Titanocene(III) mediated reduction of organic halides under photoirradiation conditions

Article abstract of DOI:10.1016/j.tetlet.2004.08.119

Photoirradiation with a xenon lamp promotes the Cp₂TiCl mediated reduction of some unsaturated alkyl halides. Photoirradiation of reaction mixtures, with a xenon lamp in the UV-vis domain, facilitates the titanocene(III) mediated reduction of some organic halides. The substrates studied include ω-halo-α,β-unsaturated esters and a 6-iodohexose derivative.

Full text of DOI:10.1016/j.tetlet.2004.08.119

Tetrahedron
Letters
Tetrahedron Letters 45 (2004) 8123–8126
Titanocene(III) mediated reduction of organic halides under
photoirradiation conditions
*
´
Gregory Hersant, Mounira Bent Sadok Ferjani and Sharon M. Bennett
´ ´ ´
Departement de chimie et de biochimie, Universite du Quebec a Montreal, Case postale 8888,
`
´ ´
succursale Centre-Ville, Montreal, Quebec, Canada, H3C 3P8
´
Received 2 July 2004; revised 19 August 2004; accepted 19 August 2004
Available online 17 September 2004
Abstract—Photoirradiation of reaction mixtures, with a xenon lamp in the UV–vis domain, facilitates the titanocene(III) mediated
reduction of some organic halides. The substrates studied include x-halo-a,b-unsaturated esters and a 6-iodohexose derivative.
Ó 2004 Elsevier Ltd. All rights reserved.
Bis(cyclopentadienyl)titanium(III) chloride is a well
known titanocene(III) reagent, that is, useful in promot-
ing radical carbon–carbon bond formation. Many of the
reported reactions involve either carbonyl compounds
or epoxides as radical precursors;1 alkyl halides have
attracted less attention. Alkyl and aromatic halides are
functionalized alkyl halides as radical precursors.
Although photoirradiation of some organometallic
compounds is known to facilitate electron transfer
to organic substrates,12,13 the use of UV–vis light
as a promoter for Cp2TiCl mediated cyclizations of
unsaturated halides has not, to the best of our knowl-
edge, previously been described in the literature.14
2
reduced with the more reactive reagents Cp2TiBH4
and [Cp2TinBu2]ÀMgCl+.3 Olefins are formed by the
Cp2TiCl reduction of 1,2-dibromides4 and the reduction
of glycosyl halides with Cp2TiBH4 or Cp2TiCl gives
either anhydroalditols,5 glycals6 or C-glycosidic com-
pounds.7,8 Allyl and benzyl halides, as well as a-halocar-
bonyl compounds, are reduced by Cp2TiCl at room
temperature but unactivated alkyl halides were found
to be inert under these conditions.9,10 The growing inter-
est in using Cp2TiCl as a selective titanocene(III) reduc-
ing agent in organic synthesis has been attributed, in
part, to its apparent requirement for alkyl halide activa-
tion as well as to the broad range of functional groups
tolerated by this reagent.11 This report summarizes
results from our laboratory on the feasibility of using
Cp2TiCl to mediate intramolecular carbon–carbon bond
formation, under photoirradiation conditions, using
We began our study with alkyl halides 1 and 2. Both
compounds contain a neighboring acetoxy group and
a pendant Michael acceptor and reaction with excess
Cp2TiCl could conceivably give rise to either the reduc-
tive cyclization products 4a and 4b, the acyclic b-elimi-
nation product 5, or the simple acyclic reduction
compound 6. The reaction of 3 with Cp2TiCl was also
investigated to determine if reduction of the conjugate
ester might be competitive with carbon–halogen bond
reduction.
In this letter, for the sake of simplicity, we have used
Cp2TiCl to represent the mixture of Ti(III) species pre-
sent in THF solutions of titanocene(III) reagents pre-
pared by either the reduction of Cp2TiCl2, with Al or
1 (X = I); R,R' = Ac
X
CO2tBu
2 (X = Br); R,R' = Ac
CO2tBu
RO
R'O
CO2tBu
RO
3 (X = OAc); R,R' = Ac
4a cis and 4b trans; R,R' = Ac
6 (X = H); R,R' = Ac
R'O
8a cis and 8b trans; R,R' = CMe2
7 (X = I); R,R' = CMe2
9 (X = I); R,R' = H
R'O
11a cis and 11b trans; R,R' = H
5 R' = Ac; 10 R' = H.
Keywords: Alkenyl halides; Cyclization; Titanium and compounds.
*
Corresponding author. Tel.: +1 514 987 3000 4320; fax: +1 514 987 4054; e-mail: bennett.sharon_m@uqam.ca
0040-4039/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2004.08.119

8124
G. Hersant et al. / Tetrahedron Letters 45 (2004) 8123–8126
I
I
I
HO
O
HO
O
O
HO
CO2tBu
AcO
CO2tBu
O
O
O
CO2tBu
O
O
O
AcO
O
HO
AcO
17
18
15
16
19a cis and 19b trans
Zn,15 or by ligand exchange from CpTl and TiCl3.16,17
Work from other groups has demonstrated that Cp2TiCl
prepared in situ is synthetically useful and that isolation
of this air sensitive reagent, prior to reaction with
organic substrates, is not required.18 The in situ reagent
prepared from Cp2TiCl2/Zn is particularly interesting in
that it is inexpensive and nontoxic. The Cp2TiCl solu-
tions were all filtered before use. This step is important
for the Cp2TiCl2/Zn preparations as activated zinc alone
reacts with 1 to give a mixture of compound 5 and com-
pounds 4a and 4b.19 The reaction of 1 with the various
Cp2TiCl solutions is chemoselective, giving only the
cyclized products, but is incomplete after 6.5h under
ambient temperature and lighting conditions—even
when a large excess of reagent is used (Table 1, entries
a and b). We saw no evidence of competition between
reduction of the carbon–iodine bond and of the conju-
gated ester.20 Although reductions of 1 with excess
Cp2TiCl prepared from either Cp2TiCl2/Al or
Cp2TiCl2/Zn were incomplete under ambient conditions,
benzyl chloride was efficiently reduced with our in situ
generated reagents (vide infra). At the time the reactions
of 1 were stopped, the mixture was still the characteristic
green color associated with Cp2TiCl in THF.21 Increas-
ing either the quantity of Zn used to reduce Cp2TiCl2 or
the time of contact between Zn and Cp2TiCl2 prior to
filtration did not dramatically improve our results. The
reduction of PhCH2Cl to bibenzyl (86% yield) with iso-
lated Cp2TiCl (prepared from Cp2TiCl2 and Al) was
reported a number of year ago;10 this reaction was a useful
test of the quality of our in situ prepared reagents. Both
the Zn and the Al methods were satisfactory.22 The
Cp2TiCl2/Al reactions gave slightly better results but
were not easy to run due to difficulties with consistent
Al activation. The UV–vis spectra of our Cp2TiCl solu-
tions are included with the Supplementary information
of this letter.
were also examined to better understand the limitations
and the issues of chemoselectivity and stereoselectivity
associated with these reactions. As expected, the
bromide substrate 2 is less reactive than 1 and, while
photoirradiation did increase the ratio of cyclized prod-
ucts:unreacted starting material, we did not see the same
level of rate enhancement as for iodide 1 (entries i–k).
Changing the hydroxyl protecting groups of 1 from ace-
tyl to a cyclic acetal (7) gave a slower reacting substrate
(entry n). In an unsuccessful attempt to improve the
reaction efficiency we increased the concentration of
substrate 7 (entry o). However, under these conditions
b-elimination becomes competitive with reductive cycli-
zation. Reaction products were isolated as their depro-
tected derivatives 9, 10, 11a, and 11b.24
There was no reaction between 6-iodo-1-phenylhex-1-
yne (12) and Cp2TiCl after 6h at rt in the dark; a portion
of the iodoalcyne was reduced, however, under the pho-
toirradiation conditions (UV–vis, 6h). GC–MS analysis
of the crude reaction mixture after workup indicated the
presence of benzylidenecyclopentane (14), 1-phenylhex-
1-yne (13) and starting material 12 in a 39:3:58 ratio.
The ratio of radical cyclization product: simple reduc-
tion product (14:13) is similar to that observed for the
reaction of 12 with SmI2/DMPU.26
Compound 15 shares some of the characteristics of sub-
strate 1 but it lacks a Michael acceptor. In the absence of
a radical trap, reductive elimination, to give the vinyl
compound 16, is the predominant reaction pathway.
We detected a higher percentage of 16 in reaction mix-
tures that have been photoirradiated; the CpTl/TiCl3
reagent gave slightly better results than the reagent pre-
pared from Cp2TiCl2/Zn (entries p–s). Finally, the
reduction of 17 with Cp2TiCl was examined. The reac-
tion of 17 with both SmI2 and Bu3SnH has previously
been studied in our laboratory.27 Under some SmI2 con-
ditions28 reduction of the c-oxygenated conjugated ester
(to give 18) competes with reduction of the carbon–halo-
gen bond, however we found no evidence of 18 in our
Ti(III) reaction mixtures. Although there was almost
no reaction between 17 and Cp2TiCl in the dark, the
reaction run under the photoirradiation conditions gave
a 62:3:34 mixture of 19b:19a:17 (entries t and u).
We found that photoirradiation of reactions mixtures of
1 and titanocene(III), prepared from either Cp2TiCl2/Zn
(entries d–f) or CpTl/TiCl3 (entries g and h), was benefi-
cial.23 We isolated the cyclized products 4a and 4b in
82% and 78% yields as a 2.6:1 mixtures of isomers. A
150W xenon lamp was used as the light source.
Broad-band UV–vis light (250–950nm) was more effec-
tive than visible light alone. The diastereoselectivity
for the Ti(III) mediated reductive cyclization of 1 is
similar to that previously observed for cyclizations
mediated by either SmI2/visible light at rt (2.3/1.0) or
Bu3SnH (1.9/1.0) but is significantly lower than that of
the corresponding SmI2/HMPA mediated cyclization
(16/1.0).24,25 The reactions of halides 2, 7, 12, 15, and
17 with Cp2TiCl solutions prepared from Cp2TiCl2/Zn,
both in the dark and under the photoirradiation condi-
tions that had been successfully used for compound 1,
In general, the Cp2TiCl mediated reductive cyclizations
described in this paper are not more efficient nor more
stereoselective than the corresponding reactions with
SmI2 or Bu3SnH. The reagent prepared from
Cp2TiCl2/Zn is however less expensive and is nontoxic.
Our preliminary results demonstrate that photoirradia-
tion of Cp2TiCl reaction mixtures enhances the reacti-
vity of this mild single electron-transfer reagent; this
relatively simple modification may prove useful for

G. Hersant et al. / Tetrahedron Letters 45 (2004) 8123–8126
Table 1. Influence of UV–vis light on Cp2TiCl reductions of various organic halides
8125
Entry
Method and substrate
Conditionsa
Ratio of s.mat:products
in crude extractb
Isolated products (% yield)c
a
b
c
d
e
f
Cp2TiCl2/Zna
Cp2TiCl2/Zn
Cp2TiCl2/Al
Cp2TiCl2/Zn
Cp2TiCl2/Zn
Cp2TiCl2/Zn
CpTl/TiCl3
1
Ambient, 6.5h
Ambient, 6.5h
Ambient, 6.5h
Dark, 2.5h
1:4a + 4b (72:28)
1:4a + 4b (54:46)
1:4a + 4b (48:52)
1:4a + 4b (80:20)
1:4a + 4b (71:29)
1:4a + 4b (4:95)
1:4a + 4b (65:35)
1:4a + 4b (1:99)
2:4a + 4b (95:5)
2:4a + 4b (82:18)
2:4a + 4b (71:29)
7:8a + 8b (77:23)
7:8a + 8b (62:38)
7:8a + 8b (44:56)
—
—
1
1
—
—
1
1
Visible light, 2.5h
UV–vis light, 2.5h
Dark, 6h
—
4a (59), 4b (23)
1
g
h
i
1
—
4a (56), 4b (22)
CpTl/TiCl3
1
UV–vis light, 6h
Dark, 2.5h
Cp2TiCl2/Zn
Cp2TiCl2/Zn
Cp2TiCl2/Zn
Cp2TiCl2/Zn
Cp2TiCl2/Zn
Cp2TiCl2/Zn
Cp2TiCl2/Znd
Cp2TiCl2/Zn
Cp2TiCl2/Zn
CpTl/TiCl3
2
—
—
j
2
UV–vis light, 2.5h
UV–vis light, 6h
Dark, 2.5h
k
l
2
2 (56), 4a (12), 4b (5)
—
7
m
n
o
p
q
r
7
UV–vis light, 2.5h
UV–vis light, 6h
UV–vis light, 6h
Dark, 6.5h
—
7 (37), 8a (16.5), 8b (26.5)
7
e
7
—
15:16 (94:6)
9 (9), 10 (10), 11a (16.5), 11b (16.5)
—
15
15
15
15
17
17
UV–vis light, 6.5h
Dark, 6h
15:16 (74:26)
15:16 (97:3)
15 (51), 16 (19)
—
s
CpTl/TiCl3
Cp2TiCl2/Zn
Cp2TiCl2/Zn
UV–vis light, 6h
Dark, 6.5h
UV–vis light, 6.5h
15:16 (45:55)
17:19a:19b (99:0:1)
17:19a:19b (34:3:62)
15 (42), 16 (48)
—
17 (25), 19b (49)
t
u
a Ar atmosphere, rt; [substrate] in THF = 0.008M (Cp2TiCl2 prep) or 0.01M (CpTl/TiCl3 prep and entries a–c). An excess of Ti(III) reagent was used
(8equiv based on the s.mat) except for entry a (2equiv). A xenon lamp was used for photoirradiation exps.
b Determined by GC–MS after workup and before chromatography; neither diastereoisomers 4a and 4b nor 8a and 8b were separable using our
method.
c Isolated yields after chromatography on silica gel. Compounds 4a and 4b, as well as isomers 8a and 8b and isomers 11a and 11b, were isolated as
diastereomeric mixtures; the yields reported for each isomer were calculated using NMR ratios.
d Reaction was run at [7] = 0.024M.
e Compounds 7, 8a, and 8b were not detected in crude after workup.
3. (a) Nii, S.; Terao, J.; Kambe, N. J. Org. Chem. 2000, 65,
5291–5297; (b) Nii, S.; Terao, J.; Kambe, N. J. Org. Chem.
2004, 69, 573–576.
4. (a) Davies, S. G.; Thomas, S. E. Synthesis 1984, 1027–
1029; (b) Qian, Y.; Li, G.; Zheng, X.; Huang, Y.-Z.
Synlett 1991, 489–490.
other types of Cp2TiCl mediated reductions. For the
examples discussed in this letter, the increased reactivity
occurred without concomitant loss of chemoselectivity.
Acknowledgements
5. Cavallaro, C. L.; Schwartz, J. J. Org. Chem. 1996, 61,
3863–3864.
We thank the Natural Sciences and Engineering
Research Council of Canada (NSERC) and the Univer-
6. (a) Cavallaro, C. L.; Schwartz, J. J. Org. Chem. 1995, 60,
7055–7057; (b) Schwartz, J.; Spencer, R. P. Tetrahedron
Lett. 1996, 37, 4357–4360; (c) Hansen, T.; Daasbjerg, K.;
Skrydstrup, T. Tetrahedron Lett. 2000, 41, 8645–8649.
7. Spencer, R. P.; Schwartz, J. J. Org. Chem. 1997, 62, 4204–4205.
8. These reductions involve formation of a glycosyl radical
that may (i) abstract a hydrogen atom to give an
anhydroalditol, or (ii) add intermolecularly to a Michael
acceptor giving a C-glycoside, or (iii) be reduced by a
second equivalent of Ti(III) to give a glycosyltitanium(IV)
complex, which then eliminates a neighboring 2-O-acetyl
group to form a glycal. See Ref. 2.
9. (a) Parrish, J. D.; Shelton, D. R.; Little, R. D. Org. Lett.
2003, 5, 3615–3617; (b) Chen, L.; Zhao, G.; Ding, Y.
Tetrahedron Lett. 2003, 44, 2611–2614.
10. Yanlong, Q.; Guisheng, L.; Huang, Y. Z. J. Organomet.
Chem. 1990, 381, 29–34.
´
´
`
´
site du Quebec a Montreal for financial support. We
thank Duc Phuong Ngo and Jean-Franc¸ois Brazeau
(undergraduate NSERC summer students) for their
assistance with preliminary work on this project.
Supplementary data
Experimental procedures, characterization data for
compounds and UV–vis spectra for Cp2TiCl solu-
tions. Supplementary data associated with this article
can be found, in the online version, at doi:10.1016/
j.tetlet.2004.08.119.
11. Spencer, R. P.; Schwartz, J. Tetrahedron 2000, 56, 2103–
2112.
References and notes
12. Hubig, S. M.; Kochi, J. M. In Electron Transfer in
Chemistry; Balzani, V., Ed.; Wiley–VCH: Weinheim,
Germany, 2001; Vol. 2, Part II, Chapter 9.
13. (a) This principle has been successfully used to promote
the reduction of organic halides by lanthanide reagents as
well. See: Ogawa, A.; Ohya, S.; Sumino, Y.; Sonoda, N.;
1. For recent reviews see: Gansa¨uer, A.; Bluhm, H. Chem.
Rev. 2000, 100, 2771–2788; Gansa¨uer, A.; Rinker, B. In
Titanium and Zirconium in Organic Synthesis; Marek, I.,
Ed.; Wiley–VCH: Weinheim, Germany, 2002, pp 435–456.
2. Liu, Y.; Schwartz, J. J. Org. Chem. 1994, 59, 940–942.
Hirao, T.
Tetrahedron Lett. 1997, 38, 9017–9018;

8126
G. Hersant et al. / Tetrahedron Letters 45 (2004) 8123–8126
(b) Ogawa, A.; Sumino, Y.; Nanke, T.; Ohya, S.; Sonoda,
21. This was true for all of the reactions described in this
manuscript except for the PhCH2Cl reactions with
1.0equiv of titanocene reagent. Note that prolonged
photoirradiation (>6.5h) of reaction mixtures resulted in
a color change from green to green-blue.
22. See Supplementary data for more details.
23. Exposure of 1 to UV–vis light alone had no effect. Parallel
control experiments were run, in the presence of Cp2TiCl,
in the dark. Literature reports suggest that Cp2TiCl2 is not
inert under photoirradiation conditions (see Ref. 14) but
we made no attempt to isolate titanocene photolysis
products.
N.; Hirao, T. J. Am. Chem. Soc. 1997, 119, 2745–2746; (c)
Ogawa, A.; Ohya, S.; Doi, M.; Sumino, Y.; Sonoda, N.;
Hirao, T. Tetrahedron Lett. 1998, 39, 6341–6342.
14. The photochemistry of Cp2TiCl2 was studied a number of
years ago; see: Tsai, Z. T.; Brubaker, C. H., Jr. J.
Organomet. Chem. 1979, 166, 199–210, and references
cited therein.
15. Coutts, R. S. P.; Wailes, P. C.; Martin, R. L. J.
Organomet. Chem. 1973, 47, 375–382.
16. Manzer, L. E. Inorg. Synth. 1982, 21, 84–86.
17. For a discussion see: Enemaerke; Larsen, J.; Skrydstrup,
T.; Daasbjerg, K. J. Am. Chem. Soc. 2004, 126, 7853–
7864, and references cited therein.
24. Salari, B. S. F.; Biboutou, R. K.; Bennett, S. M.
Tetrahedron 2000, 56, 6385–6400.
18. (a) Nugent, W. A.; Rajanbabu, T. V. J. Am. Chem. Soc.
1988, 110, 8561–8562; (b) Rajanbabu, T. V.; Nugent, W.
A. J. Am. Chem. Soc. 1994, 116, 986–997.
25. For a comparison of Cp2TiCl with SmI2 and other agents
see: Gansa¨uer, A.; Rinker, B. Tetrahedron 2002, 58, 7017–
7026.
19. GC–MS analysis of crude products of the reaction of 1
with 8equiv of activated zinc alone showed a 70:30
mixture of 4a + 4b:5. There was no reaction for the
corresponding experiment with Al.
26. Zhou, Z.; Larouche, D.; Bennett, S. M. Tetrahedron 1995,
51, 11623–11644.
27. Bennett, S. M.; Biboutou, R. K.; Zhou, Z.; Pion, R.
Tetrahedron 1998, 54, 4761–4786.
20. Crude and purified products were analyzed by GC–MS
and NMR. As an additional test, 3 was exposed to an
excess of Cp2TiCl for 6h but was found to be inert.
28. At rt, in the absence of HMPA, SmI2 and 17 give 18 as the
major product; with SmI2/HMPA 19b is formed almost
exclusively.27