Download PDF
Synlett 2019; 30(17): 1995-1999
DOI: 10.1055/s-0037-1611974
cluster
© Georg Thieme Verlag Stuttgart · New York

Thiolate-Initiated Synthesis of Dibenzothiophenes from 2,2′-Bis(methylthio)-1,1′-Biaryl Derivatives through Cleavage of Two Carbon–Sulfur Bonds

Yoshihiro Masuya
,
Yuki Kawashima
,
Takuya Kodama
,
Naoto Chatani*
,
Mamoru Tobisu  *
Department of Applied Chemistry, Graduate School of Engineering, Osaka University, Suita, Osaka 565-0871, Japan   Email: chatani@chem.eng.osaka-u.ac.jp   Email: tobisu@chem.eng.osaka-u.ac.jp
› Author Affiliations

This work was supported by Grant-in-Aid for Scientific Research (18H01978) and Scientific Research on Innovative Area "Precisely Designed Catalysts with Customized Scaffolding" (18H04259) from MEXT, Japan.
Further Information

Publication History

Received: 12 November 2018

Accepted after revision: 11 December 2018

Publication Date:
14 January 2019 (online)

 
 PDF Download Permissions and Reprints All articles of this category


Abstract

A catalytic reaction involving the cleavage of two carbon–sulfur bonds in 2,2′-bis(methylthio)-1,1′-biaryl derivatives is reported. This reaction does not require a transition-metal catalyst and is promoted by a thiolate anion. Notably, based on DFT calculations, the product-forming cyclization step is shown to proceed through a concerted nucleophilic aromatic substitution (CSNAr) mechanism.


#

Key words

C–S cleavage - C–O cleavage - metathesis - thiophene

Given the fact that heteroatom analogues of cyclopentadiene, such as siloles,[1] phospholes,[2] and thiophenes,[3] which are referred as heteroles,[4] have emerged as privileged scaffolds in the field of organic materials and pharmaceuticals, the development of synthetic methods for preparing these compounds has attracted considerable interest. In this context, we have been investigating the synthesis of heteroles that proceed through cleavage of a carbon–heteroatom bond.[5] Morandi and our group both independently reported on a palladium-catalyzed method for the synthesis of phosphole derivatives from bisphosphines through the formal metathesis of carbon–phosphorus bonds (Scheme [1a]).[6] These reactions allow for the rapid, efficient construction of a range of elaborate phosphole derivatives by using commercially available bisphosphines as substrates. We report herein on a sulfur variant of this reaction and its use in the synthesis of thiophenes (Scheme [1b]).

We hypothesized that 2,2′-bis(methylthio)-1,1′-biphenyl (1a-Me) could be cyclized by palladium catalysis to give the dibenzothiophene (2a) through a mechanism similar to that reported for the corresponding bisphosphines.[6a] Consistent with this expectation, the desired product was obtained in 97% yield.[7] To our surprise, 2a was formed, even in the absence of a palladium catalyst (Table [1]).[8] For example, the reaction of 1a-Me in the presence of a catalytic amount of KOtBu (20 mol%) in N,N-dimethylformamide (DMF) at 160 °C for 4 h afforded 2a in 56% yield (entry 3).

Zoom Image
Scheme 1 Synthesis of heteroles through the formal metathesis of ­carbon–heteroatom bonds

Screening a series of bases led to an improvement in the yield of 2a, with NaSMe being the most effective, giving 2a in 87% isolated yield[9] (Table [1], entry 5). The use of a polar aprotic solvent, such as DMF, was essential for the success of this reaction (entries 6–8), suggesting that this cyclization reaction proceeded through a nucleophilic substitution process. Having established that NaSMe is the optimal initiator for this reaction, we proceeded to explore the effect of the leaving groups on the sulfur atoms (Table [2]). Increasing the steric bulk on the sulfur substituent led to a dramatic decrease in yield, which would be predicted based on the assumption that an SN2 mechanism is involved in the C(alkyl)–S bond cleavage process. Although the desired cyclization of substrates having ethyl and phenethyl groups took place with low efficiency, it was possible to improve the yields by changing the reaction conditions (entries 1 and 2). In the case of an isopropyl group, no desired product was formed (entry 3). When a benzyl-substituted substrate was used, dibenzothiophene was successfully formed in 78% yield, along with dibenzyl sulfide (71%) (entry 4). These results suggest that the benzyl thiolate, which is generated after the cyclization reaction via a C(aryl)–S bond cleavage, also functions as a nucleophile in the cleavage of the C(alkyl)–S bond.

Table 1 Optimization of Reaction Conditionsa

Entry

Nucleophile

Solvent

NMR yield of 2a (%)

1

LiO t Bu

DMF

 0

2

NaO t Bu

DMF

 3

3

KO t Bu

DMF

56

4

NaOMe

DMF

16

5

NaSMe

DMF

99 (87b)

6

NaSMe

toluene

 0

7

NaSMe

1,4-dioxane

 0

8

NaSMe

tAmyl-OH

 0

a Reaction conditions: 1a-Me (0.20 mmol) and the nucleophile (0.04 mmol) in DMF (1.0 mL) at 160 °C for 4 h.

b Isolated yield.

Table 2 Effect of the Leaving Groupsa

Entry

R

Isolated yield of 2a (%)

1

Et

16 (79b)

2

Phenylethyl

39 (87c)

3

i Pr

 0 (15c)

4

Benzyl

78d

a Reaction conditions: 1a-R (0.20 mmol) and NaSMe (0.04 mmol) in DMF (1.0 mL) at 160 °C for 18 h.

b NaSMe (0.08 mmol) was used.

c NaO t Bu (0.40 mmol) was used instead of NaSMe.

d Dibenzyl sulfide was also obtained (71% isolated yield).

We next investigated the scope of the reaction with respect to SMe-substituted biaryl substrates (Scheme [2]). Gratifyingly, this method allowed us to synthesize various biaryl substrates containing a range of functional groups, including ketone (2b), cyano (2c), trifluoromethyl (2d), and amide (2e and 2f) groups. Interestingly, the introduction of an electron-donating group, such as a methyl group at the para-position to the SMe group was tolerated, with the cyclized product 2g being formed in 94% yield. Our formal metathesis method also allowed us to incorporate alkenes (2h and 2i), naphthalenes (2j) and a pyridine ring (2k) into the molecule, resulting in the synthesis of a variety of π-extended thiophenes. Pleasingly, it was also possible to prepare six-membered rings (2l) by using this method.

Zoom Image
Scheme 2 Substrate Scope. Reaction conditions: 1b-1l (0.20 mmol) and NaSMe (0.04 mmol) in DMF (1.0 mL) at 160 °C for 4 h. Isolated yields are shown. a 180 °C. b 18 h.

To get insights into the mechanism, we used DFT calculations to further investigate the cyclization of 1a-Me with an SMe anion (Figure [1]).[10] The energy changes at the B3LYP/6-311+G* level of theory [SCRF (pcm, solvent=N,N-dimethylformamide)] are shown in kcal/mol (Figure [1]). An exothermic reaction pathway with two transition states (TS1 and TS2) was obtained. The first step is the cleavage of a C(alkyl)–S bond via an SN2 mechanism and the calculated Gibbs energy of activation (∆G ‡) was estimated to be 33.4 kcal/mol. The calculations also indicate that the second step proceeds through a concerted nucleophilic aromatic substitution reaction (CSNAr)[11] pathway with a ∆G ‡ of 35.0 kcal/mol,[12] which is the rate-determining step. Notably, a Meisenheimer type intermediate could not be obtained by intrinsic reaction coordinate (IRC) calculations at TS2. Since the negative charge in the TS2 is also dispersed at the sulfur atoms in addition to the arene ring, the reaction would be expected to be less sensitive to the electronic effect of the arene ring, compared with a pathway that proceeds through a classical SNAr mechanism involving a Meisenheimer intermediate, in which the negative charge is accommodated over the aromatic ring. The involvement of a CSNAr mechanism is consistent with the successful cyclization of the electron-rich substrate 1g (Scheme [2]).

Zoom Image
Figure 1 Reaction pathways for carbon–sulfur bond metathesis

We subsequently investigated the possibility of extending the C–S bond metathesis reaction to the corresponding C–O bonds. A C(sp2)–O bond is typically an inert bond and transition metals are normally required to activate them.[13] It should be noted, however, that nucleophilic aromatic substitution reactions in which an OMe group serves as a leaving group have recently been reported. However, the use of substrates bearing strong electron-withdrawing groups, such as cyano groups at ortho- or para-positions are required for such reactions to proceed.[14] [15] We initially examined the reaction of 2,2′-dimethoxy-1,1′-binaphthalene (1m) in the presence of NaSMe (400 mol%) at 160 °C for 18 h, but the expected dibenzofuran derivative 2m was not formed. The lower reactivity of 1m compared with 1j can be attributed to the lower nucleophilicity of the phenoxide anion and poorer leaving ability[11c] of an OMe group compared to an SMe group. Optimization of the reaction of 1m led us to discover that when 8 equivalents of KO t Bu were used as a base at 190 °C, 2m was produced in 70% yield (Scheme [3]). These conditions can also be used for the cyclization of the more challenging biphenyl-based substrate 1n. DFT calculations revealed that the cyclization of 1m and 1n proceeds via a Meisenheimer intermediate,[16] probably because the OMe is a poorer leaving group than an SMe group.

Zoom Image
Scheme 3 Synthesis of dibenzofurans via metathesis of carbon-oxygen bonds. Reaction conditions: 1m–n (0.10 mmol) and KO t Bu (0.80 mmol) in DMF (1.0 mL) at 190 °C for 18 h. Isolated yields are shown.
Zoom Image
Scheme 4 Chemoselective ring closure via formal C-O/C-S metathesis

We also investigated substrates bearing both carbon–sulfur and carbon–oxygen bonds (Scheme [4]). Treatment of the biaryl substrate 1o, which is readily accessible from simple naphthalene derivatives,[17] with a stoichiometric amount of KO t Bu gave the dibenzofuran derivative 2m selectively in a yield of 52%. The selective O-cyclization of 1o is not surprising given that the OH group is a far poorer leaving group than the SMe group. In contrast, selective S-cyclization is possible by the reaction of the ethylated biaryl 1p under our conditions to form the dibenzothiophene derivative 2j in 71% yield. The selectivity for the cyclization of 1p is determined by the initial de-alkylation step, in which a less hindered methyl group reacts more rapidly than an ethyl group.

In summary, we report herein on the thiolate-initiated formal double carbon–sulfur bonds metathesis for use in the synthesis of dibenzothiophene derivatives. The C(aryl)–S bond cleavage process was found to proceed through a concerted nucleophilic aromatic substitution (CSNAr) pathway. Furthermore, this metathesis protocol enables the method to be expanded to double carbon–oxygen bonds and carbon–oxygen/carbon–sulfur metathesis.


#

Acknowledgment

We thank the Instrumental Analysis Center, Faculty of Engineering, Osaka University, for their assistance with HRMS.

Supporting Information

    Supporting information for this article is available online at https://doi-org.accesdistant.sorbonne-universite.fr/10.1055/s-0037-1611974.
  • Supporting Information

  • References and Notes

  • 7 Treatment of 1a-Me (0.20 mmol) with [(allyl)PdCl]2 (5.0 mol%) and NaO t Bu (200 mol%) in DMF (1 mL) under 160 °C for 18 h gave 2a in 97% NMR yield.
  • 8 Treatment of 1j with a stoichiometric amount of NaSMe was reported to give a demethylated compound, along with a small amount of 2j (11%). However, this was not investigated in detail. See: Furia FD, Licini G, Modena G, Valle G. Bull. Soc. Chim. Fr. 1990; 134
  • 9 Dibenzothiophene (2a) [CAS: 132-65-0]; Typical Procedure An oven-dried 5 mL screw-capped vial was charged with 1a-Me (49.2 mg, 0.20 mmol), NaSMe (2.8 mg, 0.04 mmol), and DMF (1 mL) under a gentle stream of nitrogen. The vessel was then sealed and heated at 160 °C for 4 h. The mixture was cooled to r.t. and filtered through a short pad of silica gel, eluting with EtOAc. The eluent was evaporated to give a residue, which was purified by flash chromatography (Rf 0.43, hexane) to give 2a as a white solid (49 mg, 87%). 1H NMR (CDCl3, 399.78 MHz): δ = 7.44-7.47 (m, 4 H), 7.84–7.87 (m, 2 H), 8.15–8.17 (m, 2 H). 13C NMR (CDCl3, 100.53 MHz): δ = 121.7, 122.9, 124.5, 126.8, 135.7, 139.6. HRMS (EI): m/z calcd for C12H8S: 184.0347; found: 184.0349.
  • 10 Calculations were performed with the Gaussian 09, Rev. D01 program (see the Supporting Information): Frisch MJ, Trucks GW, Schlegel HB, Scuseria GE, Robb MA, Cheeseman JR, Scalmani G, Barone V, Mennucci B, Petersson GA, Nakatsuji H, Caricato M, Li X, Hratchian HP, Izmaylov AF, Bloino J, Zheng G, Sonnenberg JL, Hada M, Ehara M, Toyota K, Fukuda R, Hasegawa J, Ishida M, Nakajima T, Honda Y, Kitao O, Nakai H, Vreven T, Montgomery JA. Jr, Peralta JE, Ogliaro F, Bearpark M, Heyd JJ, Brothers E, Kudin KN, Staroverov VN, Kobayashi R, Normand J, Raghavachari K, Rendell A, Burant JC, Iyengar SS, Tomasi J, Cossi M, Rega N, Millam JM, Klene M, Knox JE, Cross JB, Bakken V, Adamo C, Jaramillo J, Gomperts R, Stratmann RE, Yazyev O, Austin AJ, Cammi R, Pomelli C, Ochterski JW, Martin RL, Morokuma K, Zakrzewski VG, Voth GA, Salvador P, Dannenberg JJ, Dapprich S, Daniels AD, Farkas Ö, Foresman JB, Ortiz JV, Cioslowski J, Fox DJ. Gaussian 09, Revision D.01 . Gaussian, Inc; Wallingford CT: 2013
    Search in Google Scholar
  • 12 A pathway initiated by a single electron transfer from NaSMe to 1a is unlikely, because this process was found to be endothermic by 66.3 kcal/mol. See Supporting Information for details.
    • 16a See the Supporting Information for further information regarding the reaction pathway of the C-O/C-O bond meta­thesis.
    • 16b C-O/C-S metathesis was also investigated by DFT calculations, which indicated that the nature of the leaving group determines the mechanism, rather than the nature of the nucleo­phile. A better SMe leaving group favors CSNAr, whereas an OMe leaving group favors SNAr.
  • 17 Yanagi T, Otsuka S, Kasuga Y, Fujimoto K, Murakami K, Nogi K, Yorimitsu H, Osuka S. J. Am. Chem. Soc. 2016; 138: 14582
    CrossrefPubMedSearch in Google Scholar

  • References and Notes

  • 7 Treatment of 1a-Me (0.20 mmol) with [(allyl)PdCl]2 (5.0 mol%) and NaO t Bu (200 mol%) in DMF (1 mL) under 160 °C for 18 h gave 2a in 97% NMR yield.
  • 8 Treatment of 1j with a stoichiometric amount of NaSMe was reported to give a demethylated compound, along with a small amount of 2j (11%). However, this was not investigated in detail. See: Furia FD, Licini G, Modena G, Valle G. Bull. Soc. Chim. Fr. 1990; 134
  • 9 Dibenzothiophene (2a) [CAS: 132-65-0]; Typical Procedure An oven-dried 5 mL screw-capped vial was charged with 1a-Me (49.2 mg, 0.20 mmol), NaSMe (2.8 mg, 0.04 mmol), and DMF (1 mL) under a gentle stream of nitrogen. The vessel was then sealed and heated at 160 °C for 4 h. The mixture was cooled to r.t. and filtered through a short pad of silica gel, eluting with EtOAc. The eluent was evaporated to give a residue, which was purified by flash chromatography (Rf 0.43, hexane) to give 2a as a white solid (49 mg, 87%). 1H NMR (CDCl3, 399.78 MHz): δ = 7.44-7.47 (m, 4 H), 7.84–7.87 (m, 2 H), 8.15–8.17 (m, 2 H). 13C NMR (CDCl3, 100.53 MHz): δ = 121.7, 122.9, 124.5, 126.8, 135.7, 139.6. HRMS (EI): m/z calcd for C12H8S: 184.0347; found: 184.0349.
  • 10 Calculations were performed with the Gaussian 09, Rev. D01 program (see the Supporting Information): Frisch MJ, Trucks GW, Schlegel HB, Scuseria GE, Robb MA, Cheeseman JR, Scalmani G, Barone V, Mennucci B, Petersson GA, Nakatsuji H, Caricato M, Li X, Hratchian HP, Izmaylov AF, Bloino J, Zheng G, Sonnenberg JL, Hada M, Ehara M, Toyota K, Fukuda R, Hasegawa J, Ishida M, Nakajima T, Honda Y, Kitao O, Nakai H, Vreven T, Montgomery JA. Jr, Peralta JE, Ogliaro F, Bearpark M, Heyd JJ, Brothers E, Kudin KN, Staroverov VN, Kobayashi R, Normand J, Raghavachari K, Rendell A, Burant JC, Iyengar SS, Tomasi J, Cossi M, Rega N, Millam JM, Klene M, Knox JE, Cross JB, Bakken V, Adamo C, Jaramillo J, Gomperts R, Stratmann RE, Yazyev O, Austin AJ, Cammi R, Pomelli C, Ochterski JW, Martin RL, Morokuma K, Zakrzewski VG, Voth GA, Salvador P, Dannenberg JJ, Dapprich S, Daniels AD, Farkas Ö, Foresman JB, Ortiz JV, Cioslowski J, Fox DJ. Gaussian 09, Revision D.01 . Gaussian, Inc; Wallingford CT: 2013
    Search in Google Scholar
  • 12 A pathway initiated by a single electron transfer from NaSMe to 1a is unlikely, because this process was found to be endothermic by 66.3 kcal/mol. See Supporting Information for details.
    • 16a See the Supporting Information for further information regarding the reaction pathway of the C-O/C-O bond meta­thesis.
    • 16b C-O/C-S metathesis was also investigated by DFT calculations, which indicated that the nature of the leaving group determines the mechanism, rather than the nature of the nucleo­phile. A better SMe leaving group favors CSNAr, whereas an OMe leaving group favors SNAr.
  • 17 Yanagi T, Otsuka S, Kasuga Y, Fujimoto K, Murakami K, Nogi K, Yorimitsu H, Osuka S. J. Am. Chem. Soc. 2016; 138: 14582
    CrossrefPubMedSearch in Google Scholar

   Permissions and Reprints All articles of this category
Zoom Image
Scheme 1 Synthesis of heteroles through the formal metathesis of ­carbon–heteroatom bonds
Zoom Image
Scheme 2 Substrate Scope. Reaction conditions: 1b-1l (0.20 mmol) and NaSMe (0.04 mmol) in DMF (1.0 mL) at 160 °C for 4 h. Isolated yields are shown. a 180 °C. b 18 h.
Zoom Image
Figure 1 Reaction pathways for carbon–sulfur bond metathesis
Zoom Image
Scheme 3 Synthesis of dibenzofurans via metathesis of carbon-oxygen bonds. Reaction conditions: 1m–n (0.10 mmol) and KO t Bu (0.80 mmol) in DMF (1.0 mL) at 190 °C for 18 h. Isolated yields are shown.
Zoom Image
Scheme 4 Chemoselective ring closure via formal C-O/C-S metathesis