Direct synthesis of hydrogen peroxide using Au–Pd-exchanged and supported heteropolyacid catalysts at ambient temperature using water as solvent

Abstract

The direct synthesis of hydrogen peroxide from molecular H₂ and O₂ represents a green and economic alternative to the current anthraquinone process used for the industrial production of H₂O₂. In order for the direct process to compete with the anthraquinone process, there is a need for enhanced H₂O₂ yields and H₂ selectivity in the process. We show that Au–Pd-exchanged and supported Cs-containing heteropolyacid catalysts with the Keggin structure are considerably more effective in achieving high H₂O₂ yields in the absence of acid or halide additives than previously reported catalysts. The Au–Pd-exchanged Cs-heteropolyacid catalysts also show superior H₂O₂ synthesis activity under challenging conditions (ambient temperature, water-only solvent and CO₂-free reaction gas). Au plays a crucial role in achieving the improved performance of these heteropolyacid-based catalysts. The heteropolyacid limits the subsequential hydrogenation/decomposition of H₂O₂.

---

Introduction

The direct synthesis of hydrogen peroxide (H₂O₂) from molecular hydrogen and oxygen represents a potentially economic and environmentally benign alternative to the current industrial process. At present, H₂O₂ is produced by the sequential hydrogenation and oxidation of an alkyl anthraquinone and over 2 M tonnes per year is manufactured in this way. Most of the H₂O₂ produced globally is used in bleaching and disinfectant applications, such as in the textile/paper and pulp industry and in medical practice, respectively. Legislative pressure to limit the use of chlorine in bleaching applications means that more H₂O₂ has to be produced in order to meet the high demand for bleaches. Recent interest in the use of H₂O₂ as a green oxidant for chemical synthesis has partly accounted for the considerable increase (10% per annum¹) in its global production.

The current anthraquinone process used for the industrial production of H₂O₂ requires high capital investment and involves the production and transportation of very concentrated H₂O₂ solutions; contrarily, a direct synthesis process would allow for small scale production of dilute H₂O₂ solutions at point of use in a green and inexpensive manner. The direct process therefore eliminates the risk of accidents² associated with the transportation of such concentrated H₂O₂ solutions from the production plants to the point of use.

Several research studies have reported the use of supported monometallic Pd^3–20 and bimetallic Au–Pd catalysts^21–33 for the direct synthesis of H₂O₂. Although the monometallic Pd catalysts are the most widely studied, Au–Pd bimetallic catalysts have been shown to be significantly more effective. TiO₂- and carbon-supported Au–Pd catalysts constitute the most effective^28–30 supported bimetallic Au–Pd catalysts for the direct synthesis of H₂O₂. The carbon-based Au–Pd catalyst is the most effective, yielding significantly higher H₂O₂ productivity and H₂ selectivity compared to the monometallic Pd and other supported Au–Pd catalysts. Although we consider that the discovery of these highly active Au–Pd catalysts does represent a significant advance for the direct process; further advances are needed. To compete with the indirect anthraquinone process, there is a need to optimise the H₂O₂ yield and H₂ selectivity in the direct synthesis process. This goal can be achieved by the formulation of more effective catalysts and by tuning reaction conditions so as to achieve the optimum catalytic performance. Most of the Pd and Au–Pd supported catalysts used to date for the direct synthesis of H₂O₂ are also active for the sequential hydrogenation/decomposition of H₂O₂ to form H₂O and/or the hydrogenation of O₂ to H₂O, thus limiting the amount of H₂O₂ that can be produced using these catalysts. This represents a prime setback for the direct process and, as such, has constituted the subject of various industrial^6–8 and scientific^{11–20,31–33} studies. In order to limit the undesired side reactions that lead to low H₂O₂ yields over the monometallic Pd catalysts, acids and halides have to be added to the reaction mixture.^10–19Halide incorporation into Pd catalysts has also been employed to suppress H₂O₂ hydrogenation/decomposition.^15–17,20 However, we have recently shown^31,32 that the addition of acid and halide to the reaction mixture or the incorporation of halide into the catalyst is not required when using the most effective Au–Pd bimetallic catalyst (Au–Pd/C). Acid and halide addition or halide incorporation only lead to a subtle positive effect at low concentrations. Moreover, the effect of halide addition/incorporation is not sustained upon catalyst re-use as the presence of bromide leads to the redistribution of the Au–Pd nanoparticles and, as a result, induces Au leaching from the catalyst. We have previously reported³³ a green and cost-effective method of switching-off the undesired hydrogenation/decomposition of H₂O₂ during its synthesis over the highly effective Au–Pd/C catalyst. Acid pre-treatment of the carbon support prior to co-impregnation with Au and Pd switches off H₂O₂ hydrogenation/decomposition and significantly enhances H₂O₂ productivity. This has prompted us to consider acidic supports as a basis for improved catalyst design. Sun et al.³⁴ and Park et al.³⁵ have recently reported the use of Pd-only insoluble heteropolyacid catalysts, comprising the Keggin structure, for the direct synthesis of H₂O₂ in the absence of any acid and/or halide additives. Sun et al.³⁴ showed that Pd-based supported heteropolyacid catalysts showed higher H₂O₂ productivity and selectivity compared to other conventional Pd-only catalysts. Park et al.³⁵ studied Pd-exchanged heteropolyacids (Pd_0.15Cs_xH_2.7−xPW₁₂O₄₀) with varying Cs content (2.0–2.7) for H₂O₂ synthesis and found that the most acidic catalyst (Pd_0.15Cs_2.5H_0.2PW₁₂O₄₀) showed the best performance. Building on these earlier studies we have investigated the effect of substituting some of the Pd present in Pd_0.15Cs_2.5H_0.2PW₁₂O₄₀ with Au and in this paper we report our findings, showing that the addition of Au markedly enhances the activity. We also study the effect of using heteropolyacids with the Keggin structure as supports for Pd, Au and Au–Pd catalysts prepared by impregnation and in this paper we present the results for their use as catalysts for the direct synthesis of hydrogen peroxide.

Experimental

Catalyst preparation

A Pd-exchanged heteropolyacid catalyst (Pd_0.15Cs_2.5H_0.2PW₁₂O₄₀) was prepared using an ion-exchange method. Appropriate amounts of Pd(NO₃)₂ (Aldrich) and CsNO₃ (Aldrich) were dissolved separately in deionised water and added dropwise with stirring to an aqueous solution of H₃PW₁₂O₄₀ (Aldrich, 99.999%) containing the appropriate amount of the acid. The resulting solution was continuously stirred while heating at 60 °C for 12 h to obtain a solid product, which was then dried overnight at 70 °C, followed by calcination at 300 °C for 2 h to form the final catalyst. A Pd-free heteropolyacid (Cs_2.8H_0.2PW₁₂O₄₀) was prepared in a similar manner and used as a support material for the preparation of supported Pd, Au and Au–Pd catalysts. Au–Pd-exchanged heteropolyacids (Pd_0.1Au_0.0333Cs_2.5H_0.2PW₁₂O₄₀ and Pd_0.075Au_0.05Cs_2.5H_0.2PW₁₂O₄₀) were also prepared using a similar ion exchange method. In this case, aqueous solutions of HAuCl₄·3H₂O and Pd(NO₃)₂ were added drop-wise to a solution of H₃PW₁₂O₄₀ followed by drop-wise addition of dissolved CsNO₃ after stirring of the mixture for 5 min.

Supported Au–Pd catalysts comprising 2.5 wt% Au/2.5 wt% Pd/support were prepared using the following standard co-impregnation method (all quantities stated are per g of finished catalyst). PdCl₂ (0.042 g, Johnson Matthey) was added to aqueous HAuCl₄·3H₂O solution (2.5 ml, 5 g in 250 ml) and stirred at 80 °C until the Pd dissolved completely. The appropriate support (0.95 g; carbon (G60, Aldrich, acid-treated and untreated) or Cs_2.8H_0.2PW12O40) was then added to the solution and stirred to form a paste. The paste was dried (110 °C, 16 h) before calcination (400 °C, 3 h for the carbon catalysts and 300 °C, 2 h for the heteropolyacid-based catalysts). Monometallic Au and Pd supported heteropolyacid catalysts were also prepared using the same impregnation method.

Catalyst characterisation and testing

Hydrogen peroxide synthesis and hydrogenation was evaluated using a Parr Instruments stainless steel autoclave with a nominal volume of 100 ml and a maximum working pressure of 14 MPa. To evaluate each catalyst for H₂O₂ synthesis the following standard reaction conditions were used. The autoclave was charged with catalyst (0.01 g) and 8.5 g solvent (5.6 g MeOH and 2.9 g H₂O). The charged autoclave was then purged three times with 5% H₂/CO₂ (0.7 MPa) before filling with 5% H₂/CO₂ to a pressure of 2.9 MPa at 20 °C. The pressure was allowed to drop to 2.6 MPa as the gas dissolved in the solvent at 20 °C. This was followed by the addition of 25% O₂/CO₂ (1.1 MPa). The temperature was then allowed to decrease to 2 °C followed by stirring (at 1200 rpm) of the reaction mixture for 30 mins. The above reaction parameters represent the optimum conditions we have previously used for the synthesis of H₂O₂.^22–33 In this study, we systematically varied these reaction conditions so as to investigated the use of ambient temperature, H₂O-only solvent and 2%H₂/air for the direct synthesis of H₂O₂. H₂O₂ productivity was determined by titrating aliquots of the final solution after reaction with acidified Ce(SO₄)₂ (0.01 M) in the presence of two drops of ferroin indicator.

H₂O₂ hydrogenation experiments were carried out in a similar manner as H₂O₂ synthesis experiments but without adding the 1.1 MPa 25%O₂/CO₂. Furthermore, 0.68 g of H₂O from the 8.5 g of solvent was replaced by a 50% H₂O₂ solution to give a reaction solvent containing 4 wt% H₂O₂. The standard reaction conditions for H₂O₂ hydrogenation included: 0.01 g catalyst, 8.5 g solvent (5.6 g MeOH, 2.22 g H₂O and 0.68 g H₂O₂ (50%)), 2.9 MPa 5%H₂/CO₂, 2 °C, 1200 rpm, 30 mins. Just like for H₂O₂ synthesis, the standard conditions were varied in some experiments to study H₂O₂ hydrogenation at ambient temperature and when using H₂O-only solvent.

FT-IR spectra were measured using a Varian Excalibur 400 FT-IR spectrometer with an UMA 600 microscope. The images were recorded in reflection mode. A small portion of sample was pressed onto a single germanium crystal/microscope base and irradiated by an infra-red source. The sample was exposed to the atmosphere during analysis and so a background spectrum was processed beforehand, allowing subtraction of CO₂ peaks situated at: 1300–2000 cm⁻¹ (broad), 2300–2400 cm⁻¹ (sharp), 2840–3000 cm⁻¹ (weak) and 3500–4000 cm⁻¹ (broad).

Laser Raman spectroscopy was performed using a RENISHAW inVia Raman microscope, using a 25mW power laser set at a reflection wavelength of 514 nm. The laser was set to a power output of 5% in order to avoid damaging samples in the study and obtaining the best result clarity.

Investigation of the bulk structure of the materials was carried out using powder X-ray diffraction (XRD) on a (θ–θ) PANalytical X'pert Pro powder diffractometer using a Cu-Kα radiation source operating at 40 KeV and 40 mA. Standard analysis was performed using a 40 min scan between 2θ values of 10–80° with the samples supported on an amorphous silicon wafer. Diffraction patterns of phases were identified using the ICDD data base.

Temperature programmed desorption (TPD) profiles were recorded using a Thermo 1100 series TPDRO. 0.1 g of sample was packed into the sample tube using quartz wool and a volume reducer. The sample was then pre-treated under He while being heated from room temperature to 110 °C at 5 °C min⁻¹, where it was held for 60min. The sample was allowed to cool to room temperature, following this ammonia was passed over the catalyst for 10 min at 20 ml min⁻¹. The samples were then heated from room temperature to 900 °C at 5 °C min⁻¹. The profile was recorded using a TCD with positive polarity and a gain of 10.

XPS measurements were made on a Kratos Axis Ultra DLD spectrometer using monochromatic AlK_α radiation (source power 120–180 W). An analyser pass energy of 160 eV was used for survey scans, and 40 eV for detailed acquisition of individual elemental regions. Samples were mounted using double-sided adhesive tape, and binding energies referenced to the C(1 s) binding energy of adventitious carbon contamination taken to be 284.7 eV. Spectra were quantified using CasaXPS (Neil Fairley, UK) and data is presented as surface compositions (atom%).

Results and discussion

Comparison of carbon and heteropolyacid-based Au, Pd and Au–Pd catalysts for H₂O₂ synthesis and hydrogenation using standard reaction conditions

In our previous studies,^22–33 we employed standard conditions specified in Table 1 for H₂O₂ synthesis and hydrogenation. We have previously shown³² that a Au–Pd/C catalyst prepared by co-impregnation of Au and Pd onto an acid pre-treated carbon support, designated Au–Pd/C(2% HNO₃), was the most effective Au–Pd catalyst for the direct synthesis of H₂O₂ when using these standard reaction conditions. Table 1 compares the rate of H₂O₂ synthesis and hydrogenation over heteropolyacid catalysts and Au–Pd/C(2% HNO₃) using the standard reaction conditions. The Au–Pd supported heteropolyacid catalyst (2.5% Au/2.5% Pd/Cs_2.8H_0.2PW₁₂O₄₀) was the most active catalyst (for both synthesis and hydrogenation), yielding significantly higher H₂O₂ productivity than Au–Pd/C(2% HNO₃). The Au–Pd/C(2% HNO₃) catalyst, however, showed a higher H₂O₂ synthesis rate than the other heteropolyacid-based catalysts. Since this catalyst does not show any H₂O₂ hydrogenation/decomposition activity, it is likely to be the most selective catalyst for H₂O₂ synthesis using standard reaction conditions. Heteropolyacid-supported Pd and Au–Pd catalysts showed higher H₂O₂ formation activity compared to carbon-supported catalysts with the same Pd or Au–Pd content. Although the heteropolyacid-based Pd and Au–Pd catalysts prepared by ion exchange yield lower H₂O₂ productivities than the heteropolyacid- and carbon-based supported catalysts, they contain ca. 9–10 times less active metal (Au and/or Pd) present in the supported catalysts. Since the Pd and Au–Pd-exchanged heteropolyacid catalysts yield relatively high H₂O₂ formation rates (86–96 mol kg_cat⁻¹ h⁻¹ compared to 110–198 mol kg_cat⁻¹ h⁻¹ for the supported catalysts) irrespective of their very low Pd or Au–Pd content (ca. 0.4–0.6 wt%), we consider them to be promising catalysts for the direct synthesis of H₂O₂. Fig. 1 shows that the turnover frequency (TOF) for Pd and Au–Pd heteropolyacid catalysts made by ion exchange are ca. 3 times more than for the heteropolyacid- and carbon-supported Au–Pd catalysts. Moreover, these catalysts also show considerably lower H₂O₂ hydrogenation rates compared to the most active supported catalyst (2.5% Au/2.5% Pd/Cs_2.8H_0.2PW₁₂O₄₀).

Table 1 H₂O₂ productivity and hydrogenation using Pd and Au–Pd-exchanged heteropolyacid and Au–Pd supported catalysts under standard reaction conditions

Catalyst	H2O2 Productivity^a/mol kgcat^−1 h^−1	H2O2 Hydrogenation ^b/mol kgcat^−1 h^−1
a Rate of hydrogen peroxide production determined using reaction using standard reaction conditions: 5% H2/CO2 (2.9 MPa) and 25% O2/CO2 (1.1 MPa), 8.5 g solvent (5.6 g MeOH + 2.9 g H2O), 0.01 g catalyst, 2 °C, 1200 rpm, 30 min. b Rate of hydrogenation of H2O2 calculated from amount of H2O2 hydrogenated using standard reaction conditions: 2.9 MPa 5%H2/CO2, 8.5 g solvent (5.6 g MeOH, 2.22 g H2O and 0.68 g 50% H2O2), 0.01 g catalyst, 2 °C, 1200 rpm, 30 min.
Cs2.8H0.2PW12O40	1	162
Pd0.15Cs2.5H0.2PW12O40	96	221
Au0.1Cs2.5H0.2PW12O40	16	347
Pd0.1Au0.0333Cs2.5H0.2PW12O40	97	384
Pd0.075Au0.05Cs2.5H0.2PW12O40	86	213
5% Pd/Cs2.8H0.2PW12O40	136	281
5% Au/Cs2.8H0.2PW12O40	14	103
2.5% Au/2.5% Pd/Cs2.8H0.2PW12O40	198	704
2.5% Au/2.5% Pd/C	110	117
2.5% Au/2.5% Pd/C (2% HNO3)	160	0

---

Fig. 1 A comparison of TOF for H₂O₂ synthesis at 2 °C using heteropolyacid-based and carbon-based Au, Pd and Au–Pd catalysts [data corrected to take account of contribution from the Au and Pd-free (heteropolyacid-only) material]. HPA – heteropolyacid, AW – acid washed.

The Au-only exchanged and supported heteropolyacid catalysts showed considerably higher H₂O₂ productivity compared to Au-only supported catalysts that we have previously reported.²⁹ However, the addition of Au by ion-exchange to the Pd-only heteropolyacid did not significantly alter the catalytic performance under standard reaction conditions. This is in contrast to the synergistic effect observed^21–33 for conventional supported Au–Pd catalysts. We have shown that the addition of Au to Pd-only catalysts supported on C, Al₂O₃, TiO₂ and SiO₂ lowers H₂O₂ hydrogenation/decomposition and promotes H₂O₂ synthesis under standard reaction conditions. Unlike the Au–Pd-exchanged heteropolyacid catalysts, the Au–Pd-supported heteropolyacid catalyst showed higher H₂O₂ synthesis activity than the Pd-only supported catalyst, thus confirming the promotional effect of Au previously observed for other supported Au–Pd catalysts. This may suggest that the different methods of catalyst preparation (ion-exchange and impregnation) lead to the formation of Au and Pd nanoparticles with varying characteristics (e.g. particle size, location and nature/structure) and as such show different H₂O₂ synthesis and hydrogenation/decomposition abilities.

The heteropolyacid-only catalyst (Cs_2.8H_0.2PW₁₂O₄₀) showed the lowest rate of H₂O₂ formation amongst all catalysts tested. This catalyst was almost inactive for H₂O₂ synthesis but showed considerable hydrogenation activity. However, the use of this heteropolyacid-only material as a catalyst support for Pd-only and Au–Pd catalysts led to catalysts showing significantly higher H₂O₂ formation rates than similar Pd-only and Au–Pd catalysts supported on other conventional support materials^21–33 (C, Al₂O₃, TiO₂ and SiO₂). We have previously shown²⁹ that acidic supports with low isoelectric points (IEP) gave improved catalytic performance for supported-Pd and Au–Pd catalysts than basic supports e.g.MgO. Since the heteropolyacid-based supported catalysts are composed of a more acidic support (with lower IEP) than those in other conventional supported catalysts, we attribute their superior H₂O₂ synthesis performance under standard conditions to the intrinsically acidic character of the heteropolyacid support.

H₂O₂ synthesis and hydrogenation at ambient temperature over carbon and heteropolyacid-based Au, Pd and Au–Pd catalysts

H₂O₂ synthesis is often carried out at sub-ambient temperatures (e.g. 2 °C) in order to avoid subsequent decomposition of H₂O₂ as it is formed. We have now investigated the effect of carrying out H₂O₂ synthesis and hydrogenation at ambient temperature (Table 2) while maintaining all other reaction conditions unchanged. The H₂O₂ hydrogenation rates over all catalysts investigated (except Au_0.1Cs_2.5H_0.2PW₁₂O₄₀ and 5% Au/Cs_2.8H_0.2PW₁₂O₄₀) were considerably higher at ambient temperature compared to 2 °C. This is not surprising as thermal decomposition of H₂O₂ is favoured at ambient temperature. Surprisingly, however, Au_0.1Cs_2.8H_0.2PW₁₂O₄₀ and 5% Au/Cs_2.8H_0.2PW₁₂O₄₀ showed lower H₂O₂ hydrogenation rates at ambient temperature than at 2 °C. The fact the Au-only heteropolyacid catalyst has a significantly lower H₂O₂ hydrogenation activity compared to the heteropolyacid-only catalyst demonstrates that the incorporation of Au (by ion-exchange or impregnation) in the heteropolyacid-only material also led to significant decrease in H₂O₂ hydrogenation at ambient temperature. This suggest that Au inhibits the hydrogenation/decomposition of H₂O₂ over the heteropolyacid-based catalysts at ambient temperature. This effect is, however, unique for the elevated temperature as we observed the opposite effect (increase in H₂O₂ hydrogenation with addition of Au to heteropolyacid-only catalyst) at 2 °C (Table 1).

Table 2 H₂O₂ productivity and hydrogenation using Pd and Au–Pd-exchanged heteropolyacid and Au–Pd supported catalysts at 20 °C using CH₃OH and H₂O as solvent

Catalyst	H2O2 Productivity^a/mol kgcat^−1 h^−1	H2O2 Hydrogenation ^b/mol kgcat^−1 h^−1
a Rate of hydrogen peroxide production determined for reaction at 20 °C (all other conditions same as defined in Table 1). b Rate of hydrogenation of H2O2 calculated from amount of H2O2 hydrogenated at 20 °C (all other conditions same as defined in Table 1).
Cs2.8H0.2PW12O40	2	558
Pd0.15Cs2.5H0.2PW12O40	43	670
Au0.1Cs2.5H0.2PW12O40	35	282
Pd0.1Au0.0333Cs2.5H0.2PW12O40	107	676
Pd0.075Au0.05Cs2.5H0.2PW12O40	168	546
5% Pd/Cs2.8H0.2PW12O40	41	1105
5% Au/Cs2.8H0.2PW12O40	34	36
2.5% Au/2.5% Pd/Cs2.8H0.2PW12O40	30	1052
2.5% Au/2.5% Pd/C	98	352
2.5% Au/2.5% Pd/C (2% HNO3)	102	443

---

With the exception of the two Au–Pd-exchanged heteropolyacid and the Au-only heteropolyacid catalysts, all other catalysts showed lower H₂O₂ formation rates at ambient temperature (Table 2) compared to 2 °C (Table 1) and this is in agreement with the extent to which the H₂O₂ hydrogenation/decomposition rate over each catalyst increased at ambient temperature. Amongst all the Pd, Au and Au–Pd catalysts tested, Au–Pd/Cs_2.8H_0.2PW₁₂O₄₀ gives the best rate of H₂O₂ synthesis (198 mol kg_cat⁻¹ h⁻¹⁾ at 2 °C, but gave the lowest H₂O₂ productivity (30 mol kg_cat⁻¹ h⁻¹) at ambient temperature and this correlates with its superior H₂O₂ hydrogenation/decomposition rate (1052 mol kg_cat⁻¹ h⁻¹) at ambient temperature. Only 5% Pd/Cs_2.8H_0.2PW₁₂O₄₀ showed a slightly higher H₂O₂ hydrogenation rate than the Au–Pd/Cs_2.8H_0.2PW₁₂O₄₀. Interestingly, the use of ambient temperature enhanced H₂O₂ productivity over the Au–Pd-exchanged heteropolyacid and Au-only heteropolyacid (exchanged and supported) catalysts and the degree of enhancement was directly proportional to the Au content for the Au–Pd catalysts. Pd_0.075Au_0.05Cs_2.5H_0.2PW₁₂O₄₀ was the most effective catalyst, giving H₂O₂ productivity (168 mol kg_cat⁻¹ h⁻¹) 4 times greater than the Pd-only exchanged heteropolyacid and significantly greater than Au–Pd/C(2% HNO₃). In fact, the catalytic performance of this catalyst at ambient temperature was better than the performance of all the other catalysts (except 2.5% Au/2.5% Pd/Cs_2.8H_0.2PW₁₂O₄₀) at 2 °C. The catalyst also showed outstandingly higher TOF at ambient temperature (Fig. 2) than the supported heteropolyacid catalyst and the AuPd/C(2%HNO₃) catalyst, which we have previously shown to be the most effective catalyst for H₂O₂ synthesis. Fig. 3 compares the change in H₂O₂ formation rate and wt% H₂O₂ formed at ambient temperature as a function of time for this Au–Pd-exchanged heteropolyacid catalyst and a conventional supported Au–Pd catalyst (Au–Pd/C). Both catalysts showed similar time online profiles; the H₂O₂ formed increased with time and attained a stable maximum before subsequently decreasing, while the rate was high initially and decreased continuously over time. However, the Au–Pd-exchanged heteropolyacid catalyst led to considerably higher H₂O₂ concentrations and rate of formation over the time-range investigated.

Fig. 2 A comparison of TOF for H₂O₂ synthesis at 20 °C using HPA-based and carbon-based Au, Pd and Au–Pd catalysts [data corrected to take account of contribution from the Au and Pd-free (heteropolyacid-only) material]. HPA – heteropolyacid, AW – acid washed.

Fig. 3 The effect of reaction time on the direct formation of H₂O₂ over Pd_0.075Au_0.05Cs_2.5H_0.2PW₁₂O₄₀ (solid lines) and Au–Pd/C (broken lines) using methanol and H₂O as solvent at ambient temperature
wt% H₂O₂ formed,
H₂O₂ productivity.

The observation that H₂O₂ productivity is only enhanced at room temperature over the Au-containing heteropolyacid catalysts and not the Pd-only heteropolyacid catalysts (show decrease in H₂O₂ productivity), suggest that Au plays a critical role in achieving the superior ambient temperature performance of the Au–Pd-exchanged heteropolyacid catalysts. The role of Au in promoting H₂O₂ formation over the Au–Pd-exchanged heteropolyacid and Au-only (exchanged and supported) catalysts at ambient temperature is mainly related to its ability to inhibit H₂O₂ hydrogenation/decomposition over these catalysts. This is supported by the fact that the Au-only (both supported and exchanged) heteropolyacid catalysts show a decrease in hydrogenation activity with increase in temperature from 2 °C to 20 °C and the degree of increase in hydrogenation activity with increasing temperature (2 °C as compared with 20 °C) over the Au–Pd-exchanged heteropolyacid catalysts is lower compared to the Pd-only exchanged heteropolyacid catalyst. Since the use of ambient temperature represents an important energy-saving option when compared with operation at sub-ambient temperatures, we consider the enhanced ambient temperature performance of the Au–Pd-exchanged heteropolyacid catalysts an interesting development for the direct synthesis of H₂O₂. In order to optimise hydrogen utilization, there is still a need to limit or completely switch-off the ambient temperature hydrogenation/decomposition of H₂O₂ over these catalysts and this is a subject of on-going research.

H₂O₂ synthesis and hydrogenation at ambient temperature with water as solvent using carbon and heteropolyacid-based Au, Pd and Au–Pd catalysts

Our standard reaction conditions (Table 1) make use of a mixture of methanol (66%) and H₂O as solvent for the direct synthesis of H₂O₂. The presence of methanol improves the solubility of the reacting gases (H₂ and O₂) and stabilizes formed H₂O₂ against decomposition. However, most H₂O₂ applications require it diluted only in water. This means that if H₂O₂ is synthesised in H₂O/methanol solvent, additional cost must be incurred to separate the methanol from the final product. Furthermore, H₂O represents a non-toxic, non-flammable and cheaper solvent compared to alternative organic solvents and acids commonly used for the direct synthesis of H₂O. We have therefore studied the effect of carrying out H₂O₂ synthesis and hydrogenation using H₂O-only solvent at ambient temperature (all other reaction conditions remaining unchanged) and results are presented in Table 3. The H₂O₂ productivity over each catalyst (except the heteropolyacid-only catalyst) was considerably lower in H₂O-only solvent (Table 3) compared to H₂O/methanol solvent (Table 2). This can be attributed to the increase in H₂O₂ hydrogenation/decomposition observed upon using H₂O-only solvent (see Tables 2 and 3) coupled with the lower solubility of H₂ in H₂O-only. The increase in H₂O₂ hydrogenation/decomposition using water as solvent confirms the role of methanol as a H₂O₂ stabiliser. The heteropolyacid-based catalysts prepared by ion exchange showed significantly higher H₂O₂ formation rates than all the Au-only, Pd-only and Au–Pd-supported heteropolyacid and carbon-based supported catalysts prepared by impregnation. Au–Pd-exchanged heteropolyacid catalysts (Pd_0.1Au_0.033Cs_2.5H_0.2PW₁₂O₄₀ and Pd_0.075Au_0.05Cs_2.5H_0.2PW₁₂O₄₀) again showed the best performance in H₂O₂ formation, with H₂O₂ productivity ca. 2 and 15 times higher than H₂O₂ productivity over the Pd-only exchanged heteropolyacid and all other Au–Pd-supported catalysts, respectively. The TOF in water at 20 °C (Table 4) for the best Au–Pd-exchanged heteropolyacid catalyst (Pd_0.075Au_0.05Cs_2.5H_0.2PW₁₂O₄₀) was over 140 times higher compared to the TOF for the conventionally supported Au–Pd/C(2% HNO₃) catalyst previously shown to be the most active catalyst for H₂O₂ synthesis under our standard reaction conditions at 2 °C in water/methanol solvent. The heteropolyacid-only catalyst (Cs_2.8H_0.2PW₁₂O₄₀) showed a decrease in H₂O₂ hydrogenation using water as solvent and this was reflected in the significant increase (by an order of magnitude) in its H₂O₂ productivity. This suggests that in the absence of methanol (i.e. in H₂O-only solvent), the heteropolyacid plays a crucial role in stabilising H₂O₂ from subsequent hydrogenation/decomposition. Indeed, the increase in H₂O₂ hydrogenation/decomposition in H₂O-only solvent was higher for the carbon-based catalysts than for the heteropolyacid-based catalysts. Only the supported-Pd and Au–Pd heteropolyacid catalysts showed an increase in H₂O₂ hydrogenation comparable to the supported carbon catalysts (consistent with lower H₂O₂ productivities over these catalysts) and this can be explained by the fact that a higher number of the heteropolyacid sites in these catalysts are covered by Pd or Au–Pd sites.

Table 3 H₂O₂ productivity and hydrogenation using Pd and Au–Pd-exchanged heteropolyacid and Au–Pd supported catalysts at 20 °C using H₂O as solvent

Catalyst	H2O2 Productivity^a/mol kgcat^−1 h^−1	H2O2 Hydrogenation ^b/mol kgcat^−1 h^−1
a Rate of hydrogen peroxide production determined for reaction at 20 °C using H2O (8.5 g) as solvent (all other conditions same as defined in Table 1). b Rate of hydrogenation of H2O2 calculated from amount of H2O2 hydrogenated at 20 °C using H2O (8.5 g) as solvent (all other conditions same as defined in Table 1).
Cs2.8H0.2PW12O40	22	282
Pd0.15Cs2.5H0.2PW12O40	27	705
Au0.1Cs2.5H0.2PW12O40	18	376
Pd0.1Au0.0333Cs2.5H0.2PW12O40	58	793
Pd0.075Au0.05Cs2.5H0.2PW12O40	61	776
5% Pd/Cs2.8H0.2PW12O40	4	1369
5% Au/Cs2.8H0.2PW12O40	16	217
2.5% Au/2.5% Pd/Cs2.8H0.2PW12O40	3	1310
2.5% Au/2.5% Pd/C	4	799
2.5% Au/2.5% Pd/C (2% HNO3)	4	746

---

Table 4 Comparison of TOF for H₂O₂ synthesis at 20 °C in water as solvent using heteropolyacid-based and carbon-based Au–Pd catalysts^a

Catalyst	TOF/mol H2O2 molmetal^−1 h^−1
a Data corrected to take account of contribution from the Au and Pd-free (heteropolyacid-only) material. Hydrogen peroxide production determined for reaction at 20 °C using H2O (8.5 g) as solvent (all other conditions same as defined in Table 1).
Pd0.075Au0.05Cs2.5H0.2PW12O40	994
2.5% Au/2.5% Pd/Cs2.8H0.2PW12O40	8
2.5% Au/2.5% Pd/C (2% HNO3)	11

---

Effect of CO₂-free reaction gas (2% H₂/air) on H₂O₂ synthesis and hydrogenation over carbon and heteropolyacid-based Au, Pd and Au–Pd catalysts

Our standard reaction gas for H₂O₂ synthesis (Table 1) comprise a mixture of 5% H₂/CO₂ and 25% O₂/CO₂ with H₂ : O₂ molar ratio of ca. 1 : 2. Although we have shown²⁸ that dissolved CO₂ gives the reaction solvent the acidity required to promote H₂O₂ formation, the use of CO₂ diluent still represents an additional complexity. We have therefore investigated the direct synthesis of H₂O₂ using 2% H₂/air as the reactant gas (Table 5). The H₂O₂ formation rate over each catalyst was considerably lower (when using 2% H₂/air) compared to standard reaction conditions (Table 1). This is not surprising as the H₂/O₂ molar ratio (1/20) in 2% H₂/air is markedly different from the 1/2 ratio employed using standard reaction conditions. This result is therefore consistent with our previous study,³⁶ showing a decrease in H₂O₂ productivity with decreasing H₂/O₂ mol ratio. Au–Pd-exchanged and supported heteropolyacid catalysts were considerably more effective than the other catalysts investigated. The supported-heteropolyacid catalysts (2.5% Au/2.5% Pd/Cs_2.8H_0.2PW₁₂O₄₀ and 5% Pd/Cs_2.8H_0.2PW₁₂O₄₀) were the most active catalysts, showing slightly higher H₂O₂ productivity than the Au–Pd-exchanged heteropolyacid catalysts. We consider the Au–Pd-exchanged heteropolyacid catalysts to be the most efficient catalysts for H₂O₂ synthesis using 2% H₂/air since they contain approximately an order of magnitude lower amount of metal, and, consequently, the Au–Pd-exchanged heteropolyacid catalysts were an order of magnitude more active than the Pd-only exchanged heteropolyacid catalyst. As already discussed (Table 2 and 3), these Au–Pd-exchanged heteropolyacid catalysts also showed superior performance for H₂O₂ synthesis at ambient temperature and with water as solvent when compared to the Pd-only exchanged heteropolyacid prepared using the same ion-exchange method. This indicates that the presence of Au in these heteropolyacid-based catalysts is key in achieving high H₂O₂ productivity under such green and cost-effective conditions (ambient temperature, water only solvent and the use of CO₂-free reaction gas (2% H₂/air)) preferred for industrial production of H₂O₂.

Table 5 Synthesis of H₂O₂ using 2% H₂/Air

Catalyst	H2O2 productivity/mol kg^−1 h^−1
	2 °C, H2O + MeOH^a	20 °C, H2O + MeOH^b	20 °C, H2O-only^c
a Rate of hydrogen peroxide production determined for reaction using 2% H2/air under conditions that mimic standard reaction conditions (i.e. 4 MPa 2% H2/air, 8.5 g solvent (5.6 g MeOH + 2.9 g H2O), 0.01 g catalyst, 2 °C, 1200 rpm, 30 mins). b Rate of hydrogen peroxide production determined after reaction using 2% H2/air at 20 °C using a mixture MeOH and H2O as solvent (all other conditions unchanged) c Rate of hydrogen peroxide production determined after reaction using 2% H2/air at 20 °C using a H2O-only solvent (all other conditions unchanged).
Cs2.8H0.2PW12O40	0.7	8	6
Pd0.15Cs2.5H0.2PW12O40	4	10	7
Au0.1Cs2.5H0.2PW12O40	8	10	7
Pd0.1Au0.033Cs2.5H0.2PW12O40	43	49	11
Pd0.075Au0.05Cs2.5H0.2PW12O40	46	48	14
5% Pd/Cs2.8H0.2PW12O40	61	51	7
5% Au/Cs2.8H0.2PW12O40	6	14	6
2.5% Au/2.5% Pd/Cs2.5H0.2PW12O40	63	57	8
2.5% Au 2.5% Pd/C	22	34	6
2.5% Au 2.5% Pd/C (HNO3)	26	35	8

---

Using 2% H₂/air as reactant led to similar effects for H₂O₂ synthesis as those observed using H₂/CO₂ and O₂/CO₂ with respect to the effect of solvent and reaction temperature. The data (Table 5) confirms that Au–Pd heteropolyacid-based catalysts are more efficient for the direct synthesis of H₂O₂ using green and cost-effective conditions (ambient temperature, H₂O-only solvent) than Pd-only heteropolyacid and Au–Pd/C (untreated or acid-treated) catalysts

Effect of increase in Au–Pd loading and change in Cs content of Pd and Au–Pd-exchanged heteropolyacid catalysts

The Au, Pd and Au–Pd-exchanged heteropolyacid catalysts investigated so far only contain 0.49–0.61wt% Au, Pd or Au + Pd (Tables 1–3). We also investigated the effect of increasing the Pd and Au–Pd content of these catalysts through substitution of Cs. Pd and Au–Pd-exchanged heteropolyacid catalysts containing 2.11 and 2.35 wt% Pd and Au + Pd, respectively, were prepared and tested for H₂O₂ synthesis under standard conditions as well as using the challenging conditions (ambient temperature and water-only solvent). The results (Table 6) showed a considerable increase in H₂O₂ productivity (from 96 to 144 mol kg⁻¹ h⁻¹) with increasing Pd content for the Pd-only exchanged heteropolyacid when tested under standard conditions. However, as expected, catalysts with lower metal loading were more efficient in terms of TOFs. However, the higher loading catalyst at ambient temperature and in water-only solvent resulted in a decrease in activity. The Au–Pd-exchanged heteropolyacid catalyst showed a marked increase in activity under standard conditions (86–261 mol kg⁻¹ h⁻¹) and at ambient temperature (168–228 mol kg⁻¹ h⁻¹) with increasing Au–Pd content; the activity of the catalyst decreased from 61 to 40 mol kg⁻¹ h⁻¹ with increasing Au–Pd content in water-only solvent. The observed trends in activity appear to be related to the acidity of these catalysts and this is discussed in a subsequent section. Although the higher loading Au–Pd-exchanged heteropolyacid catalyst shows considerably higher H₂O₂ synthesis activity (228 mol kg_cat⁻¹ h⁻¹) at ambient temperature in water/methanol solvent, we do not observe the increase in activity noted for the lower loading catalysts as the temperature is increased from 2 °C to ambient temperature as there is a decrease from 261 to 221 mol kg_cat⁻¹ h⁻¹. However, the extent of the decrease in activity is considerably lower when compared to the higher loading Pd-only exchanged heteropolyacid catalyst. The absence of activity enhancement with an increase in temperature for the higher loaded Au–Pd-exchanged heteropolyacid catalyst can be related to these high activity catalysts being more active for hydrogenation of H₂O₂.

Table 6 The effects of ambient temperature and use of water-only solvent on H₂O₂ synthesis using Pd and Au–Pd-exchanged Cs-heteropolyacid catalysts containing higher metal loading

Catalyst	Metal wt (%)	H2O2 Productivity/mol kgcat^−1 h^−1
		2 °C, water + MeOH^a	20 °C, water + MeOH^b	20 °C, water only^c
a ^, b Rate of hydrogen peroxide production determined for reaction at 2 °C^a or 20 °C^b using standard reaction conditions: 5% H2/CO2 (2.9 MPa) and 25% O2/CO2 (1.1 MPa), 8.5 g solvent (5.6 g MeOH + 2.9 g H2O), 0.01 g catalyst, 1200 rpm, 30 min). c Rate of hydrogen peroxide production determined after reaction at 20 °C using water-only solvent. Reaction conditions: 5% H2/CO2 (2.9 MPa) and 25% O2/CO2 (1.1 MPa), 8.5 g H2O, 0.01 g catalyst, 1200 rpm, 30 min).
Pd0.15Cs2.5H0.2PW12O40	0.49	96	43	27
Pd0.65Cs1.5H0.2PW12O40	2.11	144	37	13
Pd0.075Au0.05Cs2.5H0.2PW12O40	0.53	86	168	61
Pd0.325Au0.217Cs1.5H0.2PW12O40	2.35	261	228	40

---

Catalyst characterisation

IR spectra recorded for the heteropolyacid catalysts (Fig. 4a and 4b) showed the characteristic features of phosphotungstic acids comprising the Keggin structure.³⁷ The Keggin structures characteristic vibrational bands were observed at 1079 cm⁻¹ [ν_as(P–O) vibration], 975 cm⁻¹ [terminal ν_as(W=O) vibration], 887 and 795 cm⁻¹ [assigned to two different ν_as(W–O–W) vibrations]. The weak bands at 2400 cm⁻¹ relate to CO₂ from the air. Some spectra also show bands centred at ca. 3400 cm⁻¹, which can be assigned to ν(O–H). The ν(O–H) vibrational bands may be due to water adsorbed on the sample. Some samples also show a band at 1710 cm⁻¹ characteristic of the presence of protonated water clusters, such as H₅O₂⁺. Since the spectra were recorded in open air without further drying of samples prior to the recording of spectra, these (O–H) vibrations cannot be exclusively assigned to the heteropolyacid structure. In all samples after Cs⁺ substitution the P–O stretch remains constant, indicating that the substitution of Cs⁺ for H⁺ does not cause strain on the P–O₄ tetrahedron at the centre of the structure. In accordance with the literature³⁷ a shift in ν_as(W=O) from 975 to 982 cm⁻¹ on incorporation of Cs⁺ (spectra B–F) indicates a strengthening of W=O bonds. This indicates that less hydrogen bonding is present, caused by the removal of H⁺ from the Keggin structure. The IR spectra recorded do not give rise to any new features on incorporation of Au and Pd by impregnation or ion exchange. However, since only small amounts of Au and Pd are incorporated by ion exchange any new signals may be below the limit of detection. The IR spectra of the impregnated catalysts (spectra C and D) show that on impregnation of 5 wt% Pd and 2.5% Pd 2.5% Au followed by calcination, the Keggin structure has been maintained. IR spectroscopy also shows that the Keggin structure was maintained when using ion exchange preparation methods.

Fig. 4 (a) IR spectra between 4000–500 cm⁻¹ of A – H₃PW₁₂O₄₀·6H₂O, B – Cs_2.8H_0.2PW₁₂O₄₀, C – 5% Pd/Cs_2.8H_0.2PW₁₂O₄₀, D – 2.5%Au/2.5%Pd/Cs_2.8H_0.2PW₁₂O₄₀, E – Pd_0.15Cs_2.5H_0.2PW₁₂O₄₀ and F – Pd_0.075Au_0.05Cs_2.5H_0.2PW₁₂O₄₀. (b) IR spectra between 1500–500 cm⁻¹ of A – H₃PW₁₂O₄₀·6H₂O, B – Cs_2.8H_0.2PW₁₂O₄₀, C – 5% Pd/Cs_2.8H_0.2PW₁₂O₄₀, D – 2.5% Au/2.5% Pd/Cs_2.8H_0.2PW₁₂O₄₀, E – Pd_0.15Cs_2.5H_0.2PW₁₂O₄₀ and F – Pd_0.075Au_0.05Cs_2.5H_0.2PW₁₂O₄₀ showing the shift in W=O peak at 975 cm⁻¹.

Raman spectra (Fig. 5) recorded for Au, Pd and Au–Pd heteropolyacid catalysts show Raman bands related to atomic vibrations characteristic of peroxocomplex fragments³⁸ in the range of Raman shifts of 0–1200 cm⁻¹. Raman bands at 996 and 1006 cm⁻¹ are associated with v(W=O) bond vibrations while a single band at 915 cm⁻¹ corresponds to the v(O–O) vibration set. A broader band at 550 cm⁻¹ has been assigned to two possible features; specifically asymmetric and/or symmetry vibrational modes of the W–(O₂) bond. In accordance with the literature,³⁸ bands located at lower frequencies (236, 216, 160, 154 and 109) arise from the deformation of complex fragments in the heteropolyacid framework. Additional Raman bands centred at 318 and 342 cm⁻¹ have been identified for the 5% Au/Cs_2.5H_0.2PW₁₂O₄₀catalyst prepared by impregnation, and these probably relate to vibrations from a Au-species. It is possible from closer inspection of the spectra to surmise that the decrease in resolution for bands centred at 540 and 915 cm⁻¹ are due to interaction of Au atoms with oxygen defects and termination points. In summary, the spectra for the heteropolyacid series shows that the key characteristic heteropolyacid bands are retained for all catalysts irrespective of the incorporation of Au, Pd or Au–Pd by ion-exchange or impregnation. Only the Au-heteropolyacid catalyst prepared by impregnation (containing ca. 8 times more Au than catalysts prepared by ion-exchange) shows additional features that might be related to the presence of Au. Pd-based heteropolyacid prepared by impregnation or ion-exchange showed no additional Raman bands.

Fig. 5 Raman spectra recorded for a series of heteropolyacid based catalysts prepared by ion-exchange and impregnation methods. Wt% loadings of Au and Pd range from 0.01 to 5% for catalysts: A – Cs_2.8H_0.2PW₁₂O₄₀, B – Pd_0.65Cs_1.5H_0.2PW₁₂O₄₀, C – Pd_0.15Cs_2.5H_0.2PW₁₂O₄₀, D – Pd_0.10Au_0.03Cs_2.5H_0.2PW₁₂O₄₀, E – Pd_0.075Au_0.05Cs_2.5H_0.2PW₁₂O₄₀ and F – 5% PdCs_2.5H_0.2PW₁₂O₄₀5% Au/Cs_2.5H_0.2PW₁₂O₄₀.

All samples (whether with Pd and Au added by either impregnation or ion exchange methods) show similar XRD spectra (Fig. 6a–c). The detected reflections are consistent with the cubic structure of the Cs₃PW₁₂O₄₀ salt (ICDD number 00-051-1857) as the catalysts prepared have similar Cs content. The XRD spectra of the Au and Pd-supported Cs_2.8H_0.2PW₁₂O₄₀ catalysts do not give rise to any changes to the reflections associated with the Cs salt. The 5% Pd supported catalyst does not show any new reflections that could be assigned to either metallic Pd or PdO, this could be attributed to the very small crystallite size of these Pd or PdO species. The XRD spectra of 5% Au supported catalyst shows reflections at 2θ values (38, 44, 64 and 77°) characteristic of metallic Au. Both monometallic and bimetallic Au and Pd ion-exchanged catalysts showed no new features when compared to Cs_2.8H_0.2PW₁₂O₄₀. This indicates that the structure of the Cs-substituted salt remained intact after ion exchange. The XRD spectra of the catalyst with lower Cs content, Pd_0.075Au_0.05Cs_1.5H_0.2PW₁₂O₄₀, showed peaks associated with Cs_2.8H_0.2PW₁₂O₄₀ along with some new weak features at 2θ = 15 and 45°assigned to the presence of H₃PW₁₂O₄₀ (see Fig. 6c). The structure of this lower Cs-containing catalyst (Pd_0.075Au_0.05Cs_1.5H_0.2PW₁₂O₄₀) lies in between the structure of Cs₃PW₁₂O₄₀ and H₃PW₁₂O₄₀.

Fig. 6 Powder X-ray diffraction paterns for (a) Au-only and Pd-only HPA-based catalysts, (b) Au–Pd HPA-based catalysts, (c) H₃PW₁₂O₄₀.

NH₃-TPD profiles (Fig. 7) recorded for Pd, Au and Au–Pd heteropolyacid catalysts prepared by ion-exchange confirmed that all catalysts have acidic character with the amount and strength of acid sites depending on both the Cs content and the type(s) of other metals (Au, Pd or Au + Pd) incorporated into the structure (see Table 7). Catalysts containing Cs_2.5 all show a small ammonia desorption at around 800 °C compared to catalysts containing less cesium, Cs_1.5, which show a large desorption at 570 °C, which is around four times larger. This indicates that lower Cs content leads to more acid sites, which are weaker in strength than Cs_2.5 containing catalysts. On incorporation of metals into the heteropolyacid structure the acidity is modified, it appears that for Cs_2.5 catalysts incorporation of one metal slightly reduces the acidity when compared to the support material, Cs_2.5H_0.5PW₁₂O₄₀. However, on incorporation of both Au and Pd more ammonia is desorbed from the catalyst, showing a slight increase in acidity. This is also true for the Cs_1.5 containing catalysts, with relative NH₃ uptakes increasing from 3.57 to 4.73 on incorporation of both Au and Pd when compared to a similar metal loading of Pd only. As noted earlier, the observed trends in activity appear to be related to the acidity of these catalysts. The higher loading Au–Pd catalysts have a higher number of acid sites that contribute to their higher activity under standard conditions and when using ambient temperature in water/methanol solvent. At ambient temperature in water-only solvent, the activity of the higher loading catalyst decreases because, although there are a larger number of acid sites, the catalysts comprise weaker acid sites compared to the lower loaded catalysts. In the absence of methanol (i.e. in water-only solvent), a strong acidic catalyst is required to stabilise the formed H₂O₂ from subsequent decomposition/hydrogenation.

Table 7 Relative NH₃ uptake

Catalyst	Relative NH3 uptake per gcat
Cs2.5H0.5PW12O40	1.00
Pd0.15Cs2.5H0.5PW12O40	0.81
Au0.1Cs2.5H0.5PW12O40	0.72
Pd0.075Au0.05Cs2.5H0.5PW12O40	1.11
Pd0.65Cs1.5H0.2PW12O40	3.57
Pd0.327Au0.217Cs1.5H0.2PW12O40	4.73

---

Fig. 7 TPD profiles of ion-exchanged HPA-based catalysts.

XPS data (Table 8) shows the presence of surface Pd and/or Au species for Pd-only, Au-only and Au–Pd heteropolyacid catalysts prepared by impregnation. Where possible we also give the data for the oxidation states of the Au and Pd, although in some of the catalysts the amount of these metals present is too little to be able to unambiguously determine the oxidation state. The Au–Pd/heteropolyacid catalyst shows a Pd/Au ratio of ca. 5.7, which suggests the formation of Au–Pd nanoparticles with a core–shell structure (Au-rich core and Pd-rich surface). With the exception of the Pd-only ion-exchanged catalyst (which showed surface Pd species), XPS data for the Au-only and Au–Pd heteropolyacid catalysts prepared by ion-exchange did not show any surface Au and/or Pd species. While the absence of these surface species might be related to the relatively low loading (0.5–0.6 wt%) of Au or Au + Pd in these catalysts, the fact that we observe surface Pd species for the Pd-only ion-exchanged heteropolyacid catalyst (which also has relatively very low Pd loading, 0.49 wt%) might suggest that most of the Au and/or Pd species present in the Au-only and Au–Pd ion-exchanged heteropolyacid catalysts are not on the surface but incorporated within the heteropolyacid porous structure. It is possible that the presence of Au together with Pd leads to the formation of Au–Pd nanoparticles, which are much smaller (than the Pd-only particles) and can therefore be easily incorporated into the porous framework. It is clear, however, that the addition of Au to these catalysts has a marked effect on the catalysis. As the gold is not on the surface it may be that the effect we observe is similar to that we have observed for supported alloy nanoparticles with Pd-rich shells and Au-rich cores for oxide-supported catalysts.³⁹ However, attempts to characterise the heteropolyacid catalysts used in this study with electron microscopy have not proved feasible due to the extreme beam-sensitivity of these materials.

Table 8 Surface stoichiometry for catalyst components determined by XPS

	Surface analysis
Sample	Au/Pd oxidation state	Stoichiometry
		Pd	Au	Cs	P	W
a nd = not detected by XPS, the metal content was too low for this method.
Pd0.15Cs2.5H0.2PW12O40	Pd^2^+(100%)	0.20	0	2.02	0.98	12.0
Pd0.1Au0.0333Cs2.5H0.2PW12O40	nd^a			2.00	1.49	12.0
Pd0.075Au0.05Cs2.5H0.2PW12O40	nd			2.12	1.17	12.0
Cs2.8H0.2PW12O40	—	—	—	2.50	1.08	12.0
2.5% Au/2.5% Pd/Cs2.8H0.2PW12O40	Pd^2^+(43%) Pd^0 (57%) Au^0(100%)	0.86	0.15	5.21	2.03	12.0
5% Pd/Cs2.8H0.2PW12O40	Pd^2^+(85%) Pd^0 (15%)			1.94	1.09	12.0
5% Au/Cs2.8H0.2PW12O40	Au^0 (100%)			2.07	2.25	12.0
Au0.1Cs2.5H0.2PW12O40	nd			1.64	1.80	12.0

---

Conclusions

Pd-only, Au-only and Au–Pd-exchanged and supported heteropolyacid catalysts have been investigated for the direct synthesis of H₂O₂ from H₂ and O₂. The heteropolyacid-based Au–Pd catalysts are more effective for H₂O₂ formation than corresponding Pd-only catalysts prepared using the same impregnation or ion-exchanged method. The supported Au–Pd heteropolyacid catalyst (2.5% Au/2.5% Pd/Cs_2.8H_0.2PW₁₂O₄₀) was the most active catalyst under standard reaction conditions. This catalyst showed significantly higher H₂O₂ productivity than the most effective catalyst (2.5% Au/2.5% Pd/C (2% HNO₃) we have previously reported for the direct synthesis of H₂O₂ using standard reaction conditions. The catalyst also showed the best H₂O₂ synthesis activity when using 2% H₂/air for direct H₂O₂ synthesis. The Au–Pd-exchanged heteropolyacid catalysts (Pd_0.1Au_0.0333Cs_2.5H_0.2PW₁₂O₄₀ and Pd_0.1Au_0.0333Cs_2.5H_0.2PW₁₂O₄₀) were slightly less active for H₂O₂ synthesis using 2% H₂/air than the Au–Pd-supported heteropolyacid catalyst but showed superior activity compared to the Pd-only exchanged heteropolyacid (Pd_0.15Cs_2.5H_0.2PW₁₂O₄₀) and supported Au–Pd/C (untreated and acid-treated) catalysts. Although the Au–Pd-exchanged heteropolyacid catalysts showed comparable activity to the Pd-only heteropolyacid catalyst and were less active than Au–Pd/C (2% HNO₃) and Au–Pd/Cs_2.8H_0.2PW₁₂O₄₀ under standard reaction conditions, these catalysts proved to be most effective for H₂O₂ synthesis at ambient temperature and when using H₂O-only solvent. Since the Au–Pd-exchanged heteropolyacid catalysts only contain ca. 3% of the total Au and Pd metal content present in the more conventional supported Au–Pd heteropolyacid and carbon catalysts, we consider that these catalysts could be the basis of more efficient catalysts for the direct H₂O₂ synthesis at ambient temperature and using air as compared with oxygen. It should be noted that the presence of Au in these heteropolyacid-type catalysts is crucial in achieving superior catalytic performance under these reaction conditions. Based on environmental and process economics considerations, the use of ambient temperature, water-only solvent and 2% H₂/air for industrial production of H₂O₂ will be preferred to low temperature (2 °C), methanol-containing solvent and 5% H₂/CO₂ + 25% O₂/CO₂ respectively. However, it should be noted that the over-riding consideration for the commercialisation of catalysts for the direct synthesis of hydrogen peroxide is the selectivity of the process and this is a factor that will require further improvement with these new heteropolyacid-supported catalysts. We therefore consider the improved performance of the Au–Pd-exchanged heteropolyacid catalysts under these preferred conditions an important advance in the development of catalysts for the direct synthesis of H₂O₂.

Acknowledgements

We thank the EPSRC for financial support.

References

1. C. W. Corti, R. J. Holliday and D. T. Thompson, Appl. Catal., A, 2005, 291, 253.
2. http://news.bbc.co.uk/1/hi/england/london/4197500.stm .
3. H. Henkel and W. Weber, US Pat., 1108752, 1914.
4. G. A. Cook, US Pat., 2368640, 1945.
5. A. I. Dalton, E. J. Greskovich and R. W. Skinner, US Pat., 4336238, 1982.
6. L. W. Gosser, US Pat., 4889705, 1989.
7. H. Nagashima, Y. Ishiuchi and Y. Hiramatsu, US Pat., 5236692, 1993.
8. K. T. Chuang and B. Zhou, US Pat., 5925588, 1999.
9. D. P. Dissanayake and J. H. Lunsford, J. Catal., 2002, 206, 173.
10. J. H. Lunsford, J. Catal., 2003, 216, 455.
11. S. Chinta and J. H. Lunsford, J. Catal., 2004, 225, 249.
12. Y. Han and J. H. Lunsford, J. Catal., 2005, 230, 313.
13. V. R. Choudhary and C. Samanta, J. Catal., 2006, 238, 28.
14. Q. Liu and J. H. Lunsford, J. Catal., 2006, 239, 237.
15. V. R. Choudhary and P. Jena, J. Catal., 2007, 246, 434.
16. C. Samanta and V. R. Choudhary, Catal. Commun., 2007, 8, 73.
17. C. Samanta and V. R. Choudhary, Appl. Catal., A, 2007, 330, 23.
18. V. R. Choudhary, C. Samanta and T. V. Choudhary, Catal. Commun., 2007, 8, 1310.
19. V. R. Choudhary, C. Samanta and P. Jena, Appl. Catal., A, 2007, 332, 70.
20. C. Samanta and V. R. Choudhary, Chem. Eng. J., 2008, 136, 126.
21. P. Landon, P. J. Collier, A. J. Papworth, C. J. Kiely and G. J. Hutchings, Chem. Commun., 2002, 2058.
22. P. Landon, P. J. Collier, A. F. Carley, D. Chadwick, A. J. Papworth, A. Burrows, C. J. Kiely and G. J. Hutchings, Phys. Chem. Chem. Phys., 2003, 5, 1917.
23. J. K. Edwards, B. Solsona, P. Landon, A. F. Carley, A. Herzing, M. Watanabe, C. J. Kiely and G. J. Hutchings, J. Mater. Chem., 2005, 15, 4595.
24. J. K. Edwards, B. Solsona, P. Landon, A. F. Carley, A. Herzing, M. Watanabe, C. J. Kiely and G. J. Hutchings, J. Catal., 2005, 236, 69.
25. A. Lopez-Sanchez, N. Dimitratos, P. Miedziak, E. Ntainjua, J. K. Edwards, D. Morgan, A. Carley, R. Tiruvalam, C. J. Kieley and G. J. Hutchings, Phys. Chem. Chem. Phys., 2008, 10, 1921.
26. J. K. Edwards, A. F. Carley, A. Herzing, C. J. Kiely and G. J. Hutchings, Faraday Discuss., 2007, 138, 225.
27. G. Li, J. Edwards, A. F. Carley and G. J. Hutchings, Catal. Today, 2007, 122, 361.
28. J. K. Edwards, A. Thomas, A. F. Carley, A. Herzing, C. J. Kiely and G. J. Hutchings, Green Chem., 2008, 10, 388.
29. E. Ntainjua, N. J. K. Edwards, A. F. Carley, J. A. Lopez-Sanchez, J. A. Moulijn, A. A. Herzing, C. J. Kiely and G. J. Hutchings, Green Chem., 2008, 10, 1162.
30. G. J. Hutchings, Chem. Commun., 2008, 1148.
31. E. Ntainjua, N. M. Piccinini, J. C. Pritchard, J. K. Edwards, A. F. Carley, J. A. Moulijn and G. J. Hutchings, ChemSusChem, 2009, 2, 575.
32. E. Ntainjua, N. M. Piccinini, J. C. Pritchard, J. K. Edwards, A. F. Carley, J. A. Moulijn, C. J. Kiely and G. J. Hutchings, ChemCatChem., 2009, 1, 479–484.
33. J. K. Edwards, B. Solsona, E. Ntainjua, N. A. F. Carley, A. A. Herzing, C. J. Kiely and G. J. Hutchings, Science, 2009, 323, 1037.
34. M. Sun, J. Zhang, Q. Zhang, Y. Wang and H. Wan, Chem. Commun., 2009, 5174.
35. S. Park, S. H. Lee, S. H. Song, D. R. Park, S. H. Baeck, T. J. Kim, Y-M. Chung, S-H. Oh and I. K. Song, Catal. Commun., 2009, 10, 391.
36. M. Piccinini, E. Ntainjua, J. K. Edwards, A. F. Carley, J. A. Moulijn and G. J. Hutchings, Phys. Chem. Chem. Phys., 2010, 12, 2488.
37. N. Essayem, A. Holmqvist, P. Y. Gayraud, J. C. Vedrine and Y. Ben Taarit, J. Catal., 2001, 197, 273.
38. Z. P. Pai, D. I. Kochubey, P. V. Berdnikova, V. V. Kanazhevskiy, I. Y. Prikhod'ko and Y. A. Chesalov, J. Mol. Catal. A: Chem., 2010, 332, 122.
39. D. I Enache, J. K. Edwards, P. Landon, B. Solsona-Espriu, A. F. Carley, A. A. Herzing, M. Watanabe, C. J. Kiely, D. W. Knight and G. J. Hutchings, Science, 2006, 311, 362.

---