DMC discoveries

26 November 2001 00:00  [Source: ICB]

EniChem has developed a new method for producing aromatic isocyanates using dimethyl carbonate as an intermediate, which cuts out the use of toxic phosgene

Amid growing industry use of dimethyl carbonate (DMC) as an intermediate for polycarbonate production, EniChem now discloses its use as a phosgene replacement in a high yield process for aromatic isocyanates (World Patent Appl. 0156977). Using tolylene diisocyanate (TDI) as an example, DMC reacts with tolylene diamine (TDA) to form the corresponding diurethane, which is subsequently pyrolysed to TDI and methanol, as shown:

The methanol is readily recovered for recycle to DMC production.

###10600###

In a specific example, an 80:20 mixture of 2,4- and 2,6-TDA is combined with excess DMC (13.2:1 molar ratio) and a catalytic amount (2.96 mol%) of zinc acetate, and held at 160-175ûC under 2.5 atm for 3.5 hours. The diurethane (TDU) is obtained in 94% selectivity at 99% TDA conversion. After catalyst removal, the crude TDU is taken up in DMC containing a small (7.5 mg/l) amount of phosphoric acid at 130ûC for two hours to remove traces of residual catalyst.

DMC is distilled away from the passivated solution and the TDU is finally passed through a tubular reactor at 456ûC under 0.09 atm pressure to produce TDI in 93.5% selectivity at 73.5% TDU conversion. A small amount of half-converted product (tolylene isocyanato methyl carbamate) is co-produced in 4.0% selectivity. Ostensibly, this can be recycled through the pyrolysis reactor to produce additional TDI in the same or greater selectivity. The key to the high yields of TDI is passivation of the TDU prior to pyrolysis by removal of any traces of catalyst.

###10601###

Although no examples are given, the EniChem development is most likely also applicable for the production of diphenylmethane diisocyanate (MDI) and its oligomers. In this instance, aniline is the likely starting material from which the urethane could be readily produced, followed by successive condensation with formaldehyde and pyrolysis.

The EniChem development is formally analogous to the phosgene-based processes in general commercial use. These processes comprise phosgenation of TDA, for example, to the corresponding carbamoyl chloride which is then thermally decomposed to TDI and hydrogen chloride (0.41 tonne/tonne of TDI). EniChem's process enables the replacement of toxic phosgene with DMC, and eliminates the necessity of disposing of HCl profitably. It also obviates the use of such solvents as o-dichlorobenzene.

###10602###

A negative feature of the DMC system, however, is an apparently higher fixed capital investment compared to phosgene-based technology, because of longer reaction times and a plant which appears to require more equipment. Higher reaction temperatures and very low operating pressures during TDU pyrolysis may also result in greater utility costs. In polycarbonate production by melt polymerisation, where DMC is finding growing use by replacing phosgene, there are savings in the elimination of solvents as well as in lower fixed capital investment.

The two principal isocyanates are TDI and MDI oligomers, whose global demand is 1.2m and 2.2m tonne/year, respectively. MDI and its oligomers are used to produce rigid polyurethane foams, such as for insulation, while TDI is employed for flexible polyurethanes. Growth rates for MDI and TDI have been averaging 11%/year and 7%/year, respectively, worldwide.

In related work, Catalytic Distillation Technologies finds that DMC can be produced in very high yield from methanol and urea when dimethyldimethoxytin is used as a catalyst in combination with triglyme (European Patent Appl. 1112120). The process is especially effective when executed in a reactive distillation reactor.

###10603###

In the process, the reboiler of a reactive distillation system is charged with methylcarbamate/methanol/triglyme/catalyst (5/4.8/3.2/ 1 w/w), and is continuously fed with a 4.6% solution of urea in methanol over a 22-hour period. The system is maintained at about 180ûC under an overhead pressure of 135-232 psig. DMC is continuously distilled from the reaction as it is formed, and is obtained in 98.2% selectivity at 98.3% conversion.

The novel catalyst and reactor configuration minimise side reactions and enable the continuous recovery of DMC and ammonia as they are formed. In addition, byproducts which are formed by thermal and catalytic decomposition of urea and methyl carbamate, a reaction intermediate, are minimised.

The primary commercial route to DMC is the catalysed coupling of carbon monoxide and methanol as developed by EniChem as shown below:

Yields of DMC by this technique are reportedly 98% on methanol and 90% from carbon monoxide. At $85/tonne, a urea-based process could be quite competitive with a carbon monoxide-based system, especially if byproduced ammonia can be effectively recovered as a credit.

###10604###

Methanol to ethylene

Methanol produced from natural gas or coal-derived synthesis gas has long been regarded as a potentially attractive route to ethylene. As an outgrowth of continuing efforts to produce ethylene from methanol, ExxonMobil now reports that selectivity improves markedly if the methanol is fed to the catalyst from multiple injection ports instead of from a single point (World Patent Appl. 0162689).

The process comprises a fluid-bed reaction system using either SAPO-34 or ZSM-34 into which methanol is fed at 375-470ûC under 1 atm pressure and a weight hourly space velocity of 0.15-0.70 hr-1. Approximately 1 tonne of water is produced per tonne of hydrocarbon products.

In a system employing a single injection port at the base of the fluidised-bed reactor, ethylene is obtained in 19.8% selectivity (weight) when using SAPO-34 as the catalyst. This is equivalent to 45.2% selectivity in hydrocarbons produced as shown in the table. Ethylene selectivity rises markedly to 24.1% (55.0% of hydrocarbon production) as the number of injection ports about the fluidised bed increase from one to three. Methanol conversion decreases only slightly from 99% to 96% with this increased number of ports.

There appears to be an increase in ethylene selectivity as the number of entry ports increases. Although the patent application only shows the results for up to three such ports, an extrapolation can be made for the anticipated results for up to ten injection ports. It appears that with ten ports, ethylene selectivity could be as high as 29% of all products including water, or 66% of all hydrocarbons.

Enhanced selectivity of methanol to ethylene also occurs in a single injector port configuration if steam is incorporated into the feedstream. Substantially, the same ethylene selectivity is obtained when a 70:30 steam:methanol feed is employed as when three injector ports are used without steam. Such steam diluted systems, however, carry an economic burden of larger reactor sizes and the costs for steam generation and condensation.

This burden would be reflected by higher fixed capital investment and higher utility costs. It would be interesting, however, to determine the effect on ethylene production by incorporating a relatively small amount of steam diluent into a multiple injection port system.

Recently, ExxonMobil also reported a significant improvement in methanol to ethylene productivity by incorporating toluene into the feedstream. In such a system, selectivity to ethylene is close to 40% at 75% conversion, and paraxylene is obtained as a coproduct (ECN 28 May 2001).

Propylene oxide

Although epoxidation of propylene with hydrogen peroxide is a potentially attractive route to propylene oxide, its commercial acceptance has been hindered by the relatively high cost of hydrogen peroxide of at least $600/tonne.

In an effort to minimise the impact of hydrogen peroxide costs due to process inefficiencies such as inadequate selectivity and incomplete hydrogen peroxide conversion, Degussa has developed a process configuration in which a tubular countercurrent flow reactor is coupled to two co-current stirred tank reactors as shown (World Patent Appl. 0157011 and 0157009).

In an example, a feedstream of 42% aqueous hydrogen peroxide and 2.0% titanium silicalite suspended in methanol (1/2.5 v/v) is charged to a stirred tank reactor while propylene is injected into the base of an associated tubular reactor. All the reactors are at 65ûC under 15 atm pressure. Propylene is maintained in excess over the hydrogen peroxide. Propylene oxide is obtained in 90.3% yield at 98.6% hydrogen peroxide conversion. Selectivity to propylene oxide among the organic products is 94.5%. Byproducts are isomeric methoxypropanols and 1,2-propanediol.

Degussa is apparently using an improved version of a titanium silicalite catalyst - TS-1 granulates that are described in its earlier European Patent Application 893158. TS-1-type titanium silicalites useful for propylene epoxidation were originally developed by EniChem more than a decade ago. The EniChem catalyst system may now be owned by Dow Chemical which recently acquired EniChem's polyurethane business. Another morphological version of titanium silicalite which is also effective for propylene oxide production is known as MFI.

Degussa is reportedly developing a hydrogen peroxide-based propylene oxide process in association with Krupp-Uhde. The attractiveness of such a process may be simplicity in plant design coupled with a largely environmentally clean system free from dependence on byproduct credits.

All current propylene oxide plants employ either the chlorohydrin process or the so-called Oxirane process which coproduces 2-plus tonne/tonne of either styrene or t-butanol of propylene oxide. The chlorohydrin process suffers from the difficulty of waste brine disposal because of contamination with organics. West European demand for propylene oxide is about 1.6m tonne, and increased 5-6% in 2000.

Fatty acid esters

Fatty acid esters derived from soya bean and other vegetable oils are of interest as diesel fuels and as base oils for lubricants. Now, Sumitomo Chemical finds that such esters can be quickly produced in very high yield by transesterification of soya bean oil with methanol under supercritical conditions (European Patent Appl. 1126011).

Heating oil and methanol (1/1.44 w/w, respectively) with a catalytic amount of sodium carbonate (0.67 wt%) in an autoclave at 300ûC for only ten minutes produces the corresponding methyl esters in 99% yield. Glycerol is also obtained in 99% yield. Substantial recycling of the methanol is necessary since it is used in 13-fold excess over theoretical.

SELECTIVITIES VIA SAPO-34 CATALYSIS (SINGLE PORT)

	Total products, Wt.%	Hydrocarbons, Wt.%
ethylene	19.8	45.2
propylene	15.7	35.8
butenes	4.1	9.4
other hydrocarbons	4.2	9.6 est.
water	56.2
total	100.0	100.0

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