Etherification of aldehydes, alcohols and their mixtures on Pd/SiO2 catalysts

Article abstract of DOI:10.1016/j.apcata.2010.03.014

Dialkyl ethers have been selectively produced from etherification of aldehydes and alcohols on supported Pd catalysts. A yield of 79% ether with a selectivity of 90% was observed when feeding 2-methylpentanal with 2-methylpentanol at a molar ratio 1:1 at 125 °C. Cross etherification of n-butanol with 2-methylpentanal shows a much higher rate than that observed when the alcohol or aldehyde is fed alone. This enhanced activity is in line with the catalyst requirement for large ensembles that allow surface alkoxide species next to η² adsorbed aldehydes. Etherification when only aldehyde or alcohol is fed arises predominantly due to aldehyde-alcohol inter-conversion to produce the necessary co-reactant. The ether yield at the same reaction conditions decreases with metal loading in the order 16 > 10 > 3 wt.% Pd. Increasing reduction temperature also increases ether yield. It is apparent that etherification is highly sensitive to metal particle morphology, consistent with needing ensembles that accommodate the two adjacent adsorption sites.

Full text of DOI:10.1016/j.apcata.2010.03.014

Applied Catalysis A: General 379 (2010) 135–140
Contents lists available at ScienceDirect
Applied Catalysis A: General
journal homepage: www.elsevier.com/locate/apcata
Etherification of aldehydes, alcohols and their mixtures on Pd/SiO catalysts
2
Trung T. Pham, Steven. P. Crossley, Tawan Sooknoi, Lance L. Lobban,
∗
Daniel E. Resasco, Richard G. Mallinson
School of Chemical, Biological and Materials Engineering, University of Oklahoma, 100 Boyd St, T-335, Norman, OK 7301, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
Dialkyl ethers have been selectively produced from etherification of aldehydes and alcohols on sup-
Received 10 December 2009
Received in revised form 2 March 2010
Accepted 7 March 2010
ported Pd catalysts. A yield of 79% ether with a selectivity of 90% was observed when feeding
◦
2
-methylpentanal with 2-methylpentanol at a molar ratio 1:1 at 125 C. Cross etherification of n-butanol
with 2-methylpentanal shows a much higher rate than that observed when the alcohol or aldehyde is fed
alone. This enhanced activity is in line with the catalyst requirement for large ensembles that allow sur-
Available online 12 March 2010
2
face alkoxide species next to Á adsorbed aldehydes. Etherification when only aldehyde or alcohol is fed
Keywords:
arises predominantly due to aldehyde–alcohol inter-conversion to produce the necessary co-reactant.
The ether yield at the same reaction conditions decreases with metal loading in the order 16 > 10 > 3 wt.%
Pd. Increasing reduction temperature also increases ether yield. It is apparent that etherification is highly
sensitive to metal particle morphology, consistent with needing ensembles that accommodate the two
adjacent adsorption sites.
Etherification
Hydrogenation
Decarbonylation
Aldehyde
Alcohol
©
2010 Elsevier B.V. All rights reserved.
1
. Introduction
Alcohol etherification on acid catalysts is a well-established
reaction [5,9–12]. Much less attention has been received to the
study of ether formation on metal catalysts [13–18]. While studies
focusing on etherification of alcohol and aldehyde have identified
the need for ensembles of sites to accomplish the condensation
[14–16], the influence of metal surface structure on ether selectivity
has not been demonstrated in a systematic way.
In recent work [19], we have shown that 2-methylpentanal
may be selectively produced from hydrogenation of 2-methyl-2-
pentenal, itself produced from aldol condensation of glycerol or
from propanal. These compounds are representative of types of
bio-derived aldehydes and analogous alcohols from processes such
as biomass pyrolysis. The present contribution investigates the
etherification of 2-methylpentanal to form di-2-methylpentylether
(DMPE) on Pd catalysts of varying particle size and morphology
to help understanding the pathways of etherification of smaller
bio-derived oxygenates by metal catalysts.
Biofuel production has attracted substantial attention in recent
years [1–4]. There is strong motivation to produce fuels from
oxygenated molecules generated from biomass conversion. Much
research has been focused toward production of fungible molecules
that may be readily incorporated into the existing transportation
fuel infrastructure. Ethers are one of the potential fuel components
made from oxygenates that might be used either in gasoline or
diesel blends. Ethers were first commercially introduced in Italy in
1
973 as a blending component in gasoline [5]. More recently, Olah
[
6] patented the use of C –C alkyl ethers as blending components
2
24
for diesel fuels, demonstrating cetane number improvement from
to 20 points. Use of short alkyl ethers such as methyl-t-butylether
MTBE), methyl-t-amylether (TAME) and ethyl-t-butylether (ETBE)
2
(
as fuel additives have had negative environmental impact due to
their high solubility in water that enhances their dispersion in the
environment. These problems are much less severe with larger
alkyl ethers, which exhibit much lower solubility. However, despite
their excellent combustion and physical properties [7,8], overcom-
ing perceived environmental issues will be required before ethers
can be added to the fuel pool.
2. Experimental
2.1. Catalyst synthesis
Different loadings (3, 10 and 16 wt.%) of Pd catalysts were pre-
pared by incipient wetness impregnation of the metal precursor
on precipitated silica (HiSil 210 obtained from Pittsburg Plate Glass
2
∗ Corresponding author. Tel.: +1 4053254378; fax: +1 4053255813.
E-mail address: mallinson@ou.edu (R.G. Mallinson).
Co.) having a BET surface area of 124 m /g. The metal loading on
each catalyst was calculated based on the amount of metal pre-
0
926-860X/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.apcata.2010.03.014

1
36
T.T. Pham et al. / Applied Catalysis A: General 379 (2010) 135–140
cursor incorporated on to the support. The liquid/solid ratio used
to achieve incipient wetness was 1.26 ml/g. The precursor was
Pd(NO ) , purchased from Sigma–Aldrich. After impregnation, the
3
2
catalysts were dried overnight in an oven and then calcined in air
◦
at 400 C for 4 h. To study the effect of alkali addition, two catalysts
were synthesized by adding controlled amounts of K onto the 10%
Pd/SiO2 catalyst.
2.2. Catalyst characterization
Dynamic pulse chemisorption was used to determine the Pd
percent exposed. CO adsorption was conducted in a dynamic
chemisorption unit, consisting of a pulse system with detection by
an SRI Model 310C Gas Chromatograph equipped with a Thermal
◦
Fig. 1. Yields of products from MPAL on 16% Pd/SiO2 reduced at 150 C and run at
◦
125 C, 1 atm, H :MPAL = 12:1.
2
Conductivity Detector (TCD). All of the catalysts were reduced in H
2
◦
◦
by heating from ambient to 150 C with a ramp rate of 10 C/min,
held for 1 h, then flushed with He for 30 min and cooled down to
room temperature. A pre-mixed gas of 5% CO in He was used for
pulse injection with a 0.1 ml loop size. Pulses of CO were sent over
3. Results
3.1. Effect of varying feed composition
◦
the catalyst at 25 C until breakthrough occurred and the size of the
pulse peak no longer increased. The SiO2 support was tested and
no CO adsorption was detected.
3.1.1. Pure aldehyde feed
-Methylpentanal was fed on the 16% Pd catalyst reduced at
2
Temperature Programmed Reduction (TPR) of 3% Pd/SiO shows
◦
2
150 C to study the conversion of the aldehyde. The W/F was var-
ied from 0.25 to 2 h to obtain conversions from 8.4% to 28.9%.
The yields of products are plotted against conversion in Fig. 1.
At low conversion, the major products are n-pentane (C5) and
MPOL that result from decarbonylation and hydrogenation, respec-
tively. As the conversion increases, the yield of the ether (DMPE)
becomes predominant, much higher than the yields of n-pentane
and MPOL. The yield of MPOL decreases, suggesting that it is con-
sumed to form the secondary product, DMPE. No other products
were observed.
◦
one peak centered at 50 C. Therefore, the changes observed after
different reduction temperatures do not reflect different degrees of
reduction, but rather morphological changes.
DRIFTS experiments were carried out in a PerkinElmer Spectrum
1
00 FTIR Spectrometer equipped with an MCT detector and a High
Temperature Reaction Chamber (HVC-DRP) fitted with KBr win-
dows (Harrick Scientific Products, Inc). The catalyst in an amount of
0
.05–0.1 g was loaded to the reaction cell and reduced in H2 flow-
◦
◦
ing at 30 sccm by heating at a 10 C/min ramp rate to 150 C or
2
◦ ◦
50 C or 350 C, then holding for 1 h followed by flushing with He
for 30 min before cooling to room temperature. CO (5% in He) then
flowed through the chamber for 30 min at room temperature and
was then flushed with He for 20 min before sample spectra were
taken. Spectra were acquired at a resolution of 4 cm averaging 64
scans.
3.1.2. Pure alcohol feed
In order to compare the behavior of the aldehyde and alcohol, 2-
methylpentanol was fed over the 16% Pd catalyst reduced at 150 C.
The reactivity of MPOL was found to be considerably lower than
that of MPAL and its conversion increases from 3.6% to 11.9% over
longer W/F, from 1 to 4 h. The product yields are shown in Fig. 2.
The products were DMPE, MPAL, n-pentane, and a small amount of
−
1
◦
2.3. Catalyst tests
2
-methylpentane (2MP). The yield of ether shows a slight induc-
tion period that could suggest that as the aldehyde increases, this
enhances the rate of etherification. Some MPAL is formed from
dehydrogenation while 2MP comes from dehydration of the MPOL,
however the equilibrium greatly favors the alcohol at these con-
ditions [19]. n-Pentane is produced from decarbonylation of the
MPAL
Catalytic activity measurements were carried out in a con-
tinuous flow 0.75 cm I.D. stainless steel tubular reactor within
◦
an electric furnace. Runs were conducted for 2–3 h at 125 C
and atmospheric pressure, using 0.05–0.15 g of catalyst and the
appropriate flow rate to achieve the desired W/F (g cat/g feed/h).
Hydrogen was used to reduce the catalyst and as a carrier gas
for the organic feed that was delivered by a syringe pump
into the gas stream, keeping a relative hydrogen to aldehyde
and/or alcohol molar ratio of 12:1. The reactants used were:
2
-methylpentanal (MPAL), 2-methylpentanol (MPOL), n-butanol
(
BOL), and n-propanal. Experiments co-feeding MPAL + MPOL,
MPAL + BOL, n-propanal + MPOL, and n-propanal + MPAL were also
conducted. The objective of carrying out different combinations
will be discussed later. Heating tapes were used on all the lines
to keep the feed and products in the vapor phase. Quantita-
tive product analysis was made online in an HP 6890A gas
chromatograph equipped with a flame ionization detector. Com-
positions of organic compounds are reported as mole fractions.
A Shimadzu GCMS-QP2010S was used for product identifica-
tion. The light gas products (mostly CO) were detected online
via gas chromatography using a Carle series 400 AGC with
TCD.
◦
Fig. 2. Yields of products from MPOL on 16% Pd/SiO2 reduced at 150 C and run at
◦
125 C, 1 atm, H :MPOL = 12:1.
2

T.T. Pham et al. / Applied Catalysis A: General 379 (2010) 135–140
137
Table 1
Etherification product mole fraction of mixed feeds of aldehydes + alcohols on 16%
◦
◦
Pd/SiO2, reduced at 150 C and run at 125 C; molar H2:feed ratio = 12:1; molar
MPAL:MPOL ratio = 1:1; molar MPAL:n-butanol ratio = 1:1.3; molar propanal:MPOL
ratio = 1:1; molar propanal:MPAL ratio = 1:1.
Feed
Ether product mole fraction
C6
C8
C9
C10
C12
MPAL + MPOL
78.5
15.9
2.28
2.11
MPAL + n-butanol
Propanal + MPOL
Propanal + MPAL
0.02
43.8
0.05
0.09
47.5
11.3
ratio of the two at very low W/F; in this case 0.25 and 0.5 h are used.
This ratio is 5.0 for C10/C12. MPAL reacts readily with BOL to form
the C10 ether. The alcohol is highly reactive for ether formation with
an aldehyde, but has low reactivity for other reactions, whereas the
aldehyde decarbonylates and hydrogenates to some extent, giving
it a higher rate of conversion.
◦
Fig. 3. Co-feed mixtures of MPAL and MPOL on 16% Pd/SiO2 reduced at 150 C and
run at 125 C, W/F = 2 h, H2:feed = 12:1.
◦
3.1.3. Mixtures of 2-methylpentanal with 2-methylpentanol at
different feed ratios
When either the aldehyde or the alcohol is fed alone, the ether
is the predominant product. Because under these reaction condi-
tions a significant extent of inter-conversion of the two molecules
may occur, both species may be present in every run. Therefore, it
is unclear to what extent the etherification reaction depends upon
the concentration of both species. To quantify the contribution of
each one, experiments with mixed feeds at different ratios of MPAL
and MPOL were carried out. The results for W/F = 2 h are shown in
Fig. 3. A much higher ether yield was observed when using the
mixtures than when using the pure feeds of either compound. The
highest ether selectivity (90%) was achieved at a 1:1 feed molar
ratio (yMPAL = 0.5) with a yield, based upon conversion of both reac-
tants, of 79%. These results show that etherification is enhanced
when significant amounts of both MPAL and MPOL are present.
3.1.5. Mixtures of n-propanal and 2-methylpentanol
A similar experiment to study cross etherification was carried
out for a mixed feed of C3AL (n-propanal) with MPOL, at a molar
◦
◦
ratio of 1:1 on 16% Pd/SiO2 reduced at 150 C and run at 125 C with
a W/F = 2 h. Thecross C9 ether yield was 47.5%, much higher than the
C6 and C12 ethers whose yields are 0.05% and 2.28%, respectively.
This result also confirms the high yield of the mixed ether, similar
to the previous experiment when feeding both BOL + MPAL. The
results of ether formation from the mixed feed experiments are
summarized in Table 1.
3
.1.6. Mixtures of n-propanal and 2-methylpentanal
From previous mixed feed experiments, it is clear that high
3
.1.4. Mixtures of n-butanol and 2-methylpentanal
Because both MPAL and MPOL have the same alkyl chain, they
yields of the cross ether are achieved when reacting an aldehyde
and an alcohol. This confirms the hypothesis that etherifica-
tion requires both adsorbed aldehyde and alcohol with abundant
amounts on the surface. However, it remains unclear if etherifica-
tion can proceed to a significant extent without an alcohol in the
system. Therefore, an experiment with a mixed feed of only aldehy-
des in the absence of hydrogenation will help demonstrate whether
this occurs.
Experiments were conducted with n-propanal and 2-
methylpentanal and with either He or H2 as the carrier gas.
In a He carrier, no hydrogenation or ether formation was observed
and the only reaction observed was decarbonylation of the alde-
hydes to hydrocarbons. In a H2 carrier, the observed products
are hydrogenated alcohols from the corresponding aldehydes, as
well as hydrocarbons and ethers. The yields of ethers are much
smaller than feeding mixed alcohol and aldehyde. Yields of C9 and
C12 ethers are 11.3% and 2.11%, respectively. The C6 ether yield is
much smaller than the others, and very little propanol is observed
compared to MPOL. This could indicate either that the propanal is
not effectively hydrogenated and the C9 ether is produced from
propanal and MPOL or that the propanal is readily hydrogenated
and reacts with the MPAL to form the C9 ether.
produce a symmetric ether and the degree to which either aldehyde
or alcohol may be incorporated is not made clear from the above
results. Therefore, n-butanol and 2-methylpentanal were co-fed at
a 1:1 volume ratio (1.3:1 molar ratio), respectively. The composi-
tion of feeds and products are shown in Fig. 4 with increasing W/F.
The results show that both C10 and C12 (DMPE) ethers are the major
products, but only a trace of C8 ether. MPAL is also consumed to a
minor extent by decarbonylation and hydrogenation. The compar-
ative rates for C10 and C12 ether formation are best shown by the
An experiment using only propanal as a feed was conducted
under the same conditions as the other experiments. As shown
in Table 2, when compared at the same conversion level (24%),
much less etherification was observed with propanal than with
methylpentanal, while the hydrogenation to propanol was substan-
tially higher than for the larger aldehyde.
Fig. 4. Cross etherification of n-butanol and 2-methylpentanal on Pd reduced at
1
◦ ◦
50 C and run at 125 C, yMPAL = 0.43. H2:feed ratio = 12:1.

1
38
T.T. Pham et al. / Applied Catalysis A: General 379 (2010) 135–140
Table 2
Table 4
◦
Etherification of propanal or 2-methylpentanal at 125 C on 16% Pd/SiO2 reduced at
Effect of reduction temperature of 3% Pd and 16% Pd catalysts on MPAL conversion
◦
◦
1
50 C, molar H2:feed = 12:1.
at 125 C, molar H2:MPAL ratio = 12:1.
Feed
Propanal
2-Methylpentanal
(
Catalyst
3% Pd
16% Pd
150a
MPAL)a
1
50a
200a
250a
200a
W/F (h)
Conversion (%)
2
23.8
1
23.8
W/F (h)
Conversion (%)
2
20.9
2
24.2
2
22.7
1
23.8
1
20.0
Yield (%)
Ethane
Propanol
Dipropylether
Yield (%)
n-Pentane
MPOL
2.5
14.6
6.7
n-Pentane
MPOL
DMPE
5.5
6.4
11.9
7.0
6.7
7.2
5.1
10.1
9.1
5.3
7.2
10.2
5.5
6.4
11.9
6.4
0.6
13.1
C12 ether
Ether/hydrocarbon ratio
2.7
2.2
Selectivity (%)
n-Pentane
MPOL
a
Data from Fig. 1.
33.4
32.0
34.6
21.0
41.6
37.4
23.2
31.7
45.1
23.1
26.8
50.1
31.7
3.0
65.3
C12 ether
3.2. Effect of metal loading on 3, 10 and 16 wt.% Pd/SiO2
a
◦
Treduction ( C).
◦
Catalysts with different metal loadings, reduced at 150 C, were
compared for etherification at similar levels of conversions of
-methylpentanal. As shown in Table 3, CO chemisorption mea-
Table 5
Effect of K addition to 10% Pd/SiO2 catalysts in the 2-methylpentanal conversion at
1
2
◦ ◦
25 C, molar H2:MPAL ratio = 12:1, after reduction at 150 C.
surements indicate that increasing the loading from 3 to 16 wt.%
resulted in increased particle size. The estimated aldehyde reac-
tion rate per surface atom is similar for the three catalysts. The
ether yield and selectivity are highest for the 16 wt.% Pd catalyst,
also shown in Table 3. The other major products are n-pentane,
which is produced by decarbonylation, and 2-methylpentanol from
hydrogenation.
Catalyst
10% Pd
0.2% K–10% Pd
2% K–10% Pd
W/F (h)
Conversion (%)
1
20.4
2
14.9
2
9.8
Selectivity (%)
n-Pentane
MPOL
26.0
36.4
37.6
55.7
25.5
18.8
95.0
4.2
0.8
C12 ether
3.3. Effect of reduction temperature
◦
3.5. In situ DRIFTS
When the 3 wt.% Pd was reduced at 150 C, the yield of ether was
.2% at W/F = 2 h, lower compared to the ether yields for the 10% Pd
7
The spectra from CO chemisorption on 3% Pd at different reduc-
tion temperatures are shown in Fig. 5. The peaks at 1980 and
(
7.7%) and 16% Pd (11.9%), as shown in Table 3. However, when the
◦
reduction temperature of the 3% Pd was increased from 150 C to
−
1
◦ ◦
1940 cm
are assigned to bridge and multibonded CO species,
2
00 C and 250 C for experiments at a conversion of about 20%,
respectively, typical of relatively large particles, in agreement
with the metal dispersions estimated from chemisorption mea-
surements [20,21]. The much weaker band ascribed to linear CO
the ether yield was found to increase from 7.2% to 9.2% and 10.2%,
respectively, as shown in Table 4. When the reduction temperature
was increased to 200 C for the 16% Pd, an increased ether yield
from 11.9% to 13.1% was observed, even while the conversion was
reduced from 23.8% to 20%.
◦
−
1
appearing at around 2070 cm
[22,23] overlaps with the two
branches resulting from residual gas phase CO remaining in the
optical path after flushing with He for 20 min. As the reduction tem-
perature increases, not only do both the bridge and multibonded
CO peaks become relatively larger, but the multibonded-to-bridge
ratio also increases. The bridge form is mostly associated with the
presence of Pd(1 0 0) faces in the metal particle [24,25] while the
multibonded species is more typical of Pd(1 1 1) faces [24,25]. This
tendency is not only consistent with an increase in metal particle
3.4. Effect of alkali addition
The results obtained on the K-containing catalysts are summa-
rized in Table 5. It is observed that even at half the space velocity,
the conversion on the K-doped catalysts is significantly lower than
that obtained on the K-free Pd catalyst. The etherification selectiv-
ity decreases with the addition of K, while that of decarbonylation
greatly increases.
Table 3
◦
Activity of 3, 10 and 16 wt.% Pd/SiO2 reduced at 150 C with product yields from
MPAL conversion at 125 C, molar H2:MPAL ratio = 12:1.
◦
Catalyst
3% Pd
10% Pd
16% Pd
Dispersion (%)
W/F (h)
7.10
2
5.30
1
3.90
1
Conversion (%)
TOF (molecule/site s)
20.9
1.44E−02
20.4
1.13E−02
23.8
1.14E−02
Yield (%)
n-Pentane
MPOL
7.0
6.7
7.2
5.3
7.4
7.7
5.5
6.4
11.9
C12 ether
Selectivity (%)
n-Pentane
MPOL
33.4
32.0
32.6
26.0
36.4
37.6
23.1
26.8
50.1
C12 ether
◦
◦
◦
Fig. 5. FTIR of CO on 3% Pd reduced at 150 C, 250 C and 350 C.

T.T. Pham et al. / Applied Catalysis A: General 379 (2010) 135–140
139
Fig. 6. Schematic reaction pathway of 2-methylpentanal on Pd catalyst.
size, but also with a smoothing of the surface caused by the crystal
annealing occurring as the reduction temperature increases.
4
. Discussion
In early work that is very relevant to the current study, Ponec
and co-worker [16] observed the production of ether when pulsing
different mixtures of alcohol and aldehyde over metal catalysts.
In agreement with that work, we show here that when fed
individually, both aldehydes and alcohols form ethers. However,
etherification is significantly enhanced when both are fed simul-
taneously. When only one is present in the feed, etherification
apparently occurs due to the inter-conversion of aldehyde to alco-
hol or vice versa. The proposed reaction pathways are shown in
Fig. 6.
As mentioned above, because both the aldehyde and alcohol
have the same methylpentyl alkyl group, the resulting ether is sym-
metric and the relative contribution of each is not discernable. The
results from the mixture experiment with MPAL and BOH clearly
show that both are required to maximize ether, confirming the
results of the MPAL–MPOL experiments. This condition suggests
that each oxygenate forms a unique surface intermediate that is
needed for the formation of the ether bond and is not efficiently
inter-converted. An alcohol molecule is typically adsorbed on metal
Fig. 7. Adsorption modes of alcohols and aldehydes on Pd(1 1 1) and sequence from
Á (C,O)-aldehyde to Á (C,C)-ketene of decarbonylation [26].
2
2
surfaces [34,35] revealed that the cleavage of the C–H bond of the
methoxide is the rate-limiting step in methanol decomposition to
2
surfaces as an alkoxide. When an aldehyde is adsorbed as an Á (C,O)
surface H CO species. They also found the same results when using
2
species adjacent to the alkoxide [17,18,26], this enables the for-
mation of a C–O–C bond generating an ether [14,16]. Accordingly,
etherification should be favored on metal catalysts that promote
the co-existence of both species. Under these conditions, the Á2
adsorbed C–O may cleave and the ether bond is formed with the
alkoxide oxygen [14,26]. The residual oxygen from the aldehyde is
removed from the surface as water in the presence of hydrogen.
The formation of this surface hydroxyl should provide the driving
force for the oxygen cleavage from the adsorbed aldehyde carbon,
allowing formation of the ether bond.
ethanol.
In contrast to alkoxides, the surface species that result from the
adsorption of aldehydes strongly depends on the type of metal used.
On Group IB metals, aldehydes adsorb exclusively in an Á config-
1
1
2
uration, but they adsorb in both Á and Á configurations on Group
VIII metals at a ratio that strongly depends on the specific metal
[27]. In the former configuration, the aldehyde binds to the sur-
face through the O lone pair, while in the latter it binds though
the ␲ orbital of the carbonyl, overlapping the d electrons from the
metal with the antibonding ␲* orbital in the carbonyl. As a result,
when there is significant back-donation from the metal, this config-
uration can be readily stabilized [26]. Therefore, the preference to
Formation of the alkoxide species is a relatively easy process that
most metal surfaces are able to promote [27–29]. On Pt-group met-
als (including Pt [30,31], Ni [32] and Pd [33]), adsorption of alcohols
results in the formation of an alkoxide on the surface via hydroxyl-
hydrogen elimination, followed by ␣-hydrogen abstraction that
leads to an aldehyde intermediate bonded onto the surface via both
1
2
form either Á or Á surface species can be explained by these elec-
tronic characteristics of the metal. For example, the much greater
2
1
Á /Á ratio observed on Ru compared to Pt has been ascribed to
a higher extent of this electron back-donation from the metal to
the ␲* orbital of carbonyl, resulting from the higher position of
the Fermi level in Ru compared to Pt. It is well known that a more
efficient transfer of an electron into the antibonding states of the
2
carbon and oxygen atoms (in an Á (C,O) configuration) or an oxy-
1
gen atom only (Á (O) configuration), as illustrated in Fig. 7. Kinetic
experiments for CH OH decomposition on Ni(1 1 1) and Ni(1 1 0)
3

1
40
T.T. Pham et al. / Applied Catalysis A: General 379 (2010) 135–140
molecule’s ␲ system enhances the strength of interaction with the
metal, while it weakens the C–O bond, favoring the Á2 mode.
In the specific case of Ru, it has been observed [29] that while
5. Conclusion
The etherification reaction has been studied for various aldehy-
des and alcohols on supported Pd catalysts. While alcohols adsorb
as alkoxide species on the surface, aldehydes adsorb as Á (C,O). For
Á2 is the preferred adsorption species on clean Ru surfaces, only
1
2
Á
is present on oxygen-covered surfaces. Those authors proposed
2
that while the hindrance of Á formation can be ascribed to simple
geometric blockage of sites, electronic effects are much more sig-
nificant in determining which configuration is preferred [27]. It is
expected that co-adsorption of an electronegative atom results in
an increase in the work function of the metal, lowering the Fermi
level and making back-donation less favorable. By contrast, they
have observed that Á2 species are favored by co-adsorption of K,
which causes a decrease in work function. It is also expected that
the specific surface plane on which the aldehyde adsorbs will affect
the type of surface species formed.
high rates of ether formation, it is necessary to have both alkox-
2
ide and Á -adsorbed species on the surface and a stoichiometric
mixture of 1:1 of aldehyde and alcohol has been found to be the
optimum.
Larger metal particles, that have been sintered and annealed by
high reduction temperatures, show lower conversion but higher
ether selectivity due to enhancement of the ensembles required
for etherification.
Acknowledgments
In this contribution, we have demonstrated that the generation
of smooth planes (e.g. Pd(1 1 1)) on the particle surface by increas-
ing either the metal loading or the reduction temperature results
in a higher etherification rate. This enhancement in rate cannot
be due to a higher concentration of Á2 species, since it cannot be
expected that Pd(1 1 1) planes result in a higher Á /Á ratio. In fact, it
has been recently shown [36–38] that in high-coordination smooth
surfaces, the d-band center is located at a lower energy, farther
from the Fermi level than rougher surfaces with lower coordination
numbers. Accordingly, one may expect that a catalyst with larger
and smoother Pd particles, such as those present in the high Pd
The support of the Oklahoma Secretary of Energy, the Okla-
homa Bioenergy Center, The National Science Foundation EPSCoR
(0814361), and the Department of Energy (DE-FG36GO88064) are
gratefully acknowledged.
2
1
References
[
[
1] J.N. Chheda, J.A. Dumesic, Catal. Today 123 (2007) 59–70.
2] R.M. West, Z.Y. Liu, M. Peter, C.A. Gärtner, J.A. Dumesic, J. Mol. Catal. A 296
(
2008) 18–27.
[3] T. Sooknoi, T. Danuthai, L.L. Lobban, R.G. Mallinson, D.E. Resasco, J. Catal. 258
(2008) 199–209.
[4] M. Snare, I. Kubickova, P. Maki-Arvela, D. Chichova, K. Eranen, D.Y. Murzin, Fuel
loading samples studied here, particularly after high temperature
reduction, should not present a higher concentration of Á2 species
87 (2008) 933–945.
than a catalyst with smaller and rougher particles. Therefore, the
enhanced etherification activity of these surfaces must be ascribed
to their ability to co-adsorb the two required species on adjacent
sites rather than an enhanced ability to adsorb either one.
The 16 wt.% Pd, with the largest particle size, may have a higher
density of large ensembles necessary for condensation to form the
ether (DMPE), thus increasing its yield, while decreasing the yields
of the other products. The increased ether yield with increasing
reduction temperature, suggests that the Pd surface anneals into
thermodynamically more stable surfaces (i.e. (1 1 1)), resulting in
a surface structure having larger ensembles that are selective for
ether formation. The FTIR results show structural changes to the
surface that are consistent with this view.
[
5] L. Karas, W.J. Piel, “Ethers”, Kirk-Othmer Encyclopedia of Chemical Technology,
John Wiley & Sons, Inc., 2004, pp. 567–583.
6] G.A. Olah, US Patent 5,520,710 (1996).
7] E.W. Flick, Industrial Solvents Handbook, 5th edition, 1998.
8] R.J.J. Nel, A. De Klerk, Ind. Eng. Chem. Res. 48 (2009) 5230–5238.
[
[
[
[9] A. Corma, M. Renz, Angew. Chem. Int. Ed. 46 (2007) 298–300.
10] R.H. Clark, W.E. Graham, A.G. Winter, J. Am. Chem. Soc. 47 (1925) 2748–2754.
11] R. Figueras Roca, L. De Mourgues, Y. Trambouze, J. Catal. 14 (1969) 107–113.
12] H. Feuer, J. Hooz, in: S. Patai (Ed.), The Chemistry of the Ether Linkage, Inter-
science, London, 1967, pp. 457–460.
[
[
[
[
[
13] T.P. Kobylinski, H. Pines, J. Catal. 17 (1970) 384–393.
14] A. van der Burg, J. Doornbos, N.J. Kos, W.J. Ultee, V. Ponec, J. Catal. 54 (1978)
2
43–253.
[15] G.M.R. van Druten, V. Ponec, Appl. Catal. A 191 (2000) 153–162.
16] G.M.R. van Druten, V. Ponec, Appl. Catal. A 191 (2000) 163–176.
17] F. Delbecq, P. Sautet, Surf. Sci. 295 (1993) 353–373.
18] R. Shekhar, M.A. Barteau, R.V. Plank, J.M. Vohs, J. Phys. Chem. B 101 (1997)
7051–7939.
[19] T.T. Pham, L.L. Lobban, D.E. Resasco, R.G. Mallinson, J. Catal. 266 (2009) 9–14.
20] A. Palazov, C.C. Chang, R.J. Kokes, J. Catal. 36 (1975) 338–350.
21] A. Arteaga, F.M. Hoffmann, A.M. Bradshaw, Surf. Sci. 119 (1982) 79–94.
22] L.L. Sheu, Z. Karpinski, W.M.H. Sachtler, J. Phys. Chem. 93 (1989) 4890–4894.
[
[
[
The decarbonylation reaction requires further transformation of
2
1
an Á (C,O) species with its conversion to an Á (C) acyl configuration
[
[
[
with the O angled away from the surface (by further H-abstraction
2
by the metal) [26]. This then may form an Á (C,C) ketene with the
adjacent C bonding to the surface [26]. C–C bond scission occurs,
leaving the alkyl fragment and CO [26]. The ether yield is higher
than the decarbonylated C5 hydrocarbon because the ether is more
readily formed when there are adjacent surface alkoxides.
[23] F. Skoda, M.P. Astier, G.M. Pajonk, M. Primet, Catal. Lett. 29 (1994) 159–168.
[
[
24] M. Fernandez-Garcia, J.A. Anderson, G.L. Haller, J. Phys. Chem. 100 (1996)
6247–16254.
25] F.M. Hoffmann, Surf. Sci. Rep. 3 (1983) 107–192.
1
[26] M. Mavrikakis, M.A. Barteau, J. Mol. Catal. A 131 (1998) 135–147.
[
27] N.R. Avery, W.H. Weinberg, A.B. Anton, B.H. Toby, Phys. Rev. Lett. 51 (1983)
82–685.
28] N.R. Avery, Surf. Sci. 125 (1983) 771–786.
The experiments conducted on K-promoted catalysts further
support the explanations given above and add another important
concept to the discussion. In the first place, as mentioned above,
the presence of K on Pd surfaces should result in an increased con-
6
[
[29] A.B. Anton, N.R. Avery, B.H. Toby, W.H. Weinberg, J. Am. Chem. Soc. 108 (1986)
84–694.
6
[
[
30] J.L. Davis, M.A. Barteau, Surf. Sci. 187 (1987) 387–406.
31] J.L. Davis, M.A. Barteau, Surf. Sci. 235 (1990) 235–248.
2
centration of Á species [27]. However, this increase does not result
in enhanced etherification activity, but rather a decrease that may
be explained by the dilution of the ensemble site necessary for the
bimolecular surface reaction. A final interesting point is the dra-
matic increase in selectivity for decarbonylation observed for the
[32] S.M. Gates, J.N. Russell Jr., J.T. Yates Jr., Surf. Sci. 171 (1986) 111–134.
[
[
[
33] B.A. Sexton, K.D. Rendulic, A.E. Hughes, Surf. Sci. 121 (1982) 181–198.
34] S.R. Bare, J.A. Stroscio, W. Ho, Surf. Sci. 150 (1985) 399–418.
35] S.M. Gates, J.N. Russell Jr., J.T. Yates Jr., Surf. Sci. 159 (1985) 233–255.
[36] B. Hammer, J.K. Nørskov, Adv. Catal. 45 (2000) 71–129.
[37] M. Mavrikakis, B. Hammer, J.K. Nørskov, Phys. Lett. 81 (1998) 2819–2822.
2
% K-doped catalyst. This selectivity change may be associated not
[
38] C.E. Tripa, T.S. Zubkov, J.T. Yates Jr., M. Mavrikakis, J.K. Nørskov, J. Chem. Phys.
11 (1999) 8651–8658.
only with an electronic modification of the Pd activity, but also to
a direct participation of K in the reaction interacting directly with
the carbonyl oxygen of the aldehyde and enhancing its C–C bond
cleavage.
1