Towards the Direct Oxidation of Benzene to Phenol

Marco Ricci, Daniele Bianchi, and Rossella Bortolo 13.1

Introduction

Phenol is one of the most important intermediates of the chemical industry. Its current global capacity is put at about 8 x 1061 yr_1 and is forecast to grow near 4.5% per year through to 2009.

The main consumption of phenol, accounting for nearly 40% of its global demand, occurs in the synthesis of bisphenol A (mainly used to produce polycarbonate for compact discs). This is followed by phenol use in the production of epoxy resins (used, for instance, in protective coatings, in composites for electrical applications and in adhesives), of phenolic resins (which have a broad range of end-uses, including circuit boards) and of caprolactam (to be converted into nylon 6 for fibers and engineering plastics). In addition, phenol is an intermediate in the syntheses of adipic acid, acetyl salicylic acid (aspirin), aniline, hydroquinone, catechol, 2,6-xylenol, alkylphenols, chlorinated phenols and diphenols, polyphenylene oxide engineering plastics and other specialty chemicals. Finally, it is used in plasticizers, in water treating (as a slimicide) and as a disinfectant and anesthetic in medicinal preparations and pharmaceuticals.

Historically, several processes have been developed to an industrial scale to produce phenol, including: (i) sulfonation of benzene and alkali fusion of the benzene sulfonate; (ii) chlorination of benzene and hydrolysis of chlorobenzene; (iii) the cumene process (Section 13.2); (iv) toluene oxidation to benzoic acid and subsequent oxidative decarboxylation of the latter to phenol; and (v) dehydrogenation of cyclohexanol-cyclohexanone mixtures. Today, however, only the cumene process and the toluene oxidation are still run on an industrial scale, all the other processes having been given up due to economic reasons or environmental problems.

Of the two commercially operated technologies, the toluene oxidation route affords not only phenol but also the specialty chemicals benzaldehyde and benzoic acid (Equation 13.1):

It is only run, quite successfully, by DSM and in few small plants licensed by DSM: this is because the amount of benzaldehyde and benzoic acid thus produced matches the market demand. Should the technology be adopted by others, the quantity of these specialty derivatives would quickly exceed their demand. Consequently, new plants are almost exclusively based on the cumene process, which, in the following, will be examined into some detail.

13.2

Cumene Process

The cumene process, sometimes referred to as the Hock process, was made possible by the discovery of cumyl hydroperoxide and of its cleavage to phenol and acetone [1]. Shortly after World War II the reaction was developed into an industrial process by the Distillers Co. (BP Chemicals) in the United Kingdom and Hercules in the USA. The first commercial plant was started in Montreal, Canada, in 1952 by M.W. Kellogg.

The process is based upon three different reactions: (i) Friedel-Crafts alkylation of benzene with propene to afford cumene (isopropylbenzene); (ii) cumene oxidation with oxygen to give cumyl hydroperoxide; and (iii) cleavage of cumyl hydroperoxide in acidic medium to afford phenol and acetone (Equation 13.2):

H3Cn ,CH3 H3c8!jfoOH QH OH C* T

13.2.1 Alkylation

Cumene is produced by alkylating benzene with propene. The reaction needs a catalyst: in recent years, zeolite-based catalysts have become almost universally used in cumene plants [2], although older plants using aluminum chloride (AlCl3) or phosphoric acid supported on silica are still operating.

Zeolite-based catalysts are definitely less hazardous materials than aluminum chloride or phosphoric acid and, being non-corrosive, they allow a reduction in plant maintenance. Furthermore, they can be regenerated offsite by burning off high molecular weight hydrocarbons deposited on them: the expected catalyst cycle length between regenerations is 36-60 months, and the expected life four or five reaction/ regeneration cycles. At the end of their life, zeolite-based catalysts can be safely disposed of as landfill after hydrocarbon removal, thus eliminating disposal problems associated with AlCl3 or phosphoric acid. Consequently, the use of

H3C/ CH3

Dipb From Cumene
Figure 13.1 Cumene plant: block scheme [3].

zeolite-based catalysts significantly improves the sustainability of the cumene production. However, it requires feedstocks of a relatively high purity and some pretreatment can be necessary.

In all cases, the main by-product of the alkylation reaction is a mixture of diisopropylbenzenes (DIPB) which, with zeolite-based catalysts, accounts for 5-25% of the whole alkylated product. Cumene can be recovered from DIPB by transalkylation with one mole of benzene to form two moles of cumene.

In a typical arrangement (Figure 13.1), alkylation occurs at 130-170 °C in liquid phase on four catalyst beds packed in two reactor shells, arranged in parallel. A mixture of fresh and recycled benzene is charged downflow through the alkylation reactors. Excess benzene is used to minimize polyalkylation and olefin oligomeri-zation. Fresh propene feed (which may contain inert propane in an amount depending on the propene source) is split between the four catalyst beds, and is completely consumed in each bed. The reaction of benzene and propene is exothermic; the temperature rise in each reactor may be controlled by recycling part of the reactor effluent, which absorbs the heat of reaction. Effluent from the alkylation reactors is first sent to a depropanizer column whose bottom stream is sent to a second column, where unreacted benzene is collected overhead and recycled, while the bottom stream is sent to a further column where cumene is recovered overhead. The bottom of the cumene column is sent to a DIPB column, where DIPBs are recovered overhead and sent to the fixed bed transalkylation reactor. In turn, the bottom product from the DIPB column, consisting primarily of heavy aromatics, is typically blended with fuel oil for burning.

Table 13.1 gives a typical material balance of the alkylation section. Notably, however, depending on the specific technology and catalyst offered by each cumene licensor, raw material consumption may be in some cases significantly better. As an example, a total amount of heavies as low as 0.0031 per t of cumene is now claimed by Polimeri Europa.

Material

Metric tonne per metric tonne of cumene

Feed

Benzene (as 100 % purity) Propene (as 100 % purity)

0.652 0.352

Product

Cumene

1.000

By-product

Heavies

0.005

Oxidation and Concentration

In the second step of the cumene process, cumene undergoes oxidation with air, possibly oxygen-enriched, to afford cumyl hydroperoxide (CHP). This is a typical radical reaction and the hydroperoxide forms at the expense of the less energetic, tertiary C—H bond. The reaction is carried out in liquid phase, at 85-120 °C and 4-10 bar. Owing to the formation of small amounts of acid by-products (mainly formic acid), unstabilized systems work at pH 3-6. However, these acids promote the decomposition of cumyl hydroperoxide to acetone and phenol: the latter is an excellent inhibitor of radical reactions and its presence is not compatible with the autoxidation. Therefore, it is common practice to stabilize the reaction medium through the addition of an emulsified weakly basic aqueous phase (sodium hydroxide or carbonate, pH 7-8). Initially, the oxidation is quite slow and several catalysts have been described that can speed up this step. However, the reaction is autocatalytic and its rate gradually increases with increasing hydroperoxide concentration. Main by-products are 2-phenyl-2-propanol and acetophenone, both arising from thermal decomposition of cumyl hydroperoxide, and dicumyl peroxide, formed by an equilibrium reaction of cumyl hydroperoxide with 2-phenyl-2-propanol.

The oxidation is mostly carried out in traditional bubble column reactors: series of two to six reactors, up to 20 m high, are common in industry. The reaction is exothermic: ~120kJ are released per mole of produced hydroperoxide, and must be removed by cooling. The final reaction mixture, containing 20-30% of cumyl hydroperoxide, is then concentrated by distilling off some unre-acted cumene to obtain a 65-85% hydroperoxide to be fed to the cleavage step (Figure 13.2).

Significant attention is paid during both the reaction and the concentration to prevent ignition or explosion of the cumene-air mixtures. Furthermore, provisions are needed for a water or a steam quench to the concentrators to prevent hydroperoxide decomposition in case of an emergency.

Phenol Production Economic
Figure 13.2 Phenol plant: block scheme [3].

Cleavage and Workup

Cumyl hydroperoxide is eventually cleaved in the presence of an acid catalyst, to yield phenol and acetone, together with minor amounts of by-products such as a-methylstyrene, arising from dehydration of 2-phenyl-2-propanol, and dicumyl peroxide. a-Methylstyrene can be recovered to cumene in a hydrogenation stage.

As the catalyst, concentrated (98%) sulfuric acid is almost exclusively used in the industrial practice. The cleavage is run in the presence of 0.2-1% of acid, under reduced pressure, at the boiling temperature of the cumyl hydroperoxide-acetone mixture, which depends on the acetone content. Again, the reaction is strongly exothermic (^250 kJ mol"1) and the heat is removed by evaporation of acetone from the reaction mixture. At 70-80 °C, cumyl hydroperoxide conversion is virtually quantitative, with a selectivity to phenol up to 94-95%.

Acetone and phenol can be recovered after the neutralization of the acidic mixture from the cleavage reactor with sodium hydroxide or phenolate solution. The neutralized mixture is then subjected to a series of distillations. Acetone is first distilled, then cumene is recovered, together with a-methylstyrene, which is either purified and marketed or hydrogenated back to cumene and recycled to the oxidation. Phenol is finally distilled with a purity up to 99.99%, suitable for the production of polycarbonate grade bisphenol A and other chemicals and polymers.

The material balances of the oxidation/cleavage section and of the overall process are reported in Tables 13.2 and 13.3, respectively. Again, raw material consumption

Table 13.2 Phenol plant: material balance [3].

Material

Metric tonne per metric

tonne of phenol

Feed

Cumene

1.330

Product

Phenol

1.000

By-products

Acetone

0.626

Heavies

0.053

Process water

0.493

Table 13.3 Overall material balance.

Material

Metric tonne per metric

tonne of phenol

Feed

Benzene

0.867

Propene

0.468

Product

Phenol

1.000

By-products

Acetone

0.626

Heavies

0.060

Process water

0.493

may in some cases be significantly better, depending on the specific phenol technology offered by each phenol licensor.

Cumene Process: Final Considerations

Since 1952, the cumene process has been improved to a fairly impressive extent as its yields to acetone and phenol, based on both benzene and propene, are currently close to the stoichiometric ones. Nor are there problems in building huge plants: 300 0001 yr—1 capacity is a common size, while still having a quite reduced environmental impact. Thus, although some precautions are needed to ensure safe operations, the process is fully satisfactory in many aspects.

Some improvements, in terms of reduction of energy consumption, could be attained by increasing the benzene conversion per pass in the alkylation and oxidation stages, thus reducing benzene and cumene recycles. By assuming a reaction molar ratio between benzene and propene as low as 2.5, currently allowed by the most competitive cumene technologies based on the latest generation of zeolite-based catalysts, the benzene conversion per pass in the alkylation stage remains below the 45% while, in the cumene oxidation stage, the conversion per pass is not higher than 35% in order to maintain a safe concentration of the intermediate cumyl hydroperoxide with good selectivity. Consequently, benzene conversion per pass over the benzene alkylation and cumene oxidation stages (cumyl hydroperoxide conversion in the cleavage stage can be considered quantitative) is never higher than about 15%, which means that the cumene process to phenol is inherently affected by relevant recycle volumes.

However, the major problem in cumene process to phenol seems to be the fixed coproduction of about 0.61 of acetone per t of phenol, while the yearly market growth for phenol applications is nowadays higher than the market growth for acetone, whose sale price is then continuously squeezed by the market. The current market situation trends to a further worsening for the acetone sale price in the future, based mainly on two facts: (i) on one hand, all volatile organic compounds (including acetone) are subjected to a strong, increasing environmental pressure, urging for their substitution in the medium term and (ii) on the other hand, the main current use of phenol is in polycarbonate synthesis, which does not actually require phenol itself but rather bisphenol A, whose production requires just one mole of acetone per two moles of phenol (Equation 13.3), inherently leading then to unbalanced acetone and phenol consumptions:

Any addition of phenol capacities adds a further unbalance on the acetone side.

A first approach to the problem has been the development, by Mitsui Petrochemical, of a recycle scheme for converting the acetone back into propene (via hydrogenation to 2-propanol and subsequent dehydration of the latter) to be added to the feed of the alkylation step (Scheme 13.1).

bisphenol A

bisphenol A

Alkylation Using Acetone

Scheme 13.1 Mitsui Petrochemical acetone recyle process.

H2 cat.

Scheme 13.1 Mitsui Petrochemical acetone recyle process.

Friedel-Crafts alkylation with 2-propanol is also possible [4], without the need of the dehydration step, whose energy requirement accounts for the most part of the overall energy balance for the acetone into propene transformation.

As the recycle of acetone to benzene alkylation to cumene leads to a corresponding saving in propylene, its profitability depends on the acetone to propene sale prices ratio.

Polimeri Europa has developed a technology based on the direct alkylation of benzene with 2-propanol (thus avoiding its dehydration stage) where the acetone recycle versus the acetone sale profitability is claimed at acetone to propylene sale prices ratio equal or lower to 0.6.

Nevertheless, much effort is currently being devoted to decouple phenol and acetone productions and, particularly, to develop effective processes for the direct oxidation of benzene to phenol.

13.3

Solutia Process

The selective insertion of an oxygen atom into a benzene carbon-hydrogen bond to yield phenol is not a classical organic chemistry reaction. The first process for such a reactions was the Solutia process, based on discoveries by Panov and coworkers at the Boreskov Institute of Catalysis in Novosibirsk and then developed in close cooperation with Monsanto. In this process, the oxidant is nitrous oxide, N2O, while an iron-containing zeolite is used as the catalyst (Equation 13.4):

The reaction is run in the gas phase at 350 ° C and, at 27% of benzene conversion, selectivity for phenol is claimed to be 98% [5]. The main by-products are dihydrox-ybenzenes (about 1%) and carbon oxides (0.2-0.3%). Selectivity is of paramount importance for this process, since 15 molecules of nitrous oxide are consumed for the total oxidation of just one molecule of benzene (Equation 13.5):

The catalyst is an iron-containing ZSM-5 zeolite. Its half-life is three to four days so that, periodically, catalytic activity must be restored by passing air through the deactivated catalyst at high temperature: no performance deterioration has been reported after more than 100 regeneration cycles.

Figure 13.3 shows a plant layout. Recycled benzene along with makeup benzene and nitrous oxide are preheated and continuously fed to a moving bed reactor utilizing the zeolite catalyst. The latter flows vertically down the reactor by gravity, while the reaction gas flows across the annular catalyst beds. The predominant reactions are exothermic: about 250 kJ are released per mole of phenol produced. In addition, significantly more heat can be generated by the deep oxidation of benzene to

Benzene

Vaporizer

A

Preheater

Benzene oxidation

Distillation

Phenol purification

Gas recycles

Phenol

Heavies

Figure 13.3 Solutia process: block scheme [6].

Benzene recycles

Gas recycles

Phenol

Heavies

Figure 13.3 Solutia process: block scheme [6].

Offgas

Cooler carbon dioxide. So, the temperature rises as the gas flows radially across the catalyst beds. The reactor effluent is collected and cooled. Offgas (mainly nitrogen) is incinerated or vented to atmosphere, whereas unreacted benzene and phenol are condensed and sent to a crude phenol column. Benzene is fractionated overhead and recycled to the reactor. The crude phenol from the bottom of the column is sent to a phenol purification column, where phenol product is taken overhead and the heavies from the bottom of the column are sent to fuel. At the same time, partially deactivated catalyst is continuously withdrawn from the bottom of the moving bed reactor and transported to the regenerator, where the accumulated coke is burnt off. The regenerated catalyst is then transported again to be fed at the top of the moving bed reactor.

Table 13.4 gives the material balance of the Solutia process.

There is significant debate about the mechanism of this reaction and, in particular, about the nature of the iron sites responsible for the unique reactivity. It is generally

Table 13.4 Solutia process: material balance [6].

Material Metric tonne per metric tonne of phenol

Feed

Benzene 0.874

Nitrous oxide 0.551

Product

Phenol 1.000 By-products

Dihydroxy benzenes 0.011

CO2 0.029

accepted that the reaction proceeds over the so-called a-sites - defect sites of the zeolite that are formed in an iron-containing zeolite matrix upon high temperature activation, and where iron atoms migrate. The iron atoms that make up a-sites are in a bivalent state, with a special affinity to nitrous oxide. N2O decomposition causes the transition Fe2 + to Fe3 +, producing the so-called atomic a-oxygen species, which can selectively insert oxygen into C—H bonds of alkanes and aromatics [7].

Despite its brilliant results, it seems unlikely that the Solutia process can become a major source of phenol. Nitrous oxide availability is quite limited and its production on-purpose (by the conventional ammonium nitrate decomposition, which enables nitrous oxide of high purity to be produced for medical anesthetic applications, or even by selective oxidation of ammonia) would result too expensive. Therefore, the only reasonable scenario to exploit the Solutia process is its implementation close to adipic acid plants, where nitrous oxide is co-produced by the nitric oxidation of cyclohexanol-cyclohexanone mixtures and where it could be used to produce phenol instead of being disposed of. However, the stoichiometry of the process is such that a relatively small phenol plant would require a world-scale adipic acid plant for its nitrous oxide supply. In fact, a pilot plant has been operated using this technology, but its commercialization has been postponed.

13.4

Direct Oxidation of Benzene to Phenol with Hydrogen Peroxide

A promising alternative to nitrous oxide is provided by hydrogen peroxide, which is finding more and more favor due to the lack of environmental impact and the easy storage and handling. Both homogeneous and heterogeneous catalysts have been developed recently to be used in the direct oxidation of benzene to phenol by hydrogen peroxide. In the first case, soluble iron complexes were used under biphase conditions, while titanium-containing zeolites were selected as heterogeneous catalysts.

Definition of the Problem and First Attempts

The direct oxidation of benzene to phenol is usually affected by a poor selectivity due to the lack of kinetic control. Indeed, phenol is more reactive towards oxidation than benzene itself and consecutive reactions occur, with substantial formation of over-oxidized products like catechol, hydroquinone, benzoquinones and tars. This is the usual output of the oxidation of aromatic hydrocarbons by the classical Fenton system, a mixture of hydrogen peroxide and an iron(II) salt, usually ferrous sulfate, most often used in stoichiometric amounts [8].

Thus, to obtain a selective synthesis of phenol via direct oxidation of benzene, suitable strategies have to be envisaged to slow down the undesired consecutive reactions and to allow phenol to accumulate. The first step in this direction was made by George Olah, who used extremely concentrated (98%) hydrogen peroxide in a superacidic medium (FSO3H-SbF5,1:1) at —78 °C and obtained phenol with a 54% yield, based both on the benzene and on the hydrogen peroxide [9]. Owing to the very harsh reaction conditions, this work has only historical value, but the concept that, in the superacidic medium, the phenol was protonated (thus being deactivated towards any further oxidation) was noteworthy.

A few years later, Hubert Mimoun discovered the rather selective oxidation of benzene by a few vanadium(V) peroxo complexes [10]. Using an excess of hydrogen peroxide under phase-transfer conditions transformed this stoichiomet-ric reaction into a true catalytic process, but the turnover numbers remained very low [11].

Homogeneous Catalysis by Iron Complexes: A Biphase Fenton Reagent

Over-oxidation problems are solved efficiently in biological systems by segregating catalysts and products into different environments. So, for instance, the active sites of several oxygenases are buried deeply into hydrophobic pockets where lipophilic substrates are readily oxidized, while the more hydrophilic reaction products, when released into the surrounding aqueous environment, do not have further access to the catalytic site [12].

A simple biphasic system can mimic these important features of biological systems and this observation is the basis for a benzene oxidation with hydrogen peroxide. According to this methodology, the reaction medium is formed by an aqueous phase, containing both the hydrogen peroxide and the oxidation catalyst, and an organic one able to dissolve most of the produced phenol. The use of the aqueous-organic reaction medium dramatically affects the selectivity of the reaction; in particular, a remarkable enhancement is obtained using a biphasic system generated by water and acetonitrile (volume ratio = 1:1) in the presence of benzene. With this particular medium, the concentration of benzene in the aqueous phase rises from 0.18% (the solubility of benzene in water) to 0.76%, and the produced phenol was extracted for the most part (85%) in the organic phase. As a consequence, the benzene/phenol molar ratio is only 0.25 in the aqueous phase of a water-benzene mixture, whereas it rises to 3.7 in the water-acetonitrile-benzene one. In this way, the biphasic operation minimizes the over-oxidation reactions by reducing the contact between the phenol and the catalyst, which segregates in the aqueous phase.

The catalyst is a water-soluble iron(II) salt, typically iron sulfate, the performances of which are improved by the addition of a suitable, bidentate ligand. N-N ligands (e. g., phenanthroline derivatives) have a detrimental effect on the hydrogen peroxide activation and result in very low conversions. O-O ligands (e.g., catechol derivatives) show a good activity but also promote the hydrogen peroxide decomposition, thus resulting in rather low selectivities. However, N-O ligands, especially pyrazinecar-boxylic acid derivatives, give the most interesting results. 2-Methylpyrazine-5-car-boxylic acid N-oxide (Figure 13.4) and the corresponding ligand without the methyl group (pyrazine-3-carboxylic acid N-oxide) are the most efficient.

COOH

Figure 13.4 2-Methylpyrazine-5-carboxylic acid N-oxide.

Using FeSO4 (1.67 x 10"3 M) in conjunction with equimolar amounts of methyl-pyrazine-5-carboxylic acid N-oxide and trifluoroacetic acid, in a water-acetonitrile-benzene (5:5:1 v/v/v) biphasic system, with benzene-H2O2-FeSO4 = 620:60:1, a benzene conversion of 8.6% is achieved (35 °C; 4h). Hydrogen peroxide conversion is almost complete (95%) and selectivities to phenol are 97% (based on benzene) and 88% (based on H2O2) [13]. These values are definitely higher than those described in the literature for the classical Fenton system [14], whereas iron complexes with pyridine-2-carboxylic acid derivatives are reported to be completely ineffective in the oxidation of benzene under the well-known Gif reaction conditions [15].

Regarding the possible mechanism, notably, toluene, ethylbenzene and tert-butylbenzene are less reactive than benzene, which is not consistent with the expected order for an electrophilic aromatic substitutions, such as that found with the classic Fenton reagent. There are also other differences with respect to the Fenton chemistry. In particular, under biphase conditions the reaction is definitely more selective: although comparisons are difficult due to the huge amount of data, sometimes inconsistent, on the Fenton system (for which most of the data have been obtained with the iron used in stoichiometric amounts) it seems that selectiv-ities close to those observed under biphase conditions are only attained at a conversion around of 1%. Furthermore, in the biphase system, only a negligible amount (<1%) of biphenyl was detected among secondary products, whereas in the classic Fenton oxidation this compound is formed by radical dimerization of hydroxycyclohexadienyl radicals in typical yields ranging from 8 to 39%.

Is a different mechanism operative under biphase conditions? Although detailed mechanistic information is not currently available, the reaction is likely to start with the oxidation of Fe(II) to Fe(III) (Equation 13.6):

Thus, it is also likely that hydroxyl radicals are present in the reaction mixture, where they probably act as oxidizing agents according to the overwhelmingly established Fenton radical mechanism, and differences could simply arise due to the peculiar biphase system: for instance, toluene is less soluble than benzene in the aqueous phase, in which the reaction takes place (98.9 and 142.3 mmol L"1, respectively) and this could help to explain its lower reaction rate. Alternatively, these differences might be the result of a competition with a second mechanism. Following several suggestions [16], the new mechanism could be triggered by the formation of an iron(III) hydroperoxo species, which could undergo heterolytic cleavage to afford an electro-philic, high-valent iron-oxo complex, stabilized by the ligand, able to oxidize aromatic hydrocarbons to phenols (Scheme 13.2).

Iron Oxo Complexes

X2Fev

Scheme 13.2 Possible catalytic cycle involving the intervention of high-valent iron species. 13.4.3

Heterogeneous Catalysis by Titanium Silicalite

X2Fev

Scheme 13.2 Possible catalytic cycle involving the intervention of high-valent iron species. 13.4.3

Heterogeneous Catalysis by Titanium Silicalite

A process based on the iron catalysis described above would be affected by a low volume productivity, in terms of the amount of produced phenol in a given time per liter of reactor volume. This is a common weakness of homogeneous versus heterogeneous catalysis. However, at the same time the iron-catalyzed reaction was being studied, a second, heterogeneous, catalyst was found that can catalyze the direct oxidation of benzene to phenol with hydrogen peroxide: titanium silicalite (TS-1). It was discovered in 1979 in the laboratories of the ENI group in San Donato Milanese and the relevant patent was filled on December 21,1979. It is a crystalline, synthetic zeolite in which tetrahedral [SiO4] and [TiO4] units are arranged into an orthorhombic MFI structure (ZSM-5 type), with titanium replacing up to 3% of silicon atoms. Owing to its structure, TS-1 shows a three-dimensional system of channels having near-circular section, which constitute the zeolitic micropores of the material: two main sub-systems are present, sinusoidal channels and straight ones, with diameters of 5.6 x 5.3 and 5.3 x 5.1 A, respectively [17].

Soon after its discovery, TS-1 was recognized as a valuable catalyst for many oxidations by hydrogen peroxide, including alkane oxidation, olefin epoxidation, alcohol oxidation, phenol hydroxylation and cyclohexanone ammoximation (Scheme 13.3) [18].

However, the activity of TS-1 in the oxidation of benzene appeared to be very poor. TS-1 does not perform very well in two-phase systems, so that only solvents able to homogenize the hydrophobic substrate and the aqueous hydrogen peroxide can be used. Even in this case, using solvents such as acetone, acetonitrile or tert-butanol, the selectivity to phenol rapidly dropped at very low benzene conversion, mainly due to the formation of dioxygenated products and tars: typically, selectivity was already less than 50% at benzene conversion as low as 5%. Even worse results were obtained using methanol, which was oxidized in competition with benzene to give formaldehyde dimethyl acetal.

Benzoquinone Formation Hydroquinone
Scheme 13.3 Some oxidation reactions catalyzed by TS-1.

In contrast, the use of sulfolane as solvent allows a conversion of benzene close to 8% while maintaining the selectivity to phenol higher than 80%. Detected byproducts are catechol (7%), hydroquinone (4%), 1,4-benzoquinone (1%) and tars

Even better results are obtained by a post-synthesis treatment of TS-1 with both hydrogen peroxide and ammonium hydrogen fluoride, NH4HF2. Upon such a treatment (H2O2/F/Ti = 10:2.5:1; 60 °C; 4h), a substantial amount of titanium (up to 75% of the initial value) is removed. Nevertheless, the crystalline structure of the zeolite remains unchanged and the catalytic activity does not decrease. On the contrary, it actually increases since the turnover frequency of residual titanium atoms rises from 31 to 80 h-1. Even more importantly, at 8.6% benzene conversion the selectivities, both on benzene and on hydrogen peroxide, also increase from 83 to 94% and from 67 to 83% respectively, with formation of catechol (4%) and hydroquinone (2%) as the only by-products, without any evidence of further oxidation reactions [19].

The treated catalyst has been named TS-1B. A preliminary characterization of it (with residual titanium 29% of the original amount) shows that it has a peculiar UV/ Vis spectrum. In particular, absorption at 48 000-50 000 cm-1 (typical of pure TS-1) was strongly reduced and a new titanium species, absorbing at 40000 cm-1, was generated by the treatment. The formation of amorphous extraframework titanium species (TiO2), absorbing at 30000-35 000cm-1, was not observed (Figure 13.5).

A possible mechanism for the TS-1-catalyzed hydroxylation of phenol with hydrogen peroxide has been proposed [20]: it is an ionic mechanism that, with

40000 30000

Wavenumber (cm1) Figure 13.5 UV/Vis DRS (diffuse reflectance spectroscopy) of TS-1 and TS-1B. TS-1 (solid line); TS-1B after extraction of 13% of TiO2 (dashed line); TS-1B after extraction of 48% of TiO2 (dot-dashed line).

minor modifications, can probably hold also for benzene hydroxylation. Alternative radical pathways are also possible.

Why is sulfolane, and only sulfolane, so effective in improving selectivities? Already in 1963, Drago suggested that sulfolane forms a complex with phenol via hydrogen bonding (Figure 13.6) [21].

Figure 13.6 Possible structure of the phenol/sulfolane complex.

Such a complex is not stable enough to be isolated but its occurrence was inferred by the IR spectra of the solutions. Drago's suggestion is confirmed by ab initio calculations. Therefore, the improved selectivity observed upon carrying out benzene oxidation in sulfolane may be due to the formation of this large species, which can not enter the titanium silicalite pores, thus allowing phenol to remain relatively protected against further oxidation.

This effect was confirmed by calculation of the loading of the free phenol molecule and the complex phenol-sulfolane (expressed as the number of loaded molecules in a crystal elementary unit of TS-1), using the software Sorption (Cerius 2), which turned out to be 13.6 and 0.8, respectively. Alternatively, the protective effect exerted by sulfolane can be evaluated by measuring the reaction rate, expressed as the turnover frequency (TOF: moles of reacted substrate/moles of Ti per hour) for the oxidation of benzene and phenol, carried out separately in acetone and sulfolane as co-solvents. In the case of acetone, the phenol oxidation (TOF = 190) was ten times faster than that measured for benzene (TOF = 19); conversely, operating in sulfolane the rate measured for phenol (TOF = 51) was only 1.6 time higher than that measured for benzene (TOF = 31), according to the higher value of the observed selectivity.

Ni/Mo H2

Scheme 13.4 Polimeri Europa TS-1/H2O2 process.

Many attempts have been carried out to find catalysts other than TS-1 able to catalyze the oxidation of benzene to phenol with hydrogen peroxide. With mesopor-ous molecular sieves (Ti-HMS, Ti-MCM-41, Ti-MSA, Ti-ERS8, Ti-AMM, all with pore diameters of 2-5 nm), and with large pores zeolites (like Ti-p or TAPO5, with pore size around 6-7 A), hydrogen peroxide decomposition was prevalent on benzene conversion, and only traces of phenol were found. Among the zeolites with pore size similar to titanium silicalite, Ti-ZSM-48 (a microporous material with a disordered structure, characterized by a tubular, monodimensional channel system with near circular pores of 5.3-5.6 A) consumed almost completely the hydrogen peroxide without any phenol production. In contrast, titanium silicalite-2, TS-2, showed a moderate activity in benzene hydroxylation. TS-2 is a synthetic zeolite with a tetragonal MEL structure (ZSM-11 type), with only straight channels with near-circular sections and diameters of 5.4 x 5.3 A [17]. The TOF so far obtained in the oxidation of benzene to phenol with hydrogen peroxide catalyzed by TS-2 is only roughly half of that found with TS-1 [22], but it is likely to be improved with a careful optimization of the catalyst (particle size control, post-synthesis treatments, etc.). However, the synthesis of TS-2 is difficult because it is not easy to obtain a pure MEL structure and this handicap can limit its industrial application. Consequently, TS-1 is still the only candidate catalyst for industrial development.

To further increase the overall yield of the process, a second step can be added in which dihydroxylated by-products, hydroquinone and catechol, are treated with hydrogen and partially deoxygenated to phenol, which is recycled back to the process (Scheme 13.4) [23].

The hydrodeoxygenation reaction (HDO) is carried out in the gas phase in a fixed bed reactor (400 ° C, 25 bar of hydrogen), using commercial nickel and molybdenum oxides supported on alumina as catalysts. The HDO allows a quantitative transformation of dioxygenated compounds into phenol with a selectivity of 96% [24]. Main by-products are heavy condensed polycyclic aromatic hydrocarbons.

In a typical arrangement, representative of the whole process (Figure 13.7), the oxidation reaction is carried out in a biphasic mixture of benzene, sulfolane and water (30/50/20 w). The hydrogen peroxide is used as an aqueous solution (35% w), with a total molar ratio of H2O2/benzene of 0.21.

The oxidation is carried out in fixed bed reactors, operating at 6 atm under adiabatic conditions with an inlet temperature of 95 ° C and outlet temperature of 110 ° C. The overall oxidation section performances per pass are:

Benzene conversion (%) 15.4

H2O2 conversion (%) 100

Selectivity on benzene (%)a 84.6

Selectivity on H2O2 (%)b 61.4

a Moles of produced phenol/moles of converted benzene x 100. b Moles of produced phenol/moles of converted H2O2 x 100.

Notably, benzene conversion is quite similar to that usually achieved in cumene process over the benzene alkylation and cumene oxidation stages.

The main by-products are catechol (0.121 pert of phenol), hydroquinone (0.0651 pert of phenol) and phenolic tars (0.021 pert of phenol).

The reaction mixture, coming from the reaction section, is sent to the separation section for the recovery of benzene, water and phenol, by consecutive distillation.

The by-products (catechol, hydroquinone and tars) are separated from the solvent by salification with NaOH and extraction, thus avoiding the complete distillation of the high boiling sulfolane.

The by-products, obtained as sodium salts in aqueous solution, are recovered by neutralization with H2SO4 and extraction with methyl isobutyl ketone (MIBK).

After separation of MIBK by distillation, the by-products, obtained in aqueous solution, are fed to the HDO section for selective hydrogenation. The HDO reaction is carried out in fixed bed reactors, operating under adiabatic conditions at 4000C inlet temperature and 25 bar of hydrogen pressure. Typically, the hydrogen/dihydroxy benzenes molar ratio is set to 20. The produced phenol is recovered by distillation and recycled to the process cycle, thus avoiding any coproduction of dihydroxybenzenes.

Phenol Process
Figure 13.7 Polimeri Europa TS-1/H2O2 process: block scheme.

The HDO section performances per pass are:

Catechol conversion: 100%

Hydroquinone conversion: 100%

Selectivity on dihydroxybenzenes: 96%

where the selectivity is given by 100 x (the number of moles of produced phenol/ moles of converted dihydroxybenzenes).

The main by-products formed in the HDO reaction are: ortho- and para-cresols (0.4 kg per t of phenol; arising from toluene impurities of benzene feeding), cyclohexylbenzene (7.2 kg per t of phenol), biphenyl (4.4kg per t of phenol), dibenzofuran (3.2 kg per t of phenol) and condensed polycyclic aromatic hydrocarbons (25.2kgpert of phenol).

All the by-products, except cresols, are separated from the phenol produced in the HDO section by distillation. The resulting crude phenol is combined with that produced in the reaction section and fed to the purification section.

Typically, the fraction of phenol, produced in the HDO section corresponds to 13% of the total produced phenol.

The overall process performance, including oxidation and HDO sections, are:

Benzene conversion (%) 100

H2O2 conversion (%) 100

Selectivity on benzene (%)a 97.7

Selectivity on H2O2 (%)b 71.0

a Moles of produced phenol/moles of converted benzene x 100. b Moles of produced phenol/moles of converted H2O2 x 100.

A block scheme and the material balance of this Polimeri Europa process are reported in Figure 13.7 and Table 13.5, respectively.

Preliminary economic evaluations suggest that direct oxidation by hydrogen peroxide is not yet competitive with the traditional cumene process as well as with

Table 13.5 Polimeri Europa TS-I/H2O2 process: material balance.

Material Metric tonne per metric tonne of phenol

Feed

Benzene 0.849

Hydrogen peroxide (100%) 0.509

Product

Phenol 1.000 By-products

Na2SO4 0.239

Heavies 0.022

Process water 6.830

the acetone recycle technology, but also that it could become convenient if the acetone sale price should further fall close to its fuel value.

13.5

Perspectives

The last developments ofhydroxylation ofbenzene to phenol are directed towards the direct oxidation with molecular oxygen, most often in the presence of a reducing agent such as hydrogen, carbon monoxide or even ammonia [25].

In several cases, the in situ formation of hydrogen peroxide is the first step of the process. Thus, phenol can be obtained from benzene, carbon monoxide (5 atm) and oxygen (65 atm) at 70 °C in a benzene-water-methyl isobutyl ketone mixture, with TS-1 and a palladium complex as catalysts [26]. Despite a 91% selectivity to phenol, benzene conversion (3.2%) and productivity are still too low for industrial application. The palladium complex is required to promote hydrogen peroxide formation upon reaction of oxygen, carbon monoxide and water [27].

Carbon monoxide has also been used as the reducing agent by Ishii and coworkers, who reported the oxidation ofbenzene to phenol using air (15 atm), CO (5 atm) and molybdovanadophosphoric acid as catalysts. The reaction is carried out at 90 °C for 15 h, with a good phenol yield (27%) and a selectivity up to 90% [28]. Daicel Chemical Industries Ltd (Tokyo) recently announced its interest in this process. In this case, however, the authors are inclined to believe that the in situ generation of hydrogen peroxide is unlikely: rather, from their experimental results, they conclude that the oxygen is activated on the catalyst and then reacts with the aromatic ring.

The use of oxygen-carbon monoxide or oxygen-hydrogen mixtures presents safety problems due to the occurrence of explosive mixtures: separate supply of the gases into the reaction medium, for example through a permeable membrane, is an approach that may solve this problem, either by producing hydrogen peroxide in situ [29] or by using the gas mixture to directly oxidize benzene. For instance, the conversion of benzene into phenol has been reported by supplying hydrogen and oxygen through a palladium membrane [30]. The reaction was carried out in continuous mode at 200 °C: phenol was produced with a selectivity of up to 80% at a benzene conversion around 10-15%.

Despite the attractive perspectives of membrane technology, many basic problems have still to be solved. Beyond the optimization of reaction conditions and catalysts, the chief obstacles to the scaling-up are membrane fragility, deterioration, high cost and manufacture complexity, which restrict, for the moment, this technique to an experimental level.

13.6

Conclusions

It is currently impossible to know if future developments in the direct oxidation of benzene will result in phenol production by a process really competitive with the cumene route. However, the results obtained so far provide a sound basis for future research. In particular, they have definitely shown that it is not unavoidable that the oxidation ofphenol proceeds at a rate higher than its formation, thus opening the way to the development of even more selective and, possibly, economic processes.

Acknowledgment

The authors wish to thank Gianni Girotti, from Polimeri Europa S.p.A., for his skilful and valuable revision of the manuscript.

References

1 Hock, H. and Lang, S. (1944) Autoxidation ofhydrocarbons. IX. Peroxides ofbenzene derivatives. Ber. Dtsch. Chem. Ges. B, 77, 257-264.

2 Bellussi, G. and Perego, C. (2000) Industrial catalytic aspects ofthe synthesis of monomers for nylon production. Cattech, 4, 4-16.

3 Meyers, R.A. (ed.) (2005) Handbook of Petrochemicals Production Processes, McGraw-Hill, Ch. 9.1.

4 Girotti, G., Rivetti, F., Ramello, S. and Carnelli, L. (2003) Alkylation of benzene with isopropanol on ß-zeolite: influence of physical state and water concentration on catalyst performances. J. Mol. Catal., 204-205, 571-579.

5 Panov, G.I. (2000) Advances in oxidation catalysis: oxidation ofbenzene to phenol by nitrous oxide. Cattech, 4, 18-32.

6 Uriarte, A.K., Rodkin, M.A., Gross, M.J., Kharitonov, A.S. and Panov, G.I. (1997) Direct hydroxylation ofbenzene to phenol by nitrous oxide. Stud. Surf. Sci. Catal., 110, 857-864.

7 Dubkov, K.A., Ovanesyan, N.S., Shteinman, A.A., Starokon, E.V. and Panov, G.I. (2002) Evolution of iron states and formation of a-sites upon activation of FeZSM-5 zeolites. J. Catal., 207, 341-352.

8 Symons, M.C.R. and Gutteridge, J.M.C. (1998) Free Radicals and Iron: Chemistry,

Biology, and Medicine, Oxford University Press Oxford.

9 Olah, G.A. and Ohnishi, R. (1978) Oxyfunctionalization of hydrocarbons. 8. Electrophilic hydroxylation of benzene, alkylbenzenes, and halobenzenes with hydrogen peroxide in superacids. J. Org. Chem., 43, 865-867.

10 Mimoun, H., Saussine, L., Daire, E., Postel, M., Fischer, J. and Weiss, R. (1983) Vanadium(V) peroxo complexes. New versatile biomimetic reagents for epoxidation of olefins and hydroxylation of alkanes and aromatic hydrocarbons. J. Am. Chem. Soc, 105, 3101-3110.

11 Bonchio, M., Conte, V., Di Furia, F. and Modena, G. (1989) Metal catalysis in oxidation by peroxides. 31. The hydro-xylation of benzene by VO(O2)(PIC) (H2O)2: mechanistic and synthetic aspects. J. Org. Chem., 54, 4368-4371.

12 See, for example, Ortiz de Montellano, P.R. (1995) Cytochrome P450. Structure, Mechanism, and Biochemistry, 2nd edn, Plenum Press, New York.

13 (a) Vignola, R., Battistel, E., Bianchi, D., Bortolo, R. and Tassinari, R. (2002) Process for the hydroxylation of aromatic hydrocarbons. European Patent EP 861688. to EniChem; (b) Bianchi, D., Bortolo, R., Tassinari, R., Ricci, M. and Vignola, R. (2000) A novel iron-based catalyst for the biphasic oxidation of benzene to phenol with hydrogen peroxide. Angew. Chem. Int. Ed., 39, 4321-4323.

14 (a) Walling, C. (1975) Fenton's reagent. V. Hydroxylation and side-chain cleavage of aromatics. J. Am. Chem. Soc., 97, 363-367; (b) Ito, S., Mitarai, A., Hikino, K., Hirama, M. and Sasaki, K. (1992) Deactivation reaction in the hydroxylation of benzene with Fenton's reagent. J. Org. Chem., 57, 6937-6941.

15 Barton, D.H.R., Halley, F., Ozbalik, N. and Mehl, W. (1989) Benzylic oxidation by the GifIV system. Tetrahedron Lett., 30, 6615-6618.

16 (a) Barton, D.R. and Launay, F. (1998) The selective functionalization of saturated hydrocarbons. Part 44. Measurement of size of reagent by variation of steric demands of competing substrates using Gif chemistry. Tetrahedron, 54, 3379-3390; (b) Chen, K., Costas, M., Kim, J., Tipton, A. K. and Que, L. Jr (2002) Olefin cis-dihydroxylation versus epoxidation by non-heme iron catalysts: Two faces of an FeIII-OOH coin. J. Am. Chem. Soc., 124, 3026-3035.

17 International Zeolite Association, Atlas of Zeolite Framework Types, http://www.iza-structure.org (last accessed June 3th, 2009).

18 Arends, I.W.C.E., Sheldon, R.A., Wallau, M. and Schuchardt, U. (1997) Oxidative transformations of organic compounds mediated by redox molecular sieves. Angew. Chem. Int. Ed. Engl., 36, 1145-1163.

19 Balducci, L., Bianchi, D., Bortolo, R., D'Aloisio, R., Ricci, M., Tassinari, R. and Ungarelli, R. (2003) Direct oxidation of benzene to phenol with hydrogen peroxide over a modified titanium silicalite. Angew. Chem. Int. Ed., 115, 5087-5090.

20 Wilkenhoner, U., Langhendries, G., van Laar, F., Baron, G.V., Gammon, D.W., Jacobs, P.A. and van Steen, E. (2001) Influence of pore and crystal size of crystalline titanosilicates on phenol hydroxylation in different solvents. J. Catal., 203, 201-212.

21 Drago, R.S., Wayland, B. and Carlson, R.L. (1963) Donor properties of sulfoxides, alkyl sulfites, and sulfones. J. Am. Chem. Soc., 85, 3125-3128.

22 Unpublished results obtained within the European Project NEOPS (Novel Eco-efficient Oxidation Processes based on H2O2 Synthesis on catalytic membranes).

23 Bianchi, D., Bortolo, R., Buzzoni, R., Cesana, A., Dalloro, L. and D'Aloisio, R. (2004) Integrated process for the preparation of phenol from benzene with recycling of by-products. European Patent EP 1424320 to Polimeri Europa.

24 Dalloro, L., Cesana, A., Buzzoni, R. and Rivetti, F. (2004) Process for the preparation of phenol by means of the hydrodeoxygenation of benzene-diols. European Patent EP 1411038 to Polimeri Europa.

25 Bal, R., Tada, M., Sasaki, T. and Iwasawa, Y. (2006) Direct phenol synthesis by selective oxidation of benzene with molecular oxygen on an interstitial-N/Re cluster/ zeolite catalyst. Angew. Chem. Int. Ed., 45, 448-452.

26 Tassinari, R., Bianchi, D., Ungarelli, R., Battistel, E. and D'Aloisio, R. (1999) Process for the synthesis of phenol from benzene. European Patent EP 894783 to EniChem.

27 Bianchi, D., Bortolo, R., D'Aloisio, R. and Ricci, M. (1999) Biphasic synthesis of hydrogen peroxide from carbon monoxide, water, and oxygen catalyzed by palladium complexes with bidentate nitrogen ligands. Angew. Chem. Int. Ed., 38, 706-708.

28 Tani, M., Sakamoto, T., Mita, S., Sakaguchi, S. and Ishii, Y. (2005) Hydroxylation ofbenzene to phenol under air and carbon monoxide catalyzed by molybdovanadophosphoric acid. Angew. Chem. Int. Ed., 44, 2586-2588.

29 (a) Choudari, V.R., Gaikwad, A.G. and Sansare, S.D. (2001) Nonhazardous direct oxidation of hydrogen to hydrogen peroxide using a novel membrane catalyst. Angew. Chem. Int. Ed., 40, 1776-1779; (b) Melada, S., Pinna, F., Strukul, G., Perathoner, S. and Centi, G. (2006) Direct synthesis of H2O2 on monometallic and bimetallic catalytic membranes using methanol as reaction medium. J. Catal., 237,213-219; (c) Dittmeyer, R. and Svajda, K. (2007) Inherent and reliable selective method for directly synthesising hydrogen peroxide from oxygen and hydrogen with the aid of a catalytically coated wettable porous membrane. World Patent WO 2007028375 to Dechema.

30 Niwa, S., Eswaramoorthy, M., Nair, J., Raj, A., Itoh, N., Shoji, H., Namba, T. and Mizukami, F. (2002) A one-step conversion of benzene to phenol with a palladium membrane. Science, 295, 105-107.

Was this article helpful?

0 0
10 Ways To Fight Off Cancer

10 Ways To Fight Off Cancer

Learning About 10 Ways Fight Off Cancer Can Have Amazing Benefits For Your Life The Best Tips On How To Keep This Killer At Bay Discovering that you or a loved one has cancer can be utterly terrifying. All the same, once you comprehend the causes of cancer and learn how to reverse those causes, you or your loved one may have more than a fighting chance of beating out cancer.

Get My Free Ebook


Post a comment