. B. Morgan, C. A. Wilkie), Wiley, Hoboken, 2014, chapter 6, p. 201.10.1002/9781118939239.ch6Search in Google Scholar

[30] M. Barbero, S. Bazzi, S. Cadamuro, S Dughera, C. Magistris, A. Smarra, P. Venturello, Org. Biomol. Chem.2011, 9, 2192.10.1039/c0ob00837kSearch in Google Scholar

Received: 2015-2-11

Accepted: 2015-4-2

Published Online: 2015-6-24

Published in Print: 2015-8-1

Articles in the same Issue

https://doi.org/10.1515/znb-2015-0029

Keywords for this article

aldol condensation; aldol reaction; boron; catalysis; dehydration

Entry	Catalyst	Amount of catalyst (mol%)	Conversion (%)	Aldol 6 (%)^b
1	3a	10	93	87
2	3b	10	90	80
3	3c	10	94	73
4	3d	10	93	91
5	3e	10	95	62
6	3f	10	94	77
7	3a	20	96	62
8	3b	20	95	77
9	3c	20	95	72
10	3d	20	96	64
11	3e	20	97	60
12	3f	20	96	56
13	3f	5	42	40
14	3f	2	25	23

Entry	Catalyst	Amount of catalyst (mol%)	Conversion (%)	Aldol 6 (%)^b
1	3a	10	93	87
2	3b	10	90	80
3	3c	10	94	73
4	3d	10	93	91
5	3e	10	95	62
6	3f	10	94	77
7	3a	20	96	62
8	3b	20	95	77
9	3c	20	95	72
10	3d	20	96	64
11	3e	20	97	60
12	3f	20	96	56
13	3f	5	42	40
14	3f	2	25	23

Entry	Catalyst	Amount of catalyst (mol%)	Conversion (%)	Aldol 6 (%)^b
1	3a	10	93	87
2	3b	10	90	80
3	3c	10	94	73
4	3d	10	93	91
5	3e	10	95	62
6	3f	10	94	77
7	3a	20	96	62
8	3b	20	95	77
9	3c	20	95	72
10	3d	20	96	64
11	3e	20	97	60
12	3f	20	96	56
13	3f	5	42	40
14	3f	2	25	23