Insights into the Oxidation State and Location of Rhenium in Re-Pd/TiO2 Catalysts for Aqueous-Phase Selective Hydrogenation of Succinic Acid to 1,4-Butanediol as a Function of Palladium and Rhenium Deposition Methods

DOI: 10.1002/cctc.201500197

Full Papers

Insights into the Oxidation State and Location of Rhenium

in Re-Pd/TiO₂Catalysts for Aqueous-Phase Selective

Hydrogenation of Succinic Acid to 1,4-Butanediol as

a Function of Palladium and Rhenium Deposition Methods

Bao Khanh Ly,^[a]Benoît Tapin,^[b]Mimoun Aouine,^[a]Pierre Delichere,^[a]Florence Epron,^[b]

Catherine Pinel,^[a]Catherine Especel,*^[b]and Michle Besson*^[a]

This paper is dedicated to Dr. Daniel Duprez and Dr. Pierre Gallezot with whom we have shared many discussions about supported

(bi)metallic catalysts.

ReO_x-Pd/TiO₂catalysts prepared from different 2 wt%Pd/TiO₂

catalysts using two protocols for the deposition of the Re pro-

moter (successive impregnation and catalytic reduction) were

characterized by different techniques to better understand the

nature of the active and selective sites implied in the aqueous-

phase hydrogenation of succinic acid to 1,4-butanediol. Re-

gardless of the support and Re introduction method, it was es-

tablished that varying amounts of Pd and Re were in very

close proximity without electronic interaction in the reduced

catalysts. A high fraction of Re always remained partially oxi-

dized to generate a bimetallic catalyst that can provide the

necessary bifunctional sites to enable the selective hydrogenol-

ysis of the intermediate g-butyrolactone to 1,4-butanediol. De-

pending on the method of promotion, the ReO_xspecies that

interact with Pd were deposited as clusters with different spa-

tial Re–Re interactions.

Introduction

Succinic acid (SUC; IUPAC

name: butanedioic acid) is a pet-

Scheme 1. Catalytic hydrogenation of SUC to BDO.

rochemical intermediate that is

currently produced by the cata-

lytic hydrogenation of maleic anhydride to succinic anhydride

and subsequent hydration or by the catalytic hydrogenation of

maleic acid. After it was the focus of many research biotechno-

logical projects in the last decade, bio-based SUC is now pro-

duced by fermentation in a sustainable and highly efficient

manufacturing process using renewable biomass as the feed-

stock and metabolically engineered micro-organisms.^[1,2]SUC

has been identified as one of the most important platform

chemicals in biorefinery, which will one day replace petroleum

refinery.^[3,4]SUC has a wide range of industrial applications.^[5]It

can be converted readily by catalytic hydrogenation to other

important bulk chemicals, such as 1,4-butanediol (BDO), THF,

and g-butyrolactone (GBL; Scheme 1). In particular, BDO, which

was until recently manufactured entirely from petroleum-

based feedstocks, is an important commodity chemical that is

used to manufacture large amounts of plastics and important

polymers such as polyesters, polyurethanes, and spandex elas-

tic fibers. A variety of catalysts have been described extensively

in patents for the catalytic hydrogenation reaction of maleic

acid, maleic anhydride, and SUC to BDO. The most frequently

used active and selective catalysts for this reaction comprise

a noble metal in combination with another metal such as Re

or Sn. These reactions are conducted predominantly in the gas

phase or in organic solvent.^[6,7]The use of a series of monome-

tallic Pd, Ru, and Re catalysts supported on different carbon

materials has been reported for the liquid-phase hydrogena-

tion of SUC to THF in dioxane at 2408C under 60 bar.^[8–13]Bi-

metallic ReÀRu catalysts showed selectivity to BDO at 2008C

under 80 bar in dioxane.^[14]

[a] Dr. B. K. Ly, Dr. M. Aouine, Dr. P. Delichere, Dr. C. Pinel, Dr. M. Besson

Institut de recherches sur la catalyse et l’environnement de Lyon

(IRCELYON)

UMR CNRS-UniversitØ Lyon1

2 Avenue Albert Einstein, 69626 Villeurbanne Cedex (France)

Fax: (+33)04-72-44-53-99

E-mail: michele.besson@ircelyon.univ-lyon1.fr

[b] Dr. B. Tapin, Dr. F. Epron, Dr. C. Especel

Institut de Chimie des Milieux et des MatØriaux de Poitiers (IC2MP)

UMR CNRS-UniversitØ Poitiers

4 rue Michel Brunet, TSA 51106, 86073 Poitiers Cedex 9 (France)

Fax: (+33)05-49-45-37-41

E-mail: catherine.especel@univ-poitiers.fr

Supporting information for this article is available on the WWW under

http://dx.doi.org/10.1002/cctc.201500197.

However, there are few examples of the hydrogenation of

SUC in an aqueous solution,^[15–21]an important medium in the

conversion of compounds obtained from renewable resources

This publication is part of a Special Issue on Palladium Catalysis. To view

the complete issue, visit:

http://onlinelibrary.wiley.com/doi/10.1002/cctc.v7.14/issuetoc

ChemCatChem 2015, 7, 2161 – 2178

2161

ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers

processing.^[4]The harsh operating temperatures usually used

(up to 2408C) favored the formation of THF and selectivity to

BDO was limited.^[15–17]However, we have reported the aque-

ous-phase hydrogenation of SUC to BDO at 1608C under

150 bar using Re-Pd catalysts supported on carbon or

TiO₂.^[18–20]Monometallic Pd catalysts were totally unselective to

BDO and produced mostly GBL.^[21]If only Re was present, no

reaction was observed in water. In turn, bimetallic catalysts

that combine both metals, namely, Re deposited on Pd cata-

catalysts were characterized extensively using various tech-

niques such as XRD, TEM, X-ray photoelectron spectroscopy

(XPS), temperature-programmed reduction (TPR), H₂chemi-

sorption, and the gas-phase model reaction of cyclohexane de-

hydrogenation, a structure-insensitive reaction catalyzed by Pd

active sites. The characterizations were correlated tentatively

to the catalytic results observed.

lysts, have demonstrated the possibility to tune the reaction Results

selectivity towards BDO. We have also observed the effects of

XRD and TEM characterizations

the method of preparation of the Pd catalyst, by the method

of introduction of the Re precursor over the monometallic Pd

catalyst, and by the Re loading. The results indicated that the

Pd and Re monometallic catalysts

The three monometallic Pd parent catalysts used in this study

(2 wt%Pd_Cl/P25, 2 wt%Pd_KCl/P25, and 2 wt%Pd_KCl/DT51) have

been described previously for the selective hydrogenation of

SUC to GBL and some of their characteristics have been

shown.^[21]

catalysts prepared by

a simple successive impregnation

method (SI) required a significantly higher amount of Re com-

pared with the catalysts prepared by catalytic reduction depo-

sition (CR) to obtain a high yield of BDO. However, the SI-pre-

pared Pd-Re catalysts exhibited a higher selectivity to BDO.

Such a marked effect on catalyst selectivity might be be-

cause of the different locations of the Re additive or the

degree of contact between the promoter and Pd. Systematic

studies on the characterization of the two-component catalysts

and the interpretation of the action of the Re promoter associ-

ated with the noble metal on the activity and selectivity of this

reaction are rather scarce.^[14,20]As an example, characterizations

of a series of Re-Ru catalysts supported on mesoporous carbon

(MC) after incipient wetness coimpregnation of metal precur-

sors in acetone showed the formation of a Re-Ru miscible

phase during the reduction process at 5008C.^[14]The reducibili-

ty, metal dispersion, and oxidation state were influenced

strongly by the Re/Ru molar ratio. The 0.3Re-0.2Ru/MC catalyst

showed the highest turnover frequency (TOF) for the forma-

tion of BDO in dioxane, which was correlated to the largest

amount of weak hydrogen binding sites determined by H₂

temperature-programmed desorption (TPD).

The Pd catalysts showed no XRD peaks other than those of

rutile or/and anatase, which indicated a high dispersion of Pd

in all prepared samples. The highest dispersion measured from

H₂chemisorption (33%) was obtained for the catalyst prepared

by the impregnation of the PdCl₂salt (2 wt%Pd_Cl/P25), whereas

the deposition–precipitation method that involves K₂PdCl₄af-

forded dispersions of 15 and 27% for 2 wt%Pd_KCl/P25 and

2 wt%Pd_KCl/DT51, respectively. By TEM, the 2 wt%Pd_Cl/P25 cat-

alyst showed a homogeneous Pd distribution with particles

that were mainly between 1 and 2 nm in diameter (average

particle size 1.8 nm).^[21]The Pd particles in 2 wt%Pd_KCl/DT51

were also dispersed homogeneously on the solid with diame-

ters between 1.5 and 3 nm (average particle size 2.4 nm; Fig-

ure S1). The average metal particle sizes determined by TEM

for 2 wt%Pd_Cl/P25 and 2 wt%Pd_KCl/DT51 (1.8 and 2.4 nm, re-

spectively) were slightly smaller than those deduced from H₂

chemisorption (2.8 and 3.5 nm, respectively).^[21]

In the XRD pattern of 1.9 wt%Re_pH1/P25 and 3.1 wt%Re_pHn

/

In a recent paper,^[20]we compared the physical and chemical

properties of two Re-Pd bimetallic catalysts synthesized from

a given 2 wt%Pd/TiO₂monometallic catalyst (which was pre-

pared by an impregnation method in an acidic aqueous

medium using PdCl₂as the precursor salt) either by CR or by

SI. This study gave us some partial information on the Re pro-

moter and suggests a different localization of the Re deposit

and a different oxidation degree of the Re species depending

on the introduction mode. However, the mechanism that un-

derlies the effect of Re on the tuning of the catalyst selectivity

to the desired pathway was not resolved. As different interac-

tions between the Pd precious metal and the Re promoter

may be expected that depend on the deposition method, we

thought it was necessary to characterize a larger set of cata-

lysts. For that purpose, Re was introduced on three different

monometallic catalysts and the resulting bimetallic solids were

characterized thoroughly to provide further insights into the

Re species present, their different oxidation states, their loca-

tion relative to the Pd particles, and the interactions between

both metals and with the support to try to explain the differ-

ences in the catalytic performances observed. In this study, the

DT51 monometallic Re catalysts, the peaks that correspond to

metallic Re or ReO_xspecies were not detected at all, which in-

dicates the high dispersion of the Re species. It was difficult to

detect Re particles in the TEM images of 1.9 wt%Re_pH1/P25

(Figure S2). Only very scarce support particles showed the

presence of highly dispersed nanoparticles (<0.7 nm). Howev-

er, the environmental transmission electron microscopy (ETEM)

pictures in the scanning transmission electron microscopy

(STEM) mode of 3.1 wt%Re_pHn/DT51 were able to show very

small and uniform Re nanoclusters (<0.7 nm) on the support

(Figure 1). These observations are consistent with the litera-

ture. Re in 1.05 wt%Re/Al₂O₃reduced at 5508C existed as very

small clusters (<1 nm) that were undetectable by TEM.^[22]Very

few nanoparticles with a size of 1 nm or less could be ob-

served in Re/TiO₂.^[23]

Re-Pd bimetallic catalysts

No discrete Pd- or Re-containing phase was evidenced in the

XRD patterns of any of the Re-Pd bimetallic catalysts, which is

ChemCatChem 2015, 7, 2161 – 2178

www.chemcatchem.org

2162

ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers

Figure 2 show evidence of well-dispersed particles in both

solids. However, as the contrast and the particle sizes were

very similar, it was not possible to draw conclusions on the re-

spective localization of Re or Pd. Some TEM pictures of the cat-

alysts prepared from 2 wt%Pd_KCl/P25 loaded with 0.6 or

2.9 wt% Re by CR show well-dispersed metallic particles (Fig-

ure S6).

Figure 1. Representative ETEM pictures of 3.1 wt%Re_pHn/DT51 in HAADF-

STEM mode. Left: scale bar=5 nm; right: scale bar=1 nm.

likely because the crystallites are too small to exhibit diffrac-

tion patterns (Figure S3).

TEM and STEM images were obtained for some “as-pre-

pared” Re-Pd samples (after reduction and passivation) and

energy-dispersive X-ray (EDX) analysis was used to investigate

the compositions on small- and large-area scans. The observa-

tion was difficult and gave only semiquantitative results as ex-

posure to the stationary electron beam during EDX resulted in

the severe degradation of the exposed area. Images and scan

areas of two bimetallic catalysts prepared from 2 wt%Pd_KCl

/

DT51 using the CR and SI protocols, 0.8 wt%Re-Pd_KCl/DT51-CR

and 3.4 wt%Re-Pd_KCl/DT51-SI, respectively, are shown in Fig-

ures S4 and S5.

In 0.8 wt%Re-Pd_KCl/DT51-CR, most of the particles were

highly dispersed with a diameter in the range 1–3 nm, which is

not significantly different from that in the monometallic parent

(Figure S4). The particles analyzed contained only Pd, and no

detectable Re content was evidenced by EDX analysis (Fig-

ure S4a). An area was found to contain a very large agglomer-

ate of Re with a diameter of around 1 mm (Figure S4b). Howev-

er, this large particle was isolated, and it does not exclude the

possibility that the majority of Re was still well dispersed in

the solid. This was rather surprising if we consider that the CR

method should deposit Re selectively on Pd. An explanation

for this unexpected distribution could find its origin in experi-

mental problems encountered during the preparation of this

solid from Pd_KCl/DT51 by CR. Indeed, the very fine granulation

of this support caused plugging of the frit and an inhomoge-

neous hydrogen flow, which could have disturbed the deposi-

tion of the promoter during preparation.

Figure 2. Representative HAADF-STEM images of a, b) 2 wt%Pd_KCl/DT51 and

c) 3.4 wt%Re-Pd_KCl/DT51-SI.

Analysis of 1.5 wt%Re-Pd_KCl/P25-CR performed using the

ETEM instrument in TEM and high-angle annular dark-field

scanning transmission electron microscopy (HAADF-STEM)

mode is shown in Figure 3. The TEM images showed the pres-

ence of Pd particles (Figure 3a). However, the Re clusters were

so small that they were not observable. EDX analysis detected

mainly Pd associated with some Re (Figure S7, TEM EDX 1–3)

and some zones that contained only Re (Figure S7, TEM EDX

4). In HAADF-STEM mode, the images confirmed that Re was

dispersed atomically all over the catalyst surface (Figure 3b

and c).

In 3.4 wt%Re-Pd_KCl/DT51-SI, which has a high Re content,

small particles in the range 1–2 nm were also detected (Fig-

ure S5). EDX analysis on large zones (50 nm) detected both Pd

and Re with a mass ratio of Re/Pd=2.03 close to the ratio of

1.75 determined by inductively coupled plasma optical emis-

sion spectroscopy (ICP-OES; Figure S5). Several spot elemental

analyses with the beam concentrated on isolated particles

showed that they contained either Pd or both Pd and Re, with

Re/Pd mass ratios that varied considerably among the particles

analyzed and that were lower than the global mass ratio. This

3.4 wt%Re-Pd_KCl/DT51-SI catalyst and the corresponding parent

catalyst were analyzed in the STEM mode with the annular

dark-field detector (ADF-STEM). The images displayed in

Finally, TEM characterization of the bimetallic catalysts pre-

pared from 2 wt%Pd_Cl/P25 has been described previously for

solids that contain 0.9 and 3.5 wt% Re deposited by CR and SI,

respectively.^[20]The bimetallic catalyst loaded by CR with

0.9 wt% Re showed metallic particles that were evidently

larger than that on the bimetallic catalyst prepared by SI that

contained 3.5 wt% Re. The average particle size of these solids

was determined to be 5.3 and 2.5 nm, respectively, which can

be compared with the parent Pd catalyst average particle size

of 1.8 nm. Re was detected on the Pd particles (Figure S7,

STEM EDX).

ChemCatChem 2015, 7, 2161 – 2178

www.chemcatchem.org

2163

ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers

Following the reduction of the precursor at 4508C (Fig-

ure 4a), fitting of the Re4f_7/2envelope resulted in the identifi-

cation of two Re components: one with a binding energy (BE)

of 40.7 eV, which matches closest with Re⁰with an atomic con-

centration of 89%, and a second one at BE=41.5 eV, which

was attributed to Re³⁺(atomic concentration 11 %; Table 1,

entry 1).^[24]As a result of the high oxygen affinity of Re, it was

expected that handling in ambient air would cause the oxida-

tion of surface Re. Indeed, after subsequent air exposure of the

reduced solid (Figure 4b), Re was reoxidized and the “as-pre-

pared” catalyst then contained a mixture of oxidized Re spe-

cies with 17% Re⁴⁺(BE=43.1 eV), 44% Re⁶⁺(BE=44.2 eV), and

39% Re⁷⁺(BE=46.2 eV; Table 1, entry 2). However, after a new

in situ hydrogen treatment (Figure 4c), only 15% Re was in the

zero-valent state, and the remainder was Re³⁺(Table 1,

entry 3). This indicates that different Re species that can be

oxidized to various degrees are present and coexist on the

support surface.

Furthermore, the Re/Ti atomic ratio initially determined at

0.196 in the solid reduced at 4508C (Figure 4a) decreased sub-

stantially to 0.054 and 0.049 (Table 1, entries 1–3), respectively,

after air exposure (Figure 4b) and subsequent rereduction (Fig-

ure 4c). In addition, in measurements of the reactivated solid,

the BEs of Re⁰and Re³⁺were 0.2 eV lower than that in the

freshly reduced sample. Both observations indicate some ag-

glomeration of Re particles on the surface, which could be at-

tributed to the presence of moisture during the reduction

step.^[25]It is highly probable that nanoclusters or monoatomic

Re species exist in the initially reduced catalysts as proposed

and evidenced on alumina and TiO₂.^[25–28]After exposure to air,

Re was detected by TEM as very small particles with a size

<0.7 nm in the “as-prepared” sample (Figure S2 and Figure 1).

Figure 3. ETEM pictures of 1.5 wt%Re-Pd_KCl/DT51in a) TEM (scale

bar=10 nm) and b, c) HAADF-STEM mode (scale bars=2 nm). The EDX anal-

ysis is shown in Figure S7.

XPS

XPS could provide valuable information about the oxidation

states of the various species present in the monometallic Re

and Pd and bimetallic Pd-Re catalytic systems. The measure-

ments were performed (i) for the monometallic Re catalyst,

either after in situ activation with H₂at 4508C of the precursor

or “as-prepared”, that is, reduced at 4508C and then handled

under ambient conditions, (ii) for the parent monometallic Pd

catalysts, either “as-prepared”, or following in situ reactivation

with H₂at 3008C, or (iii) for the two-component solids, either

after in situ reduction of the precursor at 4508C for 3 h, “as-

prepared”, or after in situ rereduction. Some curve-fitting re-

sults (binding energy of Pd3d_5/2and Re4f_7/2, distribution of the

oxidation states) are summarized in Table 1. The Re/Pd, Pd/Ti,

and Re/Ti ratios could be calculated from the surface atomic

compositions by taking into consideration the sensitivity fac-

tors of the elements. The explored depth by XPS is approxi-

mately 3–5 nm. Thus, this analysis may be considered as quite

representative of the bulk composition because of the small

particle size of the metallic particles in all of the catalysts stud-

ied. Globally, after reduction Pd was present in the metallic

state and also as Pd²⁺and Pd⁴⁺; several oxidation states were

observed for Re, and the situation was more complex.

Pd monometallic catalysts

After synthesis and storage in air, the fitting of the Pd enve-

lope for the 2 wt%Pd_KCl/DT51 “as-prepared” monometallic cat-

alyst resulted in the identification of Pd⁰(82%, Pd3d_5/2BE=

335.2 eV) and a minor amount of oxidized Pd²⁺species (15%,

BE=336.9 eV) and Pd⁴⁺(3%, BE=338.0 eV; Table 1, entry 4).

After the solid was reactivated in H₂at 3008C, Pd was nearly

completely reduced to the zero-valent state (95% Pd⁰and 5%

Pd²⁺; Table 1, entry 5). The “as-prepared” Pd_Cl/P25 contained

Pd⁰(34%, BE=335.3 eV) and Pd²⁺(66%, BE=337.0 eV; Table 1,

entry 16). In the same way, reactivation at 4508C reduced Pd

totally to the zero-valent state (100% Pd⁰, BE=335.0 eV;

Table 1, entry 17). Pd_KCl/P25 was not analyzed but similar re-

sults may be expected, that is, reduction to mainly Pd⁰.

Re-Pd bimetallic catalysts prepared by SI from Pd_KCl/DT51 and

Pd_KCl/P25

Re monometallic catalyst

The Re4f_7/2spectra of solid 1.9%[NH₄ReO₄]_pH1/P25 after im-

pregnation with ammonium perrhenate, drying, and in situ re-

duction at 4508C, the “as-prepared” catalyst, that is, handled

under ambient conditions after synthesis, and then after reacti-

vation at 4508C under H₂are illustrated in Figure 4.

The Pd and Re spectra of the bimetallic 2.6wt%Re-Pd_KCl/DT51-

SI and 2.7 wt%Re-Pd_KCl/P25-SI catalysts, which were prepared

by SI of the Pd_KCl/DT51 and Pd_KCl/P25 parent catalysts, were an-

alyzed after their synthesis and then “as-prepared”, that is, fol-

lowing exposure to air. As an illustration, the spectra of Pd

ChemCatChem 2015, 7, 2161 – 2178

www.chemcatchem.org

2164

ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers

Table 1. XPS data (BEs of the Re4f_7/2and Pd3d_5/2components, distribution of Pd and Re oxidation states, surface atomic ratios).

XPS data

Re4f_7/2BE^[a]

[eV]

([%]Reⁿ⁺

)

[d]

Entry

Catalyst

Treatment

Pd3d_5/2BE^[a]

[eV]

Pd/Ti-Re/Ti-Re/Pd

(atomic ratio)

(Re/Pd)_bulk

(atomic ratio)

([%]Pdⁿ⁺

)

1

2

1.9%[NH₄ReO₄]_pH1/P25

1.9%Re_pH1/P25

reduction^[b]

40.7 (89 Re⁰)

41.5 (11 Re³⁺

43.1 (17 Re⁴⁺

44.2 (44 Re⁶⁺

46.2 (39 Re⁷⁺

40.5 (15 Re⁰)

41.3 (85 Re³⁺

0.196

0.054

)

“as-prepared”

)

3

4

reduction^[b]

0.049

0.042

)

2%Pd_KCl/DT51

“as-prepared”

335.2 (82 Pd⁰)

336.9 (15 Pd²⁺

)

338.0 (3 Pd⁴⁺

)

5

6

7

reduction^[b]

“as-prepared”

335.2 (95 Pd⁰)

336.9 (5 Pd²⁺

0.038

)

2.6%[NH₄ReO₄]-Pd_KCl/DT51-SI

2.6%Re-Pd_KCl/DT51-SI

335.2 (100 Pd⁰)

40.1 (30 Re⁰)

41.2 (70 Re³⁺

43.3 (16 Re⁴⁺

44.2 (41 Re⁶⁺

46.1 (43 Re⁷⁺

40.4 (27 Re⁰)

41.2 (73 Re³⁺

40.0 (30 Re⁰)

41.2 (70 Re³⁺

0.032/0.029/0.90

0.025/0.033/1.32

0.77

)

335.2 (57 Pd⁰)

336.9 (43 Pd²⁺

)

8

9

reduction^[b]

335.1 (100 Pd⁰)

0.026/0.036/1.38

0.027/0.028/1.04

)

1.9%Re-Pd_KCl/DT51-CR

335.1 (95 Pd⁰)

0.62

336.5 (5 Pd²⁺

)

10

11

2%Pd_KCl/P25

2.6%[NH₄ReO₄]-Pd_KCl/P25-SI

reduction^[b]

335.0 (100 Pd⁰)

335.0 (61 Pd⁰)

40.1 (37 Re⁰)

41.3 (63 Re³⁺

43.7 (29 Re⁴⁺

44.4 (24 Re⁶⁺

46.2 (47 Re⁷⁺

40.1 (34 Re⁰)

41.2 (66 Re³⁺

40.5 (36 Re⁰)

41.5 (64 Re³⁺

40.2 (74 Re⁰)

41.3 (26 Re³⁺

0.028/0.03/1.07

0.038/0.052/1.37

0.74

0.77

)

12

2.7%Re-Pd_KCl/P25-SI

“as-prepared”

337.0 (39 Pd²⁺

)

13

14

15

16

reduction^[b]

335.0 (100 Pd⁰)

335.1 (100 Pd⁰)

335.4 (92 Pd⁰)

)

0.027/0.059/2.18

0.027/0.051/1.90

O₂+reduction^[c]

reduction^[b]

1.5%Re-Pd_KCl/P25-CR

2%Pd_Cl/P25

0.030/0.050/1.69

0.065

336.7 (8 Pd²⁺

)

“as-prepared”

335.3 (34 Pd⁰)

337.0 (66 Pd²⁺

)

17

18

reduction^[b]

335.0 (100 Pd⁰)

335.2 (100 Pd⁰)

0.037

0.037/0.028/0.78

0.9%Re-Pd_Cl/P25-CR

1.7%Re-Pd_Cl/P25-CR

40.3 (14 Re⁰)

41.3 (86 Re³⁺

0.25

0.46

)

19

“as-prepared”

335.1 (66 Pd⁰)

43.4 (8 Re⁴⁺

)

0.036/0.058/1.61

0.036/0.066/1.83

336.8 (18 Pd²⁺

337.7 (16 Pd⁴⁺

)

44.3 (29 Re⁶⁺

46.1 (63 Re⁷⁺

)

20

reduction^[b]

335.1 (95 Pd⁰)

41.1 (100 Re³⁺

)

336.5 (5 Pd²⁺

)

[a] BE: binding energy. [b] In situ reduction at 4508C. [c] In situ oxidation at 3008C then reduction at 4508C. [d] ICP-OES data.

and/or Re of 2.6wt%Re-Pd_KCl/DT51-SI are shown in Figure 5:

“as-prepared”, (Table 1, entry 7), and after reactivation under H₂

at 4508C (Table 1, entry 8). The catalysts prepared by the same

SI method on P25 showed similar profiles (data not shown,

quantitative results are given in Table 1, entries 12 and 13).

First, after the introduction of Re by SI impregnation and

drying, the solid precursors 2.6 wt%[NH₄ReO₄]-Pd_KCl/DT51-SI

and 2.6 wt%[NH₄ReO₄]-Pd_KCl/P25-SI were reduced in the XPS

treatment cell at 4508C (Table 1, entries 6 and 11). Only metal-

lic Pd was detected (3d_5/2BE=335.2 and 335.0 eV, respective-

ly). Moreover, the peaks for Pd⁰in the bimetallic solids were

observed at the same BEs as those in the monometallic Pd ma-

terial, which suggests that there is no electronic modification

of the surface Pd atoms and that the electronic interaction be-

tween Re and Pd on the TiO₂supports is not very strong. In

addition, in both catalysts, the Re spectral fits (not shown) indi-

cated a mixture of Re species with characteristic Re4f_7/2BEs of

40.1 and ꢀ41.2–41.3 eV, which match Re⁰and Re³⁺, respective-

ly. The proportions were 30% Re⁰and 70% Re³⁺, and 37% Re⁰

and 63% Re³⁺in the DT51- and P25-supported catalysts, re-

spectively, which demonstrates that Re species under the con-

ditions employed could not be reduced completely to metallic

Re.

ChemCatChem 2015, 7, 2161 – 2178

www.chemcatchem.org

2165

ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers

Figure 5. Curve fitting of the XPS data of 2.6 wt%Re-Pd_KCl/DT51-SI:

a, b) Pd3d region and Re4f region of the solid without previous treatment

and c) Re4f region of the solid after in situ reactivation.

Figure 4. Curve fit to XPS data: Re4f region of a) 1.9 wt%[NH₄ReO₄]_pH1/P25

after in situ reduction, b) 1.9 wt%Re_pH1/P25 “as-prepared”, and

c) 1.9 wt%Re_pH1/P25 “as-prepared” after in situ reactivation.

solids at BE=335.1 eV following a new reduction of the “as-

prepared” catalysts in the XPS cell (Table 1, entries 8 and 13). In

contrast, Re was difficult to reduce to the metallic state: it was

again present as Re⁰and Re³⁺with minor proportions of Re⁰of

only 27 and 34% on DT51 and P25, respectively (Figure 5c;

Table 1, entries 8 and 13). It is noticeable that although the

1.9 wt%[NH₄ReO₄]_pH1/P25 catalyst precursor could be reduced

at 4508C up to 89% metallic Re (Figure 4), the reduction of the

bimetallic catalysts prepared by SI under the same conditions

resulted in only 27–34% Re⁰, regardless of if the initial state of

Re was Re⁷⁺in the precursors or a mixture of Re⁴⁺, Re⁶⁺, and

Re⁷⁺in the “as-prepared” solids (Table 1, entries 7 and 12). A

supplementary experiment was performed in which the “as-

prepared” 2.6 wt%Re-Pd/P25-SI catalyst was oxidized initially

under pure O₂(100% Re⁷⁺) and then reduced at 4508C for 3 h.

Again, only 36% Re⁰could be detected in addition to 64%

Re³⁺(Table 1, entry 14). The presence of Pd limits the reduction

of the oxidized Re species.

Exposure to air partially reoxidized the bimetallic catalysts to

give ꢀ60% Pd⁰and ꢀ40% Pd²⁺(Figure 5a; Table 1, entries 7

and 12). Additionally, no detectable fraction of Re⁰was ob-

served in either catalyst. Three different Re BE states were

shown that correspond most closely to Re⁴⁺, Re⁶⁺, and Re⁷⁺

reference standards. After curve fitting, variable surface propor-

tions were calculated (16% Re⁴⁺, 41% Re⁶⁺, and 43% Re⁷⁺on

DT51, and 29% Re⁴⁺, 24% Re⁶⁺, and 47% Re⁷⁺on P25; Fig-

ure 5b). Thus, as observed for the monometallic 1.9 wt%Re/

P25, the Re species in the bimetallic catalysts are oxidized on

exposure to air. These observations are consistent with the

known oxophilic character of Re, which forms strong bonds

with the oxygen atoms of the support and stabilizes rhenium

oxide.^[22]

The oxidation of Pd upon exposure to air was reversible as

only the peak that corresponds to Pd⁰was detected in both

ChemCatChem 2015, 7, 2161 – 2178

www.chemcatchem.org

2166

ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers

The in situ reduction of the precursor 2.6 wt%[NH₄ReO₄]-Pd

on P25 and DT51 and prepared by SI gave a Re/Pd atomic

ratio of 0.9 and 1.07, respectively (Table 1, entries 6 and 11). Ex-

posure to air increased this ratio to 1.32 and 1.37, respectively

(Table 1, entries 7 and 12), which suggests some migration of

Re on the surface. A subsequent reactivation in H₂at 4508C

gave ratios of 1.38 and 2.18, respectively (Table 1, entries 8 and

13), which are different from those observed after the direct re-

duction of the precursors.

terface. However, the addition of Re to 1.9 wt%Pd_KCl/DT51 by

CR decreased the Pd/Ti ratio from 0.038 to 0.027 (Table 1, en-

tries 5 and 9). This implies that on this parent solid, Re might

have been deposited on Pd by CR. However, another factor to

consider for this catalyst is the experimental problem men-

tioned previously using this support.

TPR

The TPR of samples preoxidized under O₂at 3008C was per-

formed up to 7008C. TPR profiles of the series of catalysts pre-

pared on Pd_Cl/P25 are shown in Figure 6, and the profiles for

a series prepared on DT51 are shown in Figure 7. The results

are summarized in Table 2.

Bimetallic Re-Pd catalysts prepared by CR

Some bimetallic catalysts prepared by the CR method were an-

alyzed similarly after storage in air and/or after reduction in

the XPS cell at 4508C (Table 1).

The “as-prepared” 1.7 wt%Re-Pd_Cl/P25-CR catalyst contained

Re as a mixture of Re⁴⁺, Re⁶⁺, and Re⁷⁺; the highest oxidation

state species Re⁷⁺was the most abundant (63%, Table 1,

entry 19). Conversely, Pd was 66% in the zero-valent state, and

the remainder was Pd²⁺(18%) and Pd⁴⁺(16%). Compared to

the “as-prepared” Pd monometallic parent that contained only

34% Pd⁰, the presence of Re in interaction with Pd seems to

limit the reoxidation of Pd. However, as in the SI-prepared cat-

alysts, the shift of the BEs for Pd⁰and Pd²⁺to lower energies

on one part and of that of Re⁴⁺and Re⁶⁺to higher energies on

the other part compared to the “as-prepared” 2 wt%Pd_Cl/P25

and 1.9 wt%Re_pH1/P25 was small (0.2–0.3 eV), within the experi-

mental errors. Therefore, there is no evidence of an electronic

transfer from Re to Pd in this catalyst. Although the other “as-

prepared” catalysts were not characterized, one may expect

the same results.

XPS data showed that after O₂treatment at 3008C, Re was

totally oxidized to Re⁷⁺, for instance in 0.9 wt%Re-Pd_Cl/P25-CR,

3.6 wt%Re-Pd_Cl/P25-CR, and 2.7 wt%Re-Pd_KCl/P25-SI (not

shown). Pd was oxidized to Pd²⁺(62–80%) and Pd⁴⁺.

During the reduction of the oxidized 2 wt%Pd_Cl/P25 solid

only one H₂-consumption zone was clearly observed at room

temperature (Figure 6a), which was ascribed to the reduction

of oxidized Pd (n_H=2.710^À4molg_cat^À1; Table 2).^[20]In the case

2

of 1.9 wt%Re_pH1/P25, a wide consumption peak was exhibited

between 230 and 3408C, with a broad shoulder up to 4508C

(Figure 6b). According to the literature,^[25,29]H₂uptake in this

temperature range is attributed to the reduction of ReO_xoxi-

dized Re species and the reduction of the bulk titania support.

Quantitative analysis reveals that the measured H₂consump-

tion (2.310^À4molg_cat^À1) was lower than the theoretical con-

sumption calculated for the complete reduction of the Re⁷⁺

present in the calcined solid to Re⁰(3.610^À4molg_cat^À1) and

confirmed that Re could not be reduced totally to the metal

state, as shown previously by the XPS results. The reduction of

xwt%Re-Pd_Cl/P25-CR bimetallic catalysts displayed a large re-

duction peak at a low temperature^[20](Figure 6c and d). The

total amounts of H₂consumed were higher than that exhibited

by the Pd oxide in the monometallic catalyst, which suggests

that contact between Re and Pd was established (Table 2).

There is clearly a strong effect of Pd on the reduction tempera-

ture of Re⁷⁺. The amount of H₂consumed increased with the

addition of Re (0.9 to 3.6 wt%). Furthermore, a comparison of

3.6 wt%Re-Pd_Cl/P25-CR and 3.5 wt%Re-Pd_Cl/P25-SI prepared by

CR and SI, respectively, which contain very similar Re contents,

evidenced a lower H₂consumption at low temperature for the

SI-prepared solid, which nevertheless remained higher than

the consumption for the reduction of the Pd monometallic cat-

alysts (Table 2). This means that a smaller fraction of Re is in

contact with Pd in the SI-prepared catalyst. Accordingly,

a second reduction peak was detected at higher temperatures

(250–3008C) for the bimetallic SI catalysts (Figure 6e), which

could be attributed to the reduction of Re clusters isolated on

the TiO₂support.

Following in situ reduction in the XPS cell, Pd was in the

zero-valent state (>92%) whatever the nature of the support

and the precursor salt, whereas the Re species were affected

significantly by the method of preparation. If PdCl₂was used

as the precursor, most of the Re was present in the Re³⁺state;

the higher the amount of Re, the higher the proportion of

Re³⁺(Table 1, entries 18 and 20). However, if we start from

K₂PdCl₄, the nature of the support influenced the final reduc-

tion state of Re: 74% Re⁰and 26%Re³⁺, and 30% Re⁰and 70%

Re³⁺for 1.5 wt%Re-Pd_KCl/P25-CR (entry 15) and 1.9 wt%Re-

Pd_KCl/DT51-CR (entry 9), respectively. Similar to that of the cata-

lysts prepared by SI (except for 1.5 wt%Re-Pd_KCl/P25-CR), the

zero-valent state of Re was absent or represented only a minor

fraction and the remainder is Re³⁺. This probably means that

under reaction conditions in the aqueous phase under H₂pres-

sure at 1608C most of the Re is in an oxidized state.

The surface Re/Pd, Re/Ti, and Pd/Ti compositions were calcu-

lated from the XPS analysis (Table 1). In the Pd_Cl/P25 series that

contains Re deposited by CR, the determined atomic ratios of

Re/Ti and Re/Pd increased as the loading of Re increased from

0.9 to 1.7 wt% (Table 1, entries 18 and 20), which shows that

the Re loading increased at the surface. However, regardless of

the treatment and the loading of Re, the Pd/Ti ratio did not

change and was close to 0.036, which is the value in the

parent catalyst (Table 1, entry 17). This indicates that little Re is

located on the Pd, and Re was deposited at the Pd–support in-

The reduction of 2.7 wt%Re-Pd_KCl/P25-SI (not shown) also

gave a large reduction peak that was complete at around

408C (n_H=4.510^À4molg_cat^À1). In addition, a weak consump-

2

À1

tion of 0.610^À4molg_cataround 3258C was found for this

ChemCatChem 2015, 7, 2161 – 2178

www.chemcatchem.org

2167

ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers

support (Figure 7a). The reduc-

tion of oxidized Pd in

2 wt%Pd_KCl/DT51 was complete

after 408C (Figure 7b). An addi-

tional consumption between

240 and 5108C that corresponds

to the reduction of the support

catalyzed by Pd was also ob-

served.^[30]The TPR analysis of

1.9 wt% Re-Pd_KCl/DT51-CR (Fig-

ure 7c) and 3.4 wt%Re-Pd_KCl

DT51-SI (Figure 7d) revealed

consumption peak around

/

a

200–3008C in addition to the

consumption at room tempera-

ture. The consumption at room

temperature is likely attributed

to both Pd and Re in interaction

with Pd. The H₂consumptions at

higher temperatures were 0.5

10^À4and 1.910^À4molg^À1, re-

spectively, and were ascribed to

some Re that is not in contact

with Pd and that may be re-

duced as ReO_xalone.

Hydrogen chemisorption

The bimetallic catalysts prepared

by CR or SI from the three differ-

ent parent Pd catalysts were

characterized by hydrogen

chemisorption at 708C. If Pd is in

close interaction with Re, then

its

chemisorption

capacity

should be affected. As Re does

not chemisorb hydrogen,^[31]the

amount of chemisorbed hydro-

gen will quantify the fraction of

Pd atoms exposed to the surface

after Re deposition in the cata-

lysts studied.

The metallic Pd accessibility as

a function of the Re loading of

the bimetallic catalysts prepared

by the SI or CR methods from

the three monometallic parent

catalysts is shown in Figure 8.

The results of the parent mono-

metallic and blank solids are

given to help the interpretation.

The latter were treated as for

Figure 6. TPR results for the P_Cl/P25 series: a) 2 wtPd_Cl/P25, b) 1.9 wt%Re_pH1/P25, c) 0.5 wt%Re-Pd_Cl/P25-CR,

d) 1.7 wt%Re-Pd_Cl/P25-CR, e) 3.5 wt%Re-Pd_Cl/P25-SI. Samples were oxidized at 3008C before TPR analysis.

solid that could be caused by the reduction of some free iso- the deposition of Re but in the absence of the Re salt, that is,

lated Re oxides or monometallic Re particles. in acidified aqueous solution for the CR method or in neutral

The catalysts prepared on DT51 presented different profiles water for the SI method, and they are labeled blank. The effect

(Figure 7). The bare support showed a consumption around of the addition of Re on Pd is estimated systematically by com-

6408C that could correspond to the reduction of the bulk TiO₂parison with its corresponding blank sample.

ChemCatChem 2015, 7, 2161 – 2178

www.chemcatchem.org

2168

ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers

The monometallic Pd catalysts

showed different abilities to

withstand sintering during the

CR or SI treatment, depending

on the preparation method. The

accessibility to hydrogen of

blank 2 wt%Pd_KCl/P25-CR and -SI

solids obtained after the treat-

ment of 2 wt%Pd_KCl/P25 in an

acidic solution under H₂or in

neutral water, respectively, did

not change much (Figure 8a and

b). The values were 15, 18, and

15%, for the parent and the CR-

and SI-treated samples, respec-

tively. Similarly, although the dis-

persion of 2 wt%Pd_KCl/DT51 was

higher than that on Pd_KCl/P25,

the blank treatments did not

modify the metallic accessibility,

and the values were 27, 30, and

23%, for the parent and the CR-

and SI-treated solids, respective-

ly (Figure 8c and d). In contrast,

as described previously,^[20]the

CR blank protocol applied to

2 wt%Pd_Cl/P25 resulted in a sin-

tering of the Pd particles with

a

significant decrease of the

metallic accessibility from 33 to

17% (Figure 8e). However, expo-

sure of this parent catalyst to

the SI neutral solution under am-

bient air only slightly decreased

the accessibility from 33 to 26%

(Figure 8 f). The sintering of the

metallic particles in acidic

medium is in agreement with

results in the literature on the

preparation of such cata-

lysts.^[32–34]

Figure 7. TPR results for the Pd_KCl/DT51 series: a) DT51 support, b) 2 wt%Pd_KCl/DT51, c) 1.9 wt%Re-Pd_KCl/DT51-CR,

d) 3.4 wt%Re-Pd_KCl/DT51-SI. Samples were oxidized at 3008C before TPR analysis.

Table 2. H₂consumption determined from TPR profiles and XPS (after in situ oxidation at 3008C).

Catalyst

TPR data

XPS data

Pd⁴⁺/Pd²⁺

[%/%]

n_H10⁴

[molg_cat]

for Re⁷⁺reduction

OD Re^[b]

The reason why the blank

treatment by the CR protocol of

2 wt%Pd_Cl/P25 involved a signifi-

cant sintering of the metallic

2

À1

n_H10⁴

2

À1

[molg_cat

]

Low T

High T

2 wt%Pd_Cl/P25

1.9 wt%Re_pH1/P25

2.7^[a]

–

48/52

–

particles

compared

with

2.3^[a]

–

2.55

5.5

3.03

3.50

4.2

4.1

–

3.82

–

0.5 wt%Re-Pd_Cl/P25-CR

0.9 wt%Re-Pd_Cl/P25-CR

1.7 wt%Re-Pd_Cl/P25-CR

2.4 wt%Re-Pd_Cl/P25-CR

3.6 wt%Re-Pd_Cl/P25-CR

3.5 wt%Re-Pd_Cl/P25-SI

2.7 wt%Re-Pd_KCl/P25-SI

2 wt%Pd_KCl/DT51

3.2

3.8^[a]

4.1

4.2

5.1

4.1^[a]

4.5

3.1

3.8

3.6

38/62^[c]

31/69

25/75^[c]

22/78^[c]

20/80

–

0.2

0.9

1.6

1.8

2.8

–

2.3

–

2 wt%Pd_KCl/P25 may result from

different factors. The initial Pd

dispersion was much higher in

the first system, which there-

fore, may be more prone to sin-

–

1.5^[a]

0.6

0.8

0.5

1.9

20/80

–

tering. However,

metal–support

a

stronger

interaction,

1.9 wt%Re-Pd_KCl/DT51-CR

3.4 wt%Re-Pd_KCl/DT51-SI

linked to the impregnation con-

ditions, may explain the high

stability of the catalysts pre-

pared from K₂PdCl₄solutions.

–

[a] Results in Ref. [20]. [b] Average oxidation valence OD=7Àx, calculated from Re⁷⁺+x/2H₂!Re^7Àx+xH⁺, with

x=2n_H/n_Re₇₊. [c] Calculated assuming a linear correlation as a function of wt%Re.

2

ChemCatChem 2015, 7, 2161 – 2178

www.chemcatchem.org

2169

ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers

even after a 2.9 wt% Re loading.

Furthermore, if we consider an

initial accessibility of 18%, we

can calculate that approximately

0.6 wt% Re loading corresponds

to a theoretical monolayer cov-

ering of Pd. It is then likely that

the available Pd sites for direct

ReÀPd bonding become rapidly

saturated, and the Re moieties

were not only deposited on the

initial Pd surface of 2 wt%Pd_KCl

/

P25 but Re could also have been

deposited adjacent to Pd parti-

cles at the Pd–TiO₂interface,

and/or even as clusters or ag-

glomerates on particular sites.

The presence of Re on the sup-

port cannot be excluded. Never-

theless, a high fraction of Re

loading on the support that

would not be in contact with Pd

could certainly be excluded as

the preparation of a Re catalyst

with a nominal Re loading of

4 wt% by the CR method on the

bare P25 support (which con-

tains no Pd) led only to

a 0.3 wt% Re deposit on the

solid. Therefore, on the bimetal-

lic catalyst, it may be considered

that Re was deposited marginal-

ly on the support without inter-

action with Pd but mainly de-

posited at the Pd–support inter-

face in close vicinity to Pd.

Figure 8. Pd accessibility (H/Pd) of bimetallic catalysts prepared by CR (left) and SI (right) by the addition of Re

onto: a, b) Pd_KCl/P25, c, d) Pd_KCl/DT51, and e, f) Pd_Cl/P25.

In contrast to the CR protocol,

the loading of 2.7 wt% Re by

This second hypothesis is confirmed by the stability of the rela-

tively well dispersed Pd metallic phase in Pd_KCl/DT51 also ob-

tained by deposition–precipitation in alkaline medium.

The H/M ratio decreased with the addition of Re compared

to the blank treated parent catalysts for all series of catalysts

(Figure 8), which tends to demonstrate the presence of Re de-

posited on the Pd surface. However, this covering was more or

less important depending on the preparation method and on

the support used, which is discussed below.

the SI method on the same parent 2 wt%Pd_KCl/P25 (Figure 8b)

led to a significant decrease of the H/Pd ratio from 15 to 6%.

As the blank Pd_KCl/P25-CR and -SI solids showed similar disper-

sions (18 and 15%, respectively), a comparison with results for

the 2.9 wt%Re-Pd_KCl/P25-CR catalyst suggests that the covering

of the Pd surface by the SI method is more efficient than that

by the CR deposition. This is surprising if we consider the theo-

retical metal deposition by CR on the Pd surface.

Pd_KCl/DT51

Pd_KCl/P25

The results of H₂chemisorption for the bimetallic catalysts pre-

pared from the 2 wt%Pd_KCl/DT51 parent catalyst are illustrated

in Figure 8c and d. The variation of H/Pd for this series of CR

catalysts was different from the CR series prepared from the

2 wt%Pd_KCl/P25 catalyst. This time, even after the deposition of

only 0.4 or 0.8 wt% Re by CR, the Pd accessibility decreased

considerably from 30 to 14 and 12%, respectively (Figure 8c).

It then decreased slowly to 9% with further loading of Re up

On 2 wt%Pd_KCl/P25, the accessibility to hydrogen decreased

gradually from 18 (blank) to 16, 14, and 12% as Re loadings by

CR increased from 0 to 0.6, 1.5, and 2.9 wt%, respectively (Fig-

ure 8a). As the comparison was performed with the corre-

sponding blank sample, this effect should not be attributed to

a lower metal dispersion but to some coverage of the Pd sur-

face by Re. However, the decrease was not very significant

ChemCatChem 2015, 7, 2161 – 2178

www.chemcatchem.org

2170

ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers

to 3.0 wt%. The crystal structure of the support used might be

an important parameter. Although P25 is a mixture of rutile

and anatase phases with a specific surface area of 50 m²g^À1,

DT51 is constituted of anatase only and presents a specific

area of 88 m²g^À1. It cannot be excluded that the rutile phase

favors the selective deposition of Re at the Pd–support inter-

face. This may also result from the experimental problems en-

countered during the preparation using the CR protocol on

the DT51-supported catalyst mentioned above, that is, the

plugging of the frit and an inhomogeneous H₂flow, which

could lead to a different deposition. Finally, the loading of

3.4 wt% Re by SI also decreased the metallic accessibility sig-

nificantly, which decreased to 11 % (Figure 8d). This means

that the coverage of Pd by the deposition of 3.4 wt% Re by SI

on Pd_KCl/DT51 was less important than that of 2.7 wt% Re by

SI on Pd_KCl/P25. This difference might be because of the differ-

ent specific area of the supports. Notably, an approximately

3 wt% Re loading by both methods on this 2 wt%Pd_KCl/DT51

parent catalyst gave catalytic systems that present very close

H₂accessibility, that is, ꢀ10%.

Cyclohexane dehydrogenation

The catalytic activity for cyclohexane dehydrogenation at

2708C was used as a probe to investigate the exposed metal

surface as this reaction is known to be structure insensitive,

that is, the activity should be approximately proportional to

the amount of surface Pd atoms and not influenced by the

presence of Re.^[36–39]The deviation from linearity can be caused

by the differences in the electron density of Pd, which may be

affected by the addition of Re. The participation of Re on the

reaction should be negligible under the reaction conditions

chosen.^[40]

The TOF in cyclohexane dehydrogenation calculated from

the specific reaction rates based on the Pd dispersion values

(H/M) are illustrated in Figure 9 (values given for 0 wt% Re cor-

respond to those of blank samples).

Pd_Cl/P25

As for the series prepared from the P25 support, the deposi-

tion of increasing amounts of Re by CR on Pd_Cl/P25 only slight-

ly decreased the H/M ratio compared to that of the blank

treated solid: 16, 14, 12, and 11 % following the addition of 0.5,

0.9, 1.7, and 3.6 wt%, compared to 17%, respectively (Fig-

ure 8e), which evidences the deposition of Re on Pd. However,

as observed for Pd_KCl/P25, the H/Pd ratio remained very close

to the value of the blank (17%) in particular at low Re load-

ings. A limited sintering of the Pd particles caused by the CR

protocol (immersion of the catalyst in an acidic aqueous solu-

tion of pH 1 under H₂flow) in the presence of Re in solution

could be an explanation. Indeed, on Ge-Rh/SiO₂catalysts pre-

pared by the same CR protocol, the sintering process was less

important if compared with the blank treated Rh/SiO₂solid.^[35]

This behavior was explained by the consumption of adsorbed

activated hydrogen on the surface of the metallic Rh surface

during the reduction of the Re salt. Indeed, adsorbed hydro-

gen favors the sintering of metallic particles in the aqueous

phase.^[34]Then, at higher loadings of 3.6 wt% Re by the CR

protocol, the accessibility remained constant at ꢀ11 %. This

suggests that Re does not deposit uniformly on the Pd surface

but may occur at the particle–support interface and/or as ag-

gregates close to Pd as suggested above for the Re-Pd_KCl/P25-

CR catalysts, which thus leaves a non-negligible accessible Pd

metallic surface. Finally, the deposition of 3.5 wt% Re on Pd_Cl/

P25 by SI occurred on the entire available Pd surface and on

the support, as demonstrated by the dramatic decrease of the

hydrogen accessibility from 26 to 6% (Figure 8 f).

Figure 9. TOF values for cyclohexane dehydrogenation with a) xwt%Re-

Pd_KCl/P25, b) xwt%Re-Pd_KCl/DT51, and c) xwt%Re-Pd_Cl/P25.

The TOF values for the bimetallic catalysts supported on P25

(prepared from 2 wt%Pd_KCl/P25 or 2 wt%Pd_Cl/P25) that contain

Re deposited by CR were around 0.4 s^À1, which is very close to

that of blank Pd_KCl/P25-CR and blank Pd_Cl/P25-CR (0.41 and

0.38 s^À1, respectively). This results from the absence of elec-

tronic modifications of Pd in catalysts prepared under CR con-

ditions on P25, whatever the preparation method of the

monometallic catalyst. In contrast, the bimetallic catalysts that

also contain Re deposited on DT51 although prepared by CR,

showed a significant increase of the TOF to ꢀ0.7 s^À1, com-

pared with 0.28 s^À1for the blank CR solid. Moreover, the imple-

mentation of the SI protocol led to bimetallic catalysts with

high TOF values whatever the support (0.75, 0.64, and 0.55 s^À1

after the loading of 2.7, 3.4, and 3.5 wt% Re on Pd_KCl/P25,

Pd_KCl/DT51, and Pd_Cl/P25, respectively).

To explain these differences, in a first step, it was verified

whether a supported monometallic Re catalyst was able to

participate to the reaction. The 1.9 wt%Re_pH1/P25 solid after

in situ reactivation at 4508C gave a low but non-negligible

À1

specific activity of 0.84 molh

g

(compared to 1.5–

Re

À1

3.5 molh

g

Pd

for the Pd-based catalysts). It was demonstrat-

ed by XPS that this solid contained 15% Re⁰and 85% Re³⁺at

the surface of the catalyst (Table 1, entry 3). This would suggest

ChemCatChem 2015, 7, 2161 – 2178

www.chemcatchem.org

2171

ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers

that Re in the zero-valent state could effectively catalyze the

reaction.

the shift to higher energies with increasing Re content in the

bimetallic catalysts.^[44]

In a second step, we considered the reduction degree of Re

to the zero-valent state in the bimetallic catalysts as deter-

mined by XPS after in situ reduction at 4508C. In all reduced

samples, no modification of the electronic state of Pd could be

evidenced by XPS on the addition of Re. As shown previously,

depending on the support, the Pd precursor, and the deposi-

tion method, the reduced bimetallic Re-Pd catalysts examined

contained different proportions of Re⁰and Re³⁺(Table 1). How-

ever, until now we have not been able to establish a correlation

between the method of preparation and the resulting oxida-

tion degree of Re (Re deposited on Pd_KCl/DT51 by SI (2.6 wt%)

or CR (1.9 wt%), on Pd_KCl/P25 by

Catalytic performance

GBL and BDO are the main products of SUC hydrogenation.

THF, propanol, and butanol are produced as minor byproducts

(THF <10%; propanol, butanol <2%).

A comparison between the behavior of the catalysts for SUC

hydrogenation has been reported previously.^[19]Typical concen-

tration profiles observed for SUC hydrogenation in the pres-

ence of a Pd and a ReO_x-Pd catalyst are shown in Figure S8.

The catalytic results are shown schematically in Figure 10.

SI (2.7 wt%) was present as

ꢀ30% Re⁰and 70% Re³⁺). Re in-

troduced on 1.5 wt%Re-Pd_KCl

/

P25 by CR was more reduced

(74% Re⁰). However, the 0.9 wt%

Re and 1.7 wt% Re catalysts pre-

pared by CR on Pd_Cl/P25 con-

tained mainly or only 100%

Re³⁺. Nevertheless, it is noticea-

ble that catalysts that contain

Re⁰(ꢀ30%; except 1.5 wt%Re-

Pd_KCl/P25-CR, which surprisingly

contained 75% Re⁰) resulted in

Figure 10. Effect of various amounts of Re added to Pd on SUC hydrogenation. Left: selectivity to GBL (black),

BDO (light grey), THF (dark grey), others (white); right: unconverted (GBL+SUC) concentration after 48 h (open

square or circle). Reaction conditions: SUC (6 g) in water (114 g; 0.43 moll^À1), P_H=150 bar, T=1608C, 0.5 g cata-

2

a

significant increase in TOF

lyst, t=48 h.

(0.6–0.7 s^À1), which would again

suggest a possible contribution

to the dehydrogenation reaction

of metallic Re in bimetallic catalysts. Furthermore, if the activity

of catalysts that contain Re⁰is proportional to the amount of

Globally, a positive effect on the reaction rate of SUC to GBL

was usually observed as the amount of Re increased, though

the optimal amount using the CR method passed through

a maximum. Indeed, the time required to observe 50% conver-

sion of SUC to GBL in the presence of the different catalysts

was as follows: Pd_Cl/P25, 8 h; 0.5 wt%Re-Pd_Cl/P25-CR, 8 h;

0.9 wt%Re-Pd_Cl/P25-CR, 3 h; and 1.7 wt%Re-Pd_Cl/P25-CR, 5 h

for the Pd_Cl/P25 series; Pd_KCl/P25, 29 h; 0.6 wt%Re-Pd_KCl/P25-

metallic Re and if we consider the activity of 1.9 wt%Re_pH1/P25

À1

(15% Re⁰) of 0.84 molh

g

^À1, the activity attributed to Pd in,

Re

for example, 1.9 wt%Re-Pd_KCl/DT51-CR (30% Re⁰) should be

À1

1.22 molh

g

^À1, which would result in a TOF of 0.3 s^À1for

Pd

Pd⁰. This value is very close to the TOF of the monometallic

catalyst (0.28 s^À1). These results suggest that the higher activity

observed for some bimetallic catalysts in the dehydrogenation

reaction is just the sum of the contributions of both metallic

Pd and Re. The absence of change of the TOF in the absence

of Re⁰shows that the addition of Re does not modify the elec-

tronic state of Pd.

CR, 8 h; 1.5 wt%Re-Pd_KCl/P25-CR, 15 h; and 2.9 wt%Re-Pd_KCl

/

P25-CR, 12 h for the Pd_KCl/P25 series; Pd_KCl/DT51, 12 h;

0.8 wt%Re-Pd_KCl/DT51-CR, 3 h; and 1.9 wt%Re-Pd_KCl/DT51-CR,

12 h for the Pd_KCl/DT51 series. The time for the SI-prepared cat-

alysts was in the range 2.5–3 h except for 3.4 wt%Re-Pd_Cl/P25-

SI (7 h). The rate of GBL conversion to BDO is reflected by the

final selectivity as described in Figure 10. The most important

point is the remarkable effect on the selectivity to BDO.

In contrast, there are some data in the literature that dem-

onstrate a strong interaction in Re-Pd catalysts^[17,41]and other

Re-based bimetallic solids. A considerable interaction was also

concluded for reduced Re-Pd/Al₂O₃catalysts prepared by SI.^[42]

High-resolution transmission electron microscopy (HR-TEM)

images of Re-2 wt%Pd/C (Re/Pd molar ratio 0.6) prepared by

incipient wetness impregnation, drying, and reduction at

3008C showed Pd particles and some alloy bimetallic particles

with a lattice expansion associated with the incorporation of

Re into the Pd lattice.^[17]The formation of surface bimetallic

clusters on the surface of TiO₂was demonstrated in Pt-Re/TiO₂

and Pd-Re/TiO₂.^[43]The addition of Re to Pt/C resulted in a de-

crease in the electron density of both Pt and Re as shown by

Although the selectivity to BDO was almost null using Pd/

TiO₂catalysts, promotion by Re made the ring-opening and hy-

drogenation of the GBL intermediate to BDO possible. Howev-

er, the selectivity was strongly affected by the support, the

preparation method of the Pd catalyst, and the Re deposition

procedure. The optimum amount of Re promoter required by

the CR protocol (0.6–0.9 wt%) to allow high selectivities to

BDO was lower than that for the SI protocol (3–4 wt%). An in-

crease in the loading of Re by CR above this optimum value

has little effect on catalyst selectivity, although there was a gen-

ChemCatChem 2015, 7, 2161 – 2178

www.chemcatchem.org

2172

ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers

eral trend of decreasing conversion at high loadings. If we

used the SI protocol, a high Re content, high activity, and high

selectivity to BDO were achieved.

Structure and interaction between Pd and Re in the bimetallic

catalysts

The data obtained by the different techniques with regard to

the structure of the bimetallic catalysts and the nature of the

interaction of the two metals are sometimes inconsistent, how-

ever, the balance is in favor of the presence of some Pd–Re in-

teractions without significant electron modifications of Pd and

Re sites. We tried to highlight the differences, if any, when we

compared the different catalysts.

Discussion

Distribution of Re species in the catalysts

Reducibility of Re in the monometallic Re catalyst

The reducibility of Re in monometallic samples must be clari-

fied as a fraction of Re may be deposited on the support

According to the results of H₂chemisorption, the deposition

of Re onto the Pd catalyst decreased the hydrogen accessibility

of Pd, although significant differences were observed. If we

used the CR protocol, the decrease was moderate, which indi-

cates a very low coverage of the Pd surface with ReO_xspecies.

In contrast, the exposed Pd surface was modified significantly

using the SI protocol, which suggests a higher coverage of Re

on the Pd surface for a given loading of Re.

during bimetallic preparation. The TPR of the 1.9 wt%Re_pH1

/

P25 monometallic catalyst revealed a large peak of H₂con-

sumption at ꢀ3308C attributed to the reduction of various

oxidized Re species (Figure 6). Quantitative analysis by XPS

showed that the solid after reduction is still composed of

a complex mixture of surface species. Only 15% Re⁰was identi-

fied after the rereduction of the “as-prepared” solid on P25

(Figure 4c, Table 1, entry 3), compared to 89% Re⁰following

the reduction of the precursor (Figure 4a, Table 1, entry 1),

which suggests that the reduction of a mixture of partly oxi-

dized dispersed Re species is more difficult than the direct re-

duction of Re⁷⁺to Re⁰. The finding that a fraction of Re is not

completely reduced to the metallic state even after reduction

at 4508C is consistent with the results of the reducibility of

rhenium oxide in supported monometallic catalysts prepared

from NH₄ReO₄, which has been debated in the literature,

mainly on alumina.^[25]Re is oxophilic and may bond strongly to

metal oxide surfaces. The reducibility is a function of the Re

loading (the interaction might be very strong at low concen-

tration and decrease with the increasing coverage of Re), the

calcination temperature, the adsorbed moisture, and the

strength of the interaction of the Re clusters with the O atoms

of the support (which decreases in the order Al₂O₃>SiO₂~

ZrO₂>TiO₂).^{[22,25,29,45–47]}Bare et al. used several techniques to

show that Re is present as a mixture of species (unreduced

Re⁷⁺species, small clusters, and isolated atoms) after the re-

duction of 0.7 wt%Re/g-Al₂O₃.^[25]It was also suggested by ex-

tended X-ray absorption fine structure analysis (EXAFS) that

a Re/Al₂O₃catalyst after reduction at 4008C contained a fraction

of rhenium oxide in addition to highly dispersed metallic parti-

cles on the support surface.^[45]

It was difficult to detect Re particles by TEM because of the

high dispersion of Re species. Nevertheless, the EDX analysis

confirmed the presence of both Pd and Re in small or large

analysis zones. However, STEM in dark-field mode at high reso-

lution and the corresponding analysis was able to image the

Re nanoclusters (<0.7 nm) located everywhere on the Pd parti-

cles in the bimetallic catalyst and on the support. The deposi-

tion of Re located specifically on Pd could not be visualized.

The presence of Re clusters all over the catalyst is consistent

with the low Re coverage of the Pd surface of Pd_KCl/P25 with

increasing loadings of Re, even at 1.5 wt% Re addition

(Figure 8).

In the impregnated materials and after oxidation, Re was

present as Re⁷⁺only (XPS data). TPR profiles of preoxidized

samples showed that the H₂consumption at low temperature

in the bimetallic catalysts was higher than that in the parent

catalysts, which suggests that some of the Re oxide might be

reduced at lower temperatures compared with the monome-

tallic Re catalyst, if it interacts with Pd or in close vicinity of Pd.

In the Pd_Cl/P25 catalyst series, the amount of H₂consumed in-

creased with the amount of Re deposited by CR from 0.5 to

3.6 wt% (Table 2). However, even after reduction at 4508C, the

complete reduction of Re to the zero-valent state was not ob-

served by XPS, which signifies that even if Re reduction was fa-

cilitated by contact with Pd, all of the Re was not reduced to

the metal.

The reason why the Re species were more difficult to reduce

to the metallic state after exposure to air may be explained by

the weak interaction of Re with the support. It is possible that

upon exposure to ambient air these Re particles redisperse to

more highly dispersed surface rhenium oxides that could inter-

act more strongly with the support and be more resistant to

complete reduction. The metal oxidation state of Pt-Re sup-

ported on Al₂O₃and TiO₂was determined by in situ XANES in

solvents in autoclave reactors at 1208C under 10 bar H₂pres-

sure.^[41]It was harder to reduce Re on TiO₂than on alumina.

TiO₂seemed to have a fraction of Re not associated with Pt,

which is different from that on Al₂O₃. It was proposed that

some of the Re was not in contact with Pt, which makes it

harder to reduce. Additionally, the strong interaction between

ReO_xand TiO₂might play a role.

In the XPS spectra of Pd and Re in the bimetallic catalysts

measured in an in situ cell, there was no clear indication that

the addition of Re caused a shift of the BEs of reference mate-

rials, which excludes strong electronic perturbations. These re-

sults suggest that the vast majority of Re is located preferen-

tially in contact with Pd but without electronic interaction. The

sites are thought to consist of metallic Pd surface sites with

neighboring Re atoms. In the XPS spectra there is a trend that

the distribution of Re species shifts to more oxidized species

as the content of Re increases. This is shown, for instance, in

the comparison of 0.9 wt%Re-Pd_Cl/P25-CR and 1.7 wt%Re-Pd_Cl/

P25-CR (Table 1, entries 18 and 20).

ChemCatChem 2015, 7, 2161 – 2178

www.chemcatchem.org

2173

ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers

More information can be

gained if we determine the H₂

consumption during TPR, which

may be assigned to the reduc-

tion of oxidized Re in addition to

the reduction of oxidized Pd. To

calculate these values, the oxida-

tion degree of Pd and Re before

TPR was determined by XPS

after O₂pretreatment at 3008C

in the XPS cell (which mimics

the oxidation treatment before

the TPR experiments); Re was in

its highest oxidation state of +7,

whereas Pd was present as Pd²⁺

and Pd⁴⁺. The Pd⁴⁺/Pd²⁺distribu-

tion of 48/52% in Pd_Cl/P25

changed to 31/69% and 20/80%

on the addition of 0.9 and

3.6 wt% Re by CR, respectively

(Table 2). This distribution was

20/80% in 2.7 wt%Re-PdKCl/

P25-SI. Re in contact with Pd

seems to prevent Pd from

Figure 11. Schematic representation of the ReO_x-Pd bimetallic catalysts.

strong oxidation, as mentioned above.

The calculated H₂consumptions for the reduction of Re⁷⁺

characterized both by TPR and XPS are reported in Table 2. A

Re average oxidation valence (OD) was deduced from these

values and is also reported in Table 2. The average OD of Re in

the CR- and SI-prepared bimetallic catalysts evolved between

3.0 and 5.5. The 1.9 wt%Re_pH1/P25 reduced catalyst showed

a lower OD (2.5) than the bimetallic catalysts. The CR-prepared

bimetallic catalyst with the lowest Re content (0.5 wt%) dis-

played the highest OD value (5.5), which suggests that Re spe-

cies in a highly dispersed state (probably in an atomic form)

are very difficult to reduce. For the other CR-prepared bimetal-

lic catalysts of the series, the OD tends to increase with the Re

addition on Pd_Cl/P25. This means that less Re comes into direct

contact with Pd as the Re loading increases. This hypothesis is

in agreement with the hydrogen accessibility, which reached

a plateau at 10% in this series. As the Re loading increases, the

Re species are reduced, although to a higher oxidation degree

if they are deposited close to Pd, however, without decreasing

the Pd accessibility; it will also be deposited as ReO_xnanoclus-

ters isolated on the support as observed clearly by HAAD-

STEM on 1.5 wt%Re-Pd_KCl/P25 (Figure 3). The reduction of

these ReO_xnanoclusters on the support was not detected by

TPR, which signifies that they may interact strongly with the

support. In contrast, in the catalysts prepared using the SI pro-

tocol, the reduction of the ReO_xclusters isolated on the sup-

port was detected by TPR, which indicates that these aggre-

gates are larger and do not interact with the support.

tion data and explains the performance trends observed with

the different catalysts (Figure 11). Alternative locations were

considered: (i) Re is deposited on the existing Pd crystallites,

(ii) Re is in the immediate vicinity of the existing Pd, (iii) Re is

on the support surface. If Re did not contribute to the de-

crease of Pd accessibility, this could suggest that the Re depos-

its as multilayer structure or is located on the support. The dis-

tribution of reduced Re surface species must also be influ-

enced strongly by the Re content.

First, let us consider the CR method, which relies on the for-

mation of Pd-H_ads, known to be a strong reductant, during the

bimetallic preparation. At the beginning of the deposition of

Re on Pd_Cl/P25 and Pd_KCl/P25 (Figure 11 a and e), the adsorbed

À

atomic hydrogen on Pd acts to reduce ReO₄to ReO_xspecies.

Notably, the Pd particles were shown to grow during the dep-

osition of Re by CR on Pd_Cl/P25. This initial fraction of Re

should be strongly immobilized. The stoichiometry of the reac-

tion of the deposition of these Re species is not demonstrated,

but one may expect low valences. However, XPS analysis indi-

cated that most of the time, only a minor part of Re was re-

duced fully to the metallic state. These highly reduced Re sur-

face species might be more resistant to reoxidation by expo-

sure to oxygen. As the Pd accessibility (as measured by chemi-

sorption) decreased only moderately with increasing amounts

of Re by CR in both series (Figure 8a and e) and as Re is

shown to be reduced in contact with Pd (as measured by TPR),

we propose that the deposition occurs of course on the Pd

surface because of the oriented method but also mainly at the

interface of the Pd particles and the support. Moreover, we

suggest that incremental deposition on the Pd particles occurs

as clusters by stacking on the initially deposited ReO_xspecies

rather than as isolated species, which would then take place

Schematic representation of bimetallic catalysts

We can now propose a structural model for the location of Re

on the catalyst surface that is consistent with the characteriza-

ChemCatChem 2015, 7, 2161 – 2178

www.chemcatchem.org

2174

ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers

without a significant decrease of the Pd accessibility. These ad-

ditional Re layers should be less reduced than those closest to

the Pd nanoparticles. This situation would be somewhat analo-

gous to that described by Koso et al.^[48,49]who examined

MoO_x-Rh and ReO_x-Rh/SiO₂in detail. Although the amounts of

CO adsorption decreased similarly as a function of the amount

of Mo or Re, EXAFS indicated that Mo was adsorbed as isolated

MoO_x, whereas Re was deposited as ReO_xclusters. As a result,

the optimum promoter/Rh ratio for activity in tetrahydrofurfur-

yl alcohol was not the same. With a further increase of the

amount of additive, ReO_xclusters might increase or be deposit-

ed all over the TiO₂support and yet leave some non-negligible

Pd accessibility according to HAAD-STEM and chemisorption

studies. All of the added Re will not promote the Pd catalyst

further, and the selectivity to BDO will decrease with too high

a loading (Figure 10).

pletely reduced to the metal state and significant amounts of

partially oxidized Re in the range Re³⁺–Re⁷⁺are present. This

may be indicative of an active role of oxidized Re during SUC

hydrogenation during the GBL reduction.

Supported bimetallic catalysts composed of a reducible

noble metal (Pt, Ir, Rh, Pd) and an oxophilic metal (such as Re)

have been described for the conversion of biomass-derived

feedstock, the aqueous-phase reforming of biomass-derived

oxygenates,^[52,53]secondary CÀO bonds hydrogenolysis in

a broad range of polyols including glycerol^[54–58]and cyclic

ethers,^{[48,49,56,59,60]}and for amide,^[61]acid,^[62]and amino acid^[63]

hydrogenation. As described in recent publications, Re promo-

tion may generate a bifunctional catalyst that possesses both

metal and Brønsted acid sites. The presence of acid sites on

ReO_x-promoted metallic catalysts was revealed by NH₃-TPD

profiles.^[53,57]NH₃-TPD measurements also revealed that the

acid site density increased with the Re loading of Pt-ReO_xcata-

lysts.^[64]Such an acid–metal bifunctional mechanism has been

proposed for the hydrogenation of tertiary amides^[61]or (hy-

droxymethyl)tetrahydrofuran derivatives over bimetallic cata-

lysts composed of a noble metal (Rh or Pt) and Re promot-

er.^[59,65]Under hydrothermal conditions, hydroxylated Re sur-

face species are able to activate ring-opening by deprotona-

tion and interaction with the reactant molecules with proton

donation followed by metal-catalyzed hydrogenation.

The tendency followed by the catalysts synthesized by the

CR deposition method on Pd_KCl/DT51 was different from the

case of the two previous solids (Figure 11 c). Even a deposition

as low as 0.4 wt% Re decreased the Pd accessibility by a factor

of two (Figure 8c), which suggests that the Pd surface was

very efficiently covered by Re species. Increasing amounts of

Re were added in a way that scarcely modified the available

Pd surface to chemisorption, either as aggregates on the metal

or as clusters on the support. However, these samples should

be treated with some caution as one must be aware of the ex-

perimental problems encountered during the synthesis.

In the case of the SI protocol applied to the three monome-

tallic catalysts (Figure 11 b, d, and f), the important point is that

the Pd accessibility decreased more significantly for ꢀ3–

3.5 wt% Re than in the case of the materials synthesized by

the CR protocol. This could signify that modification of Pd by

the Re species occurs differently, for instance as ReO_xclusters

that cover a relatively large area of Pd atoms. Even at low load-

ings, Re is deposited randomly on the Pd particles and on the

support. We propose that Re is deposited on Pd as island clus-

ters rather than as multilayer clusters, which was the case for

the oriented CR method. Consequently, the Pd accessibility will

decrease more rapidly and, as shown by the catalytic tests

(Figure 10), a higher amount of Re is needed to obtain high

yields of BDO. The ReO_xon the support might be a larger

amount and might not interact with the support, which ex-

plains that some areas are reduced at a higher temperature

than the smaller clusters obtained by CR as observed by TPR

(Figure 6).

This type of mechanism might be considered for our system

(Scheme 2). The conversion of SUC to BDO occurs in two steps

(Figure S8). SUC is converted to GBL, which is subsequently

transformed to BDO and THF after nearly complete conversion

of GBL. Significant hydrogenolysis of GBL was only observed in

the presence of the bimetallic catalysts, which highlights the

essential role of Re in this step. The activation of GBL on pro-

tonated Re bound to Pd may occur through both the oxygen

in the cycle and the carbonyl group to yield an ester inter-

mediate. The ring-opening of this cycle then takes place in

water, and further hydrogenation over Pd will yield BDO. The

active sites are expected to be at the interface between Pd

and the ReO_xclusters. Activation through only the carbonyl

group would yield THF through the stepwise conversion of

C=O to CÀOH and then CÀOH cleavage. The latter pathway

would require a higher activation barrier energy for adsorption

and hydrogenolysis of the C=O over a Re-Pd surface.

Another important point is that, notwithstanding extensive

characterization, the specific nature of the most active sites is

still unclear. Poor information is still available on the impor-

tance of the location or which oxidation valence is the most

important to control the selectivity to BDO. Overall, there does

not appear to be a consistent trend between the fraction of Re

that has been reduced to Re⁰and the final selectivity to BDO.

Additional studies using EXAFS and CO chemisorption could

help to gain a better understanding of the local structure and

the degree of coverage of Pd with ReO_xspecies, as shown by

Tomishige et al. for ReO_x-Pt and ReO_x-Ir catalysts.^[50,51]

Tentative mechanism

One of the most evident fea-

tures is that in all formulations

of bimetallic catalysts, either “as-

prepared” or even after reduc-

tion at 4508C, Re is not com- Scheme 2. Schematic reaction pathway to BDO.

ChemCatChem 2015, 7, 2161 – 2178

www.chemcatchem.org

2175

ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers

Indeed, 2.6 wt%Re-Pd_KCl/DT51-SI and 1.9 wt%Re-Pd_KCl/DT51-CR,

which are reduced to the same extent of 30% Re⁰, gave very

different selectivities of 78 and 26%, respectively. The

1.9 wt%Re-Pd_KCl/DT51-CR catalyst was particularly poorly active

for both SUC transformation and GBL hydrogenation. This be-

havior is not related to the degree of reduction as 1.5 wt%Re-

Pd_KCl/P25-CR in a highly reduced state (70% Re⁰) was also

poorly active. The catalysts that contained a lower fraction of

Re⁰(0.9 wt%Re-Pd_Cl/P25-CR and 1.7 wt%Re-Pd_Cl/P25-CR) led to

moderate BDO yields.

nation of succinic acid by bimetallic catalysts prepared by suc-

cessive impregnation or catalytic reduction are a result of the

different structures of ReO_xclusters on the Pd surface. In the

catalysts prepared by successive impregnation, ReO_xwas

formed as islands, whereas in those prepared by catalytic re-

duction, the formation of multilayer ReO_xwas orientated by

stacking.

Experimental Section

An important parameter might be the acid strength of the

generated sites. It was shown that the acid strength was

higher for OH sites bound to multiple oxophilic metal atoms in

a threefold configuration rather than OH sites adsorbed in an

on-top configuration on one oxophilic metal.^[65]This hypothe-

sis is consistent with the suggestion of a Re deposition by

stacking using the CR protocol based on physicochemical char-

acterizations. The ReO_xclusters may yield acid sites with

a higher strength that might be less favorable for BDO selectiv-

ity.

Preparation of catalysts

The catalysts used in this study were TiO₂-supported Pd and Re-Pd

catalysts. Two commercial supports were used: TiO₂DT51 (specific

surface area 88 m²g^À1, mean mesopore diameter 12 nm, supplied

by Cristal) and TiO₂P25 (specific area 50 m²g^À1, supplied by Degus-

sa-Evonik). The Re-modified catalysts are referred to as bimetallic

catalysts throughout the manuscript, although a real alloy between

the components was not evidenced. They were prepared from the

synthesized monometallic Pd catalysts. Ammonium perrhenate

(NH₄ReO₄) was used as the source of Re.

Finally to conclude, we should be cautious that essentially

all the characterizations of the solids were performed ex situ

and might not be an exact representation of the structure and

the interactions of these metals after reduction in the working

catalysts under hydrothermal conditions. The question remains

whether the metallic phases in the bimetallic catalysts are

maintained during aqueous-phase processing. If we consider

the structure of the bimetallic catalysts after synthesis, the re-

duced state might be different from that actually present in

the hydrothermal environment under H₂. There are some indi-

cations in the literature that the reduction degree should

remain intact. Characterization by in situ XANES in water

showed that Re in PtRe at a high temperature of 2258C re-

mained reduced^[44]and that Re in RhRe/C did not oxidize

under aqueous-phase hydrogenolysis.^[57]The elucidation of the

active site and the reaction mechanism deserves further inves-

tigation.

The 2 wt%Pd_Cl/P25 catalyst was prepared using PdCl₂as precursor

salt by an impregnation method performed in an acidic aqueous

medium (pH 1, HCl 32 wt%) as described previously.^[20,21]After the

impregnation step, the solvent was evaporated, and the solid was

dried in an oven at 1208C, treated by calcination under air at

3008C for 4 h, and finally reduced in flowing H₂(3.6 Lh^À1) at

3008C.

The 2 wt%Pd_KCl/P25 and 2 wt%Pd_KCl/DT51 catalysts were prepared

by deposition–precipitation using K₂PdCl₄as the precursor.^[21]An

appropriate amount of the precursor salt was slowly added to

a suspension of the support (10 g) in water. After 30 min stirring,

the pH was adjusted and maintained at pH 11 by the addition of

KOH. The suspension was heated to reflux at 1008C for 1 h. After

cooling, the solid was collected by filtration and washed thorough-

ly several times by dispersing it in water (200 mL), decantation,

and centrifugation until the supernatant had a neutral pH. The

solid was dried under vacuum at 508C overnight, reduced in flow-

ing H₂(3.6 Lh^À1) at 3008C for 3 h, and finally passivated under

1 vol% O₂in N₂(1.8 Lh^À1, 30 min).

Conclusions

Monometallic Re catalysts were prepared using aqueous solutions

of NH₄ReO₄as the precursor. The 1.9 wt%Re_pH1/P25 catalyst was

prepared by impregnation of TiO₂P25 in acidic medium (pH 1),

evaporation, drying in an oven at 1208C, and reduction at 4508C

in H₂(to mimic the CR deposition),^[20]whereas 3.1 wt%Re_pHn/DT51

was prepared using the same procedure, except the pH of the pre-

cursor aqueous solution was not adjusted (to mimic the SI deposi-

tion).

Different techniques (XRD, TEM, high-angle annular dark-field

scanning transmission electron microscopy, X-ray photoelec-

tron spectroscopy, temperature-programmed reduction, H₂

chemisorption, and the cyclohexane dehydrogenation model

reaction) were used to characterize ReO_x-Pd/TiO₂catalysts pre-

pared by using two different methods of Re promotion (suc-

cessive impregnation or catalytic reduction) from three differ-

ent monometallic Pd/TiO₂catalysts. Significant interactions be-

tween the two metals were demonstrated, although without

electronic modification. Monomeric structures of ReO_xwere

evidenced by high-angle annular dark-field scanning transmis-

sion electron microscopy everywhere over the catalyst. The

mechanism proposed for the selective hydrogenation of suc-

cinic acid to 1,4-butanediol highlights the role of oxidized Re

clusters bound to Pd in the activation of g-butyrolactone, ring

opening, and hydrogenation to 1,4-butanediol. It is also pro-

posed that the different catalytic performances in the hydroge-

The supported Re-Pd bimetallic catalysts were prepared from the

parent Pd catalysts by the addition of different amounts of Re

using two different methods: SI and ex situ CR. In the SI method,

an aqueous solution of a predetermined amount of the Re salt was

added to a suspension of the parent Pd catalyst at RT. After 5 h

stirring, the excess water was evaporated by using a rotary evapo-

rator at 508C under vacuum. The sample was dried overnight at

508C under vacuum, reduced in H₂flow at 4508C, and finally passi-

vated in 1 vol% O₂in N₂. The CR deposition method relies on a sur-

face redox reaction between adsorbed atomic hydrogen on Pd

(Pd-H_ads) in the parent catalyst and the Re precursor according to

the overall reaction [Eq. (1)]:

ChemCatChem 2015, 7, 2161 – 2178

www.chemcatchem.org

2176

ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers

ReO₄^Àþn Pd-H_adsþð8ÀnÞ H^þ! ðPdÞ_n-Re^ð7ÀnÞþþ4 H₂O

software. Quantitative determinations were performed using the

Vision KRATOS software. The Re4f_7/2component of the Re4f char-

acteristic doublets for each species was used to determine the BE

and the chemical state of the Re species. The Re4d and Pd3d

peaks were used to calculate the Re/Pd atomic ratios.

ð1Þ

This method should direct the deposition of Re onto the pre-exist-

ing Pd particles.^[66]The parent Pd/TiO₂catalyst was placed in

a fixed-bed reactor, and after purging with inert gas, the solid was

reactivated at 3008C for 1 h and then cooled under a H₂flow

(3.6 Lh^À1). Then, the Re aqueous solution (acidified to pH 1 with

HCl and degassed with an inert gas) was poured onto the Pd cata-

lyst. The mixture was maintained for 1 h under H₂bubbling at RT.

The solution was filtered, and the catalyst was dried overnight at

1008C in the fixed-bed reactor under H₂, before reduction in H₂

flow (3.6 Lh^À1) at 4508C. The reduced catalysts were allowed to

cool under H₂. Once RT was reached, the catalyst was flushed with

N₂, followed by slow exposure to air.

H₂-TPR measurements were performed in a flow apparatus after

in situ pretreatment under pure O₂for 1 h at 3008C and after cool-

ing to RT before the reduction under a 1.0 vol% H₂/Ar gas mixture.

The temperature range was 25–7008C with a ramp of 58Cmin^À1

.

H₂consumption during the reduction was detected by using an

AutoChem II/Micromeritics apparatus using a thermal conductivity

detector.

The metallic accessibility was measured by H₂chemisorption using

a pulsed technique at 708C as described previously^[20](Supporting

Information). This temperature was chosen to avoid the formation

of a b-hydride phase. Re does not chemisorb hydrogen under

these conditions.

The SI catalysts were exposed to air and stored under Ar before

they were used for characterization or a reaction test. The catalysts

prepared by CR were stored in ambient air. They are denoted “as-

prepared” bimetallic catalysts.

Activity in cyclohexane dehydrogenation to benzene was also used

to characterize the metallic function. The reaction was performed

under atmospheric pressure in a continuous flow reactor at 2708C.

An injection of cyclohexane (2 mLh^À1) was made using a calibrated

motor-driven syringe. The partial pressures were 97 and 3 kPa for

H₂and cyclohexane, respectively. All measurements were per-

formed with a total flow rate of 100 mLmin^À1with 20 mg catalyst

to ensure a conversion lower than 13%. Samples were first reacti-

vated at 3008C (3.6 Lh^À1H₂). Analysis of products was performed

by GC with a flame ionization detector by using a HP-PLOT Al₂O₃

column (50 m).

Characterization of catalysts

The actual bulk Pd and Re loadings in the catalysts were deter-

mined with an accuracy of Æ0.1 wt% by ICP-OES by using an

ACTIVA instrument (Horiba Jobin Yvon) after acid digestion with

H₂SO₄, HCl, and HNO₃.

XRD patterns of the catalysts were recorded in the range 2q=10–

808 (scan step 0.038, step time 2 s) by using a Bruker D5005 X-ray

diffractometer and a CuK_aradiation source (l=1.54056 ). Phase

identification was made by comparison with the JCPDS database.

Crystal sizes were estimated if possible using the Scherrer equation

from the width at half maximum intensity.

Hydrogenation experiments

SUC with a purity>99% was purchased from Aldrich Chemical Co.

All aqueous catalytic hydrogenation experiments were performed

TEM and STEM were conducted by using a JEOL 2010 instrument

with a LaB6 source operated at 200 kV coupled to a Oxford Link

ISIS instrument for microanalysis EDX spectroscopy and a JEOL

2010 FEG equipped with a field-emission gun operated at an accel-

eration voltage of 200 kV. ETEM was performed on some samples

to achieve a better resolution (1.36 ) by using a FEI TITAN G2 80–

300 kV instrument. HAADF-STEM was performed in the STEM

mode with EDX to obtain elemental analysis. PdL and ReM lines

that are free from interferences by signals from other elements

were used to quantify the relative amounts of Pd and Re, respec-

tively. For the analysis of individual particles, the electron beam

was focused to a diameter close to the diameter of the particle

under investigation. Typically, monometallic Pd catalysts were ob-

served after the samples were suspended in ethanol, sonicated,

and dispersed over a copper mesh grid with a carbon film coating,

and ethanol was allowed to evaporate. As rhenium oxo species are

formed that may be soluble in water or in ethanol upon exposing

the bimetallic catalysts to air, the bimetallic catalysts were dis-

persed dry on the carbon film with a syringe.

by using

a 300 mL high-pressure batch reactor (Parr 4560)

equipped with a magnetically driven impeller and a liquid sam-

pling system described previously^[18–21](Supporting Information).

The experiments were performed at 1608C to favor the formation

of BDO. Indeed, it has been described that to obtain a high BDO/

THF ratio, the reaction should be preferably performed at tempera-

ture close to 1508C, whereas temperatures close to 225–2408C will

achieve a higher THF/BDO ratio.^[14,17,67]

Acknowledgements

The financial support for this work was provided by ANR (French

Agence Nationale de la Recherche) within the Programme Chimie

et ProcØdØs pour le DØveloppement Durable CP2D 2009

(HCHAIB). The authors acknowledge AXELERA for the support and

the CLYM for guidance in the Ly-EtTEM project, which was finan-

cially supported by the CNRS, the RØgion Rhône-Alpes, the

“Grand Lyon”, and the French Ministry of Research and Higher

Education.

The surface chemical composition and oxidation states of Pd and

Re in catalyst samples were analyzed by XPS by using a hemispheri-

cal analyzer with a Kratos Axis Ultra^DLDspectrometer with a delay

line detector and charge neutralization system. XPS spectra were

collected by using a monochromated AlK_aX-ray source (hn=

1486.6 eV). The BEs were referenced to the C1s line at 284.5 eV.

XPS profiles of “as-prepared” catalysts and after in situ pretreat-

ment in O₂or H₂in a cell connected directly to the XPS chamber

were analyzed. After pretreatment, the samples were cooled to RT

and transferred without exposure to air into the ultra-high vacuum

chamber (10^À9mbar). Peak fitting was achieved using CASA XPS

Keywords: hydrogenation · palladium · rhenium · succinic

acid · supported catalysts

[1] C. S. K. Lin, R. Luque, J. H. Clark, C. Webb, C. Du, Biofuels Bioprod. Biore-

fin. 2012 , 6 , 8 8 –1 04.

ChemCatChem 2015, 7, 2161 – 2178

www.chemcatchem.org

2177

ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers

[2] J. J. Beauprez, M. De Mey, W. K. Soetaert, Process Biochem. 2010 , 45 ,

110 3 –1 114.

[38] J. A. Cusumano, G. W. Dembinski, J. H. Sinfelt, J. Catal. 1966 , 5 , 47 1 –

475.

[3] A. Cukalovic, C. V. Stevens, Biofuels Bioprod. Biorefin. 2008 , 2 , 50 5 –5 29.

[4] C. Delhomme, D. Weuster-Botz, F. E. Kuhn, Green Chem. 2009 , 11 , 1 3 –

26.

[5] J. Xu, B. H. Guo, Biotechnol. J. 2010 , 5 , 114 9 –1 163.

[6] S. E. Pedersen, J. G. Frye Jr., T. G. Attig, J. R. Budge, US 5 698 749, 1997.

[7] M. A. Mabry, W. W. Prichard, S. B. Ziemecki, US 4 609 636, 1986.

[8] U. G. Hong, S. Hwang, J. G. Seo, J. Yi, I. K. Song, Catal. Lett. 2010 , 138 ,

2 8 –3 3.

[39] M. Boudart, A. Aldag, J. E. Benson, N. A. Doughary, C. G. Harkins, J.

Catal. 1966 , 6 , 9 2 –9 9.

[40] C. L. Pieck, P. Marecot, J. M. Parera, J. Barbier, Appl. Catal. A 1995 , 126 ,

15 3 –1 63.

[41] J. Sµ, C. Kartusch, M. Makosch, C. Paun, J. A. van Bokhoven, E. Kleyme-

nov, J. Szlatchenko, M. Nachtegaal, H. G. Manyar, C. Hardacre, Chem.

Commun. 2011, 47, 6590–6592.

[42] A. Malinowski, W. Juszczyk, M. Bonarowska, J. Pielaszek, Z. Karpinski, J.

Catal. 1998 , 177 , 15 3 –1 63.

[9] U. G. Hong, J. Lee, S. Hwang, I. K. Song, Catal. Lett. 2011 , 141 , 33 2 –3 38.

[10] U. G. Hong, S. Hwang, J. G. Seo, J. Lee, I. K. Song, J. Ind. Eng. Chem.

2011 , 17 , 31 6 –3 20.

[43] Y. Sato, K. Terada, S. Hasegawa, T. Miyao, S. Naito, Appl. Catal. A 2005 ,

296 , 8 0 –8 9.

[11] U. G. Hong, H. W. Park, J. Lee, S. Hwang, I. K. Song, J. Ind. Eng. Chem.

2012 , 18 , 46 2 –4 68.

[12] U. G. Hong, H. W. Park, J. Lee, S. Hwang, J. Yi, I. K. Song, Appl. Catal. A

2012 , 41 5 –4 16 , 14 1 –1 48.

[13] U. G. Hong, J. K. Kim, J. Lee, J. K. Lee, J. H. Song, J. Yi, I. K. Song, Appl.

Catal. A 2014 , 469 , 46 6 –4 71.

[14] K. H. Kang, U. G. Hong, Y. Bang, J. H. Choi, J. K. Kim, J. K. Lee, S. J. Han,

I. K. Song, Appl. Catal. A 2015 , 490 , 15 3 –1 62.

[44] J. Okal, W. Tylus, L. Kepinski, J. Catal. 2004 , 225 , 49 8 –5 09.

[45] A. Ciftci, D. A. J. M. Ligthart, A. O. Sen, A. J. F. van Hoof, H. Friedrich,

E. J. M. Hensen, J. Catal. 2014 , 311 , 8 8 –1 01.

[46] A. S. Fung, P. A. Tooley, M. J. Kelley, D. C. Koningsberger, B. C. Gates, J.

Phys. Chem. 1991 , 95 , 22 5 –2 34.

[47] B. Mitra, X. Gao, I. E. Wachs, A. M. Hirt, G. Deo, Phys. Chem. Chem. Phys.

2001 , 3 , 114 4 –1 152.

[48] S. Koso, H. Watanabe, K. Okumura, Y. Nakagawa, K. Tomishige, Appl.

Catal. B 2012 , 111 –1 12 , 2 7 –3 7.

[15] J.-A. T. Schwartz, US 5478952, 1995.

[16] R. E. Bockrath, D. Campos, J-A. T. Schwartz, R. T. Stimeck, US 6008384,

1999.

[49] S. Koso, H. Watanabe, K. Okumura, Y. Nakagawa, K. Tomishige, J. Phys.

Chem. C 2012 , 116 , 307 9 –3 090.

[17] Z. Shao, C. Li, X. Di, Z. Xiao, C. Liang, Ind. Eng. Chem. Res. 2014 , 53 ,

963 8 –9 645.

[18] D. Pham Minh, M. Besson, C. Pinel, P. Fuertes, C. Petitjean, Top. Catal.

2010 , 53 , 127 0 –1 273.

[50] Y. Amada, H. Watanabe, M. Tamura, Y. Nakagawa, K. Okumura, K. To-

mishige, J. Phys. Chem. C 2012 , 116 , 2350 3 –2 3514.

[51] T. Ebashi, Y. Ishida, Y. Nakagawa, S.-i. Ito, T. Kubota, K. Tomishige, J. Phys.

Chem. C 2010 , 114 , 651 8 –6 526.

[19] B. K. Ly, D. P. Minh, C. Pinel, M. Besson, B. Tapin, F. Epron, C. Especel,

Top. Catal. 2012 , 55 , 46 6 –4 73.

[52] E. L. Kunkes, D. A. Simonetti, J. A. Dumesic, W. D. Pyrz, L. E. Murillo, J. G.

Chen, D. J. Buttrey, J. Catal. 2008 , 260 , 16 4 –1 77.

[20] B. Tapin, F. Epron, C. Especel, B. K. Ly, C. Pinel, M. Besson, Catal. Today

2014 , 235 , 12 7 –1 33.

[53] D. J. King, L. Zhang, G. Xia, A. M. Karim, D. J. Heldebrand, X. Wang, T. Pe-

terson, Y. Wang, Appl. Catal. B 2010 , 99 , 20 6 –2 13.

[21] B. Tapin, F. Epron, C. Especel, B. K. Ly, C. Pinel, M. Besson, ACS Catal.

2013 , 3 , 232 7 –2 335.

[54] L. Ma, D. He, Z. Li, Catal. Commun. 2008 , 9 , 248 9 –2 495.

[55] Y. Shinmi, S. Koso, T. Kubota, Y. Nakagawa, K. Tomishige, Appl. Catal. B:

Environmental 2010 , 94 , 31 8 –3 26.

[56] M. Chia, J. Pagan-Torres, D. Hibbits, Q. Tan, H. N. Pham, A. K. Datye, M.

Neurock, R. J. Davis, J. A. Dumesic, J. Am. Chem. Soc. 2011 , 133 , 1267 5 –

12689.

[57] M. Chia, B. J. O’Neill, R. Allamino, P. J. Dietrich, F. H. Ribeiro, J. T. Miller,

J. A. Dumesic, J. Catal. 2013 , 308 , 22 6 –2 36.

[58] Y. Nakagawa, X. Ning, Y. Amada, K. Tomishige, Appl. Catal. A 2012 , 43 3 –

433 , 12 8 –1 34.

[22] J. Okal, Appl. Catal. A 2005 , 287 , 21 4 –2 20.

[23] H. Iida, A. Igarashi, Appl. Catal. A 2006 , 303 , 19 2 –1 98.

[24] J. F. Moulder, W. F. Stickle, P. E. Sobol, K. O. Bomben, Handbook of X-ray

Photoelectron Spectroscopy, PerkinElmer Corporation, USA, 1992,

pp. 174–175.

[25] S. R. Bare, S. D. Kelly, F. D. Vila, E. Boldingh, E. Karapetrova, J. Kas, G. E.

Mickelson, F. S. Modica, N. Yang, J. J. Rehr, J. Phys. Chem. C 2011 , 115 ,

574 0 –5 755.

[26] M. Rønning, D. G. Nicholson, A. Holmen, Catal. Lett. 2001 , 72 , 14 1 –1 46.

[27] I. E. Wachs, G. Deo, M. A. Vuurman, H. Hu, D. S. Kim, J. M. Jehng, J. Mol.

Catal. 1993 , 82 , 44 3 –4 55.

[28] P. Arnoldy, E. M. Van Oers, O. S. L. Bruinsma, V. H. D. De Beer, J. A. Mou-

lijn, J. Catal. 1985 , 93 , 23 1 –2 45.

[59] K. Chen, S. Koso, T. Kubota, Y. Nakagawa, K. Tomishige, ChemCatChem

2010 , 2 , 54 7 –5 55.

[60] T. Buntara, I. Melian-Cabrera, Q. Tan, J. L. G. Fierro, M. Neurock, J. G.

de Vries, H. J. Heeres, Catal. Today 2013 , 210 , 10 6 –1 16.

[61] R. Burch, C. Paun, X.-M. Cao, P. Crawford, P. Goodrich, C. Hardacre, P. Hu,

L. McLaughlin, J. Sa, J. M. Thompson, J. Catal. 2011 , 283 , 8 9 –9 7.

[62] G. Beamson, A. J. Papworth, C. Philips, A. M. Smith, R. Whyman, J. Catal.

2011 , 278 , 22 8 –2 38.

[63] M. Tamura, R. Tamura, Y. Takeda, Y. Nakagawa, K. Tomishige, Chem. Eur.

J. 2015 , 21 , 309 7 –3 107.

[64] L. Zhang, A. M. Karim, M. H. Engelhard, Z. Wei, D. L. King, Y. Wang, J.

Catal. 2012 , 287 , 3 7 –4 3.

[29] M. A. Vuurman, D. J. Stufkens, A. Oskam, I. E. Wachs, J. Mol. Catal. 1992 ,

76 , 26 3 –2 85.

[30] C. A. Gonzµlez, A. N. Ardila, C. Montes de Correa, M. A. Martanez, G.

Fuentes-Zurita, Ind. Eng. Chem. Res. 2007 , 46 , 796 1 –7 969.

[31] H. C. Yao, M. Shelef, J. Catal. 1976 , 44 , 39 2 –4 03.

[32] C. Micheaud, P. Marecot, M. Guerin, J. Barbier, Appl. Catal. A 1998 , 171 ,

22 9 –2 39.

[33] M. T. Schaal, J. Rebelli, H. M. McKerrow, C. T. Williams, J. R. Monnier, Appl.

Catal. A 2010 , 382 , 4 9 –5 7.

[34] A. Douidah, P. Marecot, S. Labruquere, J. Barbier, Appl. Catal. A 2001 ,

210 , 111 –1 20.

[65] D. Hibbitts, Q. Tan, M. Neurock, J. Catal. 2014 , 315 , 4 8 –5 8.

[66] F. Epron, C. Especel, G. Lafaye, P. Marecot, Nanoparticles and Catalysis

(Ed.: D. Astruc), Wiley-VCH, Weinheim, 2008, pp. 279–302.

[67] E. I. Du Pont De Nemours, US 6008384, 1999.

[35] G. Lafaye, C. Micheaud-Especel, C. Montassier, P. Marecot, Appl. Catal. A

2002 , 230 , 1 9 –3 0.

[36] M. Guenin, M. Breysse, R. Frety, K. Tifouti, P. Marecot, J. Barbier, J. Catal.

1987 , 105 , 14 4 –1 54.

Received: February 27, 2015

[37] R. W. Maatman, P. Mahaffy, P. Hoekstra, C. Addink, J. Catal. 1971 , 23 ,

10 5 –1 18.

Published online on June 3, 2015

ChemCatChem 2015, 7, 2161 – 2178

www.chemcatchem.org

2178

ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Article Doi

Insights into the Oxidation State and Location of Rhenium in Re-Pd/TiO2 Catalysts for Aqueous-Phase Selective Hydrogenation of Succinic Acid to 1,4-Butanediol as a Function of Palladium and Rhenium Deposition Methods

DOI: 10.1002/cctc.201500197

Source and publish data:

Authors:

Article abstract of DOI:10.1002/cctc.201500197

Full text of DOI:10.1002/cctc.201500197

Products guided by the article

R&D Labs maybe for 109-99-9

Relevant to this article

Hot Product