(Chiral) lithium-(magnesium-)zinc and lithium-cobalt combinations as dual reagents for aromatic deproto-metalation and aryl transfer to aldehydes

Article abstract of DOI:10.1016/j.tet.2012.08.008

The deprotonating ability of mixed lithium-zinc or lithium-magnesium-zinc combinations containing amido and alkyl ligands in tetrahydrofuran were compared using anisole as substrate and iodine to quantitatively trap the formed arylmetal species. The results showed that the deprotonating ability is hampered if a Grignard reagent is employed to introduce the alkyl ligand, and is reduced when 2,2,6,6-tetramethylpiperidino ligands are replaced by less hindered/basic chiral amido or alkyls. Concerning the interception of the generated lithium-zinc aryl species by aldehydes, the presence of amido ligands leads to side reactions/lower yields, and no clear improvement was observed if lithium-magnesium-zinc aryl species are used. Racemic mixtures to very low enantioselectivities were noted when chiral amido ligands were incorporated in the composition of the bases. Still with enantioselective aryl transfer to aldehyde as purpose, the deprotonating ability of mixed lithium-cobalt combinations containing amido and alkyl ligands were compared using anisole as substrate and anisaldehyde to trap the formed arylmetal species. As before, the deprotonating ability is reduced when 2,2,6,6-tetramethylpiperidino ligands are replaced by less hindered/basic alkyls or chiral amido. The trapping step using aldehydes being in this case more efficient, even in the presence of amido ligands, the alcohols were obtained in higher yields. With recourse to a lower interception temperature, and using only bis[(R)-1-phenylethyl]amino as ligands, 32 and 22% yield, and 69 and 65% ee were obtained using, respectively, anisaldehyde and 3,4,5-trimethoxybenzaldehyde to intercept the metalated anisole.

Full text of DOI:10.1016/j.tet.2012.08.008

Tetrahedron 68 (2012) 8761e8766
Contents lists available at SciVerse ScienceDirect
Tetrahedron
journal homepage: www.elsevier.com/locate/tet
(Chiral) lithiume(magnesiume)zinc and lithiumecobalt combinations as dual
reagents for aromatic deproto-metalation and aryl transfer to aldehydes
a
a
a
a
b
ꢀ
David Tilly , Katia Snegaroff , Gandrath Dayaker , Floris Chevallier , Philippe C. Gros ,
Florence Mongin a
,
*
ˇ
a
ꢀ
Institut des Sciences Chimiques de Rennes, UMR 6226, CNRS-Universite de Rennes 1, Batiment 10A, Case 1003, Campus Scientifique de Beaulieu, 35042 Rennes, France
SOR, SRSMC, Universite de Lorraine, CNRS, Boulevard des Aiguillettes, 54506 Vandoeuvre-Les-Nancy, France
b
ꢀ
ꢁ
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 14 May 2012
Received in revised form 2 August 2012
Accepted 3 August 2012
Available online 10 August 2012
The deprotonating ability of mixed lithiumezinc or lithiumemagnesiumezinc combinations containing
amido and alkyl ligands in tetrahydrofuran were compared using anisole as substrate and iodine to
quantitatively trap the formed arylmetal species. The results showed that the deprotonating ability is
hampered if a Grignard reagent is employed to introduce the alkyl ligand, and is reduced when 2,2,6,6-
tetramethylpiperidino ligands are replaced by less hindered/basic chiral amido or alkyls. Concerning the
interception of the generated lithiumezinc aryl species by aldehydes, the presence of amido ligands
leads to side reactions/lower yields, and no clear improvement was observed if lithiumemagnesiume
zinc aryl species are used. Racemic mixtures to very low enantioselectivities were noted when chiral
amido ligands were incorporated in the composition of the bases.
Still with enantioselective aryl transfer to aldehyde as purpose, the deprotonating ability of mixed
lithiumecobalt combinations containing amido and alkyl ligands were compared using anisole as sub-
strate and anisaldehyde to trap the formed arylmetal species. As before, the deprotonating ability is
reduced when 2,2,6,6-tetramethylpiperidino ligands are replaced by less hindered/basic alkyls or chiral
amido. The trapping step using aldehydes being in this case more efficient, even in the presence of amido
ligands, the alcohols were obtained in higher yields. With recourse to a lower interception temperature,
and using only bis[(R)-1-phenylethyl]amino as ligands, 32 and 22% yield, and 69 and 65% ee were ob-
tained using, respectively, anisaldehyde and 3,4,5-trimethoxybenzaldehyde to intercept the metalated
anisole.
Keywords:
Bimetallic compounds
Aromatic compounds
Deprotonative metalation
Aryl transfer to aldehyde
Chiral amides
Ó 2012 Elsevier Ltd. All rights reserved.
1. Introduction
Taking account of the reported increased reactivity of zinc
compounds toward aldehydes and ketones in the presence of
magnesium salts (due to coordination of the C]O to Lewis acids),5
Knochel and co-workers recently reinvestigated this possibility in
2010. They notably compared the reactivities of PhZnI$LiCl and
PhZnI$MgCl2$LiCl toward 2-chlorobenzaldehyde, and observed
that the former furnishes the expected alcohol in 60% yield after
72 h at rt, and the latter in 88% yield after only 1 h. The acceleration
effect of MgCl2 was rationalized by a stronger Lewis acidity of MgCl2
compared with that of RZnX, and thus a stronger ability to interact
with the C]O group of the carbonyl reagent in the six-membered
transition state (Scheme 1).6
0
0
The use of bimetal combinations of (R)n(R )n MLi-type, with M
being softer than an alkali-metal (M¼Mg, Al, Cr, Mn, Fe, Co, Cu, Zn,
Cd.), to perform the deprotonative metalation1 of sensitive aro-
matic compounds has recently appeared as an efficient tool.2 These
reagents being in general more compatible with reactive functional
groups than the corresponding monometal lithium bases, one limit
of these methods is the lack of reactivity toward electrophiles, such
as aldehydes and ketones of the generated arylmetal compounds. If
lithiumecobalt basic combinations make the corresponding aryl-
metals more reactive toward carbonylated electrophiles,3 this is far
less obvious when lithiumezinc combinations, especially those
containing amido ligands, are employed.4
Ishihara and co-workers already identified in 2006 the zincate
prepared from EtMgCl and Et2Zn, presumably Et3ZnMgCl, as
a powerful ethylating agent for benzophenone (85% addition and
6% reduction when used in THF (tetrahydrofuran) at 0 ꢀC for 2 h),
and proposed the transition state depicted in Scheme 1.7 The study
was extended in 2010 to various magnesium trialkylzincates, pre-
pared from ZnCl2 and alkylmagnesium chlorides (3 equiv), and to
* Corresponding author. Fax: þ33
2 2323 6955; e-mail address: flor-
ence.mongin@univ-rennes1.fr (F. Mongin).
0040-4020/$ e see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.tet.2012.08.008

8762
D. Tilly et al. / Tetrahedron 68 (2012) 8761e8766
X
R
X
Cl
Mg
R
Zn
X
Zn Cl
O
R1
R
R
R
Zn
R
Mg Cl
Zn
R1
R2
O
R1
R2
O
R2
Scheme 1. Six-membered transition states proposed for the addition of RZnX to R1C(O)R2 in the absence (left) or presence (middle) of MgCl2,6 and for the addition of R3ZnMgCl to
R1C(O)R2 (right).7
different aldehydes and ketones.8 The impact of the cation moiety
Table 1
Deproto-metalation of anisole (1) using in situ prepared lithiume(magnesiume)
zinc combinations (from ZnCl2$TMEDA and metal compounds) followed by trapping
with I2
on alkyl transfer from different Et3ZnM to benzophenone never-
theless proved to be in the order Liþ>ClMgþ>BrMgþ (Scheme 2).9
OMe
OMe
1) Li-(Mg-)Zn combination
THF, rt, 2 h
I
1) ethylating agent
THF, 0 °C, 2 h
O
OH
Ph
Et OH
+
Ph
Ph
2) I2
Ph
Ph
Ph
2) Hydrolysis
1
2a
Et3ZnMgBr:
Et3ZnMgCl:
Et3ZnLi:
54%
81%
94%
39%
14%
1%
Entry
Lie(Mge)Zn combination X: compound(s)
added to ZnCl2$TMEDA
(1 equiv)
Yielda (%)
Scheme 2. Ethylation of benzophenone using different Et3ZnM reagents.9
14
2
3
4
5
6
7
8
9
A: LiTMP (3 equiv)
84b
95, 72b
91
94
74
13
65
32
65
It is worth noting that such magnesiumezinc combinations can
B: Li-endo-S (1 equiv) and LiTMP (2 equiv)
C: Li-exo-S (1 equiv) and LiTMP (2 equiv)
D: Li-exo-R (1 equiv) and LiTMP (2 equiv)
E: Li-endo-S (2 equiv) and LiTMP (1 equiv)
F: Li-endo-S (3 equiv)
generate complex mixtures, as shown by Hevia and co-workers
t
with the reaction between BuMgCl and ZnCl2 in different stoichi-
ometries.10 Consequently, the species involved in the reactions with
carbonyl compounds are difficult to predict.
G: Li-exo-R (2 equiv) and LiTMP (1 equiv)
H: Li-exo-R (3 equiv)
With the aim of synthesizing enantioenriched aryl alcohols, we
recently reported the use of chiral organomagnesates as dual re-
agents for halogen-metal exchange of bromopyridines and sub-
sequent interception by aldehydes.11 We here describe our
investigation on the ability of mixed lithiumezinc (or lith-
iumemagnesiumezinc) and lithiumecobalt combinations con-
taining chiral amido ligands to play a role for aromatic deproto-
metalation and enantioselective aryl transfer to aldehydes.
I: Li-endo-S (2 equiv) then BuLi (1 equiv)
J: Li-exo-R (2 equiv) then BuLi (1 equiv)
K: Li-exo-R (1 equiv) then BuLi (2 equiv)
L1eL4: LiTMP (2 equiv) then RMgCl
10
11
12
47
22
0c
t
(R¼Me, Bu, sBu, Bu, 1 equiv)
13
14
15
M1eM4: LiTMP (2 equiv) then RMgCl
0c
0c
0c
t
(R¼Me, Bu, sBu, Bu, 2 equiv)
N1eN4: LiTMP (1 equiv) then RMgCl
t
(R¼Me, Bu, sBu, Bu, 2 equiv)
O1eO4: LiTMP (1 equiv) then RMgCl
t
(R¼Me, Bu, sBu, Bu, 3 equiv)
2. Results and discussion
a
After purification by column chromatography.
Prepared from 0.5 equiv ZnCl2$TMEDA.
Only anisole was recovered.
b
c
In order to prepare chiral mixed lithium-metal reagents, we
synthesized endo-(1S)-N-(1-phenylethyl)boranamine (H-endo-S)
(formed concomitantly, in
phenylethyl)boranamine
phenylethyl)boranamine (H-exo-R) by adapting a reported pro-
cedure,12 and considered commercial bis[(R)-1-phenylethyl]amine
(H-PEA-R) (Scheme 3).
In order to first evaluate the deprotonating ability of mixed
lithiume(magnesiume)zinc combinations, we chose anisole13 (1)
as substrate and iodine to quantitatively trap the formed arylmetal
species (Table 1). It is important to note that monometal amides,
such as lithium amides are not suitable bases to achieve the
deprotonative lithiation of anisole (1). For example, when the latter
is reacted with LiTMP (TMP¼2,2,6,6-tetramethylpiperidino,
1 equiv) in THF for 2 h at rt, and the lithio compound subsequently
trapped with iodine, the expected 2-iodo derivative 2a is isolated in
a
9:1 ratio, with exo-(1S)-N-(1-
a poor 9% yield.4 Using chiral Li-endo-S, Li-exo-S, Li-exo-R and Li-
PEA-R instead of LiTMP also afforded the iodide 2a but still in low 2,
16, 14 and 0% yield, respectively. The synergy exhibited by mixed
lithium-metal 2,2,6,6-tetramethylpiperidino-based amides as far as
deprotonation efficiencies are concerned led us to turn to this al-
ternative by incorporating chiral amino ligands in the base.
Based on the good deproto-metalating ability of the in situ
prepared mixture of ZnCl2$TMEDA (TMEDA¼N,N,N0,N0-tetrame-
thylethylenediamine) and LiTMP (3 equiv)4,14 (entry 1), heteroleptic
mixtures with LiTMP replaced by a chiral lithium amide were first
considered. The different in situ prepared bases were kept in con-
tact with the substrate in THF for 2 h before interception with
iodine.
(H-exo-S)) and exo-(1R)-N-(1-
H
Ph
N
Ph
H
H
N
H
N
Ph
Ph
Ph
HN
H
H-exo-S
H
H-exo-R
H-endo-S
H-PEA-R
Scheme 3. Chiral amines used in the study.

D. Tilly et al. / Tetrahedron 68 (2012) 8761e8766
8763
In still satisfying yields, it proved possible to replace one TMP
ligand by endo-S (95% yield, entry 2), exo-S (91% yield, entry 3), and
exo-R (94% yield, entry 4). The incorporation of additional ligands
was found deleterious, with 74 and 13% yield by, respectively,
replacing two and three TMP ligands by endo-S (entries 5, 6), and 65
and 32% yield by, respectively, replacing two and three TMP ligands
by exo-R (entries 7, 8). The reactivity of the basic mixtures F and H is
low (13 and 32% yield, respectively, entries 6, 8), barely higher than
that of the corresponding lithium amides.
While 3-chlorothiophene sometimes gave mixtures of the 2-iodo
and 2,5-diiodo derivatives, depending on the basic combination used
(e.g., in a 85:15 ratio using 1 equiv of L4), 2-chlorothiophene (3) only
afforded the 5-iodo derivative 4a. The latter was isolated in 69% yield
using the combination L3, prepared from ZnCl2$TMEDA, LiTMP
(2 equiv) and sBuMgCl (1 equiv) (Table 2, entry 1). This suitable
substrate in hands, we compared the deproto-metalation steps
performed using the combinations L3 (ZnCl2$TMEDAþ2 equiv
LiTMPþ1 equiv sBuMgCl) and L03 (ZnCl2$TMEDAþ2 equiv
LiTMPþ1 equiv sBuLi). Using iodine as electrophile, the magnesium-
free method afforded 4a in 89% yield (entry 2). Using anisaldehyde,
the expected alcohol 4b was isolated in 36 and 48% yield, re-
spectively, again showing the superiority of L03 over L3 for the
deproto-metalation step (entries 3, 4). The lithiumezinc combina-
tion giving a higher overall (deprotonationeinterception) yield, it
was decided to replace the TMP ligand by a chiral one. The combi-
nation P (ZnCl2$TMEDAþ2 equiv Li-PEA-Rþ1 equiv sBuLi) was
employed to this purpose, furnishing 4b in an increased yield (75%)
but as a racemic mixture (entry 5).
We then turned to the electrophilic trapping of the metalated
anisole, generated using a chiral lithiumezinc base (Table 3). After
2 h contact between anisole (1) and the combinations B, C, E, and I,
and subsequent interception with 4-(trifluoromethyl)benzalde-
hyde or anisaldehyde, the expected alcohols 2b,c were isolated in
very low yields, due to side reactions, such as reduction of the al-
dehyde. The incorporation of an increasing number of chiral ligands
in the triamido lithiumezinc bases led to increased degradation
(entries 1e3). An enantioselectivity, albeit very poor, was however
noted using the TMP-free combination I (entries 4e5).
Recent results showed that lithiumecobalt TMP-based mixtures
could be used as reagents to deproto-metalate aromatic compounds,
and that the resulting aryl bimetal species could be trapped by al-
dehydes.3 The outcome of the reactions using lithiumezinc species
drew us to consider cobalt reagents as an alternative to attempt
enantioselective trappings. The ability of various lithiumecobalt
mixtures to deproto-metalate anisole (1) was examined; after 2 h
contact with the base (2 equiv were used in order to discard com-
petitive formation of dimers)3 in THF, anisaldehyde was employed to
perform the quenching step at rt (Table 4). Even if the use of lith-
iumecobalt mixed butyl-TMP combinations already proved to be
unsuccessful, we evaluated the deprotonative ability of the 1:2
Lithium zincates that bear both alkyl and amido groups offer-
ing the possibility of recycling the amido ligand by alkyl-mediated
deprotonation of the formed amine,13b,15 we considered the use of
the lithiumezinc combinations IeK, prepared by successively
adding the chiral lithium amide and butyllithium to
ZnCl2$TMEDA. When compared with the previously used mixed
2:1 TMPebutyl lithiumezinc base,4 the corresponding 2:1 chiral
amido-butyl combination containing endo-S and exo-R proved less
efficient, with 65 and 47% yield (entries 9, 10), respectively, against
79%. Nevertheless, for 1:2 amido-butyl lithiumezinc bases, TMP
and exo-R led to similar results, with yields of 204 and 22% (entry
11), respectively. Surprisingly, when combinations were prepared
s
from ZnCl2$TMEDA, LiTMP (2 equiv), and RMgCl (R¼Me, Bu, Bu,
tBu, 1 or even 2 equiv), 2-iodoanisole (2a) was not observed, but
only recovered anisole (entries 12, 13); a similar result was ob-
tained using mixtures prepared from ZnCl2$TMEDA, LiTMP
(1 equiv) and RMgCl (R¼Me, Bu, sBu, tBu, 2 or even 3 equiv, entries
14, 15).
Since our aim was also to evaluate the trapping step after
deproto-metalation using lithiumemagnesiumezinc combina-
tions, we considered the reaction of more activated substrates.
Upon treatment with all the combinations obtained from
ZnCl2$TMEDA, LiTMP (2 equiv) and RMgCl (R¼Me, Bu, sBu, tBu, 1 or
2 equiv), or from ZnCl2$TMEDA, LiTMP (1 equiv) and RMgCl (R¼Me,
s
t
Bu, Bu, Bu, 2 or 3 equiv), conversions <10% were obtained from
1,3-dimethoxybenzene. When treated similarly, methyl benzoate
only reacted using the mixtures of ZnCl2$TMEDA, LiTMP (2 equiv)
and BuMgCl (2 equiv) (20% conversion), and ZnCl2$TMEDA, LiTMP
(1 equiv) and BuMgCl (3 equiv) (100% conversion); nevertheless,
the expected iodide was not observed but degradation. From 3-
chloropyridine, either starting material or degradation was ob-
served under the same reaction conditions.
Table 2
Deproto-metalation of 2-chlorothiophene (3) using in situ prepared lithiume(magnesiume)zinc combinations (from ZnCl2$TMEDA and metal compounds) followed by
electrophilic trapping
1) Li-(Mg-)Zn combination
THF, rt, 2 h
Cl
H
Cl
E
S
3
S
4
2) Electrophile
3) Hydrolysis
Entry
Lie(Mge)Zn combination X: compound(s)
added to ZnCl2$TMEDA (1 equiv)
Product
Yielda (%)
Eeb (%)
1
2
3
L3: LiTMP (2 equiv) then sBuMgCl (1 equiv)
L03: LiTMP (2 equiv) then sBuLi (1 equiv)
L3: LiTMP (2 equiv) then sBuMgCl (1 equiv)
69
89
36
d
d
d
Cl
I
4a
S
OMe
4b
Cl
S
L03: LiTMP (2 equiv) then sBuLi (1 equiv)
P: Li-PEA-R (2 equiv) then sBuLi (1 equiv)
48
75
d
OH
4
5
3
a
After purification by column chromatography.
Determined by HPLC analysis on a chiral stationary phase (OD-H column, eluent: isopropanol/hexane 8:92, 1 mL/min,
b
l¼252 nm).

8764
D. Tilly et al. / Tetrahedron 68 (2012) 8761e8766
Table 3
combination A synthesized from CoBr2 (2 equiv), and LiTMP
(6 equiv), the results obtained with 6 equiv, 4.5 equiv and 3 equiv of
anisaldehyde were compared. After subsequent hydrolysis, the al-
cohol 2c was isolated in 84, 77, and 59% yield, respectively, and it
was decided to pursue the study using 4.5 equiv of aldehyde.
In the previous experiments, the trapping step was carried out
at rt. In order to check the impact of the temperature on yield and
enantioselectivity, it was performed at ꢂ78 ꢀC, ꢂ55 ꢀC, and ꢂ20 ꢀC.
It was deduced that the reaction does not take place at tempera-
tures below ꢂ50 ꢀC while at ꢂ20 ꢀC, the yield (25%) proved too low
to allow a correct measurement of the enantioselectivity due to the
presence of impurities. It was thus decided to add the electrophile
at a temperature below ꢂ50 ꢀC, and to allow it to warm up slowly
overnight (Scheme 4). Under these conditions, we were pleased to
observe a notable improvement of the enantiomeric excess, which
reached 69% using anisaldehyde and 65% using 3,4,5-
trimethoxybenzaldehyde.
Deproto-metalation of anisole (1) using in situ prepared lithiumezinc combinations
(from ZnCl2$TMEDA and metal compounds) followed by electrophilic trapping
OMe
OMe OH
1) Li-Zn combination
THF, rt, 2 h
2)
OHC
R
1
2b: R = CF3
2c: R = OMe
60 °C
R
3) Hydrolysis
Entry
LieZn combination X: compound(s) added
to ZnCl2$TMEDA (1 equiv)
Product,
eeb,c (%)
yielda (%)
1
2
3
4
5
B: Li-endo-S (1 equiv) and LiTMP (2 equiv)
C: Li-exo-S (1 equiv) and LiTMP (2 equiv)
E: Li-endo-S (2 equiv) and LiTMP (1 equiv)
I: Li-endo-S (2 equiv) then BuLi (1 equiv)
2b, 22
2b, 18
2b, <5
2b, 22
2c, 16
0
0
dd
11
6
a
After purification by column chromatography.
b
Determined by HPLC analysis on a chiral stationary phase (OD-H column, elu-
ent: isopropanol/hexane 5:951 mL/min,
OMe OH
1) CoBr2 (2 equiv)
+ Li-PEA-R (6 equiv)
THF, rt, 2 h
OMe
l¼252 nm).
c
R
The absolute configuration was not determined unambiguously (most of the
studies found in the literature in the diphenylmethanol series are based on HPLC
elution orders and do not give unambiguous assignments).
2)
OHC
R
d
OMe
The determination proved not possible due to impurities.
R
(4.5 equiv)
-50 °C to rt
3) Hydrolysis
1
OMe
2c (R = H): 32%, 69% ee
2d (R = OMe): 22%, 65% ee
R
Table 4
Deproto-metalation of anisole (1) using in situ prepared lithiumecobalt combina-
tions (from CoBr2 and metal compounds) followed by electrophilic trapping
Scheme 4. Deproto-metalation of anisole (1) using an in situ prepared lithiumecobalt
combination (from CoBr2 and Li-PEA-R) followed by aldehyde trapping.
OMe
OMe OH
1) Li-Co combination
THF, rt, 2 h
3. Conclusion
2)
OHC
OMe
Although it did not furnish a method of high interest to per-
form enantioselective trapping of aryl bimetal compounds by al-
dehydes (notably because a large excess of base is necessary to
avoid competitive dimer formation in the case of mixed lith-
iumecobalt bases), this study offered several important discov-
eries concerning the behavior of bimetal species. First, the lower
reactivity of lithiumezinc compounds containing amido ligands
(vs organo) toward aldehydes was confirmed, and the presence of
magnesium instead of lithium in the reaction mixture does not
have a significant impact on this reactivity. Importantly, the
deproto-metalation ability of TMPealkyl lithiumezinc combina-
tions is considerably hampered if the organo group is brought by
a Grignard reagent. To a lesser extent, the replacement of TMP
ligands in lithiumezinc, lithiumecobalt and lithiumeiron bases by
the chiral amido ligands used led to decreased deproto-metalation
abilities. In the case of cobalt, such a yield decrease goes with an
enantioselectivity increase. Finally, the study shows that lith-
iumecobalt systems appear as more promising than the lith-
iumezinc or lithiumeiron ones as far as enantioselective ligand
transfers are concerned.
(6 equiv)
rt, 12 h
3) Hydrolysis
1
2c
OMe
Entry
LieCo combination X: compound(s)
added to CoBr2 (2 equiv)
A00: LiTMP (6 equiv)
Yielda (%)
eeb,c (%)
13
2
3
4
5
84
17
20
60
d
L001: LiTMP (4 equiv) then MeLi (2 equiv)
L004: LiTMP (4 equiv) then tBuLi (2 equiv)
Q00: LiTMP (4 equiv) then Li-PEA-R (2 equiv)
R00: LiTMP (2 equiv) then Li-PEA-R (4 equiv)
S00: Li-PEA-R (6 equiv)
d
d
20
43
40
6
35 (38)d
50 (16)d
a
After purification by column chromatography.
Determined by HPLC analysis on a chiral stationary phase (OD-H column, elu-
ent: isopropanol/hexane 5:95, 1 mL/min,
The absolute configuration was not determined unambiguously (most of the
studies found in the literature in the diphenylmethanol series are based on HPLC
elution orders and do not give unambiguous assignments).
b
l¼252 nm).
c
d
Using the LieFe combination prepared from FeBr2 (1 equiv) and Li-PEA-R
(3 equiv).
methyl-TMP and tert-butyleTMP mixtures for which b-elimination
is prevented; the expected alcohol 2c was isolated in yields (ꢁ20%,
entries 2, 3) clearly below that of 84% (entry 1) observed using the
‘all-TMP’ base. The incorporation of bis[(R)-1-phenylethyl]amino
(PEA-R) in lieu of TMP in the base led to decreased yields, showing
the superiority of TMP as deprotonating ligand, but to increased
enantioselectivities (entries 4e6). The higher enantioselectivity was
noted after replacement of all the TMP ligands; if iron is used in place
of cobalt in the combination S00,2f a similar yield but clearly lower ee
was observed (entry 6).
4. Experimental section
4.1. General procedure for the metalation using the in situ
prepared mixtures followed by iodination
After the substrate (2.0 mmol) has been added to the in situ
prepared mixture (see Supplementary data) at 0 ꢀC, the mixture
was stirred for 2 h at rt before introduction of a solution of I2 (1.5 g,
6.0 mmol) in THF (5 mL). The mixture was stirred overnight before
addition of an aqueous saturated solution of Na2S2O3 (20 mL) and
extraction with CH2Cl2 (3ꢃ20 mL). The combined organic layers
were dried over Na2SO4, filtered, and concentrated under reduced
Isolation issues of the expected alcohol 2c from the reaction
mixture, due to competitive degradation of the aldehyde, led us to
consider the use of lower amounts of electrophile. Using the

D. Tilly et al. / Tetrahedron 68 (2012) 8761e8766
8765
pressure before being purified by means of column chromatogra-
phy on silica gel.
reduced pressure before being purified by means of column chro-
matography on silica gel.
2-Iodoanisole (2a) was prepared from anisole (1, 0.22 mL) and
(2-Methoxyphenyl)(4-methoxyphenyl)methanol (2c) was prepared
using anisaldehyde (0.55 mL), and was isolated (eluent: heptane/Et2O
100/0 to 65/35) as a pale yellow oil (see above). The ee was de-
termined by HPLC analysis (main enantiomer at 16.7 min and minor
enantiomer at 18.3 min) on a chiral stationary phase (OD-H column,
was isolated (eluent: heptane/AcOEt 80/20) as a yellow oil: 1H NMR
(300 MHz, CDCl3)
d
3.88 (s, 3H), 6.72 (dd, 1H, J¼7.5 and 1.4 Hz), 6.83
(dd, 1H, J¼8.3 and 1.3 Hz), 7.31 (ddd, 1H, J¼8.3, 7.4 and 1.6 Hz), 7.77
(dd, 1H, J¼7.8 and 1.6 Hz); 13C NMR (75 MHz, CDCl3)
d 56.3, 86.0,
111.0, 122.5, 129.6, 139.5, 158.1. These data are analogous to those
eluent: isopropanol/hexane 10:90,1 mL/min,
l
¼252 nm). The 69% ee
previously described.16
obtained corresponds to [
a
]D¼þ14.1ꢀ (c 0.05, CHCl3, 20 ꢀC, 589 nm).
2-Chloro-5-iodothiophene
(4a)
was
prepared
from
(2-Methoxyphenyl)(3,4,5-trimethoxyphenyl)methanol (2d) was
prepared using trimethoxybenzaldehyde (0.88 g), and was isolated
(eluent: heptane/Et2O 100/0 to 65/35) as a yellow oil: 1H NMR
2-chlorothiophene (3, 0.18 mL) using L3 or L03 and was isolated
(eluent: heptane) as a yellow oil: 1H NMR (300 MHz, CDCl3)
d 6.61
(d, 1H, J¼4.1 Hz), 7.06 (d, 1H, J¼3.8 Hz); 13C NMR (75 MHz, CDCl3)
(300 MHz, CDCl3)
1H), 6.64 (s, 2H), 6.90e6.97 (m, 2H), 7.17 (dd, 1H, J¼7.5 and 1.8 Hz),
7.26e7.30 (m, 1H); 13C NMR (75 MHz, CDCl3)
55.6, 56.1 (2C), 60.9,
d 3.82 (s, 6H), 3.84 (s, 3H), 3.86 (s, 3H), 6.01 (s,
d
71.6, 129.1, 134.6, 137.6. These data are similar to those previously
described.17
d
72.2, 103.7 (2C), 110.8, 121.0, 127.9, 129.0, 131.9, 137.0, 138.8, 153.1
4.2. General procedure for the metalation using the in situ
prepared mixtures followed by trapping by aldehydes
(2C), 156.8; HRMS (ESI) calcd for C17H20NaO5 [(MþNa)þꢄ] and
C17H19Na2O5 [(MꢂHþ2Na)þ
] 327.1208 and 349.1027, found
ꢄ
327.1206 and 349.1042, respectively. The ee was determined by
HPLC analysis (main enantiomer at 19.4 min and minor enantiomer
at 33.2 min) on a chiral stationary phase (OD-H column, eluent:
After the substrate (2.0 mmol) has been added to the in situ
prepared mixture (see Supplementary data) at 0 ꢀC, the mixture
was stirred for 2 h at rt before introduction of the required aldehyde
(6.0 mmol). The mixture was stirred for 2 h at 60 ꢀC before addition
of brine (20 mL) and extraction with Et2O (3ꢃ20 mL). The combined
organic layers were dried over Na2SO4, filtered, and concentrated
under reduced pressure before being purified by means of column
chromatography on silica gel.
isopropanol/hexane 6:94, 1 mL/min,
l
¼252 nm). The 65% ee ob-
tained corresponds to [
a
]D¼þ45ꢀ (c 0.05, CHCl3, 20 ꢀC, 589 nm).
Acknowledgements
All the authors gratefully acknowledge the Agence Nationale de
la Recherche (ACTIVATE program, financial support to D.T., K.S. and
G.D.) and F.M. thanks the Institut Universitaire de France.
5-Chloro-a-(4-methoxyphenyl)-2-thiophenemethanol (4b) was
prepared from 2-chlorothiophene (3, 0.18 mL) using L3, L03 or P,
with subsequent trapping by anisaldehyde (0.74 mL), and was
isolated (eluent: heptane/AcOEt 80/20) as a yellow oil: 1H NMR
(300 MHz, CDCl3)
d
2.44 (d, 1H, J¼3.8 Hz), 3.72 (s, 3H), 5.75 (d, 1H,
Supplementary data
J¼3.8 Hz), 6.45 (d, 1H, J¼3.8 Hz), 6.60 (d, 1H, J¼3.8 Hz), 6.83 (d, 2H,
J¼8.3 Hz), 7.25 (d, 2H, J¼8.3 Hz); 13C NMR (75 MHz, CDCl3)
d 55.2,
Supplementary data associated with this article can be found in
the online version, at http://dx.doi.org/10.1016/j.tet.2012.08.008.
71.9, 113.8, 123.7, 125.5, 127.5, 129.6, 134.6, 147.1, 159.3. These data
are similar to those previously described.17
(2-Methoxyphenyl)[4-(trifluoromethyl)phenyl]methanol (2b) was
prepared from anisole (1, 0.22 mL) using B, C, E or I, with sub-
sequent trapping by 4-(trifluoromethyl)benzaldehyde (0.87 mL),
and was isolated (eluent: heptane/Et2O 100/0 to 70/30) as a yellow
References and notes
oil: 1H NMR (300 MHz, CDCl3)
d
3.13 (d, 1H, J¼5.7 Hz), 3.82 (s, 3H),
1. (a) Gschwend, H. W.; Rodriguez, H. R. Org. React. 1979, 26, 1e360; (b) Beak, P.;
Snieckus, V. Acc. Chem. Res. 1982, 15, 306e312; (c) Snieckus, V. Chem. Rev. 1990,
90, 879e933; (d) Gant, T. G.; Meyers, A. I. Tetrahedron 1994, 50, 2297e2360; (e)
Schlosser, M. Organometallics in Synthesis In. Schlosser, M., Ed., 2nd ed.; Wiley:
2002, Chapter 1.
2. (a) Mulvey, R. E. Organometallics 2006, 25,1060e1075; (b) Mulvey, R. E.; Mongin,
F.; Uchiyama, M.; Kondo, Y. Angew. Chem., Int. Ed. 2007, 46, 3802e3824; (c)
Mulvey, R. E. Acc. Chem. Res. 2009, 42, 743e755; (d) Haag, B.; Mosrin, M.; Ila, H.;
Malakhov, V.; Knochel, P. Angew. Chem., Int. Ed. 2011, 50, 9794e9824; (e) Mongin,
F.; Uchiyama, M. Curr. Org. Chem. 2011, 15, 2340e2361; (f) Nagaradja, E.;
Chevallier, F.; Roisnel, T.; Jouikov, V.; Mongin, F. Tetrahedron2012, 68, 3063e3073.
3. Dayaker, G.; Chevallier, F.; Gros, P. C.; Mongin, F. Tetrahedron 2010, 66,
8904e8910.
4. See for example the following reference: Snegaroff, K.; Komagawa, S.; Chevallier,
F.; Gros, P. C.; Golhen, S.; Roisnel, T.; Uchiyama, M.; Mongin, F. Chem.dEur. J. 2010,
16, 8191e8201.
5. Marx, B.; Henry-Basch, E.; Freon, P. C. R. Hebd. Seances Acad. Sci. 1967, 264,
527e530.
6. (a) Metzger, A.; Bernhardt, S.; Manolikakes, G.; Knochel, P. Angew. Chem., Int. Ed.
2010, 49, 4665e4668; (b) Schade, M. A.; Manolikakes, G.; Knochel, P. Org. Lett.
2010, 12, 3648e3650.
6.08 (d, 1H, J¼5.4 Hz), 6.91 (d, 1H, J¼8.2 Hz), 6.97 (td, 1H, J¼7.5 and
0.9 Hz), 7.22 (dd, 1H, J¼7.5 and 1.6 Hz), 7.30 (m, 1H), 7.51 (d, 2H,
J¼8.2 Hz), 7.58 (d, 2H, J¼8.3 Hz); 13C NMR (75 MHz, CDCl3)
d 55.4,
71.8,110.8,120.9, 124.2 (q, JF¼272 Hz),125.1 (q, 2C, JF¼3.8 Hz),126.7,
127.8, 129.1, 129.2 (q, 2C, JF¼32.3 Hz), 131.2, 147.3, 156.6. These data
are similar to those previously described.18
(2-Methoxyphenyl)(4-methoxyphenyl)methanol (2c) was pre-
pared from anisole (1, 0.22 mL) using I, and was isolated (eluent:
heptane/Et2O 100/0 to 65/35) as a pale yellow oil: 1H NMR
ꢀ
(300 MHz, CDCl3)
6.02 (d, 1H, J¼5.0 Hz), 6.88 (m, 3H), 6.95 (td, 1H, J¼7.5 and 1.0 Hz),
7.28 (m, 4H); 13C NMR (75 MHz, CDCl3)
55.2, 55.4, 71.8, 110.7, 113.5
d
3.02 (d, 1H, J¼5.2 Hz), 3.79 (s, 3H), 3.81 (s, 3H),
ꢀ
ꢀ
d
(2C), 120.7, 127.6, 127.8 (2C), 128.5, 132.1, 135.5, 156.6, 158.7. These
data are similar to those previously described.4
7. Hatano, M.; Suzuki, S.; Ishihara, K. J. Am. Chem. Soc. 2006, 128, 9998e9999.
8. Hatano, M.; Suzuki, S.; Ishihara, K. Synlett 2010, 321e324.
9. Hatano, M.; Ito, O.; Suzuki, S.; Ishihara, K. J. Org. Chem. 2010, 75, 5008e5016.
4.3. General procedure for the metalation using the in situ
prepared mixtures followed by trapping by aldehydes
ꢀ
10. Hevia, E.; Chua, J. Z.; García-Alvarez, P.; Kennedy, A. R.; McCall, M. D. Proc. Natl.
Acad. Sci. U.S.A. 2010, 107, 5294e5299.
After anisole (1, 109 mL, 1.0 mmol) has been added to the in situ
11. (a) Catel, D.; Chevallier, F.; Mongin, F.; Gros, P. C. Eur. J. Org. Chem. 2012, 2012,
53e57; (b) Catel, D.; Payen, O.; Chevallier, F.; Mongin, F.; Gros, P. C. Tetrahedron
2012, 68, 4018e4028.
12. Cain, C. M.; Cousins, R. P. C.; Coumbarides, G.; Simpkins, N. S. Tetrahedron 1990,
46, 523e544.
13. Concerning the reaction of anisole with mixed lithiumezinc bases, see Ref.4
and the following references: (a) Clegg, W.; Dale, S. H.; Drummond, A. M.;
Hevia, E.; Honeyman, G. W.; Mulvey, R. E. J. Am. Chem. Soc. 2006, 128,
prepared mixture S00 (see Supplementary data) at 0 ꢀC, the mixture
was stirred for 2 h at rt before introduction of the required aldehyde
(4.5 mmol) at ꢂ50 ꢀC. The mixture was stirred overnight with slow
warming up of the reaction mixture before addition of brine
(20 mL) and extraction with Et2O (3ꢃ20 mL). The combined organic
layers were dried over Na2SO4, filtered, and concentrated under

8766
D. Tilly et al. / Tetrahedron 68 (2012) 8761e8766
7434e7435; (b) Clegg, W.; Conway, B.; Hevia, E.; McCall, M. D.; Russo, L.;
16. L’Helgoual’ch, J. M.; Bentabed-Ababsa, G.; Chevallier, F.; Yonehara, M.;
Uchiyama, M.; Derdour, A.; Mongin, F. Chem. Commun. 2008, 5375e5377.
17. Bayh, O.; Awad, H.; Mongin, F.; Hoarau, C.; Trecourt, F.; Queguiner, G.; Marsais,
Mulvey, R. E. J. Am. Chem. Soc. 2009, 131, 2375e2384.
ꢀ
ꢀ
14. García-Alvarez, P.; Mulvey, R. E.; Parkinson, J. A. Angew. Chem., Int. Ed. 2011, 50,
9668e9671.
F.; Blanco, F.; Abarca, B.; Ballesteros, R. Tetrahedron 2005, 61, 4779e4784.
18. Snegaroff, K.; L’Helgoual’ch, J. M.; Bentabed-Ababsa, G.; Nguyen, T. T.;
ꢀ
15. Kondo, Y.; Morey, J. V.; Morgan, J. C.; Naka, H.; Nobuto, D.; Raithby, P. R.;
Uchiyama, M.; Wheatley, A. E. H. J. Am. Chem. Soc. 2007, 129,
12734e12738.
Chevallier, F.; Yonehara, M.; Uchiyama, M.; Derdour, A.; Mongin, F. Chem.dEur.
J. 2009, 15, 10280e10290.