Novel direct amination of glycerol over heteropolyacid-based catalysts

Article abstract of DOI:10.1039/c5cy01478f

The direct amination of glycerol with dimethylamine (DMA) was studied over catalysts based on salts of phosphomolybdic acid. The highest yield of (dimethylamino)acetone (33%) was observed over a catalyst containing 50 wt% of Cs_2.5H_0.5PMo₁₂O₄₀ supported on a mesoporous silica. The corresponding sample showed good dispersion of the active phase leading to an increased number of available acid sites. With respect to the observed products, a reaction scheme was established explaining notably that the formation of 1,3-bis-dimethylamine-2-propanol and (dimethylamino)propanol was not possible due to the blocking of the required strong acid sites by DMA under reaction conditions.

Full text of DOI:10.1039/c5cy01478f

Catalysis
Science &
Technology
View Article Online
View Journal
PAPER
Novel direct amination of glycerol over
heteropolyacid-based catalysts†
Cite this: DOI: 10.1039/c5cy01478f
M. Safariamin,ab S. Paul,ab K. Moonen,c D. Ulrichts,c F. Dumeignilbd
ab
*
and B. Katryniok
The direct amination of glycerol with dimethylamine (DMA) was studied over catalysts based on salts of
phosphomolybdic acid. The highest yield of (dimethylamino)acetone (33%) was observed over a catalyst
containing 50 wt% of Cs2.5H0.5PMo12O40 supported on a mesoporous silica. The corresponding sample
showed good dispersion of the active phase leading to an increased number of available acid sites. With
respect to the observed products, a reaction scheme was established explaining notably that the formation
of 1,3-bis-dimethylamine-2-propanol and (dimethylamino)propanol was not possible due to the blocking
of the required strong acid sites by DMA under reaction conditions.
Received 4th September 2015,
Accepted 28th October 2015
DOI: 10.1039/c5cy01478f
www.rsc.org/catalysis
temperature range between 250 and 290 °C. Hereby, most of
the research efforts have been focused on the use of shape-
Introduction
Amines are important products in organic synthesis.1,2 They
are used in numerous applications such as pharmaceuticals,
crop protection (pesticides and insecticides), food processing,
solvents for paints, fuel additives, or for oil & gas treatment.3
Amines can be obtained by amination of olefins, ketones
and alcohols. In the latter case, two main processes are
well established: reductive amination (often including the
borrowing hydrogen mechanism) and direct amination
(Scheme 1). Reductive amination processes involve the reac-
tion of an alcohol with ammonia or alkylamines on metal
catalysts, typically nickel or cobalt supported on silica or alu-
mina. Depending on the substrate and the product, the tem-
perature of this reaction ranges from 150 to 210 °C under a
hydrogen pressure between 17 and 20 bar.4 Two well-known
examples of reductive amination are i) the synthesis of tri-
ethylamine from ammonia and ethanol over a Co/SiO2 cata-
lyst at 200 °C and a pressure of 200 bar and ii) the synthesis
of fatty amines from long chain fatty alcohols with DMA over
Cu/Ni catalysts at 210 °C and a pressure between 50 and 200
bar.5–7 On the other hand, the direct amination uses solid
acid catalysts and proceeds at low pressure, in a typical
selective catalysts, e.g., zeolites. Zeolites are frequently used
for the synthesis of small chain aliphatic diamines,8 or cyclic
amines9 from the corresponding amino alcohols. Y-zeolites,
ZSM-5 and mordenites in their protonated form are the most
often employed molecular sieves.10 The main advantage of
direct amination notably relies on the use of such low cost
catalysts compared to the metal catalysts employed in the
reductive amination.
Recently, the use of polyols has drawn specific attention,
due to the demand in diamines and polyamines, especially
ethylenediamine and its polymers.11 One industrially interest-
ing polyol is glycerol, which is widely available, since it is a
co-product from the transesterification of vegetable oils for
biodiesel production. Hence, valorization of glycerol has
become a key issue in order to balance the overall economic
viability of the biodiesel industry, with many possible targets
(i.e., chlorination to epichlorohydrin, oxidation to fine
chemicals, dehydration to acrolein, etc.).12–18
In the present paper, we report on the direct amination of
glycerol with dimethylamine over heteropolyacid-based cata-
lysts with 3-(dimethylamino)-propane-1,2-diol (DMAPA) and
dimethylamino-2-propanone (DMA-acetone) as targets, the
a Ecole Centrale de Lille, Cité Scientifique, CS 20048, 59651 Villeneuve d'Ascq,
France. E-mail: benjamin.katryniok@ec-lille.fr; Tel: +33(0)3 20 33 54 99,
+33(0)3 20 33 54 38
b CNRS UMR 8181, Unité de Catalyse et Chimie du Solide, Université de Lille,
Cité Scientifique, Bâtiment C3, 59655 Villeneuve d'Ascq, France
c TAMINCO bvba, a subsidiary of the Eastman Chemical Company,
Pantserschipstraat 207, 9000 Ghent, Belgium
d Institut Universitaire de France, Maison des Universités, 103 Boulevard
Saint-Michel, 75005 Paris, France
† Electronic supplementary information (ESI) available. See DOI: 10.1039/
c5cy01478f
Scheme 1 Amination reaction of alcohols: reductive pathway (up) and
direct pathway (down).
This journal is © The Royal Society of Chemistry 2015
Catal. Sci. Technol.

View Article Online
Paper
Catalysis Science & Technology
former being used in the food industry, botanical extracts, oil
& gas treatment, and pharmaceutical and personal care appli-
cations, and the latter in solvents, feed additives, in the agro-
chemical industry, etc. Heteropolyacids (HPAs) are environ-
mentally friendly and economically viable Brønsted
acids.19–22 They are widely studied in the field of catalysis
and notably for the valorization of biomass.23 Hereby, HPAs
offer a unique feature to enabling fine modification of their
acidity – and thereby of their catalytic properties – via the
tuning of their composition, namely, by varying the central
atom, the addenda atoms and the counter ions.24–26
Phosphomolybdic acid was neutralized to different extents
with cesium to modulate its acidic properties. It was charac-
terized, tested, and optimized in the reaction of direct
amination of glycerol using a Design of Experiment (DoE)-
driven strategy.
The acid properties of the catalysts were determined by
temperature-programmed desorption of NH3 (NH3-TPD). A
typical NH3-TPD test was carried out as follows: 250 mg of
catalyst was pretreated in a He flow (30 mL min−1) at 250 °C
for 2 h in order to remove the physisorbed and crystal water.
Then, the sample was cooled to 130 °C, and NH3 pulsed
injections were carried out until saturation. The TPD profiles
were monitored using a TCD (Thermal Conductivity Detector)
and a mass spectrometer as detectors, and recorded from 130
to 600 °C at a 10 °C min−1 heating rate.
Raman spectra were recorded on a Jobin-Yvon LabRam
Infinity apparatus equipped with a CCD (Charge Coupled
Device) detector operating at the liquid nitrogen temperature.
A D2 filter was used to protect the catalyst structure from
thermal degradation by the laser beam heat (λ = 532 nm).
The spectra were recorded in the range between 200 and
1400 cm−1. The homogeneity of the samples was checked by
performing the analysis on at least three different particles
for each sample.
Experimental
Catalyst preparation
Elemental analysis was performed by X-ray fluorescence
energy dispersive X-ray spectroscopy (EDS) on a Hitachi
Phosphomolybdic acid (H3PMo12O40) is commercially avail-
able as crystalline hydrate (Sigma Aldrich, >99%). Neutral-
ized salts of phosphomolybdic acid (Cs3PMo12O40 and
S3600N electron microscope equipped with
a Thermo
Ultradry EDS detector using an acceleration voltage of 25 kV.
Cs2.5H0.5PMo12O40) were prepared by
a
precipitation
method. An aqueous solution of Cs2CO3 (0.1 M, Fluka) was
added via a liquid pump into an aqueous solution of
H3PMo12O40 (0.1 M). The temperature of the mixture was
maintained at 45 °C under vigorous stirring. After 2 h, the
solid was recovered by removing the solvent under reduced
pressure. The sample was then dried at 70 °C for 24 h.
The preparation of Cs2.5H0.5PMo12O40 supported on silica
was carried out in two steps. First, the silica carrier was
impregnated with cesium carbonate. 2.5 g of silica support
(CARIACT Q-10, kindly provided by Fuji Silysia) and 30 mL of
ethanol were stirred at 60 °C. Then, the desired amount of
cesium carbonate (0.1 M, Fluka) was added. After 30 minutes,
the solvent was evaporated, and the solid was dried at 60 °C
for 2 h. The as-prepared silica was then impregnated with
commercial phosphomolybdic acid; the support was
suspended in 30 mL of extra-dry ethanol. The required
amount of HPA dissolved in a minimum quantity of ethanol
was subsequently added. After 30 min, the solvent was evapo-
rated under vacuum at 60 °C, and the obtained solid was
dried at 50 °C for two hours. Catalysts supported on silica
with nominal loadings of 30, 50 and 70 wt% of
Cs2.5H0.5PMo12O40 were prepared, and the corresponding
samples were labeled 30CsPMo, 50CsPMo and 70CsPMo,
respectively.
Catalytic test
The direct amination of glycerol was carried out in a fixed
bed down-flow reactor (110 mm height and 8 mm internal
diameter) with a catalyst bed (1 g of catalyst) set at the mid-
dle of the reactor, in the isothermal zone. The catalyst was
loaded between two layers of SiC with a particle size of 0.1
mm. The temperature was controlled by a thermocouple
located near the catalyst bed. Nitrogen was used as a carrier
gas at a flow rate of 30 mL min−1, which was controlled by a
mass flow controller (Brooks). Glycerol and dimethylamine
(DMA) aqueous solutions (40 wt% each) were fed via two
HPLC pumps (Gilson) with flow rates of 0.1 mL min−1 and
2.4 mL min−1, respectively, and evaporated upfront the reac-
tor at 230 °C. The glycerol : DMA: N2 : H2O molar ratio was
6.1% : 6.4% : 17.2% : 70.4%. The reaction temperature was
fixed to 290 °C. The reaction products together with
unconverted glycerol were collected in two hour intervals in a
trap at room temperature and analyzed offline using gas
chromatography and high-performance liquid chromatogra-
phy. GC analyses were performed on an Agilent 6890 GC
using a flame ionization detector and equipped with a
Zebron CP Sil 5 CB semi-capillary column (30 m length, 3 μm
film thickness, 0.53 mm internal diameter). High Perfor-
mance Liquid Chromatography (HPLC) was used to quantify
the amount of glycerol using a refractory index detector
(Shodex RI-101) and a Rezex ROA-organic column (Phenom-
enex, 300 × 7.80 mm). Glycerol conversion and product selec-
tivities were calculated using eqn (1) and (2). The carbon bal-
ance was calculated as the percentage of the reacted carbon
atoms found in the products relative to the total number of
carbon atoms introduced in the reactor (eqn (3)).
Catalyst characterization
The textural properties were determined from N2 adsorption–
desorption isotherms recorded at the liquid nitrogen temper-
ature using an ASAP 2010 Micromeritics apparatus after
outgassing the solids at 130 °C for 2 h. The specific surface
area, the pore volume, and the pore size distribution were
obtained by using the BET and BJH methods, respectively.
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2015

View Article Online
Catalysis Science & Technology
Paper
126 m2 g−1, which is in the same range as the partially neu-
tralized salt (Cs2.5H0.5PMo12O40), exhibiting 131 m2 g−1. Both
samples also exhibited comparable pore volumes (0.184 and
0.202 cm3 g−1, respectively) and average pore radii (41 and 47
nm, respectively).
(1)
(2)
Characterization of supported HPA catalysts
In the following, the acidity of the silica-supported samples
(Cs2.5H0.5PMo12O40, this active phase being selected after the
tests on the bulk compounds) was probed by ammonia TPD
using a mass spectrometer for the detection of the desorbed
ammonia species (Fig. 1). In all the cases, the amount of acid
sites increased (Table 1). This increase is definitely not
related to the presence of acid sites on the silica support,
which is neutral, but is exclusively explained by the enhanced
dispersion of the HPA on the support. While the unsupported
Cs2.5H0.5PMo12O40 exhibited a total ammonia uptake of 0.01
mmol g−1 (Table 1), the supported samples, namely
30CsPMo, 50CsPMo and 70CsPMo, showed 0.03, 0.17 and
0.06 mmol g−1, respectively. The shape of the desorption pro-
file (Fig. 1) further shows that the samples mainly exhibit
weak acid sites (desorption temperature < 300 °C; Fig. 1).29
The fact that only weak sites were observed can be ascribed
to electronic interactions between the HPA and the support,
as described in the literature, which, while stabilizing the
HPA, decrease their acid strength.23,30,31
(3)
The experimental design was generated by the statistical
software Design-Expert, version 5.0.8, Stat-Ease Inc. The
response surface methodology was used in order to investi-
gate the influence of the varied parameters (reaction temper-
ature, DMA/GLY molar ratio and total reactant flow) on the
response variables.
Results and discussion
Characterization of bulk HPA catalysts
The catalysts were analyzed by NH3-TPD, nitrogen
physisorption, elemental analysis (EDS) and IR-Raman
spectroscopy.
The acid properties of the unsupported phosphomolybdic
acid salts were determined by NH3-TPD and the correspond-
ing results are reported in Table 1. As expected, the entirely
neutralized cesium salt of phosphomolybdic acid (Cs3PMo12O40)
exhibited no ammonia uptake (0 mmol g−1), meaning that a
total neutralization of the acidic protons was achieved. These
results are in contradiction to the literature, reporting that
stoichiometric salts partially precipitate as a mixture of the
parent HPA and cesium salt.27,28 Upon partial neutralization
with cesium (Cs2.5H0.5PMo12O40), the amount of acid sites
increased only slightly to 0.01 mmol g−1. This surprising
result could be due to the formation of a layer of entirely neu-
tralized phosphomolybdic acid (Cs3PMo12O40) on the outer
surface of the catalyst.
For the supported catalysts, the mesoporous silica
(CARiACT Q-10, Fuji Silysia) was chosen as a carrier in order
to avoid internal diffusion limitations. Correspondingly, the
impregnated catalysts show a type IV isotherm, with a hyster-
esis characteristic of mesoporous samples (Fig. S1†). Upon
impregnation of the support with the active phase, the spe-
cific surface area decreased from 268 m2 g−1 (bare support) to
234, 196 and 212 m2 g−1 for the catalysts containing 30, 50
and 70 wt% of the active phase, respectively (Table 1). In all
the cases, the decrease is rather limited to a small extent due
to the fact that the active phase (Cs2.5H0.5PMo12O40) itself
exhibits porous properties (cf. Table 1, line 2). On the other
hand, the average pore diameter of the support was
The textural properties of the bulk HPAs were measured
by N2 adsorption–desorption, and the corresponding results
are summarized in Table 1. The stoichiometric Cs salt of
phosphomolybdic acid exhibited a specific surface area of
Table 1 Textural properties of the prepared samples
Specific
Pore
Average
surface
volume
pore radius NH3 uptake
(nm)
Catalyst
area (m2 g−1
)
(cm3 g−1
)
(mmol g−1
)
Cs3PMo12O40
Cs2.5H0.5PMo12O40 131
SiO2
30CsPMo
50CsPMo
70CsPMo
126
0.184
0.202
1.21
0.77
0.54
0.32
41
47
18
12
11.2
12.2
0
0.01
—
0.03
0.17
0.06
268
234
196
212
Fig. 1 NH3-TPD profile of 30CsPMo (a), 50CsPMo (b) and 70CsPMo (c).
This journal is © The Royal Society of Chemistry 2015
Catal. Sci. Technol.

View Article Online
Paper
Catalysis Science & Technology
Table 2 Quantitative EDS and XRF analyses of supported salts of phosphomolybdic acid
Atom%
Catalyst
O (EDS) O (XRF) O theor.
Si (EDX) Si (XRF) Si theor. Mo (EDS) Mo (XRF) Mo theor. Cs (EDS) Cs (XRF) Cs theor.
30CsPMo 63.8
50CsPMo 59.5
70CsPMo 73.1
67.7
68.9
69.8
67.7
30.3
22.6
1.8
26.0
17.5
11.3
28.6
24.0
4.9
4.7
10.2
14.2
3.1
6.1
0.93
3.7
1.0
2.1
3.0
0.65
68.5
14.1
1.3
69.9 (73.3)a
17.5 (0)a 20.9
10.4 (22)a 4.9
2.1 (4.6)a
a
Theoretical amount for pure Cs2.5H0.5PMo12O40
.
significantly affected upon impregnation and decreased from
18 nm to 11–12 nm for the CsPMo-containing samples. Fur-
ther, note that the pore size distribution is rather fine for all
the samples, indicating a uniform covering of the pores,
except when too large amounts of the active phase were used
(70 wt%; Fig. 2).
In the following, we checked if the structure of the active
phase was preserved after impregnation on silica. Thus, IR-
Raman spectroscopy (Fig. 3) was used in order to verify the
presence of the heteropolyacid compounds. The Raman spec-
tra of the supported catalysts exhibited strong bands
assigned to the following vibrations: νsIJMoO) at 982 cm−1
,
Elemental analysis (EDS and XRF) was carried out on the
supported catalysts to check their elemental composition and
dispersion (Table 2 and Fig. S2†). In the case of 30CsPMo,
the as-determined composition by XRF and EDS was very
close to the theoretical one. On the other hand, for the
50CsPMo and 70CsPMo samples, the detected amount of Mo
and Cs by EDS was significantly higher than the theoretical
ones. This trend was less pronounced for the elemental com-
position determined by XRF analysis over these two samples.
In fact, the elemental composition determined by EDS for the
70CsHPA/SiO2 sample corresponded to the one of pure
Cs2.5H0.5PMo12O40. This can be explained by the low analysis
depth of EDS (100 nm). Indeed, a theoretical calculation
shows that only 62% of the surface is covered by the HPA salt
when using 30 wt% of the active phase (Scheme 2).32 On the
other hand, when increasing the amount of the active phase,
the coverage increases to 146% and 340% for 50 wt% and 70
wt% of the active phase, respectively, which corresponds to
the formation of a multilayered structure. In that case, EDS
will obviously give the elemental composition of bare CsPMo,
which is indeed the case (Table 2, line 3).
νsIJMo–Oa) at 245 cm−1 and νsIJMo–O–Mo) at 256 and 237
cm−1.33 The low intensity – but characteristic – shoulder at
994 cm−1 was assigned to the formation of MoO3 particles,
thus suggesting a very limited decomposition of the HPA.34
This confirms the preservation of the HPA structure upon
supporting on SiO2.
Nevertheless, it is noteworthy that some particles of the
catalyst based on 30CsPMo exhibited no characteristic signal
of the HPA compound, thus meaning that the active phase
was not homogeneously dispersed on all the catalyst parti-
cles, which had significant impact on the catalytic perfor-
mance. On the other hand, the catalyst based on
70CsPMo contained some particles that consisted of pure
HPA. Consequently, only the catalyst based on 50CsPMo
exhibited a homogeneous distribution of the active phase.
Catalytic results
The catalytic performance of the solids is reported in Table 3.
The stoichiometrically neutralized phosphomolybdic acid –
Cs3PMo12O40 – showed a high conversion of glycerol (25%)
but a rather low selectivity to DMA-acetone (29%). On the
other hand, the partially neutralized salt Cs2.5H0.5PMo12O40
showed a significantly higher selectivity of 63% but at a lower
conversion rate of 8% (Table 3). Since both catalysts exhibit
rather comparable specific surface areas of 126 and 131
m2 g−1 and acidities of 0 and 0.01 mmol g−1 (Table 1),
Fig. 2 Pore size distribution for bare silica (+), 30CsPMo (x), 50CsPMo
(o) and 70CsPMo (□).
Scheme 2 Silica coverage for 30 wt% (a); 50 wt% (b); and 70 wt% of
CsHPA (c).
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2015

View Article Online
Catalysis Science & Technology
Paper
Scheme 3 Reaction scheme for the formation of DMA-acetone from
glycerol.
the first step, glycerol is dehydrated over the acid catalyst to
an enol intermediate (2-propen-1-diol). The latter is not
stable and can either further react with dimethylamine to
DMAPA or tautomerize to hydroxyacetone. The formation of
DMA-acetone can then proceed via amination of hydroxy-
acetone or by dehydration of DMAPA. As DMAPA was indeed
observed during the catalytic tests, we propose that the for-
mation of DMA-acetone involves DMAPA as an intermediate.
Furthermore, it is worth mentioning that neither 1,3-bis-
dimethylamine-2-propanol (BDMAP) nor DMA-propanal was
observed. The latter could originate from the amination of
acrolein formed by dehydration of glycerol. The correspond-
ing dehydration reaction is well-described over medium to
strong acid catalysts.12–14 Nevertheless, under the reaction
conditions (reaction temperature of 290 °C), all the medium/
strong acid sites should be covered by DMA, whereby they
cannot interact with glycerol to enable the formation of acro-
lein. Consequently, the formation of DMA-propanal was
mechanistically made impossible.
Fig. 3 Raman spectra of 30CsPMo (a), 50CsPMo (b), and 70CsPMo (c).
Table 3 Catalytic test results
Selectivity (%)
X
CB
Y
Catalyst
(%)
(%)
DMA-acetone
DMAPA
(%)
Cs3PMo12O40
Cs2.5H0.5PMo12O40
30CsPMoa
25
8
61
33
8
95
95
62
85
85
29
63
2
33
30
0
0
0
0
0
7
5
1
11
2
50CsPMoa
70CsPMoa
X = conversion; CB = carbon balance; Y = (DMA-acetone + DMAPA)
yield; reaction conditions: T = 290 °C, P = 1 bar, glycerol : DMA: N2 :
H2O = 6.1% : 6.4% : 17.2% : 70.4%, contact time = 0.25 s, reaction
time = 3 h. a Same conditions but time = 0.36 s.
respectively, no correlation between the activity and the spe-
cific surface area or acidity could be found.
In the next step, we decided to study the possibility of
supporting Cs2.5H0.5PMo12O40 (CsHPA) to increase the acces-
sibility of the acid sites. The choice for Cs2.5H0.5PMo12O40 as
the active phase was motivated by its rather high selectivity
In the next step, the optimization of the reaction condi-
tions was performed by computer-assisted design of experi-
ments using a hybrid design (Design-Expert, 11 experiments).
Three factors were considered: the reaction temperature, the
DMA/GLY ratio and the total reactant flow (contact time).
The corresponding values were varied from 260 to 290 °C, 0.5
to 1.5, and 12 to 28 L h−1, respectively. 50CsPMo was selected
as the catalyst for this study, due to the fact that it exhibited
the highest DMA yield of the series.
Fig. 4 illustrates the influence of the reaction tempera-
ture and the DMA/GLY ratio on the conversion of glycerol.
At a glance, we can see that the DMA/GLY ratio had a very
limited influence on the conversion of glycerol. On the
other hand, glycerol conversion clearly increased with the
temperature and the reactant flow, reaching up to 70% at
280 °C and 18 L h−1.
The influence of the reaction conditions on the DMA-
acetone selectivity is shown in Fig. 5. At a reaction tempera-
ture of 250 °C and a relatively low reactant flow of 10 L h−1,
the selectivity exhibited its maximum at 55%. No absolute
optimum in terms of DMA-acetone selectivity was observed
for the variation of the DMA/GLY ratio in the range between
0.5 and 1.5.
Following the parameter optimization, we performed a
test using the optimal parameter predicted by the model,
namely 250 °C, a reactant flow of 10 L h−1 and a DMA/GLY
to DMA-acetone (63%), but at
a low conversion (8%).
Consequently, Cs2.5H0.5PMo12O40 was impregnated over a
commercial silica (CARiACT Q-10) at the different loadings of
30, 50, and 70 wt%. The catalytic performance of the as-
obtained supported catalysts is listed in Table 3, which shows
that the catalytic activity decreased when increasing the
amount of the active phase. Whereas 30CsPMo showed 61%
conversion, the sample containing 70 wt% of CsHPA
exhibited no more than 8% of conversion. The low activity in
this latter case was ascribed to the low dispersion of the
active phase, as observed by EDS (cf. Scheme 1), thus
resulting in a behaviour comparable to that of a bulk catalyst.
With regard to the selectivity to DMA-acetone, a significant
decrease was observed when the amount of the active phase
was low (30 wt%). In that case, the selectivity to DMA-acetone
was no more than 2%, compared to 33% and 30% for the cat-
alysts containing 50 and 70 wt% of the active phase,
respectively.
Concerning the observed products, we propose the follow-
ing reaction mechanism starting from glycerol (Scheme 3). In
This journal is © The Royal Society of Chemistry 2015
Catal. Sci. Technol.

View Article Online
Paper
Catalysis Science & Technology
Fig. 4 Glycerol conversion as a function of the reaction temperature and the reactant flow at a constant DMA/GLY ratio of 1.25 (left); glycerol
conversion as a function of the reaction temperature and DMA/GLY ratio at a constant reactant flow of 14 L h−1 (right).
Fig. 5 DMA-acetone selectivity as a function of the reaction temperature and the reactant flow at a constant DMA/GLY ratio of 1.25 (left); DMA-
acetone selectivity as a function of the reaction temperature and DMA/GLY ratio at a constant reactant flow of 14 L h−1 (right).
Fig. 6 DMA-acetone yield as a function of the reaction temperature and the reactant flow at a constant DMA/GLY ratio of 1.25 (left); DMA-
acetone yield as a function of the reaction temperature and the DMA/GLY ratio at a constant reactant flow of 14 L h−1 (right).
ratio equal to 1.5. Using these parameters, a yield of 33% in
DMA-acetone was observed (47% conversion and 70% selec-
tivity), which is significantly higher than the value predicted
by the model (Fig. 6). These results can be explained by the
fact that these parameters are at the boundaries of the value
range, where the model is significantly less precise than in
the central area.
70 wt%) resulted in a bulk-like behavior, whereas a lower
loading (i.e., 30 wt%) led to sub-monolayer coverage.
Furthermore, we demonstrated the strong influence of the
reaction conditions and notably the reaction temperature on
the catalytic performance by employing an experimental
design approach. The best performance was obtained over a
mesoporous silica-supported Cs2.5H0.5PMo12O40 catalyst
containing 50 wt% of the active phase, which exhibited a
maximum yield of 33% in DMA-acetone.
Conclusions
Acknowledgements
Various salts of phosphomolybdic acid were used as catalysts
in the direct amination of glycerol in the presence of DMA.
The use of silica as a support for the cesium salts of
phosphomolybdic acid (Cs2.5H0.5PMo12O40) significantly
increased the amount of acid sites, due to better dispersion
of the active phase, thus resulting in an increased conversion
of glycerol compared to the parent bulk Cs2.5H0.5PMo12O40
catalyst. A good distribution was notably obtained for a load-
ing of 50% of the active phase, as shown by IR-Raman and
EDS spectroscopy. On the other hand, a higher loading (i.e.,
The authors warmly thank the Eastman Chemical Company
for financial support. Furthermore, we thank G. Cambien,
J.-C. Morin and O. Gardoll for the nitrogen physisorption
experiments, technical assistance during IR-Raman analysis
and ammonia TPD, respectively. We further thank Fuji Silysia
for kindly supplying CARiACT Q-10 silica. The Chevreul Insti-
tute (FR 2638), Ministère de l'Enseignement Supérieur et de
la Recherche, Région Nord – Pas de Calais and FEDER are
acknowledged for supporting and partially funding this work.
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2015

View Article Online
Catalysis Science & Technology
Paper
References
17 B. Katryniok, S. Paul and F. Dumeignil, ACS Catal.,
2013, 3(8), 1819.
1 T. Henkel, R. M. Brunne, H. Müller and F. Reichel, Angew.
Chem., Int. Ed., 1999, 38, 643.
2 T. C. Nugent, Chiral Amine Synthesis – Developments and
Applications, Wiley-VCH, Weinheim, 2010.
3 T. A. Unger, Pesticide Synthesis Handbook, Noyes
Publications, Parkridge, 1996.
4 J. V. Martinez de Pinillos and R. L. Fowlkes, US Pat.
4314084, 1982.
5 R. Fischer and H. Mueller, US Pat. 4851578, 1989.
6 H. Kimura and H. Taniguchi, Appl. Catal., A, 2005, 287, 191.
7 H. Kimura, Catal. Rev.: Sci. Eng., 2011, 53, 1.
8 K. S. Hayes, Appl. Catal., A, 2001, 221, 187.
9 J. O. H. Peterson and H. S. Fales, US Pat. 4375002, 1983.
10 K. Eller, E. Henkes, R. Rossbacher and H. Hokf,
Encyclopedia of industrial chemistry, Wiley-VCH, Weinheim,
2012, vol. 2, DOI: 10.1002/14356007.a02_001.
11 A. Fischer, T. Mallat and A. Baiker, Catal. Today, 1997, 37,
167.
12 B. Katryniok, S. Paul, V. Bellière-Baca, P. Rey and F.
Dumeignil, Green Chem., 2010, 12, 2079.
13 B. Katryniok, S. Paul, M. Capron and F. Dumeignil,
ChemSusChem, 2009, 2, 719.
14 B. Katryniok, S. Paul and F. Dumeignil, ACS Catal.,
2013, 3(8), 1819.
15 B. Katryniok, H. Kimura, E. Skrzynska, J.-S. Girardon, P.
Fongarland, M. Capron, R. Ducoulombier, N. Mimura, S.
Paul and F. Dumeignil, Green Chem., 2011, 13, 1960.
16 M. Pagliaro and M. Rossi, The Future of Glycerol, RSC
Publishing, 2008.
18 R. Ciriminna, B. Katryniok, S. Paul, F. Dumeignil and M.
Pagliaro, Org. Process Res. Dev., 2015, 19, 748.
19 I. V. Kozhevnikov, Chem. Rev., 1998, 98, 171.
20 M. N. Timofeeva, Appl. Catal., A, 2003, 256, 19.
21 T. Okuhara, Catal. Today, 2002, 3, 167.
22 H. Firouzabadi, N. Iranpoor and A. Jafari, Tetrahedron Lett.,
2005, 46, 2683.
23 I. V. Kozheynikoy, Handbook of Green Chemistry, Wiley-VCH,
Weinheim, 2009, vol. 2.
24 M. N. Timofeeva, Appl. Catal., A, 2003, 256, 19.
25 B. Katryniok, S. Paul, M. Capron, C. Lancelot and V. Bellière-
Baca, Green Chem., 2010, 12, 1922.
26 B. Katryniok, S. Paul and F. Dumeignil, Green Chem.,
2010, 12, 1910.
27 A. Corma, A. Martinez and C. Martinez, J. Catal., 1996, 164, 422.
28 H. Atia, U. Armbruster and A. Martin, Appl. Catal., A,
2011, 393, 331.
29 K. Tanabe, M. Misono, Y. Ono and H. Hattori, Stud. Surf.
Sci. Catal., 1989, 51, 5.
30 A. M. Alsalme, P. V. Wiper, Y. Z. Khimyak, E. F. Kozhevnikova
and I. V. Kozhevnikov, J. Catal., 2010, 276, 181.
31 A. Alsalme, E. F. Kozhevnikova and I. V. Kozhevnikov, Appl.
Catal., A, 2010, 390, 219.
32 M. Kanno, T. Yasukawa, W. Ninomiya, K. Ooyachi and Y.
Kamiya, J. Catal., 2010, 273, 1.
33 Q. Huynh, A. Selmi, G. Corbel, P. Lacorre and J. Millet,
J. Catal., 2009, 266, 64.
34 A. Brückner, G. Scholz, D. Heidemann, M. Schneider and D.
Herein, J. Catal., 2007, 245, 369.
This journal is © The Royal Society of Chemistry 2015
Catal. Sci. Technol.