Poisoning and Reuse of Supported Precious Metal Catalysts in the Hydrogenation of N-Heterocycles Part?I: Ruthenium-Catalysed Hydrogenation of 1-Methylpyrrole

Article abstract of DOI:10.1007/s10562-018-2406-4

Abstract: Poisoning phenomena of heterogeneous, supported precious metal catalysts caused by nitrogen were investigated in the liquid-phase hydrogenation of 1-methylpyrrole (MP) to 1-methylpyrrolidine (MPD) over ruthenium on carbon, in non-acidic medium (methanol), at 10?bar and 25–60?°C. Reusing a spent, unregenerated 5% Ru/C catalyst, it was found that the activity of catalyst and the conversion of model substrate were strongly dependent on the amount of catalyst and the number of recycling, respectively. During the first reuse of this ruthenium catalyst, surprisingly, it showed high activity already at room temperature contrary to the fresh catalyst which worked at only 60?°C. This unexpected catalytic behaviour was studied by XRD and XPS methods which revealed the existence of a fine RuO₂ layer on the surface of the catalytic metal in the fresh catalyst. Graphical Abstract: [Figure not available: see fulltext.].

Full text of DOI:10.1007/s10562-018-2406-4

Catalysis Letters
https://doi.org/10.1007/s10562-018-2406-4
Poisoning and Reuse of Supported Precious Metal Catalysts
in the Hydrogenation of N-Heterocycles Part I: Ruthenium-Catalysed
Hydrogenation of 1-Methylpyrrole
László Hegedűs1 · Kristóf Szőke‑Molnár · István E. Sajó · Dávid Ferenc Srankó · Zoltán Schay
1
2
3
3
Received: 8 January 2018 / Accepted: 3 May 2018
©
Springer Science+Business Media, LLC, part of Springer Nature 2018
Abstract
Poisoning phenomena of heterogeneous, supported precious metal catalysts caused by nitrogen were investigated in the
liquid-phase hydrogenation of 1-methylpyrrole (MP) to 1-methylpyrrolidine (MPD) over ruthenium on carbon, in non-acidic
medium (methanol), at 10 bar and 25–60 °C. Reusing a spent, unregenerated 5% Ru/C catalyst, it was found that the activ-
ity of catalyst and the conversion of model substrate were strongly dependent on the amount of catalyst and the number of
recycling, respectively. During the first reuse of this ruthenium catalyst, surprisingly, it showed high activity already at room
temperature contrary to the fresh catalyst which worked at only 60 °C. This unexpected catalytic behaviour was studied by
XRD and XPS methods which revealed the existence of a fine RuO layer on the surface of the catalytic metal in the fresh
2
catalyst.
Graphical Abstract
H2
N
methanol
N
high activity
CH3
CH3
1
-methylpyrrole
1-methylpyrrolidine
5
% Ru/C
after recycling
reusing the catalyst
without regeneration
Keywords Hydrogenation · Poisoning · Reusing · Ruthenium · Pyrroles · Spent catalyst
1
Introduction
Dedicated to the memory of Dr. Tibor Máthé, an excellent
researcher and inventor, who passed away 17 years ago.
Among the different types of catalyst deactivation, such
as poisoning, fouling, thermal degradation, vapour–solid
and/or solid–solid reactions, attrition/crushing, poisoning
is a chemical interaction where the strong chemisorption
of species (reactants, products or impurities) on catalytic
sites blocks the surface chemical reaction [1–4]. As it has
long been known [5–11], compounds containing basic
nitrogen have inhibiting effects on the catalysts used for
Electronic supplementary material The online version of this
article (https://doi.org/10.1007/s10562-018-2406-4) contains
supplementary material, which is available to authorized users.
*
László Hegedűs
lhegedus@mail.bme.hu
Extended author information available on the last page of the article
Vol.:(0
1
123456789)
3

L. Hegedűs et al.
O
H
2
, 5% Ru/C
methanol
N
N
N
OH
N
OCH
3
CH3
MP
CH3
MPD
CH3
CH3
MPOL
MMPA
Scheme 1 Ruthenium-catalysed hydrogenation of 1-methylpyrrole
MP) to 1-methylpyrrolidine (MPD)
(
N
CH3
MP
at relatively high pressures (25–35 bar) and temperatures
130–150 °C). Thus, 2,5-dimethylpyrrole was converted
(
to cis-2,5-dimethylpyrrolidine over 5% Ru/Al O [21] or
2
3
Fig. 1 Structures of pyrrole derivatives hydrogenated previously
unsupported RuO [22], in water, at 130 °C and 28–35 bar
2
with 85–98% conversion and high cis-selectivity (>90%).
hydrogenation due to their unshared pair of electrons. This
effect can be neutralized by conversion of these reactants to
a form in which the nitrogen atom is shielded, for instance,
by adding protic acids [7]. However, this method cannot be
applied if a substrate is sensitive to acids, i.e. a side-reaction
Most recently, 5% Ru/TiO or 5% Ru/TiO catalysts have
2
2−x
been applied in the hydrogenation of MP to MPD in tet-
rahydrofuran, at 90–100 °C and 30 bar obtaining 80–95%
conversion [23].
However, handling the spent catalysts formed during the
hydrogenations is also an important technological aspect of
the heterogeneous catalytic processes in the chemical indus-
try [24–32]. Both precious [25] and base metals [26] on
different supports or in unsupported forms are usually fully
regenerated before their reusing, but the applied regeneration
methods (e.g. incineration or pyrometallurgical processes)
are typically energy-intensive and expensive. In pharmaceu-
tical industry, where carbon supported precious metal cata-
lysts (e.g. Pd/C, Ru/C) are mostly used, the usual method is
applying the completely regenerated spent catalysts due to
the very strict rules of quality assurance (e.g. GMP—Good
Manufacturing Practices [33]). Since their reusing with-
out regeneration has not solved up to now, an easy and less
expensive catalyst recycling process for the heterogenous
catalytic hydrogenations could afford to produce pharma-
ceuticals more economical.
(
[
e.g. polymerisation) takes place in the presence of them
12].
Previously we reported the liquid-phase heterogeneous
catalytic hydrogenations of several pyrrole derivatives, such
as 1-methyl-2-pyrroleethanol (MPOL), methyl 1-methyl-
2
-pyrroleacetate (MMPA) and 1-methylpyrrole (MP)
(
(
Fig. 1), over different, supported precious metal catalysts
Pd/C, Ru/C, Rh/C, Rh/Al O , Pt/C, Ir/C), in non-acidic
2
3
medium [13–16]. The corresponding pyrrolidines, which
are important and valuable pharmaceutical intermediates
[
17–19], were prepared in good yields (80–90%). In these
reductions the light platinum metals (Rh, Ru, Pd) proved
to be the most active catalysts. Although these hydrogena-
tions took place relatively easily under mild reaction condi-
tions (25–80 °C, 6 bar) with complete conversion and high
selectivity, poisoning of the catalysts was observed below
some catalyst/substrate ratios. The values of these ratio lim-
its ranged from 0.03 to 0.2, and they were dependent on
the substrates, catalytic metals and solvents. Furthermore,
the poison sensitivity of these precious metals referring to
nitrogen was also determined which decreased in the follow-
ing sequence: Pd>Ru>>Rh. This order was attributed to
electronic factors [20].
In this paper the effect of reusing the spent, unregener-
ated, carbon supported ruthenium (5% Ru/C) on its activity
and conversion of MP is discussed. An unexpected behav-
iour of this catalyst was examined by X-ray powder diffrac-
tion (XRD) and X-ray photoelectron spectroscopy (XPS)
surface analytical methods.
In this work the poisoning phenomena of heterogene-
ous, supported precious metal catalysts caused by nitrogen
and their reusing without regeneration were investigated in
detail. Based on our previous experience [16], the liquid-
phase hydrogenation of 1-methylpyrrole (MP) to 1-meth-
ylpyrrolidine (MPD) over ruthenium on carbon, in non-
acidic medium (methanol) was chosen as a model reaction
2 Experimental
2.1 Materials
1-Methylpyrrole (99%) was supplied by Merck-Schuchardt
(Hochenbrunn, Germany), while methanol (p. a.) was pur-
chased from Merck (Darmstadt, Germany).
(
Scheme 1).
Apart from our previous investigations [14–16, 20],
The 5% Ru/C catalyst was received from Aldrich (Mil-
waukee, USA), whilst anhydrous ruthenium(IV) oxide
(99.9%) was supplied by Alfa Aesar (Karlsruhe, Germany).
ruthenium was very rarely used in the heterogeneous cata-
lytic hydrogenation of pyrroles [21–23], moreover it acted
1
3

Poisoning and Reuse of Supported Precious Metal Catalysts in the Hydrogenation of…
2
.2 Hydrogenations
proportional counter. Crystallite size of ruthenium was
calculated from reflection line broadening using the Scher-
rer equation:
The hydrogenation reactions were carried out in a 250 cm3
stainless steel autoclave (Technoclave, Budapest, Hungary)
equipped with a magnetic stirrer (stirring speed: 1100 rpm),
and electric heating system, at 10 bar and 25–60 °C.
Typically, the reactor containing MP (2.0 g), 5% Ru/C
1
80 0.9ꢂ
Δ (2ꢀ)[deg] =
+
,
Δ(2ꢀ) = Δ + Δ (2ꢀ)
0 +
ꢁ
d cos ꢀ
(
1)
where d is the diameter of the crystallite, ꢀ is the X-ray
wavelength, ꢀ is the Bragg angle (in degrees), Δ(2ꢀ) denotes
the full width (FWHM) of the diffraction peak associated
3
catalyst (0.4 g) and methanol (50 cm ) was flushed with
nitrogen and hydrogen, then charged with hydrogen to the
specified pressure and heated up to the given temperature, if
necessary. After finishing the hydrogen uptake, the catalyst
was filtered off and a sample was taken from the filtrate.
It was analysed by GC–MS and the conversion value was
calculated from the hydrogen pressure drop and the gas-
with the scattering angle 2ꢀ, and Δ =0.11 deg accounts for
0
instrumental broadening.
Surface compositions of the fresh and spent catalysts, as
well as pure RuO were determined by X-ray photoelectron
2
spectroscopy (XPS) performed by a KRATOS XSAM 800
XPS machine. Mg Kα X-ray line, 40 eV pass energy (energy
steps 0.2 eV) and FAT mode were applied for recording
the XPS lines of Ru 3p, Ru 3d, O 1s, N 1s and C 1s were
applying. For charge compensation the C 1s binding energy
at 284.8 eV was used as a reference. The surface concen-
trations of the elements were calculated from the integral
intensities of the XPS lines using sensitivity factors given
by the manufacturer.
chromatographic data. The initial rate (v ) was determined
0
from the conversion curves.
1
-Methylpyrrolidine (MPD), when the conversion was
complete, was prepared in the same way as described in [16].
The MS data of the starting material and the product are the
following: MP m/z (rel%) 81(100), 55(15), 53(26), 42(24),
3
3
9(23); MPD m/z (rel%) 85(55), 84(97), 57(90), 42(100),
2(8). These analytical results are in agreement with the
literary data [34].
Dispersion of the catalyst (D
= 0.43) was deter-
% Ru/C
5
mined by H -, O - and CO-chemisorption measurements
2
2
2.3 Catalyst Recycling
using an atmospheric flow system [35, 36]. Prior to the
first adsorption of O , the catalyst sample was treated with
2
After filtering the spent catalyst, it was washed with distilled
H in argon at 90 °C for 4 h. (Ru–H) was titrated with O
2
s
2
3
water (2×5 cm ) and collected carefully to store for the next
3
injections via a calibrated loop (0.1 cm each) at 25 °C.
reaction, in that wet form. No regenerative processes were
applied prior to its reusing.
Next (Ru–O) was titrated with H in the same way. CO-
s
2
chemisorption was also measured after treating in hydrogen
When an acidic treatment of the used catalyst was imple-
mented, the procedure was as follows. The spent catalyst
was filtered, and subsequently suspended in 5% acetic acid
(
80 °C, 4.5 h). The stoichiometry of the calculations was
the following:
3
3
(
2×5 cm ), then washed with distilled water (2×10 cm ) to
•
•
•
for titration with O2
Ru − H) + 0.75O = (Ru − O) + 0.5H O
pH 7. Finally, it was also collected carefully to store for the
next reaction, in that wet form.
(
s
2
s
2
(2)
(3)
for titration with H2
Ru − O) + 1.5H = (Ru − H) + H O
2.4 Catalyst Pretreatment
(
s
2
s
2
for CO adsorption
The prehydrogenation of 5% Ru/C catalyst was carried out
in the absence of substrate (MP), in methanol, at 10 bar and
2
5 or 60 °C for 30 or 90 min. After this pretreatment, the
(
Ru − H) + CO=(Ru − CO) + 0.5H
s
s
2
(4)
starting material was added to the reaction mixture at 25 °C,
and its hydrogenation was performed at this temperature.
2.6 Analysis
2
.5 Catalyst Characterisation
The components of reaction mixtures were analysed and
identified by GC–MS. The analyses were carried out with
a Finnigan Mat/Automass II GC/MS spectrometer using a
Zebron ZB-5 ms capillary column (30 m × 0.25 mm ID,
The phase composition of the fresh or spent catalysts
was determined by X-ray powder diffraction (XRD)
measurements. The XRD patterns were obtained with
a Philips PW 3710 diffractometer based PW 1050
Bragg–Brentano parafocusing goniometer using Cu Kα
radiation (λ = 0.15418 nm), graphite monochromator and
0
4
2
.25 µm film). The temperature program was the following:
−
1
5 °C (2 min) to 300 °C at 10 °C min , then to 350 °C at
−
1
5 °C min .
1
3

L. Hegedűs et al.
Table 1 Effect of reusing the catalyst (5% Ru/C) on the initial rate
3
Results and Discussion
(
(
v ) and the reaction time in the hydrogenation of 1-methylpyrrole
0
MP)
3.1 Special Behaviour of Ruthenium
−
1
min−1
)
Run Reusing the Reaction time for
v (nL H g
0
2
Ru
catalyst
complete conversion
Conversion of MP over 5% Ru/C, depending on the number
(
min)
−
1
of reusing the catalyst, at 0.2 g g catalyst/substrate ratio,
in methanol is shown in Fig. 2, while the initial rates (v )
1*
2
3
4
5
– (fresh)
1st
2nd
3rd
4th
30
30
45
45
50
5.53
5.37
4.99
4.71
4.43
0
and the reaction time in the hydrogenation of MP are sum-
marized in Table 1.
As seen, the fresh ruthenium on carbon proved to be
a very efficient catalyst for this reduction, because the
hydrogenation of MP was fast and complete after 30 min.
3
Conditions: 2.0 g substrate, 50 cm methanol, 0.4 g catalyst, 25 °C,
*25–60 °C, 10 bar
Whereas, no H -uptake was observed at 25 °C, therefore it
2
was necessary to increase the temperature to 60 °C, at least
(
run 1). Saturation of the pyrrole ring, in each reuse of the
the reaction can readily take place on the renewed and active
surface of catalyst. Furthermore, the formed metallic ruthe-
nium can remain stable for a longer time, therefore only a
slight decrease in its activity can be observed in this hydro-
genation (Table 1). To confirm our hypothesis, X-ray pow-
der diffraction (XRD) and X-ray photoelectron spectroscopy
catalyst, was achieved with 100% conversion, even if the
initial rates of the hydrogenation (v ) were slightly lower
0
−
1
−1
(
5.37–4.43 nL H g
min ) than that of the original reac-
2
Ru
−
1
−1
tion (5.53 nL H g
min ), i.e. the complete conversion
2
Ru
of MP required a hardly longer reaction time (45–50 min)
in the repeated hydrogenations (runs 2–5). Surprisingly, it
was found that the reduction of MP took place completely
already at room temperature, during the second use of 5%
Ru/C (run 2), after the same time (30 min). Similar behav-
iour was observed in all reusing the catalyst, i.e. complete
conversions were obtained over ruthenium on carbon, with-
out heating, already at 25 °C.
(
XPS) measurements were performed on both fresh and used
catalysts, which will be discussed in Sect. 3.3.
3.2 Effect of Catalyst Pretreatment
The influence of catalyst prehydrogenation and the number
of reuse on the conversion of MP over 5% ruthenium on
Presumably, this phenomenon was caused by a fine RuO
2
−
1
carbon, at 0.2 g g catalyst/substrate ratio, in methanol is
depicted in Fig. 3.
layer formed on the surface of the catalytically active ruthe-
nium particles, which can rapidly be transformed to pure
metal around at 60–80 °C [37, 38]. This layer can avoid the
process of catalytic hydrogenation, but after its reduction,
As seen, pretreatment of the fresh catalyst at room tem-
perature (10 bar H , 30 min) resulted in complete conver-
2
sion of MP already at 25 °C (run 1), but this required a
Fig. 3 Conversion of 1-methylpyrrole (MP) vs. time over 5% Ru/C
3
Fig. 2 Reusing 5% ruthenium on carbon in the hydrogenation of
prehydrogenated at 25 °C. Conditions: 2.0 g substrate, 50 cm meth-
3
3
1
-methylpyrrole (MP). Conditions: 2.0 g substrate, 50 cm methanol,
.4 g catalyst, 25 °C, *25–60 °C, 10 bar
anol, 0.4 g catalyst, 25 °C, 10 bar; pretreatment: 30 cm methanol,
0
25 °C, 10 bar, 30 min
1
3

Poisoning and Reuse of Supported Precious Metal Catalysts in the Hydrogenation of…
Table 2 Effect of reusing the prehydrogenated catalyst (5% Ru/C)
on the initial rates (v ) and the reaction time in the hydrogenation of
0
1
-methylpyrrole (MP)
−1
min−1
)
Run Reusing the Reaction time for
v (nL H g
Ru
0
2
catalyst
complete conversion
min)
(
1
2
3
4
5
– (fresh)
1st
2nd
3rd
4th
300
70
100
110
120
0.11 (0.55a)
5.03
4.43
3.87
3.32
3
Conditions: 2.0 g substrate, 50 cm methanol, 0.4 g catalyst, 25 °C,
3
1
0 bar; pretreatment: 30 cm methanol, 25 °C, 10 bar, 30 min
a
After a transient period (30 min)
significantly longer reaction time (300 min). Behaviour of
the recycled and prehydrogenated catalyst, however, was
practically the same in this series as it was found in the
previous one (Sect. 3.1), i.e. hydrogenation of the pyrrole
ring, in each reusing the catalyst, was obtained with 100%
conversion (runs 2–5). Although the reaction times were
longer (70–120 min) in these cases (Table 2), the initial rate
−
1
−1
values (v =5.03–3.32 nL H g
Ru
min ) were very simi-
0
2
lar to those ones achieved earlier (Table 1). Whereas, the
prehydrogenation of 5% Ru/C at room temperature was not
such efficient as using the untreated catalyst after the first
run at 60 °C.
As also seen in Fig. 3, there was a transient period
−
1
−1
(
v =0.11 nL H g
Ru
min ) at the beginning of the reac-
0
2
Fig. 4 Conversion of 1-methylpyrrole (MP) vs. time over 5% Ru/C
prehydrogenated at 25 °C and 90 min (a) or at 60 °C and 30 min (b).
tion (0–30 min) which was followed by an accelerated phase
−
1
−1
(
30–180 min, v =0.55 nL H g
min ). This near linear
3
0
2
Ru
Conditions: 2.0 g substrate, 50 cm methanol, 0.4 g catalyst, 25 °C,
3
stage can refer to the slow reduction of the surface oxide
layer to metallic Ru which takes place in parallel with the
hydrogenation of MP. Thus, the effects of time and tem-
perature of prehydrogenation on the conversion and initial
rate were also investigated (Figs. 4, 5). Applying a longer
10 bar; pretreatment: 30 cm methanol, 10 bar
therefore prehydrogenation of the catalyst was also carried
out under these conditions (60 °C, 10 bar H , 30 min), in
2
catalyst pretreatment time (90 min) at 25 °C and 10 bar H ,
the absence of substrate (MP). After this pretreatment, the
2
no significant decrease in the reaction time for complete
conversion of MP (300 → 240 min) was observed during
the first run, but no transient phase was detected (Fig. 4a).
Furthermore, the sum of the reaction and prehydrogenation
times (330 min) was the same after the 30 and 90 min pre-
treatment of catalyst at 25 °C, respectively (Fig. 5a). There
was no difference between the reaction times for complete
conversion of MP (70 min) during the first reuse of Ru/C
fresh ruthenium on carbon showed a significantly higher
−
1
−1
activity (v = 6.64 nL H g
Ru
min ) in the first run at
0
2
25 °C (Fig. 5b) and, among the reactions performed over
the prehydrogenated catalysts, an appreciable shorter time
(130 min) was enough to complete the hydrogenation of MP
(Fig. 4b). Moreover, the reused catalyst provided an out-
−
1
−1
standingly high initial rate (v = 8.25 nL H g
Ru
min ),
0
2
also at 25 °C, which was accompanied by the shortest reac-
tion time (25 min) observed in all hydrogenations (Fig. 5a,
b). This indicates that the catalyst prehydrogenation at 60 °C
resulted in significant improvement in the activity of cata-
lyst during its reuse, presumably due to metallic ruthenium
formed in situ in the course of this pretreatment.
(
Fig. 5a), which was also confirmed by the practically same
−
1
−1
initial rate values (v = 5.03 and 5.25 nL H g
min )
0
2
Ru
(
Fig. 5b). This indicates that the catalyst prehydrogenation at
2
5 °C resulted in no significant improvement in the activity
of catalyst during its reuse.
As mentioned afore, the inhibitory effect of the pre-
sumed surface oxide layer could be eliminated at 60 °C,
According to these results, the prehydrogenation of
ruthenium on carbon resulted in high catalytic activity, but
1
3

L. Hegedűs et al.
activated carbon (PDF #41-1487). Although the reflec-
tions of metallic ruthenium (PDF #06-0663) at 2Θ=42.1°
and 44.0° appear in the fresh catalyst, they are very wide
and merged. Using the Scherrer Eq. (1), the Ru particle
size (d) was calculated to be 4.1 nm. However, no char-
acteristic peaks of RuO (PDF #43-1027) are recognized
x
at 2Θ=28.0°, 35.0° and 54.1°, i.e. if there is ruthenium in
oxide form in this catalyst, it should have a very fine texture.
There is no significant difference between the XRD scans
of the used (5×) catalysts (prehydrogenated or without pre-
treatment) in wet form, because their variations from the
fresh catalyst are due to the amorphous scattering of water.
Furthermore, no significant dissimilarity can be observed in
the Ru-containing phases.
These results suggest that the existence of the presumed
fine RuO layer cannot be proved by using this analytical
x
method.
3
.3.2 XPS
Figure 7a–e exhibit the XP spectra of anhydrous
ruthenium(IV)-oxide (as a reference) and the 5% Ru/C cata-
lyst in different forms: fresh, prehydrogenated at 25 °C and
used (2×), prehydrogenated at 60 °C and used (2×), as well
as prehydrogenated at 60 °C without substrate.
In the high-resolution XP spectrum of pure RuO
2
(
Fig. 7a) the peak at a binding energy of 280.3 eV is assigned
to the Ru 3d spin–orbit coupling which refers to the higher
5/2
oxidation states of ruthenium (RuO ) [39–42]. The appear-
2
ance of the characteristic shake-up satellite features and the
smaller full width at half maximum (FWHM) value of the
Fig. 5 Effects of time and temperature of prehydrogenation on the
reaction time for complete conversion (a) and the initial rate (v ) (b)
3d peak (1.37 eV, Table 3) indicate the presence of its
0
5/2
in the hydrogenation of MP over prehydrogenated 5% Ru/C. Condi-
anhydrous form [43].
3
tions: 2.0 g substrate, 50 cm methanol, 0.4 g catalyst, 25 °C, 10 bar;
As seen in Fig. 7b–d, the Ru 3d peaks are highly over-
lapped with the C 1s ones in the 5% Ru/C catalysts. The
shoulders observed at around 281.6 eV binding energies
3
pretreatment: 30 cm methanol, 10 bar
it strongly depended on the temperature of this pretreatment.
The best results were achieved after prehydrogenation of the
catalyst at 60 °C and 30 min, but using these conditions is
not practical on an industrial scale.
indicate the oxidation state of 4+ for ruthenium in RuO
2
phase before and after the catalytic cycle. The other pairs
of the spin–orbit splitting (SOS), the Ru 3d at 285.8 eV,
3
/2
and the satellites are also in overlap with the C 1s peak.
However, the SOS of 4.17 eV with an area ratio of 0.67 were
used for fittings. The Ru-content is ~4% (atomic conc.) in
the Ru/C catalysts (Table 3), except in the sample prehydro-
genated at 60 °C and used (2×). In this case the broadened
shoulder at lower binding energies can indicate some differ-
ences in the chemical composition, e.g. the altered ratio of
hydrated/anhydrous Ru-oxides. The XP spectrum of sam-
ple 5% Ru/C prehydrogenated at 60 °C without substrate
3.3 X‑ray Investigations
To find an explanation to this unexpected catalytic behaviour
of ruthenium, the fresh and used catalysts were examined by
surface analytical methods (XRD and XPS).
3.3.1 XRD
(
Fig. 7e) was recorded after its drying under reductive con-
Figure 6 shows the XRD patterns of the fresh and used (5×)
ditions (100 °C, H ) in the atmospheric chamber of the XPS
2
5
% Ru/C catalysts, as well as that of the prehydrogenated
device, namely the sample was placed into the instrument
one after also using (5×). In all cases, the most intensive
peak can be observed at 2Θ = 24.5° which refers to the
in its wet form. The lower binding energy of Ru 3d peak
5
/2
at 280.5 eV can refer either to the oxide or to metallic state
1
3

Poisoning and Reuse of Supported Precious Metal Catalysts in the Hydrogenation of…
Fig. 6 XRD patterns of 5% Ru/C catalyst in fresh and used (5×) forms
of ruthenium. Due to the position of Ru 3p3/2 at 461.6 eV,
the fresh 5% Ru/C at 25–60 °C (run 1), in turn, this required
most probably, it is the mixture of these two states with the
an appreciably longer reaction time (360 min). Although
the hydrogenation of MP started already at 25 °C in the
second run, the reaction stopped at 97% conversion. Drastic
decreases were observed in the conversion values during
the next reusing the catalyst (runs 3–4), namely they were
only 60% and 5%, respectively, even after 420 min. Lastly,
in the fifth run, the 5% ruthenium on carbon was completely
deactivated, i.e. no product (MPD) was detected in the reac-
tion mixture, at all.
presence of anhydrous and hydrated form of RuO (Figs. S1
2
and S2, Table S1).
Accordingly, these investigations can reveal the existence
of a fine ruthenium(IV) oxide layer on the surface of ruthe-
nium before and after catalysis and verify the favourable
effect of catalyst prehydrogenation on the catalytic activity.
3.4 Influence of Amount of Catalyst
To compare the effects of amount and reuse of 5% Ru/C
Conversion of MP over 5% Ru/C at lower catalyst/substrate
catalyst on the course of the hydrogenation of MP, the initial
−1
ratios (0.1 and 0.05 g g ), depending on the number of reus-
ing the catalyst, in methanol is shown in Fig. 8.
rates (v ) and the conversions at 60 min are also depicted in
0
Fig. 9.
−
1
−1
As seen, at 0.1 g g catalyst/substrate ratio, the satura-
tion of pyrrole ring of MP was also complete until the sec-
ond reuse of 5% Ru/C (runs 1–3, Fig. 8a), but this required
longer reaction times (75, 100 or 160 min) than applying a
As seen in Fig. 9a, at 0.2 and 0.1 g g catalyst/sub-
strate ratios, a similar decrease of v values was observed
0
by increasing the number of reusing the catalyst (5.53 →
−
1
−1
−1
−1
4.43 nL H g
min and 5.09 → 3.32 nL H g
min ,
2
Ru
2
Ru
−
1
−1
0
.2 g g catalyst/substrate ratio (30–45 min, Fig. 2). High
respectively), while at the 0.05 g g one a significant
conversions were also obtained (99 and 98%) during the
next hydrogenations of MP (runs 4–5), but they were not
complete even after a much longer reaction time (240 or
activity diminution was experienced (5.01 → 0 nL H
2
−1
−1
g
min ). Comparing the conversion at 60 min (Fig. 9b),
Ru
−
1
more remarkable differences can be noticed. At 0.2 g g
3
60 min). At this stage, presumably, the extent of catalyst
catalyst/substrate ratio complete conversion of MP was
obtained during all runs, while in case of a lesser amount of
poisoning reached a threshold limit which caused the incom-
plete conversion of MP.
−
1
catalyst (0.1 g g ratio) a significant conversion decrease
Further decreasing the amount of ruthenium on carbon
could be observed after the fifth cycle (97 → 62%). Apply-
−
1
−1
to 0.05 g g catalyst/substrate ratio resulted in significant
difference in the conversion of MP during the reuse of the
catalyst (Fig. 8b). The reduction of MP was complete over
ing even less catalyst (0.05 g g catalyst/substrate ratio),
the conversion drop was already occurred in the first run
(66%), and this became much more appreciably during the
1
3

L. Hegedűs et al.
Fig. 7 High-resolution and deconvolution XP spectra of anhydrous
60 °C and used (2×) (d), as well as prehydrogenated at 60 °C without
substrate (e)
RuO as a reference (a) and 5% Ru/C catalyst in different forms: fresh
2
(
b), prehydrogenated at 25 °C and used (2×) (c), prehydrogenated at
last reuse of the catalyst (0%). These phenomena refer to the
strong poisoning of ruthenium by nitrogen which becomes
more significant at lower catalyst/substrate ratios (0.1 and
amount of ruthenium on carbon, it could be applied eco-
nomically on an industrial scale considering the low price
of Ru (65 USD/tr. oz.).
−
1
0
5
.05 g g ) and room temperature, after several reusing the
% Ru/C.
3.5 Acidic Treatment of Catalyst
According to these results, it is favourable to apply a
−
1
relatively high catalyst/substrate ratio (0.2 g g ) to achieve
complete conversion of MP even after more catalytic runs
and without regeneration of the catalyst. Despite the higher
Since the product (1-methylpyrrolidine) has a strong basic
character due to its tertiary nitrogen, an acidic treatment
can be an alternative method for avoiding deactivation of
1
3

Poisoning and Reuse of Supported Precious Metal Catalysts in the Hydrogenation of…
Table 3 Ru 3d5 binding
/2
Samples
Position of
the peaks
FWHM (eV) Atomic concentra-
tions in Ru/C cata-
lysts (%)
energies, FWHM and Ru/C
surface atomic concentrations
of the fresh and spent catalysts
with the fitted line shape of AS
Ru 3d and
5
/2
Ru 3p (eV)
3
/2
(
30; 0.6)
RuO (anhydrous)
280.3 462.6 1.37
281.9 463.5 2.77
–
2
5
5
5
5
% Ru/C (fresh)
4.16
4.01
2.67
4.12
% Ru/C [prehydrogenated at 25 °C and used (2×)] 281.7 463.8 2.93
% Ru/C [prehydrogenated at 60 °C and used (2×)] 281.3 463.6 3.11
% Ru/C (prehydrogenated at 60 °C)
280.5 461.6 2.37
FWHM full width at half maximum
Fig. 9 Effects of amount and reuse of the catalyst on the initial rate
(
v ) (a) and the conversion at 60 min (b) in the hydrogenation of
0
Fig. 8 Conversion of 1-methylpyrrole (MP) vs. time over 5% Ru/C,
1
5
-methylpyrrole (MP) over 5% Ru/C. Conditions: 2.0 g substrate,
−1
−1
at different catalyst/substrate ratios: 0.1 g g (a) and 0.05 g g (b).
3
0 cm methanol, 25 °C, *25–60 °C 10 bar
3
Conditions: 2.0 g substrate, 50 cm methanol, 25 °C, *25–60 °C
1
0 bar
substrate ratio, depending on the number of reusing the
catalyst, in methanol is shown in Fig. 10.
the catalyst. As it is discussed in Sect. 3.4, a very dras-
tic decrease was observed in both the activity and the
conversion values (Fig. 9) during the reuse of 5% Ru/C
As seen, the acidic treatment of the spent 5% Ru/C cat-
alyst resulted in higher conversions of MP in every recy-
cling experiment (runs 2–5) than without it (Fig. 8b). The
reduction of MP was also complete during the first reusing
the catalyst (run 2), already at 25 °C, contrary to the result
obtained sans its acidic handling (97%). Similar decreases
were observed in the conversions in the third and fourth runs
−
1
at 0.05 g g catalyst/substrate ratio, therefore an acetic
acidic handling of the spent catalyst was also investigated.
Conversion of MP over 5% ruthenium on carbon treated
with 5% acetic acidic after each run, at 0.05 g g−1 catalyst/
1
3

L. Hegedűs et al.
Fig. 10 Conversion of 1-meth-
ylpyrrole (MP) vs. time over
5
% Ru/C using acidic treatment
after each run. Conditions: 2.0 g
3
substrate, 0.1 g catalyst, 50 cm
methanol, 25 °C, *25–60 °C,
1
2
2
0 bar; acidic treatment:
3
×5 cm 5% acetic acid, then
3
×10 cm distilled water
as did in those reactions where 5% ruthenium on carbon was
applied without acidic treatment, but better results (65 and
were strongly dependent on the amount of catalyst and the
number of recycling, respectively.
3
0%, respectively) were achieved. Finally, in the fifth run,
It was found, surprisingly, that ruthenium on carbon
showed high activity already at room temperature during
its first reuse, contrary to the fresh catalyst which worked
at only 60 °C. The prehydrogenation experiments showed
that high catalytic activity could be obtained after these
pretreatments, but this required a higher temperature
(60 °C). Whereas, using these conditions is inexpedient
on an industrial scale. This unexpected catalytic behaviour
of 5% Ru/C was clarified by XPS measurements which
a significantly higher conversion value was obtained over
the acidic treated catalyst (18%) than using the untreated
one (0%).
Figure 11a, b exhibit the initial rates (v ) and the conver-
0
sions at 60 min to compare the effects of acidic handling and
reuse of 5% Ru/C catalyst on the course of the hydrogenation
of MP.
As seen in Fig. 11a, a decrease of v values was observed
0
by increasing the number of reusing the acidic treated cata-
revealed the existence of a fine RuO layer on the surface
2
−
1
−1
lyst (5.06 → 0.44 nL H g
min ), but they were higher
of the catalytic metal in the fresh catalyst and verified
the favourable effect of catalyst prehydrogenation on the
catalytic activity.
2
Ru
than those were experienced in the presence of the untreated
−
1
−1
one (5.01 → 0 nL H g
min ). Similarly, the conver-
2
Ru
sions were also diminished (70 → 9%) during the reuse of
To obtain the complete conversion of MP over 5% Ru/C
even after more catalytic runs and without regeneration
of the catalyst, a relatively high catalyst/substrate ratio
5
% Ru/C treated with acetic acid (runs 2–5), but these val-
ues were significantly higher than those achieved with the
untreated catalyst (58 → 0%) (Fig. 11b).
−
1
(0.2 g g ) was applied, because in the presence of lesser
−
1
Although better results were obtained applying an acidic
treated 5% Ru/C catalyst, the strong poisoning of ruthenium
amounts of catalyst (0.1 or 0.05 g g ratios) significant
conversion and activity decreases were observed. This was
due to the strong poisoning of ruthenium caused by basic
nitrogen of the product (MPD). Although an acidic treat-
ment of the used catalyst resulted in better conversions at
−1
by nitrogen cannot be avoided at 0.05 g g catalyst/substrate
ratio and room temperature after its several reusing.
−
1
0
.05 g g catalyst/substrate ratio and 25 °C, the poisoning
4
Conclusions
caused by nitrogen cannot be avoided. However, reusing the
spent and unregenerated ruthenium on carbon, above a limit
catalyst/substrate ratio, can be an alternative method for Ru-
catalysed, heterogeneous catalytic hydrogenation processes
applied in fine chemical industry.
Poisoning phenomena caused by nitrogen and the reuse
of a heterogeneous, carbon supported ruthenium cata-
lyst were investigated in the liquid-phase hydrogenation
of 1-methylpyrrole (MP), as a model compound, in non-
acidic medium. Reusing the spent, unregenerated 5% Ru/C
catalyst, the activity of catalyst and the conversion of MP
Further investigations to clear poisoning mechanism of
these nitrogen containing heterocycles, over other catalytic
metals (rhodium, palladium), are in progress.
1
3

Poisoning and Reuse of Supported Precious Metal Catalysts in the Hydrogenation of…
4
5
6
7
8
.
.
.
.
.
Argyle MD, Bartholomew CH (2015) Catalysts 5:145–269
Hamilton TS, Adams R (1928) J Am Chem Soc 50:2260–2263
Maxted EB, Walker AG (1948) J Chem Soc 1093–1097
Maxted EB (1951) Adv Catal 3:129–178
Devereux JM, Payne KR, Peeling ERA (1957) J Chem Soc
2
845–2851
9
.
Maxted EB, Briggs MS (1957) J Chem Soc 3844–3847
1
0. Freifelder M (1971) Practical catalytic hydrogenation. John Wiley,
New York
1
1. Petró J (1976) In: Szabó ZG, Kalló D (eds) Contact catalysis,
vol 2.Elsevier, Amsterdam, pp 65–75
1
1
2. Hegedűs L (2007) Magy Kém Folyóirat 113:139–144
3. Hegedűs L, Máthé T, Tungler A (1996) Appl Catal A Gen
1
43:309–316
1
1
1
4. Hegedűs L, Máthé T, Tungler A (1996) Appl Catal A Gen
47:407–414
5. Hegedűs L, Máthé T, Tungler A (1997) Appl Catal A Gen
53:133–139
6. Hegedűs L, Máthé T, Tungler A (1997) Appl Catal A Gen
1
1
1
52:143–151
1
1
7. Leete E (1989) Heterocycles 28:481–487
8. Fournier AM, Brown RA, Farnaby W, Miyatake-Ondozabal H,
Clayden J (2010) Org Lett 12:2222–2225
1
9. Lee SY, Jung JW, Kim T-H, Kim H-D (2015) Arch Pharm Res
3
8:2131–2136
2
2
2
2
0. Hegedűs L, Máthé T (2002) Appl Catal A Gen 226:319–322
1. De Pablo RS, Benedict HN (1973) US Patent 3,766,089
2. Stillman NW, Osgood ER (1975) US Patent 3,864,361
3. Zhang M, Pei Q, Chen W, Liu L, He T, Chen P (2017) RSC Adv
7
:4306–4311
2
2
4. Hagelüken C (2006) Chim Oggi/Chem Today 24:14–17
5. Hagelüken C (2008) In: Ertl G, Knözinger H, Schüth F, Weitkamp
J (eds) Handbook of heterogeneous catalysis. Wiley, Weinheim,
pp 1846–1863
2
6. Axon SA, Casci JL (2008) In: Ertl G, Knözinger H, Schüth F,
Weitkamp J (eds) Handbook of heterogeneous catalysis. Wiley,
Weinheim, pp 1863–1871
2
2
7. Trimm DL (2001) Appl Catal A Gen 212:153–160
8. Grumett P (2003) Platinum Metals Rev 47:163–166
Fig. 11 Effects of acidic treatment after each run and reuse of the cat-
alyst on the initial rate (v ) (a) and the conversion at 60 min (b) in the
0
29. Marafi M, Stanislaus A (2003) J Hazard Mater B101:123–132
hydrogenation of 1-methylpyrrole (MP) over 5% Ru/C. Conditions:
3
3
3
0. Jackson SD (2006) Chem Eng J 120:119–125
3
2
.0 g substrate, 0.1 g catalyst, 50 cm methanol, 25 °C, *25–60 °C
1. Dufresne P (2007) Appl Catal A Gen 322:67–75
2. Al-Sheeha H, Marafi M, Raghavan V, Rana MS (2013) Ind Eng
Chem Res 52:12794 – 12801
3
3
1
0 bar; acidic treatment: 2×5 cm 5% acetic acid, then 2×10 cm
distilled water
3
3. EudraLex – Vol. 4, Good manufacturing practice (GMP) guide-
lines, http://ec.europa.eu/health/documents/eudralex/vol-4/index
_
en.htm
Acknowledgements The authors thank Dr. Antal Sárkány for deter-
mining the catalyst dispersion, as well as Dr. László Vida for GC–MS
measurements.
3
4. NIST Chemistry WebBook, NIST Standard Reference Database
Number 69 (2017) In: Linstrom PJ, Mallard WG (eds) National
Institute of Standards and Technology, Gaithersburg MD, 20899,
https://doi.org/10.18434/T4D303
Compliance with Ethical Standards
3
5. Sárkány A, Zsoldos Z, Furlong F, Hightower JW, Guczi L (1993)
J Catal 141:566–582
Conflict of interest The authors declare no competing financial inter-
36. Sárkány A, Stefler G, Hightower JW (1995) Appl Catal A Gen
127:77–92
est.
3
3
7. Bruce LA, Hoang M, Hughes AE (1998) J Catal 178:84–93
8. Knapp M, Crihan D, Seitsonen AP, Lundgren E, Resta A,
Andersen JN, Over H (2007) J Phys Chem C 111:5363–5373
9. Walker J, King RB, Tannenbaum R R (2007) J Solid State Chem
References
3
1
80:2290–2297
1
2
.
.
Bartholomew CH (2001) Appl Catal A Gen 212:17–60
Hegedus LL, McCabe RW (1984) Catalyst poisoning. Marcel
Dekker, New York
40. Over H (2012) Chem Rev 112:3356 – 3426
41. Wang L, Abudukelimu N, Ma Y, Qing S, Gao Z, Eli W (2014)
Reac Kinet Mech Cat 112:117–129
3
.
Moulijn JA, van Diepen AE, Kapteijn F F (2008) In: Ertl G, Knöz-
inger H, Schüth F, Weitkamp J (eds) Handbook of heterogeneous
catalysis. Wiley, Weinheim, pp 1829–1845
42. Szabados E, Srankó DF, Somodi F, Maróti B, Kemény S, Tungler
A (2016) J Ind Eng Chem 34:405–414
43. Morgan DJ (2015) Surf Interface Anal 47:1072–1079
1
3

L. Hegedűs et al.
Affiliations
László Hegedűs1 · Kristóf Szőke‑Molnár · István E. Sajó · Dávid Ferenc Srankó · Zoltán Schay
1
2
3
3
1
2
3
Department of Organic Chemistry and Technology, Budapest
University of Technology and Economics, Budafoki út 8,
Budapest 1111, Hungary
Surface Chemistry and Catalysis Department, Institute
for Energy Security and Environmental Safety, Centre
for Energy Research, Hungarian Academy of Sciences,
Konkoly-Thege Miklós út 29-33, 1121 Budapest, Hungary
Environmental Analytical and Geoanalytical Research
Group, Szentágothai Research Centre, University of Pécs,
If úság út 20, Pécs 7624, Hungary
1
3