Recent Advances in the Direct Synthesis of Hydrogen Peroxide Using Chemical Catalysis—A Review

: Hydrogen peroxide is an important chemical of increasing demand in today’s world. Currently, the anthraquinone autoxidation process dominates the industrial production of hydrogen peroxide. Herein, hydrogen and oxygen are reacted indirectly in the presence of quinones to yield hydrogen peroxide. Owing to the complexity and multi-step nature of the process, it is advantageous to replace the process with an easier and straightforward one. The direct synthesis of hydrogen peroxide from its constituent reagents is an effective and clean route to achieve this goal. Factors such as water formation due to thermodynamics, explosion risk, and the stability of the hydrogen peroxide produced hinder the applicability of this process at an industrial level. Currently, the catalysis for the direct synthesis reaction is palladium based and the research into ﬁnding an effective and active catalyst has been ongoing for more than a century now. Palladium in its pure form, or alloyed with certain metals, are some of the new generation of catalysts that are extensively researched. Additionally, to prevent the decomposition of hydrogen peroxide to water, the process is stabilized by adding certain promoters such as mineral acids and halides. A major part of today’s research in this ﬁeld focusses on the reactor and the mode of operation required for synthesizing hydrogen peroxide. The emergence of microreactor technology has helped in setting up this synthesis in a continuous mode, which could possibly replace the anthraquinone process in the near future. This review will focus on the recent ﬁndings of the scientiﬁc community in terms of reaction engineering, catalyst and reactor design in the direct synthesis of hydrogen peroxide.


Introduction
Hydrogen peroxide (H 2 O 2 ) is a colorless, odorless, and slightly acidic liquid used mainly as an oxidant in chemical synthesis [1][2][3]. It is an atom efficient, benign, and eco-friendly oxidant that produces water or oxygen as a degradation product, depending on the catalyst used [4,5]. Commercial H 2 O 2 production has a concentration range of 30-70% and the areas of application depend on the concentration of H 2 O 2 used. Usually, household, medical/dental, and cosmetic applications need diluted concentrations of Ca. 3-5% H 2 O 2 . Higher concentrations of up to 70% are needed for synthesis, wastewater treatment, mining, and bleaching applications. The electronics  [16,17] (Scheme 1c) 1914 Hugo Henkel and Walter Weber Chemical First account of direct synthesis of hydrogen peroxide using its constituent gases [18] Scheme 2a shows the reaction mechanism of the AO process and Scheme 2b depicts the practice of H2O2 manufacture in the industry. The hydrogenation chamber is usually a slurry reactor consisting of the alkylated anthraquinone dissolved in a so-called "working solution" along with a  [3,8,23]. (Image taken from Berl [23], reproduced here with the kind permission of the Journal of the Electrochemical Society) (e) Hydrazobenzene oxidation to produce H 2 O 2 (scheme modified from Walton and Filson [19], copyright 1932, American Chemical Society). (f) Georg Pfleiderer Process of producing H 2 O 2 , an adaptation of the Walton and Filson Process [20] (g) Shell 2-propanol Process to produce acetone/H 2 O 2 [3] (Scheme taken from [3] [3,8,23]. (Image taken from Berl [23], reproduced here with the kind permission of the Journal of the Electrochemical Society) (e) Hydrazobenzene oxidation to produce H2O2 (scheme modified from Walton and Filson [19], copyright 1932, American Chemical Society). (f) Georg Pfleiderer Process of producing H2O2, an adaptation of the Walton and Filson Process [20] (g) Shell 2-propanol Process to produce acetone/H2O2 [3] (Scheme taken from [3], reproduced here with the kind permission of John Wiley and Sons)

State of the Art in the Industrial Production of H2O2
On an industrial scale, the anthraquinone autoxidation (AO) process described by H.J. Riedl and G. Pfleiderer produces H2O2 in a cyclic manner (Scheme 2). Scheme 2. (a) Mechanism of the autoxidation (AO) process developed by Riedl-Pfleiderer [21,24] (Image taken from Li et al. [24], 2017, reprinted here with the kind permission of Springer Nature) and (b) block diagram of the AO process steps to synthesise H2O2 (scheme modified from Campos-Martin et al. [7], reprinted here with the kind permission of John Wiley and Sons).
Scheme 2a shows the reaction mechanism of the AO process and Scheme 2b depicts the practice of H2O2 manufacture in the industry. The hydrogenation chamber is usually a slurry reactor consisting of the alkylated anthraquinone dissolved in a so-called "working solution" along with a Scheme 2. (a) Mechanism of the autoxidation (AO) process developed by Riedl-Pfleiderer [21,24] (Image taken from Li et al. [24], 2017, reprinted here with the kind permission of Springer Nature) and (b) block diagram of the AO process steps to synthesise H 2 O 2 (scheme modified from Campos-Martin et al. [7], reprinted here with the kind permission of John Wiley and Sons). Scheme 2a shows the reaction mechanism of the AO process and Scheme 2b depicts the practice of H 2 O 2 manufacture in the industry. The hydrogenation chamber is usually a slurry reactor consisting of the alkylated anthraquinone dissolved in a so-called "working solution" along with a catalyst. Hydrogen gas is initially fed through the reactor for hydrogenation to alkyl anthrahydroquinol. After this, the contents of the hydrogenation chamber are degassed in a separate chamber to remove traces of hydrogen. Following this step, the alkyl anthrahydroquinol is transferred to the oxygenation chamber. Here, oxygen or air is used to produce H 2 O 2 and alkyl anthrahydroquninone, which is then transferred to a second degassing chamber to remove trace oxygen. The degassed solution is then extracted with water to yield H 2 O 2 and the working solution is transferred to the hydrogenation chamber to complete the synthetic cycle. In the mid-1990s, the world capacity of 100% H 2 O 2 was approximately 1.5 million tons with an average plant capacity of around 20 kt-40 kt per annum. By 2015, the capacity was 5.5 million tons with a plant capacity of 300 kt per annum [25]. Solvay (30%), Evonik (20%), and Arkema (13%) represented the global contributors of H 2 O 2 [26].

Why Is There an Increased Interest in the Direct Synthesis of Hydrogen Peroxide?
Currently, the industrial manufacture of H 2 O 2 is based on the work of Riedl and Pfleiderer, also known as the AO process [21] (Scheme 2) [24] using polynuclear hydroquinones [1]. Although this process is capable of meeting the world's H 2 O 2 demand, it does have certain drawbacks such as: • excessive use of solvents for the process • a negative environmental impact owing to the production of unwanted waste • complex and multi-step process • mass transfer limitations and low efficiency • transport limitations of reactants between reactors • organic contamination of H 2 O 2 stemming from organic solvents or hydroquinones/hydroquinols during liquid-liquid extraction [27,28].
Considering environmental issues and resource conservation aspects, academia and the industry have set out to design benign and non-polluting processes. The principles of green chemistry helps in achieving this goal. The principles outlined in the 1990s clearly state that it is necessary to design and execute industrial processes that are clean, benign, non-polluting, and safe [29][30][31][32]. Two important terms, atom utilization and E-factor, are important in assessing the greenness of a process. Atom utilization, atom efficiency or atom selectivity (AE) maybe defined as the actual mass of reactants that actually end up in the final desired product, the rest of which is termed as "waste". AE helps in assessing the amount of wastes generated by a certain process, which in turn, will determine the E-factor. The E-factor is the ratio of the amount of waste produced to the amount of desired product [33][34][35]. Considering the AO process, the amount of solvents and the alkylated hydroquinones used in the process are waste products, as the reaction is not atom efficient [1,3,7,8]. This brings up an important question: why is the AO process still practiced at an industrial level if it is not sustainable? The answer: operating the AO process is economically feasible at a scale of 1 × 10 5 tons per annum producing high concentrations of H 2 O 2 , which are diluted prior to use. For a majority of the applications mentioned previously in the introductory section (Section 1), diluted versions of H 2 O 2 (typically 3-8%) is required, in small amounts, and on site. To circumvent the issues mentioned above and promote a green H 2 O 2 production process, the direct synthesis of H 2 O 2 from H 2 and O 2 was researched, but only at the laboratory scale. Theoretically, it is clear that the direct synthesis approach is the simplest way to synthesise H 2 O 2 ; however, issues with respect to practicality limit the industrialization of this process [36,37]. This review will address the challenges and developments in the field of direct synthesis of H 2 O 2 , focussing on chemical catalytic methods.

The Direct Synthesis Approach to H 2 O 2 Production Using Chemical Catalysis
In 1914, H. Henkel and W. Weber reported the very first process that was capable of producing H 2 O 2 from hydrogen (H 2 ) and oxygen (O 2 ). The patent described the reaction of two gaseous mixtures: an oxygen species (free and bound) and hydrogen in a pressurised vessel along with water. An important aspect of the patent was the use of noble metals capable of fixing hydrogen as catalysts; e.g., palladium (Pd), platinum (Pt), nickel (Ni), etc. [18].

Pros and Cons of the Direct Synthesis Approach to H 2 O 2 Synthesis
The direct synthesis approach has not been industrially practiced due to several technological and scientific barriers [38] and Table 2 lists the advantages and disadvantages of this approach. Table 2. Analysis of the advantages/disadvantages of the direct synthesis approach [27,[39][40][41][42].

Advantages Disadvantages
Absence of organic substrates such as anthraquinones or organic solvents

Pros and Cons of the Direct Synthesis Approach to H2O2 Synthesis
The direct synthesis approach has not been industrially practiced due to several technological and scientific barriers [38] and Table 2 lists the advantages and disadvantages of this approach. Table 2. Analysis of the advantages/disadvantages of the direct synthesis approach [27,[39][40][41][42].

Advantages
Disadvantages Absence of organic substrates such as anthraquinones or organic solvents Unselective reactions leading to simultaneous side products other than H2O2, namely water (H2O) Usage of green solvents like water, methanol, or ethanol Complex process with mass transfer limitations involving three phases: gas (H2/O2), liquid (reaction medium), and solid (catalyst) Economical because of fewer downstream operations to produce H2O2 Safety: explosive nature of the H2 and O2 mixture over a wide range of concentrations (4 mol %-94 mol %) The whole process can be accomplished with a single reactor system Presence of chloride and/or bromide ions in the reaction medium Scheme 3 depicts the direct synthesis of H2O2. As seen, the reaction produces either water H2O or H2O2 depending on the reaction conditions. Scheme 3. The direct synthesis approach consisting of two parallel reactions (reduction and oxidation) during the production of H2O2. Scheme taken from Gervasini et al. [43], reprinted here with the kind permission of the American Chemical Society, Copyright 2017 , Khan et al. [44], and Seo et al. [45].

Mechanism of the Direct Synthesis of H2O2
The mechanism of hydrogen peroxide synthesis using H2 and O2 is shown in Scheme 4. A possible mechanism of such a synthesis was proposed by Bianchi et al. in 1999 [46], based on the 1980 report published by Zudin et al. [47]. While Zudin et al. used palladium triphenylphosphane in a biphasic system, Bianchi et al. found out that 2,9-dimethyl-4,7-diphenyl-1,10-phenantroline ligand was the best among other ligands tested. Based on these findings, Werner published the proposed mechanism of H2O2 synthesis by the reduction of dioxygen [48]. Stahl et al. used a bathocuproine palladium complex in order to catalyse the direct synthesis of H2O2 [49]. All three processes utilised acid halides such as hydrochloric acid or hydrogen bromide to facilitate efficient catalysis [46,47,49]. However, the actual mechanism was reported in 2001 by Stahl et al. [49] Scheme 4(i), which was confirmed by Chinta and Lunsford in 2004 using tetrachloropalladate (PdCl4 2− ) prepared prior to use (Scheme 4(ii)) by replacing nitrogen ligands with chloride [50]. Scheme 3. The direct synthesis approach consisting of two parallel reactions (reduction and oxidation) during the production of H 2 O 2 . Scheme taken from Gervasini et al. [43], reprinted here with the kind permission of the American Chemical Society, Copyright 2017, Khan et al. [44], and Seo et al. [45].

Mechanism of the Direct Synthesis of H 2 O 2
The mechanism of hydrogen peroxide synthesis using H 2 and O 2 is shown in Scheme 4. A possible mechanism of such a synthesis was proposed by Bianchi et al. in 1999 [46], based on the 1980 report published by Zudin et al. [47]. While Zudin et al. used palladium triphenylphosphane in a biphasic system, Bianchi et al. found out that 2,9-dimethyl-4,7-diphenyl-1,10-phenantroline ligand was the best among other ligands tested. Based on these findings, Werner published the proposed mechanism of H 2 O 2 synthesis by the reduction of dioxygen [48]. Stahl et al. used a bathocuproine palladium complex in order to catalyse the direct synthesis of H 2 O 2 [49]. All three processes utilised acid halides such as hydrochloric acid or hydrogen bromide to facilitate efficient catalysis [46,47,49]. However, the actual mechanism was reported in 2001 by Stahl et al. [49] Scheme 4(i), which was confirmed by Chinta and Lunsford in 2004 using tetrachloropalladate (PdCl 4 2− ) prepared prior to use (Scheme 4(ii)) by replacing nitrogen ligands with chloride [50]. The mechanism follows these steps: initially, Pd 0 reduces molecular oxygen in the presence of the nitrogen ligands (Scheme 4(i))/halide ions (Scheme 4(ii)) to form the respective Pd II complex that contains a peroxo-species. This is replaced by halides yielding H 2 O 2 . Subsequently, Pd II is reduced by molecular hydrogen to yield Pd 0 , thereby completing the catalytic cycle. The mechanism follows these steps: initially, Pd 0 reduces molecular oxygen in the presence of the nitrogen ligands (Scheme 4(i))/halide ions (Scheme 4(ii)) to form the respective Pd II complex that contains a peroxo-species. This is replaced by halides yielding H2O2. Subsequently, Pd II is reduced by molecular hydrogen to yield Pd 0 , thereby completing the catalytic cycle.

Series of Elementary Steps in H2O2 Synthesis
Several accounts of the possible elementary steps during the direct synthesis of H2O2 have been described in literature by Wilson and Flaherty [40], Plauck et al. [51] and Yi et al. [52], to name a few. Scheme 5 depicts the series of steps that lead to the synthesis of H2O2 and H2O during the direct synthesis approach using Pd catalysts in liquid solvents. In each of the steps in the process, bimolecular reactions exist between the H + and chemically adsorbed intermediates on the catalyst surface. The elementary steps are based on the following assumptions: • Free energies of H2 and O2 adsorption are negligible under saturation conditions • The adsorption and desorption of the H2O2 species is unrestricted Based on these assumptions, the first step is that the hydrogen adsorbs dissociatively on to the catalyst surface (Step (1) in Scheme 5) yielding H* and is subsequently oxidized (Step (2) in Scheme 5). In Step (3) of Scheme 5, the molecular adsorption of O2 takes place. This initiates Step (4) in Scheme 5, wherein O2** undergoes proton-electron transfer under quasi-equilibration conditions to form OOH** (hydroperoxy radical). Alternatively, O2** cleaves the O-O bond irreversibly to form O*, also known as chemi-absorbed oxygen atoms (Step (7) Scheme 5). The OOH** is then adsorbed, which then reacts further to form either H2O2** by proton-electron transfer, as shown in Step (5) of Scheme Scheme 4. (i) Mechanism of H 2 O 2 formation in water with CO and O 2 in the presence of palladium catalyst complexed with nitrogen ligands (scheme from Werner et al. [48], and reprinted here with the kind permission of John Wiley and Sons, copyright 1999). (ii) The catalytic cycle of H 2 O 2 manufacture using palladium catalysts complexed with chloride ions (Chinta and Lunsford [50], and reprinted here with the kind permission of Elsevier, copyright 2004).

Series of Elementary Steps in H 2 O 2 Synthesis
Several accounts of the possible elementary steps during the direct synthesis of H 2 O 2 have been described in literature by Wilson and Flaherty [40], Plauck et al. [51] and Yi et al. [52], to name a few. Scheme 5 depicts the series of steps that lead to the synthesis of H 2 O 2 and H 2 O during the direct synthesis approach using Pd catalysts in liquid solvents. In each of the steps in the process, bimolecular reactions exist between the H + and chemically adsorbed intermediates on the catalyst surface. The elementary steps are based on the following assumptions:

Process Conditions for the Direct Synthesis of H2O2 from H2 and O2
The direct synthesis of H2O2 requires certain operating conditions such as the ratio of H2 and O2, reaction medium for the synthesis, the reactor used, additives and/or promoters, a catalyst, and its supporting material [53][54][55]. The following sections explain the influence of each of these parameters on the conversion and yield of the direct synthesis of H2O2.  Based on these assumptions, the first step is that the hydrogen adsorbs dissociatively on to the catalyst surface (Step (1) in Scheme 5) yielding H* and is subsequently oxidized (Step (2) in Scheme 5). In Step (3) of Scheme 5, the molecular adsorption of O 2 takes place. This initiates Step (4) in Scheme 5, wherein O 2 ** undergoes proton-electron transfer under quasi-equilibration conditions to form OOH** (hydroperoxy radical). Alternatively, O 2 ** cleaves the O-O bond irreversibly to form O*, also known as chemi-absorbed oxygen atoms (Step (7) Scheme 5). The OOH** is then adsorbed, which then reacts further to form either H 2 O 2 ** by proton-electron transfer, as shown in Step (5)  The direct synthesis of H 2 O 2 requires certain operating conditions such as the ratio of H 2 and O 2 , reaction medium for the synthesis, the reactor used, additives and/or promoters, a catalyst, and its supporting material [53][54][55]. The following sections explain the influence of each of these parameters on the conversion and yield of the direct synthesis of H 2 O 2 .

Ratio of the Gaseous Mixture
During direct synthesis, one would expect the reaction of H 2 and O 2 on a catalytic surface to form only H 2 O 2 as the product. However, unwanted side reactions also occur that reduce the productivity of this otherwise green process. The first side reaction is the oxidation of H 2 to H 2 O instead of H 2 O 2 and the second one is the reduction of the H 2 O 2 produced to H 2 O (Scheme 3). One can influence the selectivity of the process by optimising the parameters mentioned previously, but it is self-explanatory that the amount of H 2 and O 2 in the reaction mixture directly influences the H 2 O 2 output. Three combinations of H 2 /O 2 are possible for this reaction excess H 2 , excess O 2 , and stoichiometric amounts. Using excess H 2 would favour the reduction of H 2 O 2 , while using stoichiometric amounts would increase H 2 O 2 concentration during synthesis. However, an excess of oxygen, up to three times compared to hydrogen, would also increase the selectivity and yield of the direct synthesis approach [56]. It is worth mentioning that the flammable and explosive nature of these two gases over a wide concentration range at 25 • C and 0.1 (MPa) (1 atmospheric pressure) is a point of great concern. The flammability limit for H 2 in O 2 is 4% (lower flammability limit) to 94% (upper flammability limit), while the detonation limit is at 15% to 95%, with an increased risk of explosion with increasing pressure [42,54,57]. DuPont faced frequent explosions in their pilot plant by feeding 10% H 2 in O 2 to their process, which led to the discontinuation of the pilot plant studies [27,54].
In order to minimise the explosion risk, it is suitable to perform the reaction at lowered feed rates of H 2 and O 2 , diluted with inert gases [54]. The most common diluents are helium (He), argon (Ar), nitrogen (N 2 ), or carbon dioxide (CO 2 ) [26]. Most of the recent literature indicates the use of either CO 2 [44,58,59] or N 2 [39,60,61]. The work of Wilson and Flaherty described the use of N 2 and CO 2 as the diluents during synthesis using palladium catalysts (Pd) supported on silica (Si). By doing so, the researchers reported an overall selectivity of 31% towards H 2 O 2 using only CO 2 as the diluent. However, the combined effect on the overall selectivity was not reported [40]. Using CO 2 as a diluent is advantageous as it can expand different solvents during the reaction and increase H 2 solubility. Secondly, CO 2 dissolves in water to form carbonic acid (HCO 3 − ), which makes the medium acidic. The acidic condition is helpful as it is the most commonly used storage condition for H 2 O 2 [5]. The chemists and engineers in the field have agreed that increasing the solubility of H 2 and O 2 in the reaction medium would also lead to a better adsorption of the gases on to the catalytic surface. This, in turn, would lead to a better yield of H 2 O 2 [62]. Selinsek et al. recently reported a process design with two separate tanks containing H 2 and O 2 dissolved in water with a two-fold benefit. First, an explosion is circumvented due to the separate feeding of gases. Second, the H 2 to O 2 ratio in the reaction cell can be easily controlled by varying the flow rate of the pump, ensuring that at any given point in time, the process operates in a safe manner [4]. The recent work of Urban et al. describes the use of an electrochemical sensor system to detect the hydrogen and oxygen amounts present in the system during direct synthesis. Additionally, the sensor is capable of monitoring the H 2 O 2 levels in the reactor as well. This sensor is capable of performing under high analyte concentrations and high pressures. The authors also claim that the usage of this novel electrochemical sensor could minimise the risk of explosion due to high accuracy of detection [63].

Reaction Medium
As mentioned previously, the direct synthesis of H 2 O 2 from H 2 and O 2 in the gaseous state without any reaction medium is highly dangerous. This is because the gases form an explosive mixture over a wide range of concentrations [64]. Hence, performing the synthesis of H 2 O 2 at lowered temperatures in highly pressurised environments, in an appropriate reaction medium, prevents explosions and produces high yields [27]. Therefore, the choice of the reaction medium is crucial to the success of the process. Most of the existing literature uses water as the reaction solvent, with some exceptions where pure methanol or ethanol is used. Additionally, using water with co-solvents such as methanol or ethanol favours a higher dissolution of H 2 when compared to pure water as a reaction medium [56,65]. In 2001, Hâncu and Beckmann reported the use of CO 2 as a reaction medium for the direct synthesis of H 2 O 2 using a CO 2 soluble ligand-supported Pd catalyst. The researchers worked on the assumption that the H 2 O 2 solubility in CO 2 is considerably less than the conventional working solutions of the AO process, i.e. organic solvents. Furthermore, the CO 2 used was liquid under the reaction conditions (298 K, 17 MPa) and the presence of a CO 2 -phillic catalyst would minimize the contact time of H 2 O 2 on Pd, thereby increasing the selectivity of the process (Figure 1) [28,66]. reaction medium would also lead to a better adsorption of the gases on to the catalytic surface. This, in turn, would lead to a better yield of H2O2 [62]. Selinsek et al. recently reported a process design with two separate tanks containing H2 and O2 dissolved in water with a two-fold benefit. First, an explosion is circumvented due to the separate feeding of gases. Second, the H2 to O2 ratio in the reaction cell can be easily controlled by varying the flow rate of the pump, ensuring that at any given point in time, the process operates in a safe manner [4]. The recent work of Urban et al. describes the use of an electrochemical sensor system to detect the hydrogen and oxygen amounts present in the system during direct synthesis. Additionally, the sensor is capable of monitoring the H2O2 levels in the reactor as well. This sensor is capable of performing under high analyte concentrations and high pressures. The authors also claim that the usage of this novel electrochemical sensor could minimise the risk of explosion due to high accuracy of detection [63].

Reaction Medium
As mentioned previously, the direct synthesis of H2O2 from H2 and O2 in the gaseous state without any reaction medium is highly dangerous. This is because the gases form an explosive mixture over a wide range of concentrations [64]. Hence, performing the synthesis of H2O2 at lowered temperatures in highly pressurised environments, in an appropriate reaction medium, prevents explosions and produces high yields [27]. Therefore, the choice of the reaction medium is crucial to the success of the process. Most of the existing literature uses water as the reaction solvent, with some exceptions where pure methanol or ethanol is used. Additionally, using water with co-solvents such as methanol or ethanol favours a higher dissolution of H2 when compared to pure water as a reaction medium [56,65]. In 2001, Hâncu and Beckmann reported the use of CO2 as a reaction medium for the direct synthesis of H2O2 using a CO2 soluble ligand-supported Pd catalyst. The researchers worked on the assumption that the H2O2 solubility in CO2 is considerably less than the conventional working solutions of the AO process, i.e. organic solvents. Furthermore, the CO2 used was liquid under the reaction conditions (298 K, 17 MPa) and the presence of a CO2-phillic catalyst would minimize the contact time of H2O2 on Pd, thereby increasing the selectivity of the process (Figure 1) [28,66]. Moreno et al. reported the use of a supercritical CO 2 (scCO 2 ) with methanol (MeOH) water mixture at an operating temperature range of 283 K-318 K and pressure of 16.7 MPa to synthesize H 2 O 2 from H 2 and O 2 . N 2 was used as a diluent in this reaction to achieve a yield between 11.6% and 45.9% [42]. Except for these works, almost every direct synthesis of H 2 O 2 using chemical catalysis is documented in water, alcohol, or a defined ratio of both. Landon et al. reported that at 31.1 • C, which is the critical temperature of CO 2 , more decomposition of H 2 O 2 was observed than formation. The researchers suggested that the synthesis was performed at a temperature just below the critical temperature [67]. This could be one of the many reasons as to why the use of CO 2 as a solvent has not been widely researched. Abate et al., on the other hand, used scCO 2 -expanded methanol (a solution operating below the triple point in the presence of a solvent to form a two-phase fluid system) as the solvent in the presence of a Pd catalyst supported on mesoporous silica for the direct synthesis of H 2 O 2 . On doing so, a selectivity of 40% towards H 2 O 2 and a productivity of 0.11 mol H2O2 m −2 ·Pd·h −1 at the end of 3 h [65].

Additives/Promoters
Along with the reaction medium, special additives termed "promoters" are often used to stabilize H 2 O 2 production and to increase the process yield. The most commonly used promoters are acids or halides. Edwards et al. published the effect of acids and halides on the outcome of the direct synthesis of H 2 O 2 . In this report, the researchers classified the promoters into two groups: (i) oxyacids such as acetic acid, perchloric acid, phosphoric acid (H 3 PO 4 ), nitric acid, and sulphuric acid (H 2 SO 4 ) (ii) halide acids such as hydrochloric acid (HCl), hydrobromic acid (HBr), and hydroiodic acid (HI).
The authors reported that the reactions could accept potassium and sodium salts of halide acids as well. The presence of halide ions in the reaction could influence the H 2 O 2 conversion in the following order: KF > no halide ions > KCl > KBr and selectivity in the following order: KBr > KCl > no halide ions > KF. The report also claimed that the presence of iodide ions eventually poisoned the catalyst [55].
The next pioneering work in the field was that of Wilson and Flaherty, pointing out that the presence of protons (H + ) from mineral acids such as HCl, H 2 SO 4 , H 3 PO 4 and HNO 3 aids in reducing molecular oxygen. They also suggested that the corresponding counter ions such as Cl − , SO 4 2− , PO 4 2− , and NO 3 − modify the structure of the metal catalyst. By adding halide groups to the Pd catalyst and performing the synthesis in ethanol at an acidic pH, the selectivity would eventually increase from 60% to 80% [40]. The recent report of Gallina    The researchers formulated that the absence of promoters (H 3 PO 4 or NaBr) led to either (i) a much more prominent water forming reaction than the H 2 O 2 forming one or (ii) subsequent reaction of H 2 O 2 further to form H 2 O. The researchers also concluded that NaBr/H 3 PO 4 had a combined effect on the leaching of the Pd catalyst and that there was no fixed NaBr to H 3 PO 4 ratio to improve the direct synthesis of H 2 O 2 [68]. The authors also suggested that a pH of 2.0 in the reaction medium would favour a better selectivity towards H 2 O 2 production from H 2 and O 2 [68].
Another method of acidifying the reaction medium to suit the direct synthesis of H 2 O 2 is to use solid acid catalysts (SAC) as reported by the publication of Lewis et al. [57]. The problem of the acidic additives and halide salts being soluble in aqueous medium making the recovery of these a tedious downstream operation was addressed in this work. By using a SAC such as caesium substituted phosphotungstic acid (HPA) in the presence of a Pd or Au-Pd alloyed catalyst, the productivity of H 2 O 2 could be increased and the degradation to H 2 O could be decreased [69].

Reactor Design
One of the most important parameters to be discussed for the direct synthesis of H 2 O 2 is the type of reactor used for the reaction. Until today, slurry reactors, plugged flow reactors, microreactors or trickle bed reactors are often used to synthesise H 2 O 2 directly from H 2 and O 2 . One of the major requirements when choosing a reactor is that the vessel should withstand high pressure. It is known that the prolonged exposure of H 2 O 2 to H 2 in the presence of a Pd catalyst would lead to the formation of water (Scheme 3). One possible way to overcome this phenomenon is to limit the exposure of H 2 O 2 to H 2 on the catalytic surface.
Microreactors present a unique way to operate this process in a continuous manner, with defined flow characteristics, large surface area to volume ratio, promising heat and mass transfer rates, and excellent process safety. A microreactor is "a device that contains micro structured features with a sub millimetre dimension, in which chemical reactions are performed in a continuous manner." The microreactors are constructed from silicon, quartz, glass, metals, polymers, and ceramics, to name a few [70]. The work of Shang and Hessel describes the operational-and reaction-based benefits of using microreactors for the direct synthesis of H 2 O 2 in their work [71]. One drawback of using a microreactor for the direct synthesis of H 2 O 2 is the incorporation of the metal catalyst within the capillaries of such a reactor [72]. Kanungo et al. described a technique to incorporate an Au-Pd alloy on to the walls of a silica coated capillary microreactor. The innovation in the design lies in the fact that the catalytic particles were formed in situ on the walls of the microreactor by a layer-by-layer self-assembly creating a multi-layer catalyst. By using this approach, the researchers were able to produce 210 mol H2O2 ·kg cat  [75]. The reviews of Kolehmainen et al. [56] and Dittmeyer et al. [53] summarize the innovations in the field of catalyst design and reactor engineering in a detailed manner.
Another technique to overcome H 2 O 2 decomposition by reactor engineering is the use of a trickle bed reactor or a plugged flow reactor. Almost all works of Biasi et al. uses trickle bed reactors (Figure 3) to improve the selectivity towards H 2 O 2 [68,[76][77][78][79][80], with the maximum being 80% using Pd on a sulfated ceria (CeS) catalyst. Another technique to overcome H2O2 decomposition by reactor engineering is the use of a trickle bed reactor or a plugged flow reactor. Almost all works of Biasi et al. uses trickle bed reactors ( Figure  3) to improve the selectivity towards H2O2 [68,[76][77][78][79][80], with the maximum being 80% using Pd on a sulfated ceria (CeS) catalyst.

Influence of the Catalytic Material
A major hindrance in the direct synthesis approach is the process' low selectivity, as the formation of H2O2 is not thermodynamically favoured. Although promoters such as H2SO4, H3PO4, NaBr, and KBr help enhance the selectivity of the process, one might consider the catalyst and its supporting material to be the most influencing parameter [81]. Furthermore, the presence of noble metals or noble metal alloys as catalysts aid in the hydrogenation and subsequent decomposition of  Another technique to overcome H2O2 decomposition by reactor engineering is the use of a trickle bed reactor or a plugged flow reactor. Almost all works of Biasi et al. uses trickle bed reactors ( Figure  3) to improve the selectivity towards H2O2 [68,[76][77][78][79][80], with the maximum being 80% using Pd on a sulfated ceria (CeS) catalyst.

Influence of the Catalytic Material
A major hindrance in the direct synthesis approach is the process' low selectivity, as the formation of H2O2 is not thermodynamically favoured. Although promoters such as H2SO4, H3PO4, NaBr, and KBr help enhance the selectivity of the process, one might consider the catalyst and its supporting material to be the most influencing parameter [81]. Furthermore, the presence of noble metals or noble metal alloys as catalysts aid in the hydrogenation and subsequent decomposition of

Influence of the Catalytic Material
A major hindrance in the direct synthesis approach is the process' low selectivity, as the formation of H 2 O 2 is not thermodynamically favoured. Although promoters such as H 2 SO 4 , H 3 PO 4 , NaBr, and KBr help enhance the selectivity of the process, one might consider the catalyst and its supporting material to be the most influencing parameter [81]. Furthermore, the presence of noble metals or noble metal alloys as catalysts aid in the hydrogenation and subsequent decomposition of H 2 O 2 to water (Scheme 3) [4]. On surveying the recent literature in the direct synthesis of H 2 O 2 , it is certain that the majority of scientists in the field are dedicated towards developing new, robust and stable catalysts. Existent catalysts and the newly developed versions of noble metal catalysts are characterized based on two criteria: conversion and selectivity. Pd catalysts are almost exclusively used for such reactions either as obtained or alloyed with other metals and/or supports to enhance the selectivity of the process.
Edwards et al. published the importance of alloying Pd with other metals to increase the efficiency of H 2 O 2 production. Their observation stemmed from the highly reactive nature (25 times more active) of an Au-Pd catalyst that was capable of oxidising alcohols better than the corresponding monometallic catalysts. They also reported that when Au-Pd alloys were attached to a titanium dioxide (TiO 2 ) or alumina (Al 2 O 3 ) support, core-shell structures were formed. With the development of the new catalyst, the researchers still faced the problem of H 2 O 2 being reduced to water [82]. The same working group developed a new tin-based Pd alloy to stop the hydrogenation of H 2 O 2 . With the new alloyed catalyst accompanied by a heat treatment cycle, the hydrogenation reactions were prevented and selectivities of more than 95% were reported [37]. Ntainjua et al. used ruthenium (Ru) alloyed with Au and Pd to perform the direct synthesis of H 2 O 2 . The authors investigated the Ru-Au, Ru-Pd and Ru-Au-Pd catalyst for H 2 O 2 synthesis. The amount of Ru added to the alloy, along with the calcination conditions, had an effect on the catalyst activity and reusability. So far, this report is the only one using an Ru catalyst for the direct synthesis of H 2 O 2 [83]. Besides Au, only silver [44,84], tellurium [85], tin [37], and zinc [86] are described as possible metals for alloying with Pd. Xu et al. reported the possible increase of H 2 O 2 production by using different metals such as tungsten (W), lead (Pd), molybdenum, etc. and validated their results with density functional theory (DFT). The researchers suggested that all these metals were superior to platinum (Pt) as a promoter. The research was a computational model and experimental evidence to substantiate this fact is needed [87]. Tian et al. reported that by increasing the amount of Pd in the system, the H 2 conversion increases as well, which is self-explanatory. However, the selectivity and productivity increases with decreasing Pd content. From these results, they were able to conclude that having a Pd particle size in the range of 2.5 nm to 1.4 nm would yield a selectivity of approximately 94% with 0.5% Pd loading [88]. Most recently, Howe et al. used microwaves to prepare an Au-Pd alloy supported on TiO 2 for synthesizing H 2 O 2 from H 2 and O 2 . The authors claim that the catalysts were capable of maintaining their activity for four reaction cycles. Compared to other alloyed Pd particles, these particles have a core-shell structure and can be prepared in 0.25 h [89]. The recent 2017 patent of Desmedt et al. used metallic catalysts supported on sulphate and phosphate to reduce the amount of inorganic acid content in the reaction medium. The inventors varied the metal content between 0.001 mass % and 10 mass % (0.62 to 2 wt % Pd) to obtain conversions ranging between 26.9% and 46% and selectivities between 19.9% and 74% [90]. Table 4 below summarises the reaction conditions and the catalysts for the direct synthesis of H 2 O 2 along with the catalyst used.

Summary, Conclusions and Future Perspectives
More than a century has passed since the first documented work of Henkel and Weber in 1914 producing H 2 O 2 directly from H 2 and O 2 using Pd catalysts, and the direct synthesis has only reached to pilot plant scale of production. This is because the synthesis needs to be operated beyond the explosive range of H 2 and O 2 ratios and the thermodynamic favoring of water formation over hydrogen peroxide formation. The conversion and selectivity of the process depends on several parameters such as the H 2 /O 2 ratio, the diluent used in the process, the reaction medium used to prevent explosion, the catalyst and its supporting material, reactor design, and the operating temperature and pressure, to name a few. From recent literature, it is understood that it is an advantage to perform the synthesis at reduced temperatures (263 K to 283 K) and pressures (most commonly 2.0 MPa to 4.0 MPa). To synthesise H 2 O 2 selectively from H 2 and O 2 , oxygen is to be used in excess to avoid hydrogenation of H 2 O 2 . With the new developments in microreactor technology, great advancements are being achieved in increasing the selectivity of the process. Moreover, robust and stable catalysts have been the research focus of the scientific community ever since this reaction was reported. The use of Pd in its pure form or as an alloy presents an opportunity towards industrializing this process. Additionally, additives such as NaBr, H 3 PO 4 , H 2 SO 4 etc. aid in stabilising the synthesised H 2 O 2 . Finally, with the scientific community focusing on green and sustainable processes, ably supported by the advancements in the field of direct synthesis, the first commercial plant producing H 2 O 2 using direct synthesis technology is not that far in the future.