Vapor-phase self-aldol condensation of butanal over Ag-modified TiO2

Article abstract of DOI:10.1016/j.apcata.2016.05.018

Vapor-phase self-aldol condensation of butanal was performed over various solid catalysts. Among the tested catalysts, SiO₂-Al₂O₃, Nb₂O₅ and TiO₂ showed relatively high catalytic activity for the formation of aldol condensation product, 2-ethyl-2-hexenal, whereas all the catalysts deactivated rapidly. In order to stabilize the catalytic activity, metal-modified catalysts were investigated in hydrogen flow, and it was found that Ag-modified TiO₂ showed the best catalytic performance. Characterizations such as XRD, TPD, TPR, TG-DTA, and DRIFT were performed for investigating the effect of the additive Ag and analyzing the coke component. The loaded Ag metal inhibited the formation of carbon accumulated on catalyst surface, and H₂ carrier gas was indispensable in the inhibition. Ag would work as a remover of the products on the catalyst surface together with H₂ to prevent dehydrogenation followed by coke formation. Self-aldol condensation of butanal was stabilized over Ag-modified TiO₂ at Ag₂O loadings higher than 3 wt.% at 220 °C in H₂ flow. TiO₂ with Ag₂O of 5 wt.% showed the best catalytic performance and gave a 72.2% selectivity to 2-ethyl-2-hexenal at 72.1% conversion in H₂ flow at 220 °C.

Full text of DOI:10.1016/j.apcata.2016.05.018

Accepted Manuscript
Title: Vapor-phase self-aldol condensation of butanal over
Ag-modified TiO2
Author: Daolai Sun Shizuka Moriya Yasuhiro Yamada
Satoshi Sato
PII:
S0926-860X(16)30271-X
DOI:
Reference:
http://dx.doi.org/doi:10.1016/j.apcata.2016.05.018
APCATA 15882
To appear in:
Applied Catalysis A: General
Received date:
Revised date:
Accepted date:
6-4-2016
14-5-2016
19-5-2016
Please cite this article as: Daolai Sun, Shizuka Moriya, Yasuhiro Yamada, Satoshi
Sato, Vapor-phase self-aldol condensation of butanal over Ag-modified TiO2, Applied
Catalysis A, General http://dx.doi.org/10.1016/j.apcata.2016.05.018
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.

A revised manuscript for the original paper submitted to the journal of Appl. Catal. A: Gen.
Vapor-phase self-aldol condensation of butanal over Ag-modified TiO2
Daolai Sun, Shizuka Moriya, Yasuhiro Yamada, Satoshi Sato*
Graduate School of Engineering, Chiba University, Chiba, 263-8522, Japan
* Corresponding author. Tel. +81 43 290 3376; Fax: +81 43 290 3401
E-mail address: satoshi@faculty.chiba-u.jp (S. Sato)
1

Graphic abstract
100
80
60
40
20
0
100
80
60
40
20
TiO 2
5 wt.% Ag2O loaded TiO 2
5 wt.% Ag2O loaded TiO 2
O
O
Aldol condensation
220 oC in H2 flow
2
2-Ethyl-2-hexenal
Butanal
TiO 2
0
1
2
3
4
5
0
1
2
3
4
5
Time on stream / h
Time on stream / h
2

Highlights
Ag-modified TiO2 was found to be efficient for stabilizing the catalytic activity.
The loading of Ag2O on TiO2 can inhibit carbon deposition onto the catalyst surface.
H2 carrier gas was indispensable for stabilizing the catalytic activity together with Ag.
3

Abstract
Vapor-phase self-aldol condensation of butanal was performed over various solid catalysts.
Among the tested catalysts, SiO2-Al2O3, Nb2O5 and TiO2 showed relatively high catalytic activity for
the formation of aldol condensation product, 2-ethyl-2-hexenal, whereas all the catalysts deactivated
rapidly. In order to stabilize the catalytic activity, metal-modified catalysts were investigated in
hydrogen flow, and it was found that Ag-modified TiO2 showed the best catalytic performance.
Characterizations such as XRD, TPD, TPR, TG-DTA, and DRIFT were performed for investigating
the effect of the additive Ag and analyzing the coke component. The loaded Ag metal inhibited the
formation of carbon accumulated on catalyst surface, and H2 carrier gas was indispensable in the
inhibition. Ag would work as a remover of the products on the catalyst surface together with H2 to
prevent dehydrogenation followed by coke formation. Self-aldol condensation of butanal was
stabilized over Ag-modified TiO2 at Ag2O loadings higher than 3 wt.% at 220 oC in H2 flow. TiO2
with Ag2O of 5 wt.% showed the best catalytic performance and gave a 72.2% selectivity to 2-ethyl-
2-hexenal at 72.1% conversion in H2 flow at 220 oC.
Keywords: aldol condensation, butanal, 2-ethyl-2-hexenal, Ag-modified TiO2, catalytic
deactivation.
4

1. Introduction
Aldol condensation of aldehydes and/or ketones is one of the most powerful methods for
the formation of the C-C bond, and has been widely used for the production of industrially important
chemicals, such as 2-ethylhexanal, isophorone, mesityl oxide and crotonaldehyde [1-2]. In recent
years, biomass conversion into energy and useful chemicals has attracted much attention. Dumesic
et al. have developed a process of sugar conversion into monofunctional compounds such as
alcohols, ketones, carboxylic acids and heterocyclic compounds at the range of C4-C6 [3-5]. These
compounds can be converted to large chemicals by the C-C bond formation through aldol
condensation and ketonization, and biomass-derived liquid fuel can be produced by the following
hydrodeoxygenation process [3-5]. Aldol condensation of aldehydes can be catalyzed by acid, base
or acid-base bifunctional catalyst [1, 6], and industrial aldol condensation processes are performed in
the presence of sulfuric acid or sodium hydroxide in a liquid phase [7-10]. In the current processes,
however, there are many disadvantages, such as the disposal of acid or base containing in the waste
solution and the complicated separation processes. Aldol condensations over solid catalysts have
been investigated to replace the current processes in the past twenty years. In a liquid-phase reaction,
acid catalysts, such as zeolites [11], sulfate-modified ZrO2 [8], Nb2O5 [12], and TiO2 [13-14], as well
as base catalysts, such as alkaline earth metal oxides [15] and supported alkaline metal oxides [15-
17], have been found to be efficient to catalyze the aldol condensation. In a vapor-phase reaction,
aldol condensation also proceeds over both acid catalysts such as zeolites [18] and base catalysts
such as MgO [19], whereas the catalysts deactivate rapidly because of the serious carbon
accumulation on the catalyst surface.
Self-aldol condensation of butanal, has been studied intensively because the aldol
condensation product of 2-ethyl-2-hexenal (2E2H) can be readily hydrogenated into 2-ethylhexanal
(2EH) and 2-ethylhexanol, which are important raw materials for the production of perfumes,
cosmetics, and plasticizers [20-23]. Self-aldol condensation of butanal is also a model reaction for
the study of aldol condensation. Hamilton et al. performed vapor-phase self-aldol condensation of
butanal over Pd/Na/SiO2 catalyst at 350 oC: a high conversion of 81.8% with an 80.7% selectivity to
5

aldol condensation products, which contained 2E2H, 2EH and 2-ethylhexanol, was obtained at a
time on stream of 4 h, whereas the conversion decreased to 49.1% at a time on stream of 8 h [24]. In
another example, Pd/K/zeolite gives an initial butanal conversion of ca. 75% with an initial 2E2H
formation rate of ca. 31 mmol g-1 h-1 at 150 oC, whereas the conversion decreased to ca. 30% at a
time on stream of 5 h [21]. Vapor-phase aldol condensation of butanal was also performed over
MgO/SiO2: the conversion at a time on steam of 15 min was 17.4% at 300 oC, and it decreased to
5.1% at a time on stream of 120 min [25].
Deactivation of catalysts is an important problem in catalytic conversion under continuous
flow conditions because the stability of the catalyst is highly required in a practical industrial process.
In our group, an effective operation on the stabilization of catalytic activity of acidic catalysts has
been reported [26-29]. For example, vapor-phase cyclodehydration of diethylene glycol (DEG) into
1,4-dioxane can be catalyzed by acidic Al2O3, whereas the conversion of DEG over Al2O3 is
seriously deteriorated at 250 °C. We developed an Ag-modified Al2O3 catalyst, which exhibited
stable catalytic activity with high selectivity to 1,4-dioxane under hydrogen flow conditions. Prior to
the reaction, Ag2O was reduced to metallic Ag, which works as a product remover together with
hydrogen to prevent coke formation [29].
In this paper, vapor-phase aldol condensation of butanal was performed over various solid
catalysts. Metal-modified catalysts were prepared for inhibiting carbon accumulation and stabilizing
the catalytic activity in the conversion of butanal. The suitable reaction conditions for the formation
of aldol condensation products and the effect of metal loading were also investigated.
2. Experimental
2.1 Samples
Butanal was purchased from Wako Pure Chemical Industries, Japan, and was used for the
catalytic reaction without further purification. SiO2-Al2O3 (N632HN, SiO2/Al2O3=4, SBET=389 m2 g-
1) was purchased from Nikki Chemical Co., Ltd. MFI zeolite (Si/Al=90.2, SBET=431 m2 g-1), Nb2O5
(SBET=108 m2 g-1), and anatase TiO2 (CS-750-24, SBET=40 m2 g-1) were supplied by Clariant
6

Catalysts Industries Co., Ltd, CBMM Asia Co., Ltd., and Sakai Chemical Industry Co., respectively.
ZrO2 (JRC-ZRO-4, SBET=30 m2 g-1), CeO2 (JRC-CEO-3, SBET=81 m2 g-1) and SiO2-MgO (JRC-
SM-2, SBET=642 m2 g-1) were supplied by Catalyst Society of Japan. γ-Al2O3 (DC-2113, SBET=125
m2 g-1) and Ca10(PO4)6(OH)2 were supplied by Dia Catalyst & Chemicals Ltd. and Wako Pure
Chemical Industries Ltd., Japan, respectively. AgNO3 and Cu(NO3)2, which were used as the
precursors of the additive metals, were purchased from Wako Pure Chemical Industries, Ltd., Japan.
Metal-modified catalysts were prepared by an incipient wetness impregnation method using
a solution with a prescribed amount of metal nitrate dissolved in distilled water. The precursor
solution with an amount of ca. 0.3 ml was dropped onto a support of 3 g, and the water was
evaporated at ambient pressure and 70 oC by being illuminated by 350-W electric light bulb. The
operation was repeated until all the precursor solution was added. After the impregnation process,
the samples were dried at 110 oC for 12 h, and then calcined for 3 h. The calcination temperature of
o
metal-modified catalysts was 400 C. The weight percentages of metal in the catalyst were
calculated using the weight of Ag2O and CuO in the calcined catalysts. Hereafter, the supported
catalysts are expressed as A-B-x, where A means the support; B means the additive metal; x means
the weight percentage of the additive metal oxide of B. For example, TiO2-Ag-1 means 1 wt.%
Ag2O loaded on TiO2.
2.2 Catalytic reaction
The self-aldol condensation of butanal was performed in a fixed-bed down-flow glass reactor
with an inner diameter of 17 mm at an ambient pressure of either H2 or N2. Prior to the reaction, a
catalyst of either 1.0 or 2.0 g was placed in the catalyst bed and heated at 250 oC for 1 h. After the
temperature of the catalyst bed had been stabilized at a prescribed temperature, butanal was fed
through the top of the reactor at a liquid feed rate of 2.0 cm3 h-1 together with either an H2 or N2 flow of
20 cm3 min-1. The liquid effluents collected in a dry ice-acetone trap (-78 oC) every hour were analyzed
by a FID-GC (GC-14B, Shimadzu) with a 60-m capillary column of TC-WAX (GL-Science, Japan).
A GC-MS (QP5050A, Shimadzu) was used for identification of the products in the effluent. 2-Octanol
7

was used as an internal standard substance.
2.3 Characterization of catalysts
The specific surface area of the catalysts was calculated by the BET method using N2
isotherm measured in a self-made gas adsorption apparatus at -196 oC. Temperature-programmed
desorption (TPD) and temperature-programmed reduction (TPR) were performed by using self-
made apparatuses. The TPD of adsorbed NH3 and CO2 was measured by neutralization titration
using an electric conductivity cell immersed in an aqueous solution of H2SO4 and NaOH to estimate
the acidity and the basicity of the catalysts, respectively, as has been described in the TPD
experiment in our previous studies [28,30]. The samples were preheated at 500 oC in vacuum
conditions for 1 h before the adsorption. The TPR measurements were performed from 25 to 900 oC
at a heating rate of 5 oC min-1 and the details are described elsewhere [31]. The X-ray diffraction
(XRD) patterns of the catalyst samples were recorded on a D8 ADVANCE (Bruker, Japan) using
Cu Kα radiation. The thermogravimetry analysis (TG) was performed using Thermoplus 8120E2
(Rigaku) under the conditions: sample weight, ca. 10 mg; the rate of the temperature increase, 5 oC
min-1; heating range, from the room temperature to 900 oC. The diffuse reflectance infrared Fourier-
transform (DRIFT) spectra of the catalysts were recorded on a spectroscopy using FT/IR-4200
(JASCO).
3. Results
3.1. Self-aldol condensation of butanal over various solid acidic catalysts.
Table 1 shows the reaction results of self-aldol condensation of butanal over various solid
acid catalysts. The reactions were performed at 200 oC in an N2 flow at a flow rate of 20 cm3 min-1. It
is proposed that the formation routes of each product as shown in Scheme 1. Aldol condensation
product of 2E2H was the major product in the reactions, and 2EH, which was generated by the
further hydrogenation of 2E2H, was detected at small amounts. Butanoic acid (BA), 1-butanol
8

(BuOH) and butyl butyrate (BB) were also detected as by-products. BB would be generated through
Tishchenko reaction of butanal, BA and BuOH are supposed to be produced from the hydrolysis of
BB. Among all the tested catalysts, Nb2O5, SiO2-Al2O3 and TiO2 showed the relatively high
conversion of butanal and selectivity to 2E2H. Although MgO/SiO2, γ-Al2O3 and Ca10(PO4)6(OH)2
gave 2E2H selectivity higher than 70%, the conversion of butanal was lower than 10%. The initial
conversion of butanal was 19.4% over γ-Al2O3, while it deactivated rapidly and the conversion of
butanal decreased to ca. 2% at a time on stream of 5 h. ZrO2 showed poor catalytic activity for
butanal conversion and gave relatively high selectivity to BA.
Fig. 1a and 1c shows the NH3- and CO2-TPD profiles of Nb2O5, SiO2-Al2O3 and TiO2,
respectively. The NH3 desorption peaks of SiO2-Al2O3, Nb2O5 and TiO2 were observed at 363, 223
and 272 oC, respectively. The order of the number of acid site was as follows: SiO2-Al2O3> Nb2O5>
TiO2 and the order of the averaged acid strength was SiO2-Al2O3> TiO2> Nb2O5. Fig. 1b shows the
NH3-TPD profiles per unit surface area, which present the acidity density of each catalyst. The order
of the acidity density was as follows: TiO2> Nb2O5> SiO2-Al2O3. An inverse peak at ca. 700 oC was
observed in NH3-TPD profile of TiO2, and it was also observed in NH3-TPD profile of TiO2 without
NH3 adsorption. Thus, the inverse peak in NH3-TPD profile is proposed to attribute to the acid
component such as SO2 containing in the original TiO2, which could not be removed completely in
the catalyst preparation process. Because of the same reason, the peak observed at a temperature
higher than 700 oC in CO2-TPD profile of TiO2 would not attribute to CO2 desorption (Fig. 1c).
Since there were no CO2 desorption observed in Fig. 1c, it was confirmed that there was almost no
basic sites in SiO2-Al2O3, Nb2O5 and TiO2.
Fig. 2 shows the changes in the butanal conversion and the selectivity to aldol condensation
products of 2E2H and 2EH with time on stream over Nb2O5, SiO2-Al2O3 and TiO2. The conversion
of butanal decreased rapidly with time on stream in all the reactions. TiO2 showed the highest initial
selectivity to the aldol condensation products, which decreased rapidly with time on stream. Because
H2 carrier gas was efficient for inhibiting catalytic deactivation in our previous studies [26-29], the
reactions over Nb2O5, SiO2-Al2O3 and TiO2 in H2 carrier gas were performed. Fig. 3 compares the
9

changes in the conversion of butanal and the selectivity to the aldol condensation products with time
on stream in both N2 and H2 flow. No significant differences of conversion and selectivity were
observed in H2 and N2 carriers in the reactions over Nb2O5 and SiO2-Al2O3. On the other hand, there
was a difference in the conversion of butanal over TiO2: the initial conversion of butanal over TiO2
was 58% in H2 flow, which was higher than 42% in N2 flow. However, the conversion of butanal
over TiO2 in H2 flow decreased rapidly with time on stream comparing with that in N2 flow.
Fig. 4 shows the effect of reaction temperature over TiO2 in H2 flow. The selectivity to the
aldol condensation products was relatively stable at temperatures of 180 oC and lower. The reaction
performed at 200 oC gave the highest initial selectivity to the aldol condensation products, which
decreased with time on stream. On the other hand, the reaction performed at 240 oC showed the
lowest selectivity to the aldol condensation products, and it also decreased rapidly with time on
stream. The conversions of butanal at all the temperatures decreased with time on stream. The initial
conversion was relatively low at a temperature of 160 oC, and it decreased slowly with time on
stream. Although the initial conversions were as high as ca. 58% at reaction temperatures of 180 oC
and higher, they decreased rapidly with time on steam.
3.2. Self-aldol condensation of butanal over metal-modified catalysts.
In our previous studies, the loading of Ag onto SiO2-Al2O3 and the loading of Cu onto
Al2O3 were found to be effective for stabilizing the vapor-phase dehydration of 1,2-propanediol into
propanal [28] and the dehydration of tetrahydrofurfuryl alcohol into 3,4-2H-dihydropyran [26],
respectively. Thus, Ag- and Cu-modified catalysts were prepared and applied for the self-aldol
condensation of butanal. Table 2 and Fig. 5c show the reaction results of butanal self-aldol
condensation over TiO2, 3 wt.% Ag2O- (TiO2-Ag-3) and 5 wt.% CuO-modified TiO2 (TiO2-Cu-5)
catalysts at 200 oC in H2 flow, and the appropriate loading of each metal oxide was determined
through the individual activity tests. The loading of CuO onto TiO2 slightly decreased the averaged
selectivity to the aldol condensation products, but increased the averaged conversion of butanal. The
selectivity to the aldol condensation products over TiO2-Cu-5 was more stable than that over
10

unmodified TiO2, whereas the conversion of butanal still decreased with time on stream steeply.
TiO2-Ag-3 showed a lower selectivity to the aldol condensation products than those over TiO2 and
TiO2-Cu-5, while gave a relatively high and stable conversion of butanal. The amount of carbon
accumulated on the catalysts was calculated from the results of TG by comparing the weight loss in
used catalysts with fresh one. Because Ag2O was reduced to Ag in the reactions in H2 flow, Ag in
the used catalyst would be oxidized to Ag2O during the TG analysis. In a similar way, Cu would be
oxidized to CuO during the TG analysis. Table 2 shows the corrected value of the amount of
accumulated carbon after considering the increment of mass caused by the oxidation of Ag and Cu.
The amount of carbon accumulated on TiO2-Ag-3 was 2.2 wt.%, which was the lowest value
among all the tested catalysts. The reaction over TiO2-Ag-3 in an N2 carrier gas was also performed
and the reaction results are shown in Table 2 and Fig 5d. The reaction over TiO2-Ag-3 in N2 flow
gave much lower butanal conversion and lower selectivity to the aldol condensation products, but
higher amount of accumulated carbon than that in H2 flow. Fig. 5a and 5b shows the effects of metal
loading in Nb2O5 and SiO2-Al2O3, respectively. No significant differences were observed over 3
wt.% Ag2O- and 5 wt.% CuO-modified Nb2O5 comparing with Nb2O5 without modification. The
loading of Ag onto SiO2-Al2O3 decreased both the conversion of butanal and the selectivity to the
aldol condensation products. The loading of Cu onto SiO2-Al2O3 significantly increased the
conversion of butanal, but significantly decreased the selectivity to the self-aldol condensation
products. Butanal hydrogenation into 1-butanol proceeded over SiO2-Al2O3-Cu-5 in H2 flow, and
the selectivity to 1-butanol was 55%.
Fig. 6 shows the TPR profiles of Ag-modified TiO2 at different Ag loadings. Two reduction
peaks at ca. 100 and 180 oC were observed, which indicated that Ag could exist as the form of
Ag2O. The intensity of the peaks increased with increasing the loading of Ag2O. Fig. 7 shows the
XRD profiles of Ag-modified TiO2 at different Ag loadings after the pretreatment in H2 flow. TiO2
without Ag loading showed typical anatase TiO2 peaks [JCPDSfile 4-0477]. No diffraction peak
attributed to Ag was observed at Ag loadings of 5 wt.% and lower, which indicated that Ag was
11

highly dispersed on TiO2 surface. A diffraction peak attributed to Ag was observed at 2θ= 38.4o
[JCPDSfile 4-0783] in the diffraction profile of TiO2-Ag-10.
Table 3 and Fig. 8 show the reaction results of butanal self-aldol condensation over TiO2
with different Ag loadings in H2 flow at 220 oC. The conversion of butanal increased with increasing
the loading of Ag, whereas the selectivity to 2E2H decreased. Both the conversion of butanal and the
selectivity to the aldol condensation products decreased with time on stream over TiO2 and TiO2-
Ag-1. On the other hand, the catalytic activity was stable when the loading of Ag2O was higher than
3 wt.%. Fig. 9 shows the TG profiles of the Ag-modified TiO2 catalysts after used in the catalytic
reaction. The increment of mass at ca. 200 oC and 350-500 oC in the TG profiles of TiO2-Ag-3,
TiO2-Ag-5 and TiO2-Ag-10 was caused by the oxidation of Ag. Table 3 summarizes the amount of
carbon accumulated in each catalyst after compensation of the oxidation of Ag. The amount of
accumulated carbon decreased with increasing the loading of Ag. Fig. 10 shows the DRIFT spectra
of TiO2-Ag-3 before and after the reactions. The reaction over TiO2-Ag-3 was performed in H2 flow
-1
at 200 oC and a flow rate of 20 cm3 min-1. The peaks at 1697, 2877, 2935 and 2964 cm were
observed in the TiO2-Ag-3 used in the reaction, but they were not observed in the fresh one. The
-1
peaks at 1697 and 2964 cm are assigned to the stretching vibration of C=O [32] and CH3 [32],
-1
respectively. The peaks at 2877 and 2935 cm are attributed to the stretching vibration of CH2 [33].
Table 4 and Fig. 11 show the effect of the flow rate of H2 over TiO2-Ag-5 at 200 oC. At a
low H2 flow rate of 20 cm3 min-1, although the highest selectivity to 2E2H was achieved, the
conversion of butanal was low together with rapid deactivation with time on stream. The conversion
of butanal was relatively stable at H2 flow rates of 40 cm3 min-1 and higher, and was maximized at a
H2 flow rate of 40 cm3 min-1. The amount of accumulated carbon decreased with increasing the flow
rate of H2, but the selectivity to 2EH increased at high H2 partial pressure, which accelerated the
hydrogenation of 2E2H into 2EH.
4. Discussion
12

4.1. The effects of properties of catalyst and reaction conditions on self-aldol condensation of
butanal.
According to the prior study, vapor-phase reaction of butanal proceeds through two types of
reactions: aldol condensation and Tishchenko esterification [15]. It has been reported that both aldol
condensation and Tishchenko esterification proceed over acid, base or acid-base bifunctional
catalysts [15]. Thus, Tishchenko esterification always competes with aldol condensation. In this
study, BA, BuOH and a Tishchenko esterification product of BB are detected as the major by-
products. We propose that BA and BuOH are the further hydrolysis products of BB, which agrees
with the proposition of Shen et al [25]. On the other hand, BA and BuOH are also possible to be
produced by the direct oxidation and hydrogenation, respectively, of butanal with small amounts
according to the property of the catalysts. Among all the tested catalysts, Nb2O5, SiO2-Al2O3 and
TiO2 show the relatively high conversion of butanal and high selectivity to aldol condensation
products (Table 1). Based on the results of NH3- and CO2-TPD analysis (Fig. 1), it is clear that
Nb2O5, SiO2-Al2O3 or TiO2 works as an acid catalyst. In many dehydration processes, such as the
dehydration of 1,2-propanediol into propanal, strong acid catalysts always show high initial catalytic
activity [28]. However, in this study, the order of catalytic activity does not agree with the order of
the acid strength. SiO2-Al2O3 shows the strongest acid strength, but gave the lowest initial
conversion of butanal, which indicates that strong acidity is not preferable for butanal condensation.
Shylesh et al. performed self-aldol condensation of butanal using SiO2-supported amine catalysts
prepared by an impregnation method, and concluded that weak acid sites were important for the
formation of 2E2H [34]. Thus, we propose that the catalyst with weak and medium acid strength is
appropriate for the formation of aldol condensation products. Because all the catalysts deactivated
rapidly with time on steam, here, we evaluate the catalytic activity by comparing the initial formation
-
rates of 2E2H per unit surface area: the formation rates of 2E2H are 0.185, 0.042 and 0.007 mmol h
-2
1 m for TiO2, Nb2O5 and SiO2-Al2O3, respectively, which were calculated from the data in Table 1.
The order of catalytic activity only agrees with the order of acid density. Thus, the relatively high
catalytic activity of TiO2 is not only attributed to its appropriate acid strength, but also attributed its
13

high acid density. The initial formation rates of 2E2H per unit acid amount are 162, 45.3 and 16.4
mol molacid-1 h-1 for TiO2, Nb2O5 and SiO2-Al2O3, respectively. This clearly indicates that a turnover
frequency based on an acid site of TiO2 is much higher than those of Nb2O5 and SiO2-Al2O3.
Nb2O5 and SiO2-Al2O3 show similar results in H2 and N2 flow, whereas TiO2 shows higher
conversion of butanal and higher selectivity to aldol condensation products in H2 flow than those in
N2 flow (Fig. 3). In a recent report of our group, the dehydration of 1,2-propanediol into propanal
was studied over SiO2-supported WO3 catalyst, which also shows better catalytic performance in H2
flow than in N2 flow: the averaged conversion of 1,2-propanediol in H2 and N2 flow are 99.9% and
87.1%, respectively, at 370 oC [35]. Although the effect of H2 is not well understood, it is reasonable
that H2 can be adsorbed on the surface of some solid acid catalysts and participates in the reaction.
On the other hand, TiO2 still deactivates rapidly in H2 flow, and further operation is necessary for
inhibiting the catalytic deactivation. Reaction temperature also significantly affects the stability of
TiO2: the conversion of butanal conversion decreases rapidly with time on stream at high reaction
temperatures. A similar result is also reported in self-aldol condensation of butanal over Pd/K/Zeolite
catalyst [21]. It is considered that high reaction temperatures could accelerate the proceeding of
oligomerization of butanal, which causes the accumulation of carbon on the catalyst surface. The
accumulated carbon would poison the active acid sites and lead to the catalytic deactivation. We
regenerated a used TiO2 catalyst (entry 1 in Table 3), which gave a butanal conversion decreased
from 63.7 to 8.8% during 5 h as shown in Fig. 8a, at 500 oC for 3 h. The regenerated TiO2 gave a
conversion of 44.6% with a 2E2H selectivity of 84.3% at the initial 1 h of the reaction, which
demonstrated that the deactivation is mainly due to carbon deposition.
4.2. The effects of metal loading on solid acidic catalysts on self-aldol condensation of butanal.
Catalytic deactivation in vapor-phase aldol condensation has attracted much attention, and
some studies dealing with this issue has been reported [7,25]. Rode et al. performed vapor-phase
self-condensation of butanal over Cs-Na/zeolite catalyst at 150 oC [7]. They have found that the
pretreatment of Cs-Na/zeolite with propylene can block the strong Lewis acid sites and stabilize the
14

catalytic activity in the initial 12 h of the reaction. Although the details of conversion and selectivity
are not reported, the formation rate of 2E2H over propylene-pretreated Cs-Na/zeolite is 0.20 mmol
g-1 h-1, which is much lower than that of 12.1 mmol g-1 h-1 over TiO2-Ag-3 at 220 oC in H2 flow in
the present study (Table 3, entry 3). Shen et al. investigated the vapor-phase aldol condensation of
butanal over SiO2-supported alkaline earth metal oxide catalysts, such as MgO/SiO2 and SrO/SiO2
[25]. The conversion could be maintained at ca. 40% with a 2E2H selectivity of ca. 50% and a
2E2H formation rate of ca. 13 mmol g-1 h-1 in the initial 2 h over MgO/SiO2 at 400 oC. On the other
hand, the catalysts deactivate at temperatures lower than 300 oC, whereas the catalytic activity is
stable at 400 oC. In order to study the deactivation mechanism, they prepared samples adsorbed BA
and measured BA desorbed from the samples by TPD analysis. Temperatures higher than 350 oC
are found to be required for the decomposition of BA on the catalyst surface. Thus, it is concluded
that the catalytic deactivation is due to the poisoning of BA, which is generated as a by-product
during the aldol condensation. The alkaline earth metal oxide-supported catalysts are acid-base
bifunctional catalysts, and BA is reasonable to adsorb and poison the base sites. In the present study,
although BA is also a by-product, it would not be the substance leading to the catalytic deactivation
because no basic sites were observed in TiO2, Nb2O5 or SiO2-Al2O3. DRIFT spectra indicate that the
accumulated carbonaceous compounds contain CH2, CH3 and C=O groups (Fig. 10). Although the
structure of the accumulated carbon is not confirmed, the possible component is considered to be the
oligomerization products, which would be produced by the aldol condensation, Tishchenko
esterification or ketonization of aldehydes and carboxylic acids.
In our group, several studies dealing with the catalytic deactivation have been reported [26-
29], and it is found that the modification of acid catalysts by loading metal contents is effective for
inhibiting the catalytic deactivation. In particular, Cu-modified Al2O3 exhibits stable catalytic activity
with high selectivity to 3,4-2H-dihydropyran in the dehydration of tetrahydrofurfuryl alcohol [26];
Co-modified Al2O3 stabilizes the conversion of pinacolone to produce 2,3-dimethyl-1,3-butadiene
[27]; Ag-modified SiO2-Al2O3 gives stable and increased selectivity to propanal in 1,2-propanediol
dehydration [28], the loading of Ag onto Al2O3 inhibits carbon deposition and stabilize the catalytic
15

activity in the dehydration of diethylene glycol into 1,4-dioxane [29]. All the above-mentioned
catalysts were effective only under H2 flow conditions. It is well known that both Cu and Ag has the
ability to adsorb with H2 and work as hydrogenation catalysts. Cu has relatively high hydrogenation
activity and has been widely used in many hydrogenation processes, such as levulinic acid
hydrogenation to -valerolactone [36] and glycerol hydrogenolysis to 1,2-propanediol [37-38] and
propylene [39]. On the other hand, because of the relatively low hydrogenation activity of Ag, Ag is
generally used in partial hydrogenation processes such as the hydrogenation of acrolein to allyl
alcohol [40]. In this study, Ag as a promoter showed a better performance than Cu, and is only
effective in H2 flow (Fig. 5). Because of the high hydrogenation activity of Cu, Cu-modified SiO2-
Al2O3 even catalyzes the hydrogenation of butanal to BuOH, while aldol condensation of butanal
almost dose not proceed.
TPR profiles of Ag-modified TiO2 show that the reduction temperature for Ag2O is lower
than 200 oC (Fig. 6). Because Ag-modified TiO2 catalysts are pretreated at 250 oC in H2 flow, Ag2O
must be reduced to Ag before the reaction. XRD profiles have proved it because the diffraction peak
of Ag is observed in pretreated TiO2-Ag-10 (Fig. 7). Thus, it is clear that Ag works as a metal to
inhibit the catalytic deactivation during the reaction. The catalytic activity is stable at Ag loadings
higher than 3wt.% (Fig. 8) and the amount of accumulated carbon decreased with increasing the Ag
loading (Table 3), which indicates that the modification of Ag is effective for inhibiting the carbon
accumulation and stabilizing the catalytic activity. On the other hand, high loadings of Ag decreases
the selectivity to aldol condensation products, which is possibly because Ag metal also provides
active sites to catalyze the side reaction. Thus, the suitable loading of Ag2O is considered to be 3-5
wt.%. H2 plays an important role in such a process for inhibiting the catalytic deactivation. A high
flow rate of H2 is effective for decreasing the amount of accumulated carbon and stabilizing the
catalytic activity (Table 4 and Fig. 11). The role of the additive Ag in TiO2 in H2 flow is considered
to be the same as those of the additive Cu and Ag reported in our previous researches [26-29]. We
have concluded that the additive metals in solid acid catalysts could adsorb with the strong acid sites
and work as a remover of the product from the catalyst surface together with H2 to prevent carbon
16

deposition before dehydrogenation of adsobates [26-29]. In particular, it is proposed that H2 is firstly
adsorbed on Ag metal and cleaved to H atoms, and then the H atoms shift from Ag to the surface of
TiO2 to form H+ that can be explained as spillover effect (Scheme 2). Over pure TiO2 surface,
carbonaceous products are accumulated through dehydrogenation. On theother hand, the generated
H+ must have extremely weak acidity and is considered to give little contribution to the dehydration
of butanal to 2E2H. However, the existence of the large amount of H+ on TiO2 surface would inhibit
the polymerization of the products such as 2E2H because the formation of carbonaceous
compounds is generally through a dehydrogenation process. The H adsorbed on the surface of TiO2
assists desorption of the products from the catalyst surface before the oligomerization of the products
proceeds. The amount of accumulated carbon is effectively inhibited at high loadings of Ag and high
flow rates of H2 because that would increase the amount of H on TiO2 surface. It is probable that
coke accumulation could be inhibited under such as a mechanism, and stable catalytic activity can be
achieved.
5. Conclusions
Vapor-phase self-aldol condensation of butanal was performed over various solid catalysts
and metal-modified solid catalysts. Among the tested catalysts without metal loading, SiO2-Al2O3,
Nb2O5 and TiO2 showed relatively high catalytic activity for aldol condensation products formation,
whereas all the catalysts deactivated rapidly. The order of catalytic activity is TiO2> Nb2O5> SiO2-
Al2O3, which agrees with the order of the acid density of the catalysts. Metal-modified catalysts were
prepared for stabilizing the catalytic activity, and Ag-modified TiO2 showed the best catalytic
performance on the stabilization. The XRD and TPR profiles indicate that Ag2O is reduced to Ag
before the reaction and works as a metal during the reaction. The loading of Ag onto TiO2 inhibited
the amount of carbon accumulated on catalyst surface, and H2 as a carrier gas was indispensable for
the stabilization. It is considered that Ag works as a remover of the product from the catalyst surface
together with H2 to prevent coke formation. DRIFT spectra showed that the accumulated carbon
contains CH2, CH3 and C=O groups. It is speculated that the possible components of accumulated
17

carbon are the oligomers of butanal. Self-aldol condensation of butanal was stabilized over Ag-
modified TiO2 at Ag2O loadings higher than 3 wt.% at 220 oC in H2 flow. High loadings of Ag into
TiO2 inhibited carbon accumulation, whereas decreases the selectivity to aldol condensation
products. A stable butanal conversion of 72.1% with a 72.2% selectivity to 2-ethyl-2-hexenal was
achieved over 5 wt.% Ag2O-modified TiO2 in H2 flow at 220 oC.
18

References
[1] R.N. Hayes, R.P. Grese, M.L. Gross, J. Am. Chem. Soc. 111 (1989) 8336-8341.
[2] J.E. Rekoske, M.A. Barteau, Ind. Eng. Chem. Res. 50 (2011) 41-51.
[3] D.A. Simonetti, J.A. Dumesic, ChemSusChem 1 (2008) 725-733.
[4] E.L. Kunkes, D.A. Simonetti, R.M. West, J.C. Serrano-Ruiz, C.A. Gartner, J.A. Dumesic,
Science 322 (2008) 417-421.
[5] E.I. Gurbuz, E.L. Kunkes, J.A. Dumesic, Green Chem. 12 (2010) 223-227.
[6] L.M. Baigrie, R.A. Cox, H. Slebocka-Tilk, M. Tencer, T.T. Tidwell, J. Am. Chem. Soc. 107
(1985) 3640-3645.
[7] E.J. Rode, P.E. Gee, L.N. Marquez, T. Uemura, M. Bazargani, Catal. Lett. 9 (1991) 103-114.
[8] W. Ji, Y. Chen, H.H. Kung, Appl. Catal. A: Gen. 161 (1997) 93-104.
[9] V. Serra-Holm, T. Salmi, J. Multamäki, J. Reinik, P. Mäki-Arvela, R. Sjöholm, L.P. Lindfors,
Appl. Catal. A: Gen. 198 (2000) 207-221.
[10] H. Naka, Y. Kaneda, T. Kurata, J. Oleo Sci. 50 (2001) 813-821.
[11] E. Dumitriu, V. Hulea, I. Fechete, A. Auroux, J.-F. Lacaze, C. Guimon, Micropor. Mesopor.
Mater. 43 (2001) 341-349.
[12] M. Paulis, M. Martín, D.B. Soria, A. Díaz, J.A. Odriozola, M. Montes, Appl. Catal. A: Gen.
180 (1999) 411-420.
[13] H. Idriss, K.S. Kim, M.A. Barteau, J. Catal. 139 (1993) 119-133.
[14] S. Luo, J.L. Falconer, J. Catal. 185 (1999) 393-407.
[15] H. Tsuji, F. Yagi, H. Hattori, H. Kita, J. Catal. 148 (1994) 759-770.
[16] F. King, G.J. Kelly, Catal. Today 73 (2002) 75-81.
[17] S.K. Sharma, P.A. Parikh, R.V. Jasra, J. Mol. Catal. A: Chem. 278 (2007) 135-144.
[18] Y.-C. Chang, A.-N. Ko, Appl. Catal. A: Gen. 190 (2000) 149-155.
[19] J.I. Di Cosimo, C.R. Apesteguía, J. Mol. Catal. A: Chem. 130 (1998) 177-185.
[20] P. Moggi, G. Albanesi, Appl. Catal. 68 (1991) 285-300.
[21] A.-N. Ko, C.H. Hu, J. Chen, Appl. Catal. A: Gen. 184 (1999) 211-217.
19

[22] G. Chang, Z. Bao, Z. Zhang, H. Xing, B. Su, Y. Yang, Q. Ren, J. Colloid Interface Sci. 412
(2013) 7-12.
[23] R. Both, A. Cormos, P. Agachi, C. Festila, Comp. Chem. Eng. 52 (2013) 100-111.
[24] C.A. Hamilton, S.D. Jackson, G.J. Kelly, Appl. Catal. A: Gen. 263 (2004) 63-70.
[25] W. Shen, G.A. Tompsett, R. Xing, W.C. Conner Jr., G.W. Huber, J. Catal. 286 (2012) 248-259.
[26] S. Sato, J. Igarashi, Y. Yamada, Appl. Catal. A: Gen. 453 (2013) 213-218.
[27] S. Sato, N. Sato, Y. Yamada, Chem. Lett. 41 (2012) 831-833.
[28] D. Sun, R. Narita, F. Sato, Y. Yamada, S. Sato, Chem. Lett. 43 (2014) 450-452.
[29] D. Sun, J. Wang, Y. Yamada, S. Sato, Appl. Catal. A: Gen. 505 (2015) 422-430.
[30] S. Sato, R. Takahashi, T. Sodesawa, A. Igarashi, H. Inoue, Appl. Catal. A: Gen. 328 (2007)
109-116.
[31] T. Nakayama, N. Ichikuni, S. Sato, F. Nozaki, Appl. Catal. A: Gen. 158 (1997) 185-199.
[32] H.-K. Jeong, H.-J. Noh, J.-Y. Kim, M.H. Jin, C.Y. Park, Y.H. Lee, Europhys. Lett. 82 (2008)
67004.
[33] H. Sun, Y. Yang, Q. Huang, Integr. Ferroelectr. 128 (2011) 163-170.
[34] S. Shylesh, D. Hanna, J. Gomes, S. Krishna, C.G. Canlas, M. Head-Gordon, A.T. Bell,
ChemCatChem 6 (2014) 1283-1290.
[35] D. Sun, Y. Yamada, S. Sato, Appl. Catal. A: Gen. 487 (2014) 234-241.
[36] P.P. Uprare, J. Lee, Y.K. Hwang, D.W. Hwang, J. Lee, S.B. Halligudi, J.S. Hwang, J. Chang,
ChemSusChem 4 (2011) 1749-1752.
[37] M. Akiyama, S. Sato, R. Takahashi, K. Inui, M. Yokota, Appl. Catal. A: Gen. 371 (2009) 60-
66.
[38] D. Sun, Y. Yamada, S. Sato, Appl. Catal. A: Gen. 47 (2014) 63-68.
[39] D. Sun, Y. Yamada, S. Sato, Appl. Catal. B: Environ. 174 (2015) 13-20.
[40] H. Wei, C. Gomez, J. Liu, N. Guo, T. Wu, R. Lobo-Lapidus, C.L. Marshall, J.T. Miller, R.J.
Meyer, J. Catal. 298 (2013) 18-26.
20

O
OH
O
-H2O
+H2
O
O
O
Aldol consendation
Butanal
2-Ethyl-3-hydroxy-hexanal
O
2-Ethylhexanal
2-Ethyl-2-hexenal
O
O
+H2O
+
OH
CH3
O
OH
Butyl butyrate
Butanoic acid
1-Butanol
Scheme 1 Proposed formation routes of each product.
(2-Ethyl-3-hydroxy-hexanal was not detected in the reaction)
Scheme 2 Proposed mechanism of the inhibition of coke formation over Ag-modified TiO2.
21

Table 1Aldol condensation of butanal over various catalysts at 200 oC in N2 flow of 20 cm3 min-1.
Catalyst
Conversiona Selectivity /%a
/ %
48.5
29.3
18.9
10.1
7.4
2E2H
72.7
70.5
81.7
76.5
72.1
38.1
26.1
76.6
46.3
2EH
1.0
BA
3.3
BuOH
1.9
BB
4.4
Others
16.7
21.7
12.7
9.7
b
Nb2O5
b
SiO2-Al2O3
2.0
0.2
2.5
3.2
c
TiO2
0.1
3.5
0.2
1.9
c
MgO/SiO2
0.1
10.8
0.9
2.0
b
Al2O3
0.2
0.0d 9.3
29.0 14.1
0.0d 10.5
2.0
16.4
11.8
53.0
5.2
b
ZrO2
4.6
0.1
6.9
b
CeO2
4.1
0.0d
0.0d
0.0 d
10.4
0.0d
5.4
c
Ca10(PO4)6(OH)2
MFI zeolitec
3.6
18.2
14.9
0.0d
2.9
1.4
32.0
aAveraged activity in the initial 5 h. b Catalyst weight, 2.0 g. c Catalyst weight, 1.0 g. dThe value is
lower than 0.05%. (2E2H, 2-ethyl-2-hexenal; 2EH, 2-ethylhexanal; BA, butanoic acid; BuOH, 1-
butanol; BB, butyl butyrate)
Table 2Aldol condensation of butanal over metal modified TiO2 catalysts a.
Catalyst
Carbon Conv. b Selectivity /% b
/wt. %
3.5
/%
28.5
33.9
44.6
7.7
2E2H
86.9
86.7
82.5
54.6
2EH
0.0d
BA
2.4
0.0d
0.2
BuOH
0.0d
0.8
BB Others
TiO2
2.2
2.3
2.3
0.4
8.4
9.8
TiO2-Cu-5
TiO2-Ag-3
TiO2-Ag-3 c
3.1
0.4
0.5
2.2
0.0d
0.2
14.5
32.3
3.5
0.0d 12.6
a Reaction conditions: reaction temperature, 200 oC; flow rate of H2, 20 cm3 min-1; catalyst weight, 1
g. b Averaged activity in the initial 5 h. c The reaction was performed in N2 flow of 20 cm3 min-1. dThe
value is lower than 0.05%. (2E2H, 2-ethyl-2-hexenal; 2EH, 2-ethylhexanal; BA, butanoic acid;
BuOH, 1-butanol; BB, butyl butyrate)
22

Table 3Aldol condensation of butanal overAg modified TiO2 at differentAg loadings a.
Catalyst
Carbon Conv. b Selectivity /% b
/wt. %
3.9
/%
2E2H
81.7
80.0
77.1
72.2
70.0
2EH
BA
3.6
BuOH
0.5
BB Others
TiO2
26.0
27.4
70.9
72.1
72.4
0.1
2.7
2.8
2.5
2.5
2.8
11.4
11.7
17.5
23.3
22.0
TiO2-Ag-1
TiO2-Ag-3
TiO2-Ag-5
TiO2-Ag-10
2.9
0.0c 5.3
0.3
2.1
0.6
1.0
3.4
2.1
2.1
1.4
0.1
1.4
0.0c
0.4
1.4
a Reaction conditions: reaction temperature, 220 oC; flow rate of H2, 20 cm3 min-1; catalyst weight, 1
g. b Averaged activity in the initial 5 h. cThe value is lower than 0.05%.
(2E2H, 2-ethyl-2-hexenal; 2EH, 2-ethylhexanal; BA, butanoic acid; BuOH, 1-butanol; BB, butyl
butyrate)
Table 4Aldol condensation of butanal over TiO2-Ag-5 at different H2 flow rates a.
Flow rate of H2 Carbon Conv. b Selectivity /% b
/ cm3 min-1
/wt. %
2.2
/%
2E2H
84.8
75.0
76.1
2EH
0.0c
BA
3.4
BuOH
0.0c
0.4
BB Others
20
44.0
66.8
62.0
2.3
2.0
2.2
9.5
21.3
18.1
40
1.5
1.4
3.0
0.0c
0.1
100
1.3
0.5
a Reaction conditions: reaction temperature, 200 oC; catalyst weight, 1 g. b Averaged activity in the
initial 5 h. cThe value is lower than 0.05%.
(2E2H, 2-ethyl-2-hexenal; 2EH, 2-ethylhexanal; BA, butanoic acid; BuOH, 1-butanol; BB, butyl
butyrate)
23

Figure captions
Fig. 1 NH3-TPD (a), NH3-TPD per surface area (b) and CO2-TPD (c) profiles of SiO2-Al2O3, Nb2O5
and TiO2.
Fig. 2 Changes of conversion and the selectivity with time on stream over Nb2O5, SiO2-Al2O3 and
TiO2.
Closed symbols, conversion of butanal; open symbols, selectivity to the aldol condensation products
of 2-ethyl-2-hexenal and 2-ethylhexanal. (Reaction conditions are same as shown in Table 1)
Fig. 3 Comparison of N2 and H2 flow in self-aldol condensation of butanal over Nb2O5 (a), SiO2-
Al2O3 (b) and TiO2 (c).
Closed symbols, conversion of butanal; open symbols, selectivity to the aldol condensation products
of 2-ethyl-2-hexenal and 2-ethylhexanal. (Reaction conditions: reaction temperature, 200 oC; catalyst
weight, 1 g; flow rate of H2 or N2, 20 cm3 min -1)
Fig. 4 Aldol condensation of butanal over TiO2 in H2 flow at different temperatures.
Closed symbols, conversion of butanal; open symbols, selectivity to the aldol condensation products
of 2-ethyl-2-hexenal and 2-ethylhexanal. (Reaction conditions: catalyst weight, 1 g; flow rate of H2,
20 cm3 min -1)
Fig. 5 Self-aldol condensation of butanal over metal modified catalysts in H2 flow.
Closed symbols, conversion of butanal; open symbols, selectivity to the aldol condensation products
of 2-ethyl-2-hexenal and 2-ethylhexanal. (Reaction conditions: reaction temperature, 200 oC; catalyst
weight, 1 g; flow rate of H2 or N2, 20 cm3 min -1)
Fig. 6 TPR analysis of TiO2 at different Ag2O loadings.
24

Fig. 7 XRD profiles of TiO2 at different Ag loadings.
Fig. 8 Aldol condensation of butanal over Ag modified TiO2 at different Ag loadings.
Closed symbols, conversion of butanal; open symbols, selectivity to the aldol condensation products
of 2-ethyl-2-hexenal and 2-ethylhexanal. (Reaction conditions: reaction temperature, 220 oC; catalyst
weight, 1 g; flow rate of H2, 20 cm3 min -1)
Fig. 9 TG profiles of the used Ag-modified TiO2 catalysts at different Ag loadings.
(The reaction conditions are the same as shown in Table 3)
Fig. 10 DRIFT spectra of fresh and used TiO2-Ag-3.
(The reaction conditions of used TiO2-Ag-3 was as follows: reaction temperature, 200 oC, flow rate
of H2, 20 cm3 min-1)
Fig. 11 Aldol condensation of butanal over TiO2-Ag-5 at different H2 flow rates.
Closed symbols, conversion of butanal; open symbols, selectivity to the aldol condensation products
o
of 2-ethyl-2-hexenal and 2- ethylhexanal. (Reaction conditions: reaction temperature, 200 C;
catalyst weight, 1 g)
25

(a)
SiO2-Al2O3
Nb2O5
TiO2
TiO2 without NH3 adsoption
200
400
600
800
Temperature / °C
(b)
SiO2-Al2O3
Nb2O5
TiO2
TiO2 without NH3 adsoption
200
400
600
800
Temperature / °C
(c)
SiO2-Al2O3
Nb2O5
TiO2
600
200
400
800
Temperature / °C
Fig. 1 NH3-TPD (a), NH3-TPD per surface area (b) and CO2-TPD (c) profiles of SiO2-Al2O3, Nb2O5
and TiO2.
26

100
80
SiO2-Al2O3
Nb2O5
TiO2
60
40
20
0
1
2
3
4
5
Time on stream / h
Fig. 2 Changes of conversion and the selectivity with time on stream over Nb2O5, SiO2-Al2O3 and
TiO2. Closed symbols, conversion of butanal; open symbols, selectivity to the aldol condensation
products of 2-ethyl-2-hexenal and 2-ethylhexanal. (Reaction conditions are same as shown in Table
1)
27

100
80
100
80
60
40
20
100
80
60
40
20
(a)
(c)
(b)
60
N2
H2
40
N2
H2
20
N2
H2
0
1
2
3
4
5
0
1
2
3
4
5
0
1
2
3
4
5
Time on stream / h
Time on stream / h
Time on stream / h
Fig. 3 Comparison of N2 and H2 flow in self-aldol condensation of butanl over Nb2O5 (a), SiO2-
Al2O3 (b) and TiO2 (c). Closed symbols, conversion of butanal; open symbols, selectivity to the aldol
condensation products of 2-ethyl-2-hexenal and 2-ethylhexanal. (Reaction conditions: reaction
temperature, 200 oC; catalyst weight, 1 g; flow rate of H2 or N2, 20 cm3 min -1)
28

100
80
60
40
20
240℃
200℃
180℃
160℃
0
1
2
3
4
5
Time on stream / h
Fig. 4 Aldol condensation of butanal over TiO2 in H2 flow at different temperatures.
Closed symbols, conversion of butanal; open symbols, selectivity to the aldol condensation products
of 2-ethyl-2-hexenal and 2-ethylhexanal. (Reaction conditions: catalyst weight, 1 g; flow rate of H2,
20 cm3 min -1)
29