Volume 30, Issue 13 e202303130
Research Article
Open Access

H2O ⋅ B(C6F5)3-Catalyzed para-Alkylation of Anilines with Alkenes Applied to Late-Stage Functionalization of Non-Steroidal Anti-Inflammatory Drugs

Dr. Laura Winfrey

Dr. Laura Winfrey

School of Chemistry, University of Leicester, Leicester, LE1 7RH United Kingdom

Search for more papers by this author
Lei Yun

Lei Yun

School of Chemistry, University of Leicester, Leicester, LE1 7RH United Kingdom

Search for more papers by this author
Ginevra Passeri

Ginevra Passeri

School of Chemistry, University of Leicester, Leicester, LE1 7RH United Kingdom

Search for more papers by this author
Dr. Kogularamanan Suntharalingam

Corresponding Author

Dr. Kogularamanan Suntharalingam

School of Chemistry, University of Leicester, Leicester, LE1 7RH United Kingdom

Search for more papers by this author
Dr. Alexander P. Pulis

Corresponding Author

Dr. Alexander P. Pulis

School of Chemistry, University of Leicester, Leicester, LE1 7RH United Kingdom

Search for more papers by this author
First published: 15 January 2024

Graphical Abstract

We describe a new method for the selective para-alkylation of anilines. The approach has been applied to the late-stage functionalization of non-steroidal anti-inflammatory drugs (NSAIDs) and has allowed the identification of new structures with improved anti-inflammatory properties.

Abstract

Anilines are core motifs in a variety of important molecules including medicines, materials and agrochemicals. We report a straightforward procedure that allows access to new chemical space of anilines via their para-C−H alkylation. The method utilizes commercially available catalytic H2O ⋅ B(C6F5)3 and is highly selective for para-C-alkylation (over N-alkylation and ortho-C-alkylation) of anilines, with a wide scope in both the aniline substrates and alkene coupling partners. Readily available alkenes are used, and include new classes of alkene for the first time. The mild reaction conditions have allowed the procedure to be applied to the late-stage-functionalization of non-steroidal anti-inflammatory drugs (NSAIDs), including fenamic acids and diclofenac. The formed novel NSAID derivatives display improved anti-inflammatory properties over the parent NSAID structure.

Introduction

Anilines are N-aryl substituted amines that are commonly found in far reaching and important applications such as pharmaceuticals, agrochemicals, antioxidants and precursors for materials.1 In the drug discovery setting, N-aryl- or N,N-diarylamines represent around a third of compounds used,2 and the aniline moiety is found in a quarter of the top 200 small molecule drugs by sales in 2022, of which para-C functionalized anilines are well represented (Scheme 1a).3 In addition, the aniline core constitutes several non-steroidal anti-inflammatory drugs (NSAIDs) used to treat pain and inflammation in humans and animals, and include those based on fenamic acids (e. g. mefenamic, flufenamic, tolfenamic and meclofenamic acid), and from the arylacetic acids class of NSAIDs (e. g. diclofenac, brand name Voltaren/Voltarol) (Scheme 1a).4 These NSAIDs can have serious side effects including disruption of the gastrointestinal and cardiovascular systems. Exploration of new chemical space around these NSAIDs, as well as other aniline bioactive compounds, is therefore critical to the development of new and more selective medicines.

Details are in the caption following the image

Importance of anilines and their C−H alkylation using alkenes.

Synthetic methodologies capable of late-stage functionalization (LSF) – the selective and direct transformation of a complex molecule – are highly prized in drug discovery.5 They offer an efficient approach to exploring new chemical space from known biologically active structures without resorting to inefficient pre-functionalization of the parent structure. The C−H functionalization of anilines6 offers the potential to realise LSF methodologies, particularly for the introduction of heteroatoms and sp2- or sp-hybridized carbon atoms, yet the C−H alkylation of anilines remains challenging.

Classic Friedel-Crafts alkylation of anilines (using alkyl halides and Lewis acid catalysts) is limited due to competitive Lewis acid-nitrogen coordination.7 Alkylation of anilines with alkenes (also know as alkene hydroarylation with anilines) offers an atom economical approach and has been reported using catalysts derived from transition metals,8 main group elements9 and Brønsted acids.10, 11 For example, Colomer reported the para-alkylation of anilines with styrene derivatives using NaOAc(cat) and hexafluoroisopropanol (HFIP) as the solvent.10b Gandon, Lebœuf and coworkers discovered a Ca(NTf2)2 catalyzed process, using HFIP as the solvent, which delivered ortho-alkylated anilines from primary (ArNH2) and secondary (ArNH) anilines, and para-alkylated anilines from tertiary (ArNR2) anilines.10c However, all aniline alkylation methods using alkenes possess at least some of the following limitations: 1) are limited to a subset of alkenes (typically styrene derivatives); 2) do not encompass all aniline classes (i. e. primary [ArNH2], secondary [ArNHR] and tertiary [ArNR2] anilines); 3) have poor selectivity between ortho and para functionalization of the aniline; 4) in the case of NH and NH2 anilines, have low selectivity between N-alkylation (alkene hydroamination) and C-alkylation (alkene hydroarylation); 5) require high temperatures, and; 6) use non-commercially available catalysts (Scheme 1b). Therefore, LSF via C−H alkylation of anilines is currently underdeveloped.

Herein we report the H2O ⋅ B(C6F5)3 catalyzed para-selective C−H alkylation of anilines with alkenes (Scheme 1c). The catalyst is commercially available and uniquely provides a weakly coordinating counterion that is essential for reaction efficiency. The mild conditions allow the process to have wide scope in both the aniline and alkene coupling partners, which includes primary, secondary and tertiary anilines, as well as aryl alkenes, aliphatic alkenes, and for the first time, allyl silanes and enamides. The use of enamides has delivered valuable N-benzyl amides directly from anilines for the first time. The methodology is also suitable for the LSF of aniline-based drugs. We demonstrate that fenamic acid-based NSAIDs and diclofenac are efficiently derivatized via this Friedel-Crafts-type process to access new chemical space. Subsequent testing of the new NSAID derivatives for COX-2 inhibitory effects revealed improved anti-inflammatory properties

Results and Discussion

Optimization

We began our investigation by considering Brønsted acids that produce weakly coordinating conjugate bases.12 We settled upon H2O ⋅ B(C6F5)313, 14 as a potential catalyst because it is commercially available.15 Due to the stability of B(C6F5)3 with respect to oxidation and hydrolytic bond cleavage, it can be weighed on the open bench. Recently, Qu, Chatterjee and coworkers reported a H2O ⋅ B(C6F5)3 catalyzed para-homoallylation of anilines using dienes.10e, 16, 17 However, a general study of H2O ⋅ B(C6F5)3 catalyzed aniline alkylation beyond dienes has not been reported. N-Benzyl aniline (1 a) was reacted with 1-methyl styrene (2 a) under a range of conditions (Table 1 and SI). Initial screening of solvent, temperature, time and stoichiometry identified acetonitrile at 100 °C using H2O ⋅ B(C6F5)3 (10 mol %) as a starting point for further optimization studies (Entry 1).18

Table 1. Reaction optimization using N-benzylaniline (1 a) and 1-methylstyrene (2 a).[a]

image

Entry

Catalyst

Solvent

Additive (equiv.)

Yield 3 a (%)[b]

1

H2O ⋅ B(C6F5)3

CH3CN

n/a

35

2

HCl

CH3CN

n/a

12

3

TFA

CH3CN

n/a

22

4

pTsOH ⋅ H2O

CH3CN

n/a

23

5

TfOH

CH3CN

n/a

71

6

HNTf2

CH3CN

n/a

43

7

H2O ⋅ B(C6F5)3

CH3CO2H

n/a

37

8

H2O ⋅ B(C6F5)3

HFIP

n/a

45

9

H2O ⋅ B(C6F5)3

CF3CH2OH

n/a

28

10

H2O ⋅ B(C6F5)3

CH3CN

CH3CO2H (1.5)

66

11

H2O ⋅ B(C6F5)3

CH3CN

HFIP (2)

64

12

H2O ⋅ B(C6F5)3

CH3CN

CF3CH2OH (10)

75

13

CF3CH2OH ⋅ B(C6F5)3

CH3CN

CF3CH2OH (10)

53

14

H2O ⋅ B(C6F5)3

CH3NO2

CF3CH2OH (10)

63

15

none

CH3CN

CF3CH2OH (10)

0

  • [a] Reactions were performed using 1 a (0.20 mmol) and 2 a (0.44 mmol). [b] Yields determined by 1H NMR spectroscopy using an internal standard.

We also screened other commercially available Brønsted acids for this Friedel-Crafts-type process (Entries 2–6). Hydrochloric acid, trifluoroacetic acid (TFA), para-toluenesulfonic acid (TsOH) were found to be inferior acids to H2O ⋅ B(C6F5)3 (12–23 %). Whereas trifluoromethanesulfonic acid (TfOH) and bis(trifluoromethane)sulfonimide (HNTf2) gave 3 a in 71 % and 43 % yield respectively. The high reaction efficiency with TfOH was surprising given that TfOH was previously used to catalyse the alkylation of primary para-methyl anilines with styrenes at 160 °C.10i However, when using either TfOH or HNTf2 as the catalyst, we found the reaction scope was highly limited with respect to both the alkene and aniline,19 and returned to H2O ⋅ B(C6F5)3 for further investigation.

We then investigated different proton sources by using various protic solvents in conjunction with H2O ⋅ B(C6F5)3, including alcohols, carboxylic acids and amines (Entries 7–9 and SI). Use of acetic acid and 1,1,1,3,3,3-hexafluoroisopropanol (HFIP) as the solvent lead to marginally increased yields of 3 a (37 % and 45 % resp.), whereas 2,2,2-trifluoroethanol (CF3CH2OH) decreased the yield (28 %).

Using these protic solvents as additives/co-solvents with acetonitrile as the bulk solvent led to significantly increased yields, with 10 equivalents of CF3CH2OH being optimal and giving 3 a in 75 % yield (Entries 10–12 and SI). When CF3CH2OH ⋅ B(C6F5)3 was used in place of H2O ⋅ B(C6F5)3, the yield of 3 a was reduced to 53 % (Entries 12 versus 13). As for TfOH and HNTf2, the scope of the reaction with respect to other alkenes and anilines when catalysed by CF3CH2OH ⋅ B(C6F5)3 was limited.19 Control experiments were performed throughout and proved that H2O ⋅ B(C6F5)3 was critical to the reaction success (Entry 15 and SI).

para-Alkylation of Anilines: Aniline Scope

We explored the scope of the H2O ⋅ B(C6F5)3 catalyzed alkylation with respect to the aniline input 1 using 1-methyl styrene (2 a) as the alkylating agent (Scheme 2). In contrast to other methods, we found that the process encompassed primary, secondary, and tertiary anilines and delivered products 3 in generally good to excellent yields (up to 99 %). Complete selectivity for C-alkylation products was observed in all cases (see mechanistic discussion). Highlights from the aniline scope include important saturated heterocycles, N-aryl pyrrolidines (3 e, 3 f, 3 h) and morpholine (3 g), substituents in ortho (3 hv) and meta (3 w) positions of the aniline, and versatile functional groups such as halides (3 h, 3 j, 3 m, 3 q, 3 u, 3 w, 3 aa, 3 af), trifluoromethyl (3 r), ethers (3 k, 3 n, 3 ae), ketone (3 o), carboxylic acids (3 s, 3 am) and nitro (3 z, 3 ad). Of note, the reaction tolerated electron withdrawing groups (cf. 3 o, 3 r, 3 s, 3 z, 3 am, and Scheme 5 below) on the aniline ring undergoing functionalization. The reaction was selective for para-functionalization of the aniline, as demonstrated with 3 ac and 3 ae: despite ortho positions being available in the more electron rich ring, functionalization occurred on the less electron rich ring in the para position.

Details are in the caption following the image

H2O ⋅ B(C6F5)3-catalyzed para-alkylation of anilines: Aniline scope. Reactions were performed using aniline 1 (0.20 mmol), 1-methyl styrene 2 a (0.44 mmol), H2O ⋅ B(C6F5)3 (10 mol %) in 30 % CF3CH2OH/co-solvent (0.5 mL) at 100 °C. Yields determined after 1H NMR spectrum analysis of the crude reaction mixture with an internal standard. Analytically pure samples were obtained after purification and used for characterization. [a] Using MeNO2 as co-solvent. [b] Using MeCN as co-solvent. [c] ortho- and para-bisalkylation was observed in 12 % yield. [d] ortho- and para-bisalkylation was observed in 22 % yield. [e] Using 4 equiv. of alkene. [f] Using alkene 2 a (1 equiv.) and Ph2NH (2 equiv.) for 1 h.

Dialkylated aniline 3 al is an efficient antioxidant for polymeric materials.1a We were pleased to find that dialkylated products 3 al and 3 aj, and novel dialkylated anilines 3 ak, 3 am and 3 ay (Scheme 2 and 3) were formed from the corresponding diaryl amines using the H2O ⋅ B(C6F5)3 catalyzed procedure. Monoalkylated products could be selectively formed by adjusting the stoichiometry so that the alkene was the limiting reagent (2 equiv. of aniline relative to alkene, e. g. 99 % yield for 3 ab). We also employed less nucleophilic anilides which are absent from previously reported para-aniline alkylation procedures.20 In this case, para-substituted product 3 an was formed, albeit in lower yields (34 %).

Details are in the caption following the image

H2O ⋅ B(C6F5)3-catalyzed para-alkylation of anilines: Alkene scope. Reactions were performed using aniline 1 b (0.20 mmol), alkenes 2, 4, 5 or 6 (0.44 mmol), H2O ⋅ B(C6F5)3 (10 mol %) in DCE (0.5 mL) at 60 °C. Yields determined after 1H NMR spectrum analysis of the crude reaction mixture with an internal standard. Analytically pure samples were obtained after purification and used for characterization. [a] Using N-methyl-N-phenylaniline. [b] Using 2-methylhex-1-ene. [c] Using 2,4,4-trimethylpent-1-ene. [d] Using methylenecyclohexane. [e] Using 1-methylcyclopent-1-ene. [f] 35 % of protodesilylation product also formed. [g] Using 6 (5 equiv.) at 80 °C. [h] 1 : 1 dr.

In all cases we did not observe products of α-nitrogen C−H hydride abstraction,15a, 15b presumably due to the H2O quenching the Lewis acidity at boron.

para-Alkylation of Anilines: Alkene Scope

We also explored different alkene based alkylating agents in the H2O ⋅ B(C6F5)3 catalyzed para-alkylation of anilines and found the scope to be broad (Scheme 3). We found that lower temperature (60 °C) and the use of DCE as solvent was a more general set of conditions when investigating the alkene scope. A variety of aromatic alkenes 2 were efficient and gave products containing tertiary (3 ao, 3 ap, 3 ay) and quaternary carbons (Scheme 2 and 3aq-ax), and including moieties such as dihydroindene (3 ao), thiophene (3 as), xanthene (3 av), thioxanthene (3 aw), and fluorene (3 ax). Aliphatic alkenes 4 that contained 1,1-disubstitution, including open chain (7 a, 7 b) and cyclic alkenes (7 c, 7 d), were also effective alkylating agents. 2-Hexene, cyclohexene, cyclooctene and styrene did not produce the desired products.

We also explored allyl silanes 5 and enamides 6 as they have not previously been reported as alkene coupling partners in aniline alkylation. The use of allyl silanes is challenging due to their propensity to protodesilylate under acidic conditions.21 Nevertheless, allyl silanes including trimethyl- (8 a), trisopropyl- (8 b), triphenyl- (8 c) and triethoxy- (8 d) silyl groups performed well and delivered the corresponding silyl alkane para-functionalized anilines (61–80 % yield).

The use of enamides in Friedel-Crafts-type arene alkylation will deliver valuable N-benzyl amides, yet their use is challenging as oligomerization side reactions can occur. Only very recently have Lebœuf and Moran reported a general method for coupling arenes with enamides.22 They elegantly showed that HFIP-iminium adducts form and serve as a reservoir for the desired iminium, thus reducing unwanted oligomerizations. However, anilines were absent from the scope of their process. We were pleased to find that enamides 6 were successfully utilized in our H2O ⋅ B(C6F5)3-catalyzed para-alkylation of anilines, and delivered valuable N-benzyl amides (9 ae) directly from anilines for the first time.

Mechanistic Studies and Proposed Mechanism

We investigated some of the mechanistic features of the H2O ⋅ B(C6F5)3 catalyzed alkylation of anilines (Scheme 4). The pKa of H2O ⋅ B(C6F5)3 is similar to that of HCl in acetonitrile (8.4 versus 8.5 resp.).23 In acetonitrile, typical pKa’s of the conjugate acids of mono aryl amines and diaryl amines are approx. 12 and 6, respectively.24 Therefore it is reasonable to expect that some aniline starting materials 1 or their corresponding products 3 will be protonated under the reaction conditions. To determine if a protonated aniline could be a possible intermediate or catalyst in the H2O ⋅ B(C6F5)3 catalyzed alkylation, we synthesized ammonium salt, [PhNMe2H][BArCl] (11), where the weakly coordinating anion BArCl serves as a surrogate for the conjugate base of H2O ⋅ B(C6F5)3 (BArCl=[B(3,5-Cl2-C6H3)4], Scheme 4a).25 The alkylation of 1 a with 2 a using either H2O ⋅ B(C6F5)3 or [PhNMe2H][BArCl] (11) proceeded with virtually the same efficiency (75 % vs. 71 % of 3 a resp.). The same reaction using HCl or the analogous chloride salt, [PhNMe2H][Cl] (10), provided 3 a with significantly lower yields (35 % and 39 %). Under the optimised conditions for the aniline scope, CF3CH2OH might displace water from H2O ⋅ B(C6F5)3. When CF3CH2OH ⋅ B(C6F5)3 adduct (prepared in situ from water-free B(C6F5)3 and CF3CH2OH) was used as the catalyst, 3 a was formed in 53 % yield (Table 1 and Scheme 4a). In addition, the reaction using H2O ⋅ BPh3 as a catalyst did not provide the intended product and returned only quantitative starting materials 1 a and 2 a. These results underscore the critical nature of the weakly coordinating anion in the aniline para-C-H alkylation method described herein.

Details are in the caption following the image

Mechanistic investigation and proposed catalytic cycle. [a] Using 4 M HCl in dioxane. BArCl=[B(3,5-Cl2−C6H3)4].

We also investigated if N-alkylated anilines were formed in the reaction via alkene hydroamination with NH bearing anilines (cf. 12), and if they could be converted to the corresponding para-alkylated anilines 3 under the reaction conditions via the Hoffmann-Martius rearrangement (Scheme 4b).10a, 26 N-Alkyl anilines 12 a and 12 b (the theoretical products of 1-methyl styrene (2 a) hydroamination with Ph2NH (1 c) and PhNH2 (1 d) resp.) were synthesized independently. Anilines 12 a and 12 b were subjected to the reaction conditions and gave the corresponding para-alkylated products 3 ab and 3 i in moderate yields. In situ NMR spectra analysis of the reaction between alkene 2 a and PhNH2 (1 d) revealed the initial formation of N-alkylation product 12 b which was smoothly converted to the intended para-alkylated product 3 i (see SI). The formation of analogous hydroamination intermediates (cf. 12) in other alkene classes used herein (i. e. 4, 5 and 6) will require further investigation but were not observed by in situ NMR spectra analysis of the reaction mixtures.

Based on these experiments, we propose that the alkene is protonated to form carbocation 13 that bears a weakly coordinating anion (Scheme 4c). The exact nature of the acid is unknown and is likely dependant on the basicity of the aniline. In addition to H2O ⋅ B(C6F5)3, we propose that further hydrated borane, nH2O ⋅ B(C6F5)3,23 CF3CH2OH ⋅ B(C6F5)3 adduct, ammonium borate salts 15, or primary or secondary aniline B(C6F5)3 adducts10e 16 might protonate the alkene. The carbocation then engages in a typical electrophilic aromatic substitution with aniline 1 to form the desired products. Whilst hydroamination products 12 may form in the reaction between aryl alkenes 2 and primary and secondary anilines, their formation is reversible and they are efficiently converted to the desired para-alkylation products 3 under the reaction conditions, accounting for the complete C- vs N-selectivity observed in the final products.

Applications in Late-Stage Functionalization of NSAIDs

The fenamic acid family of NSAIDs, including mefenamic, flufenamic, tolfenamic and meclofenamic acid, as well as diclofenac (Voltaren/Voltarol), are used to treat pain and inflammation in humans and animals.4 Given that NH and CO2H must be conserved to retain biological activity,4 a method capable of para-C−H alkylation of the parent drug is a potentially attractive approach to explore new chemical space from the established bioactive molecules. para-Alkylation may improve drug bioavailability and reduce the required dosage by preventing metabolism via known para-oxidation of the anthranilic ring.4, 27

In addition, para-alkylation may improve COX-2 selectivity and therefore reduce the serious side effects associated with these NSAIDs: para-methylation is essential for COX-2 selectivity in a related NSAID, lumiracoxib.28 However, the para-C−H functionalization of the aforementioned NSAIDs has rarely been achieved and has only been reported in isolated examples.29 To our knowledge, the effect of para-alkyl substituents of these fenamic acid based NSAIDs on anti-inflammatory properties has not previously been reported.

The utility of the H2O ⋅ B(C6F5)3 catalyzed para-alkylation of anilines was demonstrated in the late-stage functionalization of aniline based NSAIDs (Scheme 5). The H2O ⋅ B(C6F5)3 catalyzed procedure successfully coupled aryl alkene 2 a and the NSAIDs mefenamic, flufenamic, tolfenamic and meclofenamic acid, and diclofenac to form the corresponding para-functionalized products 18 ae (91–23 %). Importantly, the mild conditions of the process meant that decarboxylation, which has previously been observed with anthranilic acid derivatives (cf. 18 ad), did not occur.30 We also prepared the mefenamic acid derivative 18 a on a preparative scale (1.63 g) where workup and purification simply involved removal of the volatiles from the reaction mixture followed by recrystallization.

Details are in the caption following the image

Late-stage functionalization of NSAIDs. Reactions were performed using NSAID (0.20 mmol), alkene 2 a (0.44 mmol), H2O ⋅ B(C6F5)3 (10 mol %) in 30 % CF3CH2OH/CH3CN at 100 °C. Yields determined after 1H NMR spectrum analysis of the crude reaction mixture with an internal standard. Yields in parenthesis are after isolation from preparative scale reactions.

To demonstrate the ease with which the method could explore biological activity through late-stage functionalization, we tested the ability of the novel compounds 18 ae to downregulate and inhibit cyclooxygenase-2 (COX-2). An established osteosarcoma stem cell (OSC) model was used to determine the COX-2 inhibitory effect of 18 ae.31 Prior to measuring the effect of 18 ae on COX-2, their toxicity towards the OSC line, U2OS-MTX was determined using the MTT assay. The dose-response curves (Figure S1) give an indication on the non-lethal concentration range to conduct the COX-2 associated studies (the IC50 values for 18 ae were >28 μM). U2OS-MTX cells pre-treated with lipopolysaccharide (LPS) (2.5 μM for 24 h), to increase basal COX-2 levels, were treated with 18 ae or their corresponding parent NSAID analogues at a non-lethal dose (20 μM for 48 h), and the COX-2 expression was determined by flow cytometry. COX-2 expression significantly decreased upon treatment with 18 ae, with 18 a, 18 c, and 18 e outperforming or matching their corresponding parent NSAID analogues (Figure 1ac and S2).

Details are in the caption following the image

COX-2 inhibitory effects of novel NSAID derivatives and their parent NSAIDs. a)–c) Representative histograms displaying the green fluorescence emitted by anti-COX-2 Alexa Fluor 488 nm antibody-stained HMLER-shEcad cells treated with LPS (2.5 μM) for 24 h followed by 48 h in fresh media (red), or media containing parent NSAID (20 μM, blue) or NSAID derivative (20 μM, orange) [a) mefenamic acid versus 18 a (20 μM, orange); b) meclofenamic acid versus 18 c; c) diclofenac versus 18 e]. d) Representative bar chart for the inhibition of COX-2 activity by 18 a, 18 c, 18 e and their corresponding parent NSAID analogues at specific concentrations (15.6 and 62.5 μM).

The COX-2 inhibitory properties of 18 a, 18 c, and 18 e were further investigated using an enzyme immunoassay. The para-functionalized derivative 18 c inhibited COX-2 activity to a better extent than its corresponding parent NSAID, meclofenamic acid (Figure 1d). Contrastingly, 18 a and 18e displayed reduced COX-2 inhibition compared to their corresponding parent NSAIDs, mefenamic acid and diclofenac, respectively. Collectively, the flow cytometry and enzyme immunoassay studies show that para-functionalization of mefenamic acid and diclofenac (to generate 18 a and 18 e resp.) improves not only their ability to downregulate COX-2 expression but also to directly inhibit COX-2 activity. Therefore H2O ⋅ B(C6F5)3 catalyzed late-stage para-C−H functionalization of NSAIDs could yield derivatives with improved anti-inflammatory properties.

Conclusions

In summary, we have developed a new operationally simple approach for the C−H alkylation of anilines using alkene alkylating agents. We utilise H2O ⋅ B(C6F5)3 as an air stable, commercially available catalyst. H2O ⋅ B(C6F5)3 uniquely offers easy access to a weakly coordinating conjugate base which is critical for the success of the Friedel-Crafts-type alkylation process. The reaction has excellent selectivity for para-alkylation of the anilines, and neither N- or ortho-alkylation is observed in the final products. Unlike most reported methods, the H2O ⋅ B(C6F5)3 catalyzed approach encompasses all aniline classes (primary, secondary and tertiary anilines). The mild conditions allow a variety of aryl and aliphatic alkenes to be used, and it also enables the use of allyl silanes and enamides as new classes of aniline alkylating agents for the first time. We have shown that the method is capable of mediating late-stage C−H functionalization and has allowed the identification of novel fenamic acid based NSAID derivatives with improved anti-inflammatory properties.

Experimental

General procedure

A flask was charged with H2O ⋅ B(C6F5)3 (10.2 mg, 20.0 μmol, as received from the supplier and weighed in air), aniline 1 (0.20 mmol), alkene 2, 4, 5 or 6 (0.46 mmol) and solvent (0.5 ml) under a nitrogen atmosphere. The flask was stirred at the stated temperature for 21 h. After cooling to ambient temperature, sat. NaHCO3(aq) (1.5 ml) was added and the mixture vigorously stirred. The organic phase separated and the aqueous phase extracted with CH2Cl2 (3×2 mL). The combined organic phases were dried over MgSO4, filtered, and the solvent removed in vacuo. An NMR spectroscopic yield was obtained using 1,3,5-trimethoxybenzene as an internal standard. Pure products were obtained after purification by flash column chromatography and were characterised to confirm identity. See Supporting Information for full experimental details.

Supporting Information

The authors have cited additional references within the Supporting Information.32-56, 57-59

Author Contributions

LW, LY and AP designed and performed the synthetic chemistry. GP and KS designed and performed the biological studies. All authors contributed to the preparation of the manuscript

Acknowledgments

We thank the School of Chemistry, University of Leicester for their generous support. LBW thanks EPSRC for PhD funding. LY thanks the China Scholarship Council for PhD funding. APP thanks the Royal Society for a Research Grant (RGS\R1\191082), and EPSRC for grants EP/W02151X/1 and EP/Y00146X/1.

    Conflict of interests

    The authors declare no conflict of interest.

    Data Availability Statement

    The data that support the findings of this study are available in the supplementary material of this article.