Elf Atochem finds H2O2 route

22 November 1999 00:00  [Source: ICB]

Direct combination of hydrogen and oxygen to produce hydrogen peroxide has been sought for some time. Elf Atochem looks to have found the answer, reports Art Brownstein*

A means to produce hydrogen peroxide efficiently and safely by direct combination of hydrogen and oxygen has been long sought. In the latest effort to achieve this, Elf Atochem discloses a technique in which hydrogen peroxide is so obtained and in 88% selectivity (World Patent Appl 99/41190).

Elf Atochem's approach is to combine the hydrogen and oxygen as very small bubbles (less than 3mm in diameter) in an aqueous phase in which the molar ratio of hydrogen to oxygen falls within the explosive range, but reactants in the gas phase lie outside this range (H2:O2 < 0.0416 molar). Execution of the reaction within the explosive range is regarded as essential for productivity at an economical level (see diagram right).

Specifically, hydrogen and oxygen (1:1 v/v) are passed into an acidic aqueous slurry of phosphoric and sulphuric acids containing finely divided particles of platinum and palladium (0.8 and 0.4wt%, resp.) supported on silica. At 20ûC and over a five-hour period, a 13% hydrogen peroxide solution is obtained in 88% selectivity based on hydrogen.

At lower liquid phase hydrogen to oxygen ratios (0.4 v/v) and otherwise similar conditions, hydrogen peroxide concentrations as high as 20% are obtained in three hours at slightly lower selectivities of 83-85%. The patent does not disclose reaction pressures and conversions. The latter could be as high as 31% at 1 atm, or substantially less at higher pressures, which quite likely are required for efficiency.

Elf Atochem's results are similar to those reported by BASF (ECN 14 September 1998) which obtained 82% selectivity to hydrogen peroxide at 76% hydrogen conversion. BASF employed a catalyst in the shape of a cylindrical monolith coated with palladium nitrate. Although its hydrogen peroxide concentration was much lower (7%) than the 13-20% obtained by Elf Atochem, conversion of hydrogen appears to be much greater.

In a markedly different approach, Eka Chemicals describes a vapour phase technique to produce hydrogen peroxide directly from hydrogen and oxygen (World Patent Appls 99/32398 and 99/32399). Eka employs a fluidised bed catalyst comprised of a thin layer (50nm) of palladium on polytetrafluoroethylene. The palladium, in turn, is coated with an even thinner layer (10nm) of siloxane.

A gaseous feed of H2/O2/N2 (3/5/47 v/v) under 15 bar pressure is first passed through a 1% sulphuric acid solution containing NaBr (20 ppm) and then over the catalyst at 45ûC. Productivity of hydrogen peroxide appears to be quite low. Specific data on selectivity and conversion are not disclosed.

The appeal of a route to hydrogen peroxide based on direct combination of hydrogen and oxygen is the elimination of ethyl anthraquinone (EAQ) or its homologues as intermediates. These are used in virtually all commercial processes which rely upon their successive hydrogenation and oxidation to produce hydrogen peroxide. Alkyl anthraquinones typically cost more than $10/kg and their losses can account for about 10% of the total variable production cost of hydrogen peroxide. In addition, their elimination could offer a sizable savings in the investment cost for a new plant.

###8330###

Global demand for hydrogen peroxide is about 2m tonne/year with laundry and bleaching chemicals and pulp and paper accounting for nearly 80% of its uses. A potentially low-cost process such as direct combination of hydrogen and oxygen could create new markets for hydrogen peroxide, such as the production of propylene oxide. Although there has been substantial research for its use in such new applications, commercialisation has been hindered by the relatively high cost of hydrogen peroxide.

Investigators at Inha University in South Korea find that the nitration of aromatics proceeds in very high yields by using nitrogen dioxide and oxygen in the presence of inorganic oxide catalysts, such as silica (World Patent Appl 99/42433). Nitrobenzene, for example, is obtained in nearly 100% yield by reacting benzene with a 2 molar excess of liquid NO2 at 45ûC under 40 psi oxygen pressure in the presence of silica gel (1.3% based on benzene).

###8331###

Under similar conditions, dinitrotoluene is obtained in 99% yield. The distribution of 2,4- and 2,6-isomers (83:17 respectively) is similar to that obtained when using conventional technology. The key to the development is a catalyst which has a surface area of at least 100m2/gramme and a pore size of at least 0.5nm. The process offers an alternative to the use of mixed concentrated sulphuric and nitric acids which are used in nearly all commercial systems. By so doing, the Inha University development supplants nitric acid with cheaper nitrogen dioxide and eliminates the need to recover and recycle spent sulphuric acid.

The stoichiometry for the process suggests that 0.5 mol of oxygen are required per mol of benzene and that 0.5 mol of water are produced per mol of nitrobenzene. Consequently, nitric acid and nitric oxide are likely by-products in accordance with the following reaction:

###8332###

The nitric oxide is probably converted to the dioxide under the reaction conditions and can be either recycled or sold off as a credit in the form of nitric acid. Unless air can be used efficiently, the process must also bear the cost of oxygen at about $40/tonne.

Nitrobenzene and dinitrotoluene are intermediates in the production of aniline and tolylene diamine, which in turn are used to produce isocyanates. Global demand for tolylene diisocyanate (TDI) and polymeric (MDI) isocyanates is about 1.7m and 2.3m tonne/year respectively.

Bromosulphonated polyphenylene oxides are effective as membranes for removal of carbon dioxide from natural gas according to UK-based BG (World Patent Appl 99/42204). The permeability of these membranes to pure carbon dioxide relative to methane is 59:1. When applied to a natural gas stream containing 20.3% CO2 and under 100 psig pressure, essentially pure natural gas is recovered while the amount of CO2 in the permeate is 83.3%.

The membranes are derived from commercially available polyphenylene oxide which is first 20% brominated and then treated with a stoichiometric amount of chlorosulphonic acid. The bromosulphonated polymer so produced is employed as a hollow fibre in which a polyetherimide is the support. Natural gas in many parts of the world has an inordinately high carbon dioxide content which must be lowered in order to raise the caloric value of the gas to a commercially acceptable level. Membranes are attractive for such applications on offshore platforms because of their compactness and lighter weight compared to the use of cryogenic systems.

Although nearly 100% of world capacity for chlorine is based on the electrolysis of brine, there is continuing interest in its production from hydrogen chloride. This is because many processes which use chlorine produce hydrogen chloride as a by-product. This is true in the case of vinyl chloride, isocyanates or polycarbonate production. Hydrogen chloride must be either recycled, if possible, or disposed of as a low value by-product or as a waste product.

Now Sumitomo Chemical reports that ruthenium oxide (1.9%) supported on titanium dioxide is an effective catalyst for the oxidation of hydrogen chloride to chlorine in very high yield (European Patent Appl 936,184).

###8333###

Passing a stream of HCl and O2 (96:53 v/v) over the Ru2O/TiO2 catalyst at 361ûC and 1 atm pressure produces chlorine in 93% conversion at ostensibly 100% selectivity.

The high yield is attractive since this reaction is known to be equilibrium limited, more so at progressively higher temperatures. Productivity appears to be relatively low, however, with a maximum rate of chlorine production of 3.8kg/kg cat/hr. This suggest a large reactor size and correspondingly greater cost for fixed investment.

Vinyl chloride is the largest single use for chlorine and accounts for about 40% of consumption. An HCl oxidation process, such as Sumitomo's, must be cost competitive with the oxychlorination section of vinyl chloride plants to find use for this particular application.

In related work, Bayer discloses a gas phase electrolytic process for processing HCl into high purity chlorine (World Patent Appl 99/31297). Bayer's system produces chlorine in 99.9+% purity by linking gas phase and liquid phase electrolytic processes. Residual HCl from the gas phase electrolytic cell (84% conversion; 10kA/m2 current density) is scrubbed from the chlorine product with concentrated HCl which is then sent to a liquid phase electrolytic cell as shown in the figure below.

Bayer has been operating an 80 000 tonne/year plant at Baytown, Texas, US, which processes by-product HCl from nearby isocyanates production.

Lonza has found that butyrolactone can be obtained in very high yield by vapour phase hydrogenation of maleic anhydride (World Patent Appl 99/35139).

###8334###

Passage of hydrogen and maleic anhydride over a Cu-Zn catalyst (Sud Chemie T-4322) at 230ûC and 5 atm pressure produces butyrolactone in selectivities as high as 97.7% at 100% conversion. Ratios of hydrogen to maleic anhydride in the feed range from 109 to 165 (molar).

###8335###

Yields in pilot plant runs are essentially constant over 765 hours. The fixed bed catalyst is activated by treatment with hydrogen and nitrogen in a programmed manner such that the hydrogen content is gradually increased from 0% to 100%. Similar yields of butyrolactone are obtained with Cu/Cr catalysts (Sud Chemie T-4466).

###8336###

In a related effort, Eurodiol describes the multistep production of tetrahydrofuran (THF) and 1,4-butanediol by hydrogenation of maleate esters (World Patent Appls 99/35113 and 99/35114). Butyrolactone is an intermediate in its system.

Although butyrolactone can be readily converted to THF and butanediol, it has a niche market of its own as an environmentally friendly solvent or as intermediate for one, such as N-methylpyrrolidone. Production of butyro-lactone, THF and butanediol has undergone a rapid change in the last few years with new capacity largely based on maleic anhydride to the exclusion of long standing Reppe tech-nology which uses acetylene and formaldehyde.

Novo Nordisk has developed an enzyme system which is effective in organic media such as heptane (World Patent Appl 99/33964). Immobilisation of H lanuginosa (a lipase) on a zeolite enables the transesterification of trilaurin with myristic acid in heptane/aq NaCl (20/3 v/v) to lauric acid in 53% conversion at 40ûC in 24 hours. The co-product is mixed myristyl and lauryl triglycerides. The enzyme is added to the zeolite by atomisation and the resulting system is used as a fluidised bed. Such zeolites as Degussa's Wessalith MS330 and silica (Celite) are effective.

Although immobilised enzyme systems are well established, a disadvantage to their use is that most reactions require the use of aqueous media. This not only limits the applicability of enzymes as catalysts for reactions which proceed in water, but also generally entails the costly recovery of products from dilute solutions. The ability to use enzymes in organic solvents, especially at higher temperatures, enables a broadening of their applicability to a large number of synthetic processes.

For the latest chemical news, data and analysis that directly impacts your business sign up for a free trial to ICIS news - the breaking online news service for the global chemical industry.

Get the facts and analysis behind the headlines from our market leading weekly magazine: sign up to a free trial to ICIS Chemical Business.

Printer Friendly