Download PDF
Synlett 2023; 34(11): 1241-1246
DOI: 10.1055/a-2030-7826
letter
Special Edition Thieme Chemistry Journals Awardees 2022

Organophotoredox-Catalyzed Oxidative C(sp2)–H Alkylation of N-Heteroarenes with Dihydroquinazolinones by C–C Cleavage

Pinku Prasad Mondal
,
Amit Pal
,
Subham Das
,
Sariga Mangalamundackal Vijayan‡
,
Anagha Veluthanath Nair‡
,
Shubham Ojha‡
,
Basudev Sahoo∗

This work was supported by SERB, India (File: SRG/2021/000572). P.P.M. and A.P. thank the Ministry of Education, India for their Prime Minister’s Research Fellowship. S.D. thanks UGC, India for research fellowship.
› Further Information
 
 PDF Download Permissions and Reprints All articles of this category


Dedicated to Professor Matthias Beller on his 60th birthday

Abstract

We report a visible-light-mediated, organophotoredox-catalyzed, C(sp2)–H alkylation of N-heteroarenes with dihydroquinazolines, prepared from aliphatic ketones, under oxidative conditions. This protocol represents a metal-free approach to the effective construction of C–C bonds through a Minisci-type reaction, formally activating the native C–H bond of the N-heteroarene and an α-C–C bond of a readily available ketone. The mild nature of this method accommodates a wide variety of N-heteroarenes and ketones, tolerating a wide range of functional groups.


#

Key words

photoredox catalysis - heteroarenes - ketone - C–C bond cleavage - alkylation

N-Heteroarenes form the privileged core structures of many naturally occurring alkaloids,[1] pharmaceutical drugs,[2] agrochemicals,[3] and functional materials.[4] Encouraged by the beneficial effects of an enhanced C(sp3) fraction in prodrugs or lead structures for drug discovery,[5] the selective alkylation of N-heteroarenes has recently gained prominence as a method for building up structural complexity in molecular skeletons.[6] In this regard, the catalytic C–H alkylation of electron-deficient N-heteroarenes (the Minisci-type reaction) is typically used for the direct installation of a C(sp3) fragment onto an N-heteroaryl core.[6] Whereas the classic Minisci reaction of N-heteroarenes with aliphatic carboxylic acids has limitations under relatively harsh conditions,[7] the monumental development of visible-light photocatalysis[8] has enabled a broadening of the range of alkyl-radical precursors to include alkyl halides,[9`] [b] [c] alkylborates[9d,e] or alkylsilicates,[9f] alcohols,[9`] [h] [i] aldehydes[9j] [k] aldehyde-derived dihydropyridines,[9l] carboxylic acids[9`] [n] [o] or their redox-active esters,[9`] [q] [r] amine-derived pyridinium salts,[9s] or even native C–H bonds,[9`] [u] [v] [w] [x] [y] among others[9z] for the alkylation of heteroarenes (Scheme [1]).[9]

Zoom Image
Scheme 1 An overview of C(sp2)–H alkylations of N-heteroarenes with alkyl-radical precursors

The ketone group is recognized as a versatile functionality and an integral part of various commodity chemicals, natural products, and pharmaceuticals.[10] However, due to the lack of an effective orbital interaction, α-C–C bond functionalization of aliphatic ketones has remained a thought-provoking task[11] compared with conventional nucleophilic additions or α-C–H functionalizations. Recently, the tactic of converting ketones into proaromatic dihydroquinazolinones and their subsequent aromatization-driven C–C bond cleavage has emerged as a promising technique for the functionalization of ketone C–C bonds.[12] In this context, the Zhu group disclosed a photocatalytic activation of ketone-derived dihydroquinazolinones toward a Giese reaction with activated alkenes.[12a] Furthermore, the groups of Martin and Liao have evaluated such dihydroquinazolinones as efficient alkyl-radical precursors in metallaphotoredox-catalyzed C–C cross-couplings with aryl or alkyl bromides[12b] [c] or electrophilic Togni’s reagent.[12d] We recently reported a thermal activation of dihydroquinazolinones towards catalytic alkylation of native C–H bonds of N-heteroarenes under oxidative metal catalysis.[12e] Here, we report an alternative approach to the oxidative C(sp2)–H alkylation of various N-heteroarenes by using dihydroquinazolinones under visible-light photoredox catalysis. In this protocol, the reaction is performed without any metal catalyst by using an organic dye as a photosensitizer and visible light as a renewable energy source.

As dictated by our previous studies,[12e] at the outset of this investigation we optimized the reaction by treating lepidine (1a) with the cyclohexyl methyl ketone-derived dihydroquinazolinone 2a (2.5 equiv) in the presence of 2,4,5,6-tetrakis(9H-carbazol-9-yl)isophthalonitrile (4CzIPN; 2 mol%) and K2S2O8 (2.0 equiv) in DMF (0.2 M) under visible-light irradiation from blue LEDs (λmax = 456 nm) for 18 hours, affording the alkylated product 3aa in 76% isolated yield (Table [1], entry 1). Surprisingly, the use of rose bengal or eosin Y as an organic dye instead of 4CzIPN gave 3aa in reduced yields (entries 2 and 3). A survey of the sacrificial oxidant showed that (NH4)2S2O8 or N-fluorobenzenesulfonimide (NFSI) gave 3aa in slightly reduced yields, whereas di-tert-butyl peroxide (DTBP) was completely ineffective (entries 4–6). Altering the solvent from DMF to N-methyl-2-pyrrolidone (NMP) or 1,2-dichloroethane (DCE) led to a marked reduction in the yield of 3aa (entries 7 and 8). Furthermore, a 23 W compact fluorescent lamp (CFL) was found to be ineffective in this transformation (entry 9). Moreover, the reduction of the amount of 2a or the 4CzIPN reduced the efficiency of this transformation (entries 10 and 11). Control experiments confirmed the critical roles of 4CzIPN, K2S2O8, and light in achieving the desired reactivity (entries 12–14).

Table 1 Optimization Studiesa

Entry

Deviation from standard conditions

Yield (%) of 3aa

1

None

76

2

rose bengal instead of 4CzIPN

49

3

eosin Y instead of 4CzIPN

40

4

(NH4)2S2O8 instead of K2S2O8

67

5

NFSI instead of K2S2O8

71

6

DTBP instead of K2S2O8

–

7

NMP instead of DMF

<5

8

DCE instead of DMF

<5

9

23 W CFL instead of blue LEDs

–

10

2a (2.0 equiv) and K2S2O8 (2.0 equiv)

62

11

4CzIPN (1 mol%)

53

12

no 4CzIPN

12

13

no K2S2O8

–

14

no blue light

–

a Standard conditions: 1a (0.10 mmol), 2a (0.25 mmol), 4CzIPN (2 mol%), K2S2O8 (0.20 mmol), DMF (0.5 mL), rt, 18 h, blue LEDs.

b Isolated yield.

Having identified the optimal conditions, we set out to evaluate the scope of the dihydroquinazolinone in reactions with lepidine (1a; Scheme [2]). Preferential cleavage of the C–C bond of the dihydroquinazolinone is dictated by the bond strength and stability of the incipient radical.[12f] Initially, an acyclic isopropyl methyl ketone-derived dihydroquinazolinone 2b was found to be a competent partner, delivering the alkylated product 3ab in 64% yield. Furthermore, the dihydroquinazolinone 2c, featuring a heterocyclic tetrahydropyran moiety, exhibited the desired reactivity, furnishing the product 3ac in 55% yield. Likewise, the dihydroquinazolinone 2d furnished the expected product 3ad in 47% yield. Pleasingly, the dihydroquinazolinone 2e, prepared from benzyl phenyl ketone, served as a viable substrate in this transformation, providing the product 3ae in 41% yield through preferential cleavage of the C(sp3)–C(sp3) bond over the C(sp3)–C(sp2) bond. Gratifyingly, the dihydroquinazolinones 2f and 2g, synthesized from 1-(1-arylcycloalkyl)ethan-1-ones, were effective substrates under the optimal conditions, affording the products 3af and 3ag in yields of 77 and 82%, respectively; these products resemble the lead structures of a selectin inhibitor.[13]. Notably, the pivalaldehyde-derived dihydroquinazolinone 2h exhibited competency in this transformation, delivering the desired product 3ah in 68% yield, indicating a proclivity for the cleavage of the C–C bond in preference to the C–H bond. In view of the scalability and practicability of this protocol, the dihydroquinazolinone 2b was synthesized in situ from cyclohexyl methyl ketone and anthranilamide and used directly, without purification, in a C–H alkylation of lepidine (1.0 mmol scale) under the standard conditions, giving product 3aa in 65% yield (Scheme [2]).

Zoom Image
Scheme 2 Reaction conditions: 1a (0.2 mmol), 2a–h (0.5 mmol), 4CzIPN (2 mol%), K2S2O8 (0.4 mmol), DMF (1 mL), rt, 18 h, blue LEDs. Isolated yields are reported. a R2 = Me. b R2 = Ph. c R2 = H.

Next, we turned our attention to a study of the scope of the N-heteroarene in reactions with the dihydroquinazoline 2a (Scheme [3]). As predicted from their Fukui indices, several N-heteroarenes, including pyridine and quinoline, underwent C–H alkylations at a C-2 or C-4 site.[] Quinoline 1b, containing a p-(trifluoromethyl)phenyl group at the C-4 position, was effectively alkylated at the C-2 site to give product 3ba in 64% yield. Interestingly, the ester (1c) and amide (1d) derivatives of the quinoline-2-carboxylic acid obtained from l-menthol and piperidine, were also viable substrates, affording the products 3ca and 3da in yields of 52 and 51%, respectively, upon C–H alkylation at the C-4 site. Gratifyingly, the benzoxazole 1e and the N-benzyl-protected benzimidazole 1f exhibited the desired reactivity, delivering products 3ea and 3fa in yields of 59 and 41%, respectively. As expected, the isoquinolines 1g and 1h with phenyl and p-anisidinyl substituents, respectively, at the C-6 position participated successfully in this transformation, affording the corresponding products 3ga and 3ha, in yields of 75% and 76% respectively. Notably, 1-methylquinolin-2(1H)-one (1i) was found to be an effective substrate under the optimal conditions, providing product 3ia in 39% yield. To our delight, 4-(trifluoromethyl)pyridine (1j) was also a successful substrate under this optimized protocol for C–H alkylation, delivering the dialkylated product 3ja in 40% yield. Encouraged by the critical role of N-donor ligands (e.g., bipyridines) in transition-metal catalysis, we applied the current protocol to the bipyridines 1k and 1l, featuring an electron-rich tert-butyl and an electron-withdrawing ester functionality, respectively, and we obtained the corresponding structurally modulated bipyridines as separable monoalkylated [3ka (51%) and 3la (27%)] and dialkylated products [3ka′ (9%) + 3la′ (7%)].

Zoom Image
Scheme 3 Reaction conditions: 1a–l (0.2 mmol), 2a (0.5 mmol), 4CzIPN (2 mol%), K2S2O8 (0.4 mmol), DMF (1 mL), rt, 18 h, blue LEDs. Isolated yields are reported.

To elucidate the mechanistic intricacies of this photochemical transformation, we conducted some preliminary mechanistic experiments and spectroscopic studies. Our initial control experiments indicated the critical roles of the 4CzIPN photocatalyst, light, and the K2S2O8 oxidant in this transformation (Table [1], entries 11–13). In a radical-inhibition experiment, when the standard reaction was executed in the presence of TEMPO as a radical scavenger, the formation of product 3aa was significantly suppressed, and an adduct between the cyclohexyl radical and the TEMPO radical was detected by HRMS analysis, suggesting the involvement of a radical intermediate (Scheme [4a]). Furthermore, the reaction was also inhibited when it was carried out in the presence of 1,1-diphenylethene (DPE) as a radical-trapping agent (Scheme [4b]). Luminescence-quenching plots and a Stern–Volmer plot suggested that the 4CzIPN excited state is quenched by the dihydroquinazoline 2a,[12b] rather than by lepidine (1a) or K2S2O8 (Scheme [4c]).

Zoom Image
Scheme 4 (a) Radical-trapping experiment with TEMPO. (b) Radical-trapping experiment with DPE. (c) Luminescence-quenching studies of 4CzIPN with 1a, 2a, and K2S2O8, and a Stern–Volmer plot.

On the basis of the preliminary mechanistic studies and previous reports,[9] we suggest the tentative mechanism shown in Scheme [5]. Initially, photoexcited 4CzIPN* is generated from ground-state 4CzIPN by visible-light irradiation. Quenching of 4CzIPN* (E red = +1.35 V vs SCE)[14] with 2a (E ox = +0.98 V vs SCE)[15] gives the reduced photocatalyst 4CzIPN∙– and the intermediate Int-I, which rapidly undergoes C–C bond cleavage, driven by aromatization, to liberate a nascent cyclohexyl radical and the quinazolinone 6 as a byproduct.[9] Subsequently, the addition of the cyclohexyl radical to the activated lepidine generates intermediate Int-II. Concurrently, the single-electron reduction of S2O8 2– by 4CzIPN∙– regenerates 4CzIPN in the ground state and SO4 2– along with an SO4 ∙– radical anion that engages in hydrogen atom transfer (HAT) with Int-II,[9w] [12e] furnishing the final product 3aa.

Zoom Image
Scheme 5 Mechanistic rationale

In summary, we have developed a mild protocol for C(sp2)–H alkylation of a variety of N-heteroaromatic compounds with ketone-derived dihydroquinazolinones under a visible-light organophotoredox catalytic regime.[16] Activation of the N-heteroarene C–H bond and the ketone C–C bond was amalgamated under net oxidative conditions to accomplish the targeted C–C bond formation through a Minisci reaction. Due to mild nature, this methodology is applicable to a broad range of N-heteroarenes and ketones, as well as tolerating distinct functionalities and demonstrating scalability and practicability.


#

Conflict of Interest

The authors declare no conflict of interest.

Acknowledgment

Infrastructural support from IISER Thiruvananthapuram is gratefully acknowledged.

Supporting Information

    Supporting information for this article is available online at https://doi-org.accesdistant.sorbonne-universite.fr/10.1055/a-2030-7826.
  • Supporting Information

Corresponding Author

Basudev Sahoo
School of Chemistry, Indian Institute of Science Education and Research Thiruvananthapuram
Maruthamala PO, Vithura, Thiruvananthapuram 695551, Kerala
India   

Publication History

Received: 24 December 2022

Accepted after revision: 08 February 2023

Accepted Manuscript online:
08 February 2023

Article published online:
09 March 2023

© 2023. Thieme. All rights reserved

Georg Thieme Verlag KG
Rüdigerstraße 14, 70469 Stuttgart, Germany


   Permissions and Reprints All articles of this category
Zoom Image
Scheme 1 An overview of C(sp2)–H alkylations of N-heteroarenes with alkyl-radical precursors
Zoom Image
Scheme 2 Reaction conditions: 1a (0.2 mmol), 2a–h (0.5 mmol), 4CzIPN (2 mol%), K2S2O8 (0.4 mmol), DMF (1 mL), rt, 18 h, blue LEDs. Isolated yields are reported. a R2 = Me. b R2 = Ph. c R2 = H.
Zoom Image
Scheme 3 Reaction conditions: 1a–l (0.2 mmol), 2a (0.5 mmol), 4CzIPN (2 mol%), K2S2O8 (0.4 mmol), DMF (1 mL), rt, 18 h, blue LEDs. Isolated yields are reported.
Zoom Image
Scheme 4 (a) Radical-trapping experiment with TEMPO. (b) Radical-trapping experiment with DPE. (c) Luminescence-quenching studies of 4CzIPN with 1a, 2a, and K2S2O8, and a Stern–Volmer plot.
Zoom Image
Scheme 5 Mechanistic rationale