Novel tridentate diamino organomanganese(II) complexes as homogeneous catalysts in manganese(II)/copper(I) catalyzed carbon-carbon bond forming reactions

Article abstract of DOI:10.1016/S0022-328X(97)00675-X

The new, paramagnetic arylmanganese(II) complex Li[MnCl₂(NCN)] (2, NCN [C₆H₃(CH₂NMe₂)₂-2,6] ^-) has been obtained in high yield from the reaction of MnCl₂ and [Li(NCN)]₂ in a 2:1 molar ratio. In THF solution, 2 is likely an ionic species [Li(THF)_n] [MnCl₂(NCN)] (molecular weight determination and conductivity measurements), while magnetic measurements indicate that a high spin d⁵ manganese(II) center is present. Subsequent reaction of 2 with RLi afforded [MnR(NCN)] (R=Me (3a), n-Bu (3b)). Complex 2, using CuCl as a co-catalyst, is an effective catalyst system for cross-coupling of Grignard reagents with alkyl bromides and the 1,4-addition of organomagnesium halides to α,β-unsaturated ketones. No further additives or co-solvents are necessary. For both reactions a dramatic decrease in reaction times is observed when compared to standard manganese/copper systems. Alkyl bromides with unsaturated or heteroatom functionalities can be cross-coupled. Also, excellent reactivity towards normally unreactive β,β-disubstituted ketones has been observed in the 1,4-addition reaction.

Full text of DOI:10.1016/S0022-328X(97)00675-X

Journal of Organometallic Chemistry 558 (1998) 61–69
Novel tridentate diamino organomanganese(II) complexes as
homogeneous catalysts in manganese(II)/copper(I) catalyzed
carbon–carbon bond forming reactions
Johannes G. Donkervoort a, Jose´ L. Vicario a, Johann T.B.H. Jastrzebski a,
Robert A. Gossage a, Ge´rard Cahiez b, Gerard van Koten a,*
a Department of Metal-Mediated Synthesis, Debye Institute, Utrecht Uni6ersity, Padualaan 8, 3584 CH Utrecht, The Netherlands
b Ecole Supe´rieure de Chimie Organique et Minerale, De´partement de Chimie, 13 Boule6ard de l’Hautil, F-95092, Cergy-Pontoise, France
Received 30 August 1997
Abstract
The new, paramagnetic arylmanganese(II) complex Li[MnCl2(NCN)] (2, NCN [C6H3(CH2NMe2)2-2,6]−) has been obtained in
high yield from the reaction of MnCl2 and [Li(NCN)]2 in a 2:1 molar ratio. In THF solution, 2 is likely an ionic species [Li(THF)n]
[MnCl2(NCN)] (molecular weight determination and conductivity measurements), while magnetic measurements indicate that a high
spin d5 manganese(II) center is present. Subsequent reaction of 2 with RLi afforded [MnR(NCN)] (R=Me (3a), n-Bu (3b)).
Complex 2, using CuCl as a co-catalyst, is an effective catalyst system for cross-coupling of Grignard reagents with alkyl bromides
and the 1,4-addition of organomagnesium halides to h,i-unsaturated ketones. No further additives or co-solvents are necessary.
For both reactions a dramatic decrease in reaction times is observed when compared to standard manganese/copper systems. Alkyl
bromides with unsaturated or heteroatom functionalities can be cross-coupled. Also, excellent reactivity towards normally
unreactive i,i-disubstituted ketones has been observed in the 1,4-addition reaction. © 1998 Elsevier Science S.A. All rights reserved.
Keywords: Manganese; Aryl-manganese; Cross-coupling catalyst; Grignard reagents
1. Introduction
excellent functional group tolerance, allows the selective
conversion of, for example, ketoaldehydes to ketoalco-
hols, keto-carboxylic acid chlorides to diketones, un-
symmetrical diketones to ketoalcohols as well as the
substitution of ꢀ halogenoketones or ꢀ ketoalkenyl
halides ([1]d).
The binding of a non-transferable, chiral ligand and
a transferable organic anion to the manganese center is
required to make organomanganese(II) reagents useful
for the enantioselective synthesis of organic products.
This strategy, which is well-known in cuprate chemistry
[2], has so far not been applied in manganese(II) chem-
istry as suitable non-transferable groups were not
available1.
Recent studies have shown that organomanganese(II)
complexes are excellent reagents in a range of C–C
bond forming reactions [1]. These reagents are more
stable than many organotransition metal complexes
and are easily prepared from the reaction of (cheap)
manganese(II) halides with Grignards or organolithium
compounds ([1]a,d). Reactions that can be performed
using organomanganese(II) reagents (with CuX in cata-
lytic amounts) include acylation, 1,2-addition, 1,4-addi-
tion, enolization, alkylation, alkenylation and arylation
([1]d,e). These reactions proceed with good to excellent
chemoselectivity under mild conditions (ether solvents,
0°C to ambient temperature). The high chemoselectivity
of organomanganese(II) reagents, combined with an
1 In contrast, this strategy has led to the use of chiral mangane-
se(III) Schiff-base complexes in, for example, enantioselective epoxi-
dation reactions [3].
* Corresponding author.
0022-328X/98/$19.00 © 1998 Elsevier Science S.A. All rights reserved.
PII S0022-328X(97)00675-X

62
J.G. Donker6oort et al. / Journal of Organometallic Chemistry 558 (1998) 61–69
and easier to study as, except for CuX, no further
We have set out to explore the use of our monoan-
ionic, bis(amino)aryl [C6H3(CH2NMe2)2-2,6]− (=
NCN; Fig.
A, R1=Me, R2=R3=H) and
additives or co-solvents are needed, and (iii) require
much shorter reaction times when compared to litera-
ture procedures2.
1
bis(phosphino)aryl [C6H3(CH2PPh2)2-2,6]− (=PCP;
Fig. 1 B, R=Ph) ligands as auxiliaries (i.e. non-trans-
ferable groups) for catalytically-active metal centers.
For example, the nickel complex [Ni(NCN)X] (X=Cl,
Br), has already found use in the selective 1:1 Kharasch
addition reaction of haloalkanes to alkenes ([4]a–e) as
well as in the controlled living radical polymerization of
olefins ([4]f). Some NCN– and PCP–RuII complexes,
such as [RuCl(NCN)(PPh3)], [RuCl(PCP)(PPh3)] and
[RuCl(OTf)(PCP)], are excellent catalysts for selective
transfer hydrogenation and hydrogenation catalysis [5].
The bis(phosphino)aryl–rhodium and –iridium com-
plexes [M{C6H3(CH2P(t-Bu)2)2-2,6}(H4)] (M=Rh, Ir)
are unique transfer dehydrogenation catalysts for the
conversion of alkanes to alkenes using, for example,
3,3-dimethylbutene as hydrogen acceptor [6].
2. Results and discussion
2.1. Synthesis and characterization of two
NCN–manganese(II) complexes
The transmetallation reaction of [Li(NCN)]2 (1) with
two equivalents of MnCl2 affords a novel arylman-
ganese compound (2), which was isolated as a yellow
green solid (Eq. 1). Elemental analysis of 2 indicated a
molecular stoichiometry corresponding to Li[Mn-
Cl2(NCN)], i.e. to a manganate complex. The reaction
of 2 with D2O afforded quantitatively 2-deuterio-1,3
bis[(dimethylamino)methyl]benzene. This is further evi-
dence for the presence of a metal–carbon bond. Unfor-
tunately, the paramagnetic character of 2 prevented
structural characterization by NMR spectroscopy.
However, by using the Evans method [8], a magnetic
moment of 5.95 BM was measured for a THF solution
of 2, indicating the presence of a high spin d5 mangane-
se(II) metal center.
In the present study, we report on the synthesis of
NCN–manganese(II) complexes and the use of the
NCN ligand as a non-transferable group in manganese/
copper-mediated C–C bond forming reactions. The
effect of the use of the NCN ligand is that the reactions
become (i) catalytic in manganese, (ii) highly selective
Complex 2 is only slightly soluble in non-polar sol-
vents like benzene but readily soluble in THF. The
value of the measured degree of association was 0.5
(cryoscopy in THF [9]). This strongly suggests that 2
exists in THF either as a 1:1 mixture of two neutral
species, [MnCl(NCN)] and LiCl, or as a manganate
consisting of a [MnCl2(NCN)] anion and a [Li(THF)n]
cation. Conductivity measurements of THF solutions
of 2 at room temperature revealed a molar conductivity
(Lm) value of 0.0206 cm2 V−1 mol−1. This is signifi-
cantly larger than that of LiCl and [Li(NCN)]2 (see
Section 3). This result strongly suggests that 2 is likely
an ionic species such as [Li(THF)n][MnCl2(NCN)].
Subsequent reaction of 2 with MeLi or n-BuLi af-
fords the corresponding heteroleptic diorganoman-
ganese compounds 3a and 3b, respectively (see Eq. 1).
After work-up (see Section 3) these compounds were
isolated as extremely air sensitive, paramagnetic, green
colored oils. The compounds are readily soluble in both
polar and non-polar solvents. The absence of LiCl in
samples of 3a–b became evident from the qualitative
Fig. 1. Structures of bis(amino)aryl and bis(phosphino)aryl ligand
systems.
2 A preliminary report concerning the cross-coupling reaction has
appeared [7].

J.G. Donker6oort et al. / Journal of Organometallic Chemistry 558 (1998) 61–69
63
Table 1
analysis for both Li and halogen. Like 2, the reaction of
3a–b with D2O affords 2-deuterio-1,3-bis[(dimethy-
lamino)methyl]benzene in quantitative yield. Due to the
extreme sensitivity of 3 to air, no molecular weight
determinations (cryoscopy) were performed.
Product yields of R–R% obtained from cross-coupling reactions of
Grignard reagents with alkyl bromidesa
Entry Alkyl bromide
Grignard reagent
Yield (%)
1
n-C8H17Br
n-C8H17Br
n-C8H17Br
n-C8H17Br
n-C8H17Br
n-C8H17Br
n-C8H17Br
n-C8H17Br
n-BuMgCl
n-BuMgCl
n-BuMgCl
n-BuMgCl
n-BuMgCl
n-BuMgCl
i-PrMgCl
t-BuMgCl
s-BuMgCl
n-EtMgCl
n-BuMgCl
83
83
83
62
55
70
75
84
92
94
91
89
91
89
Although the structural features of 2 and 3 are not
known, it is likely that 2 has a five-coordinate Mn-cen-
ter as a result of additional intramolecular Mn–N
coordination. The formation of organometallic aggre-
gates with lithium halides like LiCl is not unprece-
dented. For example, the lutetium complex
[LuCl2(NCN)LiCl]2 has been isolated and the structure
established by an X-ray crystal structure determination
[10]. The proposed monomeric structure of 3 would
resemble that of analogous [SnIIR(NCN)] complexes
[11]. However, aggregate structures are also feasible, as
shown by examples of organomanganese complexes in
which organic groups are bridging two Mn centers [12].
In species of general formula [MnR(NCN)], the R
group could be involved in the latter kind of binding, in
addition to the NCN ligand [13], e.g. the aggregated
organocopper/copper halide complex [Cu4Br2(NCN)2]
[14].
2b
3c
4d
5e
6f
7
8
9
n-C8H17Br
n-C4H9CH(Me)Br
n-C4H9CO(CH2)10Br
10
11
12
13
14
n-C2H5OC(O)(CH2)2Br n-BuMgCl
CH2ꢀCH(CH2)2Br
Me3SiCꢁC(CH2)2Br
n-C14H29MgCl
n-C14H29MgCl
a Unless otherwise stated; catalyst 10 mol% of in situ prepared 2, 5
mol% CuCl. b 10 mol% pure 2, 5 mol% CuCl. c 10 mol% pure 3b, 5
mol% CuCl. d 10 mol% 2, 2.5 mol% CuCl. e 12 mol% 2, 2 mol%
CuCl. f 10 mol% 2, 10 mol% CuCl.
forded essentially the same products in identical yields,
further reactions were carried out simply using in situ
prepared 2 (i.e. addition of one equivalent of MnCl2 to
1).
The catalytic activity of the new organomanganese
species 2 and 3 (either as pure compounds or prepared
in situ) were tested in cross-coupling reactions of Grig-
nard reagents with organic bromides as well as in
1,4-addition reactions of Grignard reagents to h,i-un-
saturated ketones.
Primary, secondary and tertiary alkyl Grignard
reagents reacted successfully (see entries 7–9 in Table
1). Specifically, the reaction with secondary alkyl Grig-
nard reagents (entry 7) is notable. Our newly developed
procedure affords the cross-coupled product in excel-
lent yield (92%), whereas a similar reaction using the
classical Li2CuCl4 reagent afforded the cross-coupled
product in only 24% yield [15]. Also, the reaction of a
Grignard reagent with a secondary alkyl bromide gives
rise to the formation of the cross-coupled product in
high yield (entry 10). This is in contrast with the
manganese-mediated [16] or copper-catalyzed proce-
dures [17], in which these substrates do not react.
It has been well established ([l]d) that organoman-
ganese reagents are highly chemoselective and tolerant
to the presence of a variety of functional groups. These
advantages are retained in our catalyst system (see
entries 11 to 14 in Table 1). The presence of unsatu-
rated functionalities (alkene or alkyne) does not influ-
ence either the selectivity or the yield of the C–C
coupled product. The corresponding unsaturated cross-
coupled products were obtained in excellent yields.
None or only small traces (B5%) of olefin rearrange-
ment products could be detected in the crude reaction
mixtures. This result is of particular interest for the
preparation of ꢀ-unsaturated acids, compounds which
can be used for the preparation of ultra thin-layer
photoresists and ꢀ-radio-iodonated fatty acids, which
2.2. Homogeneous catalysis with
[Li(THF)n][MnCl2(NCN)] and [MnR(NCN)]
Addition of a Grignard reagent to a solution of an
organic bromide in THF at 5°C in the presence of 10
mol% of 2 and 5 mol% of CuCl affords after 15 min
reaction time and work-up (see Section 3), the cross-
coupled products with excellent chemoselectivity and
high yield, see Table 1.
Under the reaction conditions (all Grignard reagent
is added at once) it is most likely that, prior to the
cross-coupling reaction, the organomanganese halide 2
is converted to a heteroleptic diorganomanganese com-
pound 3 (see Eq. 1). These compounds are probably
involved in the subsequent cross-coupling reaction. In
order to test this hypothesis, we have carried out the
reaction of n-C8H17Br with n-BuMgCl under identical
reaction conditions but now using either pure or in situ
prepared 2, and pure 3b, (see entries 2, 1 and 3,
respectively, in Table 1). Because these reactions af-

64
J.G. Donker6oort et al. / Journal of Organometallic Chemistry 558 (1998) 61–69
2.3. Mechanistic considerations
find use as myocardial imaging agents [18]. Synthetic
procedures for ꢀ-unsaturated acids described in litera-
ture involve many steps and the products are not
always of the required purity for photoresists synthesis
because of, for example, the presence of small amounts
of compounds resulting from double bond migration
[18,19].
Substrates containing heteroatom functionalities, e.g.
ketones and esters, are tolerated by our catalyst system
(entries 11 and 12, respectively). Of special interest is
the excellent yield of heptanoic ethyl ester which was
obtained by reacting ethyl 3-bromopropionate with n-
BuMgCl (entry 12). Previous attempts to alkylate this
alkyl bromo ester resulted in low yields (28%) [20].
The applicability of NCN–manganate 2/CuX as a
catalytic system for the conjugate addition reactions of
Grignard reagents towards enones (Eq. 3) was studied.
Excellent yields and chemoselectivities for the 1,4-addi-
tion products were obtained, which appeared to be
independent of the type of electrophile used (endo-
cyclic, exo-cyclic or acyclic h,i-unsaturated ketones;
Table 2).
The present results show that with the monoanionic
NCN ligand the above mentioned C–C bond forming
reactions become catalytic in manganese. Moreover, the
reactions are much faster than the reactions mediated
by stoichiometric amounts of MnCl2 and much easier
to study because only a single solvent, THF or toluene,
can be used and additives or activators are not
required.
It was shown that a D2O quench of completed reac-
tion mixtures provided the 2-deuterio derivative of
NCN quantitatively. This suggests that the NCN ligand
remains connected via Cipso to a metal center. There-
fore, the C6H3(CH2NMe2)2-2,6 anion is a potentially
terdentate ligand with excellent non-transferable prop-
erties in the reactions studied here. This is not only
interesting from the point of view that manganese can
be used in catalytic amounts, but also that the Mn
center is contained in the ligand framework and thus
becomes subject to the steric and electronic influences
exerted by the NCN ligand. Preliminary experiments
with an enantiomerically enriched, modified NCN-pin-
cer ligand [26] (see Section 3), suggest that chiral infor-
mation can be transferred from the ligand in the
product forming step. Unfortunately, the enantiomeric
excesses of the products that are obtained are quite low.
Finally, the experiment with Li2MnCl4 and the free
arene 1,3-bis[(dimethylamino)methyl]benzene failed to
give the C–C coupled products. This corroborates the
view that the NCN ligand is bound in its deprotonated
form, i.e. a metal–carbon bond is present.
The next question to be addressed is the meaning of
the 2:1 atomic ratio of manganese and copper that is
required. In a series of experiments, it was found that a
higher (4:1 or 6:1) or lower (1:1) Mn:Cu ratio, lowers
the yield of the C–C coupled product (entries 4–6 in
Table 1). In addition, in separate experiments using
[Cu4Br2(NCN)2] as the copper(I) source, a much lower
yield of cross-coupled product was obtained as com-
pared to the NCN–Mn/CuCl system (see Section 3).
This indicates that it is not simply a NCN–Cu species
that is the active catalyst.
In particular, the high yields of products obtained
from i,i-disubstituted h,i-unsaturated ketones are no-
table (entries 1–3 in Table 2), since it is well known
that these substrates undergo conjugate addition reac-
tions with difficulty using organo-copper or cuprate
reagents [21].
Table 2
Product yields of the 1,4-addition products obtained from the reac-
tion of BuMgCl with conjugated enonesa
Entry
Substrate
Isolated yieldb
It must also be noted, that substrates that have been
tested with our catalyst system (entries 1–3 in Table 2)
are not reactive enough to undergo a conjugate addi-
tion with a cuprate reagent. Therefore, a magnesium
cuprate species can also be excluded as the kinetically
active species. Finally, the stoichiometric reaction of 3b
with octyl bromide in the absence of copper(I) did not
proceed. All this information points to a reactive inter-
mediate that at least comprises of MnII, NCN and
Cu(I) as well as the substrate and the transferable
group R−.
Catalyst 10 mol% of in situ prepared 2, 5 mol% CuCl. b Isolated
yield of the corresponding 1,4-addition products.
a
With the present information, further conclusions
about the nature of the reactive intermediate are impos-

J.G. Donker6oort et al. / Journal of Organometallic Chemistry 558 (1998) 61–69
65
Fig. 3. Molecular structure of the isolated end-product (8) of the
1,2-addition reaction of a methylmanganese(II) reagent with ben-
zaldehyde.
The cross-coupling and conjugate addition reactions
differ in at least one important aspect, i.e. the cross-
coupling reaction produces the end-product without
hydrolysis, whereas in the conjugate addition reaction a
magnesium enolate is the end-product. The latter sug-
gests that, at least in the conjugate addition reactions,
magnesium cations play a role. A model of this end
product has recently been proposed by Floriani et al.
[23]. They isolated and characterized (X-ray analysis)
an alkoxymanganate complex 8 (Fig. 3), which is the
end-product of the 1,2-addition reaction of an
organomanganese(II) reagent to benzaldehyde [23]. Al-
though this manganese reagent is not catalytically ac-
tive in the 1,2-addition reaction, the proposed
formation of 8 [23] provides a possible intermediate
structure for our catalytic system in the 1,4-addition
reaction. Analogous to 8, we propose a kinetically
active species 9 (Fig. 4). In this intermediate, the sub-
strate is anchored in a bidentate fashion to the CuMn2
unit through the oxygen atom of the carbonyl unit to
one of the manganese centers and via p2-binding of the
CꢀC double bond to copper. This type of intermediate
has also been postulated for the conjugate addition
reactions catalyzed by arenethiolatocopper(I) com-
plexes ([2]f). The R-group from the other manganese
center of the aggregate is then ideally situated to be
Scheme 1. Possible equilibria in the formation of the kinetically active
species in 2/CuX and 3/CuX catalyzed C–C bond forming reactions.
sible. However, the fact that the manganate species 2 or
the diorganomanganese species 3 can be used catalyti-
cally, can be explained when the equilibria in Scheme 1
are considered. Under the reaction conditions (i.e. first
addition of CuCl to 2 or 3) it is likely that initially a
manganese–copper species is formed (e.g. 5 or 6). In
the presence of excess RMgX, these intermediates could
then form an organomanganate (e.g. 7; note that NCN
is non-transferable probably because of chelation). A
possible structure of 7 is shown in Fig. 2A, which is
based on the versatile coordination properties of the
NCN anion3. A precedent for the bridging of Cipso of
different Lewis acidic metal centers has been obtained
for
the
tantalum(V)
complex
[TaCl2(CH(t-
Bu)(NCN)ZnCl] (established by an X-ray crystal struc-
ture determination) [22], of which the schematic
structure is shown in Fig. 2B.
Fig. 2. A: Proposed structure of [CuMn2R3(NCN)2], 7; B: Molecular
structure of the NCN–tantalum(V) complex [TaCl2ꢀ(CH(t-
Bu))(NCN)ZnCl].
3 Based on experimental evidence (during the reaction an excess of
RMgX is present), the formation of negatively charged manganate
species containing a higher number of R-groups may, a priori, not be
excluded.
Fig. 4. Proposed intermediate (9) in the 1,4-addition reaction of
[CuMnR3(NCN)2] with h,i-unsaturated ketones. The CH2NMe2 sub-
stituents have been omitted for clarity.

66
J.G. Donker6oort et al. / Journal of Organometallic Chemistry 558 (1998) 61–69
Scheme 2. Proposed catalytic cycle of the 1,4-addition reactions of [CuMn2R3(NCN)2] with h,i-unsaturated ketones (the CH2NMe2 arms of the
NCN ligand have been omitted for clarity).
measurements were carried out using a Philips PW 9501
conductometer with a YSI model 3402 microcell. The
transferred to the 4-position of the substrate to afford a
manganese enolate complex 10 (Scheme 2). Complex 10
probably reacts with one equivalent of a Grignard
reagent to afford a magnesium enolate and 7, which is
capable of entering a new catalytic cycle.
The present study has provided an excellent new
catalytic system for C–C coupling reactions. In addi-
tion, some information regarding the kinetically active
species, involved in manganese–copper catalyzed reac-
tions, has been obtained. However, studies to give a
further insight into the kinetically active species and
mechanism of these reactions should be part of future
work.
compounds
[Cu4Br2(NCN)2]
[14],
4-bromo-1-
trimethylsilyl-1-butyne [24], [Li(NCN)]2 ([13]a) and (R,
R%)–[LiC6H3(CH(Et)NMe2)2-2,6}]2 [25] were prepared
according to literature procedures. Commercially ob-
tained manganese(II) chloride tetrahydrate was dehy-
drated in vacuo at 120°C and stored under a nitrogen
atmosphere.
3.2. Preparation of [LiMnCl2{C6H3(CH2NMe2)2-2,6}],
2
To a suspension of MnCl2 (11.60 mmol) in THF (25
ml), a solution of [Li(NCN)]2 (5.8 mmol) in THF (5 ml)
was added after which the mixture was stirred for 15
min at room temperature. The solvent was then evapo-
rated in vacuo to give a dark oil. The oil was washed
with pentanes (2×25 ml) and dried in vacuo to give
2.76 g (73%) of 2 as a yellow/green colored solid. Anal.
Calc. (found) for C12H19Cl2LiMnN2: C, 44.47 (44.40);
H, 5.91 (6.05); N, 8.64 (8.66).
3. Experimental section
3.1. General section
All organometallic syntheses were carried out using
standard Schlenk techniques under an atmosphere of
dry, oxygen-free nitrogen. Solvents were carefully dried,
distilled from sodium benzophenone ketyl and deoxy-
genated prior to use. 1H (200 or 300 MHz) and 13C{1H}
(50 or 75 MHz) NMR spectra were recorded on a
Bruker AC-200 or AC-300 spectrometer. Conductivity
3.2.1. Cryoscopy measurements
The cryoscopic measurements were conducted ac-
cording to a literature procedure [9] using a S2541

J.G. Donker6oort et al. / Journal of Organometallic Chemistry 558 (1998) 61–69
67
The combined organic fractions were washed with a
saturated aqueous NaHCO3 solution (1×25 ml) and
dried (MgSO4). The products were isolated as colour-
less liquids by flash distillation and were identified by
comparison of their GC, MS and NMR data with those
of authentic samples. Dodecane: yield 1.41 g (83%).
2-Methyldecane: yield 1.17 g (75%). 2,2-Dimethylde-
cane: yield 1.43 g (84%). 3-Methylundecane: yield 1.20
g (94%). 3-Methyloctane: yield 1.20 g (94%). 5-
Methyldec-1-ene: yield 1.36 g (87%). Heptanoic ethyl
ester: yield 1.41 g (89%). Octadec-1-ene: yield 2.29 g
(91%). 1-(Trimethylsilyl)-octadec-1-yne: yield 2.87 g
(89%).
thermolyzer and a metal-mantled Pt-100 sensor. For
the calibration, naphthalene was used to give the fol-
lowing cryoscopic constant Ek=1.52 K kg mol−1. The
addition of 2 as weighed tablets took place under
nitrogen atmosphere and after dissolution, the freezing
point was determined (three repetitions). The concen-
tration of 2 was varied between 0 and 50 mmolal.
3.2.2. Determination magnetic moment of 2 by the
E6an’s method
In a typical experiment, a NMR tube was filled with
ca. 25 mg of 2, 1 ml of THF and 1 mg of durene. The
shift of the durene signals was determined by compar-
ing the durene signals with those of a solution of 1 mg
of durene in THF (1 ml). The formulae as derived by
Evans [8] allowed for the calculation of the magnetic
moment of 2.
3.5. [MnBu(NCN)]/CuCl-catalyzed cross-coupling
reaction with octyl bromide
The procedure was identical to that described above
for the 2/CuCl-catalyzed cross-coupling reactions ex-
cept that a solution of [Mn(n-Bu)(NCN)] (1.0 mmol)
and CuCl (0.5 mmol) in THF (25 ml) was used. The
product(s) were identified by comparison of the GC,
MS and NMR data with those of an authentic sample.
3.2.3. Conducti6ity measurements of 2
In a typical experiment, the conductivity (Vm) of a
solution of 3 (ca. 10−3 M) in THF at room tempera-
ture was found to be 2.06×10−2 cm2 V−l mol−l. For
comparison LiCl and [Li(NCN)]2 were measured in the
same system, which gave molar conductivity values of
9×10−5 and 1.3×10−5 cm2 V−l mol−l, respectively.
3.6. Use of [Cu4Br2(NCN)2] as catalyst
3.3. Preparation of [Mn{C6H3(CH2NMe2)2-2,6}R], 3
n-Octyl bromide (2.86 mmol) and i-PrMgCl (2.86
mmol, 1.43 ml of a 2.0 M solution in THF) were added
to a solution of [Cu4Br2(NCN)2] (0.14 mmol) in THF
(25 ml) at a temperature of 5°C. After stirring the
reaction mixture for 1 h at this temperature, the reac-
tion mixture was hydrolyzed with 50 ml of 4 M HCl
(aq.). The work up procedure was identical as that
described for the corresponding 2/CuCl-catalyzed cross-
coupling reaction. Yield: 0.24 g (54%). The product was
identified by comparison of the GC, MS and NMR
data with those of an authentic sample.
In a typical experiment a solution of either RLi (5.0
mmol; R=Me, n-Bu) in hexanes or RMgCl (5.0 mmol;
R=n-Bu) in THF, was added to a solution of 2 (5.0
mmol) in C6H6 (25 ml) at ambient temperature. The
mixture was stirred for 5 min. The resulting salt that
was formed was removed by centrifugation of the reac-
tion mixture and decantation of the organic layer. The
remaining solid was extracted with C6H6 (3×25 ml).
The combined organic fractions were then evaporated
in vacuo to give 3 as off-green colored oils. Yield: 1.20
g (92%) 3a, 1.44 g (95%) 3b. Due to the extreme
sensitivity to air, no elemental analysis have been per-
formed of these products.
3.7. General procedure for
[Li(THF)n][MnCl2(NCN)]/CuCl-catalyzed conjugate
1,4-addition reactions
3.4. General procedure for cross-coupling reactions
catalyzed by the 2/CuCl catalytic system
To a solution of 2 (1.0 mmol), copper(I) chloride (0.5
mmol) and the appropiate electrophile (10 mmol) in 25
ml of THF, was added drop-wise the Grignard reagent
(10 mmol) dissolved in THF at such a rate that the
temperature of the reaction mixture did not exceed 7°C.
Complete addition of the Grignard reagent was fol-
lowed by stirring the reaction mixture for 45 min at
5°C. The mixture was then quenched with 100 ml of 4
M HCl (aq.). The organic layer was separated and the
aqueous layer extracted with Et2O (3×30 ml). The
combined organic fractions were washed with a satu-
rated aqueous NaHCO3 solution (30 ml) and dried
In a typical experiment, copper(I) chloride (0.5
mmol) was added to a clear solution of 2 prepared in
situ (1.0 mmol) in THF (25 ml). The reaction mixture
was cooled to 5°C. Addition of the appropriate alkyl
bromide (10 mmol) was followed by the rapid addition
(30 s) of the Grignard reagent (10 mmol) dissolved in
THF. The resulting mixture was then stirred for an-
other 15 min at 5°C and then hydrolyzed with 50 ml of
4M HCl (aq.). The organic layer was separated and the
aqueous layer was extracted with pentanes (4×25 ml).

68
J.G. Donker6oort et al. / Journal of Organometallic Chemistry 558 (1998) 61–69
Acknowledgements
(MgSO4). The crude product(s) were purified by flash
distillation and identified by comparing their GC, MS
and NMR data with those of authentic samples. 3-
Butyl-3-methyl-5,5-dimethyl-cyclohexanone: yield 1.76
g (90%). 3-(1,1-Dimethyl-pentyl)-5 methyl-cyclohex-
anone: yield 2.28 g (92%). 4,4-Dimethyloctan-2-one:
yield 1.36 g (87%). 4-Phenyloctan-2-one: yield 1.75 g
(86%). 3-Butylcyclohexanone: yield 1.40 g (91%).
This work was supported in part (JGD) by the
Netherlands Foundation for Chemical Research
(SON) with financial aid from the Netherlands Orga-
nization for Scientific Research (NWO).
References
3.8. General procedure for enantioselecti6e 1,4-addition
reactions
[1] (a) J.F. Normant, G. Cahiez, in: R. Scheffold (Ed.), Modern
Synthetic Methods, vol. 3, Wiley, Chichester, UK, 1983, p. 173.
(b) M.T. Reetz, K. Ro¨fing, N. Griebenow, Tetrahedron Lett. 35
(1994) 1969. (c) C. Kiefl, A. Mannschreck, Synthesis (1995) 1033.
(d) Review: G. Cahiez, in: L. Paquette (Ed.), Encyclopedia of
Reagents for Organic Synthesis, Wiley, Chichester, 1995, p. 925.
(e) I. Klement, H. Stadtmu¨ller, P. Knochel, G. Cahiez, Tetrahe-
dron Lett. 38 (1997) 1927.
(A) In THF: the procedure was identical to that
described for the 2/CuCl-catalyzed 1,4-addition reac-
tions,
starting
from
in
situ
prepared
[MnCl2{C6H3(CH(Et)NMe2)2-2,6}] (9) by reacting (R,
R)–[Li{C6H3(CH(Et)NMe2)2-2,6}]2 (9) (0.5 mmol) and
MnCl2 (1.0 mmol) in THF (25 ml) at ambient tem-
perature. (B) In toluene: the arylmanganese(II) com-
plex 9 was prepared in an identical way to that
described for 2, starting from 10 (0.5 mmol) and
MnCl2 (1.0 mmol). After 30 min, all volatiles were
removed from the reaction mixture in vacuo. The
residue was washed with pentanes (2×25 ml) and
then dissolved in toluene (25 ml). The work-up proce-
dure was identical to that described above for the
2/CuCl-catalyzed 1,4-addition reaction. 4-Phenyloctan-
2-one: yield 1.75 g (86%). The % e.e. value (4% in
THF, 8% in C7H8) was determined from the 13C{1H}
NMR data of the diastereoisomeric ketal, which is
formed from the crude product according to a litera-
ture procedure [26,27]. 3-Butylcyclohexanone: yield
1.40 g (91%). The % e.e. value (11% in C7H8) was
determined in a similar way.
[2] (a) B.H. Lipshutz, G. Bulow, R.F. Lowe, K.L. Stevens, J. Am.
Chem. Soc. 118 (1996) 5512. (b) A.H.M. de Vries, A. Meetsma,
B.L. Feringa, Angew. Chem. Int. Ed. Engl. 35 (1996) 2374. (c)
A. Pfaltz, Acta Chem. Scand. 50 (1996) 189. (d) U. Koop, G.
Handke, N. Krause, Liebigs Ann. 10 (1996) 1487. (e) J.E.
Ba¨ckvall, Acta Chem. Scand. 50 (1996) 661. (f) M. van Klaveren,
E.S.M. Persson, A. DelVillar, D.M. Grove, J.E. Ba¨ckvall, G.
van Koten, Tetrahedron Lett. 36 (1995) 3059.
[3] (a) W. Zhang, J.L. Loebach, S.R. Wilson, E.N. Jacobsen, J. Am.
Chem. Soc. 112 (1990) 2801. (b) E.N. Jacobsen, W. Zhang, M.L.
Gu¨ler, Ibid 113 (1991) 6703.
[4] (a) G. van Koten, Pure Appl. Chem. 61 (1989) 1681. (b) M.H.P.
Rietveld, D.M. Grove, G. van Koten, New J. Chem. 21 (1997)
751. (c) D.M. Grove, A.H.M. Verschuuren, G. van Koten,
J.A.M. van Beek, J. Organomet. Chem. 372 (1989) C1. (d)
J.A.M. van Beek, G. van Koten, M.J. Ramp, et al., Inorg.
Chem. 30 (1991) 3059. (e) J.W.J. Knapen, A.W. van der Made,
J.C. de Wilde, P.W.M.N. van Leenwen, P. Wijkens, D.M.
Grove, G. van Koten, Nature 372 (1994) 659. (f) C. Granel, Ph.
Dubois, R. Je´roˆme, Ph. Teyssie´, Macromolecules 29 (1996) 8576.
[5] P. Dani, T. Karlen, G. van Koten, in preparation.
[6] (a) M. Gupta, C. Hagen, R.J. Flesher, W.C. Kaska, C.M.
Jensen, Chem. Commun., (1996) 2083 (Corrigendum: Edam,
Ibid (1996) 2687). (b) M. Gupta, W.C. Kaska, C.M. Jensen, Ibid
(1997) 461.
3.9. Attempted 1,4-addition reaction using a
[Li2MnCl4/NC(H)N] system
[7] J.G. Donkervoort, J.L. Vicario, J.T.B.H. Jastrzebski, G. Cahiez,
G. van Koten, Recl. Trav. Chim. Pays-Bas 115 (1996) 547.
[8] (a) D.F. Evans, J. Chem. Soc. (1959) 2003. (b) J. Loliger, R.
Scheffold, J. Chem. Ed. 49 (1972) 646.
The 1,3-bis[(dimethylamino)methyl]benzene (14.5
mmol) was added to a freshly prepared solution of
[Li2MnCl4] (14.5 mmol) in THF (50 ml). Immediately,
an off-white precipitate formed. To this suspension
copper(I) chloride (1.72 mmol) and pulegone (48.3
mmol) were subsequently added at 5°C. The Grignard
reagent (25.6 ml of a 1.95 M solution in THF) was
added drop-wise at such a rate that the temperature
did not exceed 7°C. Complete addition of the Grig-
nard reagent was followed by stirring the reaction
mixture for 45 min at 5°C. The reaction mixture was
quenched with 100 ml of 4 M HCl (aq.). The work
up procedure was identical to the corresponding 2/
CuCl-catalyzed 1,4-addition reaction to give an in-
tractable mixture of high boiling products.
[9] A. Gerold, J.T.B.H. Jastrzebski, C.M.P. Kronenburg, N.
Krause, G. van Koten, Angew. Chem. Int. Ed. Engl. 36 (1997)
755.
[10] M.P. Hogerheide, D.M. Grove, J. Boersma, J.T.B.H. Jastrzeb-
ski, H. Kooijman, A.L. Spek, G. van Koten, Chem. Eur. J. 1
(1995) 343.
[11] J.T.B.H. Jastrzebski, P.A. van der Schaaf, J. Boersma, G. van
Koten, Organometallics 8 (1989) 1373.
[12] T.C. Flood, in: C.P. Casey (Ed.), Comprehensive Organometallic
Chemistry, vol. 6, Oxford, Elsevier, 1995.
[13] (a) J.T.B.H. Jastrzebski, G. van Koten, M. Konijn, C.H. Stam,
J. Am. Chem. Soc. 104 (1982) 5490. (b) P. Steenwinkel, S.L.
James, D.M. Grove, N. Veldman, A.L. Spek, G. van Koten,
Chem. Eur. J. 2 (1996) 1440.
[14] E. Wehman, G. van Koten, C.J.M. Erkamp, D.M. Knotter,
J.T.B.H. Jastrzebski, C.H. Stam, Organometallics 8 (1989) 94.
[15] S. Nunomoto, Y. Kawakami, Y. Yamashita, J. Org. Chem. 48
(1983) 1912.

J.G. Donker6oort et al. / Journal of Organometallic Chemistry 558 (1998) 61–69
69
[16] G. Cahiez, S. Marquais, Synlett. (1993) 45.
[17] (a) M. Tamura, J. Kochi, Synthesis (1971) 303. (b) J.H. Nor-
mant, J. Organomet. Chem. Library 1 (1976) 219.
[18] S.B. Mirviss, J. Org. Chem. 54 (1989) 1948.
[19] R.E. Bowman, W.D. Fordham, J. Chem. Soc. (1952) 3945
[20] R.A. Volkmann, J.T. Davis, C.N. Meltz, J. Org. Chem. 48
(1983) 1767.
[21] G. Cahiez, M. Alami, Tetrahedron Lett. 30 (1989) 3541.
[22] H.C.L. Abbenhuis, N. Feiken, H.F. Haarman, D.M. Grove, E.
Horn, A.L. Spek, M. Pfeffer, G. van Koten, Organometallics 12
(1993) 2227.
[23] E. Gallo, E. Solari, C. Floriani, A. Chiesi-Villa, C. Rizzoli,
Organometallics 14 (1995) 2156.
[24] L. Brandsma, in: Preparative Acetylenic Chemistry, 2nd ed.,
Elsevier, Amsterdam, 1988, pp. 26 and 256.
[25] J.T.B.H. Jastrzebski, G. van Koten, M. Konijn, C.H. Stam, J.
Am. Chem. Soc. 104 (1982) 5490.
[26] J.G. Donkervoort, J.L. Vicario, E. Rijnberg, J.T.B.H. Jastrzeb-
ski, H. Kooijman, A.L. Spek, G. van Koten, J. Organomet.
Chem. 547 (1997) 349.
[27] H. Hiemstra, H. Wijnberg, Tetrahedron Lett. 25 (1997) 2183.
.