Propargylamine Synthesis by a Two-Step Flow Process
FULL PAPER
To investigate the performance of the process, the reac-
tion sequence was carried out under different regimes:
A) the catalytic potentials of PBCR1 and PBCR2 for the
three-component reaction were investigated independently
of one another; B) solutions of all three building blocks
were mixed and fed sequentially through PBCR1 and
PBCR2 at different temperatures; C) the aldehyde and
amine building block solutions were mixed and fed through
PBCR1, and then the alkyne solution was added to the in-
termediate product stream before implementation of
PBCR2. These set-up modes are illustrated in Scheme 2. In
the case of regime C, each solution was fed at the same flow
rate. However, the flow rate for PBCR2 was twice that for
PBCR1. In other words, the residence time in PBCR2 was
half of that in PBCR1.
small amounts at about 1208C (Table 1, entries 5 and 6).
Hence, the catalytic capabilities of both catalysts were low
or absent when they were used independently. Reaction
regime B involved subjecting the premixed building blocks
to both catalysts sequentially, which enabled the formation
of intermediate 3 under MM K-10 catalysis in the presence
of phenylacetylene (4). When the intermediate reaction mix-
ture passed the second PBCR2, formation of the desired
propargylamine (5) under Au-NP@Al2O3 catalysis was an-
ticipated. However, the observed process performance was
low, with only 13% conversion to 5 (Table 1, entry 7). When
the reaction was investigated under flow regime C, the de-
sired product was obtained in a fair yield of about 40%
(Table 1, entry 8). Hence, regime C clearly gave the best re-
action performance.
The product profiles generated under the different reac-
tion regimes were analyzed by LC-MS and studied for par-
ticular combinations of building blocks. The regime giving
the best reaction performance was optimized in greater
detail. Therefore, different reaction temperatures, residence
times (flow rates), stoichiometries, and solvents were investi-
gated. The optimized conditions were applied to produce
about 25 representative propargylamines using different
building blocks.
The conditions for the different reaction regimes were op-
timized using benzaldehyde (1a), piperidine (2a), and phe-
nylacetylene (4). For reaction regimes A–C, 0.5 molLÀ1
ethanolic solutions of these compounds were applied at
50 mLminÀ1. The obtained results are summarized in
Table 1. The thermal reaction under regime A without any
We further optimized the reaction regime C by tuning the
reaction conditions for both PBCRs separately. First, the ini-
tial condensation reaction was investigated. To this end, a
premixed ethanolic solution of benzaldehyde (1a) and piper-
idine (2a) was fed through the first PBCR1 and the inter-
mediate reaction product 3 was analyzed by GC-MS. Reac-
tant concentrations of 0.5 molLÀ1 were used. The reaction
profile of the intermediate aldimine (3) was analyzed with
respect to the flow rate and the temperature. It was found
that near-quantitative conversion was achieved at ambient
temperature (about 258C) and 60 min residence time
(50 mLminÀ1) (Figure 1a). Elevated temperatures (40–
1008C) led to significantly decreased yields. Hence, PBCR1
was used at ambient temperature.
The second reaction step catalyzed by Au-NP@Al2O3 was
investigated with respect to the reaction temperature, flow
rate, and catalyst loading. Motivated by literature reports,
we carried out the reaction with 1.5 equivalents of phenyla-
cetylene (4) at 808C. The obtained conversions are shown in
Figure 1b. It was found that complete conversion could not
be achieved within 30 min residence time (PBCR2) at a cat-
alyst loading of about 1.25 wt.% Au. However, a loading of
about 2.5 wt.% Au led to full conversion of the aldehyde
(1a) within 30 min (Table 1, entry 9). Variation of the reac-
tion temperature led to lower performance at both loading
levels (Figure 2). That is to say, the best yield of the desired
propargylamine (5) was obtained at 808C. Finally, we sought
to ascertain whether a 1.5-fold stoichiometric excess of phe-
nylacetylene (4), as mentioned in many literature reports,
was in fact necessary. Reduction of the 1.5 equivalent excess
to a stoichiometric ratio proved to have no influence; con-
versions of about 97% were observed in both cases. Differ-
ent solvents (methanol, acetonitrile, toluene, water) were
also tested for the overall process, but ethanol was found to
be the most effective. In previous studies, the lifetime per-
formance of the Au-NP@Al2O3 catalyst was investigated
over 48 h. A marginal decrease in the catalytic performance
was observed after 12 h of continuous use.[22] Hence, we re-
newed the catalyst in PBCR2 after 12 h of operation time.
To identify any leaching of the Au catalyst, a batch of about
1 mmol of propargylamine (5a) was synthesized using fresh-
ly prepared catalyst. The crude product was concentrated in
Table 1. Reaction performance in dependence on the reaction regime as
shown in Scheme 2 at constant flow rate (0.5 mol LÀ1, ethanolic solution,
50 mLminÀ1 per feed[a], 1.25 wt.% Au-NP@Al2O3).
Entry
Reaction
regime
Catalyst
Temp.
[8C]
Conv.
[%][b]
AHCTUNGTRENNUNG
1
2
3
4
5
6
80
150
25
80
25
80
25
80
25
80
25
80
0
0
0
0
0
5
A
A
A
B
C
C
no catalyst
MM K-10
1.25% Au-NP@Al2O3
MM K-10
1.25% Au-NP@Al2O3
MM K-10
1.25% Au-NP@Al2O3
MM K-10
2.5% Au-NP@Al2O3
7
8
9
13
44
97
[a] Fluidic residence times: PBCR1 60 min; PBCR2 30 min. [b] Calculat-
ed for benzaldehyde.
catalytic support was attempted at up to 1508C (Table 1, en-
tries 1 and 2), but the desired propargylamine (5a) was not
observed. The use of Montmorillonite K-10 alone under
regime A did not yield the desired product either (Table 1,
entries 3 and 4). The use of Au-NP@Al2O3 alone (at a load-
ing of 1.25wt.% Au) yielded the desired product in very
Chem. Eur. J. 2011, 17, 3005 – 3010
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3007