Free Radical-Mediated Hydroxymethylation Using CO and HCHO

372 CHIMIA 2012, 66, No. 6 Organic Free radicals

doi:10.2533/chimia.2012.372 Chimia 66 (2012) 372–376 © Schweizerische Chemische Gesellschaft Free Radical-mediated Hydroxymethylation Using CO and HCHO

Takuji Kawamoto and Ilhyong Ryu*

Abstract: Tin-free radical hydroxymethylations of haloalkanes using CO and HCHO as a C1 unit proceed efficiently in the presence of borohydrides as radical mediators. In the approach using CO, the formation of aldehydes by radical carbonylation and their subsequent reduction by hydrides lead to alcohols. On the other hand, the use of formaldehyde is more straightforward, in which the key reaction is alkyl radical addition to formaldehyde to give alkoxy radical, which abstracts hydrogen from borohydride reagents. The cascade sequences were observed in the reaction of cholesteryl bromide with HCHO, which displays the diverse applications of HCHO in radical chemistry. Keywords: Borohydrides · Carbon monoxide · Formaldehyde · Hydroxymethylation · One-carbon Homologation · Radical reaction

1. Introduction butyltin hydride and 2,2’-azobisisobutyro- NaBH CN as a reducing agent allowed 3 nitrile (AIBN, Scheme 1).[10] This work us to carry out the triphasic workup (H O/ 2 Radical cascade reactions are powerful served as a breakthrough of radical carbon- CH Cl /perfluoro hexane) to separate in- 2 2 tools in organic synthesis.[1] In the last two ylation reactions, which has since emerged organic byproducts, hydroxymethylated decades, the potential of CO as a radical as a standard method for the synthesis of a products, and fluorous tin hydride. Later C1 synthon has been well established with variety of carbonyl compounds. we noticed that an organic/fluorous hy- applications yielding a significant variety In 1993, the application of the radical brid ether, F-626: (2-perfluorohexyl)ethyl of carbonyl compounds.[2,3] In contrast formylation for hydroxymethylation of a 1,3-dimethylbutyl ether) is a good solvent formaldehyde (HCHO) has rarely been sugar iodide was reported by Gupta and for hydroxymethylation using fluorous utilized in radical chemistry after initial Kahne.[11] They used the combination of tin hydride. The separated F-626 solution pioneering studies.[4–9] In this short review, a catalytic amount of triphenylgermanium containing fluorous tin hydride was suc- we focus on two types of hydroxymethyl- hydride (Ph GeH) and excess NaBH CN cessfully used for the second iteration of 3 3 ation reaction of organohalides, in which (Scheme 2). The reaction comprises of i) experiments.[14] The recovered fluorous tin CO and HCHO were utilized as radical C1 radical carbonylation followed by ii) in hydride can be used over and over again. synthons. The reactions are coupled with situ hydride reduction of the resulting alde- All these studies contribute to creating borohydride reagents which act as a radical hydes. Although sugar iodide gave a rather greener reaction systems with a minimum mediator thus enabling these reactions to low yield of the desired one-carbon homol- use of tin reagents, however, we decided be run in the absence of toxic tin hydride ogated alcohol, a simple alkyl iodide like to investigate the possibility of conducting reagents. 1-iodoadamantane gave the corresponding hydroxymethylation of haloalkanes with- alcohol in 75% yield. In spite of the lower out using any tin reagents at all. hydrogen-donating ability of triphenylger- 2. Radical Hydroxymethylation of man compared to tributyltin hydride, it Alkyl Halides: Earlier Work still required pressurized CO conditions to 3. Tin-free Radical compete with premature quenching of the Hydroxymethylation of Alkyl In 1990, one of us demonstrated that key alkyl radicals by triphenylgerman. Iodides Using CO CO can be used for formylation of alkyl In collaboration with the Curran group, halides, when coupled with the use of tri- we carried out a similar hydroxymethyl- An opportunity arose from our discov- ation of organic halides employing cata- ery that the Giese reaction and the related lytic amounts of fluorous tin hydride.[12,13] carbonylation reaction can be carried out Combining fluorous tin hydride with using tetrabutylammonium cyanoboro-

Scheme 1. Free- Bu3SnH, AIBN O radical formylation. RX + CO R H C6H6,80ºC

*Correspondence: Prof. I. Ryu CO Department of Chemistry, Osaka Prefecture O University R Sakai, 599-8531, Osaka, Japan R Tel. +81-72-254-9695 E-mail: [email protected] Organic Free radicals CHIMIA 2012, 66, No. 6 373

Me O OMe I OEt + CO hν O NaBH CN I OBz + 3 O OBz 1400 psi OEt

Ph3GeH (10 mol %) Me O OMe O AIBN (10 mol %) O NaBH3CN (2.9 equiv.) OEt OBz CO C6H6/THF (50/1) OH OBz AIBN 105 ºC,12h n-BH NBH CN 37% 4 3

Scheme 3. Tin-free Giese reaction and relative carbonylation. OH I Ph3GeH cat. AIBN (10 mol %) OH NaBH3CN (1.5 equiv.) I + CO

n-Bu4NBH4 75% + CO MeCN, 3h (C6F13CH2CH2)3SnH 0.5 M (5 mol %) OH I AIBN (10 mol %) 85 atm, AIBN (20mol %), 80 ºC 88% NaBH3CN (1.5 equiv.) + CO 1atm, AIBN (20 mol %), 80 ºC 27% F-626, t-BuOH 90 ºC,3h 80 atm 1atm, blacklight (15 W, pyrex), 25 ºC 74% 85% (ﬁrst run) Scheme 4. Tin-free hydroxymethylation using CO. 80% (second run)

80 or 1atm BH3CN R OH R I ++CO n-Bu4NBH4 CH3CN O H O

CO M-H 91% (80 atm) OH 82% (80 atm) OH 70% (1 atm) 64% (1 atm) Scheme 2. Radical hydroxymethylation with catalytic amount of metal hydrides.

H 70% (1 atm) 48% (1 atm) OH H hydride as a substitute for tin hydride H (Scheme 3).[15] exo/endo =93/7 cis/trans =50/50 HO The development of this borohydride- based tin-free approach led us to reexamine Scheme 5. Hydroxymethylation of various alkyl iodides. that hydroxymethylation reaction. Indeed hydroxymethylation of alkyl iodides using tetrabutylammonium borohydride gave the cholesteryl radical gave a 50:50 cis/trans 4. HCHO in Radical Chemistry: expected hydroxymethylation product in mixture of hydroxymethylation product. Earlier Work good yield (Scheme 4).[16] The hydroxymethylation reaction was also Interestingly, less than 2% yield of the applied to prepare a key intermediate in the Successful hydroxymethylation of reduction product, adamantane, was de- synthesis of (±)-communiol E.[17] haloalkanes, through the use of HCHO in tected, suggesting that the H-abstraction of We propose a mechanism involving radical addition chemistry is straightfor- adamantyl radical from borohydride anion radical carbonylation and the subsequent ward and highly desirable (Scheme 7). is sluggish and therefore does not compete trapping of the resulting acyl radical by H Trapping formaldehyde by alkyl radi- with the carbonylation step. Consequently, abstraction from hydroborate anion. The cal was first reported in 1958 by Fuller and we found that by employing photo-irra- borate anion radical may undergo one- Rust.[4] They observed that the reaction of diation conditions, hydroxymethylation electron donation to the alkyl iodide to cyclohexyl radical with formaldehyde re- could proceed under atmospheric pressure generate radical and iodo borate anion,[18] sulted in the production of cyclohexane of CO. thus sustaining the radical chain (Scheme methanol (Scheme 8). Having optimized conditions for hy- 6). In 1965, Oyama reported that reac- droxymethylation of alkyl iodides in hand, tion of methanol with formaldehyde in the we applied the method to a variety of io- presence of di-tertiary butyl peroxide gave doalkanes (Scheme 5). The reaction of ethylene glycol.[5] Later, this work was re- 374 CHIMIA 2012, 66, No. 6 Organic Free radicals

t-BuOOt-Bu O (0.4 mmol) HO HO OH + H H H 2BH4 140 °C

CO 9mol 15% aqueous solution O 1.8 mol 11% RCH2OH

PhCO2Ot-Bu O (0.016 mol) O OH + H (HCHO)n O 100 °C, 6h O BH3 1.1 mol 0.33 mol 47% selectivity

BH3I R I Scheme 9. α-Hydroxymethylation of methanol and 1,3-dioxolane.

Scheme 6. Proposed mechanism. 1) air,THF O 0 °C, 90 min O C H OH + C H OH Bu3B + 5 11 4 9 H H 2) 3N NaOH aq H H O 30% H2O2 aq R 10 mmol 7.8 mmol 20.3 mmol R mineral oil 120-140 ºC Scheme 7. Key radical species of hydroxy- (HCHO)n methylation using HCHO. 100 mmol

t-BuOOt-Bu H O H O (40 mol%) 2 2 OH O Bu B H H 135 ºC H H 3 Bu Bu Bu O B 5equiv. 37% aqueous solution 38% Bu O Bu

t-BuO Scheme 10. Oxygen-promoted hydroxymethylation of tributylborane.

O O + ZnCl2 H H + (HCHO)n O N THF,reﬂux N 75%

Scheme 8. First report using HCHO as a radical acceptor. O

H H investigated by Kollar from the industrial O point of view.[6] Sanderson and coworkers examined the regiochemistry of hydroxymethylation of 1,3-dioxolanes and reported that hydrogen abstraction from the methy- Scheme 11. Intermolecular cycloaddition of formaldehyde. lene attached to two oxygen atoms showed higher reactivity than that of a methylene with a single oxygen bond (Scheme 9).[7] A rather peculiar but interesting exam- could act as radical mediators, we decid- In 1972, Brown and coworkers re- ple of cycloaddition reaction was reported ed to explore hydroxymethylation using ported the reaction of tributylborane with in 1985 by Little and coworkers, in which HCHO. formaldehyde under air to give a mixture they used HCHO as a trap of 2-methylene of 1-pentanol and 1-butanol after hydrolyt- 1,3-diyl (Scheme 11).[9] Zinc chloride was ic oxidation (Scheme 10).[8] This reaction essential in obtaining the high level of re- 5. Radical Hydroxymethylation of is thought to involve butyl radical addition gioselectivity. Alkyl Halides using HCHO to formaldehyde to give 1-pentyloxy radi- After these earlier efforts, the radical cal, which then reacts with tributylborane chemistry of formaldehyde has been only Initially we attempted the hydroxy- to form borate ester as the precursor to the rarely utilized for a quarter century. With methylation of 1-iodoadamantane us- alcohols. our knowledge that borohydride reagents ing paraformaldehyde in the presence Organic Free radicals CHIMIA 2012, 66, No. 6 375

OH X n-Bu4NBH3CN (4.0 equiv.) n-Bu4NBH3CN (4 equiv) low pressure Hg lamp initiator (6 W, quartz) + (HCHO) n H MeCN, 4h H + (HCHO) n MeCN/ C6H6 =1/2 0.25 M 15 equiv. H conv.71% 15 equiv. X=I black light, 25 ºC 85% Br X=I AIBN (20 mol %), 90 ºC 87% H X=I none, 90 ºC 0% H H

X=Br black light, 25 ºC 0% H H H + + X=Br low pressure Hg lamp, 25 ºC 85% O O OH Scheme 12. Radical hydroxymethylation using HCHO. AB C OH 16% 7% 14% (HCHO)n (dr =93/7) (dr =85/15) n-Bu4NBH3CN AIBN OH Scheme 14. Hydroxymethylation of cholesteryl bromide. I MeCN, 90 ºC HO HO 77% BH3CN (from cis) A O OH I H H 5-exo I HO 77% (from trans) O

I OH 50% OMe OMe trans/cis BH CN (HCHO) 3 n =78/22 B n-Bu4NBH3CN O low pressure Hg lamp quartz Br BH2CNBr OH MeCN, 25 ºC 57% O R Br Scheme 13. Hydroxymethylation of various alkyl halides. H H BH2CN of tetrabutylammonium borohydride. O O Unfortunately, the reaction was unsuccess- ful due to rapid reduction of formaldehyde C BH3CN by the borohydride anion. As the cyano- Scheme 15. Proposed mechanism. borohydride is known to be a more mild reducing agent, we again tested the reaction using n-Bu NBH CN and found that 4 3 it worked beautifully (Scheme 12). tylammonium borohydride. In this case the dergo two alternative reactions: i) abstrac- Irrespective of whether the radical re- 5-exo cyclization of the acyl radicals is un- tion of hydrogen from cyanoborohydride action was initiated by black-light irradia- favorable due to steric congestion, and the to give B or ii) addition to a second mol- tion or thermal initiation using AIBN, the product resulting from cyclization is not ecule of HCHO to give C.[21–23] reaction gave good yield of the desired observed. On the other hand, the reaction 1-adamantanemethanol.[19] The reaction of cholesteryl bromide with HCHO and of bromoalkanes was more difficult, how- tetrabutylammonium cyanoborohydride 6. Conclusion and Future ever, UV light irradiation resulted in the gave not only hydroxymethylated product Directions smooth conversion and good yields of the but also cyclized product (Scheme 14). desired product (Scheme 13). The hydroxymethylation product A We have shown that radical mediated As discussed above, the reaction of was formed with high cis stereoselectiv- hydroxymethylation of alkyl halides can cholesteryl iodide using CO and tetrabu- ity to the angular Me group. The alkoxy be achieved by the use of CO and tetra- tylammonium borohydride gave a 50:50 radical with the trans geometry undergoes butylammonium borohydride under tin- cis/trans mixture of hydroxymethylation 5-exo radical cyclization onto the C–C free reaction conditions. Due to a slower product. Both α/β-stereoisomers of the ac- double bond (Scheme 15).[20] The subse- H donation ability of the borohydride re- yl radical abstract hydrogen from tetrabu- quently formed THF radical can then un- agent compared to tributyltin hydride, the 376 CHIMIA 2012, 66, No. 6 Organic Free radicals

reaction can be conducted even under an [3] For radical C1 synthons having C=N moieties, [17] S. Kobayashi, T. Kinoshita, T. Kawamoto, M. atmospheric pressure of CO without suf- see: a) S. Kim, S. Kim, Bull. Chem. Soc. Jpn. Wada, H. Kuroda, A. Masuyama, I. Ryu, J. Org. 2007, 80, 809; b) G. K. Friestad, Tetrahedron Chem. 2011, 76, 7096. fering from premature quenching of key 2001, 57, 5461; c) H. Miyabe, E. Yoshioka, S. [18] For a kinetic study on halogen atom abstraction alkyl radicals. We also found that HCHO, Kohtani, Curr. Org. Chem. 2010, 14, 1254. by borane radical anion, see: B. Sheeller, K. U. when coupled with cyanoborohydride as [4] G. Fuller, F. F. Rust, J. Am. Chem. Soc. 1958, Ingold, J. Chem. Soc., Perkin Trans. 2 2001, radical mediator, can serve as a useful re- 80, 6148. 480. agent for hydroxymethylation of alkyl ha- [5] M. Oyama, J. Org. Chem. 1965, 30, 2429. [19] T. Kawamoto, T. Fukuyama, I. Ryu, J. Am. [6] a) J. Kollar, U.S. Patent. No. US4393252, 1983; Chem. Soc. 2012, 134, 875. lides. The results with cholesteryl bromide b) J. Kollar, U.S. Patent. No. US4 337371, [20] For a review on alkoxy radical cyclization, hold promise that HCHO could become a 1982; c) J. Kollar, U.S. Patent. No. US4412085, see: a) J. Hartung, in ‘Radicals in Organic promising C1 synthon in radical cascade 1983; d) J. Kollar, Chemtech 1984, 504. Synthesis’, Eds. P. Renaud, M. P. Sibi, Wiley- chemistry and we are now engaged in the [7] a) J. R. Sanderson, E. L. Yeakey, J. J. Lin, R. VCH: Toronto, 2001, Vol. 2, p. 246–278. For a Duranleau, E. T. Marquis, J. Org. Chem. 1987, recent example, see: b) M. Rueda-Becerril, J. C. development of multicomponent reactions 52, 3243; b) J. R. Sanderson, J. J. Lin, R. T. Leung, C. R. Dunbar, G. M. Sammis, J. Org. using HCHO. Duranleau, E. L. Yeakey, E. T. Marquis, J. Org. Chem. 2011, 76, 7720. Chem. 1988, 53, 2859. [21] For hydrogen abstraction of alkoxy radicals Acknowledgement [8] N. Miyaura, M. Itoh, A. Suzuki, H. C. Brown, from borohydrides, see: a) J. R. M. Giles, B. P. Financial support from JSPS and M. M. Midland, P. Jacob, III J. Am. Chem. Soc. Roberts, J. Chem. Soc., Chem. Commun. 1981, Monbukagakusho is gratefully acknowledged. 1972, 94, 6549. 360; b) J. R. M. Giles, B. P. Roberts, J. Chem. [9] R. D. Little, H. Bode, K. J. Stone, O. Wallquist, Soc., Perkin Trans. 2 1982, 1699. R. Dannecker, J. Org. Chem. 1985, 50, 2401. [22] For examples of radical hydrogen abstraction Received: March 9, 2012 [10] I. Ryu, K. Kusano, A. Ogawa, N. Kambe, N. from borane based reagents, see: a) J. A. Baban, Sonoda, J. Am. Chem. Soc. 1990, 112, 1295. B. P. Roberts, J. Chem. Soc., Perkin Trans. [1] ‘Radicals in Organic Synthesis’, Vol. 1, Vol. [11] V. Gupta, D. Kahne, Tetrahedron Lett. 1993, 34, 2 1988, 1195; b) D. H. R. Barton, M. Jacob, 2, Eds. P. Renaud, M. P. Sibi, Wiley-VCH: 591. Tetrahedron Lett. 1998, 39, 1331; c) S.-H. Weinheim, 2001. For a retrosynthetic analysis [12] D. P. Curran, S. Hadida, S.-Y. Kim, Z. Luo, J. Ueng, A. Solovyev, X. Yuan, S. J. Geib, L. of cascade radical reactions, see: D. P. Curran, Am. Chem. Soc. 1999, 121, 6607. Fensterbank, E. Lacôte, M. Malacria, M. Synlett, 1991, 63. [13] I. Ryu, T. Niguma, S. Minakata, M. Komatsu, Newcomb, J. C. Walton, D. P. Curran, J. Am. [2] For reviews on radical carbonylations, see: a) I. S. Hadida, D. P. Curran, Tetrahedron Lett. 1997, Chem. Soc. 2009, 131, 11256. Ryu, N. Sonoda, Angew. Chem. Int. Ed. 1996, 38, 7883. [23] For a recent review of NHC-borane in radical 35, 1050; b) I. Ryu, N. Sonoda, D. P. Curran, [14] H. Matsubara, S. Yasuda, H. Sugiyama, I. Ryu, reactions, see: D. P. Curran, A. Solovyev, M. Chem. Rev. 1996, 96, 177; c) C. Chatgilialoglu, Y. Fujii, K. Kita, Tetrahedron 2002, 58, 4071. M. Brahmi, L. Fensterbank, M. Malacria, E. D. Crich, M. Komatsu, I. Ryu, Chem. Rev. 1999, [15] I. Ryu, S. Uehara, H. Hirao, T. Fukuyama, Org. Lacôte, Angew. Chem. Int. Ed. 2011, 50, 10294. 99, 1991; d) I. Ryu, Chem. Soc. Rev. 2001, 16; Lett. 2008, 10, 1005. e) C. H. Schiesser, U. Wille, H. Matsubara, I. [16] S. Kobayashi, T. Kawamoto, S. Uehara, T. Ryu, Acc. Chem. Res. 2007, 40, 303. Fukuyama, I. Ryu, Org. Lett. 2010, 12, 1548.