Effect of Substituents and Stability of Transient Aluminum-Aminals in the Presence of Nucleophiles

Abstract

Disubstituted hydroxylamines were synthesized and used to form aluminum–amide complexes. These reagents masked carbonyl groups in situ from nucleophilic addition. The stability and utility of the aluminum–aminals are presented in the context of selectively controlling nucleophilic addition on substrates with multiple carbonyl groups.

Graphical Abstract

Carbonyl groups are prominent functional groups on many organic compounds and naturally occurring molecules. The careful manipulation of these groups and control of their reactivity as electrophiles is critical in the synthesis of complex targets.^1–3 The primary methods to address these issues are the use of carbonyl protecting groups or the implementation of oxidation/reduction strategies. An alternative tactic is the incorporation of a Weinreb amide, an aldehyde or ketone precursor, into the synthesis.^4,5 The process of transforming a Weinreb amide into a carbonyl group requires a nucleophilic addition that results in the formation of a transient, but stable aminal (Scheme 1).^6,7 This aminal is resistant to subsequent nucleophilic attack but collapses to a carbonyl in the presence of aqueous acid.

Scheme 1.

Strategies for the formation of stable aminals.

Recently, Colby and coworkers demonstrated the effectiveness of trapping reactive carbonyl groups as aminals using an aluminum–amide complex rather than forming them from Weinreb amides (Scheme 1).⁸ This aluminum complex is prepared with diisobutylaluminum hydride (DIBALH) and N,O-dimethylhydroxylamine. It is an alternative to prior methods of using highly basic lithium–amides.^9–12 The intermediate aluminum–aminal is resistant to different types of reactive nucleophiles such as Grignard reagents and organolithiums. Using this in situ masking method, selective additions to esters were accomplished in the presence of aldehydes and ketones (Table 1). In a similar fashion, selective additions to ketones in the presence of aldehydes were executed. Specifically, the more reactive carbonyl group was trapped after treatment with a mixture of DIBALH and N,O-dimethylhydroxylamine. An equivalent of i-PrMgCl was used as a base to remove the last acidic proton and stabilize the aminal. Substrates 1–3, each containing two distinct carbonyl groups, readily participated and the products 4–6 were isolated in good yields with either an organolithium or Grignard reagent adding selectively to the less reactive carbonyl group.

Table 1.

Stability of Aluminum–Aminals to Nucleophilic Additions

Entry	Substrate	R-M	Product	Yielda
1		n-BuLib		87%
2	1	MeMgBr		70%
3		MeLi	5	63%
4	2	MeMgBr	5	86%
5		MeLib		83%
6	3	MeMgBr	6	76%

^aIsolated yields.

^bOrganolithium was used instead of i-PrMgCl.

This methodology enables the aluminum–amide to add to the more reactive carbonyl group and eliminate its electrophilic nature by creating a stable aminal following addition of a base, usually i-PrMgCl. A drawback of this approach is the use of a base to remove the last acidic proton and stabilize the aminal intermediate. By re-designing the generation of the aminal and adding the base earlier in the sequence, we speculated that the scope of compatible reactions would be enhanced. Specifically, we aimed to create a discrete aluminum reagent that would form the aminal in one step.¹³ In order to achieve this objective, adding n-BuLi or i-PrMgCl to HN(OMe)Me•HCl succeeded in consuming the acidic proton of the hydrochloride and then subsequent treatment with Me₃Al formed the requisite aluminum–amide complex. Substrates 1–3, each with two distinct carbonyl groups, participated in this more efficient method and provided access to products 4–11 upon addition of organolithiums, Grignard reagents, reducing reagents, Wittig reagents, and enolate anions (Table 2). High levels of chemoselectivity were observed, good yields of products were isolated, and the scope of the reaction was enhanced as more nucleophiles participated in the process.

Table 2.

Stability of Aluminum–Aminals Formed by a Reagent

Entry	Substrate	R-M	Product	Yielda
1	1	n-BuLi	4	91%
2	1	allylMgBrb		83%
3	1	DIBALHb		89%
4	2	MeLi	5	86%
5	2	EtMgBrb		85%
6	2	Ph3P=CH2		86%
7	2			70%
8	3	MeMgBr	6	70%

^aIsolated yields.

^bi-PrMgCl was used instead of n-BuLi.

Next, the stability of the aluminum-amide complex was investigated across three weeks (Table 3).¹⁴ A 0.5 M solution of the aluminum–amide reagent in THF was prepared using HN(OMe)Me•HCl, n-BuLi, and Me₃Al and stored at 0 °C. This reagent was used to transform methyl 4-formylbenzoate 1 into the alcohol 4 following treatment with n-BuLi. The reagent performed superbly at the time points: one hour, one day, one week, two weeks, and three weeks when compared to the baseline reaction yield of 91% following immediate use. A slight enhancement in yields was observed and this result may be attributed to the additional time that the reagent has to form the aluminum–aminal.

Table 3.

Effect of Time on Reagent Prior to Formation of Aminal

Entry	Storage Time for Reagent at 0 °C (step i)a	Yieldb
1	immediate use	91%
2	one hour	99%
3	one day	99%
4	one week	97%
5	two weeks	98%
6	three weeks	99%

^aA 0.5 M solution of the aluminum–amide reagent in THF was prepared using HN(OMe)Me•HCl, n-BuLi, and Me₃Al and stored at 0 °C.

^bIsolated yields.

Following the collection of these promising data, the aminal was re-designed to understand its stability when the methyl groups on the N,O-dimethylhydroxylamine were replaced by other large alkyl groups (Scheme 2). We have already demonstrated the aluminum species can be derived from M₃Al or DIBALH and the respective methyl or isobutyl group is compatible with the process. The next site to investigate was the contribution of substituents on the nitrogen and oxygen of the hypothesized aluminum–aminal intermediate.

Scheme 2.

Design of aluminum–amide reagents to evaluate the effects of substituents on the nitrogen and oxygen atoms.

Two synthetic strategies for the N,O-dialkylhydroxylamines were undertaken according to the literature precedents for these new stability studies (Scheme 3).^15,16 A combination of N-hydroxyphthalimide 12 and a benzylhalide followed by treatment with hydrazine generated the corresponding O-benzylhydroxylamine 13. The amine 13 was protected and the intermediate carbamate was treated with NaH and an alkyl halide to give the carbamate 14. Acidolysis of carbamate with TFA provided the corresponding N,O-dialkylhydroxylamines 15a–d. Alternatively, the N,O-dibenzylhydroxylamine 17 was synthesized from O-benzylamine hydrochloride 16 with benzyl bromide in the presence of potassium carbonate.¹⁶

Scheme 3.

Two synthetic routes for substituted hydroxylamines.

The utility of the new aluminum–amide precursors was investigated by treating the amines with Me₃Al and using them for the in situ masking of carbonyl groups on substrates 1 and 2 (Table 4). The selective addition of n-BuLi and EtMgBr to the less reactive carbonyl group was examined and compared to the isolated yields when N,O-dimethylhydroxylamine was the precursor. Typically, the major product was isolated in modest to good yields, and these yields were lower than observed with the N,O-dimethylhydroxylamine. Specifically, the use of amine 15a that displays an ethyl and benzyl group provided a moderate yield of product 4. Exchanging the ethyl group for an allyl group (i.e., amine 15b) had a minimal effect on the yield compared to 15a. Similarly, the amine 15c with a para-methoxybenzyl substituent provided yields similar to the benzyl amine 15b. The highest yield was obtained using the naphthyl substituted amine 15d; however, the lowest yield was obtained when using the dibenzyl amine 17. Overall, the yields of products 4 and 9 were in the range of 51–72% and these results are somewhat lower than using N,O-dimethylhydroxylamine that provided 85–91%. The conclusion is that exchanging the methyl substituents of N,O-dimethylhydroxylamine with other alkyl groups is well tolerated with only a modest effect on yields.

Table 4.

Effects of Substitution Pattern of Hydroxylamines on the Stability of the Aminal.

Entry	Substrate	Amine	R-M	Product	Yielda
1	1	HN(OMe) Me•LiClb	n-BuLi	4	91%
2	1		n-BuLi	4	58%
3	1		n-BuLi	4	67%
4	1		n-BuLi	4	66%
5	2	HN(OMe) Me•LiClb	EtMgBr	9	85%
6	2		EtMgBr	9	72%
7	2		EtMgBr	9	51%

^aIsolated yields

^bFormed from n-BuLi and HN(OMe)Me•HCl.

In conclusion, we have demonstrated that aluminum complexes derived from N,O-dialkylhydroxylamines can serve as reagents to trap reactive carbonyl groups as aminals from reactive nucleophiles and that there is substantial flexibility in the implementation of this method. The broad scope of this process has been established by synthesizing and reacting N,O-dialkylhydroxylamines with different carbonyl group-containing substrates. The effect on using different substituents on the nitrogen atom was modest despite exchanging methyl groups with substantially larger substituents. These data are compelling as they lay a foundation to design new classes of reagents in which these substituents can potentially participate in synthetic transformations or direct addition reactions along with masking the carbonyl group from nucleophiles.

Compounds 4–6⁸ and 7–11¹³ have been previously described. Syntheses of amines 15a,¹⁷ 15b,¹⁸ 15c,¹⁹ 15d,²⁰ and 17¹⁶ were conducted according to the literature methods. Characterization data for each of the known compound matched previous reports.

General Procedure for Formation of Aluminum–Aminal followed by Addition of Organolithium Reagent

To a 0 °C solution of amine 15a¹⁷ (74.9 mg, 0.49 mmol) in THF (2 mL) was added Me₃Al (250 µL, 2.0 M in hexanes). The mixture was warmed 0 °C to rt across 30 minutes. Next, the resultant mixture was added to a 0 °C solution of methyl 4-formylbenzoate 1 (67 mg, 0.41 mmol) in THF (10 mL), and the mixture was stirred for 2 hours at rt. Following cooling to –78 °C, n-BuLi (650 µL, 2.5 M in hexanes) was added, and the mixture was stirred for 35 minutes at –78 °C. The reaction was quenched with sat. aqueous NH₄Cl (10 mL) and extracted with EtOAc (3 × 15 mL). The organics were dried over Na₂SO₄, filtered, concentrated under reduced pressure. SiO₂ flash chromatography (8:2 hexanes/EtOAc) afforded product 4, (58.8 mg, 58% yield). Characterization data was identical with the reported data.⁸

General Procedure for Formation of Aluminum–Aminal followed by Addition of Grignard Reagent

To a 0 °C solution of amine 17¹⁶ (103 mg, 0.48 mmol) in 4:1 Et₂O/THF (5.5 mL) was added Me₃Al (484 µL, 1.0 M in heptane). The mixture was warmed from 0 °C to rt across 30 minutes. Next, the resultant mixture was added to a 0 °C solution of 4-acetyl benzaldehyde 2 (65 mg, 0.439 mmol) in Et₂O (22 mL), and the mixture was stirred for 2 hours at rt. Following cooling to 0 °C, EtMgBr (293 µL, 3.0 M in Et₂O) was added, and the mixture was stirred for 40 minutes with warming from 0 °C to rt. The reaction was quenched with sat. aqueous NH₄Cl (20 mL) and extracted with EtOAc (3 × 25 mL). The organics were dried over Na₂SO₄, filtered, and concentrated under reduced pressure. SiO₂ flash chromatography (3:1 hexanes/EtOAc) afforded product 9 (40 mg, 51% yield). Characterization data was identical with the reported data.¹³