Regioselective alkylation of 2,4-dihydroxybenzyaldehydes and 2,4-dihydroxyacetophenones

Abstract

We report a cesium bicarbonate-mediated alkylation of 2,4-dihydroxybenzyaldehyde and 2,4-dihydroxyacetophenone to generate 4-alkylated products in acetonitrile at 80 °C with excellent regioselectivity, up to 95% isolated yields, and broad substrate scope.

Graphical Abstract

Introduction

4-Alkoxy-2-hydroxybenzaldehydes and 4-alkoxy-2-hydroxyacetophenones represent privileged building blocks that have found applications in a broad range of fields. As key fragments and synthetic precursors of small molecule therapeutics, 4-alkoxy-2-hydroxybenzaldehydes and 4-alkoxy-2-hydroxyacetophenones have been used in numerous chemical scaffolds with promising therapeutic effects for the treatment of human diseases including different types of cancers,^[1] bacterial^[2] and fungal infections,^[3] type 2 diabetes,^[4] and Alzheimer’s Disease^[5]. Additionally, 4-alkoxy-2-hydroxybenzaldehydes and 4-alkoxy-2-hydroxyacetophenones have been used in the construction of compounds that can be used as fluorescent probes and imaging agents,^[6] selective labeling agents of protein^[7] and DNA,^{[1d, 8]} antioxidants,^[9] ligands for metal-based catalysts for organic transformations,^[10] and linkers for solid-phase peptide synthesis.^[11]

Available syntheses of 4-alkoxy-2-hydroxybenzaldehydes and 4-alkoxy-2-hydroxyacetophenones often involve a direct alkylation of 2,4-dihydroxybenzyaldehyde or 2,4-dihydroxyacetophenone using alkyl halides as an electrophile in the presence of an inorganic base such as Na₂CO₃, NaHCO_3, K₂CO₃, and KHCO₃ under vigorous heating in polar aprotic solvents including DMF, DMSO, acetone, THF, and CH₃CN.^[1–12] Although these conditions lead to the formation of the desired products, they usually suffer from limitations from low isolated yields to extended reaction times. Importantly, upon prolonged heating, the reactions commonly result in a complicated mixture of the target product contaminated by side products resulting from double alkylation and compound decomposition, which can create additional challenges in the purification step, especially for large reactions with multi-gram scales.

During our ongoing efforts in the development of novel antibiotics for infections by the gram-negative opportunistic pathogen Pseudomonas aeruginosa, we aimed to synthesize gallium salophen-based complexes as inhibitors for the P. aeruginosa extracellular hemophore HasAp, via a route employing 4-alkoxy-2-hydroxybenzaldehydes and 4-alkoxy-2-hydroxyacetophenones as key building blocks.^[13] Here we report our discovery of a practical CsHCO₃-mediated alkylation of 2,4-dihydroxybenzyaldehydes and 2,4-dihydroxyacetophenones that efficiently provide 4-alkylated products with excellent regioselectivity, good isolated yields, and a broad substrate scope.

Results and discussion

We began our studies with a solvent screen using a model reaction employing 2,4-dihydroxybenzyaldehyde (1a) as the starting material, KHCO₃ as the base, and 1,2-dibromoethane as the alkylating agent (Table 1). The reactions in either DMSO (entry 1) or DMF (entry 2) indicated complete consumption of starting material after 4 h. However, these reactions produced complicated mixtures containing multiple side products in addition to the desired product 2a^{[6a, 12i]} and the bis-alkylation product 2a’. Moreover, the reactions in MeOH, acetone, THF, and DCE led to minimum formation of product 2a after heating at 60 °C for 4 h (entries 3–6). On the other hand, the reaction in CH₃CN, although produced product 2a in low yields, indicated a clean conversion without the formation of any side products (entry 7). It was also encouraging to see that with increased temperature, a significant improvement of yields was observed (entries 8–9). Thus, we decided to keep our further optimization in CH₃CN at 80°C.

Table 1.

Solvent screen for the formation of compounds 2a and 2a’ from starting material 1a^a

entry	solvent	temp. (°C)	yield of 2ab (%)	yield of 2a’b (%)
1	DMSO	60	37	6
2	DMF	60	35	5
3	MeOH	60	< 3	NDd
4	acetone	60	< 3	NDd
5	THF	60	< 3	NDd
6	DCE	60	-c	NDd
7	CH3CN	60	9	NDd
8	CH3CN	70	20	NDd
9	CH3CN	80	40	NDd

^aA mixture of substrate 1a (0.5 mmol), 1,2-dibromoethane (3.0 equiv) in solvent (5.0 mL) was stirred at the given temperature for 4 h. DMSO = dimethyl sulfoxide, DMF = dimethylformamide, MeOH = methanol, THF = tetrahydrofuran, DCE = 1,2-dichloroethane, CH₃CN = acetonitrile.

^bNMR yields.

^cThe starting material 1a was not soluble, and the corresponding yield was not determined.

^dND = not detected

Next, the effects of different bases used in the reaction was investigated (Table 2). When the potassium ion of KHCO₃ was replaced by a harder sodium ion, the reaction gave no significant formation of product 2a (entry 1). Interestingly, when a softer cesium ion was used to replace the potassium ion of KHCO₃, the reaction indicated a clean conversion of the starting material 1a to the product 2a with a yield of over 60% (entry 2). We reasoned that the significantly improved yield of the CsHCO₃ reaction was due to the higher solubility and decreased coordination interaction of the softer cesium ion to the phenoxide anion that could favor ligand exchange.^[14] The more basic carbonate salts were also evaluated (entries 3–5). Reactions using Na₂CO₃ and Cs₂CO₃ gave low yields and while the one employing K₂CO₃ provided modest yields. However, the usage of all three carbonate bases led to the formation of a significant amount of bis-alkylated side product 2a’, through deprotonation of both hydroxy groups of compound 1a. Furthermore, organic bases TEA (entry 6) and DIPEA (entry 7) were also tested, and the results indicated that both tertiary amine-mediated reactions led to low yields.

Table 2.

The effects of bases used in the formation of compound 2a from 1a^a

entry	base	yield of 2ab (%)	yield of 2a’b (%)
1	NaHCO3	< 3	NDc
2	CsHCO3	61	4
3	Na2CO3	7	NDc
4	K2CO3	53	14
5	Cs2CO3	15	16
6	TEA	11	NDc
7	DIPEA	26	NDc

^aA mixture of substrate 1a (0.5 mmol), 1,2-dibromoethane (3.0 equiv) in CH₃CN (5.0 mL) was stirred at 80 °C for 4 h. NaHCO₃ = sodium bicarbonate, Na₂CO₃ = sodium carbonate, KHCO₃ = potassium bicarbonate, K₂CO₃ = potassium carbonate, Cs₂CO₃ = cesium carbonate, CsHCO₃ = cesium bicarbonate, TEA = triethylamine, DIPEA = N,N-diisopropylethylamine.

^bNMR yields.

^cND = not detected

We then assessed the reagent ratio and reaction time (Table 3). Increasing the equivalence of CsHCO₃ from 1.5 equiv to 3.0 equiv led to a significant increase of the yield (entry 1) although further increasement of CsHCO₃ indicated no additional benefit to the reaction outcome (entry 2). The effects of increasing equivalence of 1,2-dibromoethane were also studied (entries 3–4), and the results showed that extra equivalence of the electrophile had no obvious contribution to the yields. Extension of the reaction time to 6–8 h also indicated minimum change in reaction yields (entries 5–6). Interestingly, prolonged reaction time was detrimental to the yield of the reaction (entries 7–8).

Table 3.

Impact of reagent equivalence and reaction time on the formation of 2a from 1a^a

entry	dibromoethane/equiv	CsHCO3/equiv	time	yield of 2ab (%)
1	3.0	3.0	4 h	83
2	3.0	4.5	4 h	86
3	6.0	3.0	4 h	81
4	9.0	3.0	4 h	87
5	3.0	3.0	6 h	78
6	3.0	3.0	8 h	81
7	3.0	3.0	24 h	55
8	3.0	3.0	48 h	40

^aA mixture of substrate 1a (0.5 mmol), 1,2-dibromoethane, and CsHCO₃ in 5.0 mL CH₃CN was stirred at 80 °C for the given reaction time.

^bNMR yields.

With the optimal conditions in hands, we next explored the substrate scope of alkyl bromides in the reaction on a 5 mmol scale (Scheme 1).^[15] Under the optimized conditions, the product of the model reaction, 2a, was isolated in 73% yield. Alkylations using dibromoalkanes with different linkers led to the generation of compounds 3a,^{[1b, 1d, 2b, 6b–d, 10a, 10b]} and 4a with high regioselectivity in similar modest yields. Using 1-bromopropane or 1-bromobutane, the corresponding alkylation reactions went smoothly to give compounds 5a^{[1a, 1g]} and 6a^{[12a, 12b, 12d, 12f, 12j]} in good yields. In addition, substrates containing an ester group provided compound 7a^{[2a, 7a, 7b, 11]} in good yields, while the substrate containing a Boc-protected amino group led to compound 8a with a low isolated yield. This low yield is likely due to the decomposition of the tert-butyl-carbamate group under heating and the cyclization of the resulting 2-bromoethylamine. Alkyl bromides containing either an ether or an alcohol groups were well-tolerated to generate the desired product 9a,^[12k] 10a,^{[4a, 12k]} and 11a^[2a] in good yields. Moreover, the alkyne- and alkene-containing substrates generated compounds 12a^{[1e, 2c, 4b, 6e, 6f, 10c, 12l]} and 13a in good yields. Furthermore, alkylation reactions involving substrates including phenyl or cyano groups led to the generation of compounds 14a, 15a, and 16a^[2a] in good yields. Finally, the reaction using a secondary alkyl bromide was studied. Although the reaction proceeded slower than the ones employing primary alkyl bromides, it produced the target compound 17a in high yields after 24 h with excellent regioselectivity.

Scheme 1. Substrate scope for the alkylation of 2,4-dihydroxybenzyaldehyde^[15].

.

The substate scope of the alkylation of 2,4-dihydroxyacetophenone was also studied (Scheme 2).^[15] Using 1,2-dibromoethane, 1,3-dibromopropane, or 1,4-dibromobutane as the substrate, the corresponding alkylation of 2,4-dihydroxyacetophenone went smoothly with high regioselectivity to generate compounds 2b,^{[1h, 1i]} 3b,^{[5a–d, 6d, 7c]} and 4b^{[1j, 1k, 3b, 5]} in good yields. Although the alkylations of 2,4-dihydroxyacetophenone took longer time to complete compared to those of 2,4-dihydroxybenzyaldehyde, no obvious formation of either the 2,4-dialkyl or double-alkylated side products was detected. Alkylation using 1-bromopropane and 1-bromobutane gave compounds 5b^{[1f, 1g]} and 6b^[3a] in excellent yields. The reactions employing alkyl bromides containing an ester functionality indicated clean conversion of the starting material to provide compound 7b in high yields. Similar to the alkylation of 2,4-dihydroxybenzyaldehyde, the Boc-protected amino-containing substate went slower and gave product 8b in low yield. Moreover, the reactions utilizing ether- or alcohol-containing substrates indicated clean conversion to give 4-selective alkylated products 9b, 10b, and 11b^[9] in good yields. The alkyne and alkene groups were also well-tolerated in the reactions to provide compounds 12b^{[2e, 12c, 12e, 12g]} and 13b in good yields. Furthermore, the alkyl bromides containing phenyl and cyano groups also brought the desired 4-alkylated products 14b,^[12m] 15b, and 16b in good yields. Finally, secondary alkyl bromide was tolerated to give excellent isolated yield after extended reaction time.

Scheme 2. Substrate scope for the alkylation of 2,4-dihydroxyacetophenone^[15].

.

In summary, 4-alkoxy-2-hydroxybenzaldehydes and 4-alkoxy-2-hydroxyacetophenones are important building blocks that are widely used in different fields. Although their synthesis is known, available conditions usually require a long reaction time, and generate unsatisfactory regioselectivity and low yields. By screening different reaction conditions, we have developed a CsHCO₃-mediated regioselective alkylation of 2,4-dihydroxybenzyaldehydes and 2,4-dihydroxyacetophenones. Under our conditions, the reactions proceed smoothly to generate the desired mono-alkylation product in up to 95% isolated yield. Compared to stronger carbonate bases, CsHCO₃ offers the favorable basicity for selective alkylation on the 4-hydroxy group of the substrates to give the desired products with the minimum formation of bis-alkylated site products. In addition, compared to sodium and potassium bicarbonate salts, the CsHCO₃ reaction indicated significantly improved yield due to its higher solubility and decreased coordination interaction of the softer cesium ion to the phenoxide anion that could favor ligand exchange.