HOTf-Catalyzed Alkyl-Heck-type Reaction.

The Heck reaction, along with other cross-coupling reactions, led to a revolution in organic chemistry. In the last 50 years, metal-catalyzed, photo-induced, or base-mediated Heck and Heck-type reactions have been elegantly developed. Brønsted acid-catalyzed Heck (or Heck-type) reactions are still unknown, however. By introducing alkyl peroxides as the key intermediates, primary, secondary, and tertiary aliphatic carboxylic acids are therefore applied here in a one-pot Brønsted acid-catalyzed Heck-type reaction, to deliver E-alkenes exclusively in most cases. The use of HOTf is vital to the reaction, whose mechanism is supported by both experimental and computational results. This method can be expanded to the direct alkylation of complex natural products.

Introduction

The Heck reaction, pioneered by Heck and Mizoroki in the late 1960s and the early 1970s (Heck, 1968, Mizoroki et al., 1971, Heck and Nolley, 1972), along with other cross-coupling reactions, led to a revolution in organic chemistry (Johansson Seechurn et al., 2012). In the last 50 years, many types of Heck and Heck-type reactions, including metal-catalyzed (Heck, 1968, Mizoroki et al., 1971, Heck and Nolley, 1972, Littke and Fu, 2001, Farrington et al., 2002, Na et al., 2004, Loska et al., 2008, Delcamp et al., 2013, Nishikata et al., 2013, Standley and Jamison, 2013), photo-induced (Iqbal et al., 2012, Liu et al., 2013, Paria et al., 2014, Yu et al., 2014), or base-mediated (Rueping et al., 2011, Shirakawa et al., 2011, Sun et al., 2011) reactions, have been elegantly developed (Beletskaya and Cheprakov, 2000, Dounay and Overman, 2003, Wu et al., 2010, Le Bras and Muzart, 2011, Mc Cartney and Guiry, 2011, Tang et al., 2015). Notwithstanding these classical reaction modes, there is no precedent of Brønsted acid-catalyzed or Brønsted acid-promoted Heck (or type) reaction being realized. Moreover, compared with aryl Heck reactions, the alkyl-Heck (type) reaction has been developed less. This is due mainly to the potential accompanying side reactions. Significant breakthroughs in alkyl-Heck-type reactions have, however, been made (Ikeda et al., 2002, Liu et al., 2012, Liu et al., 2015, Nishikata et al., 2013, Mcmahon and Alexanian, 2014, Zou and Zhou, 2014, Kurandina et al., 2018, Wang et al., 2018) (Scheme 1A), and in this article, we report a Brønsted acid-catalyzed alkyl-Heck-type reaction.

Scheme 1

Intermolecular Alkyl-Heck-Type Reaction of General Alkyl Groups and Decarboxylative Vinylic Alkylation of Aliphatic Acids

(A) Previous alkyl-Heck-type reactions by Oshima, Alexanian, Zhou, Li, Fu, Lei, and Nishikata.

(B) Previous decarboxylative vinylic alkylation with aliphatic acid derivatives.

(C) This work: Brønsted acid-catalyzed alkyl-Heck-type reaction.

As is well known, alkyl halides are one of the most frequently used alkyl functionalities for alkyl-Heck-type reactions (Kambe et al., 2011, Weix, 2015, Tellis et al., 2016, Choi and Fu, 2017). However, their shortcomings, such as limited availability and perceived instability might prevent more extensive applications (Qin et al., 2016). Furthermore, there are still significant challenges remaining for alkyl-Heck-type reactions such as E/Z selectivity, use of metal catalysis, and diversity of alkyl sources (Scheme 1A). Carboxylic acids are inexpensive, stable, non-toxic, and structurally diverse feedstock chemicals that have been widely used in numerous reactions. For example, they have been utilized in cross-coupling with prefunctionalized alkenes such as vinyl halides or their derivatives to generate alkenes (Mai et al., 2013, Noble et al., 2015, Toriyama et al., 2016, Wang et al., 2016, Edwards et al., 2017, Xu et al., 2017, Zhang et al., 2017) (Scheme 1B). However, the decarboxylative cross-couplings of aliphatic acids or their derivatives with alkenes (X = H) are very rare (Wang et al., 2018). As part of our ongoing interest in the application of aliphatic acids as the alkyl source (Li et al., 2016, Ge et al., 2017, Jian et al., 2017, Qian et al., 2017, Ye et al., 2017, Zhu et al., 2017) and our interests in the discovery of different reaction models of alkyl peroxides, we have developed the first Brønsted acid-catalyzed alkyl-Heck-type reaction of alkenes with aliphatic acids via alkyl peroxide intermediates (Scheme 1C).

Results and Discussion

Optimization Study

We commenced our studies by screening a variety of Brønsted acids for the alkyl-Heck-type reaction of styrene with aliphatic acid. The aliphatic acid was converted into alkyl peresters in the presence of trifluoroacetic anhydride (TFAA) and tert-butyl hydroperoxide (TBHP) and used in situ for the subsequent step (Donchak et al., 2006). The best Brønsted acid was found to be HOTf, which offered the desired alkylated alkene 3 exclusively as a single E-isomer in 88% yield, determined by 1H nuclear magnetic resonance (NMR) analysis (Equation 1 and Table 1, entry 1). Studies of acids showed that Tf2O had a lower efficiency (Table 1, entry 2) and other Brønsted acids such as TsOH⋅H2O, CF3COOH, HOAc, and MeSO3H were ineffective in this reaction (Table 1, entries 3–6). When performed at room temperature (rt), the reaction afforded the desired product in 70% yield (Table 1, entry 7). To exclude the possibility of interference of trace amount of metal in HOTf, the HOTf was used after redistillation and the product was obtained in the same yield (Table 1, entry 8). The role of light was investigated by conducting the reaction in the dark, but no difference in the yield was observed (Table 1, entry 9). In the absence of HOTf, the alkyl peroxide decomposed completely and the styrene remained unchanged (Table 1, entry 10).

Table 1

Optimizations of Reaction Condition

Entry	Variation from the Standard Conditions	Yield(%)a,b
1	None	88(75c)
2	Tf2O instead of HOTf	78
3	TsOH⋅H2O instead of HOTf	Trace
4	CF3COOH instead of HOTf	Trace
5	HOAc instead of HOTf	Trace
6	MeSO3H instead of HOTf	Trace
7	Room temperature instead of 50°C	70
8	Fresh distilled HOTf	88
9	In dark	90
10	Without HOTf	Trace

Reaction conditions: First, 2-ethylhexanoic acid 1 (1.5 mmol), TBHP (1.5 mmol), and TFAA (1.5 mmol) at 0°C–rt for 4 hr, and then THF (2 mL), styrene 2 (0.5 mmol), and HOTf (0.05 mmol) were added. The mixture was stirred at 50°C for 8 hr.

1H NMR yield.

Yield of the isolated product.

Scope of the Investigation

With the identified conditions in hand, we studied the scope of alkenes for this one-pot process (Figure 1). In most of the cases, the products were obtained as a single E-isomer. Reactions of vinyl arenes containing carbon substituents at the o-, m-, and p-positions afforded the corresponding products (4–10) in good yield (68–84%). Vinyl arenes containing halides reacted with 2-ethylhexanoic acid to give the desired products (11–15) in moderate to good yield (54%–80%). Functional groups, such as dimethylaminomethyl, and even free carboxylic acid and boronic acid were compatible with the reaction conditions (20–23). α-Methylstyrene and α-phenylstyrene participated in the reaction smoothly, providing the products (24, 25) in 82% and 93% yields, respectively. Furthermore, an enyne was a suitable substrate for this reaction, and the corresponding terminal-cross-coupled product (26) was obtained in good yield (71%). 1-Octene, an unactivated alkene, examined under the standard reaction conditions was not reactive to this reaction.

Figure 1

Alkyl-Heck-Type Reaction of Alkenes

Top: One-pot process from aliphatic acid: First, 2-ethylhexanoic acid 1 (1.5 mmol), TBHP (1.5 mmol), and TFAA (2.0 mmol) at 0°C–rt for 4 hr, and then THF (2 mL), alkene (0.5 mmol), and HOTf (0.05 mmol) were added. The mixture was stirred at 50°C for 8 hr; yields of isolated products.

Bottom: HOTf (0.35 mmol) was added for 16, 18, 19, and 21.

The acetyl group on oxygen atom was removed under the reaction conditions for 19.

HOTf (0.75 mmol) was added for 22.

See also Figures S45–S94.

Next, we proceeded to study the scope of the reaction with respect to secondary and tertiary aliphatic carboxylic acids (Figure 2). The desired products (28–43) were obtained in moderate to high yields, using acyclic or cyclic aliphatic acids. The compatibility of various functional groups was good, and many functional groups, such as carbonyl (42), imide (38), amine (43), and ether (36), were tolerated. Most importantly, the E/Z selectivity of this reaction was excellent and only E-alkenes were observed. We then tried to expand this reaction to primary aliphatic acids, but the desired products were obtained in low yields as the methylated vinylic products were observed as by-products (Zhu et al., 2017). To overcome this problem, primary aliphatic acids were converted into alkyl diacyl peroxides and then subjected to the reaction (Figure 3). With the similar reaction conditions (please see Table S4 for details), generic primary aliphatic acids afforded the corresponding products (44–48) in good yields (60–77%). Primary aliphatic acids with functionalities, e.g., the bromide (49), chloride (50), ketones (51 and 52), ester (53), or the alkene (54) were well tolerated in the protocol, delivering the corresponding products in moderate to good yields. In every case, the E-alkene was obtained exclusively.

Figure 2

Alkyl-Heck-Type Reaction of Secondary and Tertiary Aliphatic Acids

Top: One-pot process from aliphatic acid: First, acid (1.5 mmol), TBHP (1.5 mmol), and TFAA (2.0 mmol) at 0°C–rt for 3–5 hr, and then THF (2 mL), styrene 27 (0.5 mmol), and HOTf (0.05 mmol) were added. The mixture was stirred at 50°C for 8 hr; yields of isolated products.

Bottom: Styrene 27 (0.5 mmol), perester (1.25 mmol), and HOTf (0.1 mmol) at 80°C for 8 hr for 35, 36, 38, 42, and 43.

See also Figures S95–S127.

Figure 3

Alkyl-Heck-Type Reaction of Primary Aliphatic Acids

Top: Reaction conditions: alkyl diacyl peroxide (synthesized from acid, 1.0 mmol), styrene 2 (0.5 mmol), and HOTf (0.1 mmol) in THF (1 mL); yields of isolated products.

Bottom: Alkyl diacyl peroxide (synthesized from acid, 1.0 mmol), styrene 2 (0.5 mmol), and HOTf (0.25 mmol) in THF (2 mL) for 49 and 50.

See also Figures S128–S149.

Synthetic Applications

To highlight the synthetic utility of this methodology (Scheme 2), the perester (55), which is readily derived from the corresponding steroidal carboxylic acid, was coupled with styrene in the presence of HOTf. The decarboxylative Heck-type coupling product (56) was obtained in 48% yield as a single isomer. The configuration of the product (56) was reversed, and this was confirmed by X-ray crystallographic analysis (please see Tables S5 and S6 for details). The reaction of 57 afforded the desired product (58) in 65% yield with E-selectivity. Gemfibrozil 59, an oral drug used to lower lipid levels, could also be converted into the vinylated product (60). These examples demonstrated that this reaction is potentially useful for the functionalization of complex molecules in the late stage.

Scheme 2

Synthetic Applications

See also Figures S150–S155.

Mechanistic Study

To probe the mechanism of the reaction, a series of control experiments were performed. The reaction of α-phenylstyrene with 2-cyclopropylacetic acid under the standard conditions afforded the ring-opening product (61) in 62% yield (Scheme 3A), supporting the assumption of the involvement of radical species in the reaction. The competitive reaction of styrene and d8-styrene used in 1:1 ratio in the presence of HOTf and lauroyl peroxide (LPO) offered an identical yield of the corresponding products (Scheme 3B). When the reaction of d8-styrene with perester 62 was performed in tetrahydrofuran (THF), the desired product (-3) was isolated (Scheme 3B). Interestingly, the deuterated side products d(OD)-butanol were detected by gas chromatography-mass spectrometry (GC-MS). To further explore the mechanism, possible intermediates 63 and 64 were synthesized and tested with or without HOTf (Scheme 3C). Compounds 63 and 64 are thermally stable in the absence or presence of one equivalent of C11H23COOH. Even though compounds 63 and 64 can be converted to the desired alkene products in the presence of 0.2 equivalent of HOTf, it is unlikely that they are competent intermediates because the formation of 63 or 64 was not observed using GC-MS when the corresponding Heck reaction was conducted no matter with or without HOTf (Ge et al., 2017).

Scheme 3

Mechanism Studies

(A) Radical clock reaction.

(B) Deuterium labeling experiment.

(C) Exclusion of possible intermediates.

See also Figures S156–S164.

Plausible Reaction Mechanism

As the result shown in entry 10 of Table 1, no desired product was observed in the absence of HOTf, implying that HOTf must play a vital role in the reaction. Density functional theory (DFT) calculations were carried out to gain further insight into the reaction mechanism. As can be seen from Scheme 4, before the catalytic cycle R⋅ radical I-5 can be formed by homolytic dissociation of the alkyl diacyl peroxide, which is a very slow step with a high barrier of 27.5 kcal/mol. However, this is considered as the trigger to invoke the following catalytic cycle. Attack on the styrene substrate by the active species R⋅ radical to form a benzyl radical (I-6) leads to energies lower by 31.3 kcal/mol with a small barrier of 2.8 kcal/mol, indicating that such a reaction is both thermodynamically and kinetically favorable. In the beginning of the catalytic cycle, LPO binding a molecule of HOTf forms a complex I-1 with a strong hydrogen bonding of 10.2 kcal/mol. This complex oxidizes benzyl radical (I-6) to yield a benzyl cation species (I-2), a radical (I-3), and an OTf− anion, which is exothermic by 4.4 kcal/mol. Meanwhile, the generated OTf− deprotonates I-2 to yield the product and regenerate the acid HOTf with a reaction energy of −13.4 kcal/mol. Thus, from the reactions of LPO and I-6 with the product and I-3, a proton-coupled electron transfer process is promoted by HOTf, which serves as the driving force and proton source for the reaction. Thereby, homolytic dissociation of I-3 leads to RCOO⋅ radical (I-4) and RCOOH, which is exothermic by 2.3 kcal/mol without any barrier. Subsequently, C-C cleavage of I-4 is exothermic by 3.8 kcal/mol, which releases the active species R⋅ radical (I-5) and CO2 to close the catalytic cycle. Alternatively, in the absence of HOTf formation of this radical I-4 with carboxylic acid RCOOH requires high energies (>27 kcal/mol, See Scheme S1), indicating that the strong acidity of HOTf plays a significant role in the formation of I-4. A similar mechanism of reaction starting from perester was also calculated and presented in Scheme S2.

Scheme 4

Plausible Reaction Mechanism

Conclusion

We have developed a Brønsted acid-catalyzed radical alkyl-Heck-type reaction of alkenes with aliphatic acids. This HOTf-catalyzed process has been shown to be an efficient method to deliver only E-alkenes in most cases. Relatively simple and available starting materials are used, and wide substrate scope and good functional group tolerance are observed. Preliminary mechanistic studies illustrated the vital role of HOTf in the reaction, whose proposed mechanism is supported by both the experimental and computational results.

Methods

All methods can be found in the accompanying Transparent Methods supplemental file.