Rhodium-Catalyzed Bis-Hydroaminomethylation of Linear Aliphatic Alkenes with Piperazine

Article abstract of DOI:10.1002/adsc.201500896

An efficient protocol was developed to prepare a series of dialkylpiperazines via Rh-catalyzed bis-hydroaminomethylation of linear aliphatic alkenes with piperazine. The well-known Rh/Biphephos catalytic system was applied, yielding the desired dialkylpiperazines within six tandem catalytic steps, already at low catalyst loadings of 0.1 mol%. For the model alkene 1-octene, good yields and linearities of 80% and 77:23, respectively, were achieved under optimized conditions. Influences on the catalytic system regarding n/iso ratio, possible side reactions and the reaction path are discussed on the basis of yield vs. time plots and parameter optimization. With the developed general protocol, other linear, functionalized and branched substrates were effectively transformed to the corresponding linear N,N-disubstituted piperazines.

Full text of DOI:10.1002/adsc.201500896

FULL PAPERS
DOI: 10.1002/adsc.201500896
Rhodium-Catalyzed Bis-Hydroaminomethylation of Linear
Aliphatic Alkenes with Piperazine
Thomas Seidensticker,a Jonas M. Vosberg,a Karoline A. Ostrowski,a
and Andreas J. Vorholta,*
a
Lehrstuhl Technische Chemie, Fakultät Bio- und Chemieingenieurwesen, Technische Universität Dortmund,
Emil-Figge-Straße 66, 44227 Dortmund, Germany
Fax: (+49)-231-755-2311; phone: (+49)-231-755-2313;
e-mail: andreas.vorholt@bci.tu-dortmund.de (homepage: http://www.tc.bci.tu-dortmund.de)
Received: September 24, 2015; Revised: November 11, 2015; Published online: January 26, 2016
Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/adsc.201500896.
Abstract: An efficient protocol was developed to tions and the reaction path are discussed on the basis
prepare a series of dialkylpiperazines via Rh-cata- of yield vs. time plots and parameter optimization.
lyzed bis-hydroaminomethylation of linear aliphatic With the developed general protocol, other linear,
alkenes with piperazine. The well-known Rh/Biphe- functionalized and branched substrates were effec-
phos catalytic system was applied, yielding the de- tively transformed to the corresponding linear N,N-
sired dialkylpiperazines within six tandem catalytic disubstituted piperazines.
steps, already at low catalyst loadings of 0.1 mol%.
For the model alkene 1-octene, good yields and line-
arities of 80% and 77:23, respectively, were achieved Keywords: carbonylation; high-pressure chemistry;
under optimized conditions. Influences on the cata- homogeneous catalysis; hydroaminomethylation; hy-
lytic system regarding n/iso ratio, possible side reac- droformylation; tandem catalysis
Introduction
Tandem catalytic systems triggered by hydroformy-
lation for the transformation of olefins have emerged
Tandem reactions and related concepts involving tran- as a powerful platform for developing and expanding
sition metal catalysts that allow for the selective and tandem catalytic strategies, not least due to the versa-
atom-efficient synthesis of various functional groups tile chemistry of the aldehyde moiety.[19–22] One of the
and complex carbon scaffolds, within one reaction most prominent examples for tandem reactions under
step, have attracted growing attention in recent hydroformylation condition is hydroaminomethyla-
years.[1–16] Their inherent advantage of reducing both tion (HAM).[23,24] In this three-step tandem reaction
waste and time is very attractive for academic re- amines are formed from olefins, typically catalyzed by
searchers as well as industrial applications. Hence, ligand modified rhodium catalysts (Scheme 1). HAM
more and more general concepts for the merger of consists of an initial hydroformylation of an olefin in
two or more catalytic steps without the need for inter- the presence of syngas to form an aldehyde, which
mittent work-up of the intermediates have been de- subsequently undergoes condensation with the amine
veloped. Of particular usefulness are tandem catalytic substrate present. The thus formed enamine/imine is
systems that allow for the selective synthesis of target finally hydrogenated to the corresponding amine, cat-
molecules, which would not be accessible by a multi- alyzed by the applied rhodium species. Hence, accord-
step approach, because the respective intermediate is ing to the taxonomy of dos Santos and Fogg, HAM is
not or only impracticably isolable. Besides the devel- referred to as an auto-tandem catalytic reaction.[6]
opment of new methodologies for the combination of The development of HAM goes back to the pioneer-
reaction steps, including necessary strategies to ad- ing work of W. Reppe on carbonylation of alkynes
dress catalyst incompatibilities and time resolved and alkenes in the 1940s at BASF/Ludwigshafen
transformations,[17,18] the extension of existing tandem (Germany).[25,26] But only in the last two decades, it
catalytic systems for example, for the conversion of became a general tool for the production of various
renewable resources, has also great potential.[16]
amines (fine chemicals, pharmaceuticals, etc.) from
610
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2016, 358, 610 – 621

FULL PAPERS
Rhodium-Catalyzed Bis-Hydroaminomethylation of Linear Aliphatic Alkenes
Scheme 1. General hydroaminomethylation (HAM) of an alkene with a secondary amine.
olefins, particularly due to the work of Eilbracht,[27–33] (ii) By employing a substrate bearing two reactive
Beller,[23,34–36] and Zhang.[37–42]
sites, a two-fold hydroaminomethylation is con-
ceivable, in which two new independent C N
À
Generally, hydroaminomethylations are, however,
limited to the aforementioned three auto-tandem cat-
alytic steps and only scattered examples deal with the
extension of this sequence. Among them are combi-
nations of HAM with concurrent esterification of the
applied amino acid,[43] formation of the required
syngas atmosphere from CO2 by preliminary reversed
water-gas shift (RWGS) reaction[31] and the access to
linear amines from internal alkenes by isomerizing
hydroaminomethylation.[38,44–46]
bonds are formed under the same reaction condi-
tions [reactions (II) and (III), Scheme 2]. This
can either be accomplished by applying a diolefin
or a diamine in combination with the correspond-
ing stoichiometric amount of amine or olefin, re-
spectively. We refer to the term “bis-hydroamino-
methylation (bis-HAM)” for these reactions, in
accordance with Eilbracht et al.[27]
Another methodology for synthetic intensification
Only a few examples can be found in which either
of HAM is by principally “doubling” the reaction of the aforementioned concepts for the two-fold
À
steps, which allows for the formation of two new C N HAM have been achieved, and no general protocol
bonds under the same reaction conditions. Particular- has been developed so far.[19] In particular, bis-hydro-
ly for the formation of more complex structures, this aminomethylations applying a diamine and a mono-
appears a suitable approach. Two general principles olefin are especially rare. Table 1 summarizes bis-hy-
exist in the two-fold HAM:
droaminomethylations in which piperazine (3), the
most frequently used diamine component, was ap-
(i) The application of primary amines potentially plied.
leads to the formation of tertiary amines via the
Notably, most of the olefins previously used
intermediate secondary amines, if a stoichiometric (Table 1) readily formed linear aldehydes in the initial
amount of olefin is present [reaction (I), hydroformylation with high n:iso selectivities, as in-
Scheme 2]. For differentiation, we term this pro- tended by the authors, due to their structure (cyclic or
cess “dihydroaminomethylation (diHAM)”. If a-branched). Thus, no ligands were needed. In only
ammonia is applied as the amine substrate, a single example, a terminal alkene with no substitu-
a three-fold hydroaminomethylation is possible, ent at the a-position was applied but with major
and under some conditions also very likely, which modification.[30] The initial hydroformylation of an N-
analogously is referred to as “trihydroaminome- allyloxazolidin-2-one derivative was conducted in the
thylation (triHAM)”
absence of piperazine, under the influence of Biphe-
Scheme 2. Distinction between dihydroaminomethylation [diHAM, (I)], and bis-hydroaminomethylation [bis-HAM, (II) and
(III)].
Adv. Synth. Catal. 2016, 358, 610 – 621
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
611

FULL PAPERS
Thomas Seidensticker et al.
Table 1. Bis-hydroaminomethylations of piperazine (3). NBD=norbornadiene; cod=cyclooctadiene.
Olefinic substrate
Category
Catalytic system
Tandem catalyt- Yield Linearity Reference
ic?
[%]
À
internal
alkene
[Rh(NBD)((CH3)2PPh)3]+PF6 ,
no ligands
[47]
[48]
yes
70
–
no, no bis-HAM
products ob-
served
vinylidene
compound
not
given
HRh(CO)(PPh3)3, no ligands
[Rh(cod)Cl]2, no ligands
89
vinylidene
compound
not
given
[30]
[33]
[30]
yes
81–91
vinylidene
compound
not
given
[Rh(cod)Cl]2 no ligands
yes
95
no, piperazine
allylic
compound ligand
[Rh(cod)Cl]2 Biphephos as
not present in 1st 95
step
87:13
linear ali-
phatic 1-
alkene
presented in this work
phos as ligand, to achieve reasonable selectivities for applying a ligand modified rhodium catalytic system.
the linear aldehyde of 87% after 48 h with 1 mol% Additional to the development of a general protocol
[Rh(cod)Cl]2. In a second step, piperazine and addi- for the conversion of terminal olefins by optimizing
tional [Rh] were added and the reductive amination the reactions conditions, the limitations and possible
of the linear aldehyde then furnished the desired bis- side reactions are discussed as well.
HAM product in a one-pot manner.
In our ongoing studies concerning hydroaminome-
thylations and the intensification of tandem catalytic Results and Discussion
protocols, we were interested to see whether bis-
HAM of linear, unbiased alkenes with piperazine as For optimizing the reaction conditions, we chose 1-
the diamine component is possible, without the need octene (1a) and piperazine (3) as our model substrate
for intermittent perturbation. We chose the [Rh]/Bi- system. Scheme 3 displays the anticipated reaction
phephos system for our investigations, as this was al- paths and thus the desired product 7a and corre-
ready shown to catalyze both the hydroformylation sponding intermediates. Besides the linear aldehyde
with high linear selectivities and the reductive amina- 2a, branched aldehydes may be formed in the initial
tion.[30,49,50] Moreover, 1,4-bis-alkylated piperazines hydroformylation, which potentially result in the for-
are accessible in bis-HAM yields, which have been mation of regioisomers of each intermediate and
otherwise synthesized by employing mono-alkylated products, which are not displayed in Scheme 3 for the
piperazines under hydroaminomethylation condi- sake of clarity. The final hydrogenation of enamines
tions.[34,44,49–51] Additionally, symmetrical bis-alkylated to amines in HAM reactions has already been re-
piperazines exhibit biological activity and hence, their vealed as the rate-limiting step in preliminary contri-
straightforward synthesis is of considerable inter- butions. Hence, we envisioned the reaction to proceed
est.[52–54]
via bis-dienamine intermediates 5a and 6a (Path A,
We herein wish to report our investigations towards Scheme 3). However, a reaction path via mono-HAM
the selective, bis-hydroaminomethylation of linear 1- intermediate 4a is also plausible (Path B, Scheme 3).
alkenes with piperazine as the diamine component The catalytic species was derived in situ from the ap-
612
asc.wiley-vch.de
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2016, 358, 610 – 621

FULL PAPERS
Rhodium-Catalyzed Bis-Hydroaminomethylation of Linear Aliphatic Alkenes
Scheme 3. Intended reaction paths for the formation of linear bis-HAM product 7a. Path A: via bis-dienamine 5a; Path B:
via mono-HAM intermediate 4a. Possible branched derivatives and stoichiometric coefficients omitted for clarity.
plied precursor Rh(acac)(CO)2 and Biphephos as the intermediate aldehyde 2a [reaction (II)] to aldol
ligand. Other ligands (Figure 1) did not show suffi- condensate 9a are typical side and consecutive reac-
cient chemoselectivity in preliminary investigations. tions under hydroformylation conditions. However,
Toluene was chosen as solvent at a temperature of a third yet literature unknown reaction proceeded
1208C and a syngas pressure of 40 bar, both repre- under our chosen reaction conditions, which led to
sented typical HAM conditions (see the Supporting products with higher molecular mass (by GC-MS)
Information for experiments with other solvents).
and longer retention time (GC-FID) as the desired
In initial investigations, we noticed a strong influ- bis-HAM product 7a. We assumed the formation of
ence of the syngas ratio upon the outcome of the ex- higher condensates 10a (more than two equivalents
periment. Therefore, the syngas ratio was first opti- aldehyde per piperazine molecule) by consecutive re-
mized systematically in a range of H2/CO from 1 to 4 actions of the intermediate mono-HAM product 4a
(Table 2). Besides the anticipated intermediates and and the aldol condensation product 9a [reaction
products displayed in Scheme 3, side products result- (III)]. This transformation and the resultant structure
ing from unintended side reactions were identified in of the higher condensates 10a will be explored in
the reaction mixture after bis-hydroaminomethylation detail later on.
(Scheme 4). Hydrogenation of the starting alkene 1a
In all experiments from the initial syngas ratio var-
[reaction (I)] to alkane 8a and aldol condensation of iation, both 1-octene (1a) and piperazine (3) were
fully converted and bis-HAM product 7a represented
the major product in the reaction mixture. As expect-
ed, hydrogenation of the starting alkene was more
prominent the higher the syngas ratio (H2/CO) was,
with 18% yield for n-octane (8a) at a ratio of 4
(entry 1.7). A comparable trend was found for the
mono-HAM product 4a, which was yielded in 14% at
a syngas ratio of 4 (entry 1.7). Both hydrogenation
and mono-HAM product formation were only low
until a syngas ratio of 2. Highest yields for the desired
bis-HAM products 7a were achieved with a syngas
ratio between 1 and 2 (entries 1.1–1.5) ranging from
70% to 86%. Nevertheless, especially the n:branched
ratios within the products seemed to be highly sensi-
tive to the syngas ratio: with increasing hydrogen con-
tent from 1 to 1.35, linearity increased while higher
ratios of >1.5 resulted in decreasing linearity. Highest
linear selectivity of 77:23 was observed at a syngas
ratio of 1.35 H2/CO representing the best compromise
Figure 1. Ligands applied in the bis-hydroaminomethylation
of 1-octene with piperazine.
Adv. Synth. Catal. 2016, 358, 610 – 621
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
613

FULL PAPERS
Thomas Seidensticker et al.
Table 2. Bis-hydroaminomethylation of 1-octene with piperazine: Initial screening of syngas ratio.[a]
Substrates
Intermediates
Product
Side Products
Entry
H2/CO ratio
X1a
X3
Y4a
Y5a+6a
Y7a
linear:branched
Y8a
Y9a
Y10a
1.1
1.2
1.3
1.4
1.5
1.6
1
>99
>99
>99
>99
>99
>99
>99
>99
>99
>99
>99
>99
>99
>99
>99
>99
>99
>99
3
7
5
2
0
1
<1
0
0
6
5
10
<1
82
70
80
86
77
50
37
35
82
61:39
66:34
77:23
58:42
63:37
59:41
35:65
82:18
65:35
4
4
3
4
<1
<1
<1
<1
<1
<1
<1
2
8
1.2
1.35
1.5
2
3
4
10
10
4
5
3
1
4
10
6
6
13
14
16
2
12
18
11
5
1,7
1.8[b]
1.9[c]
1.35
1.35
<1
[a]
Reaction conditions: n1a =5 mmol, 3/1a=0.5, 0.1 mol% Rh(acac)(CO)2, 0.3 mol% Biphephos, 5 mL toluene, ptotal =40 bar,
T=1208C, t=16 h, 750 rpm. Conversions (X) and yields (Y) in [%], determined via GC-FID with dibutyl ether as inter-
nal standard.
[b]
[c]
ptotal =20 bar.
total =50 bar,
p
between overall yield and linearity (entry 1.3). Both
Considering typical n:iso ratios for the hydroformy-
product linearity and yield showed to be influenced lation of linear a-olefins with the [Rh]/Biphephos
not only by the ratio but also by the total pressure of system of about 95%,[55,56] these rather low linearities
the syngas. Whereas a low pressure of 20 bar favored were unexpected. Hence, we performed test reactions
product linearity (82%) at low yield for bis-HAM to explore whether hydroformylation under our
product 7a (35%), a higher pressure of 50 bar yielded chosen conditions is generally unselective or if the
82% bis-HAM product 7a at only moderate linearity low regioselectivity is based on other phenomena
(65%) (entries 1.8 and 1.9). At 20 bar syngas (H2/CO (Table 3).
1.35), the activity of the whole reaction sequence was
Indeed, in the absence of an amine component, a re-
lower, resulting in an incomplete hydroaminomethyla- gioselectivity in the hydroformylation of 1-octene (1a)
tion (entry 1.8).
of 95:5 with full conversion after 2 h was achieved
under similar reaction conditions compared to
entry 1.3. Only minor hydrogenation to octane (8a,
5%) and negligible yield for aldol condensation prod-
uct 9a were observed (entry 2.1). Comparable yields
have been observed when equimolar amounts of
water were added (entry 2.2). Hence, a drastic change
in the regioselectivity, for example, by substantial hy-
drolysis of the diphosphite ligand by water liberated
during condensation under HAM conditions, is not
observed. The presence of equimolar amounts of a ter-
tiary diamine component (tetramethylethylenedia-
mine), which intrinsically is unsuitable for HAM, low-
ered the hydroformylation activity (entry 2.3). Under
these conditions, conversion of 1a was not completed,
but regioselectivity was almost untouched (97%,
entry 2.2). We therefore deduce that diamines do not
coordinate to the rhodium metal center to such an
extent that regioselectivity changes significantly. With
N-methylpiperazine, the corresponding HAM product
was formed in 67% yield with a linearity of 93:7
(entry 2.3). We consequently assumed the regioselec-
tivity in the hydroformylation under our chosen bis-
Scheme 4. Observed side-products and their formation in
the bis-hydroaminomethylation of 1a and 3.
614
asc.wiley-vch.de
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2016, 358, 610 – 621

FULL PAPERS
Rhodium-Catalyzed Bis-Hydroaminomethylation of Linear Aliphatic Alkenes
Table 3. Test reactions for investigating n:branched ratios under different conditions.[a]
Substrates Products
Side Products
Y9a
Entry
Additional reagent
X1a
Xdiamine linear:branched
Ymain product
Y8a
Y10a
2.1[b]
2.2[b]
2.3[c]
2.4[d]
none
H2O
>99
>99
94
–
–
–
88
94
93
86
67
95:5
94:6
97:3
93:7
5
6
7
4
<1
1
1
–
–
–
18
tetramethylethylenediamine
N-methylpiperazine
>99
5
[a]
Reaction conditions: n1a =5 mmol, nadditional reagent =5 mmol, 0.1 mol% Rh(acac)(CO)2, 0.3 mol% Biphephos, 5 mL toluene,
total =40 bar (H2/CO=1.35), T=1208C, t=2 h, 750 rpm. Conversions (X) and yields (Y) in [%], determined via GC-FID
p
with dibutyl ether as internal standard.
Main product: C9 aldehydes.
Main product: C9 aldehydes, diamine/olefin 0.5.
Main product: 1-methyl-4-nonylpiperazine, diamine/olefin=1.
[b]
[c]
[d]
HAM conditions to be 93:7. Noteworthy, by applying by subsequent reactions of the linear aldehyde 2a to
the piperazine derivative (entry 2.3), aldol condensa- either aldol condensation product 9a or higher con-
tion products 9a and higher condensates 10a occurred, densates 10a, leading to an overall lower linearity
this led to the assumption that these side reactions within the desired bis-HAM products 7a.
were organocatalyzed rather than Brønsted base-cata-
lyzed.
With these assumptions, a general optimization of
the reaction conditions was performed, in order to
In the contemplation of bis-HAM product linearity, maintain high linearity within hydroformylation and
only those products are designated “linear” that to promote the path towards the desired bis-HAM
result from the two-fold condensation of n-nonanal products. In Table 4, a summary of the performed var-
(2a) with piperazine. Hence, statistically a maximum iations from the general conditions (Table 1,
linearity for bis-HAM products of 0.93”0.93=0.86 entry 1.3) is shown, for detailed results please see the
(=^ 86:14) can be achieved if the regioselectivity of Supporting Information.
93:7 from entry 2.3 is considered. In fact we observed
Highest yields of 91% for the desired products 7a
a maximum selectivity to the linear product of 77% were achieved at low ligand loadings of 0.1 mol% due
(entry 1.3). Consequently, n-nonanal (2a) preferably to a high hydroformylation activity of the ligand-free
reacted in the observed side reactions [reactions (II) Rh-centers, but at the expense of linearity (entry 3.2).
and (III), Scheme 4], in comparison to branched alde- Ligand excess almost completely inhibited the forma-
hydes. Indeed, GC-FID and GC-MS analyses re- tion of bis-HAM products in favor of higher conden-
vealed that aldol condensation products 9a were ex- sates 10a (entry 3.3). Substoichiometric amounts of pi-
clusively and higher condensates 10a were almost perazine of 0.2 equivalents as expected furnished only
fully formed from linear aldehyde 2a. As a result, the low yield (33%) for bis-HAM product 7a, although
n:branched ratio was in situ lowered (D in Figure 2) high linearities were observed (87%, entry 3.4). 0.4
Figure 2. Formation of linear vs. branched bis-HAM products.
Adv. Synth. Catal. 2016, 358, 610 – 621
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
615

FULL PAPERS
Thomas Seidensticker et al.
Table 4. Screening of various reaction parameters.[a]
Substrates
Intermediates
Products
Side Products
Entry Exception from general conditions X1a
X3
Y4a
Y5a+6a
Y7a linear:branched Y8a Y9a
Y10a
3.1
3.2
none
>99
>99
>99 72
>99
>99
>99
>99
>99
>99
>99
5
2
3
<1
<1
9
<1
<1
13
1
<1
12
0
80
91
4
33
75
39
67
47
77:23
62:38
(46:54)
87:13
68:32
56:44
55:45
62:38
3
4
10
8
4
5
4
5
<1 10
0.10 mol% Biphephos
0.62 mol% Biphephos
0.2 equiv. piperazine
0.4 equiv. piperazine
T=808C
<1
1
12
6
2
26
7
3.3
3.4[b]
3.5[c]
3.6
>99
>99
>99
>99 12
>99 16
5
5
4
3.7
3.8
T=1008C
T=1408C
<1
<1
9
3
0
[a]
General conditions: n1a =5 mmol, 3/1a=0.5, 0.1 mol% Rh(acac)(CO)2, 0.3 mol% Biphephos, 5 mL toluene, ptotal =40 bar
(H2/CO=1.35), T=1208C, t=16 h, 750 rpm. Conversions (X) and yields (Y) in [%], determined via GC-FID with dibutyl
ether as internal standard.
33% aldehydes were observed.
5% aldehydes
[b]
[c]
equivalents piperazine gave 75% yield for the desired (entry 3.6) while at 1008C linearity was low (entry 3.7,
bis-HAM products, despite lower selectivity to the 55%). A higher reaction temperature of 1408C also
linear product (68%, entry 3.5). Under these condi- led to lower bis-HAM yield of only 47% at moderate
tions, aldehyde intermediates 2a and aldol products linearity (62%, entry 3.8).
9a were observed for the first time to a significant
At the optimized reaction conditions, 80% of the
extent in lieu of a reaction partner towards HAM desired bis-HAM products 7a were obtained with
products. A lower reaction temperature of 808C was a linearity of 77% within six auto-tandem catalytic
not sufficient for complete enamine hydrogenation steps. During the course of optimization, some rele-
Figure 3. Yield (Y) and conversion (X) vs. time plot in the bis-HAM of 1-octene (1a) with piperazine (3). Reaction condi-
tions: n1a =68.18 mmol, 3/1a=0.5, 0.1 mol% Rh(acac)(CO)2, 0.3 mol% Biphephos, 68 mL toluene, ptotal =40 bar (H2/CO=
1.35), T=1208C, 750 rpm. Conversions (X) and yields (Y) in [%], determined via GC-FID with n-dodecane as internal stan-
dard.
616
asc.wiley-vch.de
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2016, 358, 610 – 621

FULL PAPERS
Rhodium-Catalyzed Bis-Hydroaminomethylation of Linear Aliphatic Alkenes
Figure 4. Yield (Y) vs. time plot of the C8/C9 fraction in the bis-HAM of 1-octene (1a) with piperazine (3). Conditions: see
Figure 3.
Figure 5. Yield (Y) vs. time plot of the bis-product fraction in the bis-HAM of 1-octene with piperazine. Conditions: see
Figure 3.
vant aspects of the reaction were revealed and insight conversion and yield vs. time plots were recorded on
into the promotion of different paths by relevant re-
a larger reaction scale in a 300-mL autoclave
action parameters was gained. In order to understand (Figure 3, Figure 4, and Figure 5).
this complex reaction network in even more detail,
Adv. Synth. Catal. 2016, 358, 610 – 621
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
617

FULL PAPERS
Thomas Seidensticker et al.
Conversion (X) of both starting materials was fast et al.[55] The formation of the hydrogenation product
with 1-octene (1a) being almost fully consumed al- 8a followed a linear trend, presumably due to its for-
ready after about 40 min. Piperazine (3) was fully mation from internal olefins, which have almost equal
converted after only 100 min. Bis-HAM product for- tendency to hydrogenation.
mation already started after about 30 min but only
One could conclude that for highest linearity of bis-
became significant after 100 min, and reached already HAM products 7a, a termination of the reaction after
74% yield after 260 min. Prolongation of the reaction 100 min would be sufficient. However, the detailed
time to 70 h (4162 min) only led to an increase in analysis of the reaction progress within the fraction of
yield of about 5%. Aldol condensation product 9a be- bis-condensation products (5a, 6a, 7a) revealed that
haved like a typical intermediate, proving that indeed hydrogenation of the enamine intermediates re-
a subsequent reaction consumed it. Parallel to the mained the rate-limiting step and, hence, hydrogena-
conversion of aldol intermediate 9a, the yield for the tion was only completed after 200 min (Figure 5). Ad-
higher condensates 10a increased, giving evidence ditionally, a minor amount of bis-monoenamine spe-
that indeed higher condensates contain aldol inter- cies 6a remained in the product mixture. We assumed
mediates as proposed in Scheme 4.
that hydrogenation of branched enamines, which re-
The fact that 1-octene (1a) and piperazine (3) were sulted from the condensation of 3 with branched alde-
not converted in parallel motivated us to take a more hydes 2a-branched, is slower (k7 >k8; k9 >k10,
detailed look in the progress of the C8 and C9 fraction Figure 5). As a result, conversion of bis-monoenamine
of the product mixture (Figure 4).
species 6a after 200 min furnished branched bis-HAM
We were quite surprised to see a good share of the products 7a-branched, rather than linear products,
converted alkene 1a being isomerized to internal ole- which matched the further decrease in linearity after
fins 1a-isomers. After only 35 min, 35% internal iso- most of the branched aldehyde and the bis-dienamine
mers were formed, which subsequently reacted fur- was converted. On the other hand, one can conclude
ther until no alkenes were detected in the product that bis-HAM products 7a indeed arose from two dif-
mixture after 460 min. However, considering the early ferent reactions paths:
consumption of 1a as seen in Figure 3, formation of
2a is faster than isomerization (k1 >k2, Figure 4). In (i) At first, hydroformylation of 1a led to n-nonanal
contrast, no linear aldehyde 2a was detected through-
out the reaction sequence, despite intermediate yields
for the branched aldehydes 2a-branched of up to
16%. Hence, condensation of n-nonanal (2a) with 3 is
much faster than condensation of branched aldehydes
(k5 > > >k6, Figure 4). A very interesting insight into
the progress of the reaction was given by the selectivi-
(2a) which directly condensed with 3 to form bis-
dienamine intermediate 5a as intended in
Scheme 3, Path A and shown in Figure 5. This in-
termediate is subsequently hydrogenated to the
desired bis-HAM product. Via this Path A, most
of the desired linear bis-HAM product 7a was
formed.
ty of linear bis-HAM products 7a: until a reaction (ii) Via the other, subsequent reaction path, octene
time of about 100 min, selectivity towards linear prod-
ucts was on a high level with 87%. Noteworthy, this is
the level that would result if a n:branched ratio of
93% in the initial hydroformylation is considered
(Figure 2). After 100 min, branched aldehydes started
to react and selectivity decreased until no branched
aldehydes were left and the final n:branched ratio for
the bis-HAM products 7a was maintained throughout
the reaction sequence.
isomers (1a-isomers) were converted into
branched aldehyde intermediates (2a-branched),
which reacted more slowly with piperazine (3) or
an intermediate mono-HAM product 4a, respec-
tively (k5 > > >k6, Figure 4). Hydrogenation of
the resulting branched enamine intermediates led
to branched bis-HAM-products 7a-branched at
the expense of linear HAM product selectivity.
Remarkably, although 35% internal olefins 1a-iso-
In summary, bis-HAM of linear aliphatic 1-alkenes
mers were present, not all of these were transformed is a very complex reaction network, in which different
to branched aldehydes, as can be seen from the final reaction paths and intermediates occur, and are influ-
linearity. Hence, not only the isomerization of termi- enced by different reaction parameters to different
nal into internal olefins was feasible under the chosen extents, which makes a linear correlation of the condi-
conditions, but also the back reaction of 1a-isomers to tions very difficult. In consideration of the variety of
1a, which was subsequently converted into linear al- possible side products, intermediates and bis-HAM
dehyde 2a. However, once internal alkenes have been products, the formation of the linear bis-HAM prod-
formed, the formation of branched aldehydes over uct 7a from 1-octene (1a) and piperazine 3 within six
the linear aldehyde 2a dominated (k4 >k3, Figure 4). tandem catalytic steps at a yield of 80% with 77% lin-
This behavior of isomerizing hydroformylation of ter- earity, encouraged us to explore other substrates
minal olefins into linear aldehydes with a [Rh]/Biphe- under our optimized reaction conditions (Table 5).
phos system was previously described by Carpentier
618
asc.wiley-vch.de
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2016, 358, 610 – 621

FULL PAPERS
Rhodium-Catalyzed Bis-Hydroaminomethylation of Linear Aliphatic Alkenes
Table 5. Substrate scope under the optimized reaction conditions.[a]
Substrates
Intermediates
Y4a–k Y(5+6)a–k
Products
Y7a–k
Side Products
Entry
4.1
Olefin
X1a–k
>99
>99
>99
>99
X3
linear:branched
77:23
Y8a–k
Y9a–k
<1
<1
<1
<1
Y10a–k
10
>99
>99
>99
>99
5
<1
<1
<1
<1
80
73
95
66
3
4.2
7
66:34
<1
<1
<1
17
4.3
<1
1
47:53
<1
28
4.4
75:25
4.5
4.6
>99
>99
>99
>99
9
<1
<1
87
65
>99:1
<1
<1
<1
<1
<1
<1
12
5:95
4.7
90[a]
50
11
30
11
–
<1
–
–
4.8
74
70
13[b]
10
56
<1
88
–
<1
<1
5
–
–
4.9
>99
>99
>99
>99
<1
<1
60:40
72:28
<1
8
<1
17
4.10
<1
70
4.11
>99
>99
<1
<1
72
70:30
4
6
17
[a]
Reaction conditions: n1a–k =5 mmol, 3/1a–k=0.5, 0.1 mol% Rh(acac)(CO)2, 0.3 mol% Biphephos, 5 mL toluene, ptotal
=
40 bar (H2/CO=1.35), T=1208C, t=16 h, 750 rpm. Conversions (X) and yields (Y) in [%], determined via GC-FID with
dibutyl ether as internal standard.
35% aldehyde 2g was observed.
Non-hydrogenated 4h.
[b]
[c]
1-Hexene (1b) gave under the optimized reaction was directly applied as the substrate, slightly lower
conditions slightly lower yields for the bis-HAM yields for the bis-HAM products were observed in
product 7b with 73% besides also lower linearity comparison to 1b (65%, entry 4.6). Nevertheless, iso-
(66%, entry 4.1) compared to 1a. Branched 1-hexene merizing hydroformylation yields 5% linear bis-HAM
derivatives such as 3-methyl-1-pentene (1d) and 3,3- product 7a after all, which corresponds to an
dimethyl-1-butene (1e) led to higher linearity within n:branched ratio within the initial hydroformylation
the bis-HAM products with 75% for 7d and >99% of 22%. Cyclic substrates such as cyclohexene (1g)
for 7e, respectively. The double bond in 1e is not able and cyclooctene (1h) did not lead to promising bis-
to isomerize and is in proximity to a bulky substitu- HAM yields (entries 4.7 and 4.8). Hydroformylation
ent. Hence, the linear aldehyde 2e was formed almost of these substrates under the chosen conditions
exclusively, yielding the corresponding bis-HAM seemed to be less active resulting in incomplete con-
product 7e in a very good yield of 87% with almost version of the substrates. For cyclohexene (1g), 35%
perfect linearity. For 1-dodecene (1c), an excellent of the intermediate aldehyde 2g was found in the
bis-HAM product 7c yield was observed with 95%. product mixture, proving that branched aldehydes are
However, linearity for this long-chain linear alkene less active in condensation with piperazine. Although
was rather low with only 47% (entry 4.3), presumably 11% bis-HAM product 7g can be found for cyclohex-
due to the increased possible internal isomers via iso- ene, in contrast to cyclooctene (1h),which did not
merization. If an internal linear alkene (2-hexene 1f) gave any bis-HAM products 7h at all. In both cases,
Adv. Synth. Catal. 2016, 358, 610 – 621
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
619

FULL PAPERS
Thomas Seidensticker et al.
a magnetic stirring bar under air. After closing the reactor
with its upper part and replacing the air by argon, 0.79 mL
1-octene (561.1 mg, 5 mmol, 100 mol%) and 5 mL toluene
(4.35 g) were subsequently introduced by syringe in an
argon counter stream. The sealed reactor was placed over
a stirring plate and left for 30 min. The pressure was adjust-
ed to 34 bar syngas (CO/H2 1:1) and then to 40 bar with H2
(H2:CO 23:17 bar=1.35:1). The reactor was placed in a pre-
heated aluminum plate on a commercial magnetic stirrer at
1208C for 16 h at a stirring rate of 750 rpm. After that, the
reactor was cooled to room temperature using a water/ice
bath and the pressure was carefully released in a fume-
hood. The reactor was opened and for quantitative GC anal-
ysis, an aliquot was withdrawn. In the case of preparative
isolation of the product, the reaction mixture was trans-
ferred to a 100-mL round-bottom flask, the autoclave was
rinsed with toluene and the solvent of the combined solu-
tions was removed using a rotary evaporator. The crude re-
action product was subjected to column chromatography
using cyclohexane/EtOAc 10/1!2/1. The isolated linear bis-
HAM product 1,4-dinonylpiperazine was obtained as a color-
less liquid; yield: 54%. 1H NMR (CDCl3, 258C,
enamine intermediates (5+6g and 5+6h) were the
main products from the reaction path, once again
showing that branched enamines are less active in hy-
drogenation. Styrene (1i) gave very good yields of
88% bis-HAM product 7i, at a remarkable linearity
of 60% (entry 4.9). The developed protocol was also
applicable to the remotely functionalized terminal
olefins methyl 10-undecenoate (1j) and 10-undecenol
(1k). The corresponding a,w-diester 7j and a,w-diol
7k were obtained in high yields with 70% and 72%,
respectively (entries 4.10 and 4.11).
Conclusions
The bis-hydroaminomethylation of linear aliphatic 1-
alkenes was successfully performed for the first time
with high yields and selectivities towards linear prod-
ucts. In the model system of 1-octene and piperazine
as the diamine component, the corresponding bis-al-
kylated piperazine was furnished in 80% yield with
a linear selectivity of 77% within six auto-tandem cat-
alytic steps. Key to success was on the one hand, the
incorporation of the very efficient [Rh]/Biphephos
catalyst system at low loadings of only 0.1 mol% to
ensure high chemo- and regioselectivity. On the other
hand, careful adjustment of the reaction parameters
proved to be crucial to achieve high yields and selec-
tivities and to suppress unwanted side reactions. By
thorough parameter optimizations and recording yield
vs. time plots, a detailed understanding of the reaction
system, the reaction path leading to the desired prod-
ucts and the relation between side products, inter-
mediates and hydroaminomethylation products was
gained. Under optimized reaction conditions, other
olefins such as branched 1-alkenes, internal alkenes,
styrene, cycloalkenes and olefins with alcohol and
ester groups were successfully transformed into the
corresponding bis-hydroaminomethylation products
with moderate to excellent yields and linearities. Es-
pecially the possibility to elegantly link two ester or
alcohol functionalities via a piperazine ring under the
formation of linear a,w-bifunctional components
opens up a new tandem-catalytic methodology for the
atom-economic formation of potential polyester mon-
omers.
3
500.13 MHz): d=0.87 (t, JH,H =7.0 Hz, 6H), 1.27 (m, 24H),
3
1.48 (m, 4H), 2.31 (t, JH,H =8.0 Hz, 4H), 2.47 (br. s, 8H);
13C NMR (CDCl3, 258C, 125.77 MHz): d=14.46 (2C), 23.04
(2C), 27.34 (2C), 28.05 (2C), 29.66–29.98 (6C), 32.26 (2C),
53.71 (4C), 59.31 (2C); EI-MS (70 eV): m/z=338 (M+,
12.52%), 225 (100), 226 (26.17), 70 (18.43), 197 (13.59), 338
(12.52), 168 (9.6), 170 (8.92), 99 (7.71), 156 (7.32), 98 (6.1),
184 (5.64), 182 (4.9); ESI-HR-MS: m/z=339.37313, calculat-
+
+
À
ed for C22H47N2 [M H] : 339.37338.
General Procedure for Yield vs. Time-Plot
Investigations in the Bis-Hydroaminomethylation of
1-Octene with Piperazine
According to the above mentioned procedure, the reaction
was performed in a 300-mL Parr reactor and the scale of the
reaction was increased by the factor 13.636, which equals
the increase in reactor size. For quantitative GC-FID analy-
sis, 4680.3 mg n-dodecane were introduced to the reaction
mixture from the beginning. Aliquots were regularly with-
drawn from the autoclave via a valve by a capillary dipping
into the solution with the aid of the internal pressure. The
capillary was flushed twice with reaction mixture before
0.5 mL solution were filled into a liquid nitrogen precooled
GC vial and immediately frozen and stored in liquid nitro-
gen. For GC-FID analysis, the melted sample was diluted
with 2-propanol and analyzed.
For more detailed experimental information please see
the Supporting Information.
Experimental Section
General Procedure for Bis-Hydroaminomethylation
of 1-Octene with Piperazine
Acknowledgements
In a typical experiment, 1.3 mg Rh(acac)(CO)2 (5 mmol,
0.1 mol%), 11.8 mg Biphephos (15 mmol, 0.3 mol%) and
213.5 mg piperazine (2.5 mmol, 50 mol%) were weighed into
a custom-made stainless steel autoclave equipped with
The authors are very grateful to the “Fonds der Chemischen
Industrie” for financial support and granting a PhD scholar-
ship to T.S. Additionally, we also thank Umicore AG &Co.
KG for the donation of precious metal catalysts.
620
asc.wiley-vch.de
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2016, 358, 610 – 621

FULL PAPERS
Rhodium-Catalyzed Bis-Hydroaminomethylation of Linear Aliphatic Alkenes
References
[31] V. K. Srivastava, P. Eilbracht, Catal. Commun. 2009, 10,
1791–1795.
[1] R. Bunce, Tetrahedron 1995, 51, 13103–13159.
[2] S. Ikeda, Acc. Chem. Res. 2000, 33, 511–519.
[3] G. Poli, G. Giambastiani, A. Heumann, Tetrahedron
2000, 56, 5959–5989.
[32] A. Behr, M. Fiene, C. Buß, P. Eilbracht, Eur. J. Lipid
Sci. Technol. 2000, 102, 467–471.
[33] F. KoÅ, P. Eilbracht, Tetrahedron 2004, 60, 8465–8476.
[34] M. Ahmed, A. M. Seayad, R. Jackstell, M. Beller, J.
Am. Chem. Soc. 2003, 125, 10311–10318.
[4] A. Ajamian, J. L. Gleason, Angew. Chem. Int. Ed.
2004, 43, 3754–3760.
[5] A. Ajamian, J. L. Gleason, Angew. Chem. 2004, 116,
3842–3848.
[35] S. Gülak, L. Wu, Q. Liu, R. Franke, R. Jackstell, M.
Beller, Angew. Chem. 2014, 126, 7448–7450; Angew.
Chem. Int. Ed. 2014, 53, 7320–7323.
[6] D. E. Fogg, E. N. dos Santos, Coord. Chem. Rev. 2004,
248, 2365–2379.
[7] J. M. Lee, Y. Na, H. Han, S. Chang, Chem. Soc. Rev.
2004, 33, 302–312.
[8] J.-C. Wasilke, S. J. Obrey, R. T. Baker, G. C. Bazan,
Chem. Rev. 2005, 105, 1001–1020.
[9] C. Chapman, C. Frost, Synthesis 2007, 1–21.
[10] N. Shindoh, Y. Takemoto, K. Takasu, Chem. Eur. J.
2009, 15, 12168–12169.
[36] X.-F. Wu, X. Fang, L. Wu, R. Jackstell, H. Neumann,
M. Beller, Acc. Chem. Res. 2014, 47, 1041–1053.
[37] G. Liu, K. Huang, C. Cai, B. Cao, M. Chang, W. Wu, X.
Zhang, Chem. Eur. J. 2011, 17, 14559–14563.
[38] G. Liu, K. Huang, B. Cao, M. Chang, S. Li, S. Yu, L.
Zhou, W. Wu, X. Zhang, Org. Lett. 2012, 14, 102–105.
[39] S. Li, K. Huang, J. Zhang, W. Wu, X. Zhang, Org. Lett.
2013, 15, 1036–1039.
[40] S. Li, K. Huang, J. Zhang, W. Wu, X. Zhang, Org. Lett.
[11] J. Zhou, Chem. Asian J. 2010, 5, 422–434.
[12] L. M. Ambrosini, T. H. Lambert, ChemCatChem 2010,
2, 1373–1380.
2013, 15, 3078–3081.
[41] X. Zheng, B. Cao, X. Zhang, Tetrahedron Lett. 2014,
55, 4489–4491.
[13] H. Clavier, H. Pellissier, Adv. Synth. Catal. 2012, 354,
3347–3403.
[42] G. Liu, Z. Li, H. Geng, X. Zhang, Catal. Sci. Technol.
2014, 4, 917–921.
[14] X. Zeng, Chem. Rev. 2013, 113, 6864–6900.
[15] H. Pellissier, Tetrahedron 2013, 69, 7171–7210.
[16] A. Behr, A. J. Vorholt, K. A. Ostrowski, T. Seidenstick-
er, Green Chem. 2014, 16, 982–1006.
[17] T. L. Lohr, T. J. Marks, Nat. Chem. 2015, 7, 477–482.
[18] L. Li, S. B. Herzon, Nat. Chem. 2014, 6, 22–27.
[19] P. Eilbracht, L. Bärfacker, C. Buss, C. Hollmann, B. E.
Kitsos-Rzychon, C. L. Kranemann, T. Rische, R. Rog-
genbuck, A. Schmidt, Chem. Rev. 1999, 99, 3329–3366.
[20] E. Petricci, E. Cini, Top. Curr. Chem. 2013, 342, 117–
149.
[43] A. Behr, T. Seidensticker, A. J. Vorholt, Eur. J. Lipid
Sci. Technol. 2014, 116, 477–485.
[44] M. Ahmed, R. P. J. Bronger, R. Jackstell, P. C. J.
Kamer, P. W. N. M. Van Leeuwen, M. Beller, Chem.
Eur. J. 2006, 12, 8979–8988.
[45] J. F. Hartwig, Science 2002, 297, 1653–1654.
[46] A. Seayad, M. Ahmed, H. Klein, R. Jackstell, T. Gross,
M. Beller, Science 2002, 297, 1676–1678.
[47] F. Jachimowicz, J. W. Raksis, J. Org. Chem. 1982, 47,
445–447.
[48] C. S. Graebin, V. L. Eifler-Lima, R. G. da Rosa, Catal.
[21] K. A. Ostrowski, T. A. Faßbach, A. J. Vorholt, Adv.
Synth. Catal. 2015, 357, 1374–1380.
[22] K. A. Ostrowski, T. A. Faßbach, D. Vogelsang, A. J. Vo-
rholt, ChemCatChem 2015, 7, 2607–2613.
[23] A. Bçrner, M. Beller, B. Wünsch, in Houbel-Weyl, Sci-
ence of Synthesis, Vol. 40a, Georg Thieme Verlag, Stutt-
gart, 2009, pp 111–117.
Commun. 2008, 9, 1066–1070.
[49] E. Petricci, A. Mann, J. Salvadori, M. Taddei, Tetrahe-
dron Lett. 2007, 48, 8501–8504.
[50] J. Briggs, G. T. Whiteker, J. Klosin, Hydroaminomethy-
lation of Olefins, WO Patent WO 2005/077884A2, 2005.
[51] L. Wu, I. Fleischer, R. Jackstell, M. Beller, J. Am.
Chem. Soc. 2013, 135, 3989–3996.
[24] D. Crozet, M. Urrutigoïty, P. Kalck, ChemCatChem
[52] L. M. Zhou, X. S. He, G. Li, B. R. de Costa, P. Skolnick,
2011, 3, 1102–1118.
J. Med. Chem. 1995, 38, 4891–4896.
[25] W. Reppe, Experientia 1949, 5, 93–110.
[26] W. Reppe, H. Vetter, Justus Liebigs Ann. Chem. 1953,
583, 133–161.
[27] C. L. Kranemann, P. Eilbracht, Synthesis 1998, 1, 71–77.
[28] P. Eilbracht, C. L. Kranemann, L. Bärfacker, Eur. J.
Org. Chem. 1999, 1907–1914.
[53] L. Salmon-Chemin, E. Buisine, V. Yardley, S. Kohler,
M. A. Debreu, V. Landry, C. Sergheraert, S. L. Croft,
R. L. Krauth-Siegel, E. Davioud-Charvet, J. Med.
Chem. 2001, 44, 548–565.
[54] F. Devinsky, I. Lacko, D. Mlynarcik, L. Krasnec, Col-
lect. Czech. Chem. Commun. 1982, 47, 1130–1138.
[55] J. Ternel, J.-L. Couturier, J.-L. Dubois, J.-F. Carpentier,
Adv. Synth. Catal. 2013, 355, 3191–3204.
[29] C. Kranemann, P. Eilbracht, Eur. J. Org. Chem. 2000,
2367–2377.
[30] M. A. Subhani, K.-S. Müller, P. Eilbracht, Adv. Synth.
Catal. 2009, 351, 2113–2123.
[56] Y. Brunsch, A. Behr, Angew. Chem. 2013, 125, 1627–
1631; Angew. Chem. Int. Ed. 2013, 52, 1586–1589.
Adv. Synth. Catal. 2016, 358, 610 – 621
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
621