Generation of allyl Grignard reagents via titanocene-catalyzed activation of allyl halides

Article abstract of DOI:10.1016/j.tetlet.2010.02.144

A protocol for the generation of allyl Grignard reagents via the catalytic activation of allyl halides is described herein. Subsequent nucleophilic addition to carbonyl derivatives provided the desired homo allylic alcohols in excellent yields (84-99%). E

Full text of DOI:10.1016/j.tetlet.2010.02.144

Tetrahedron Letters 51 (2010) 2427–2430
Contents lists available at ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Generation of allyl Grignard reagents via titanocene-catalyzed activation
of allyl halides
*
Lauren M. Fleury, Brandon L. Ashfeld
Department of Chemistry and Biochemistry, University of Notre Dame, Notre Dame, IN 46556, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
A protocol for the generation of allyl Grignard reagents via the catalytic activation of allyl halides is
described herein. Subsequent nucleophilic addition to carbonyl derivatives provided the desired homo
allylic alcohols in excellent yields (84–99%). Evidence suggests that titanocene dichloride catalyzes the
formation of an allyl Grignard species which reacts solely with the carbonyl electrophile as evidenced
by the complete absence of Würtz coupling. This methodology will have wide-ranging applicability in
the generation of highly reactive organometallic reagents.
Received 10 February 2010
Revised 19 February 2010
Accepted 22 February 2010
Available online 26 February 2010
Ó 2010 Elsevier Ltd. All rights reserved.
Given the ready availability of organohalides, metallation of
carbon-halogen bonds with magnesium is still one of the most
common methods of generating highly reactive organomagnesium
reagents.1 Quite possibly the most useful and well-known of the
main group organometallics are Grignard reagents.2–4 Although
the synthetic utility of Grignard reagents is undeniable, their prep-
aration can be rather tenuous.5 Conventional methods rely on the
initiation of magnesium turnings or powder with iodine or dibro-
moethane, or the use of highly reactive, pre-activated RiekeÒ mag-
nesium.6 All too often, these conditions are unsuitable for highly
functionalized substrates, and even the generation of simple Grig-
nard reagents can be complicated by homocoupling and degreda-
tion.7 The generation of allyl Grignard reagents can be especially
difficult due to their propensity to undergo Würtz coupling, result-
ing in a low yield of the desired organometallic species.8 To cir-
cumvent these issues, we surmised that the allyl halide itself
could be activated in a catalytic fashion by a titanocene complex
in the presence of magnesium turnings for an efficient, high-yield-
ing generation of allyl Grignard reagents, even in the presence of
carbonyl electrophiles (Scheme 1).9,10 We have recently reported
the successful implementation of this reductive transmetallation
protocol in the generation of organozinc reagents via the titano-
cene-catalyzed activation of alkyl halides.10
turnings as the reductant. Herein, we describe its successful imple-
mentation toward the facile generation of allyl Grignard reagents.
To test this hypothesis, we began by examining the allylation of
simple ketones with allyl bromide in the presence of a titanocene
catalyst and unactivated magnesium turnings. Thus, we were grat-
ified to see that when a solution of enone 1a and allyl bromide (2a)
was treated with Cp2TiCl2 (1 mol %) and magnesium turnings
(1.5 equiv) alcohol 3a was obtained in 99% yield at room tempera-
ture after only 10 min (Eq. 1). Based on the reported slow rate at
which allyltitanocene reagents undergo addition to ketones, this
result would seem to indicate that the allyl bromide underwent a
rapid metallation event to generate the corresponding allyl Grig-
nard reagent, even in the presence of enone 1a.9g
Although, allyl bromide underwent metallation rapidly at room
temperature, we quickly discovered that the efficiency of the pro-
cess was highly dependent on the nature of the halogen. Both allyl
chloride and allyl iodide required gentle heating (50 °C) to reach
completion, but still provided the desired adduct in 88% and 56%
yield, respectively, (Eq. 1). The low yield of the allylation with allyl
iodide is intriguing in that often this species is not a good metalla-
tion substrate due to the propensity for Würtz-coupling byprod-
ucts. However, considering that the 1,5-diene that would result
from this unproductive side reaction was not observed, the lower
It is our contention that the titanocene-catalyzed method for
the generation of organozinc reagents11 could be expanded to in-
clude other organometallic reagents with broad reactivity profiles
by simply exchanging the reducing metal to provide the desired
organometallic reagent. We hypothesized that Grignard reagents
could be generated in a mild and efficient manner utilizing our
reductive transmetallation protocol by employing magnesium
MIIX
Stoichiometric Generation of
Allylmetal Reagents
Br
Substrate Activation
M(0)
MX2
TiIVLn
TiIIILn
Reductive Transmetallation
TiIIILn
Scheme 1. Titanium-catalyzed allyl halide activation.
* Corresponding author.
E-mail address: bashfeld@nd.edu (B.L. Ashfeld).
0040-4039/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2010.02.144

2428
L. M. Fleury, B. L. Ashfeld / Tetrahedron Letters 51 (2010) 2427–2430
Table 1
yield of allyl alcohol 3a may be due to an inherent instability of the
iodo allyltitanocene intermediates. Likewise, the lower yield when
employing allyl chloride as the organometallic reagent precursor is
likely the result of an inefficient initial metallation coupled with
the thermal instability of low valent titanocene complexes. Based
on this supposition, it did not come as a surprise that allyl acetate
led to a quantitative recovery of the starting materials. Increasing
the reaction concentration failed to provide an improvement.
Temperature and additive effect on metallation
Entry
Temp (°C)
Additive
Time (min)
Yield (%)
1
2
3
4
5
rt
0
ꢀ20
ꢀ40
ꢀ40
—
—
—
—
5
5
10
300
20
>99
99
74 (94)a
NR
5 mol % dppp
98
a
Yield in parentheses based on recovered starting material.
Cp2TiCl2 (1 mol %)
Mg, THF, 10 min
O
HO Me
X
+
Ph
Ph
Me
2
1a
3a
restored the catalytic activity at ꢀ40 °C to provide alcohol 3a in
98% yield after only 20 min. Similar results were obtained with
various other phosphine additives.12 This unusual effect which
the addition of phosphines has on the metallation was also ob-
served with zinc dust as the reducing metal when generating
organozinc reagents. Although it is unclear what role phosphine
is playing in the metallation event, the step in the catalytic cycle
to benefit most from phosphine additives at low temperatures
would be the reduction of allyltitanocene(IV) to allyltitanocene(III).
However, any determination at this stage would be premature.
ð1Þ
2a 99% (rt)
X = Br
2b 88% (50 °C)
Cl
I
2c 56% (50 °C)
Based on our empirical observations and previous mechanistic
work with zinc as the reducing metal, a catalytic cycle for the
titanocene-catalyzed activation/metallation procedure is proposed
(Scheme 2).10 Initial reduction of Cp2TiCl2 to Cp2TiCl by Mg(0) pre-
cedes metallation of allyl bromide to yield the allyltitanium(IV)
species 4. A second reduction of (allyl)Ti(IV) 4 to (allyl)Ti(III) 5 en-
ables the transmetallation to magnesium thereby generating a
stoichiometric quantity of the allyl Grignard 6. The titanocene(III)
complex that is released can then re-enter the catalytic cycle and
react with another equivalent of allyl bromide (2a). Finally, addi-
tion of the allyl Grignard reagent 6 to the carbonyl substrate 1
leads to the observed allylation product following protonation of
magnesium alkoxide 7.
Table 2
Allylation of aldehydesa
Br
O
OH
2a
Cp2TiCl2 (1 mol%), Mg
THF, rt
R
H
R
8
9
The efficiency of the reductive transmetallation of allyl bromide
with catalytic titanocene and magnesium turnings proved sensi-
tive to the reaction temperature in a similar fashion to what we ob-
served with the in situ generation of organozinc reagents (Table
1).10 The titanocene-catalyzed activation/metallation/allylation se-
quence proceeded rapidly, and in near quantitative yield, when
performed at room temperature or 0 °C (entries 1 and 2). If the
reaction was cooled to ꢀ20 °C, allylic alcohol 3a was obtained in
a diminished 74% yield (entry 3). Interestingly, when the reaction
was cooled further to ꢀ40 °C, the reductive transmetallation pro-
cess appeared to completely shutdown, resulting in a near quanti-
tative recovery of the starting enone 1a (entry 4). However, the
addition of 5 mol % 1,3-bis(diphenylphosphino)propane (dppp)
Entry
Aldehyde
Yield (%)
1
2
3
4
5
6
7
8
9
Benzaldehyde (8a)
4-Tolualdehyde (8b)
4-Anisaldehyde (8c)
4-Dimethylaminobenzaldehyde (8d)
4-Trifluoromethylbenzaldehyde (8e)
4-Chlorobenzaldehyde (8f)
4-Fluorobenzaldehyde (8g)
(E)-Cinnamaldehyde (8h)
(E)-Crotonaldehyde (8i)
93
94
95
>99
84
96
87
>99
88
87
91
10
11
Cyclohexylcarboxaldehyde (8j)
Hexanal (8k)
a
Allylation reactions were carried out using 0.8 mmol 8, 0.96 mmol 2a, 1 mol %
Cp2TiCl2, and 1.2 mmol Mg turnings in THF (0.1 M) at room temperature for 10–
30 min.
2 Cp2TiIVCl
Mg0
Table 3
MgIICl2
Allylation of ketonesa
Br
HO R2
2a
Br
O
2 Cp2TiIIICl
XMgIIO
R1 R2
R1
R1
R2
2a
Cp2TiCl2 (1 mol%), Mg
THF, rt
1
3
7
Entry
Ketone
Acetophenone (1b)
Yield (%)
Cp2TiIVBrCl
1/2 Mg0
1
2
3
4
5
6
7
8
9
97
97
99
97
96
94
96
98
97
94
94b
Cl
O
(E)-4-MeO–C6H4–CH@CH–C(O)CH3 (1c)
(E)-2-HO–C6H4–CH@CH–C(O)CH3 (1d)
(E)-4-CF3–C6H4–CH@CH–C(O)CH3 (1e)
(E)-4-Cl–C6H4–CH@CH–C(O)CH3 (1f)
(E)-c-C6H11–CH@CH–C(O)CH3 (1g)
(E)-c-C3H5–CH@CH–C(O)CH3 (1h)
2-Cyclohexenone (1i)
Me2C@CH–C(O)CH3 (1j)
Me2C@CH–C(O)C6H5 (1k)
4-tert-Butylcyclohexanone (1l)
TiIVCp2
II
1/2
Mg Cl
R1 R2
4
1
Br
X
MgII
1/2 Mg0
10
11
6
TiIII
Cp
5
1/2 MgIICl2
MgIIX2
Cp
a
Reaction conditions: 0.8 mmol 1, 0.96 mmol 2a, 1 mol % Cp2TiCl2, and
1.2 mmol Mg in THF (0.1 M) at room temperature for 10–30 min.
b
syn/anti = 1.3:1.
Scheme 2. Proposed catalytic cycle.

L. M. Fleury, B. L. Ashfeld / Tetrahedron Letters 51 (2010) 2427–2430
2429
Table 4
Next, we examined the effect that the substitution on the allyl
moiety had on the regiochemical outcome of the allylation (Table
5). In particular, we were intrigued to see whether allylmetal re-
agents generated in situ would undergo addition to ketone 1a in
an SE2 or SE20 fashion. Treatment of prenyl bromide (2b) and ke-
tone 1a to our titanocene-catalyzed Grignard generation protocol
resulted in clean SE20 1,2-addition, despite the sterically hindered
gem-dimethyl motif, to provide the corresponding alcohol in 91%
yield (entry 1). Crotyl (2c) and cinnamyl (2d) bromides also under-
went SE20 addition, resulting in homoallyl alcohols in 89% and 91%
yield, respectively (entries 2 and 3). Substitution at the 2-position
of the allyl motif, as demonstrated by methallyl bromide (2e), did
not diminish the efficiency of the allylation event (entry 4).
In conclusion, we have demonstrated that Würtz-coupling
byproducts in the stoichiometric generation of allyl Grignard re-
agents can be avoided under very mild conditions through the cat-
alytic activation and subsequent reductive transmetallation of the
corresponding allyl halide in the presence of 1 mol % Cp2TiCl2 and
unactivated magnesium turnings. The conditions described herein
allow for the in situ allylation of a wide variety of carbonyl deriv-
atives, including aldehydes, ketones, and esters, in excellent yield
within minutes at room temperature. This protocol effectively cir-
cumvents the traditional use of carcinogenic activators and often
exothermic conditions required for Grignard reagent generation.
Further, examination of the scope, mechanistic studies to elucidate
the effect of phosphine additives on the metallation process, and
the development of an enantioselective protocol are in progress,
and will be reported in due course.
Diallylation of estersa
Br
O
2a
OH
11
R1
OR2
R1
Cp2TiCl2 (1 mol%), Mg
THF, rt
10
Entry
Ester
Yield (%)
1
2
3
4
(E)-PhCH@CHCO2Me (10a)
2-Me-C6H4CO2Me (10b)
(E)-4-MeO–C6H4CH@CHCO2Me (10c)
(E)-4-Cl–C6H4CH@CHCO2Me (10c)
96
93
92
97
a
Reaction conditions: 0.4 mmol 10, 1.0 mmol 2a, 1 mol % Cp2TiCl2, and
1.2 mmol Mg in THF (0.1 M) at room temperature for 30–90 min.
In an effort to establish the scope and versatility of the titano-
cene-catalyzed Grignard generation, a series of aldehydes were
examined for their propensity to provide the corresponding homo-
allylic alcohols by treatment with allyl bromide (2a) (Table 2). Both
electron-rich and electron-deficient benzaldehyde derivatives
proved to be excellent substrates providing benzylic alcohols in
high yield (entries 1–7).
a,b-Unsaturated aldehydes underwent
1,2-allylation to yield the corresponding allylic alcohols (entries
8 and 9), and aliphatic aldehydes also proved to be viable sub-
strates (entries 10 and 11).
The titanocene-catalyzed allyl halide activation protocol proved
just as reliable for the allylation of ketones, providing exceptional
yields of the desired homoallylic alcohols (Table 3).13 Benzyl deriv-
atives (electron rich and electron deficient) and
a,b-unsaturated
ketones underwent allylation in excellent yields (entries 1–10).
The allylation of ketone 1d is particularly noteworthy in that the
presence of the free ortho-phenol moiety did not hinder the allyla-
tion event (entry 3). Aliphatic ketones were also viable substrates
as is evidenced by the allylation of 4-t-butylcyclohexanone to pro-
vide alcohol 3l in 94% yield and a syn/anti ratio of 1.3:1 (entry 11).
This diastereomeric outcome is consistent with the addition of an
allyl Grignard reagent and not an allyltitanocene(IV) species.14,15
We next examined the titanocene-catalyzed generation of allyl
Grignard reagents in the presence of ester derivatives (Table 4).
This class of substrates is particularly interesting due to their resis-
tance to nucleophilic attack by allyltitanocene(IV) reagents. A ser-
Acknowledgment
The authors would like to thank the University of Notre Dame
for their generous financial support of this research.
Supplementary data
Supplementary data (detailed experimental procedures and
spectral data for all new compounds) associated with this article
can be found, in the online version, at doi:10.1016/j.tetlet.2010.02.
144.
ies of
a,b-unsaturated esters were examined, and in each case an
References and notes
excellent yield (P92%) of the diallylation product was observed.
Regardless of the substrate or relative amount of allyl halide and
magnesium used, the diallylation products were obtained exclu-
sively. These results lend support to our hypothesis that the active
allylmetal reagent is not an allyltitanocene(IV) derivative, but
rather the more reactive allyl Grignard species.
1. (a) Ashby, E. C.; Laemmle, J. T.; Neumann, H. M. Acc. Chem. Res. 1974, 7, 272; (b)
Ashby, E. C.; Laemmle, J. T. Chem. Rev. 1975, 75, 521; (c) Shirley, D. A. Org. React.
1954, 8, 28; (d) Huryn, D. M. Comp. Org. Syn. 1991, 1, 49; (e) Gradén, H.; Kann,
N. Curr. Org. Chem. 2005, 9, 733.
2. Wakefield, B. J. Organomagnesium Methods in Organic Synthesis; Academic
Press: London, 1995; (b)Handbook of Grignard Reagents; Silverman, G. S., Rakita,
P. E., Eds.; Marcell Dekker: New York, 1996.
3. For recent reviews on allylations, see: (a) Yamamoto, Y.; Asao, N. Chem. Rev.
1993, 93, 2207; (b) Denmark, S. E.; Fu, J. Chem. Rev. 2003, 103, 2763; (c) Bower,
J. F.; Kim, I. S.; Patman, R. L.; Krische, M. J. Angew. Chem., Int. Ed. 2009, 48, 34; (d)
Tietze, L. F.; Kinzel, T.; Brazel, C. C. Acc. Chem. Res. 2009, 42, 367.
4. For reviews on allyl metal carbonyl additions, see: (a) Kennedy, J. W. J.; Hall, D.
G. Angew. Chem., Int. Ed. 2003, 42, 4732; (b) Roush, W. R. Actual. Chim. 2004, 21;
(c) Hall, D. G. Synlett 2007, 1644.
5. (a) Ward, H. R.; Lawler, R. G.; Marzilli, T. A. Tetrahedron Lett. 1970, 11, 521; (b)
Ashby, E. C.; Laemmle, J.; Neumann, H. M. Acc. Chem. Res. 1974, 7, 272; (c)
Ashby, E. C. Pure Appl. Chem. 1980, 52, 545; (d) Maeda, K.; Shinokuno, H.;
Oshima, K. J. Org. Chem. 1996, 61, 6770; (e) Krasovskiy, A.; Knochel, P. Angew.
Chem., Int. Ed. 2004, 3955; (f) Boymond, L.; Rottlander, M.; Cahiez, G.; Knochel,
P. Angew. Chem. 1998, 110, 1801; (g) Rottlander, M.; Boymond, L.; Cahiez, G.;
Knochel, P. J. Org. Chem. 1999, 64, 1080; (h) Inoue, A.; Shinokubu, H.; Oshima, K.
Synlett 1999, 1582; (i) Inoue, A.; Shinokubu, H.; Oshima, K. Org. Lett. 2000, 2,
651.
Table 5
Effect of allyl substitutiona
R3
R3
O
HO Me
Cp2TiCl2 (1 mol%), Mg
THF, rt
R1
Br
+
Ph
Ph
Me
R1 R2
R2
2
3
1a
Entry
Allyl bromide
R1
R2
R3
Yield (%)
1
2
3
4
2b
2c
2d
2e
Me
Me
Ph
H
Me
H
H
H
H
H
Me
91
89b
91b
95
6. (a) Rieke, R. D.; Hudnall, P. M. J. Am. Chem. Soc. 1972, 94, 7178; (b) Burns, T. P.;
Rieke, R. D. J. Org. Chem. 1987, 52, 3674; (c) Rieke, R. D.; Xiong, H. J. Org. Chem.
1991, 56, 3109.
7. Hoffmann, R. W.; Knopff, O.; Kusche, A. Angew. Chem., Int. Ed. 2000, 39, 1462.
8. (a) DeWolfe, R. H.; Young, W. G. Chem. Rev. 1956, 56, 753; (b) Hutchison, D. A.;
Beck, K. R.; Benkeser, R. A.; Grutzner, J. B. J. Am. Chem. Soc. 1973, 95, 7075; (c)
H
a
Allylation reactions were carried out using 0.8 mmol 1a, 0.96 mmol 2a, 1 mol %
Cp2TiCl2, and 1.2 mmol Mg turnings in THF (0.1 M) at room temperature for 20 min.
b
syn/anti = 1:1.

2430
L. M. Fleury, B. L. Ashfeld / Tetrahedron Letters 51 (2010) 2427–2430
Hill, E. A.; Boyd, W. A.; Desai, H.; Darki, A.; Bivens, L. J. Organomet. Chem. 1996,
514, 1.
12. Other phosphines that restore catalytic activity at ꢀ40 °C: triphenylphosphine
(95%), 1,2-bis(diphenylphosphino)ethane (94%), rac-BINAP (82%), tris(o-
tolyl)phosphine (99%), tributylphosphine (95%).
13. Allylation of ketones can be problematic, see: (a) Fürstner, A.; Shi, N. J. Am.
Chem. Soc. 1996, 118, 12349; (b) Bandini, M.; Cozzi, P. G.; Umani-Ronchi, A.
Polyhedron 2000, 19, 537; (c) McManus, H. A.; Cozzi, P. G.; Guiry, P. J. Adv. Synth.
Catal. 2006, 348, 551; (d) Hargaden, G. C.; Müller-Bunz, H.; Guiry, P. J. Eur. J.
Org. Chem. 2007, 4235.
9. (a) Ding, Y.; Zhao, G. J. Chem. Soc., Chem. Commun. 1992, 941; (b) Ding, Y.; Zhao,
G. Tetrahedron Lett. 1992, 33, 8117; (c) Ding, Y.; Zhao, Z.; Zhou, C. Tetrahedron
1997, 53, 2899; (d) Gänsauer, A. Synlett 1998, 801; (e) Szymoniak, J.; Moise, C.
Titanium Zirconium Org. Synth. 2002, 451; (f) Rosales, A.; Oller-López Juan, L.;
Justica, J.; Gänsauer, A.; Oltra, J. E.; Cuerva, J. M. Chem. Commun. 2004, 2628; (g)
Estévez, R. E.; Justica, J.; Bazdi, B.; Fuentes, N.; Paradas, M.; Choquesillo-Lazarte,
D.; Garcia-Ruiz, J. M.; Robles, R.; Gänsauer, A.; Cuerva, J. M.; Oltra, J. E. Chem.
Eur. J. 2009, 15, 2774; (h) Kosal, K.; Ashfeld, B. L. Org. Lett. 2010, 12, 44.
10. Fleury, L. M.; Ashfeld, B. L. Org. Lett. 2009, 11, 5670.
14. (a) Gaudemar, M. Tetrahedron 1976, 32, 1689; (b) Ireland, R. E.; Armstrong, J.
D.; Lebreton, J.; Meissner, R. S.; Rizzacasa, M. A. J. Am. Chem. Soc. 1993, 115,
7152.
11. For the transition metal-catalyzed generation of Grignard reagents, see: (a)
Takai, K.; Ueda, T.; Hayashi, T.; Moriwake, T. Tetrahedron Lett. 1996, 37, 7049;
(b) Bogdanoviae, B.; Schwickardi, M. Angew. Chem., Int. Ed. 2000, 39, 4610.
15. The addition of an allyltitanium complex has been shown to proceed with in a
diastereomeric ration of 2:1: Ireland, R. E.; Armstrong, J. D.; Lebreton, J.;
Meissner, R. S.; Rizzacasa, M. A. J. Am. Chem. Soc. 1993, 115, 7152.