Rhodium-catalyzed codimerization of n-butenes with allylic halides

Article abstract of DOI:10.1016/j.apcata.2014.02.013

This paper describes the codimerization of n-butenes with allylic halides such as cinnamyl chloride and crotyl chloride. The highly active catalyst RhCl₃·3H₂O was investigated, resulting 70% yield of codimers with cinnamyl chloride and 1-butene. Furthermore, the reaction is dependent on the molar ratio of allylic chloride to RhCl₃· 3H₂O ratio, where a ratio of 75:1 is optimal at a catalyst concentration of 2 mol%. Additionally allylic compounds such as cinnamyl chloride lead to higher yields of the codimer than other short chained allylic chlorides. So the codimerization of an allylic compound with 1-butene represents a simple method to produce short chained dienes at very mild reaction conditions.

Full text of DOI:10.1016/j.apcata.2014.02.013

Applied Catalysis A: General 476 (2014) 68–71
Contents lists available at ScienceDirect
Applied Catalysis A: General
journal homepage: www.elsevier.com/locate/apcata
Rhodium-catalyzed codimerization of n-butenes with allylic halides
Arno Behra,∗, Zeynep Bayraka, Stephan Peitzb,1
,
Dietrich Maschmeyerb,1, Guido Stochniolb,1
a Department of Biochemical and Chemical Engineering, Chair of Technical Chemistry, Technische Universität Dortmund, Emil-Figge-Str. 66,
44227 Dortmund, Germany
b Evonik Industries AG, Advanced Intermediates, Paul-Baumann-Straße 1, 45772 Marl, Germany
a r t i c l e i n f o
a b s t r a c t
Article history:
This paper describes the codimerization of n-butenes with allylic halides such as cinnamyl chloride and
crotyl chloride. The highly active catalyst RhCl3·3H2O was investigated, resulting 70% yield of codimers
with cinnamyl chloride and 1-butene. Furthermore, the reaction is dependent on the molar ratio of
allylic chloride to RhCl3·3H2O ratio, where a ratio of 75:1 is optimal at a catalyst concentration of 2 mol%.
Additionally allylic compounds such as cinnamyl chloride lead to higher yields of the codimer than other
short chained allylic chlorides. So the codimerization of an allylic compound with 1-butene represents a
simple method to produce short chained dienes at very mild reaction conditions.
Received 21 October 2013
Received in revised form 5 February 2014
Accepted 6 February 2014
Available online 18 February 2014
Keywords:
Codimerization
1-Butene
© 2014 Published by Elsevier B.V.
Homogenous catalysis
Rhodium catalyst
Allylic compounds
1. Introduction
manner. For this reason, Behr et al. have proposed a mechanism
that will be explained in more detail below (Fig. 1) [11–13].
The oligomerization of alkenes, especially ethene, propene and
butenes has been studied quite extensively. Most of the investi-
gated catalyst systems consist of a transition metal complex of
metals like nickel, palladium, chromium, titanium and rhodium.
The catalyst selection depends on which alkene is used and on the
desired product [1–7].
The cooligomerization of ethene with butadiene to 1,4-
hexadiene is a highly selective, homogenously catalyzed reaction.
Cramer et al. demonstrated that RhCl3·3H2O is a very active
cooligomerization catalyst for dienes and alkenes [7–10]. Recent
research by Behr et al. has demonstrated that it is also possible to
cooligomerize ethene with conjugated linoleic acids, an oleochem-
ical diene [11–13].
A possible mechanism for the reaction of butadiene with ethene
for the rhodium catalyzed co-oligomerization was explained in
detail by Cramer. Cramer describes that the active species is a
rhodium hydride species [HRhCl2] formed by the reaction of ethene
with RhCl3·3H2O [7–10]. It is assumed that the cooligomerization
of ethene with conjugated linoleic acid proceeds in an analogous
The mechanism consists of two cycles. Cycle I describes the
cooligomerization and cycle II the regeneration of rhodiumIII which
is believed to be the active oxidation state of rhodium. In the first
step, the rhodium hydride reacts as a catalytically active species
under a 1,4-addition with the diene (step A). The diene is bound
in this case as an ␩3-crotyl ligand to the rhodium center. An
associated ethene molecule is inserted between rhodium and the
␩3-crotylbond in subsequent steps (steps B and C). The catalytic
active rhodium hydride and the 1,4-diene may be generated after
a ␤-hydride elimination (step D) [11–14].
As the rhodiumIII hydride could simply be reduced to a rhodiumI
complex (step E), rhodiumI must be reoxidated to rhodiumIII, as
rhodiumI complexes do not catalyze the oligomerization of alkenes.
The reoxidation can be generated by an allylic chloride (RCl), which
leads to the formation of rhodiumIII via an oxidative addition (step
F). In the following steps G and H, a ␤-hydride elimination and
subsequent ethene elimination leads to the formation of the cat-
alytically active rhodium hydride [11–14].
2. Experimental
2.1. Typical codimerization experiment
∗
Corresponding author. Tel.: +49 231 755 2310; fax: +49 231 755 2311.
E-mail address: arno.behr@bci.tu-dortmund.de (A. Behr).
Tel.: +49 2365 49 6047; fax: +49 2365 49 806047.
In
a
representative experiment 0.0469 g of RhCl3·3H2O
1
(0.0178 mmol, 1 mol%) were dissolved in 5 g 1,4-dioxane. Then
http://dx.doi.org/10.1016/j.apcata.2014.02.013
0926-860X/© 2014 Published by Elsevier B.V.

A. Behr et al. / Applied Catalysis A: General 476 (2014) 68–71
69
Fig. 1. Postulated mechanism of the cooligomerization of ethene with conjugated dienes [11–14].
0.8069 g of crotyl chloride (8.911 mmol, 50 mol%) were drop wise
added to the solution. The solution was then transferred to an auto-
clave and 1 g (17.8 mmol) of 1-butene was added. The reaction was
carried out at 70 ◦C for 18 hours and analyzed with GC to determine
the codimer yields.
(13), 116 (4), 115 (27), 103 (4), 102 (6), 91 (19), 89 (5), 79 (4), 78
(6), 77 (14), 76 (4), 75(3), 65 (9), 64 (2), 63 (9), 53 (5), 52 (4), 51 (14),
50 (6).
3. Results and discussion
2.2. Product characterization
In the present paper we have investigated for the first time that
allylic halides can be codimerized with alkenes such as 1-butene.
The postulated mechanism for the codimerization of 1-butene with
crotyl chloride is described in Fig. 2. In the first step, the rhodium
hydride is reduced to a rhodiumI complex by releasing HCl (step A).
The crotyl chloride leads to a reoxidation to rhodiumIII via an oxida-
tive addition (step B). Then a 1-butene molecule is bound to the
rhodium center (step C). Instead of a ␤-hydride elimination in step
D, an insertion of the crotyl ligand between rhodium and 1-butene
occurs. The catalytic active rhodium hydride and the codimer may
be generated after a ␤-hydride elimination (step E).
2.2.1. Codimer 2: (E)-octa-2,5-diene, (E)-octa-2,6-diene,
(E)-5-methylenehept-2-ene
MS (GC): 110 (2) [M+], 95 (6), 82 (2), 80 (12), 79 (3), 68 (5), 67
(9), 56 (5), 55 (100), 54 (11), 53 (13), 52 (3), 51 (7), 50 (4).
2.2.2. Codimer 4: (1E,4E)-hepta-1,4-dien-1-ylbenzene,
(1E,5E)-hepta-1,5-dien-1-ylbenzene,
(E)-(4-methylenehex-1-en-1-yl)benzene
MS (GC): 172 (20) [M+], 157 (4), 155 (2), 144 (12), 81 (7), 143
(100), 142 (9), 141 (14), 139 (2), 130 (5), 129 (40), 128 (63), 127
Fig. 2. Postulated mechanism of the codimerization of crotyl chloride with 1-butene.

70
A. Behr et al. / Applied Catalysis A: General 476 (2014) 68–71
Table 1
Table 2
Results of the catalyst and allylic compounds screening.
Results of the solvent screening.
Catalyst
Allylic compound
Ycodimer (%)
Solvent
Y3 (%)
RhCl(PPh3)3
Rh(CO)2(acac)
[RhCl(C2H4)]2
Cinnamyl chloride
Cinnamyl chloride
Cinnamyl chloride
0
0
0
1,4-Dioxane
44
35
30
15
35
0
Dichloromethane
1,2-Dichloroethane
Toluene
Tetrahydrofurane
Methanol
RhCl3·3H2O
RhCl3·3H2O
RhCl3·3H2O
RhCl3·3H2O
Crotyl bromide
Allyl chloride
Crotyl chloride
Cinnamyl chloride
0
0
12
44
Propylene carbonate
n-Heptane*
0
0
[Ni(MeCN)6][BF4]2
[Ni(MeCN)6][BF4]2
[Ni(MeCN)6][BF4]2
[Ni(MeCN)6][BF4]2
Cinnamyl chloride
Allyl chloride
Crotyl chloride
Crotyl bromide
0
0
0
0
m(solvent)/m(1-butene) = 5, RhCl3·3H2O = 1 mol%, T = 70 ◦C, 750 rpm, t = 18 h, *low
solubility of the catalyst precursor, y: based on n(1-butene).
Table 3
[Ni(MeCN)6][AlCl4]2
[Ni(MeCN)6][AlCl4]2
[Ni(MeCN)6][AlCl4]2
Allyl chloride
Crotyl chloride
Crotyl bromide
0
0
0
Results of the n(cinnamyl chloride)/n(RhCl3·3H2O) ratio variation.
n(Cinnamyl chloride)/
n(RhCl3·3H2O)
n(1-Butene)/
n(cinnamyl chloride)
Y3 (%)
m(1,4-Dioxane)/m(1-butene) = 5, c(cat.) = 1 mol%, T = 70 ◦C, 750 rpm, t = 18 h, y:
based on n(1-butene).
10:1
15:1
25:1
50:1
75:1
100:1
1:0.10
1:0.15
1:0.25
1:0.50
1:0.75
1:1.00
8
11
23
44
46
10
3.1. Variation of the catalyst and the comonomer
Different allylic halides were used to codimerize with 1-butene
in order to define what kind of allylic halides react with 1-butene.
For this purpose, we ran reactions with allyl chloride, crotyl chlo-
ride, cyrotyl bromide and cinnamyl chloride, yielding the codimers
2 and 3. The codimer yields of 2 and 3 represents the total yield of
the isomers in each box (Fig. 3). Both isomers are formed in each
reaction, but could not be separated by GC.
m(1,4-dioxane)/m(1-butene) = 5, RhCl3·3H2O = 1 mol%, T = 70 ◦C, 750 rpm, t = 18 h, y:
based on n(1-butene).
Table 4
Results of the n(crotyl chloride)/n(RhCl3·3H2O) ratio variation.
n(Crotyl
n(1-Butene)/
Y2 (%)
chloride)/n(RhCl3·3H2O)
n(crotyl chloride)
The three metal complexes RhCl3·3H2O, [Ni(MeCN)6][BF4]2 and
[Ni(MeCN)6][AlCl4]2,[11–14] which are known to be catalytically
active in oligomerization reactions, were used as precursors. The
results in Table 1 show that the codimerization of 1-butene and an
allylic compound cannot be catalyzed by these nickel complexes. In
comparison, the reactions with 1 mol% of RhCl3·3H2O only led to a
product formation for crotyl chloride and cinnamyl chloride, where
the codimerization of cinnamyl chloride and 1-butene yielded a
maximum yield of 44% of the codimers. Additionally, an increasing
yield of the codimers was observed with larger alkyl chains in the
allylic position of the educt. This can be explained by the postulated
mechanism in which an allylic rhodium species is formed in step
B (Fig. 2). The allylic anion can be stabilized most effectively by a
benzyl chain, due to its resonance stabilization.
Looking back on the mechanism of the reaction, we investigated
whether it is possible to catalyze the reaction with a rhodiumI
complex. In reactions with rhodiumIII complexes, a reduction to
rhodiumI must first occur in order to form the allylic rhodium com-
plex. The direct use of rhodiumI obviates this step. However, the
reaction with three different rhodiumI complexes shows no active
Rh species is formed as no product formation was observed.
As the reaction with cinnamyl chloride and RhCl3·3H2O led to
the best results, all subsequent optimization reactions were carried
out using these compounds.
10:1
25:1
50:1
75:1
1:0.10
1:0.25
1:0.50
1:0.75
–
3
11
12
10
0
50, without 1-butene
m(1,4-dioxane)/m(1-butene) = 5, RhCl3·3H2O = 1 mol%, T = 70 ◦C, 750 rpm, t = 18 h, y:
based on n(1-butene).
3.3. Influence of the n(allylchloride)/n(RhCl3·3H2O) ratio
As a further influence factor the ratio of n(allylchloride)/
n(RhCl3·3H2O) was investigated for cinnamyl chloride and crotyl
chloride. The n(allylchloride)/n(RhCl3·3H2O) ratio is of great impor-
tance as the C Cl bond of the allylchloride is split by an oxidative
addition of rhodiumI. It is therefore crucial to determine the best
ratio of n(allylchloride)/n(RhCl3·3H2O) in order to avoid a catalyst
blockage.
The results with cinnamyl chloride are listed in Table 3. As
expected an increased yield of 3 with an increased n(cinnamyl
chloride)/n(RhCl3·3H2O) ratio from 10:1 to 50:1 was observed.
When viewing the entire results in Table 3, it is noteworthy that the
cinnamyl chloride was nearly entirely converted to 3 at n(cinnamyl
chloride)/n(RhCl3·3H2O) ratios between 10:1 and 50:1 yielding the
codimer 3. Hypothetically 10–50% of the codimer 3 could be built.
The yield of 3 is 8% at the lowest ratio of 10:1 and 44% at a ratio of
50:1. No significant change in the yield was obtained by increasing
the ratio further to 75:1 (46% of 3). An additional increasing of the
ratio to 100:1 leads to a blockage of the catalyst and only 10% of
codimer 3 was built.
3.2. Solvent screening
Since the solvent plays an important role in homogeneous catal-
ysis, the solvent for the codimerization of 1-butene and an allylic
halide was investigated, too. Solvents with different polarities were
selected. As shown in Table 2, the solvents methanol, propylene
carbonate and n-heptane are not suitable for the codimeriza-
tion of cinnamyl chloride and 1-butene. Comparable yields of the
codimer 3 were obtained with dichlormethane, 1,2-dichlorethane
and tetrahydrofurane. The highest yield of 3 under these conditions
is 44% in 1,4-dioxane.
Different results were observed for crotyl chloride (Table 4). The
n(crotyl chloride)/n(RhCl3·3H2O) ratio had no great influence on
the yield of 2. An increase in yield from 3% to only 10% at a n(crotyl
chloride)/n(RhCl3·3H2O) ratio of 10:1 and 75:1 was observed. These
results confirm the hypothesis of chapter 3.1 that crotyl anions are
less stable than cinnamyl anions.

A. Behr et al. / Applied Catalysis A: General 476 (2014) 68–71
71
Fig. 3. Reactions of 1-butene with allylic halides.
Table 5
75:1 at 2 mol% rhodium led to maximum yield of 70% 3. But similar
to the reaction with 1 mol% catalyst the yield of 3 decreases at a
molar ratio of cinnamyl chloride to RhCl3·3H2O of 100:1 to 25%. A
high n(cinnamyl chloride)/n(RhCl3·3H2O) ratio seems to lead to a
blockage of the catalyst.
Variation of the educt concentration.
m(1,4-Dioxane)/m(cinnamyl chloride)
Y3 (%)
1
2
3
4
5
6
11
14
41
45
44
43
4. Conclusion
(c(Cinnamyl chloride) = 50 mol%, RhCl3·3H2O = 1 mol%, T = 70 ◦C, 750 rpm, t = 18 h, y:
In summary, allylic chlorides and 1-butene could be codimer-
ized, resulting in a low chained codimer with two double bonds.
Furthermore, we have demonstrated that allylic compounds such
as cinnamyl chloride are more suitable than other allylic chlorides,
because the intermediate rhodium allylic complex is more stable
with a benzyl chain than with a simple alkyl chain. The codimer
yield depends on several parameters such as the catalyst concen-
tration, as well as the mass ratio of the solvent to cinnamyl chloride
and molar ratio of cinnamyl chloride to catalyst. Finally, 2 mol% of
the catalyst RhCl3·3H2O had to be used to obtain a high yield of
70% 3.
based on n(1-butene).
Table 6
Results with 2 mol% RhCl3·3H2O.
n(Cinnamyl chloride)/n(RhCl3·3H2O)
Y3 (%)
25:1
50:1
75:1
100:1
20
49
70
25
m(1,4-Dioxane)/m(1-butene) = 5, RhCl3·3H2O = 2 mol%, T = 70 ◦C, 750 rpm, t = 18 h, y:
based on n(1-butene).
Acknowledgement
3.4. Influence of the educt concentration
Z. Bayrak thanks Evonik Industries AG for financial support and
helpful discussions.
In order to establish a diluting effect, further experiments were
carried out at different m(solvent)/m(cinnamyl chloride) ratios
with cinnamyl chloride (Table 5). The results show that concen-
trating the reaction solution is not advantageous as the yield of 3
decreased from 45% at an m(solvent)/m(cinnamyl chloride) ratio of
4–11% at a ratio of 1. A possible explanation for this is a blockage of
the rhodium due to the increasing concentration of reactants and
catalyst in the solution. The solubility of the catalyst decreases with
low solvent concentrations, whereby maybe not every rhodium
atom could be activated for codimerization. Furthermore, diluting
the solution between an m(solvent)/m(cinnamyl chloride) ratio of
3–6 led to similar yields of 4 (41–45%).
References
[1] A. Behr, Angewandte homogene Katalyse, Wiley-VCH, Weinheim, 2008.
[2] A. Behr, P. Neubert, Applied Homogeneous Catalysis, Wiley-VCH, Weinheim,
2012.
[3] P. Kuhn, D. Sémeril, C. Jeunesse, D. Matt, M. Neuberger, A. Mota, Chem. Eur. J.
12 (2006) 5210–5219.
[4] J. Zimmermann, I. Tkatchenko, P. Wasserscheid, Adv. Synth. Catal. 345 (2003)
402–409.
[5] J.T. Dixon, M.J. Green, F.M. Hess, D.H. Morgan, J. Organomet. Chem. 689 (2004)
3641–3668.
[6] Y. Chauvin, D. Commereuc, Y. Glaize, European Patent No. 200.654, 1986.
[7] R. Cramer, J. Am. Chem. Soc. 87 (1965) 4717–4727.
[8] R. Cramer, Inorg. Chem. 1 (1962) 722–723.
[9] R. Cramer, J. Am. Chem. Soc. 86 (1964) 217–222.
[10] R. Cramer, J. Am. Chem. Soc. 89 (1967) 1633–1639.
[11] Q. Miao, Rhodium Catalyzed Cooligomerization of Ethylene with Conjugated
Dienes (dissertation TU Dortmund), 2006.
[12] H. Witte, Entwicklung, Projektierung und Untersuchung eines kontinuierlichen
Miniplantverfahrens zur Herstellung verzweigter Fettstoffe (dissertation TU
Dortmund), 2014.
3.5. Increasing of the catalyst concentration
Reactions with higher catalyst concentration of 2 mol% were
carried out in order to increase the codimer yield. The results
of the n(cinnamyl chloride)/n(RhCl3·3H2O) ratio variation showed
that the cinnamyl chloride was not completely converted in reac-
tions with high ratios like 75:1 due to a catalyst blockage. Thus
an increase of the catalyst concentration from 1 mol% to 2 mol%
should lead to an increase of the yield. As can be seen in Table 6,
the reaction with an n(cinnamyl chloride)/n(RhCl3·3H2O) ratio of
[13] C. Fängewisch, Entwicklung eines Recycling-Konzepts in der Mehrphasenkatal-
yse
– Ein Verfahren zur Synthese verzweigter Fettstoffe (dissertation TU
Dortmund), 2002.
[14] A. Laufenberg, Homogenkatalysierte C,C-Verknüpfung von Oleochemikalien
mit petrochemischen Verbindungen (dissertation RWTH Aachen), 1990.