A Scrutiny on the Reductive Amination of Carbonyl Compounds Catalyzed by Homogeneous Rh(I) Diphosphane Complexes

Article abstract of DOI:10.1002/1615-4169(200202)344:2<200::AID-ADSC200>3.0.CO;2-4

The reductive amination of a series of aldehydes with secondary amines and H₂ in the presence of a homogeneous Rh-diphosphane catalyst was studied in order to establish a general mechanism of this reaction and to identify conditions for the improvement of the amine/alcohol ratio in the product. Several possible intermediates as constituents of changing equilibria like half-aminals, N,O-acetals and aminals were observed in the reaction mixture by means of ¹H NMR spectroscopy. In individual trials, these compounds could be successfully hydrogenated under the conditions applied for reductive amination (50 bar H₂ pressure, MeOH). Some evidence is accumulated that half-aminals and N,O-acetals might be key intermediates of the reductive amination. Moreover, it was found that the formation of the undesired product alcohol is likely based on the reduction of the starting carbonyl compound. However, due to numerous equilibria consisting of several intermediates, general conclusions are hard to be drawn. Proof will be given that, in several cases, the efficiency of the reductive amination of aliphatic aldehydes can be significantly improved by prehydrogenation of the cationic [Rh(dppb)(COD)]⁺ complex.

Full text of DOI:10.1002/1615-4169(200202)344:2<200::AID-ADSC200>3.0.CO;2-4

FULL PAPERS
A Scrutiny on the Reductive Amination of Carbonyl Compounds
Catalyzed by Homogeneous Rh(I) Diphosphane Complexes
Vitali I. Tararov,*,a Renat Kadyrov,b Thomas H. Riermeier,b Armin Bˆrner*,a
a
Institut f¸r Organische Katalyseforschung an der Universit‰t Rostock e.V., Buchbinderstr. 5/6, 18055 Rostock, Germany,
Fax: ( ‡ 49)-381-4669324; e-mail: vitali.tararov@ifok.uni-rostock.de; armin.boerner@ifok.uni-rostock.de
b
Degussa AG, Project House Catalysis, Industriepark Hoechst, G 830, 65926 Frankfurt/Main, Germany
Received: November 11, 2001; Accepted: February 15, 2002
Abstract: The reductive amination of a series of N,O-acetals might be key intermediates of the
aldehydes with secondary amines and H2 in the reductive amination. Moreover, it was found that
presence of a homogeneous Rh-diphosphane catalyst the formation of the undesired product alcohol is
was studied in order to establish a general mechanism likely based on the reduction of the startingcarbonyl
of this reaction and to identify conditions for the compound. However, due to numerous equilibria
improvement of the amine/alcohol ratio in the consistingof several intermediates, egneral conclu-
product. Several possible intermediates as constitu- sions are hard to be drawn. Proof will be given that, in
ents of changing equilibria like half-aminals, N,O- several cases, the efficiency of the reductive amina-
acetals and aminals were observed in the reaction tion of aliphatic aldehydes can be significantly
mixture by means of 1H NMR spectroscopy. In improved by prehydrogenation of the cationic
‡
individual trials, these compounds could be success- [Rh(dppb)(COD)] complex.
fully hydrogenated under the conditions applied for
reductive amination (50 bar H2 pressure, MeOH). Keywords: amination; homogeneous catalysis; phos-
Some evidence is accumulated that half-aminals and phane; reduction; rhodium.
Introduction
from the economic and ecological point of views. Up to
now, a plethora of examples is known usingH and
2
The conversion of aldehydes and ketones into primary heterogeneous catalysts.[2] Some of these catalytic
or secondary amines is an important reaction in modern systems have seen even application on an industrial
organic chemistry with a great synthetic potential for scale. In contrast, much less is known about the
application in academia and industry. The most straight- homogeneous version of this reaction. Thus, in 1974,
¬
forward reaction for this chemical transformation Marko and Bakos tested successfully cobalt and rho-
comprises the reductive amination of appropriate dium carbonyls.[4] However, these typical hydroformy-
carbonyl compounds. Usually, it involves the reduction lation catalysts require rather severe conditions. Just
of imines and enamines that are easily available by prior recently, Blaser et al.[5] and Fernandez et al.[6] showed
condensation of ketones and aldehydes, respectively, the interestingpotential of iridium(I) diphosphane
with the appropriate amines. This method requires two complexes for the reductive amination of ketones with
steps. The direct conversion (aminoalkylation), when a substituted anilines.
mixture of carbonyl compound and amine is treated in
the presence of a reducingagent, seems to be superior.
In a preliminary communication, we reported for the
first time that cationic complexes of the type [Rh(P-
However, in this approach the situation might be more P)(COD)]BF4 (P-P: diphosphane, diphosphinite) are
complicated due to the large number of possible efficient precatalysts for the homogeneous DRA of
equilibria and intermediates involved. In general, for aldehydes and a-keto acids under relatively mild con-
the semantic differentiation between both methodolo- ditions (ca. 50 bar H2 pressure, room temperature, molar
gies recently the term indirect reductive amination ratio substrate/precatalyst 500:1, methanol) affording
(IRA) was introduced for the former and direct the desired amines in good yield (Scheme 1).[7]
reductive amination (DRA)[1] or single stage amina-
We could also provide proof that even an asymmetric
version of the homogeneously catalyzed DRA is
tion[2] for the later reaction type.
Hitherto, a range of chemical reducing agents has possible. A first trial to reduce phenylpyruvic acid
been shown to be valuable for DRA giving rise to the together with benzylamine in the presence of a chiral
alkylated amines in good yields.[3] In particular, the Rh(I)-diphosphane complex and hydrogen afforded N-
catalytic DRA with molecular hydrogen is promising benzyl (S)-phenyl alanine in 38% ee.
200
¹ VCH Verlagsgesellschaft mbH, D-69451 Weinheim, 1615-4150/02/34402-200 208 $ 17.50-.50/0
Adv. Synth. Catal. 2002, 344, No. 2

Reductive Amination of Carbonyl Compounds
FULL PAPERS
HNR3R4, H2 (50 bar),
methanol, rt.
acetals 4, aminals 5, imines 6 and enamines 7 (in specific
cases, for example, formaldehyde and ammonia, some
additional intermediates can be found in the equilibrium
which are not considered in this work). The occurrence
of each individual intermediate is influenced by the
nature of carbonyl compound and amine as well as by
the reaction conditions. Recently, we presented evi-
dence that imines[8] as well as enamines[9] possessing
double bonds of different natures can be cleanly
reduced with the same homogeneous Rh-catalyst. It is
R3
R1
R4
O
OH
[Rh(P-P)(COD)]BF4
N
+
R1
R2
R1
R2
R2
R1= Alkyl, Aryl
R2, R3, R4= Alkyl, H
Amine
Alcohol
Scheme 1.
An important characteristic of DRA is the selectivity worth mentioningthat the hydrogenation of enamines
of the process which can be expressed as a ratio of proceeded at lower H2 pressure than the reduction of
product amine to alcohol formed (Pam/Pal). In our imines. In contrast to these reports, much less is known
investigations we found selectivities ranging from 0.05/ about the homogeneous hydrogenation of half-aminals
1 to 12/1.[7] Best selectivities were obtained with steri- 3, N,O-acetals 4 and aminals 5 which lack double bonds.
¬
cally non-hindered aldehydes and amines as reagents. Marko and Bakos postulated half-aminals 3 in order to
Moreover, a good correlation between the selectivity of rationalize amine formation in homogeneously cata-
the reaction and the basicity of the amine was observed. lyzed DRA of PhCHO with secondary amines.[4] Also in
The work detailed here was undertaken to improve heterogeneous catalysis only a very few examples are
the formation of the product amine in the DRA using concerned with the hydrogenation of N,O-acetals[10a, b]
cationic Rh(dppb)-catalysts [dppb ˆ 1,2-bis(diphenyl- and aminals.[10b]
phosphanyl)butane]. For this purpose DRA processes
were studied alongwith the hydrogenation of potential
intermediates.
It has been generally accepted that in the first stage of
the reaction between the carbonyl compound and the
amine the relevant half-aminal 3 is formed. The other
intermediates 4 7 are produced later on the reaction
time scale. This time-dependent formation of inter-
mediates is important in DRA since all equilibria shown
in Scheme 2 are not immediately established in the
Results and Discussion
It is reasonable to assume that the production of amines initial mixture of amine 2 and carbonyl compound 1.
8 based on DRA in methanol is associated with the Moreover, the formation of these intermediates could
formation of adducts and condensation products of be seriously perturbed by the irreversible formation of
carbonyl compound 1 and amine 2 (Scheme 2). Such product amine 8 emerging in the reaction mixture due to
intermediates are preferentially half-aminals 3, N,O- the reduction process.
R3
N
R2
R1
6
O
HO NR3R4
-H2O
+
HNR3R4
R2
R1
R2
R1
1
2
3
R4
R3
R3
R1
R4
R2
H2, cat.
N
+H2O
+MeOH
N
-MeOH -H2O
R2
-MeOH
MeO NR3R4
R5
+MeOH
8
7
R2
R1
4
+HNR3R4
-MeOH
+MeOH
-HNR3R4
R3R4N NR3R4
R2
R1
5
Scheme 2.
Adv. Synth. Catal. 2002, 344, 200 208
201

Vitali I. Tararov, et al.
FULL PAPERS
+
CHO
N
RO
Ph
N
fast
H2, cat.
OH
N
H
2a
Ph
+
Ph
Ph
1a
+ MeOH
3a R = H
4a R = Me
9a
8a
slow
N
Ph
7a
Scheme 3.
In a preliminary experiment on DRA of hydratropal-
dehyde (1a) with piperidine (2a) (1:1 ratio) using
[Rh(dppb)(COD)]BF4 as a precatalyst in methanol
under standard conditions (ca. 50 bar H2 initial pressure,
room temperature, 5 mmol of aldehyde, 0.01 mmol of
the precatalyst in 10 mL of MeOH) we observed
complete conversion of the aldehyde within 2 h
(Scheme 3). The product amine 8a and the correspond-
N
N
N
On-Bu
OH
N
OMe
N
Ph
Ph
Ph
3b
Ph
4b
4c
5b
In a first experiment half-acetal 3b was reduced with
ingalcohol 9a were formed in a ratio of 0.5/1. 1H NMR H2 producinga 1.5/1 mixture of N-benzylpiperidine (8b)
investigations using the same concentrations of 1a and and benzyl alcohol (9b) (Scheme 4).
2a in CD3OD revealed that already after 9 min the
whole amount of the aldehyde was consumed. A heterogeneously catalyzed DRA as a common inter-
complex mixture consistingpreferentially of diastereo-
mediate for amine and alcohol production.[2] Unfortu-
Recently, the half-aminal 3b was postulated in the
meric half-aminals 3a and N,O-acetals 4a were found. nately compound 3b is rather unstable. Slow dispropor-
Surprisingly, the formation of enamine 7a was found to tionation into PhCHO and 5b occurs even in the solid
be rather slow. Only ca. 10% of the enamine was state.[10] Due to this feature it was impossible to show
detected in the reaction mixture after 24 hours and ca. unambiguously whether 3b was responsible for the
30% after 12 days. Interestingly, the individual hydro- formation of alcohol 9b or not.
genation of enamine 7a catalyzed by [Rh(dppb)-
To clarify this problem, the presumably more stable
(COD)]BF4 was extremely slow as compared with the model compound 4c was employed. Hydrogenation of
correspondingDRA process (after 20 h only 80% of
amine 8a was produced).
These findings are in agreement with the participation 3 min). The H NMR experiment with an excess of
4c in MeOH in the presence of [Rh(dppb)(COD)]BF4
gave N-benzylpiperidine (8b) as the sole product (t1/2 ca.
1
of half-aminals or N,O-acetals rather than enamines in CD3OD as solvent showed exchange of butoxide in the
DRA. It is interestingto note that half-aminals in the
N,O-acetal 4c and the formation of the corresponding
heterogeneous DRA were already proposed in the methoxy derivative 4b (Scheme 5). Nevertheless, hydro-
past,[2] but hitherto no direct evidence was given.
genation of independently synthesized 4b yielded again
exclusively 8b (t1/2 ca. 30 sec).
As shown in Scheme 5two different products might be
expected from N,O-acetals 4 through reductive C-O and
C-N bond cleavage, respectively, amine 8b or/and ether
10. The absence of even traces of PhCH2O(n-Bu) and
Reduction of Half-Aminals, N,O-Acetals and
Aminals
To show the generality of half-aminals and N,O-acetals,
respectively, in DRA we initiated a more detailed study
concerningthe hydrogenation of these particular sub-
strates. Due to their facile preparation, half-aminal
3b,[11] N,O-acetals 4b, 4c[12] and aminal 5b[13] were chosen
to represent possible intermediates in the DRA of
PhCHO with piperidine.
[Rh(dppb)(COD)]BF4,
H2 (50 bar),MeOH, rt.
N
OH
OH
N
+
Ph
Ph
Ph
8b
9b
3b
Scheme 4.
202
Adv. Synth. Catal. 2002, 344, 200 208

Reductive Amination of Carbonyl Compounds
FULL PAPERS
In THF when formation of N,O-acetal 4b from aminal
5b is not possible the former is hydrogenated much
faster than the latter (after 7 h quantitative hydro-
genation of 4b: after 30 h only 33% conversion of 5b).
Due to this fact it is reasonable to propose that in MeOH
fast methanolysis of 5b affording 4b takes place and only
the latter is subjected to hydrogenation.
The preferred hydrogenation of N,O-acetals finds
support by the fact that the cyclic N,O-acetal 4d is
reduced under standard conditions, whereas the cyclic
aminal 5c is not affected. This observation may serve as
an indirect proof that aminal 5b is not a pivotal
intermediate in the DRA.
OMe
N
Ph
C-N bond
4b
cleavage
C-O bond
cleavage
OR
N
X
n-BuOH
On-Bu
Ph
Ph
10
R = Me, n-Bu
8b
N
Ph
4c
Scheme 5.
PhCH2OMe was confirmed by GLC analysis in addi-
tional to NMR spectroscopy. Since no C-N bond
cleavage was observed in the hydrogenation of 4b or
4c, half-aminal 3b is also unlikely to be a direct precursor
for alcohol in the homogeneously catalyzed DRA.
Nevertheless, the possibility of the decomposition of
half-aminal 3 mediated by cationic rhodium complexes
giving rise to the starting compounds 1 and 2 and the
N
O
N
N
Ph
Ph
4d
5c
On the basis of these results and takinginto account
subsequent reduction of the aldehyde 1 under the that the hydrogenation of imines and enamines affords
formation of the alcohol cannot be ruled out. exclusively amines, the overall DRA process can be
It is quite surprisingthat, although aminals 5 have written in general as displayed in Scheme 7. Principally,
been shown to be hydrogenated over heterogeneous it can be concluded that alcohol in DRA is produced
catalyst affordingthe correspondingamines 8,[10b] their only by the reduction of the startingcarbonyl compound
intermediacy in DRA has never been discussed. Now we and that the rate of this undesired hydrogenation can be
report for the first time that under the conditions given influenced by amines (vide infra). Nevertheless, with an
above aminal 5b is cleanly and rapidly hydrogenated in appropriate catalyst that is selective for the N-inter-
the presence of [Rh(dppb)(COD)]BF4 (t1/2 ca. 5 min) mediate hydrogenation rather than for the hydrogena-
giving N-benzylpiperidine 8b (Scheme 6). Nevertheless, tion of aldehydes or ketones, high yields of tertiary
its intermediacy in DRA is still questionable for the amines should be achieved. Moreover, as will be shown
followingreasons. Aminal 5b is remarkable stable in the below, water has no influence on the outcome of the
solid state and in aprotic solvents, while in MeOH it DRA in this case.
forms an equilibrium mixture consistingof N,O-acetal
4b and piperidine with the equilibrium beingshifted
towards the N,O-acetal (in accordance with NMR Direct Reductive Amination of Aldehydes with
investigations in CD3OD).
Secondary Amines
After studyingthe reduction of relevant intermediates
we turned our attention to the DRA of aldehydes with
secondary amines.
N
N
Ph
N
5b
H2, cat.
H2, cat.,
MeOH
all N-
intermediates
C
O
HC
N
N
+ MeOH
quant.
-
Ph
8b
N
H
H2, cat.
X
N
MeO
no common intermediates
H2, cat.
Ph
HC OH
4b
Scheme 6.
Scheme 7.
Adv. Synth. Catal. 2002, 344, 200 208
203

Vitali I. Tararov, et al.
FULL PAPERS
H2 (50 bar), MeOH, 10 min
[Rh(P-P)(COD)]BF4
[Rh(P-P)(MeOH)2]BF4
HO
[Rh(dppb)(COD)]BF4,
H2, (50 bar), MeOH
CHO
Ph
N
A
B
+
+
Scheme 9.
N
H
Ph
9b
Ph
8b
2a
1b
0.9/1 up to 2.6/1
mentioned that in our experiments usually a Rh(I)-
diphosphane precatalyst instead of the catalyst was
used. The former is stabilized by an ancillary COD
ligand. It is well-known from other hydrogenation
Scheme 8.
First, DRA of benzaldehyde (1b) with piperidine (2a) reactions that in MeOH complete formation of the
was studied. Immediate hydrogenation of a 1:1 mixture catalytically active species B by reductive replacement
in the presence of [Rh(dppb)COD]BF4 after mixing of the di-olefin by two alcohol molecules in the
gave a fast reaction (t1/2 ca. 17 min) with the formation of precatalysts A can require a considerable period
amine 8b and alcohol 9b in a ratio of 0.9/1 (Scheme 8).
(Scheme 9).[15] Thus, the time for the establishment of
In contrast, hydrogenation of the same mixture kept pre-equilibria in the DRA and the time for prehydroge-
for 30 min before treatment with H2 gave a 2.6/1 mixture nation of the precatalyst proceedingin parallel should
of amine 8b and alcohol 9b. No significant change in the influence the outcome of the DRA.
rates was observed. To rationalize the change in
In order to verify this hypothesis we converted the
selectivity, NMR investigations of the mixture were precatalyst into the catalytically active species by
undertaken. These measurements revealed the time- hydrogenation with H2 (10 min, 50 bar) in MeOH
dependent formation of half-aminal 3b, N,O-acetal 4b before the DRA was initiated.
and aminal 5b (Table 1).
As listed in Table 2 in several instances (runs 3, 6 8),
As shown in Table 1, the sum of the concentrations of a significant influence on the selectivity of DRA of
N-intermediates 3b, 4b and 5b increases whereas the aldehydes with piperidine was observed when the
concentration of PhCHO decreases with the time. In precatalyst was treated with H2 prior to the DRA.
particular, the dramatic increase in the concentration of
In the series of aromatic aldehydes the effect of
4b is worthy of note. This feature is obviously the main substituents in the 4-position was impossible to ration-
reason (see Scheme 7) for the improvement in the amine alize in both cases (Table 2, runs 1 5). When the
selectivity of the DRA with the time. The enhancement precatalyst A is employed the reaction is only slightly
of the amine/alcohol ratio (to 1.4/1) with increase of the sensitive to electronic changes. The nitro group does not
initial piperidine concentration (PhCHO:piperidine ˆ survive under the conditions of the DRA. Aromatic
1:2) can be also rationalized from this point of view.
amines derived from this side reaction subsequently
As outlined above, the establishment of pre-equilibria react with the aldehyde affordinga complex mixture of
is strongly time-dependent and of relevance to the DRA condensation products. The friendly behaviour in DRA
and the hydrogenation of the carbonyl compound. of the phenolic OH group is worthy of mention.
Consequently, different amine/alcohol ratios were ob-
served due to varyingtimes of pre-equilibration. Similar
In the series of aliphatic aldehydes (Table 2, runs 6 8)
a considerable dependency of the selectivity on the
features should also hold for the catalyst. It should be degree of a-alkyl branchingwas observed. The effect
Table 1. Time-dependent composition of an equimolar mixture of PhCHO (1b)/piperidine (2a) in CD3OD.
N
CD3OD
H2O
CHO
Ph
OD
N
OCD3
N
N
N
H
+
CD3OD
N
H
Ph
4b(D)
Ph
3b(D)
Ph
5b(D)
1b
2a
Time
1a [%]
3b [%]
4b [%]
5b [%]
ca. 10 min
30 min
69
36
18
6
3
2
22
58
78
4.69 (s)
3
3
2
6 h
d [ppm][a]
9.9 (br, s)
5.56 (br, s)
3.57 (s)
[a]
Chemical shifts indicated are those of PhCH. Similar chemical shifts measured in DMSO-d6 for compounds of type 3 and 5
were reported in Ref.[14]. Values are closely related to shifts of 3b and 5b in THF-d8 detailed in this work. Chemical shifts in
CDCl3 of 4b: 4.72 ppm and of 5b: 3.48 ppm.
204
Adv. Synth. Catal. 2002, 344, 200 208

Reductive Amination of Carbonyl Compounds
FULL PAPERS
Table 2. The influence of the aldehyde structure on the selectivity of DRA with piperidine (aldehyde/piperidine ratio 1:2) using
precatalyst A and prehydrogenated precatalyst B.[a]
Run
Aldehyde
Pam/Pal without prehydrogenation
Pam/Pal with prehydrogenation
1
2
3
4
5
6
7
8
4-HOC6H4CHO
4-MeOC6H4CHO
PhCHO
27.0
0.8
1.4
1.6
>99
2.1
4.7
1.2
4-ClC6H4CHO[b]
4-NO2C6H4CHO[b]
PhCHMeCHO
EtCHMeCHO
n-C7H5CHO
complex mixture
0.9
2.4
complex mixture
7.7
18.0
12.0
>99.5
[a]
Conditions: H2 pressure 50 bar, 5 mmol aldehyde, 0.01 mmol precatalyst, 10 mL MeOH.
[b]
Formation of dimethyl acetals observed was not due to Rh(I) catalysis as was confirmed by separate experiments. Acetal
formation was only found when relevant aromatic aldehydes were kept for 10 15 min in MeOH. However, in the presence
of piperidine no acetals were detected.
Table 3. Comparison of DRA of PhCHO with amines catalyzed with precatalyst A and prehydrogenated precatalyst B.[a]
Amine
pKa of amine
Pam/Pal without prehydrogenation
Pam/Pal with prehydrogenation
pyrrolidine
piperidine
Me2NH
Et2NH
2-methylpiperidine
11.27
11.02
10.73
10.49
10.99
2.9
1.4
0.4
0.1
<0.01
>99.5
4.7
0.7
0.1
<0.01
[a]
Conditions: 5 mmol of aldehyde, 10 mmol of amine, 0.1 mmol of Rh complex, 10 mL of MeOH, 50 bar initial H2 pressure.
became more pronounced when the solvent complex B were found by 1H NMR spectroscopy. Consequently, no
was used instead of the precatalyst A.
The pKa of amines has a strikinginfluence upon the
amine was formed in the DRA. However, it is important
to note that the amine may affect the hydrogenation of
selectivity of the DRA of PhCHO as shown with a series the aldehyde. Thus, in the presence of Et3N the half-time
of simple secondary amines. This effect was already of the reduction of PhCHO was dramatically reduced
noted in our preliminary publication by application of from 85 to 15 min.
the precatalyst A.[7] The same, but more pronounced,
Interestingly, the superior selectivity of the solvent
trend is present usingthe solvent complex B. A complex B was observed only when hydrogenation was
comparison of the results is given in Table 3. started immediately after the addition of PhCHO and
1H NMR investigations of a mixture containing amine. When a mixture of PhCHO and amine was kept
PhCHO and Et2NH (1:1) in CD3OD revealed that the for 30 min prior to the addition of the catalyst B the
rate of the formation of deuterated half-aminal 3c and subsequent hydrogenation gave a lower Pam/Pal ratio.
N,O-acetal 4e was slower than with piperidine as amine This effect was observed when piperidine and pyrroli-
(Scheme 10). Only 6% and 5%, respectively, of these dine were used as amines (compare Pam/Pal ratios 3.5 and
intermediates were observed after 10 min. The corre- 2.2 with the correspondingdata for complex B given in
spondingaminal 5d was not found in the mixture.
Obviously, this effect is related to the increased steric
Table 4).
To find explanations for the decreased selectivity with
demand of the amine in the formation of the half- progressing DRA the influence of water and aminal was
aminal. Indeed, in the case of 2-methylpiperidine having considered. Thus, the DRA of PhCHO with pyrrolidine
pKa 10.99 (10.02 for piperidine) no N-intermediates usingcomplex B was conducted in the presence of
0.55 mmol of water. This amount corresponds to 11% of
all water that is expected after completion of the DRA
process. Under the conditions given in Table 2 we
slow
RO
NEt2
Et2N
NEt2
CHO
Ph
CD3OD
observed no influence of water on the selectivity. It is
interestingthat the addition of an excess of water, e.g.,
11.1 mmol, decreased the Pam/Pal ratio to 11. But this
selectivity is still rather high and thus the conclusion
outlined in Scheme 7 is confirmed. The next candidate
+ HNEt2
X
Ph
Ph
1b
3c(D): R = D
4e(D): R = CD3
5d
Scheme 10.
Adv. Synth. Catal. 2002, 344, 200 208
205

Vitali I. Tararov, et al.
FULL PAPERS
for the decrease of selectivity may be aminal 5 which is Experimental Section
not produced in the first stage of the reaction between
PhCHO and amine. Unfortunately, due to the instability
of this compound in MeOH (vide supra) we were unable
General Remarks
All solvents and liquids used in hydrogenations were distilled
and kept under Ar. NMR spectra were recorded with Bruker
ARX 400. Chemical shifts (d, in ppm) are given for 1H relative
to TMS as internal standard and for 13C relative to residual
CDCl3 peak (77.36 ppm). Spin-spin couplingconstants ( J) are
given in Hz. a-Piperidinobenzyl alcohol (3b),[11] 1-(butoxy-
phenylmethyl)piperidine (4c),[12] 1,1'-benzylidenedipiperidine
(5a)[13] were prepared accordingpublished procedures. All
operations were conducted under Ar.
to propose any experimental evidence for this hypoth-
esis.
The last question to be discussed concerns the
mechanism of the reduction of half-aminal 3, N,O-
acetal 4 and aminal 5. Since diacetal PhCH(OMe)2 can
be excluded as precursor of the product amine it is
necessary to assume that hydrogenation of N,O-adducts
should not proceed via a substitution mechanism. The
only intermediate for hydrogenation is therefore an
‡
iminium cation C ˆ N which can also be produced from
half-aminal 3 and aminal 5. This has been confirmed in
the present work by observation of the interconversion
of these species in MeOH. The time-dependence of the
outcome of DRA processes found suggest that half-
aminal 3, N,O-acetal 4 and aminal 5 may be coordinated
to the Rh centre with different rates (or equilibria).
Moreover, also different rates producingthe iminium
species are possible.
1-(Methoxyphenylmethyl)piperidine (4b)
1,1'-Benzylidenebispiperidine (5a; 17.1 g, 66 mmol) was dis-
solved in 100 mL of n-hexane containingMeOH (5 mL,
123 mmol) and cooled to 0 8C. A 4 M solution of HCl in
dioxane (16.5 mL, 66 mmol) was slowly added and the mixture
was stirred at ambient temperature for 2 h. The precipitated
piperidine hydrochloride was filtered off and washed with an
additional portion of hexane. The clear solution was evapo-
rated and the residual liquid was distilled in vacuum. Yield:
1
74%; bp 78 8C/0.06 mbar; H NMR (CDCl3): d ˆ 1.35 1.44
Conclusions
(m, 2H, g-CH2), 1.44 1.59 (m, 4H, b-CH2), 2.54 (t, 4H, J ˆ
5.4 Hz, a-CH2), 3.37 (s, 3H, OCH3), 4.72 (s, 1H, N-CH-O), 4.74
7.38 (m, 5Harom); 13C NMR (CDCl3): d ˆ 25.1 (g-CH2), 26.5
(b-CH2), 48.9 (a-CH2), 56.7 (OCH3), 98.7 (N-CH-O), 127.6
(CH), 127.8 (CH), 128.1 (CH), 138.7 (C).
The reductive amination (DRA) of a range of aldehydes
and aliphatic secondary amines with cationic rho-
dium(I) complexes was studied. Ratios of product
amine:alcohol by up to 99:1 were achieved. The role of
possible intermediates like half-aminals, N,O-acetals,
aminals in the DRA was investigated by individual
hydrogenations. It is shown that in dependence upon the
structure of the startingcompounds in principal all N-
intermediates could play a role in the DRA. However,
due to the fast and dominant formation of relevant N,O-
General Procedure for DRA
Method A using precatalyst A: A glass beaker with 0.01 mmol
of the precatalyst and a stirringbar was placed in a standard
stainless autoclave (25 mL inner volume) equipped with a
valve and connected to a vacuum pump, Ar and H2 lines. The
acetals there is some evidence that they are key vessel was evacuated and then filled with Ar. The cycle was
repeated 2 3 times. Under a flow of Ar through the open valve
10 mL of MeOH and liquid reaction components (or solution
of components in 10 ml of MeOH) were added by means of
syringes. (In the case of solid aldehydes they were placed
together with the precatalyst.) The valve was closed and the
autoclave was pressurized with H2. The contents of autoclave
was stirred with a magnetic stirrer.
Method Busing prehydrogenated precatalyst B: The auto-
clave containinga glass beaker with the precatalyst and the
stirringbar was deoxygenated as described above. 10 mL of
intermediates in DRA processes. An interestingrela-
tion between the generation of the catalytically active
species by pretreatment of the COD complex with H2
and the DRA proceedingin parallel was found. Thus, by
application of the catalyst in several examples improved
amine:alcohol ratios were achieved. However, in gen-
eral it has to be stated that due to the large variety of pre-
equilibria involved, includingthe acceleratingeffect of
amines on the undesired hydrogenation of the carbonyl
compound, DRA processes are very difficult for mech- MeOH were added through the open valve in an Ar flow and
the precatalyst was hydrogenated for 10 15 min under 50
52 bar of H2 pressure. Then the valve was slowly opened. When
the pressure had dropped to normal a flow of Ar was
immediately allowed to pass. Liquid reaction components
were introduced and then the valve was closed and the
autoclave pressurized with H2. With solid aldehydes the
precatalyst was hydrogenated in 3 mL of MeOH and then
the aldehydes were introduced in the autoclave as a solution in
7 mL of MeOH. (Additionally when usingPhCHO/piperidine
and PhCHO/pyrrolidine mixtures it was confirmed that the last
procedure did not result, within the experimental error, in a
anistic studies. The development of more selective and
active catalysts in the future will, therefore, be mainly
based on trial and error. The data presented above show,
however, that careful screeningand consideration of all
relevant influences allow one to find efficient and
selective homogeneous catalysts for DRA.
206
Adv. Synth. Catal. 2002, 344, 200 208

Reductive Amination of Carbonyl Compounds
FULL PAPERS
Table 4. NMR characterization of alcohols of the general formula R1CH2OH (chemical shifts of aromatic H and C are
omitted) recorded in CDCl3.
R1
CH2OH
1H
R1
1H
13C
13C
4-HOC6H4
4-MeOC6H4 4.47 (s)
Ph
4-ClC6H4
4.43 (s)
64.3
64.7
65.5
64.6
68.9
3.69 (s, 3H, OCH3)
55.6 (OCH3)
4.51 (s)
4.55
Ph(Me)CH 3.55 (dd, 1H, J ˆ 7.0 and 10.7); 3.59 (dd,
1.22 (d, 3H, J ˆ 7.0, CH3), 2.85 (ddd, 1H, 18.1 (CH3),
J ˆ 7.0, CH) 42.8 (CH)
1H, J ˆ 7.0 and 10.7)
Et(Me)CH[a] 3.33 (m)
68.9
62.5
[a]
n-C7H15
3.51 (t, J ˆ 6.7)
[a]
Only characteristic resonances are listed.
Table 5. NMR characterization of tertiary amines of general formula R1CH2NR2 (chemical shifts of aromatic H and C are
2
omitted) recorded in CDCl3.
R1
R2
CH2NR2
1H
R1
13C 1H
63.1
R2
1H
2
13C
13C
4-HOC6H4 C5H10 3.31 (s)
1.27 1.39 (m, 2H, g-CH2), 24.3 (g-CH2),
1.42 1.56 (m, 4H, b-CH2), 25.4 (b-CH2),
2.31 (m, 4H, a-CH2)
54.1 (a-CH2)
4-MeOC6H4 C5H10 3.29 (s)
63.7 3.66 (s, 3H, CH3O)
55.5 (CH3
O)
1.25 1.37 (m, 2H, g-CH2), 24.9 (g-CH2),
1.39 1.52 (m, 4H, b-CH2), 26.4 (b-CH2),
2.24 (m, 4H, a-CH2)
54.8 (a-CH2)
Ph
C5H10 3.37 (s)
C5H10 3.34 (s)
64.4
63.5
1.27 1.40 (m, 2H, g-CH2), 24.9 (g-CH2),
1.42 1.57 (m, 4H, b-CH2), 26.5 (b-CH2),
2.27 (m, 4H, a-CH2)
55.0 (a-CH2)
4-ClC6H4
1.27 1.40 (m, 2H, g-CH2), 24.8 (g-CH2),
1.43 1.57 (m, 4H, b-CH2), 26.4 (g-CH2),
2.27 (m, 4H, a-CH2)
55.0 (a-CH2),
Ph(Me)CH C5H10 2.32 (d,
67.4 1.18 (d, 1H, J ˆ 7.0), 2.85 20.3 (CH3), 1.25 1.36 (m, 2H, g-CH2), 24.8 (g-CH2),
J ˆ 7.0)
(6 lines, 1H, J ˆ 7.0, CH) 37.8 (CH)
1.37 1.52 (m, 4H, b-CH2), 26.3 (b-CH2),
2.12 2.39 (m, 4H, a-CH2) 55.2 (CH2N)
Et(Me)CH C5H10 2.66
66.9 0.80 (t, 6H, J ˆ 7.2, CH3), 11.9 (CH3), 1.27 1.40 (m, 2H, g-CH2), 25.1 (g-CH2),
0.92 1.18 18.4 (CH3), 1.41 1.57 (m, 4H, b-CH2), 26.5 (b-CH2),
(m, 1H, CH), 1.93 (dd, 28.4 (CH2), 2.21 (m, 4H, a-CH2) 55.6 (a-CH2)
2.79 (m)
1H, J ˆ 7.7 and 12.1,
32.5 (CH)
MeCHH),
2.03 (dd, 1H, J ˆ 6.7 and
12.1, MeCHH)
n-C7H15
C5H10 2.19 (t,
59.9 0.80 (t, 3H, J ˆ 6.8, CH3), 14.3 (CH3), 1.30 1.46 (m, 2H, g-CH2), 24.7 (g-CH2),
J ˆ 7.8)
1.13 1.27
22.9 (CH2), 1.46 1.51 (m, 4H, b-CH2), 26.2 (b-CH2),
(m, 10H, 5CH2), 1.30
1.46 (m, 2H, CH2),
27.2 (CH2), 2.27 (m, 4H, a-CH2)
28.0 (CH2),
54.9 (a-CH2)
29.5 (CH2),
29.8 (CH2),
32.1 (CH2)
Ph
C4H8 3.52 (s)
61.3
1.62 1.78 (m, 4H, b-CH2), 23.9 (b-CH2),
2.34 2.51 (m, 4H, a-CH2) 54.7 (a-CH2)
Ph
Ph
Me
Et
3.31 (s)
3.46 (s)
64.7
57.5
2.09 (s, 6H, CH3)
45.6 (CH3)
0.94 (t, 6H, J ˆ 7.1, CH3); 11.53 (CH3),
2.40 (q, 4H, J ˆ 7.1, CH2); 46.7 (CH2)
Adv. Synth. Catal. 2002, 344, 200 208
207

Vitali I. Tararov, et al.
FULL PAPERS
change of the Pam/Pal ratio. When 4-ClC6H4CHO and
4-NO2C6H4CHO were applied as substrate duringtheir
dissolution in MeOH the formation of correspondingdime-
thoxy acetals was observed. Storage of these solutions for 10
min increased the yields of acetals.) The same approach
(prehydrogenation in 3 mL of MeOH) was used when the
mixture of an aldehyde and an amine in 7 mL of MeOH was
pre-equilibrated for 30 min.
The consumption of H2 was monitored as a decrease of
pressure by means of a pressure detector. When the consump-
tion of hydrogen ceased the autoclave was opened, the solution
evaporated in vacuum and the residue analyzed by integration
of 1H NMR signals. Additionally, the structure of the products
was confirmed by 13C NMR spectroscopy. NMR spectral data
for produced alcohols and amines are given in Tables 4 and 5.
rardo, A. G. Giumanini, P. Strazzolini, M. Poiana,
Synthesis 1993, 121; Zn(BH4)2-ZnCl2: S. Bhattacharyya,
A. Chatterjee, S. K. Duttachowdhury, J. Chem. Soc.
Perkin Trans. 1 1994, 1 2; NaBH(OAc)3: see Ref.[1]; Et3
¬
SiH-CF3COOH: D. Dube, A. A. Scholte, Tetrahedron
Lett. 1999, 40, 2295 2298; NaBH4-NiCl2: I. Saxena, R.
Borah, J. C. Sharma, J. Chem. Soc. Perkin Trans. 1 2000,
503 504; TiCl4/(dialkylamino)dimethylsilane: K. Miura,
K. Ootsuka, S. Suda, H. Nishikori, A. Hosomi, Synlett
2001, 1617 1619. Chemicals for one-pot reductive
amination (when intermediates are preformed, but not
isolated); NaBH4-Mg(ClO4)2: J. Brussee, R. A. T. M.
van Benthem, C. G. Krusse, A. van der Gen, Tetrahedron
Asymmetry, 1990, 1, 163 166; NaBH3CN-Ti(OPr-i)4:
R. J. Mattson, K. M. Pham, D. J. Leuck, K. A.Cowen, J.
Org. Chem. 1990, 55, 2552 2554.
¬
[4] L. Marko, J. Bakos, J. Organomet. Chem. 1974, 81, 411
Acknowledgements
414.
[5] H.-U. Blaser, H.-P. Buser, H.-P. Jalett, B. Pugin, F.
Spindler, Synlett 1999, 867 868.
[6] R. Margalef-Catala, C. Claver, P. Salagre, E. Fernandez,
Tetrahedron Lett. 2000, 41, 6583 6588.
[7] V. I. Tararov, R. Kadyrov, T. H. Riermeier, A. Bˆrner,
Chem. Commun. 2000, 1867 1868.
Financial support from Degussa AG, Project House Catalysis
(Frankfurt, Germany) and the Fonds der Chemischen Industrie
is gratefully acknowledged. We thank Mrs. Z. Kadyrova for
skilled technical assistance.
¡
[8] V. I. Tararov, R. Kadyrov, T. H. Riermeier, J. Holz, A.
Bˆrner, Tetrahedron Asymmetry 1999, 10, 4009 4015.
[9] V. I. Tararov, R. Kadyrov, T. H. Riermeier, J. Holz, A.
Bˆrner, Tetrahedron Lett. 2000, 41, 2351 2355.
[10] a) A. Pohland, H. R. Sullivan, R. E. McMahon, J. Am.
Chem. Soc. 1957, 79, 1442-1444; b) N. Sakura, K. Ito, M.
Sekiya, Chem. Pharm. Bull. 1972, 20, 1156 1163.
[11] A. Dornow, H. Thies, Liebigs Ann. Chem. 1953, 581,
219 224.
[12] T. Stewart, C. R. Hauser, J. Am. Chem. Soc. 1955, 77,
1098 1103.
[13] E. Staple, E. C. Wagner, J. Org. Chem. 1949, 14, 559
578.
[14] L. Forlani, E. Marianucci P. E. Todesco, J. Chem. Res. (S)
1984, 126 127.
[15] A. Bˆrner, D. Heller, Tetrahedron Lett. 2001, 42, 223
225; D. Heller, J. Holz, S. Borns, A. Spannenberg, R.
Kempe, U. Schmidt, A. Bˆrner, Tetrahedron Asymmetry
1997, 8, 213 222.
References
[1] A. F. Abdel-Magid, K. G. Carson, B. D. Harris, C. A.
Maryanoff, R. D. Shah, J. Org. Chem. 1996, 61, 3849
3862.
[2] J. J. Birtill, M. Chamberlain, J. Hall, R. Wilson, I.
Costello, in Catalysis of Organic Reactions, (Ed.: F. E.
Herkes), Dekker, New York, 1998, pp. 255 271.
[3] Reagents for direct reductive amination; NaBH3CN:
R. F. Borch, M. D. Bernstein, H. D. Durst, J. Am. Chem.
Soc. 1971, 93, 2897 2904; BH3-Py: A. Pelter, R. M.
Rosser, S. Mills, J. Chem. Soc. Perkin Trans. 1 1984, 717
720; NaBH3CN-ZnCl2: S. Kim, C. H. Oh, J. S. Ko, K. H.
Ahn, Y. J. Kim, J. Org. Chem. 1985, 50, 1927 1932; Zn-
AcOH: I. V. Micovic, M. D. Ivanovic, D. M. Piatak, V. D.
Bojic, Synthesis 1991, 1043 1045; borohydride exchange
resin: N. M. Yoon, E. G. Kim, H. S. Son, J. Choi, Synth.
Commun. 1993, 23, 1595 1599; NaBH4-H2SO4: G. Ve-
208
Adv. Synth. Catal. 2002, 344, 200 208