Low-Valent Titanium-Mediated Radical Conjugate Addition Using Benzyl Alcohols as Benzyl Radical Sources

Article abstract of DOI:10.1021/acs.orglett.8b02305

A concise method to directly generate benzyl radicals from benzyl alcohol derivatives has been developed. The simple and inexpensive combination of TiCl₄(collidine) (collidine = 2,4,6-collidine) and manganese powder afforded a low-valent titanium reagent, which facilitated homolytic cleavage of benzylic C-OH bonds. The application to radical conjugate addition reactions demonstrated the broad scope of this method. The reaction of various benzyl alcohol derivatives with electron-deficient alkenes furnished the corresponding radical adducts.

Full text of DOI:10.1021/acs.orglett.8b02305

Letter
pubs.acs.org/OrgLett
Cite This: Org. Lett. XXXX, XXX, XXX−XXX
Low-Valent Titanium-Mediated Radical Conjugate Addition Using
Benzyl Alcohols as Benzyl Radical Sources
Takuya Suga,* Shoma Shimazu, and Yutaka Ukaji*
Division of Material Chemistry, Graduate School of Natural Science and Technology, Kanazawa University, Kakuma, Kanazawa,
Ishikawa 920-1192, Japan
S
* Supporting Information
ABSTRACT: A concise method to directly generate benzyl
radicals from benzyl alcohol derivatives has been developed. The
simple and inexpensive combination of TiCl4(collidine)
(collidine = 2,4,6-collidine) and manganese powder afforded a
low-valent titanium reagent, which facilitated homolytic cleavage
of benzylic C−OH bonds. The application to radical conjugate
addition reactions demonstrated the broad scope of this method.
The reaction of various benzyl alcohol derivatives with electron-
deficient alkenes furnished the corresponding radical adducts.
omolytic carbon−heteroatom bond cleavage is a
Scheme 1. Titanium-Mediated C−OH Bond Cleavage
H
promising strategy to introduce carbon radicals into
organic synthesis. Generally, the hydroxy group has been
regarded as a carbon radical synthon owing to its role as a
synthetic precursor to halides and various esters that can
undergo homolysis (e.g., xanthate,1 oxalate,2 benzoate,3 and
phosphate4). However, the necessity of additional synthetic
steps for their preparation inevitably elongates synthetic
schemes.5 Despite its potential usefulness, the direct use of
the hydroxy group in radical reactions remains underexplored.6
In particular, its practical application to C−C bond formation
is almost unknown. To realize the one-step radical C−C bond
formation reaction via C−OH bond cleavage, we focused on
low-valent titanium complexes that possess sufficient oxophi-
licity to extract oxygen atoms from organic compounds (Ti−O
= ca. 660 kJ/mol, C−OH = ca. 400 kJ/mol7) and are capable
of single-electron reduction.8 The intriguing precedent for
titanium-mediated homolytic C−O bond cleavage emerged
more than 50 years ago (Scheme 1a). In 1965, van Tamelen
and Schwartz reported that the thermal decomposition of low-
valent benzyl titanate afforded a benzyl radical dimer.9
Subsequently, a few improved procedures were reported in
the 1960s and 1970s.10 However, these reactions received little
attention as possible foundations for C−C bond formation.
Low-valent titanium reagents are also known deoxygenating
agents (Scheme 1b).11 Although more than one mechanism is
possible, it is likely that deoxygenation proceeds through a
carbon radical intermediate.11c,12 Accessing carbon radicals for
C−C bond formation via these methods for purposes other
than dimerization has not yet been reported. For the most
relevant example, Zheng recently reported the first one-step
dehydroxylative radical conjugate addition reactions of N-
carbonyl hemiaminals by means of a Cp2TiCl2/Mg/TMSCl
system (Scheme 1c).13,14 Owing to the role of the alkyl
chloride as an intermediate, this reaction should be
distinguished mechanistically from van Tamelen and
Schwartz’s reaction.13a Presumably, the titanium-mediated
homolytic C−O bond cleavage has not been applied actively
to organic synthesis on account of the reported harsh reaction
conditions9,10 and the undesired deoxygenation pathway11 as
depicted in Scheme 1a,b. Therefore, by addressing these
problems, the titanium-mediated homolytic C−O bond
cleavage would become a new promising tool for radical C−
C bond formation reactions. We report here a general method
for generating benzyl radicals from the corresponding benzyl
Received: July 22, 2018
© XXXX American Chemical Society
A
DOI: 10.1021/acs.orglett.8b02305
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
alcohols by means of an inexpensive low-valent titanium
reagent (Scheme 1d). This method works regardless of the
electronic nature of the aromatic moiety and facilitates
conjugate addition to a variety of electron-deficient alkenes.15
In the preliminary investigation, we found that harsh
reductants (e.g., K, LiAlH4) were unnecessary and that the
simple and cheap TiCl4/Mn system was sufficient to cleave the
benzyl alcohol C−O bonds (Table 1). For instance, a
and the predominant formation of 3a (58% yield) with 4b are
consistent with the results reported in Table 1 (entry 2).
Although 4a was ultimately deemed the optimal reductant
among the complexes that we examined, further screening of
titanium complexes revealed that this reaction does not require
the design of a specific titanium reagent under the tested
conditions. For instance, the results obtained using
TiCl4(tmeda) (4c) (tmeda = N,N,N′,N′-tetramethylethylene-
diamine) and TiCl4(thf)2 (4d) were comparable to those
obtained with 4a (entries 3 and 4, 86 and 64% yields).
a
Table 1. Dimerization versus Deoxygenation
This method was applicable to various benzyl alcohol
derivatives (Table 3). To improve the yields, 1.2 equiv of 2,4,6-
Table 3. Scope of Benzyl Alcohols
b
yield (%)
entry
1
[Ti]
x (equiv)
2a
47
0
3a
TiCl4(collidine) (4a)
Cp2TiCl2 (4b)
Cp2TiCl2 (4b)
1.1
2.1
1.1
23
82
41
c
2
c
3
trace
a
b
See SI for preparation of low-valent titanium reagents. NMR yield.
c
Reproduction of experiment in ref 11c.
combination of TiCl4(collidine) (4a) (collidine = 2,4,6-
collidine) and manganese powder effectively cleaved the C−
O bond of 2-naphthalenemethanol (1a) to afford dimer 2a and
deoxygenated product 3a in 47% and 23% yields, respectively
(entry 1). The preferential formation of 2a suggested that the
benzyl radical was generated at a sufficiently high concen-
tration to dimerize because the deoxygenation pathway was
not operative. In contrast, the reaction with two equivalents of
Cp2TiCl2 (4b) and manganese afforded only 3a in 82% yield,
as previously reported (entry 2).11c The use of one equivalent
of 4b halved the yield of 3a (41% yield) and did not generate
dimer 2a (entry 3). Compound 4a was prepared nearly
quantitatively from inexpensive TiCl4 and collidine.16 This
material had sufficient quality to use without purification.
Compound 4a is suitable for brief use under an air atmosphere
and can be stored in a refrigerator for several months.
On the basis of these results, we applied the Ti/Mn system
to Giese-type radical C−C bond formation reactions employ-
ing acrylonitrile (5a) as a radical acceptor (Table 2).15 In this
case, 2.0 equiv titanium reagent was necessary to obtain
satisfactory yield (see SI). As expected, the radical conjugate
addition product 6aa was obtained in 92% yield when 4a was
used (entry 1). The substantially low yield of 6aa (13% yield)
a
Table 2. Application to Radical Conjugate Addition
a
b
Isolated yield of 6. 4 equiv of alkene was used.
b
yield (%)
entry
[Ti]
6aa
2a
3a
collidine was added. Its effect was summarized in the SI. Both
2-naphthalenemethanol (1a) and benzyl alcohol (1b) gave
their corresponding products in very good isolated yields (88
and 85% yields, respectively). para- and meta-Alkyl sub-
stitutions did not significantly affect the yield (1c−1e, 63−77%
yields), and the para-substituted halide series also gave suitable
substrates (1f−1h, 63−74% yields). Both electron-rich and
1
2
3
4
TiCl4(collidine) (4a)
Cp2TiCl2 (4b)
TiCl4(tmeda) (4c)
TiCl4(thf)2 (4d)
92
13
86
64
<5
0
0
<5
58
<5
<5
<5
a
b
See SI for experimental detail. NMR yield.
B
DOI: 10.1021/acs.orglett.8b02305
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
electron-deficient benzyl alcohols underwent the desired
reaction. For instance, 4-methoxycarbonyl-, cyano-, and
trifluoromethyl-substituted benzyl alcohols afforded the
products in good yields (1i−1k, 53−71% yields). The
electron-rich 4-methoxy- and tert-butyldimethylsiloxy-benzyl
alcohols were also good substrates (1l and 1m, 82 and 85%
yields).17 The complete retention of the phenolic silyl ether
moiety highlighted the mildness of the reaction conditions.
Additionally, heteroaromatic 2-thiophenemethanol (1n) af-
forded the expected product in 63% yield. ortho-Substituted
alcohols 1o−1q reacted smoothly (61−72% yields), and the
yield significantly increased when methyl groups occupied both
the 2- and 6-positions (1r, 92% yield). The good yields
obtained with secondary and tertiary benzyl alcohols promise a
further expansion of the reaction scope (1s−1v, 59−92%
yields).
Scheme 2. Mechanistic Study
Several electron-deficient alkenes 5 were reacted with
alcohol 1a (Table 4). In addition to acrylonitrile (5a), ethyl
Table 4. Scope of Alkenes
Scheme 3. Proposed Mechanism
a
b
Isolated yield of 6. 3 equiv of alkene was used.
acrylate (5b) underwent the conjugate addition to afford the
expected product (79% yield). The successful utilization of
methacrylate 5c (83% yield) suggested that this reaction
tolerates substitution at the internal position of the alkene
moiety. N,N-Dimethylacrylamide (5d) was a potentially
suitable radical acceptor (38% yield). The considerable
decreases in yields are likely due to the relatively low reactivity
of the benzyl radical. Although phenyl vinyl sulfone (5e) was
less reactive than 5a, the radical adduct was obtained in
reasonable yield (68% yield).
another low-valent titanium species to give alkyltitanium
D.11c,12a Protonation subsequently furnishes the conjugate
addition product. Alternatively, C is directly converted to the
final product through hydrogen atom transfer (HAT).12b
In summary, we have developed a concise method for
generating benzyl radicals from benzyl alcohol derivatives and
explored their synthetic application by exploiting the attractive
nature of low-valent titanium species. This simplified access to
carbon radicals will provide an opportunity to employ alcohols
in a greater number of radical reactions.
Finally, we conducted several mechanistic studies to exclude
the possible intermediacy of benzyl chloride as observed in the
previously described Zheng’s reaction (Schemes 1c and 2).13a
We found that the reaction with 4-tert-butylbenzyl chloride
(7d) afforded the conjugate addition product 6da under
conditions identical to those used with benzyl alcohol 1d
(Scheme 2a);18 however, its conversion after 10 h was
noticeably lower than that observed for 1d. In the reaction
with 1d, only a trace amount of 7d (ca. 0.2% yield) was
observed. Additionally, the treatment of 1a with 4a in the
absence of manganese did not afford chloride 7a at all, and the
initial 1a was recovered quantitatively (Scheme 2b). These
experiments support the theory that the benzyl chloride
pathway is not predominant, at least in our reaction.
The proposed reaction mechanism is depicted in Scheme 3.
Reaction between the benzyl alcohol and the low-valent
titanium complex initiates the in situ formation of titanate A.
The thermal decomposition of A formally generates benzyl
radical B and titanium oxide and is followed by radical
conjugate addition. The generated radical C is trapped by
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.or-
glett.8b02305.
■
S
Experimental procedures, mechanistic studies, and
analytical data (PDF)
1H, 13C, and 19F NMR spectra for all new products 6
(PDF)
AUTHOR INFORMATION
Corresponding Authors
■
*E-mail: suga-t@se.kanazawa-u.ac.jp.
*E-mail: ukaji@staff.kanazawa-u.ac.jp.
ORCID
Takuya Suga: 0000-0002-7527-9611
Yutaka Ukaji: 0000-0003-3185-3113
C
DOI: 10.1021/acs.orglett.8b02305
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
́
Oltra, J. E.; Bunuel, E.; Justicia, J.; Cardenas, D. J.; Cuerva, J. M. J.
̃
Notes
Am. Chem. Soc. 2010, 132, 12748−12756.
The authors declare no competing financial interest.
(13) (a) Zheng, X.; Dai, X.-J.; Yuan, H.-Q.; Ye, C.-X.; Ma, J.; Huang,
P.-Q. Angew. Chem., Int. Ed. 2013, 52, 3494−3498. (b) Zheng, X.; He,
J.; Li, H.-H.; Wang, A.; Dai, X.-J.; Wang, A.-E.; Huang, P.-Q. Angew.
Chem., Int. Ed. 2015, 54, 13739−13742.
ACKNOWLEDGMENTS
■
This work was supported in part by Mitsui Chemicals Award in
Synthetic Organic Chemistry, Japan, UBE Industries Founda-
tion award, and Kanazawa University SAKIGAKE Project.
(14) The reactions utilizing Cp2TiIIICl are intensively studied.
(a) Nugent, W. A.; Rajanbabu, T. V. J. Am. Chem. Soc. 1988, 110,
8561−8562. (b) Tarantino, K. T.; Miller, D. C.; Callon, T. A.;
Knowles, R. R. J. Am. Chem. Soc. 2015, 137, 6440−6443. (c) Zhang,
Y.-Q.; Jakoby, V.; Stainer, K.; Schmer, A.; Klare, S.; Bauer, M.;
REFERENCES
■
(1) (a) Barton, D. H. R.; McCombie, S. W. J. Chem. Soc., Perkin
Trans. 1 1975, 1574−1585 For reviews, see:. (b) Crich, D.; Quintero,
L. Chem. Rev. 1989, 89, 1413−1432. (c) Chenneberg, L.; Ollivier, C.
Chimia 2016, 70, 67−76.
̈
Grimme, S.; Cuerva, J. M.; Gansauer, A. Angew. Chem., Int. Ed. 2016,
55, 1523−1526. (d) Hao, W.; Wu, X.; Sun, J. Z.; Siu, J. C.;
MacMillan, S. N.; Lin, S. J. Am. Chem. Soc. 2017, 139, 12141−12144.
̈
(e) Zhang, Y.-Q.; Vogelsang, E.; Qu, Z.-W.; Grimme, S.; Gansauer, A.
(2) (a) Dolan, S. C.; MacMillan, J. J. Chem. Soc., Chem. Commun.
1985, 1588−1589. (b) Barton, D. H. R.; Crich, D. Tetrahedron Lett.
1985, 26, 757−760. (c) Coppa, F.; Fontana, F.; Lazzarini, E.; Minisci,
F.; Pianese, G.; Zhao, L. Chem. Lett. 1992, 21, 1295−1298.
(d) Lackner, G. L.; Quasdorf, K. W.; Overman, L. E. J. Am. Chem.
Soc. 2013, 135, 15342−15345. (e) Nawrat, C. C.; Jamison, C. R.;
Slutskyy, Y.; MacMillan, D. W. C.; Overman, L. E. J. Am. Chem. Soc.
2015, 137, 11270−11273.
(3) (a) Boar, R. B.; Joukhadar, L.; McGhie, J. F.; Misra, S. C.;
Barrett, A. G. M.; Barton, D. H. R.; Prokopiou, P. A. J. Chem. Soc.,
Chem. Commun. 1978, 0, 68−69. (b) Saito, I.; Ikehira, H.; Kasatani,
R.; Watanabe, M.; Matsuura, T. J. Am. Chem. Soc. 1986, 108, 3115−
3117. (c) Prudhomme, D. R.; Wang, Z.; Rizzo, C. J. J. Org. Chem.
Angew. Chem., Int. Ed. 2017, 56, 12654−12657 For the recent review,
́
see:. (f) Rosales, A.; Rodríguez-García, I.; Mun
Molina, E.; Padial, N. M.; Morales, L. P.; García-Ocan
E. Eur. J. Org. Chem. 2015, 2015, 4567−4591.
̃
oz-Bascon, J.; Roldan-
a, M.; Oltra, J.
̃
(15) (a) Giese, B. Angew. Chem., Int. Ed. Engl. 1983, 22, 753−782
For the review, see:. (b) Srikanth, G. S. C.; Castle, S. L. Tetrahedron
2005, 61, 10377−10441.
(16) The structure of analogous TiCl4(lutidine) has been
determined by XRD. Hensen, K.; Lemke, A.; Bolte, M. Z. Naturforsch.,
B: J. Chem. Sci. 2000, 55, 877−881.
(17) Regarding the mechanism of the radical formation step, the
good reactivity of both electron-deficient and rich benzyl alcohols
indicated that a simple homolysis of C−O bond is more plausible
than single-electron transfer from low-valent titanium to carbocation.
Sai, K. K. S.; Tokarz, M. J.; Malunchuk, A. P.; Zheng, C.; Gilbert, T.
M.; Klumpp, D. A. J. Am. Chem. Soc. 2008, 130, 14388−14389.
(18) Barrero, A. F.; Herrador, M. M.; Quílez del Moral, J. F.;
́
1997, 62, 8257−8260. (d) Lam, K.; Marko, I. E. Org. Lett. 2008, 10,
2773−2776.
(4) (a) Zhang, L.; Koreeda, M. J. Am. Chem. Soc. 2004, 126, 13190−
́
13191. (b) Lam, K.; Marko, I. E. Org. Lett. 2011, 13, 406−409.
(5) Recent reviews for direct C−C bond formation reactions from
alcohols: (a) Kumar, R.; Van der Eycken, E. V. Chem. Soc. Rev. 2013,
42, 1121−1146. (b) Obora, Y. ACS Catal. 2014, 4, 3972−3981.
(c) Corma, A.; Navas, J.; Sabater, M. J. Chem. Rev. 2018, 118, 1410−
1459.
́
Arteaga, P.; Akssira, M.; El Hanbali, F.; Arteaga, J. F.; Dieguez, H. R.;
́
Sanchez, E. M. J. Org. Chem. 2007, 72, 2251−2254.
(6) (a) Sato, M.; Oshima, K. Chem. Lett. 1982, 11, 157−160.
(b) Nishino, T.; Nishiyama, Y.; Sonoda, N. Tetrahedron Lett. 2002,
43, 3689−3691. (c) Nishino, T.; Nishiyama, Y.; Sonoda, N. Bull.
Chem. Soc. Jpn. 2003, 76, 635−641. (d) Ankner, T.; Hilmersson, G.
Tetrahedron 2009, 65, 10856−10862. (e) Kasner, G. R.; Boucher-
Jacobs, C.; McClain, J. M., II; Nicholas, K. M. Chem. Commun. 2016,
52, 7257−7260. (f) Hikawa, H.; Mori, Y.; Kikkawa, S.; Azumaya, I.
Adv. Synth. Catal. 2016, 358, 765−773. (g) Steffensmeier, E.;
Nicholas, K. M. Chem. Commun. 2018, 54, 790−793 Net C−OH
bond cleavage via radical reaction has also been reported:. (h) Jin, J.;
MacMillan, D. W. C. Nature 2015, 525, 87−90.
(7) Luo, Y.-R. Comprehensive Handbook of Chemical Bond Energies;
CRC Press: Boca Raton, FL, 2007; pp 255−261, 667−678.
́
(8) (a) Furstner, A.; Bogdanovic, B. Angew. Chem., Int. Ed. Engl.
̈
1996, 35, 2442−2469. (b) Streuff, J. Chem. Rec. 2014, 14, 1100−
1113. (c) Okamoto, S. Chem. Rec. 2016, 16, 857−872 See also ref
14f..
(9) van Tamelen, E. E.; Schwartz, M. A. J. Am. Chem. Soc. 1965, 87,
3277−3278.
(10) (a) Sharpless, K. B.; Hanzlik, R. P.; van Tamelen, E. E. J. Am.
Chem. Soc. 1968, 90, 209−210. (b) van Tamelen, E. E.; Åkermark, B.;
Sharpless, K. B. J. Am. Chem. Soc. 1969, 91, 1552−1554. (c) McMurry,
J. E.; Silvestri, M. J. Org. Chem. 1975, 40, 2687−2688. (d) Karunakar,
G. V.; Periasamy, M. Tetrahedron Lett. 2006, 47, 3549−3552.
(11) (a) Ledon, H.; Tkatchenko, I.; Young, D. Tetrahedron Lett.
1979, 20, 173−176. (b) Sato, F.; Tomuro, Y.; Ishikawa, H.; Oikawa,
́
́
T.; Sato, M. Chem. Lett. 1980, 9, 103−106. (c) Dieguez, H. R.; Lopez,
A.; Domingo, V.; Arteaga, J. F.; Dobado, J. A.; Herrador, M. M.;
Quílez del Moral, J. F.; Barrero, A. F. J. Am. Chem. Soc. 2010, 132,
254−259.
(12) (a) Prieto, C.; Delgado, J. A. G.; Arteaga, J. F.; Jaraíz, M.;
́
́
Lopez-Perez, J. L.; Barrero, A. F. Org. Biomol. Chem. 2015, 13, 3462−
́
3469. (b) Paradas, M.; Campana, A. G.; Jimenez, T.; Robles, R.;
̃
D
DOI: 10.1021/acs.orglett.8b02305
Org. Lett. XXXX, XXX, XXX−XXX