Novel [ruthenium(substituted-tetramethylcyclopentadiene) (2-quinolinecarboxylato)(allyl)] hexafluorophosphate complexes as efficient catalysts for highly regioselective nucleophilic substitution of aliphatic allylic substrates

Article abstract of DOI:10.1002/adsc.200800135

Stable [ruthenium(R-substituted-tetramethylcyclopentadiene)(2- quinolinecarboxylato) (1-R′-substituted-allyl) hexafluorophosphate (R = Me, R′ = H, Me, n-Pr, Ph; R = t-Bu, R′ = Me) and [ruthenium(pentamethylcyclopentadiene)(2-quinolinecarboxylato)(1-n-propylallyl)] tetrafluoroborate (4′a), as allylruthenium(IV) complexes, have been synthesized in one step, starting from [ruthenium(R-substituted- tetramethylcyclopentadiene)tris(acetonitrile) hexafluorophosphate or tetrafluoroborate complexes, quinaldic acid, and allylic alcohols. Single stereoisomers are obtained and the X-ray single crystal structure determinations of 3b (R = t-Bu, R′ = Me) and 4′a allow one to specify the preferred arrangement. Complexes 3a (R = R′ = Me) and 3b are involved as precatalysts favoring the formation of branched products in regioselective nucleophilic allylic substitution reactions, starting from ethyl 2-(E)-hexen-1-yl carbonate and chlorohexene as unsymmetrical aliphatic allylic substrates. Phenols, dimethyl malonate, and primary (aniline) and secondary (pyrrolidine, piperidine) amines have been used as nucleophiles under mild basic conditions. For the first time, the regioselectivity in favor of the branched product obtained from purely aliphatic allylic substrates is close to the high regioselectivity previously reached starting from cinnamyl-type substrates in the presence of ruthenium catalysts.

Full text of DOI:10.1002/adsc.200800135

FULL PAPERS
DOI: 10.1002/adsc.200800135
Novel [Ruthenium(substituted-tetramethylcyclopentadiene)
(2-quinolinecarboxylato)(allyl)] Hexafluorophosphate Complexes
AHCTREUNG
as Efficient Catalysts for Highly Regioselective Nucleophilic
Substitution of Aliphatic Allylic Substrates
Hui-Jun Zhang,a,b Bernard Demerseman,b Loïc Toupet,b Zhenfeng Xi,a,*
and Christian Bruneaub,*
a
Beijing National Laboratory for Molecular Sciences (BNLMS), Key Laboratory of ioorganic Chemistry and Molecular
Engineering of Ministry of Education, College of Chemistry, Peking University, Beijing 100871, Peopleꢀs Republic of
China
Fax : (+86)-10-6275-1708; e-mail: zfxi@pku.edu.cn
Sciences Chimiques de Rennes, UMR 6226 CNRS-UniversitØ de Rennes 1, Catalyse et OrganomØtalliques, UniversitØ de
b
Rennes 1, 35042 Rennes Cedex, France
Fax : (+33)-2-2323-6939; e-mail: christian.bruneau@univ-rennes1.fr
Received: March 5, 2008; Published online: May 21, 2008
Supporting information for this article is available on the WWW under http://asc.wiley-vch.de/home/.
Abstract: Stable [ruthenium(R-substituted-tetrame- products in regioselective nucleophilic allylic substi-
thylcyclopentadiene)(2-quinolinecarboxylato)(1-R’-
tution reactions, starting from ethyl 2-(E)-hexen-1-yl
substituted-allyl) hexafluorophosphate (R=Me, R’= carbonate and chlorohexene as unsymmetrical ali-
H, Me, n-Pr, Ph; R=t-Bu, R’=Me) and [ruthe- phatic allylic substrates. Phenols, dimethyl malonate,
nium(pentamethylcyclopentadiene)(2-quinolinecar-
and primary (aniline) and secondary (pyrrolidine, pi-
boxylato)(1-n-propylallyl)] tetrafluoroborate (4’a), as peridine) amines have been used as nucleophiles
allylruthenium(IV) complexes, have been synthe- under mild basic conditions. For the first time, the re-
sized in one step, starting from [ruthenium(R-substi- gioselectivity in favor of the branched product ob-
tuted-tetramethylcyclopentadiene)tris(acetonitrile)
tained from purely aliphatic allylic substrates is close
hexafluorophosphate or tetrafluoroborate complexes, to the high regioselectivity previously reached start-
quinaldic acid, and allylic alcohols. Single stereoiso- ing from cinnamyl-type substrates in the presence of
mers are obtained and the X-ray single crystal struc- ruthenium catalysts.
ture determinations of 3b (R=t-Bu, R’=Me) and
4’a allow one to specify the preferred arrangement.
Complexes 3a (R=R’=Me) and 3b are involved as Keywords: allyl ligands; allylation; homogeneous
precatalysts favoring the formation of branched catalysis; regioselectivity; ruthenium
Introduction
Since the discovery of the catalytic activity of the
[Ru
ophilic
[Ru(C5Me5)
interest as it provided higher regioselectivities in ty.[5] Thus, the regioselective formation of branched
favor of the chiral branched products.[2]
derivatives from aromatic cinnamyl-type linear sub-
ACHTREUNG
A
N
ACHTREUNG
Furthermore, the substitution of one or two aceto- strates is almost perfect in several reactions involving
nitrile ligands in the cationic complex 1a by various carbon-carbon, oxygen-carbon, and nitrogen-carbon
nitrogen[3] or phosphorus[4] ligands (Scheme 1) has led bond formation. However, the regioselectivity re-
to a family of precatalysts allowing one to extend the mained unsatisfactory starting from unsymmetrical
scope of the catalytic process and to enhance reactivi- aliphatic allylic substrates.[4]
Adv. Synth. Catal. 2008, 350, 1601 – 1609
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1601

Hui-Jun Zhang et al.
FULL PAPERS
AHCTREUNG
Results and Discussion
The addition of 2-quinolinecarboxylic acid to an
orange solution of [Ru
A
G
N
dichloromethane resulted within a few minutes in a
deep violet color accounting for the formation of the
expected intermediate (C). The subsequent addition
of allyl alcohol and further stirring for several hours,
Scheme 1. Ruthenium(II) precatalysts related to complex
1a: (A) P=P
N
N
dine or a-diimine type ligands; (C) N,O =2-quinolinecar-
boxylic acid.
led to a yellow precipitate of [Ru
A
carboxylato)(h3-CH2CHCH2)]
AHCTREUNG
Very recently, it has been shown that the formation
of phenyl 1-vinylbutyl ether from 1-chloro-2-butene
and phenol in the presence of a ruthenium complex
related to type (A) gave a high 20:1 branched to
linear ratio, but with a moderate yield of 36%.[6] Vari-
ous transition metal complexes have recently been in-
voked to tackle the challenge of regioselectivity with
aliphatic substrates. Remarkably, iridium,[7] palladi-
um,[8] and platinum,[9] complexes bearing appropriate
ligands have been successfully used as catalyst precur-
sors to reach good regioselectivies in favor of
branched derivatives. Alternately, the formation of
branched products has also been achieved starting
from branched allylic substrates, but according to a
distinct regiospecific process, using rhodium,[10] or
ruthenium,[11] complexes as catalyst precursors.
On the other hand, 2-quinolinecarboxylic acid in a
transient cationic [Ru(Cp)(N,O-chelate)
(MeCN)]+
U
species analogous to (C) in Scheme 1, has recently
proved to cooperate with the metal center to achieve
the activation of allyl alcohol.[12] From the hydroxy
group of allyl alcohol and the acidic proton of 2-qui-
nolinecarboxylic acid, water was produced in this re-
Scheme 2. Synthesis of [Ru
A
lato)
AHCTREUNG
action besides the stable [Ru(Cp)(2-quinoline- The more soluble complexes [Ru
A
carboxylato) [PF6], 3a (R’=Me), 4a
A
ACHTREUNG
complex which is a key intermediate for the rutheni- (R’=n-Pr), and 5a (R’=Ph) were similarly prepared
um-catalyzed protection of alcohols as their allyl using crotyl alcohol, 2-(E)-hexen-1-ol, and isomeric
ethers, as well as their subsequent deprotection.[13] cinnamyl and a-vinylbenzyl alcohols, respectively. The
These remarkable results also suggested that a gener- tetrafluoroborate salt [Ru
al route leading to [Ru
(h5-C5R5)(2-quinoline- ylato)(h3-CH2CHCHn-Pr)]
carboxylato)
(h3-allyl)]+ as allylruthenium(IV) cations same cation as 4a but arising from [Ru
might be possible. (MeCN)3][BF4] 1’a instead of 1a, was also prepared
We report herein the synthesis of new [Ru- and, more peculiarly, allowed an X-ray single crystal
(C5Me4R)(2-quinolinecarboxylato)(h3-CH2CHCHR’)]- structural study. Carrying a distinct cyclopentadienyl
ACHTREUNG
A
ACHTREUNG
A
ACHTREUNG
A
ACHTREUNG
ACHTREUNG
1602
asc.wiley-vch.de
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 1601 – 1609

Highly Regioselective Nucleophilic Substitution of Aliphatic Allylic Substrates
FULL PAPERS
ligand, the analogous complex [Ru(C5Me4t-Bu)(2-qui-
nolinecarboxylato)(h3-CH2CHCHMe)]
ACHTREUNG
AHCTREUNG
ACHTREUNG
The new allylruthenium(IV) complexes have been
1
characterized by H and 13C NMR spectroscopy, and
elemental analysis. The 1H NMR spectra gave a
straightforward evidence for the presence of the allyl
and cyclopentadienyl ligands besides a quinoline frag-
ment. Only complex 2a involves a symmetrical allyl
ligand whereas two stereoisomers (both as enantio-
meric pairs) might be considered for the other com-
plexes, depending on the position of the R’ group rel-
1
ative to the N,O-chelate. H and 13C NMR spectrosco-
py gave evidence for the selective formation of only
one of them. Complex 5a was obtained as very thin
Figure 2. ORTEP drawing of 4’a showing 50% probability
thermal ellipsoids. The BF4 anion and the CH2Cl2 molecule
needles that retained both acetonitrile and dichloro-
1
methane as disclosed by H NMR spectroscopy. Re- are omitted for clarity. Selected bond lengths [Š] and angles
À
À
À
[8]: Ru1 O1 2.065(1), Ru1 N1 2.165(2), Ru1 C11 2.195(2),
Ru1 C12 2.168(2), Ru1 C13 2.315(2), C11 C12 1.408(3),
markably, the same stereoisomer 5a was obtained
starting from cinnamyl alcohol or a-vinylbenzyl alco-
hol, as linear and branched allylic isomers.
X-Ray single crystal analysis studies of 3b and 4’a
allowed us to specify the preferred arrangement.
Ortep drawings of 3b and 4’a are shown in Figure 1
À
À
À
À
À
À
À
À
C12 C13 1.400(3); O1 Ru1 N1 77.40(6), O1 Ru1 C13
À
À
À
À
80.34(7), N1 Ru1 C11 87.07(7), O1 Ru1 C11 127.17(7),
À
À
À
À
N1 Ru1 C13 118.98(7), C11 Ru1 C13 63.88(8).
and Figure 2, respectively, and selected bond distances boxylate anion at basal positions acting as a chelate
are given in the captions. Cations of 3b and 4a dis- to form a nearly planar metallacycle, and the terminal
played a similar square-pyramidal structure with the carbon atoms of an endo h3-allyl ligand, the usual con-
oxygen and nitrogen atom from the 2-quinolinecar- figuration for comparable allylruthenium(IV) com-
plexes.[5] No more surprising was the observation of
À
only slightly longer CH CH2 bond length [1.418(3) Š
in 3b and 1.408(3) Š in 4’a] relative to the CH CHR’
À
one [1.404(3) Š in 3b and 1.400(3) Š in 4’a], which
accounts for a true allylic coordination. Complexes 3b
and 4’a differ in the presence of a tert-butyl group at
the cyclopentadienyl ring in 3b, and in the presence
of a n-propyl group at the substituted allylic terminal
carbon atom in 4’a. The ruthenium-oxygen and ruthe-
nium-nitrogen bonds in 3b [2.056(1) and 2.173(2) Š,
respectively] are slightly shorter and longer, respec-
tively, relative to 4’a [2.065(1) and 2.165(2) Š, respec-
tively].
The comparison of the ruthenium-carbon bond
lengths involving the unsubstituted terminal allylic
carbon atom [2.196(2) Š in 3b vs. 2.195(2) Š in 4’a] as
well as those involving the medium allylic carbon
atom [2.176(2) Š in 3b vs. 2.168(2) Š in 4’a] and the
substituted allylic terminal carbon [2.327(2) Š in 3b
vs. 2.315(2) Š in 4’a], revealed even weaker differen-
ces between 3b and 4’a. Of interest also is the com-
Figure 1. ORTEP drawing of 3b showing 50% probability
thermal ellipsoids. The PF6 anion is omitted for clarity. Se-
lected bond lengths [Š] and angles [8]: Ru1-O2 2.056(1),
parison of these structural features to those of the
allylruthenium(IV)
complex
[Ru
A
A
ACHTREUNG
À
À
Ru1-N1 2.173(2), Ru1 C14 2.196(2), Ru1 C15 2.176(2),
Ru1 C16 2.327(2), C14 C15 1.418(3), C15 C16 1.404(3);
indeed, already disclosed a very similar ruthenium-
À
À
À
À
allyl tetrahedral arrangement [Ru CH2: 2.191(4) Š,
À
À
À
À
À
À
O2 Ru1 N1 76.87(5), O2 Ru1 C16 78.45(6), N1 Ru1 C14
85.96(6), O2 Ru1 C14 126.04(6), N1 Ru1 C16 116.39(6),
C14 Ru1 C16 64.33(7).
À
À
À
À
À
À
À
Ru CH: 2.198(4) Š, Ru CH-n-Pr: 2.339(4) Š, CH
[4]
À
À
À
CH2: 1.406(6) Š, CH CH-n-Pr: 1.395(6) Š]. As the
Adv. Synth. Catal. 2008, 350, 1601 – 1609
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1603

Hui-Jun Zhang et al.
FULL PAPERS
preferred stereoisomer in this case, the substituted al-
lylic terminal carbon is located in a cis position rela-
tive to a chlorine atom whereas the unsubstituted ter-
minal allylic carbon atom is located in a cis position
relative to the bulkier phosphorus ligand. Remarka-
bly, 3b and 4’a similarly showed the substituted allylic
terminal carbon located in a cis position relative to
the oxygen atom from 2-quinolinecarboxylate anion
whereas the unsubstituted terminal allylic carbon
atom is located in a cis position relative to the nitro-
gen atom from the bulkier quinoline moiety.
Starting from cinnamyl-type allylic substrates, nu-
cleophilic allylic substitution reactions catalyzed by 1a
and related complexes as precatalysts, most often oc-
curred according to a highly regioselective formation
of branched products.[5,15] By contrast, the moderate
The results given in Table 1, obtained using 3a and
regioselectivities as yet reached starting from unsym- 3b as precatalyst, disclose a spectacular increase of
metrical aliphatic allylic substrates emphasized the in- the regioselectivity in favor of the branched products,
terest to evaluate the potential of new species of type and also reveal the beneficial effect of a tert-butyl
(C) (Scheme 1). In the light of the putative mecha- group at the cyclopentadienyl ring in 3b. Using aceto-
nism for these ruthenium-catalyzed reactions,[5] the nitrile as solvent, 6 was obtained in a 94:6 B/L ratio
new
[Ru
A
CH2CHCHR’)]
[PF6] derivatives as allylruthenium(IV) with 3b (entries 1 and 2). Furthermore, increased 95:5
complexes might be considered as alternative precata- B/L and 99:1 B/L selectivities, respectively, were
lysts, as compared to previously reported [Ru- reached when dichloromethane was used as solvent
A
N
or Me) cinnamylruthenium(IV) complexes.[16]
were obtained starting from para-cresol (entries 5 and
6) and para-chlorophenol (entries 7 and 8). Precata-
lyst 3a was less efficient for the allylation of ortho-
chlorophenol (entry 9) and phenol (entry 11) leading
to 75:25 B/L and 88:12 B/L selectivities, respectively.
Allylation Reactions Catalyzed by Complexes 3a,b
The low solubility of 2a in usual solvents while the However, the use of the precatalyst 3b resulted in an
crystals of 4a retain dichloromethane, led us to select excellent regioselectivity in favor of the branched
3a and 3b as precatalysts for nucleophilic allylic sub- products, as shown by the achievement of a 94:6 B/L
stitution catalytic experiments. Chlorohexene as a 4:1
mixture of linear n-PrCH=CHCH2Cl and branched n-
PrCH(Cl)CH=CH2 isomers resulting from the reac-
tion between PCl3 and 2-(E)-hexen-1-ol,[17] was select-
Table 1. Allylation of ArOH with chlorohexene in the pres-
ence of 3a and 3b.[a]
ed as a typical aliphatic allylic substrate for the syn-
thesis of allyl aryl ethers, according to a previously re-
ported protocol involving ruthenium-catalyzed nucle-
ophilic substitution with phenols under basic condi-
tions [Eq. (1)].[18] Thus, the addition of various
phenols (1 equiv.) to chlorohexene (0.5 mmol) in the
presence of the ruthenium precatalyst 3a or 3b (3
mol%) and K2CO3 (1 equiv.), in 4.0 mL of solvent for
18 h at ambient temperature, resulted in the complete
consumption of chlorohexene and quantitatively led
to the formation of the allyl aryl ethers 6–10, as mix-
Entry Ar
Catalyst Solvent Product B/L[b]
1
p-MeOC6H4 3a
MeCN
MeCN
CH2Cl2
CH2Cl2
MeCN
MeCN
MeCN
MeCN
MeCN
MeCN
6
6
6
6
7
7
8
8
9
9
94:6
97:3
95:5
99:1
94:6
95:5
93:7
97:3
75:25
94:6
88:12
97:3
2
3
4
p-MeOC6H4 3b
p-MeOC6H4 3a
p-MeOC6H4 3b
5
6
7
8
p-MeC6H4
p-MeC6H4
p-ClC6H4
p-ClC6H4
o-ClC6H4
o-ClC6H4
C6H5
3a
3b
3a
3b
3a
3b
3a
3b
9
tures of branched n-PrCH(OAr)CH=CH2 and linear
G
10
11
12
CH2Cl2 10
CH2Cl2 10
n-PrCH=CHCH2OAr isomers [Eq. (1)]. Under simi-
lar conditions, the involvement of 1a as precatalyst
for these reactions, had already been studied. The for-
mation of the allyl aryl ether 6 from the reaction with
p-methoxyphenol, gave a moderate 61:39 B/L regiose-
lectivity.[18]
C6H5
[a]
Experimental conditions: chlorohexene 0.5 mmol, ArOH
0.5 mmol, K2CO3 0.5 mmol, precatalyst 3a or 3b 3 mol%,
room temperature, 18 h.
[b]
Conversion and B/L ratio as determined by 1H N MR
spectroscopy and GC analysis.
1604
asc.wiley-vch.de
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 1601 – 1609

Highly Regioselective Nucleophilic Substitution of Aliphatic Allylic Substrates
FULL PAPERS
ratio for ortho-chlorophenol (entry 10) and a 97:3 B/L dimethyl malonates 11 [Eq. (2)]. A reaction time of
ratio for phenol (entry 12). 18 h ensured a complete consumption of the allylic
This level of regioselectivity for the formation of carbonate and the linear allylic derivative was pro-
allyl aryl ethers via nucleophilic substitution starting duced as the major compound as indicated by a 26:74
from aliphatic substrates, represents the best result B/L ratio (entry 1). To compare the reactivities of 3a
ever obtained with ruthenium catalysts and competes and 3b to those of the precatalysts 1a or 1’a, ethyl cin-
with the results obtained with other metal catalysts. namyl carbonate was then used as allylic substrate.
The efficiency of precatalysts 3a and 3b for the allyla- The results given in Table 2 show a B/L ratio located
tion of phenols with chlorohexene led us to evaluate in the range 94:6 with 1’a as precatalyst (entry 2) and
their activity for other ruthenium-catalyzed nucleo- 96:4 with 3b as precatalyst (entry 4). Thus, the preca-
philic allylic substitution reactions. The allylation of talysts 1’a, 3a, and 3b maintained the excellent regio-
sodio dimethyl malonate with ethyl cinnamyl and selectivity in favor of the branched isomer of 12 [Eq.
ethyl 2-(E)-hexen-1-yl carbonates [Eq. (2)] was thus (2)] arising from ethyl cinnamyl carbonate, as already
investigated.
reached using 1a as precatalyst.[2]
Attempts to achieve the direct allylation of dimeth-
yl malonate without prior deprotonation confirmed a
lack of reactivity, which could be overcome by addi-
tion of K2CO3 as a base according to path (ii) [Eq.
(2)]. Acetonitrile can be used as solvent and these
conditions promoted the formation of the branched
isomer of 11 starting from ethyl 2-(E)-hexen-1-yl car-
bonate. The best regioselectivity was obtained in the
presence of the precatalyst 3a allowing us to reach a
89:11 B/L ratio (entry 5). Under these conditions, the
precatalyst 3b featuring the tert-butyl-substituted cy-
clopentadienyl ligand failed to enhance this remark-
Preliminary experiments were conducted at room able regioselectivity starting from an unsymmetrical
temperature in THF as solvent according to path (i) aliphatic allylic substrate (entry 6).
[Eq. (2)] using a solution of sodio dimethyl malonate
For comparison, the involvement of 1a as precata-
in THF prepared by reacting dimethyl malonate with lyst under these mild conditions led to the favored
NaH (Table 2). Thus, using 3a as precatalyst, ethyl 2- formation of the linear product as indicated by a
(E)-hexen-1-yl carbonate led to the monosubstituted 23:77 B/L ratio (entry 7). The involvement of K2CO3
as a base to generate the dimethyl malonate anion as
a soft carbonucleophile, might be reasonably assumed
to ensure a large presence of dimethyl malonate rela-
Table 2. Formation of monosubstituted dimethyl malonates tive to dimethyl malonate anion. The preserved di-
11 and 12.[a]
methyl malonate might act as a source of protons for
the regeneration of the type (C) intermediate allow-
Entry
R
Path Catalyst Solvent Product B/L[b]
ing the release of the ethoxide anion as EtOH and
making easier the activation of ethyl 2-(E)-hexen-1-yl
carbonate. Accordingly, the addition of ethyl 2-(E)-
hexen-1-yl carbonate to a solution of 1a and quinaldic
acid in dichloromethane as solvent, led to the quanti-
tative formation of 4a, thus providing an alternate
synthesis of 4a (experiment details in Supporting In-
formation).
Under the conditions (ii), a longer reaction time
was clearly required when starting from ethyl cinnam-
yl carbonate as the allylic substrate. More unexpect-
edly, the cinnamyl substrate led to a 81:19 B/L regio-
selectivity with 1a as precatalyst (entry 8) and to a
still lower 58:42 B/L ratio with 3a as precatalyst
(entry 9), which represent markedly reduced regiose-
lectivities as compared to the experiments performed
under conditions (i) (entries 3 and 4). The regioselec-
tivity was even worse in THF as solvent (entries 10
and 11).
1
2
3
4
n-Pr (i)
3a
1’a
3a
3b
3a
3b
1a
1a
3a
1a
3a
THF
THF
THF
THF
MeCN
MeCN
MeCN
MeCN
MeCN
THF
11
12
12
12
11
11
11
12
12
12
12
26:74
94:6
95:5
Ph
Ph
Ph
n-Pr (ii)
n-Pr (ii)
n-Pr (ii)
Ph
Ph
Ph
Ph
(i)
(i)
(i)
96:4
5
89:11
86:14
23:77
81:19
58:42
80:20
50:50
6[c]
7
8[d]
9[d]
10
11
(ii)
(ii)
(ii)
(ii)
THF
[a]
Experimental conditions: substrate 0.5 mmol, precatalyst
3 mol%, room temperature, 18 h, (i) sodio dimethylmalo-
nate 0.6 mmol or (ii), dimethyl malonate 0.5 mmol,
K2CO3 0.5 mmol.
[b]
Conversion and B/L ratio as determined by 1H N MR
spectroscopy and GC analysis.
[c]
[d]
Conversion: 93%.
40 h.
Adv. Synth. Catal. 2008, 350, 1601 – 1609
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1605

Hui-Jun Zhang et al.
FULL PAPERS
Finally, complexes 3a and 3b were evaluated as pre- importance of the choice of the solvent, conversion
catalysts for allylic substitution reactions leading to and regioselectivity were both disfavored when the
the formation of a carbon-nitrogen bond. Aniline was use of THF as solvent was attempted (entry 6). Under
selected as a primary amine and pyrrolidine and pi- the most favorable conditions for the formation of the
peridine were selected as secondary amines. Starting branched allylamine from pyrrolidine (entry 3), piper-
from ethyl 2-(E)-hexen-1-yl carbonate as the allylic idine afforded the allylic amines 15 with a lower re-
substrate [Eq. (3)], the results are given in Table 3.
gioselectivity as determined by a 73:27 B/L ratio
(entry 7).
The formation of allylruthenium(IV) intermediates
is expected to be easier starting from chlorohexene
relative to ethyl 2-(E)-hexen-1-yl carbonate. For this
reason, the synthesis of the allylic amines 13–15 was
also investigated starting from chlorohexene as the al-
lylic substrate [Eq. (4)].
In acetonitrile as solvent, the reaction with aniline
in the presence of 3a (3 mol%) as precatalyst, result-
ed in a modest consumption of 49% of the allylic car-
bonate after a 16 h reaction time, and revealed a fa-
vored formation of the linear isomer of the allylic
amines 13 (entry 1). No conversion at all had been
previously observed using [Ru
A
Bu)(h3-CH2CHCHPh)]
AHCTREUNG
analogous tert-butyl 2-(E)-hexen-1-yl carbonate as the
allylic substrate.[16a] Under similar conditions, the con-
sumption of ethyl 2-(E)-hexen-1-yl carbonate with
Two distinct procedures were introduced. In path
pyrrolidine was complete and led to a favored forma- (i) [Eq. (4)], the initial presence of one equivalent of
tion of the branched isomer of the allylic amines 14. K2CO3 with respect to the amine provided the forma-
The 85:15 B/L ratio (entry 2) was increased to an ex- tion of the free allylic amines 13–15. In path (ii), no
cellent 93:7 B/L regioselectivity by using dichlorome- K2CO3 was added so that the allylic amines remained
thane instead of acetonitrile as solvent (entry 3). Sur- protected as their hydrochloride form. Whatever the
prisingly, the precatalyst 3b achieved a markedly re- conditions used, the results given in Table 4 show the
duced regioselectivity (entry 4) whereas the involve- favored formation of the branched isomers of the al-
ment of 1a as precatalyst showed the linear product lylic amines 13–15. As compared to the results ob-
to be formed selectively (entry 5). Pointing out the tained using ethyl 2-(E)-hexen-1-yl carbonate as the
Table 3. Formation of allylic amines 13–15 starting from ethyl 2-(E)-hexen-1-yl carbonate.[a]
Entry
R1R2NH
Catalyst
Solvent
Conversion[b]
Product
B/L[b]
1
2
3
4
5
6
7
Aniline
3a
3a
3a
3b
1a
3a
3a
MeCN49
MeCN100
CH2Cl2
CH2Cl2
CH2Cl2
THF
13
14
14
14
14
14
15
33:67
85:15[c]
93:7
60:40
2:98
Pyrrolidine
Pyrrolidine
Pyrrolidine
Pyrrolidine
Pyrrolidine
Piperidine
100
100
100
68
44:56
CH2Cl2
100
73:27[c]
[a]
Experimental conditions: allylic carbonate 0.5 mmol, amine 0.75 mmol, catalyst 3 mol%, room temperature, 16 h.
[b]
[c]
1
Conversion (%) and B/L ratio as determined by H NMR spectroscopy and GC analysis.
1
An unidentified additional minor product was detected by H NMR spectroscopy.
1606
asc.wiley-vch.de
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 1601 – 1609

Highly Regioselective Nucleophilic Substitution of Aliphatic Allylic Substrates
FULL PAPERS
Table 4. Formation of allylic amines 13–15 starting from chlorohexene.[a]
Entry
R1R2NH
Catalyst
Solvent
Path
Conversion[b]
Product
B/L[b]
1
2
3
4
5
6
7
8
Aniline
Aniline
Aniline
Pyrrolidine
Pyrrolidine
Pyrrolidine
Piperidine
Piperidine
3a
3b
3b
3a
3a
3b
3b
3b
Acetone
Acetone
Acetone
CH2Cl2
CH2Cl2
CH2Cl2
Acetone
CH2Cl2
(i)
(i)
100
100
77
100
100
100
100
100
13
13
13
14
14
14
15
15
85:15
88:12
87:13
71:29
82:18
83:17
69:31
64:36
(ii)
(i)
(ii)
(ii)
(ii)
(ii)
[a]
Experimental conditions: chlorohexene 0.5 mmol, amine 0.75 mmol, catalyst 3 mol%, K2CO3 0.75 mmol for path (i) and
none for path (ii), room temperature, 16 h.
Conversion (%) and B/L ratio as determined by H NMR spectroscopy and GC analysis.
[b]
1
allylic substrate, the reaction with aniline now led to vent and the basic conditions. Therefore, a large
the formation of the branched isomer of the allylic choice is offered for variation to reach the optimal
amines 13 as the major product. The 85:15 B/L regio- conditions with respect to regioselectivity.
selectivity reached when 3a was used as precatalyst
The novel and most attractive property of this
(entry 1) was enhanced to a 88:12 B/L regioselectivity family of catalysts is their efficiency to provide, espe-
with 3b as precatalyst (entry 2). The reaction required cially from phenols and soft carbonucleophiles, high
the presence of K2CO3 to reach completion as com- regioselectivities in favor of branched products start-
pared to path (ii), which gave a moderate 77% con- ing from linear aliphatic allylic substrates.
sumption of chlorohexene (entry 3).
On the other hand, the 71:29 B/L regioselectivity
provided by the more basic pyrrolidine and 3a as pre- Experimental Section
catalyst in the presence of K2CO3 (entry 4) could be
increased to a 82:18 B/L selectivity according to path Typical Synthesis of Complexes; [Ru
ACHTREUNG
(ii) using 3a or 3b as precatalyst (entries 5, 6). As was linecarboxylato)(h3-CH2CHCHMe)]
observed from ethyl 2-(E)-hexen-1-yl carbonate as al-
ACHTREUNG
As a typical procedure, dichloromethane (35 mL) was added
to a stirred mixture of 1a (2.01 g, 3.98 mmol) and quinaldic
acid (0.69 g, 3.98 mmol),), to rapidly afford a dark violet-
purple solution to which crotyl alcohol (0.60 mL, an excess)
was added. The mixture was further stirred overnight and
was then evaporated to dryness under vacuum. The remain-
ing solid was dissolved in dichloromethane (20 mL) and the
orange-red solution was then covered with diethyl ether
(100 mL) to afford thin orange needles that were collected
by filtration; yield: 1.97 g (81%).
lylic substrate, piperidine afforded a lowered regiose-
lectivity as compared to pyrrolidine, in acetone
(entry 7) or in dichloromethane (entry 8) as solvent.
Conclusions
New stable allylruthenium(IV) complexes [Ru-
A
Complete experimental details concerning the synthesis of
[PF6] were synthesized as single stereoisomers, in one
step starting from [Ru(C5Me4R)(MeCN)3] [PF6], qui-
1
complexes, H and 13C NMR data and elemental analysis are
A
ACHTREUNG
given as Supporting Information.
naldic acid, and allylic alcohols. They were demon-
strated to be efficient precatalysts in nucleophilic al-
lylic substitution and provided highly regioselective
formation of branched products from allylic substrates
X-Ray Crystallography
The samples were studied with an Oxford Diffraction Xcali-
such as ethyl 2-(E)-hexen-1-yl carbonate and chloro- bur Saphir 3 diffractometer with graphite monochromator.
Crystallographic data are given in Table 5. Data reduction
was carried out with CrysAlis RED.[19] The structures were
solved with SIR-97 which revealed the non-hydrogen
atoms.[20] After anisotropic refinement, many hydrogen
atoms may be found with Fourier difference calculations.
The whole structures were refined with SHELXL97 by full-
matrix least-squares methods on F2 (x , y, z, bij for Ru, P, N,
Cl, F, C and O atoms; x, y, z in riding mode for H atoms).[21]
ORTEP views were prepared with PLATON98.[22] CCDC
668006 (3b) and CCDC 666465 (4’a) contain the supplemen-
hexene. The selective formation of carbon-oxygen,
carbon-carbon, and carbon-nitrogen bonds, was thus
achieved under mild basic conditions using phenols,
dimethyl malonate and primary and secondary
amines as nucleophiles, respectively. With amines, the
regioselectivity strongly depends on the nature of the
amine. Variations of regioselectivity were also ob-
served arising as well from a structural modification
of the C5Me4R cyclopentadienyl ligand, from the
nature of the allylic substrate, the nature of the sol- tary crystallographic data for this paper. These data can be
Adv. Synth. Catal. 2008, 350, 1601 – 1609
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1607

Hui-Jun Zhang et al.
FULL PAPERS
Table 5. Crystallographic data for complexes 3b and 4’a.[a]
(0.015 mmol, 3 mol%), K2CO3 (0.5 mmol, 1 equiv.), and di-
chloromethane (4.0 mL). The mixture was stirred at room
temperature for 18 h and the resulting slurry was filtered for
GC analysis. The filtrate was then concentrated under
vacuum and H NMR spectroscopy analysis solely disclosed
3-(4-methoxyphenoxy)-1-hexene as the branched isomer of
Complex
3b
4’a
Empirical for-
mula
C27H34F6NO2PRu
C27H34BCl2F4NO2Ru
1
Molecular
weight [gmolÀ1]
Crystal size
[mm]
650.59
663.33
allyl aryl ether 6 (B/L ratio=99:1).
Complete experimental details concerning catalytic ex-
0.22”0.20”0.20
0.25”0.12”0.10
1
periments and H NMR spectra and data for compounds 6–
15 are given as Supporting Information.
Crystal system monoclinic
monoclinic
P21/n
Space group
a [Š]
P21/c
8.7309(6)
14.4825(7)
21.2213(9)
96.097(8)
2668.2(3)
4
13.7437(5)
8.2230(2)
25.6883(8)
104.320(3)
2813.0(2)
4
b [Š]
Acknowledgements
c [Š]
b [8]
Volume [Š3]
Z
H.J.Z. is grateful to the Ambassade de France in China and
to the International Program of the Natural Science Founda-
tion of China for Ph. D. financial support.
Density [gcmÀ3] 1.620
1.566
Temperature
[K]
110(2)
110(2)
F
(000)
1328
0.71073
1352
0.71073
References
Mo-Ka radia-
tion, l [Š]
[1] a) T. Kondo, T. Mitsudo, in: Ruthenium in Organic Syn-
thesis, (Ed.: S.-I. Murahashi), Wiley-VCH, Weinheim,
2004, pp 129–151; b) T. Kondo, H. Ono, N. Satabe, T.
Mitsudo, Y. Watanabe, Organometallics 1995, 14, 1945–
1953.
[2] B. M. Trost, P. L. Fraisse, Z. T. Ball, Angew. Chem.
2002, 114, 1101–1103; Angew. Chem. Int. Ed. 2002, 41,
1059–1061.
[3] a) M. D. Mbaye, B. Demerseman, J.-L. Renaud, L.
Toupet, C. Bruneau, Angew. Chem. 2003, 115, 5220–
5222; Angew. Chem. Int. Ed. 2003, 42, 5066–5068;
b) M. D. Mbaye, J.-L. Renaud, B. Demerseman, C. Bru-
neau, J. Organomet. Chem. 2005, 690, 2149–2158.
[4] B. Demerseman, J.-L. Renaud, L. Toupet, C. Hubert,
C. Bruneau, Eur. J. Inorg. Chem. 2006, 1371–1380.
[5] C. Bruneau, J.-L. Renaud, B. Demerseman, Chem. Eur.
J. 2006, 12, 5178–5187.
Absorption co- 0.716
0.800
efficient [mmÀ1]
q range [8]
Index ranges
2.74 to 33.52
2.61 to 32.10
À20<h<19,
À6<k<12,
À38<l<37
26113
À13<h<12,
À21<k<22,
À31<l<31
Reflections col- 47802
lected
Independent re- 9359 (Rint =0.0234) 8933 (Rint =0.0211)
flections
Reflections I> 8075
2s(I)
6961
Data/restraints/ 9359/0/344
parameters
8933/0/344
1.062
Goodness-of-fit 0.854
on F2
Final R indices R1 =0.0309,
R1 =0.0353,
wR2 =0.1041
R1 =0.0481,
wR2 =0.1078
1.295 and À0.982
[6] K. Onitsuka, H. Okuda, H. Sasai, Angew. Chem. 2008,
120, 1478–1479 ; Angew. Chem. Int. Ed. 2008, 47,
1454–1457.
[I> 2s(I)]
R indices (all
data)
Largest diff.
peak/hole
[e·ŠÀ3]
wR2 =0.0864
R1 =0.0386,
wR2 =0.0917
0.802 and À0.846
[7] a) R. Takeuchi, M. Kashio, J. Am. Chem. Soc. 1998,
120, 8647–8655; b) R. Takeuchi, N. Ue, K. Tanabe, K.
Yamashita, N. Shiga, J. Am. Chem. Soc. 2001, 123,
9525–9534; c) F. López, T. Ohmura, J. F. Hartwig, J.
Am. Chem. Soc. 2003, 125, 3426–3427; d) C. Shu, J. F.
Hartwig, Angew. Chem. 2004, 116, 4898–4901; Angew.
Chem. Int. Ed. 2004, 43, 4794–4797; e) C. Shu, A. Leit-
ner, J. F. Hartwig, Angew. Chem. 2004, 116, 4901–4904;
Angew. Chem. Int. Ed. 2004, 43, 4797–4800; f) A. Leit-
ner, S. Shekhar, M. J. Pouy, J. F. Hartwig, J. Am. Chem.
Soc. 2005, 127, 15506–15514; g) D. Polet, A. Alexakis,
K. Tissot-Croset, C. Corminboeuf, K. Ditrich, Chem.
Eur. J. 2006, 12, 3596–3609.
[a]
2
2
w=1/[s
N
G
(0.0695P)2 +0.7219P] (4’a), where P=(F0 +2Fc2)/3.
2
obtained free of charge from The Cambridge Crystallo-
graphic Data Centre via www.ccdc.cam.ac.uk/data_request/
cif.
[8] a) S.-L. You, X.-Z. Zhu, Y.-M. Luo, X.-L. Hou, L.-X.
Dai, J. Am. Chem. Soc. 2001, 123, 7471–7472; b) S. F.
Kirsch, L. E. Overman, N. S. White, Org. Lett. 2007, 9,
911–913.
Typical Catalytic Experiment; Synthesis of Allyl Aryl
Ether 6
As a typical experiment, a sample of p-methoxyphenol
(0.5 mmol, 1 equiv.) was added to a stirred mixture consist-
ing of hexenyl chloride (0.5 mmol, 1 equiv.), precatalyst 3b
[9] A. J. Blacker, M. L. Clarke, M. S. Loft, M. F. Mahon,
M. E. Humphries, J. M. J. Williams, Chem. Eur. J. 2000,
6, 353–360.
1608
asc.wiley-vch.de
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 1601 – 1609

Highly Regioselective Nucleophilic Substitution of Aliphatic Allylic Substrates
FULL PAPERS
[10] P. A. Evans, D. K. Leahy, J. Am. Chem. Soc. 2000, 122,
5012–5013.
Chem. 2005, 117, 4471–4474; Angew. Chem. Int. Ed.
2005, 44, 4397–4400.
[11] M. Kawatsura, F. Ata, S. Hayase, T. Itoh, Chem.
Commun. 2007, 4283–4285.
[12] S. Tanaka, H. Saburi, M. Kitamura, Adv. Synth. Catal.
[17] C. D. Hurd, R. W. McNamee, J. Am. Chem. Soc. 1932,
54, 1648–1651.
[18] M. D. Mbaye, B. Demerseman, J.-L. Renaud, L.
Toupet, C. Bruneau, Adv. Synth. Catal. 2004, 346, 835–
841.
[19] Oxford Diffraction Ltd., Abingdon, Oxfordshire, UK,
2004.
[20] SIR-97 – A New Tool for Crystal Structure Determina-
tion and Refinement: A. Altomare, M. C. Burla, M. Ca-
malli, G. Cascarano, C. Giacovazzo, A. Guagliardi,
A. G. G. Moliterni, G. Polidori, R. Spagna, J. Appl.
Crystallogr. 1998, 31, 74–77.
[21] G. M. Sheldrick, SHELXL97 – Program for the Refine-
ment of Crystal Structures, University of Gçttingen,
Germany, 1997.
[22] A. L. Spek, PLATON – A Multipurpose Crystallo-
graphic Tool, University of Utrecht, The Netherlands,
1998.
2006, 348, 375–378.
[13] a) H. Saburi, S. Tanaka, M. Kitamura, Angew. Chem.
2005, 117, 1758–1760; Angew. Chem. Int. Ed. 2005, 44,
1730–1732; b) S. Tanaka, H. Saburi, Y. Ishibashi, M.
Kitamura, Org. Lett. 2004, 6, 1873–1875.
[14] H. Schumann, K. Zietzke, F. Erbstein, R. Weimann, J.
Organomet. Chem. 1996, 520, 265–268.
[15] a) J.-L. Renaud, B. Demerseman, M. D. Mbaye, C. Bru-
neau, Curr. Org. Chem. 2006, 10, 115–133; b) T.
Kondo, T. Mitsudo, Curr. Org. Chem. 2002, 6, 1163–
1179.
[16] a) I. Fernµndez, R. Hermatschweiler, P. S. Pregosin, A.
Albinati, S. Rizzato, Organometallics 2006, 25, 323–
330; b) R. Hermatschweiler, I. Fernµndez, F. Breher,
P. S. Pregosin, L. F. Veiros, M. J. Calhorda, Angew.
Adv. Synth. Catal. 2008, 350, 1601 – 1609
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1609