A combined experimental-theoretical study on the lithiation/electrophilic quench reaction of benzylic position of (η5- tetramethylcyclohexadienyl)-Mn(CO)3

Article abstract of DOI:10.1039/c1nj20178f

An [(η⁶-1,2,4,5-tetramethylbenzene)Mn(CO)₃] ⁺ BF₄^- complex treated with hydride gives a neutral (η⁵-1,2,4,5-tetramethylcyclohexadienyl)Mn(CO)₃ complex whose X-ray analysis is described. Functionalization of the latter complex at a "benzylic" position using a lithiation/electrophilic quench sequence affords two regioisomers of which the major one corresponds to the functionalization at the C9 carbon. Theoretical computations show that the regioselectivity of lithiation is governed by the stability of the carbanion and thus by its conjugation.

Full text of DOI:10.1039/c1nj20178f

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
NJC
Cite this: New J. Chem., 2011, 35, 2375–2381
www.rsc.org/njc
PAPER
A combined experimental–theoretical study on the lithiation/electrophilic
quench reaction of benzylic position of (g5-tetramethylcyclohexadienyl)-
Mn(CO)3wz
a
b
Eric Rose,*a Julien Dubarle-Offner,a Franc
¸ oise Rose-Munch and Helene Gerard*
´
´
Received (in Montpellier, France) 28th February 2011, Accepted 16th August 2011
DOI: 10.1039/c1nj20178f
ꢀ
An [(Z6-1,2,4,5-tetramethylbenzene)Mn(CO)3]+ BF4 complex treated with hydride gives a
neutral (Z5-1,2,4,5-tetramethylcyclohexadienyl)Mn(CO)3 complex whose X-ray analysis is
described. Functionalization of the latter complex at a ‘‘benzylic’’ position using a lithiation/
electrophilic quench sequence affords two regioisomers of which the major one corresponds to the
functionalization at the C9 carbon. Theoretical computations show that the regioselectivity of
lithiation is governed by the stability of the carbanion and thus by its conjugation.
Introduction
The presence of the electron-deficient Cr(CO)3,1 [Mn(CO)3+]2 or
[(C5H5)Fe+]3 fragment attached to an arene ring decreases the
electronic density of the arene and enhances the electrophilic
character of the ligand as well as the acidity of its protons. Thus,
addition of a nucleophile Nuꢀ to neutral (Z6-arene)Cr(CO)3
+
and isoelectronic cationic (Z6-arene)-Mn(CO)3 or -FeCp+
complexes leads toꢀth1ae,1cformation of anionic [(Z5-6-Nu-cyclo-
hexadienyl)Cr(CO)3
]
which can be stabilized by trapping
with ClSnPh3 (Scheme 1a),4 and of stable neutral exo-substituted
(Z5-6-Nu-cyclohexadienyl)-[M] complexes [M] = Mn(CO)3 or
-Fe(C5H5)2 (Scheme 1b). The chemistry of Cr5 and Mn6 complexes
has been extensively developed by our group which opened an
access to a wide range of organometallic molecules.
Scheme 1 Addition of Nuꢀ to (a) (Z6-arene)Cr(CO)3 complexes,
(b) M = Mn(CO)3+, Fe(C5H5)+
.
Recently, we wondered whether the lithiation/electrophilic
quench sequence method which allows functionalization of
the ring of (Z5-cyclohexadienyl)Mn(CO)3 complexes6c could
be applied to substitute ‘‘benzylic’’ positions, i.e. the positions alpha
to the Z5-p system. Indeed, the study of the benzylic position has
already been reported for the cationic [(Z6-alkylarene)Mn(CO)3]+
complexes7 and has been well developed in the (Z6-arene)Cr(CO)3
series8 as well as in the [(Z6-alkylarene)(Z5C5H5)]Fe+ series.9
Deprotonation of organometallic complexes at a benzylic position
of [(Z6-alkylarene)Mn(CO)3]+ complexes readily occurred
using a large excess of a strong base.7e,7f The authors observed
the formation of (Z5-pentamethylbenzyl)Mn(CO)3 complexes
presenting
a
highly activated exocyclic double bond
a
(Scheme 2).
´
UPMC Univ Paris 06, Institut Parisien de Chimie Moleculaire
IPCM, CNRS UMR 7201, Equipe de Chimie Organique et
Very recently, we reported a lithiation/electrophilic quench
sequence at a ‘‘benzylic’’ position of (Z5-tetramethyl) and (Z5-
pentamethyl)-6-exo-phenylcyclohexadienyl)Mn(CO)3 complexes 3
´
ˆ
Organometallique, Bat. F, Case 181, 4, Place Jussieu, F-75252 Paris
Cedex 05, France. E-mail: eric.rose@upmc.fr;
Tel: +33 (0)144276235
and 4 obtained by addition of PhMgBr to (Z6-tetramethyl)- and
b
´
UPMC Univ Paris 06, Laboratoire de Chimie Theorique, CNRS
´
UMR 7616, Tour 12-13, 4e`me etage, Case 137, 4, Place Jussieu,
F-75252 Paris Cedex 05, France. E-mail: helene@lct.jussieu.fr
+
(Z6-pentamethyl)-benzeneMn(CO)3
(Scheme 3).
complexes 1 and 2
w Dedicated to Prof. Didier Astruc on the occasion of his 65th
birthday.
In both cases, as evidenced by NMR and X-ray analyses, we
observed a regioselective deprotonation at the methyl groups
C9 or C10 (Scheme 3) substituting the C2 or C4 carbons of the
cyclohexadienyl ring eclipsed by a Mn–CO bond by trapping
the lithiated product with different electrophiles (ClSiMe3,
PPh2Cl and DMF).10
z Electronic supplementary information (ESI) available: Tables listing
detailed X-ray crystallographic data of complex 7 and Cartesian
coordinates and electronic energies for compounds 7 and 8 in neutral,
deprotonated and lithiated forms. CCDC reference number 812616.
For ESI and crystallographic data in CIF or other electronic format
see DOI: 10.1039/c1nj20178f
c
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2011
New J. Chem., 2011, 35, 2375–2381 2375

View Article Online
Scheme 2 Deprotonation of [(Z6-alkylarene)Mn(CO)3]+ complexes.
Fig. 1 Molecular structure of complex 7 with thermal ellipsoids at
the 50% probability level. Hydrogen atoms are omitted for clarity.
Selected bond lengths (A): Mn–C1: 2.2218(16); Mn–C2: 2.1536(15);
Mn–C3: 2.1145(15); Mn–C4: 2.1575(15); Mn–C5: 2.12335(16);
Mn–C6: 2.7445(19); C1–C2: 1.396(2); C2–C3: 1.414(2); C3–C4:
1.419(2); C4–C5: 1.389(2); C5–C6: 1.506(2).
Table 1 Crystallographic data and details of the refinement of the
structure of 7
Scheme 3 Lithiation of (Z5-tetramethyl)- and (Z5-pentamethylcyclo-
hexadienyl)tricarbonylmanganese complexes.
Complex 7
Empirical formula
Formula weight
Temperature/K
Wavelength/A
Crystal system
Space group
a/A
C13H15MnO3
274.19
200(2)
0.71073
Monoclinic
P21/c
11.887(1)
b/A
c/A
a/1
9.228(2)
12.585(3)
90
Scheme 4 Theoretically studied complexes.
Here we report on the synthesis and structure of the (Z5-
1,2,4,5-tetramethylcyclohexadienyl)Mn(CO)3 complex 7 bearing
two hydrogen atoms at the sp3 carbon and the study of its
deprotonation. The results obtained are then analyzed on a
theoretical basis using computations on the lithiation products
of compound 7 compared to those of pentamethyl derivative 8
and non-methylated compound 9 (Scheme 4). Steric effects due
to the Ph group substituting H at C6 in 7 (yielding, respectively,
compound 3) are also examined (see below).
b/1
108.932(7)
90
1305.8(4)
4
1395
g/1
V/A3
Z
Dcalculated/mg mꢀ3
m/mmꢀ1
F(000)
1.004
568
Crystal size/mm
Diffractometer
y range/1
Limiting indices
0.3 ꢁ 0.15 ꢁ 0.1
k-CCD
3.62 to 30.00
ꢀ16 r h r 13
ꢀ12 r k r 8
ꢀ17 r l r 17
10505/3772[R(int) = 0.0303]
Full matrix square on F2
3772/0/158
1.05
Reflections collected/unique
Refinement method
Data/restraints/parameters
Goodness-of-fit on F2
Results and discussion
The [(Z6-1,2,4,5-tetramethylbenzene)Mn(CO)3]+ BF4ꢀ complex 1,
obtained in 90% yield usꢀing the ‘‘silver (I) method’’ with
BrMn(CO)5 and AgBF4 and the corresponding free
arene,10,11 treated with LiAlH4 afforded the (Z5-tetramethyl-
cyclohexadienyl)Mn(CO)3 complex 7 in 66% yield. Pleasingly,
monocrystals of complex 7 were readily obtained by slow
evaporation of a diethylether/pentane solution of the complex.
The ORTEP view shown in Fig. 1 indicates an Z5-cyclohexa-
dienyl structure, with five coplanar sp2 carbons and the
remaining sp3 carbon located 42.6(1)1 above this plane. The
conformation of the Mn(CO)3 tripod is in agreement with
what is usually observed, the sp3 carbon being eclipsed by one
Mn–CO bond.2f,2g,6f The Mn–C bond lengths lie in the range
2.1145(15) to 2.2235(16) A with the Mn–C3 bond being the
shortest one. Crystal data are gathered in Table 1.
Final R indices [I 4 2s(I)]
R1 = 0.0333
wR2 = 0.0727
R1 = 0.0593
wR2 = 0.0801
0.297 and ꢀ0.234
812616
R indices (all data)
Largest difference in peak/hole [e Aꢀ3
CCDC deposition number
]
derivatives 3 and 4.10 It appeared that deprotonation of complex
7 could be achieved at low temperatures and that the
corresponding ‘‘benzylic’’ carbanion was stable up to ꢀ20 1C.
The best conditions for this lithiation/electrophilic quench se-
quence correspond to the following protocol. Introduction of
4 equivalents of nBuLi and TMEDA at ꢀ40 1C in solution of
complex 7 in THF, 1 h at ꢀ40 1C, then introduction of the
electrophile, again 1 h at ꢀ40 1C followed by warming the
reaction mixture to rt within 1 h. Chlorotrimethylsilane and
chlorodiphenylphosphine were tested as electrophiles. Chromato-
graphy on silica gel afforded in each case two mono-function-
alized regioisomers at the C7 and C9 carbons in the 3 : 1 ratio
A lithiation/electrophilic quench sequence was performed on
(Z5-1,2,4,5-tetramethylcyclohexadienyl)Mn(CO)3 complex
under the same experimental conditions as those for deprotona-
7
tion of benzylic positions of Z5-6exo-Ph-methylcyclohexadienyl
c
2376 New J. Chem., 2011, 35, 2375–2381
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2011

View Article Online
Table 3 Energies (in kcal molꢀ1, computed with respect to the cis
product with Li at the C7 benzylic carbon) of the lithiated regio-
isomers at the C7, C9 and C3 carbons of 7 and at C7, C9 and C11 of 8
for the two possible spatial arrangements as well as energies computed
with respect to the anionic product deprotonated at the C7 benzylic
carbon of the anionic form optimized in the absence of any Li+
counterion. Energies in parentheses are computed in the presence of a
continuum model for solvation by THF
Regio.
C7
C9
C3
C11
7
8
Trans
Cis
Anion
Trans
Cis
11.7 (6.8)
0.0 (0.0)
0.0
12.5
0.0
16.3 (14.7)
10.5 (12.1)
12.4
17.4
10.7
19.2 (20.2)
—
—
—
21.0
16.2
16.7
36.7
—
—
—
Anion
0
12.2
center12 and, coherently, the smaller the CO vibrational
frequencies. This highlights the noticeable donor effect of the
Me groups toward the cyclohexadienyl unit and thus to the
manganese atom.
Reaction products of these three complexes toward lithia-
tion were then examined. In order to represent to the best the
experimentally formed species, benzylic H atoms number 7, 9
and 11, Scheme 4, were replaced by a Li cation coordinated to
a Me2N–CH2–CH2–NMe2 ligand (TMEDA) so that both
steric and electronic effects are evaluated. Under these condi-
tions, no additional molecule can be added to lithium due to
the steric hindrance of the TMEDA ligand so that further
explicit solvation of the molecule is necessary.
Scheme 5 Lithiation/electrophilic quench sequence on complex 7.
as evidenced by NMR spectroscopy. These new complexes 10,
11 and 12, 13, which could not be separated, were isolated in
73 and 82% yields respectively (Scheme 5). As previously
observed, no product due to the lithiation of the cyclo-
hexadienyl ring at the C3 carbon was formed.
The origin of the experimentally observed regioselectivity
was examined using theoretical computations on the lithiation
products (Schemes 3 and 4) of tetramethyl derivatives 3 and 7.
Comparison to the pentamethylated derivatives 4 and 8 and to
the compound without the methyl group 9, which was examined
in our previous studies,6d will be proposed for better
understanding.
When lithiation takes place at a benzylic carbon, two coordi-
nation modes can be proposed for the lithium cation: either
under the cyclohexadienyl plane, cis to the Mn(CO)3 moiety, or
above the cyclohexadienyl unit trans to the Mn(CO)3 tripod.
When lithiation takes place at the arylic position (C3 in 7), the
lithium cation lies within the cyclohexadienyl plane and only one
isomer is obtained. Both arrangements were computed in the
case of 7 and 8 for the three possible regioisomers and results are
gathered in Table 3, represented in Fig. 2 and Cartesian coordi-
nates are given in ESI.z
In a first step, the three complexes 7, 8 and 9 were geometrically
optimized and their geometric, energetic and electronic properties
were examined to evaluate the effect of the Me substituents on
the geometrical properties of these species (Table 2 and
Cartesian coordinates in ESIz). In all three cases, the distances
of the Mn center to the cyclohexadienyl carbon atoms decrease
as the carbon is further from the sp3 center; in addition, the
distances for a given carbon lie within 0.03 A between the three
complexes, which highlights the small effect of the Me groups
on the geometrical properties of the cyclohexadienyl moiety.
In contrast, a clear cut electronic effect is evidenced using
charge and frequency analysis. The larger the number of
methyl groups the more negative the charge at the Mn(CO)3
Considering the cis/trans spatial arrangement of the lithium
binding, it appears that the lithiation product cis to the
Mn(CO)3 moiety is always more stable than the trans one.
In contrast, shorter Li–C bonds are observed for the trans
conformers (between 2.08 and 2.18 A) than for the cis
conformers (between 2.21 and 2.34 A). This side preference
is found to be unaffected by the adjunction of a continuum
model for solvation by THF. This conformation preference
can be easily understood on the basis of the electrostatic
properties of the complexes 7 and 8, as illustrated by the
dipole moment and the electrostatic potential associated with
complex 8 (see Fig. 3): the side of the ring trans to the
Mn(CO)3 moiety is the more positively charged13 and on the
positive side of the dipole so that binding of the lithium cation
is clearly preferred cis to the Mn(CO)3 moiety.
Table 2 Mn–C distances (in A), Mulliken (qM) charges of the
¨
Mn(CO)3 moiety (in electrons) and the three harmonic nCO vibrational
frequencies (in cmꢀ1, repeated when equal from degeneracy) for three
of the complexes described in Scheme 5
7
8
9
Mnꢂ ꢂ ꢂC1
Mnꢂ ꢂ ꢂC2
Mnꢂ ꢂ ꢂC3
qM(Mn(CO)3)
nCO
2.282
2.200
2.144
ꢀ1.34
2008
2010
2072
2.267
2.193
2.170
ꢀ1.44
2005
2006
2068
2.265
2.172
2.158
ꢀ0.88
2024
2024
2087
Transfer of this side preference from the H substituted 7 and
8 to 3 and 4, where C6 is substituted by R0 = Ph instead of H,
is straightforward as the less stable conformer (trans) is also
more destabilized by the steric hindrance due to the Ph
substituent. Quantitative evaluation of steric effects between
the R0 group and the Li(TMEDA) moiety is nevertheless
c
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2011
New J. Chem., 2011, 35, 2375–2381 2377

View Article Online
Fig. 3 Electrostatic properties of 8. (a) Electrostatic potential
mapped on the electron density isosurface (isodensity value = 0.04,
more positive regions in blue and more negative regions in red).
(b) Dipole moment direction (carbon atoms in green, hydrogen in
white, oxygen in red and manganese in blue).
Table 4 Effect of the substituent R0 at C6 on the relative energies of
the lithiation products of 3 and 7 in the trans arrangement (computed
with respect to the product of lithiation of Li at the C7 carbon in the
trans form, in kcal molꢀ1
)
R0
Li at C7
Li at C9
Li at C3
7
3
H
Ph
0.0
0.0
+4.6
+3.2
+7.5
+7.6
the arylic position, for which only one arrangement is possible
(lithium bonded in the plane), and little steric effects between R0
and the Li(TMEDA) moiety are expected. As little or no change
in the energy difference between Li at the carbon C7 and Li at the
carbon C3 or C9, Table 4, can be observed upon transformation
of a R0 = phenyl into a R0 = hydrogen in the trans products, it
can be concluded that there is no significant steric effect between
Li(TMEDA) and the phenyl substituent located at C6. As no
effect of R0 is found in the case of the trans product and as no
effect is expected in the cis ones, the change in the regioselectivity
of the reaction observed experimentally when replacing a phenyl
group by a hydrogen cannot be justified by the difference in the
stability of the lithiation product.
Lithiated products in the cis arrangement will thus be
considered in the rest of our study. The relative stabilities of
the three regioisomers are the following: Li7 o Li9 o Li3
(where o stands for ‘‘more stable’’). This order is found to be
unaffected by inclusion of implicit solvation in the case of
compound 7 (see Table 3). The product lithiated in position 7
is preferred by over 10 kcal molꢀ1, which should be associated
to a fully selective lithiation in this position if this intermediate
is the source for regioselectivity. As this preference is expected
to be independent of the nature of the substituent at C6 (see
above), recovery of products substituted in position 7 should
be favoured under thermodynamic control of the lithiation step
for both 7 and 8. Comparison of experimental results gives
interesting insights within these results. Indeed, from compu-
tational data, lithiation on the benzylic carbon 7 or 8 should
always be preferred, either for 3 and 7 or for 4 and 8. As the
energy difference between the most stable intermediate and the
other possible one is large, we can trust the computationally
predicted selectivity and thus propose that recovery of products
substituted on the carbon 9 or 10 results from a kinetically
controlled lithiation sequence. These kinetic products are the sole
Fig. 2 Cis and trans spatial arrangements of the regioisomers of
Table 3 (a) 7, (b) 8 (energies in kcal molꢀ1 in parentheses).
obtained. This is done by examining the stabilities of the two
regioisomers in the trans arrangement (Li at C7 and C9)
relative to that of the regioisomer with Li at C3 in compounds
7 (R0 = H, no steric effect expected) and 3 (R0 = Ph, Table 4).
Indeed in these reactants, lithiation in position 3 takes place at
c
2378 New J. Chem., 2011, 35, 2375–2381
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2011

View Article Online
products in the presence of a R0 = phenyl in position 6 on the
cyclohexadienyl moiety (Scheme 3). When R0 = H, further
equilibration of the kinetic product, i.e. the product of lithiation
at C9 or C10, leads to the 3/1 ratio of products as shown in
Scheme 5. This equilibration is believed to be incomplete.
We learned more about this regioselectivity by performing
the reaction using ClSiMe3 as an electrophile at lower tem-
perature. Indeed, at ꢀ78 1C, instead of ꢀ40 1C, the NMR
spectrum of the crude mixture indicates the presence of only
one regioisomer 11 at the C9 carbon, in the presence of the
starting material as the major product. This confirms that
lithiation at the benzylic carbon C9 is under kinetic control
while the preference for the lithiation at the benzylic carbon
C7 results from a thermodynamic control.14
Another remarkable point is that lithiation at the site opposite to
the sp3 carbon is disfavoured over all other lithiations, whatever the
compound (and thus the nature of the site, arylic vs benzylic) at
stake. Both benzylic C11 and arylic C3 lithiations are disfavoured by
19.2 and 16.2 kcal molꢀ1, respectively, with respect to the lithiation
in position C7 (Table 3). This justifies the total absence of substitu-
tion at the carbon C3 in the course of the experimental procedure.
Understanding of the high energies for regioisomers
lithiated at C3 and C11 was obtained by decomposing the
lithiation process energies in two effects: (i) impact of the
lithium counterion; (ii) intrinsic stability of the deprotonated
species. In order to apply this decomposition scheme, the
deprotonated species have been computed (Cartesian coordi-
nates are given in ESIz) and their relative stabilities are
compared to those of the lithiated species. As shown in
Table 3, the relative stabilities of the anions in 8 (0.0, 12.2
and 16.7 kcal molꢀ1 for the C7, C9 and C11 deprotonated
species respectively) follow exactly those of the cis lithiated
species in 8 (0.0, 10.7 and 16.2 kcal molꢀ1 for the C7, C9 and
C11 lithiated species respectively) whereas in 7 additional
destabilization (from +19.2 to +36.7 kcal molꢀ1) is obtained
in position 3 only by removing the counterion: lithium
coordination to the arylic position is thus more stabilizing
than to the benzylic position. This is associated to a shorter
Li–C bond in the arylic position (2.04 A) compared to those
obtained in the benzylic position, either in trans (2.08 A) or
even more in cis (2.21 A) conformations. This can be con-
nected to delocalization of the lone pair when deprotonation
takes place in the benzylic position: the electrons of the
carbanion are thus less adapted to ensure lithium binding.
This is in line with the long Li–C distances observed in this
compound, characteristic of clearly ionic and delocalized
Liꢂ ꢂ ꢂC distances. Another origin can be the strong elongation
and thus partial breaking of the Mnꢂ ꢂ ꢂarylic carbon bond
observed upon deprotonation at the corresponding benzylic
position.14
Fig. 4 Deprotonation energies (kcal molꢀ1) for compounds 9, 7 and
8 in positions C3 and C11. For 8 (2 bottom lines) the transition state
ꢀ
(TS) for rotation of the CH2 moiety is used as an intermediate state
to decompose the overall deprotonation in two steps.6d,15
four Me groups does not alter significantly the stability on the
anion in the arylic position 3 as shown by the very small
energy for the isodesmic proton exchange reaction between
deprotonated 9 and 7. In conclusion, thermodynamically, the
anionic structure is not less stable in 7 than it is in 9, for which
deprotonation in position 3 was observed experimentally
together with deprotonation in position 2 in the ratio
20 : 80.6d The proton exchange between the arylic position in
9 and the benzylic position in 8 is then examined (Fig. 4,
bottom). The benzylic anion is found to be favoured by nearly
20 kcal molꢀ1 over the arylic position. The origin of this
preference can in turn be decomposed in two terms, as shown
by the cycle at the bottom of Fig. 4. A structure, in which
there is no conjugation of the carbanion with the ring as the
CH2 moiety is perpendicular to the cyclohexadienyl plane, has
6d,15
been located:
it is searched as a transition state for
rotation along the C3–C11. The minimum is the conjugate
form and the transition state the non-conjugated one, the
difference between both can be used as an evaluation of the
conjugation energy of the anionic double pair with the cyclo-
hexadienyl moiety. A 34.1 kcal molꢀ1 value is found for
conjugation (Table 5 and Fig. 4). The energy for the proton
exchange between this transition state anion and 9 can thus be
considered as the energy difference between abstraction of a
H+ from a non-conjugated sp3 carbon and a sp2 carbon. As
expected, the sp2 anion is the more stable structure by
15.1 kcal molꢀ1. The negative value for the isodesmic reaction
for proton transfer between the anionic form of 9 and 8 is thus
due to the stabilization of the carbanionic form of 8 by
conjugation.
This same decomposition (H+ transfer to a transition state
followed by conjugation, see Fig. 4, bottom) can also be used to
examine the origin of the regioselectivity. H+ transfer in position
C7 is slightly more energetic than in positions C9 and C11, but
the difference is less than 6 kcal molꢀ1. In contrast, strong
differences can be observed in the conjugation term: conjugation
Nevertheless, the major effect determining the relative stability
of the lithiated species is the intrinsic stability of the carbanion
formed. This effect (referred to ii in the paragraph above) can
be examined using isodesmic reactions allowing us to compare
the deprotonation of species 7 or 8 with that of the analogue 9
bearing no methyl group at all on the cyclohexadienyl ring,
which is used as a reference (Fig. 4). The aim of this reaction is
not to reproduce a true chemical process but to compare the
relative stability of the anionic forms. The adjunction of the
c
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2011
New J. Chem., 2011, 35, 2375–2381 2379

View Article Online
Table 5 Energies (kcal molꢀ1) for the proton transfer sequence (see
Fig. 4, bottom for description, compound 9 deprotonated in position 3
is used as a transfer reference in all cases) of 8 in positions 7, 9 and 11
complex 7(182 mg) in 66% yield. IR (neat) 1997, 1901 cmꢀ1 Mn–CO.
1H NMR, 200 MHz, CdCl3, d 1.58 (s, 6H, Me7, Me8), 1.85 (s, 6H,
Me9, Me10), 2.41 (d, J = 13 Hz, 1H, H6exo), 2.51 (d, J = 13 Hz, 1H,
H6endo), 5.43 (s, 1H, H3). Anal. calcd for C13H15MnO3: C, 56.94; H,
5.51%; found: C, 56.78; H, 5.34%.
C7
C9
C11
H+ transfer
Conjugation
Overall transfer
+19.8
+55.5
ꢀ35.7
+16.0
+39.5
ꢀ23.5
+15.1
+34.1
ꢀ19.0
(g5-1-Trimethylsilylmethyl-2,4,5-trimethylcyclohexadienyl)-
Mn(CO)3 and (g5-2-trimethylsilylmethyl-1,4,5-trimethylcyclo-
hexadienyl)Mn(CO)3 complexes 10 and 11. (Z5-1,2,4,5-Tetra-
methylcyclohexadienyl)Mn(CO)3 complex 7 (0.274 g, 1 mmol)
is dissolved in freshly distilled dry THF (10 mL). Distilled
TMEDA (4 equiv.) is added and the solution cooled down at
ꢀ78 1C. n-BuLi (1.6 M in hexane, 4 equiv.) is added slowly and
the reaction mixture stirred for 2 h at ꢀ40 1C. Then an excess
of ClSiMe3 is added (0.434 g, 4 equiv.) and the reaction
mixture is stirred for 2 h at ꢀ40 1C and then at rt for 2 h.
Then water is added. In this flask, Et2O is poured and the
organic phase recovered and washed with a saturated aqueous
NaCl solution and dried over MgSO4. The solvent is removed
under reduced pressure and the crude material purified by a
silica gel chromatography column. Using a 9 : 1 mixture of
PE/Et2O we eluted complexes 10 and 11 in 73% yield in the
ratio 3 : 1, 0.253 g. HRMS (ESI, positive mode): m/z
369.0689 M + Na+, calcd for C16H23MnO3SiNa: 369.0695.
Anal. calcd: C, 55.48; H, 6.69%; found: C, 55.32; H, 6.49%.
IR (neat) n (cmꢀ1): 1996, 1898 [Mn(CO3)]. Complex 10 1H
NMR (200 MHz, CdCl3): d 0.00 (s, 9H, SiMe3), 1.55 (m, 8H,
CH2, Me9 and Me10), 1.85 (s, 3H, Me8), 2.42 (d, J = 13 Hz,
1H, H6exo), 2.52 (d, J = 13 Hz, 1H, H6endo), 5.41 (s, 1H, H3).
Complex 11 1H NMR (200 MHz, CdCl3): d 0.11 (s, 9H,
SiMe3), 1.55 (m, 8H, CH2, Me7 and Me8), 1.80 (s, 3H,
Me10), 2.40 (d, J = 13 Hz, 1H, H6exo), 2.50 (d, J = 13 Hz,
1H, H6endo), 5.27 (s, 1H, H3).
in position C7 is much stronger than in position C9 or C11,
thus leading to the experimentally observed thermodynamic
preference for position C7.
In summary, a neutral (Z5-1,2,4,5-tetramethylcyclohexadienyl)-
Mn(CO)3 complex 7 has been prepared and characterized by
X-ray crystallography. Its reactivity toward a strong base has
been studied. Thus, deprotonation of 7 at a ‘‘benzylic’’
position using a lithiation/electrophilic quench sequence affords
functionalized complexes 10–13. The reaction is not regioselective
and the experimental results can be interpreted by a kinetic
deprotonation at the C9 benzylic carbon at very low tempera-
tures whereas thermodynamic equilibration leads to major
deprotonation at the C7 benzylic carbon at higher tempera-
tures. This is confirmed and decomposed in terms of lithium
binding and electron delocalization by theoretical computa-
tions on the lithiation products of compound 7.
Experimental section
Syntheses
[(g6-1,2,4,5-Tetramethylbenzene)Mn(CO)3]+BF4
compleꢀx
ꢀ
1.10 The ‘‘silver (I) method’’ using BrMn(CO)5, and AgBF4
and the corresponding free arene was used.16 In a Schlenk tube
is introduced successively AgBF4 (0.973 g, 0.5 mmol),
BrMn(CO)5 (1.374 g, 0.5 mmol), dry freshly distilled CH2Cl2,
10 mL under Ar and the mixture is stirred for 2 h under reflux.
Distilled tetramethylmethylbenzene (2.682 g, 0.5 mmol) is
added and the reaction mixture stirred under reflux for 16 h.
The solution is cooled and filtered on Celite and concentrated
under a N2 stream. It is better to avoid evaporation of all the
solvents under reduced pressure because the yield decreases
dramatically. The resulting solution is mixed with diethylether
and the complex precipitates as a yellow powder (1.620 g,
0.45 mmol, 90% yield). IR (neat) 2058, 1994 cmꢀ1 Mn–CO.
1H NMR, 200 MHz, C3D6O, d 2.54 (s, 12H, Me), 6.83 (s, 2H,
ArH). 13C NMR (100 MHz, C3D6O): d 17.5 (CH3), 103.5 (C3
and C6), 115.4 (C1, C2, C4, C5), 217.3 Mn(CO). HRMS (ESI,
positive mode): m/z calcd for C13H14MnO3+, 273.0323; found,
273.0318. Anal. calcd for C13H14MnO3BF4: C, 37.01; H,
3.34%; found: C, 36.89; H, 3.22%.
(g5-1-Diphenylphosphinomethyl-2,4,5-trimethylcyclohexadie-
nyl)Mn(CO)3 and (g5-1,4,5-trimethyl-2-diphenylphosphino-
methylcyclohexadienyl)Mn(CO)3 complexes 12 and 13. (Z5-
1,2,4,5-Tetramethylcyclohexadienyl)Mn(CO)3 complex
7
(0.274 g, 1 mmol) is dissolved in freshly distilled dry THF
(10 mL). Distilled TMEDA (4 equiv.) is added and the
solution cooled down at ꢀ78 1C. n-BuLi (1.6 M in hexane,
4 equiv.) is added slowly and the reaction mixture stirred for
2 h at ꢀ40 1C. Then an excess of ClPPh2 is added (2.20 g,
10 equiv.) and the reaction mixture stirred for 2 h at ꢀ40 1C
and then at rt for 2 h. Then water is added. In this flask, Et2O
is poured and the organic phase recovered and washed with a
saturated aqueous NaCl solution and dried over MgSO4. The
solvent is removed under reduced pressure and the crude
material purified by a silica gel chromatography column.
Using a mixture of PE/Et2O in the ratio 9 : 1 we eluted
complexes 12 and 13 in 82% yield in the ratio 3 : 1, 0.376 g.
HRMS (ESI, positive mode): m/z 481.0737 M + Na+, calcd
for C25H24MnO3PNa: 481.0741. Anal. calcd: C, 55.48; H,
6.69%; found: C, 55.32; H, 6.49%. IR (neat) n (cmꢀ1): 1998,
1905 [Mn(CO3)]. Complex 12 1H NMR (200 MHz, CdCl3):
d 1.49 (s, 3H, Me8), 1.54 (s, 3H, Me10), 1.83 (s, 3H, Me9), 2.45
(d, J = 13 Hz, 1H, H6exo), 2.55 (d, J = 13 Hz, 1H, H6endo),
2.70 (d, J = 12 Hz, 1H, H7), 2.90 (d, J = 12 Hz, 1H, H7),
(g5-1,2,4,5-Tetramethylcyclohexadienyl)Mn(CO)3 complex 7.
[(Z6-1,2,4,5-Tetramethylmethylbenzene)Mn(CO)3]+BF4ꢀ complex 1
(0.360 g, 1 mmol) in dry THF (10 mL) treated with an excess
of LiAlH4 (19 mg, 0.5 mmol) at ꢀ78 1C afforded a yellow solution
which was stirred for 2 h. HCl (1 M, 1 mL) is added and the reaction
mixture was treated with Et2O (20 mL). The organic phase is poured
into a saturated NaCl aqueous solution (20 mL) and the resulting
organic phase dried over MgSO4. The solvent was removed under
reduced pressure and purified on 15–40 mm silica gel with petroleum
ether, PE, which yields the (Z5-tetramethylcyclohexadienyl)Mn(CO)3
1
5.35 (s, 1H, H3), 7.45 (m, 10H, PPh2). Complex 13 H NMR
c
2380 New J. Chem., 2011, 35, 2375–2381
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2011

View Article Online
(200 MHz, CdCl3): d 1.28 (s, 3H, Me7), 1.29 (s, 3H, Me8),
1.74 (s, 3H, Me10), 2.43 (d, J = 13 Hz, 1H, H6exo), 2.53 (d, J =
13 Hz, 1H, H6endo), 2.80 (d, J = 12 Hz, 1H, H9), 2.87 (d, J =
12 Hz, 1H, H9), 5.10 (s, 1H, H3), 7.45 (m, 10H, PPh2).
5 (a) J. C. Boutonnet, F. Rose-Munch, E. Rose, Y. Jeannin and
F. Robert, J. Organomet. Chem., 1985, 297, 185;
(b) J. C. Boutonnet, F. Rose-Munch and E. Rose, Tetrahedron
Lett., 1985, 26, 3989; (c) F. Rose-Munch, E. Rose and A. Semra,
J. Chem. Soc., Chem. Commun., 1986, 1551; (d) F. Rose-Munch,
E. Rose and A. Semra, J. Chem. Soc., Chem. Commun., 1987, 942;
(e) J. C. Boutonnet, F. Rose-Munch, E. Rose and A. Semra, Bull.
Soc. Chim. Fr., 1987, 640; (f) F. Rose-Munch, E. Rose, A. Semra
and C. Bois, J. Organomet. Chem., 1989, 363, 103; (g) F. Rose-
Munch, E. Rose, A. Semra, J. Garcia-Oricain and C. Knobler,
J. Organomet. Chem., 1989, 363, 297; (h) F. Rose-Munch,
K. Aniss, E. Rose and J. Vaisserman, J. Organomet. Chem.,
1991, 415, 223; (i) J. P. Djukic, F. Rose-Munch, E. Rose,
F. Simon and Y. Dromzee, Organometallics, 1995, 14, 2027.
6 (a) F. Balssa, V. Gagliardini, F. Rose-Munch and E. Rose, Organo-
metallics, 1996, 15, 4373; (b) A. Auffrant, D. Prim, F. Rose-Munch,
E. Rose, S. Schouteeten and J. Vaissermann, Organometallics, 2003,
22, 1898; (c) B. Jacques, M. Chavarot, F. Rose-Munch and E. Rose,
Angew. Chem., Int. Ed., 2006, 45, 3481; (d) B. Jacques, A. Chanaewa,
Computational details
Full geometry optimizations were systematically conducted
with no symmetry constraints using the Gaussian 03 pro-
gram17 within the framework of the Density Functional
Theory (DFT) using the hybrid B3LYP exchange–correlation
functional18 and the 6-31+G(d,p) basis set for all atoms
including Mn as implemented in the Gaussian program.
Frequencies were evaluated within the harmonic approxima-
tion and used unscaled to confirm the nature of all minima (no
complex frequency and fulfilment of the criteria on the forces)
having two H atoms on C6 (frequencies for compounds
derived from 3 with a phenyl group at C6 were not computed).
For key structures, importance of adding an implicit solvation
was examined using single point computations on gas phase
optimized geometry. For these computations, the PCM model
using the dielectric constant implemented for THF (eR = 7.58)
and the default implemented in Gaussian 03 are used.19
M. Chavarot-Kerlidou, F. Rose-Munch, E. Rose and H. Gerard,
´
Organometallics, 2008, 27, 626; (e) A. Eloi, F. Rose-Munch and
E. Rose, J. Am. Chem. Soc., 2009, 131, 14178; (f) F. Rose-Munch,
A. Marti, D. Cetiner, J. P. Tranchier and E. Rose, Dalton Trans.,
2011, 40, 1567.
7 (a) J. W. Johnson and P. M. Treichel, J. Chem. Soc., Chem. Commun.,
1976, 688; (b) P. M. Treichel and J. W. Johnson, Inorg. Chem., 1977,
16, 749; (c) P. M. Treichel, K. P. Fivizzani and K. J. Haller, Organo-
metallics, 1982, 1, 931; (d) R. J. Bernhardt and D. P. Eyman, Organo-
metallics, 1984, 3, 1445; (e) R. J. Bernhardt, M. A. Wilmoth,
J. J. Weers, D. M. LaBrush and D. P. Eyman, Organometallics,
1986, 5, 883; (f) D. M. LaBrush, D. P. Eyman, N. C. Baenziger and
L. M. Mallis, Organometallics, 1991, 10, 1026; (g) J. W. Hull,
K. J. Roesselet and W. L. Gladfelter, Organometallics, 1992, 11, 3630.
8 (a) G. Jaouen, Ann. N. Y. Acad. Sci., 1977, 295, 59;
(b) S. G. Davies, S. J. Coote and C. L. Goodfellow, in Advances
in Metal Organic Chemistry, ed. L. S. Liebeskind, JAI Press,
Greenwich, 1991, vol. 2, pp. 1–57; (c) S. G. Davies and
T. D. McCarthy, in Comprehensive Organometallic Chemistry II,
ed. E. W. Abel, F. G. A. Stone and G. Wilkinson, Pergamon Press,
Oxford, UK, 1995, vol. 12, pp. 1039–1070; (d) T. Volk,
D. Bernicke, J. W. Bats and H.-G. Schmalz, Eur. J. Inorg. Chem.,
1998, 1883.
Acknowledgements
We thank CNRS for financial support, He
the determination of the X-ray structures, Centre de Re
tion de Structures, IPCM, UMR 7201 and Noemie Elgrishi for
´
lene Rousseliere for
´
solu-
´
a preliminary study.
9 (a) D. Astruc, E. Roman, J. R. Hamon and P. Batail, J. Am. Chem.
Soc., 1979, 101, 2240; (b) J. R. Hamon, D. Astruc, E. Roman,
P. Batail and J. J. Mayerle, J. Am. Chem. Soc., 1981, 103, 2431;
(c) D. Astruc, J. R. Hamon, E. Roman and P. Michaud, J. Am.
Chem. Soc., 1981, 103, 7502; (d) C. M. Casado, T. Wagner and
D. Astruc, J. Organomet. Chem., 1995, 502, 143; (e) H. A. Trujillo,
C. M. Casado, J. Ruiz and D. Astruc, J. Am. Chem. Soc., 1999,
121, 5674; (f) D. Astruc, Acc. Chem. Res., 2000, 33, 287.
10 J. Dubarle-Offner, F. Rose-Munch, E. Rose, N. Elgrishi and
H. Rouseliere, Organometallics, 2010, 29, 4643.
11 J. D. Jackson, S. J. Villa, D. S. Bacon, R. D. Pike and
G. B. Carpenter, Organometallics, 1994, 13, 3972.
12 For theoretical justification of the negative charge at the Mn(CO)3
centre, see (a) R. Hoffmann and P. Hofmann, J. Am. Chem. Soc.,
1976, 98, 598; (b) O. Eisenstein, W. M. Butler and A. J. Pearson,
Organometallics, 1984, 3, 1150.
13 An identical feature is observed for the isoelectronic chromium
complex: C. A. Merlic, M. M. Miller, B. N. Hietbrink and
K. N. Houk, J. Am. Chem. Soc., 2001, 123, 4909.
14 (a) A. Pfletschinger, T. K. Dargel, J. W. Bats, H.-G. Schmalz and
W. Koch, Chem.–Eur. J., 1999, 5, 537; (b) C. A. Merlic,
J. C. Walsh, D. J. Tantillo and K. N. Houk, J. Am. Chem. Soc.,
1999, 121, 3596.
References
1 (a) M. F. Semmelhack, J. H. T. Hall, R. Farina, M. Yoshifuji,
G. Clark, T. Bargar, K. Hirotsu and J. Clardy, J. Am. Chem. Soc.,
1979, 101, 3535; (b) M. F. Semmelhack, in Comprehensive Organo-
metallic Chemistry II, ed. G. Wilkinson, E. W. Abel and F. G. A.
Stone, Pergamon Press, Oxford, UK, 1995, vol. 2.4, p. 517;
(c) A. Fretzen, A. Ripa, R. Liu, G. Bernardinelli and E. P. Kundig,
¨
Chem.–Eur. J., 1998, 4, 251; (d) F. Rose-Munch and E. Rose, Curr.
Org. Chem., 1999, 3, 445; (e) F. Rose-Munch and E. Rose, in Modern
Arene Chemistry, ed. D. Astruc, Wiley-VCH, 2002, ch. 11, p. 368;
´
(f) M. Rosillo, G. Dominguez and J. Perez-Castells, Chem. Soc. Rev.,
2007, 36, 1589; (g) D. Astruc, Organometallic Chemistry and Cata-
lysis, Springer, Heidelberg, 2007, p. 243.
2 (a) K. F. McDaniel, in Comprehensive Organometallic Chemistry II,
ed. E. W. Abel, F. G. A. Stone and G. Wilkinson, Pergamon Press,
Oxford, 1995, vol. 6, p. 93; (b) A. R. Pape, K. P. Kaliappan and
E. P. Kundig, Chem. Rev., 2000, 100, 2917; (c) F. Rose-Munch and
¨
E. Rose, Eur. J. Inorg. Chem., 2002, 1269; (d) D. Prim, B. Andrioletti,
F. Rose-Munch, E. Rose and F. Couty, Tetrahedron, 2004, 60, 3325;
(e) D. A. Sweigart, J. A. Reingold and S. U. Son, in Comprehensive
Organometallic Chemistry, ed. R. H. Crabtree and D. M. P. Mingos,
Elsevier, Oxford, 2006, 3rd edn, vol. 5, ch. 10, pp. 761–814;
(f) F. Rose-Munch, E. Rose and A. Eloi, in Cationic (Z6-arene)-
and neutral (Z5-cyclohexadienyl)-tricarbonylmanganese complexes:
Synthesis and reactivity. Patai’s Chemistry of Functional Groups,
ed. Marek and Rappoport, The Chemistry of Organomanganese
Compounds, 2011, pp. 489–558; (g) F. Rose-Munch and E. Rose,
Org. Biomol. Chem., 2011, 9, 4725.
´
15 I. Chataigner, C. Panel, H. Gerard and S. R. Piettre, Chem.
Commun., 2007, 3288.
16 A. J. Pearson and I. C. Richards, J. Organomet. Chem., 1981, 204, C25.
17 M. J. Frisch et al., Gaussian 03, Revision C.02, Gaussian Inc.,
Wallingford CT, 2004.
18 (a) C. Lee, W. Yang and R. G. Parr, Phys. Rev. B, 1988, 37, 785;
(b) B. Miehlich, A. Savin, H. Stoll and H. Preuss, Chem. Phys. Lett.,
1989, 157, 200; (c) A. D. J. Becke, Chem. Phys., 1993, 98, 5648.
19 M. Cossi, G. Scalmani, N. Rega and V. Barone, J. Chem. Phys.,
2002, 117, 43 and references therein.
3 (a) D. Astruc, Tetrahedron, 1983, 39, 4027; (b) J. C. Boutonnet and
E. Rose, J. Organomet. Chem., 1981, 221, 157.
4 J. P. Djukic, F. Rose-Munch, E. Rose and Y. Dromzee, J. Am.
Chem. Soc., 1993, 115, 6434.
c
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2011
New J. Chem., 2011, 35, 2375–2381 2381