DS

TMa

b

c

a

ARRAA

KHDPS

1

oa9tdthiol

tpdaArb

p

0d

Applied Catalysis A: General 386 (2010) 28–33

Contents lists available at ScienceDirect

Applied Catalysis A: General

journa l homepage: www.e lsev ier .com/ locate /apcata

irect synthesis of H2O2 in methanol at low pressures over Pd/C catalyst:emi-continuous process

eresa Morenoa,b, Juan García-Sernaa,∗, Pawel Plucinskib,aría Jesús Sánchez-Monteroc, María José Coceroa

High Pressure Processes Group, Chemical Engineering & Environmental Technology Department, University of Valladolid, Prado de la Magdalena sn, 47011 Valladolid, SpainDepartment of Chemical Engineering, University of Bath, BA 7AY Bath, UKDepartment of Physical Chemistry, University of Salamanca, Salamanca, Spain

r t i c l e i n f o

rticle history:eceived 15 December 2009eceived in revised form 9 July 2010ccepted 12 July 2010

a b s t r a c t

The production of H2O2 via direct synthesis has been studied in depth over the last decades, due tothe possibilities given by the discovery of active catalysts based on nanotechnology and selective activemetals. However, the process is also complicated because of the coexistence of three phases, where

vailable online 13 August 2010

eywords:2O2

irect synthesis

mass transfer between gas and liquid (solvent) limits the concentration of O2 and H2 in contact withthe solid catalyst sites, and subsequently the final H2O2 productivity. High pressures are normally usedto enhance mass transfer by increasing the solubility. In this work, we explore the influence of lowpressures, in order to optimise the reaction reducing mechanical requirements. Pressures from 0.1 to0.9 MPa in a semi-continuous reactor have been tested, obtaining H2O2 concentrations up to 1.33 wt.%

.9% t

d catalystlurry reactor

and a selectivity from 44been used.

. Introduction

Finding a green process for the production of hydrogen per-xide by direct synthesis as an alternative to the traditionalnthraquinone route has fascinated scientists for decades [1]. Over5% of the current industrial production of H2O2 is carried out viahe anthraquinone route or Riedl–Pfleiderer process which avoidsirect contact between H2 and O2 and offers continuous produc-ion at moderate temperatures [2,3]. Nevertheless, and this is aackneyed argument, large amounts of by-products are produced

n the traditional process, and the process requires several stepsf separation and concentration, consequently requiring a ratherarge energy input.

Analysing the chemistry, one can easily find that the direct reac-ion between hydrogen and oxygen would be the simplest way toroduce H2O2. However, the real reaction scheme is more complexue to the occurrence of simultaneous and consecutive reactions,ll of them thermodynamically favoured and highly exothermic.mong these reactions are: decomposition of hydrogen peroxide,

eduction of hydrogen peroxide and direct water formation by com-ustion.

The well-known safety issue associated with the direct synthesisrocess is the explosive nature of the hydrogen–oxygen mixtures.

∗ Corresponding author. Tel.: +34 983184934; fax: +34 983423013.E-mail address: jgserna@iq.uva.es (J. García-Serna).

926-860X/$ – see front matter © 2010 Elsevier B.V. All rights reserved.oi:10.1016/j.apcata.2010.07.019

o 69.0%. A nano-Pd/C commercial catalyst and methanol as solvent have

© 2010 Elsevier B.V. All rights reserved.

In the traditional route this problem is avoided because H2 and O2are not in direct contact. Under atmospheric pressure conditions,the explosive range falls between 4.0 and 95 mol% H2 at 25 ◦C [4].Although the direct synthesis could be conducted over the UpperFlammability Limit (UFL), the operation under Low FlammabilityLimit (LFL) is preferred for economic (H2 is the limitant reagent),safety (the lower the quantity of H2, the lower is the power of anundesired explosion) and process reasons (to minimise hydrogena-tion of H2O2). This forces the use of large amounts of diluents,usually nitrogen (pure or from air), although CO2 has also beenstudied [5].

An aqueous reaction medium provides the safest conditions[6–8]; however, its main drawback is the low solubility of the react-ing gases, which strongly limits mass transfer and thus the rate ofperoxide production. The use of alcohols [9,10] or alcohol–watermixtures [11,12] increases the solubility of the gases. As an exam-ple, Degussa/Headwaters recently announced the construction ofthe first pilot plant for direct synthesis of H2O2 to be used as an insitu source integrated into a propylene oxide plant, disclosing thatmethanol will be used as solvent [13,14].

The other main issue associated with the direct synthesis ofhydrogen peroxide is the discovery of an active and highly selective

catalyst. Most catalysts described in the literature are supported oncarbon, zeolites, Al2O3 or SiO2. Ntainjua et al. [15] studied the roleof the support, finding that supports with low isoelectric pointssuch as carbon and silica give the highest rates of synthesis. Theactive metals are usually palladium [16–18], gold [8,19] or com-

alysis

btthpiate[somtohtTsl[sp

(atmbds(q

piIcmmhsbp

2

2

auimpcSwfuc

XXT

T. Moreno et al. / Applied Cat

inations of Pd + Au [10,11,20,21] or Pd + Pt [9,21]. Unfortunately,hese catalysts are also responsible for the side reactions, namelyhe combustion of hydrogen to water and the decomposition ofydrogen peroxide. The role of size and crystalline type of the metalarticles in the selectivity to H2O2 is still uncertain, but the effect

s clear. Several authors claim the beneficial effects of nanoparticlend nanocrystals use [12,20,22]. In addition to the catalyst selec-ion, the use of promoters and H2O2 protective additives is crucial,specially when using monometallic Pd catalysts. Pospelova et al.23] early reported the beneficial effect of adding a mineral aciduch as HCl in order to inhibit the decomposition of the peroxidever a Pd catalyst. More recently, the role of these so-called pro-oters has been analyzed in detail, and it is now generally accepted

hat halide anions increase the selectivity towards hydrogen per-xide, while the role of protons is to facilitate the adsorption ofalide anions on the catalyst and to inhibit peroxide decomposi-ion by decreasing its adsorption on decomposition sites [6,24,25].he promoters are most commonly added homogenously to theolvent, although they might also be incorporated onto the cata-yst via catalyst modification with similar or slightly better results26]. On the other hand, Hutchings and his coworkers have demon-trated that it is possible to achieve high selectivities without anyromoters by combining metals, such as Au–Pd catalysts [20,27].

Finally, the existence of three phases in the reactor, i.e. gasesH2, O2 and an inert), liquid (solvent) and solid (catalyst), usu-lly involves important mass transfer limitations and complicateshe process. Mass transfer can be enhanced by: (1) increasing

ass transfer coefficient or reducing the thickness of stagnant filmy stirring, (2) increasing surface area of the bubbles using low-iameter-bubble spargers or injection systems, (3) increasing theolubility of the gases by increasing the partial pressure of the gasesHenry’s law); this implies the use of high pressures and subse-uently the use of more expensive equipment.

In a previous work, we have demonstrated the possibility ofroducing H2O2 in both H2 and methanol using CO2 and N2 as

nert gases at high pressures and moderate temperatures [28].n this work we demonstrate that H2O2 can be produced semi-ontinuously in methanol at low pressures, up to 0.9 MPa, andoderate temperatures, near ambient conditions, using a com-ercial Pd/C catalyst. The effect of key variables such as O2/H2,

alide/Pd and acid/Pd ratios, reaction time and pressure has beentudied. The catalyst has been characterized by TEM and SEM–EDXefore and after the use in order to elucidate the effect of the sup-ort and active metal.

. Experimental

.1. Materials and methods

The catalyst used in this study was fine particles with an aver-ge of 5 wt.% Pd over carbon support purchased from Aldrich andsed fresh for each experiment. It has been selected because it

s well-known as an active catalyst, commercially available andonometallic. Research grade oxygen, hydrogen and nitrogen were

urchased from BOC Gases (UK) and used without further modifi-ation. Methanol with a purity of 99.8% was used as solvent (Fischercientific). KI (Fluka), H2SO4 (Sigma) and Na2S2O3·5H2O (Fluka)ere used for iodometric titration of hydrogen peroxide [11]. Water

ormation was determined by coulometric Karl-Fischer titrationsing Hydranal-Coulomat reagents (Fluka) and a Mettler-Toledo

oulometer (DL37).

The bulk palladium phase in the catalyst was determined by-ray diffraction (XRD) method using a Holland Phillips PW/1710-ray generator with Cu K� radiation (� = 1.5406 A; 40 kV, 30 mA).he surface area and porosity of the catalysts were obtained by the

A: General 386 (2010) 28–33 29

BET method by measuring the adsorption of nitrogen at 77 K usinga Micromeritics ASAP 2010 analyzer. TEM analysis was performedusing a JEOL® JEM-1011 electron microscope (100 kV). SEM imageswere obtained with a FEI® Quanta 200 F ESEM microscope equippedwith Everhart–Thornley and backscattered electron detectors.

2.2. Experimental set-up

Reactions were performed in a 0.1 L AISI 316 SS agitated reactor.A crosshead magnetic stirrer bar (diameter 4.5 cm) and a magneticstirrer operating at 600 rpm were used. The gases (O2, H2 and N2)were bubbled directly into the reaction medium using Brooks®

mass flow controllers (Smart Mass Flow 5850). Pressure in the sys-tem was controlled by a Brooks® back pressure regulator (model5688).

2.3. H2O2 synthesis

All tests were carried out at room temperature and pressuresfrom 0.3 to 0.9 MPa. The reactor was initially passivated using35% HNO3 for 4 h to minimise H2O2 decomposition. The reac-tion medium was 0.03 L of methanol plus promoters (0.0125 molH3PO4 and 1.2 × 10−4 mol NaBr unless otherwise stated). The cata-lyst (3 × 10−2 g Pd/C) was carefully added to the liquid phase beforeintroducing the gases with a total flow of 0.125 L/min (care shouldbe taken to avoid spontaneous combustion of methanol in contactwith the catalyst). The H2 content of the mixture was kept to 4.0%mol, the O2/H2 ratios varied between 2.0 and 24.0 mol/mol, and thenitrogen flow was adjusted accordingly. When the desired pressurewas obtained, stirring was initiated in order to start the reaction.When the reaction was complete (reaction time 1.0 h unless oth-erwise stated), the products were quantified by iodometric (H2O2)and Karl-Fischer titration (water). Water content in the reactionmedium before catalyst addition was also determined prior to eachexperiment.

H2O2 yield and selectivity are defined as follows:

H2O2 yield (%) = H2O2 formed (mol)H2 added to reactor (mol)

(1)

H2O2 selectivity (%) = H2O2 formed (mol)H2 consumed (mol)

= H2O2 formed (mol)H2O2 formed (mol) + H2O formed (mol)

(2)

3. Results and discussion

3.1. Catalyst characterization

The bulk and surface properties of the fresh Pd/C catalyst areshown in Table 1. It is a highly microporous material, which pro-vides a high specific surface area (1314 m2 g−1) and a significantmicropore volume (0.503 cm3 g−1). SEM analysis of the catalystshowed a nanotube carbon structure (see Fig. 1). The size of thepalladium particles on the fresh catalyst was found to be ca. 5 nm(see Fig. 2). The presence of the bulk metallic Pd (Pd0) phase inthe catalyst was confirmed by X-ray diffraction (XRD) as shown in

Fig. 3.

A total of 15 runs have been carried out studying the effect ofpromoters, O2/H2 ratio, reaction time and pressure in yield, selec-tivity, conversion and peroxide concentration. A summary of theseexperiments is shown in Table 2.

30 T. Moreno et al. / Applied Catalysis A: General 386 (2010) 28–33

Table 1Surface and bulk properties of the Pd/C catalyst.

Pd loading (wt.%) Pd bulk phase Pd particle size (nm) Specific surfacearea (m2 g−1)

Micropore volume(cm3 g−1)

Total pore volume(cm3 g−1)

Average porewidth (nm)

5 Pd0 3–5 1314 0.503 0.684 1.229

Table 2Experimental summary for H2O2 production.a.

Run # Br− (103 mol L−1) Br/Pd H+ (mol L−1) H+/Pd O2/H2 Time (h) Pressure (MPa) H2O2 (wt.%) Yield (%) Select. (%) H2 conv. (%)

01 3.33 7.1 0.42 887 12 1.0 0.9 0.56 38.2 51.0 74.902 4.00 8.5 0.42 887 12 1.0 0.9 0.56 38.3 69.0 55.503 4.67 9.9 0.42 887 12 1.0 0.9 0.53 36.5 62.9 58.004 4.00 8.5 0.33 709 12 1.0 0.9 0.52 35.7 65.0 55.005 4.00 8.5 0.50 1064 12 1.0 0.9 0.55 38.0 60.5 62.806 4.00 8.5 0.42 887 2 1.0 0.9 0.29 20.1 48.6 41.307 4.00 8.5 0.42 887 10 1.0 0.9 0.27 18.7 43.1 43.308 4.00 8.5 0.42 887 15 1.0 0.9 0.37 25.7 55.6 46.209 4.00 8.5 0.42 887 24 1.0 0.9 0.37 25.7 58.9 43.610 4.00 8.5 0.42 887 12 2.0 0.9 0.92 31.7 58.8 53.911 4.00 8.5 0.42 887 12 3.0 0.9 1.03 23.5 48.5 48.512 4.00 8.5 0.42 887 12 4.0 0.9 1.33 22.9 47.0 48.613 4.00 8.5 0.42 887 12 1.0 0.3 0.35 23.8 44.9 53.014 4.00 8.5 0.42 887 12 1.0 0.5 0.42 28.9 51.0 56.715 4.00 8.5 0.42 887 12 1.0 0.7 0.51 35.0 59.9 58.4

a Reaction conditions: 3.0 × 10−2 g 5 wt.% Pd/C catalyst, 0.030 L MeOH, room temperature, 0.125 L min−1 total gas flow rate, 4.0 mol% H2 in gas flow.

Fig. 1. ESEM image of fresh 5 wt.% Pd/C catalyst.

Fig. 3. X-ray diffraction pattern obtained for 1%, 3% and 5% Pd/C catalyst.

Fig. 2. TEM images of fresh 5 wt.% Pd/C catalyst (left) and spent 5 wt.% Pd/C catalyst after run #01 (middle) and run #03 (right).

T. Moreno et al. / Applied Catalysis A: General 386 (2010) 28–33 31

pent 5 wt.% Pd/C catalyst after run #01 (middle) and run #03 (right).

3

ipapctHiaeetum#cppwTntrttwC

awrdaouc

w

3

dtaHm

Fig. 4. ESEM–BSED images fresh 5 wt.% Pd/C catalyst (left) and s

.2. Influence of promoters

The role of acidic and halide promoters has been thoroughlynvestigated in the literature. Choudhary and Samanta have pro-osed that the enhancement in selectivity obtained with halides ischieved due to halide–Pd interactions that modify the electronicroperties of Pd in the catalyst, resulting in the inhibition of theleavage of O–O bonds in both H2O2 and O2 during their adsorp-ion on the catalytic sites and thus avoiding H2 combustion and2O2 decomposition/hydrogenation; however, an excess of halide

ons results in catalyst poisoning, with the consequent decrease inctivity [29]. It is therefore crucial to adjust the amounts of promot-rs to optimise the process, and in our specific case promoters weressential since no hydrogen peroxide production was achieved inheir absence, in accordance with previous results by other authorssing the same Pd/C catalyst [30]. The effect of the amount of bro-ide in yield and selectivity towards H2O2 is shown in Table 2 (runs01–03). The best results were obtained for 4.0 × 10−3 mol L−1 Br−,orresponding to a Br−/Pd molar ratio of 8.5. Interestingly, in ourrevious work we found that the appropriate Br/Pd ratio for thisarticular Pd/C catalyst was 8.26, although the reaction conditionsere entirely different, i.e. high pressure and scCO2 as diluent [28].

his may indicate that the interaction between halide and Pd doesot depend on the reaction conditions but rather on the nature ofhe catalyst. Regarding the protons, it can be observed comparinguns #02, #04 and #05 (see Table 2) that the highest yield and selec-ivity were obtained when using 0.42 mol L−1 of H3PO4. In this case,he molar ratio acid/Pd is much higher than the one in our previousork (887 vs. 27), confirming the important acidification effect ofO2, which behaves as a promoter itself.

TEM images from the spent catalyst (runs #01 and #03, lowestnd highest Br− concentration, respectively) are shown in Fig. 2,here it can be observed that the particle size remains ca. 5 nm after

un #01 but increases to ca. 15 nm after run #03. This is probablyue to an agglomeration of palladium particles coupled with Br−

bsorption on the surface, since EDX analysis revealed the presencef Br− on the catalyst. These halide ions activate the catalyst, butnder high concentrations inhibit the H2 and O2 adsorption on theatalyst sites, reducing the H2O2 productivity.

The difference in particle size can also be observed in Fig. 4,hich shows SEM–BSED images of the catalysts.

.3. Influence of O2/H2 ratio

The reaction between H2 and O2 under the conditions studiedoes not only yield to H O as the only final product. H O is a reac-

2 2 2 2ion intermediate, and H2O is the final product. Thus, as indicatedbove, H2O2 decomposes itself and can also be reduced by H2 giving2O. The extent of these secondary reactions can be controlled byodifying O2/H2 ratios. An excess of O2 is recommended to max-

Fig. 5. Effect of O2/H2 ratio on yield (�) and selectivity (�.). Reaction conditions:0.9 MPa, time on stream 1 h, catalyst 3.0 × 10−2 g 5 wt.% Pd/C catalyst, 0.030 L MeOH,room temperature, 4 × 10−3 mol L−1 NaBr, 0.42 mol L−1 H+, 4.0 mol% H2 in gas flow.

imize H2O2 production. By testing molar O2/H2 ratios from 2.0 to24.0 (runs #02, #06–09) we have found that hydrogen conversionis similar in all cases except for a ratio of 12.0 (run #02), when selec-tivity and yield also reach their highest value (see Table 2). This canbe clearly observed in Fig. 5. Note that the results obtained witha ratio of 24.0 (i.e. in the absence of nitrogen) are very similar tothose obtained with a ratio of 15 in terms of yield and selectivity.Note that an industrial reaction using air as diluent (21 mol% O2and 79 mol% N2) and a 4 mol% H2 implies a O2/H2 = 5.04, which notbeing the optimum for process conditions could be the cheapestoption, although the use of enriched air is also an alternative.

3.4. Influence of reaction time

Due to the occurrence of above-mentioned secondary reactionsassociated with the hydrogen peroxide formed (i.e. decompositionand hydrogenation), reaction time is a crucial variable in the sys-tem under study (runs #02, #10–12). Although the concentrationof peroxide clearly builds up with time reaching 1.33 wt.% H2O2after 4 h (see run #12 in Table 2), this fact favours the decompo-sition and hydrogenation of the product, and both selectivity andyield are higher at shorter reaction times. For instance, after 3 h theconcentration of peroxide was increased by 84% with respect to a1 h reaction time, but the selectivity decreased by 30% and yielddecreased by 39% (see Fig. 6). However, after 4 h both yield andselectivity remained constant, although the peroxide concentration

continued to increase.

The selection of the optimum residence time will depend onthe minimum concentration required at reactor outlet, this will beconnected with the subsequent H2O2 concentration unit.

32 T. Moreno et al. / Applied Catalysis A: General 386 (2010) 28–33

F lectivc .0 mo

3

wi(a(gimV

tccf

ea#pHit

Fd0H

ig. 6. Effect of reaction time on: (A) H2O2 concentration and (B) yield (�) and seatalyst, 0.030 L MeOH, room temperature, 4 × 10−3 mol L−1 NaBr, 0.42 mol L−1 H+, 4

.5. Influence of pressure

The solubility of the reacting gases in the liquid solvent increasesith pressure and, for the case of a three phase slurry reaction, it

ncreases H2O2 yield. Eqs. (3–5) show the estimated solubility, ci*

mol L−1), of H2, N2 and O2 in the system depending on temper-ture, 20 ◦C < T (◦C) < 40 ◦C and the partial pressure of the gas, PiMPa) for total pressures up to 9.0 MPa and an initial mixture ofases with 4.0 mol% H2, 8.0 mol% O2 and the rest of N2. The solubil-ties have been estimated using Aspen Plus 2006® with the property

odel RKSMHV2 (Redlich–Kwong–Soave EoS with modified Huronidal mixing rules).

c∗H2

= (0.00107 · T−0.0737) · yH2 · Pt (3)

c∗N2

= (0.00184 · T−0.1383) · yN2 · Pt (4)

c∗O2

= (0.00224 · T−0.0404) · yO2 · Pt (5)

As it can be seen, there is a linear relationship (Henry’s law-ype) between the solubility and the partial pressure. Assuming aonstant mass transfer coefficient within this pressure interval, itan be deduced that mass transfer rate increases with pressure byavouring gas solubility.

Fig. 7 shows that both yield and selectivity increase almost lin-arly with increasing reaction pressure and the highest values arechieved at the highest examined pressure of 0.9 MPa (runs #02,

13–15). This effect confirms that shorter reaction times give betterroductivity, as H2O2 observed reaction rate decreases with time.igh pressures increase gas solubility enhancing mass transfer,

ncreasing the observed reaction rate and thus, decreasing reactionime and subsequently reducing secondary reactions.

ig. 7. Effect or reaction pressure on yield (�) and selectivity (�.). Reaction con-itions: O2/H2 = 12, time on stream 1 h, catalyst 3.0 × 10−2 g 5 wt.% Pd/C catalyst,.030 L MeOH, room temperature, 4 × 10−3 mol L−1 NaBr, 0.42 mol L−1 H+, 4.0 mol%2 in gas flow.

ity (�.). Reaction conditions: 0.9 MPa, O2/H2 = 12, catalyst 3.0 × 10−2 g 5 wt.% Pd/Cl% H2 in gas flow.

4. Conclusions

In this work we have successfully carried out the direct synthe-sis of hydrogen peroxide under mild pressure conditions and roomtemperature, using a commercial nano-Pd/C catalyst. The influenceof promoters, O2/H2 ratio, reaction time and reaction pressure hasbeen evaluated. Based on our previous experiments using the samecatalyst, under high pressure and using scCO2, we have found thatthe interaction between Br− anions and Pd does not depend onthe reaction conditions but solely on the nature of the catalyst.Therefore, the appropriate Br−/Pd molar ratio is approximately8.5 (corresponding to 4.0 × 10−3 mol L−1). An acidification effectis also necessary, in our case the optimum amount of phospho-ric acid was 0.0125 mol (corresponding to 0.4 mol L−1 and a H+/Pdmolar ratio of 887), which is sensibly higher than in our previousexperiments, confirming that the acid character of CO2 makes ita promoter itself. High O2/H2 ratios are more favourable for thisreaction system; specifically, the best results were obtained witha ratio of 12, whilst the presence of N2 did not have a significanteffect. The choice of reaction time is essential since both selectiv-ity and yield decrease rapidly with increasing reaction time, dueto the occurrence of side reactions, although the concentration ofperoxide continues to build up. Finally, reaction pressure also hasan important role in the process design, and, as expected, the pro-duction of H2O2 increases linearly with increasing pressure due toan enhanced solubility of the reactant gases. A selectivity of 69.0%and a yield of 38.3% were obtained under the best conditions, whilea maximum concentration of H2O2 of 1.33 wt.% was obtained at thelongest reaction time (4 h). As a general conclusion, the increase ofthe H2O2 observed reaction rate using short residence times has apositive effect in both yield and selectivity. Future work includesa semi-continuous set-up to work under high pressure in order toenhance the selectivity towards hydrogen peroxide.

Acknowledgements

The authors wish to thank the Spanish Science and InnovationMinistry, Project Reference: CTQ 2006-0299/PPQ and CTQ2009-14183-C02-01 for funding and for the FPI fellowship granted, theUniversity of Bath for their collaboration in the European PhD ofTeresa Moreno, Dr. Francisco Sobrón for the numerical filtering ofthe X-ray diffraction pattern and the Laboratorio de Técnicas Instru-mentales of the University of Valladolid for their invaluable helpwith the analysis of the catalyst.

References

[1] C. Samanta, Appl. Catal. A 350 (2008) 133–149.[2] H.-J. Riedl, G. Pfleiderer, US Patent 2,158,525 (1939).

alysis

[

[

[

[[

[

[[[

[[

[

[

[

[[

T. Moreno et al. / Applied Cat

[3] K. Weissermel, H.-J. Arpe, Industrial Organic Chemistry, 4th ed., Wiley–VCH,Weinheim, 2003.

[4] M.G. Zabetakis, Bureau of Mines 627 (1965).[5] Q. Chen, E.J. Beckman, Green Chem. 9 (2007) 802–808.[6] R. Burch, P.R. Ellis, Appl. Catal. B 42 (2003) 203–211.[7] D.P. Dissanayake, J.H. Lunsford, J. Catal. 206 (2002) 173–176.[8] T. Ishihara, Y. Ohura, S. Yoshida, Y. Hata, H. Nishiguchi, Y. Takita, Appl. Catal. A

291 (2005) 215–221.[9] Q. Liu, J.C. Bauer, R.E. Schaak, J.H. Lunsford, Appl. Catal. A 339 (2008) 130–136.10] F. Menegazzo, P. Burti, M. Signoretto, M. Manzoli, S. Vankova, F. Boccuzzi, F.

Pinna, G. Strukul, J. Catal. 257 (2008) 369–381.11] J.K. Edwards, A. Thomas, B.E. Solsona, P. Landon, A.F. Carley, G.J. Hutchings,

Catal. Today 122 (2007) 397–402.12] P. Landon, P.J. Collier, A.F. Carley, D. Chadwick, A.J. Papworth, A. Burrows, C.J.

Kiely, G.J. Hutchings, Phys. Chem. Chem. Phys. 5 (2003) 1917–1923.13] T. Haas, G. Stochniol, R. Jahn, US Patent 7,241,908 (2007).

14] Headwaters-Incorporated, Degussa/Headwaters Incorporated Building

Demonstration Plant for Direct Synthesis of Hydrogen Peroxide, 2009 (accessed02.03.2009) http://www.headwaters.com/data/upimages/press/Degussa%20PR.pdf.

15] E. Ntainjua, J.K. Edwards, A.F. Carley, J.A. Lopez-Sanchez, J.A. Moulijn, A.A. Herz-ing, C.J. Kiely, G.J. Hutchings, Green Chem. 10 (2008) 1162–1169.

[[

[[[

A: General 386 (2010) 28–33 33

16] C. Samanta, V.R. Choudhary, Appl. Catal. A 330 (2007) 23–32.17] Q. Liu, J.H. Lunsford, Appl. Catal. A 314 (2006) 94–100.18] S. Melada, R. Rioda, F. Menegazzo, F. Pinna, G. Strukul, J. Catal. 239 (2006)

422–430.19] G. Li, J. Edwards, A.F. Carley, G.J. Hutchings, Catal. Today 114 (2006) 369–371.20] J.K. Edwards, A. Thomas, A.F. Carley, A.A. Herzing, C.J. Kiely, G.J. Hutchings,

Green Chem. 10 (2008) 388–394.21] G. Bernardotto, F. Menegazzo, F. Pinna, M. Signoretto, G. Cruciani, G. Strukul,

Appl. Catal. A 358 (2009) 129–135.22] J.K. Edwards, B. Solsona, E.N. Ntainjua, A.F. Carley, A.A. Herzing, C.J. Kiely, G.J.

Hutchings, Science 323 (2009) 1037–1041.23] T.A. Pospelova, N.I. Kobozev, E.N. Eremin, Russ. J. Phys. Chem. 35 (1961)

143–147.24] Q. Liu, J.H. Lunsford, J. Catal. 239 (2006) 237–243.25] V.R. Choudhary, Y.V. Ingole, C. Samanta, P. Jana, Ind. Eng. Chem. Res. 46 (2007)

8566–8573.

26] V.R. Choudhary, P. Jana, J. Catal. 246 (2007) 434–439.27] E.N. Ntainjua, M. Piccinini, J.C. Pritchard, J.K. Edwards, A.F. Carley, J.A. Moulijn,

G.J. Hutchings, ChemSusChem 2 (2009) 575–580.28] T. Moreno, J. García-Serna, M.J. Cocero, Green Chem. 12 (2010) 282–289.29] V.R. Choudhary, C. Samanta, J. Catal. 238 (2006) 28–38.30] V.R. Choudhary, C. Samanta, P. Jana, Appl. Catal. A 317 (2007) 234–243.