Direct synthesis of hydrogen peroxide with no ionic halides in solution

Abstract

We have studied the direct synthesis of hydrogen peroxide from H₂/O₂ mixtures using a palladium catalyst loaded on a sulfonic acid-functionalized resin in methanol as a solvent and using alternative bromine promoters in the absence of ionic halides in the reaction medium. Three different bromo-organic compounds were tested, 2-bromo-2-methylpropane (t-ButBr), 2-bromopropane (i-PrBr), and bromobenzene (PhBr), and compared to HBr as the reference. The direct synthesis of hydrogen peroxide was tested under 9.5 and 5 MPa overall pressures. The use of t-ButBr and PhBr precursors yielded H₂O₂ concentrations higher than 9 wt% in less than 3 h of reaction, while the H₂O₂ concentration obtained in the presence of i-PrBr was very low. A higher concentration of H₂O₂ and higher selectivity are obtained with t-ButBr than with PhBr, but when using a t-ButBr promoter some hydrolysis is expected to occur in the course of the reaction because bromide ions are detected in the liquid samples, but no bromide ions are detected when PhBr is used.

---

Introduction

Hydrogen peroxide is a green oxidant and an important commodity chemical with extensive applications in a large variety of processes, such as its use as a whitening agent in the pulp and paper industry and as reactant for chemical synthesis in the pharmaceutical and waste water treatment industries.¹ At the industrial scale, hydrogen peroxide is produced mainly by the sequential hydrogenation and oxidation of an alkyl anthraquinone. However, this process requires high-volume reactors and costly separation steps, thus making the process economical only for large units.²

A very attractive alternative to this process is the production of hydrogen peroxide directly from the reaction between hydrogen and oxygen.² The direct reaction between hydrogen and oxygen has been a research target for many years,³ but no commercial process has been devised, mostly because of its very hazardous nature (the explosive limits of hydrogen/oxygen gas mixtures are very broad, but interestingly, the lower flammability limit (LFL) decreases with increasing pressure) in air or even pure oxygen, and the reported conversions and selectivity are poor.^2–5 The best catalysts for this reaction are based on Pd nanoparticles deposited on solid substrates.^1–3 It has been demonstrated that the incorporation of small amounts of H⁺ and Br⁻ into the liquid phase significantly enhances the yield of H₂O₂ production.^6–12 These ions can be obtained from HBr or from a combination of strong acids and inorganic bromides.^6,13 However, working with highly concentrated acid solutions requires the use of special equipment to avoid corrosion and additional separation steps. The presence of strong acids in solution can be avoided by using supports functionalized with acidic groups.^14–20 Catalysts based on palladium-loaded sulfonic acid-functionalized resin proved to be very efficient in the synthesis of hydrogen peroxide by the direct reaction between hydrogen and oxygen in a neutral medium with the addition of minute amounts of HBr.^14–16,21 However, the presence of this small amount of ionic halides in solution still requires the use of special equipment to avoid corrosion, as well as removal from the final product.

For this reason, this work was undertaken to test the behavior of Pd-PhSO₃H catalyst in the absence of ionic halides, by changing the source of the bromine precursor from hydrogen bromide (HBr) to an organic compound that contains bromide. For this purpose, three possible candidates have been studied: 2-bromo-2-methylpropane (t-ButBr), 2-bromopropane (i-PrBr), and bromobenzene (PhBr) (Fig. 1).

Fig. 1 Molecular structure of the organic bromide employed: 2-bromo-2-methylpropane (t-ButBr), 2-bromopropane (i-PrBr), and bromobenzene (PhBr).

Experimental methods

Materials

Methanol (≥99.9%), organic compounds containing bromide (t-ButBr (≥98% GC), i-PrBr (≥99% GC), and PhBr (≥99.5% GC)), and HBr (≥99.99% based on trace metals analysis) were purchased from Sigma-Aldrich. Pd(II) acetate was purchased from Alfa-Aesar. LANXESS AG kindly provided strongly acidic macroreticular resin, functionalized with sulfonic groups (LK2621).

Catalysts

The catalyst used in this study, Pd-PhSO₃H, was prepared using LK2621 mesoporous ion exchange resins.^14,15 First, the resin was washed three times with acetone, using equal volumes of solvent and resin. A resin suspension (4 g) was then stirred with 50 mL of acetone. A Pd(II) acetate solution (86 mg) in acetone (20 mL) was added dropwise to this suspension. The suspension was concentrated to half of the solvent in a rotary evaporator, using a bath temperature of 318 K. The remaining solution was filtered off, and the solid obtained was washed and air-dried at 333 K for 2 h. The amount of palladium incorporated was 2 wt%.

Hydrogen peroxide direct synthesis

The catalytic tests were performed in a high-pressure stirred reactor (Autoclave Engineers) working in semi-batch conditions, that is the liquid was kept inside the reactor for whole experiment while a continuous gas flow was fed during the experiment, the pressure in the reactor was kept constant by means of a pressure controller. In a typical run, a total of 3.38 g of the Pd-PhSO₃H catalyst was placed in the autoclave with 320 g of methanol and a bromide source; the system was then pressurized under N₂ flow at the reaction pressure (9.5 or 5.0 MPa) and heated at 313 K. Then, flows of oxygen and hydrogen were added successively without stirring to avoid the reaction. The total flow of gas fed was 5300 mL (STP) min⁻¹, with a molar hydrogen concentration of 3.6% (outside flammability limits), an oxygen concentration of 46.4% and a nitrogen concentration of 50.0%. After a short period of stabilization, the reaction was started by stirring the mixture (1500 rpm). Hydrogen consumption was determined by on-line GC using a Varian CP-4900 microGC device. The hydrogen peroxide and water concentrations were determined by potassium permanganate and Karl Fisher standard titrations, respectively. Liquid samples at the end of the reaction were analyzed by ionic chromatography to detect the presence of bromide ions in solution, using a Dionex ICS-5000 equipped with an As24 column, this methodology implies a detection limit for bromide ions of 0.5 ppb.

Results and discussion

The physico-chemical characteristics of the catalyst have been previously reported by us.^14,15 XPS analysis of the samples showed only the presence of Pd^II ions interacting with the –SO₃H groups¹⁴ Then, we study the replacement of HBr by t-ButBr (Fig. 1). The behaviors of these two promoters were tested at 9.5 MPa overall pressure with 95 µmol of promoter added to the reactor. For comparative purposes, a blank experiment with no promoter in the reaction mixture was also performed.

The use of bromine promoter yields hydrogen peroxide (Fig. 2), but no H₂O₂ was detected after 2 h of reaction in the blank experiment (without bromine promoter). The hydrogen peroxide concentration profile (Fig. 2) shows that the use of HBr gives slightly higher concentration than the use of t-ButBr. Similarly, the selectivity for hydrogen peroxide was slightly higher using the inorganic bromide (ca. 80%) than the organic bromide (ca. 75%).

Fig. 2 Influence of the bromide source on the H₂O₂ concentration and H₂O₂ selectivity profiles versus time in the direct reaction of H₂ and O₂ at 313 K and 9.5 MPa.

These preliminary results encouraged us to pursue the study of this bromine promoter. We previously showed the critical importance of the catalyst-bromine promoter ratio.^15,16 For this reason, we studied the effect of the amount of organic bromide (t-ButBr) added to the reaction medium and tried to optimize this amount to obtain the maximum yield of hydrogen peroxide. The amount of t-ButBr in the reactor varied from 95 to 190 µmol.

As expected, changes in the promoter concentration produce changes in the hydrogen peroxide concentration profiles (Fig. 3). The H₂O₂ yield first increases with the t-ButBr amount, reaches a maximum and then decreases at higher concentrations. The selectivity for hydrogen peroxide undergoes clear changes in this concentration range, following the same trend as the hydrogen peroxide yield. The maximum yield of hydrogen peroxide is reached with 160 µmol of t-ButBr.

Fig. 3 Influence of bromide quantity on the H₂O₂ concentration and H₂O₂ selectivity profiles versus time for the direct reaction of H₂ and O₂ at 313 K and 9.5 MPa.

The reaction results observed with the optimum concentration of t-ButBr (160 µmol) and the HBr counterpart (95 µmol) are similar (Fig. 2 and 3), allowing us to obtain high concentrations of hydrogen peroxide with no ionic halides.

We have shown that this system allows the direct synthesis of hydrogen peroxide at 9.5 MPa of overall pressure, but working at this pressure has some practical problems at the industrial scale. If it is possible to work at lower overall pressure, the reduction of CAPEX and OPEX in an industrial plant will be high. To account for this possibility, tests were conducted at 5 MPa, holding the other reaction conditions constant.

The use of low overall pressure yields hydrogen peroxide for both promoters used (Fig. 4), and the hydrogen peroxide profile and selectivity for H₂O₂ were very similar (Fig. 4).

Fig. 4 Direct synthesis of H₂O₂ reaction at P = 5 MPa, with two different promoters at optimized concentrations.

The slope of hydrogen peroxide profile is slightly lower for the experiments at 5 MPa (Fig. 4) than their counterparts at 9.5 MPa (Fig. 2 and 3). This effect was expected because the solubility of the gases in the solvent is lower at lower overall pressure. As a consequence, a longer reaction time is required to reach the same level of H₂O₂ concentration. The selectivity for hydrogen peroxide is slightly lower at 5 MPa (approximately 72–75% at 8 wt% of H₂O₂) than at 9.5 MPa (75–80% at 8 wt% of H₂O₂).

These interesting results obtained with the t-ButBr precursor encourage us to use other Br-containing organic compounds, such as 2-bromopropane (i-PrBr) and bromobenzene (PhBr) (Fig. 1), using the same reaction conditions as for t-BuBr. We detected high hydrogen peroxide concentrations when t-ButBr or PhBr were employed as promoter, while the hydrogen peroxide concentration was very low with i-PrBr as the promoter (Fig. 5). The final concentration reached with t-ButBr as promoter was slightly higher than with PhBr. A similar trend was observed for the hydrogen peroxide selectivity profiles.

Fig. 5 Direct synthesis of H₂O₂ reaction results at 313 K and 5.0 MPa using a different organic bromine promoter.

The exceptional results and molecular structure of t-ButBr suggested the possibility of decomposition of the bromo-compounds in the reaction media, yielding the formation of bromide ions in the liquid phase. For this reason, the reaction mixture samples were analyzed by ionic chromatography to detect the presence of bromide ions. When t-ButBr was employed as promoter, bromide ions were detected, but no bromide was detected when PhBr was the promoter used. This effect may be related to the molecular structure of t-ButBr (Fig. 1), a bromine bonded to a tertiary carbon atom, which makes the compound more reactive for hydrolysis, yielding HBr in the liquid phase under the reaction conditions used in this study. We have analyzed the stability of t-BuBr in a 10 wt% of hydrogen peroxide in methanol. In this experiment, t-butanol was formed and complete conversion of t-BuBr was measured by NMR. This reactivity was not observed with the other promoters.

In view of these results, where the use of t-ButBr finally yielded bromide ions, we decide to continue the study with the PhBr promoter, which yields high hydrogen peroxide concentrations, though slightly lower than the ones obtained using t-ButBr or HBr. Again, the concentration of PhBr was optimized in order to maximize the yield of hydrogen peroxide. The concentration range that was explored varied between 95 and 200 µmol. Varying the promoter concentration alters the hydrogen peroxide production profiles (Fig. 6).

Fig. 6 Reaction results in the direct synthesis of hydrogen peroxide at 313 K and 5.0 MPa, using different amounts of PhBr promoter.

The final hydrogen peroxide concentration increases with promoter amount to a maximum at approximately 120–160 µmol of PhBr, then decreases for higher amounts of promoter. A similar trend was observed for the selectivity of hydrogen peroxide: the maximum selectivity was reached when using between 120 and 160 µmol of PhBr.

To summarize, comparing the reaction results of the three Br-containing promoters with the best behavior (HBr, t-ButBr and PhBr) and using the optimal concentration for each one reveals that the yield and selectivity for H₂O₂ depend on the promoter that was used (Fig. 7). After 180 minutes of reaction, the selectivity for hydrogen peroxide is 66% for PhBr and approximately 75% for HBr and t-ButBr. Similarly, the H₂O₂ concentration reached is ca. 9.4% for PhBr and above 10% for HBr and t-ButBr.

Fig. 7 Influence of the promoter on the selectivity for H₂O₂ (top) and H₂O₂ and water concentration (bottom) at 313 K and 5.0 MPa after 180 minutes of reaction.

Density functional theory (DFT) studies of the H₂–O₂ reaction on Pd catalysts distinguish two types of sites: more unsaturated sites, such as corners or edges, and more saturated sites, such as a (111) face.¹⁰ The energy profiles of the H₂ + O₂ reaction suggest that H₂O₂ would be smoothly produced at more saturated sites, whereas the formation of H₂O and the decomposition of H₂O₂ would be preferred at more unsaturated sites.

The role of bromide ions is to block the unsaturated sites and reduce the secondary reactions because the adsorption energy of bromide and protons is higher than H₂ and O₂.¹⁰ This role of the bromide species can explain the effect of the different promoters employed in this work. Firstly, hydrogen bromide may be adsorbed easily on unsaturated sites and yields good selectivity for hydrogen peroxide. The use of t-ButBr promoter yields its hydrolysis to form bromide ions, which act similarly to the HBr promoter and gives similar reaction results (Fig. 7). The PhBr may be adsorbed at the unsaturated sites, but its molecular size hinders its adsorption, yielding lower selectivity for hydrogen peroxide because the secondary reactions at the unsaturated sites are less depressed. A dual effect with ring and Br adsorption as well as electron transfer inside the molecule could be involved but the way by which PhBr promotes the H₂O₂ synthesis is still unclear.

Conclusions

There are alternatives to inorganic bromides as promoters in the direct synthesis of hydrogen peroxide. Replacing HBr with organic bromo-compounds can yield similar reaction results. The reduction of the overall pressure from 9.5 MPa to 5 MPa reduces the hydrogen peroxide concentration slope, but the same concentration can be reached with a small increase in reaction time.

The best results have been obtained with 2-bromo-2-methylpropane (t-ButBr), but analysis of the liquid phase by ionic chromatography detected bromide ions. This compound is therefore not a real alternative for the direct synthesis of hydrogen peroxide in the absence of bromide ions. However, the use of PhBr is a more promising alternative, as no bromide ions were observed in the liquid phase, and good values of hydrogen peroxide concentration and selectivity were reached.

Acknowledgements

The authors acknowledge financial support from Solvay (Brussels). Lanxess AG is gratefully acknowledged for providing the K2621 resin.

Notes and references

1. J. M. Campos-Martin, G. Blanco-Brieva and J. L. G. Fierro, Angew. Chem., Int. Ed., 2006, 45, 6962–6984.
2. C. Samanta, Appl. Catal., A, 2008, 350, 133–149.
3. Y. Yi, L. Wang, G. Li and H. Guo, Catal. Sci. Technol., 2016, 6, 1593–1610.
4. G. Centi and S. Perathoner, Catal. Today, 2009, 143, 145–150.
5. Q. Liu, J. C. Bauer, R. E. Schaak and J. H. Lunsford, Appl. Catal., A, 2008, 339, 130–136.
6. N. M. Wilson and D. W. Flaherty, J. Am. Chem. Soc., 2016, 138, 574–586.
7. V. R. Choudhary, C. Samanta and P. Jana, Appl. Catal., A, 2007, 332, 70–78.
8. V. R. Choudhary and P. Jana, Appl. Catal., A, 2007, 329, 79–85.
9. P. Centomo, C. Meneghini, S. Sterchele, A. Trapananti, G. Aquilanti and M. Zecca, Catal. Today, 2015, 248, 138–141.
10. T. Deguchi and M. Iwamoto, J. Phys. Chem. C, 2013, 117, 18540–18548.
11. P. Centomo, C. Meneghini, S. Sterchele, A. Trapananti, G. Aquilanti and M. Zecca, ChemCatChem, 2015, 7, 3712–3718.
12. P. Biasi, S. Sterchele, N. Gemo, A. Shchukarev, A. R. Rautio, J. P. Mikkola, P. Canu, P. Centomo, M. Zecca, K. Kordas and T. Salmi, AIChE J., 2016 DOI:10.1002/aic.15382.
13. R. Burch and P. R. Ellis, Appl. Catal., B, 2003, 42, 203–211.
14. G. Blanco-Brieva, E. Cano-Serrano, J. M. Campos-Martin and J. L. G. Fierro, Chem. Commun., 2004, 1184–1185.
15. G. Blanco-Brieva, M. C. Capel-Sanchez, M. P. de Frutos, A. Padilla-Polo, J. M. Campos-Martin and J. L. G. Fierro, Ind. Eng. Chem. Res., 2008, 47, 8011–8015.
16. G. Blanco-Brieva, M. P. De Frutos Escrig, J. M. Campos-Martin and J. L. G. Fierro, Green Chem., 2010, 12, 1163–1166.
17. J. Kim, Y.-M. Chung, S.-M. Kang, C.-H. Choi, B.-Y. Kim, Y.-T. Kwon, T. J. Kim, S.-H. Oh and C.-S. Lee, ACS Catal., 2012, 2, 1042–1048.
18. Y. M. Chung, Y. T. Kwon, T. J. Kim, S. H. Oh and C. S. Lee, Chem. Commun., 2011, 47, 5705–5707.
19. C. Burato, P. Centomo, M. Rizzoli, A. Biffis, S. Campestrini and B. Corain, Adv. Synth. Catal., 2006, 348, 255–259.
20. C. Burato, S. Campestrini, Y.-F. Han, P. Canton, P. Centomo, P. Canu and B. Corain, Appl. Catal., A, 2009, 358, 224–231.
21. G. Blanco-Brieva, J. M. Campos-Martin, M. P. De Frutos and J. L. G. Fierro, Catal. Today, 2010, 158, 97–102.

---

Footnote

† Present address: Departamento de Química Física Aplicada, Facultad de Ciencias, Universidad Autónoma de Madrid, Francisco Tomás y Valiente, 7, 28049 Madrid, Spain.

---