4-Aminobenzotriazole (ABTA) as Removable Directing Group for Palladium-Catalyzed Aerobic Oxidative C-H Olefination

Abstract

The 4-aminobenzotriazole (ABTA) was applied as an effective removable directing group (DG) in Pd-catalyzed C-H activation for the first time. Compared with the widely applied pyridine and quinoline analogs, the ABTA showed significantly improved reactivity, achieving aerobic-oxidative C-H olefination in excellent yields (up to 95% vs. <50% with other reported DGs under identical conditions). Using this new strategy, macrocyclization was achieved to give cyclic peptides in good yields with easy ABTA removal under mild condition, highlighting the promising potential of this new DG.

Graphical Abstract

With the ability to achieve late-stage structural modification, the metal catalyzed C-H functionalization has been considered one of the most important developments in chemical synthesis over the past two decades.¹ To achieve good selectivity, the incorporation of a removable directing group (DG) has been applied as one general strategy.² The benchmarks of good directing groups are A) clear reactivity enhancement and B) easy removal under mild conditions. The amino N-heterocycles (ANH) are important auxiliary DGs for the substrate containing COOH.³ The bidentate binding site of resulting amide could boost the reactivity of metal center for selected C-H activation (Scheme 1A). Subsequent deprotection of DGs gives the C-H functionalization with the initial COOH moiety. According to literature, both aliphatic amines (pyridine-PIP and triazole-TAM) and aromatic amine (8-amino quinoline AQ) were used.⁴ Compared with aliphatic amine, aromatic amine is much easier to remove, which makes AQ a popular DG with numerous successful applications reported.⁵

Scheme 1.

Direct C-H functionalization with ANH DGs

In 2014, Ackermann first introduced triazole amine (TAM) in the iron catalyzed C-H functionalization.^4e Our group later disclosed the TAM directed Pd catalyzed C-H alkynylation under silver free conditions.⁶ Compared with pyridine and quinoline, 1,2,3-triazole is a more stable heterocycle under redox conditions, making it an unique DG in promoting C-H activation.⁷

Herein, we report the synthesis of ABTA derivatives with the confirmation of various N-isomers and application of this new DG in the Pd catalyzed aerobic C-H olefination and macrocyclization (cyclic peptides, Scheme 1B). Impressively, the ABTA gave a significantly improved reactivity over AQ and PIP with easy removal under mild conditions, revealing ABTA as a promising new member of auxiliary ANH-DGs for metal catalyzed C-H functionalization

We initiated our investigation with the development of regio-selective synthesis of N-substituted 4-amino-benzotriazole (ABTA). As shown in Figure 1A, 4-nitro-benzotriazole 1b could be prepared from benzotriazole (BTA, < $ 0.1/g) with high yield. After testing various reduction conditions, Pd/C promoted hydrogenation is selected as the practical condition for the NO₂ reduction. The key challenge for N-substituted triazole synthesis is the identification of various N-isomers. Both N-arylation and N-alkylation were explored (see details in SI).⁸ In summary, reaction of 4-amino-BTA 1c under copper catalyzed amination works well to reach N-aryl-substituted ABTA, giving N-1 and N-2 isomers in 3:1 ratio (no N-3). Alternatively, for N-alkyl ABTA, the optimal synthetic sequence is 4-nitro-BTA 1d alkylation (SN₂) followed by NO₂ reduction. All these isomers have been isolated with structures confirmed by X-ray analysis. Overall, this work offered a systematic investigation on BTA substitution, which set up the foundation to synthesize ABTA library as a new building block for future applications (Figure 1B).⁹ With the substituted ABTA (20–50 gram scale) prepared, we investigated their reactivity toward Pd mediated C-H activation.

Figure 1.

Synthesis of various ABTA isomers and Pd mediated C-H activation.

Encouraged by this result, we explored the Pd-catalyzed C-H olefination, which is known undergoing Pd(II/0) cycles with the possibility of using O₂ as the oxidant (for Pd⁰ oxidation). According to literature, the DG-assisted C-H olefination has been problematic and often require strong external oxidants (over 1 eq. Ag salts or BQ).¹⁰ Thus, achieving reaction under aerobic condition is of great interest for practical synthesis. Reactions between Pd-complexes 4a/4c and alkene 5a were performed. While the 5,5-bicyclic complex 4a produce no reaction even at 120 °C, the 6,5-bicyclic complex 4c gave the desired olefination product 6a in 75% yield (16% 3e, C-Pd to C-H). This lack of reactivity of complex 4a is consistent with literature reports, likely due to the high energy transition state associated with the 5,5-bicyclic system in the migratory intersection step.¹¹ Nevertheless, the clean reaction obtained with the 6,5-bicyclic complex 4c is exciting since it offered a potential new route to achieve C-H olefination using molecular oxygen as oxidant. It is important to note that other DGs showed significantly reduced reactivity under the identical conditions (see details in SI), which highlighted the “boosting effect” of ABTA.

The catalytic conditions were explored by reacting 3e and 5a with Pd catalyst. After screening various conditions, 10% Pd(OAc)2, 20% AgOAc and 5.0 eq KOAc in DCE at 120 °C for 24 h (sealed tube) with N-Bn modified ABTA were identified as the optimal conditions (see detailed screening in SI), giving the desired product 6a in 92% isolated yield. Alternative N1-substituted ABTAs gave similar results (see SI). The N1-Bn ABTA was selected due to the relatively easy synthesis. Results from representative alternative conditions are summarized in Table 1.

Table 1.

Screening Table^{a, b}

entry	variation from “standard conditions”	conv. (%)	yield of 6a (%)
1	None	100%	93%(92%)
2	no AgOAc	73%	67%
3	no AgOAc and KOAc	20%	15%
4	NaOAc instead KOAc	90%	81%
5	K2CO3 instead KOAc	25%	19%
6	toluene as solvent	81%	73%
7	dioxane as solvent	60%	53%
8	Ar protection	25%	17%
9	air	34%	27%
10	100 °C	92%	84%
11	AQ as Directing Group	51%	43%
12	PIP as Directing Group	<10%	trace
13	TAM as Directing Group	53%	47%

^aConditions: 3e (0.2 mmol), 5a (0.8 mmol), base (1.0 mmol), Pd (OAc)₂ (0.02 mmol), AgOAc (0.04 mmol) and solvent (2.0 mL) in a 25 mL Schlenk tube under O₂ atmosphere for 24 h.

^b1H NMR yields using 1,3,5-tribromobenzene as an internal standard (isolated yields).

Although the reaction could occur using O₂ as the only oxidant, the starting material cannot reach complete conversion using only Pd (entry 2). This is likely due to the Pd-ABTA coordination in the product, causing slow catalyst turnover. Addition of catalytic amount of silver solved this problem. The KOAc base is crucial for good reactivity (entries 4 and 5). Other solvents (toluene and dioxane) gave lower yields than DCE (entries 6 and 7). Significantly reduced conversion was observed if conducting the reaction under Ar, confirming the important role of O₂ as the oxidant (entries 8 and 9). Lowering the reaction temperature to 100 °C led to slightly lower conversion (entry 10). It is important to note that other DGs (AQ, PIP, TAM) all gave much poorer results under identical conditions (entries 11–13), highlighting the boosting effect of this new ABTA DGs. With the optimal reaction conditions developed, substrate scope was explored as summarized in Table 2.

Table 2.

ABTA substrate scope^{a, d, e}

^aReaction conditions: 1 (0.2 mmol), alkene (0.8 mmol), Pd(OAc)₂ (0.02 mmol), AgOAc (0.04 mmol), KOAc (1.0 mmol) in DCE (2.0 mL) under O₂ atmosphere for 24 h. Isolated yield.

^b36 h.

^c48 h.

^dalkene (2.4 mmol), 48 h for di-olefination.

^eThe ratio of mono and di substitution was determined by ¹HNMR.

The EDG modified arenes (6a-6b) ran smoothly, giving the desired products in excellent yields (>90%). Substrates with EWG modified arenes (6c-6e) required longer reaction time (36 h) to reach completion. For the meta-substituted benzenes, substitution occurred at the less hindered ortho-position (6f-6h). The selectivity of mono- and di-olefination was less ideal with substrates containing both two accessible ortho-C–H bonds (6i-6n). Application of excess alkene (12 equiv) gave the di-olefination product in excellent yields (7a-7j). The OTBS and NPhth groups were compatible with this transformation, offering new synthetic handles (6m, 6n, 6r). Some valuable carboxylic derivatives, such as 2-thenoic acid (6p), 2-phenylpropionic acid(6q), and (S)-Ibuprofen (6s), all worked well and gave the desired products in good yields.

In addition to acrylate, styrene and 2-vinylnaphthalene also worked well in this transformation (8a-8g). Impressively, alternative EWG-activated alkenes, such as acrylamide (8h), acrylonitrile (8i), vinyl phosphate (8j), and vinyl sulfone (8k), are all suitable, giving the desired olefins in excellent yields. The coupling with an estrone derivative was successfully achieved, showing the application of this method for late-stage functionalization of drug-like molecules (8l). Notably, the amino acid modified ABTA (6t) worked well, with the corresponding olefination product obtained in excellent yield. The feasibility of introducing various functional groups highlighted the advantage of ABTA over other DGs for the preparation of bio-active compounds.

With the confirmation of ABTA as a new DG in C-H olefination, we focused on the more challenging macrocyclization.¹² Recently, Wang reported the first example of amino acid directed peptide macrocyclization through C-H olefination.¹³ While this seminal work initiated a new strategy to achieve cyclic peptides, the transformation relied on α-t-Bu amino acid DG, likely due to the needed conformation control for a better Pd coordination. We compared reactions of ABTA with AQ, PIP, and the L-t-Bu-Leucine (Wang’s strategy) substrates. As expected, no cyclization was observed with either AQ or PIP substrates (see SI). With the ABTA-DG, the cyclization product was observed in 77% yield. Notably, under the identical conditions, the Wang’s amino acid DGs gave a much lower overall yield (22%) due to slow reaction rate (Table 3). In 2010, Yu reported Pd(II) catalyzed C-H olefination aerobically with the free carboxylic acids.^1g We also carried out the macrocyclization reaction with free carboxylic acid. No reaction was observed even under their optimal conditions. These results highlighted the importance of using DGs to promote challenging macrocyclization with boosted Pd stability and reactivity (see SI). Encouraged by this result, various ABTA-substrates were prepared for this macrocyclization. The result is summarized in Table 3.

Table 3.

ABTA promoted macrocyclization^a

^aReaction conditions: 9 (0.1 mmol), Pd(OAc)₂ (0.01 mmol), AgOAc (0.2 mmol) in DCE (10 mL) for 12 h. Isolated yield.

^bDMF (10 mL).

Products with different ring sizes (14 to 22) were prepared with good yields (10a-10e). Impressively, linkers containing Gly, Abu, Ala, Val and Ile residues worked well (10f-10l), suggesting the possibility of applying this method for peptide cyclization. In fact, using DMF as solvent, the cyclic peptides were produced in good yields with ring sizes of 16 (10m-10o) and 19 (10p). Notably, the ring size have a large impact on the E/Z selectivity. Smaller ring (14–16) favoured for Z conformation to release the ring strain (see SI for detail). 13-membered polyamides gave a messy reaction mixture, which might be due to the higher ring strain. Other DGs could not promote this transformation under either aerobic or silver oxidative condition, giving significantly reduced yields.

Besides enhanced reactivity, the other critical feature of a good DG is the feasibility to remove under mild conditions.¹⁴ With the electron deficient benzotriazole, ABTA is expected as better leaving group than other DGs. As shown in Figure 2A, ABTA could be removed upon methanolysis of Boc-protected amide with less nucleophilic MeOH under mild condition. The Boc protected 1c could be recycled in >90% yield. Notably, while more stable aliphatic amide TAM, PIP and amino acid DGs required much harsher conditions to be removed, the aromatic amine AQ-DG could not be deprotected under this mild condition. Through hydrolysis using LiOH, ABTA can be easily removed, giving corresponding acid derivative 11b in 90% yield. Similarly, treating cyclic compound 10j with 3.0 equiv. of 2-iodoxybenzoic acid (IBX) in HFIP/H₂O (1:1) at 60 °C gave the amide 11c in 73% yield, while TAM-DG failed to be deprotected under identical conditions. This result clearly emphasized the unique reactivity of ABTA in directed C-H functionalization with both boosted reactivity and easier deprotection.

Figure 2.

Removal of directing group

In conclusion, we have unveiled ABTA as a novel effective directing group in promoting aerobic C-H oxidative olefination. This study confirmed the clearly improved reactivity of ABTA over other ANH auxiliary DGs in promoting Pd catalyzed C-H activation with improved reactivity. Furthermore, the success in macrocyclization bearing amino acid chains allowed the facile synthesis of cyclic peptides in excellent yields. Considering that other DGs lead to poor result under identical conditions, it is expected that this new ABTA directing group strategy would have a promising potential to achieve more challenging chemical transformations via C-H functionalization, which is currently under investigation in our lab.