Journal Archive

Platinum Metals Rev., 1983, 27, (2), 72

A New Platinum Catalyst for the Hydrogenation of Halonitroaromatics

  • By G. G. Ferrier
  • F. King
  • Johnson Matthey Group Research Centre

Article Synopsis

A platinum catalyst has been developed for the selective hydrogenation of halonitroaromatics. The catalyst performs favourably under a variety of operating conditions without the need for process additives. Satisfactory selectivities have been obtained with a wide variety of substrates and further product purification may be unnecessary.

The wide application of haloaromatic amines in the production of pesticides, herbicides, synthetic dyes, pharmaceutical preparations and other substances has resulted in considerable interest in their mode of preparation. The principal route to halogenated amines is from the nitro compound, either by reduction with iron-hydrochloric acid or by catalytic reduction with hydrogen over a noble metal catalyst. Until recently the preferred route in Europe, and to a certain extent in the United States of America, was the iron-hydrochloric acid reduction route. The reduction usually proceeds smoothly with a high yield of the haloamine, but requires that a steam distillation step be inserted into the preparation to separate the haloamine from the iron oxide sludge. The high energy consumption by this process and the environmental problems associated with disposing of the iron oxide sludge have led haloamine producers to seriously consider the catalytic reduction of halonitroaromatics. However, the catalytic reduction of the corresponding halogen-substituted aromatic nitro compounds is complicated because the conversion to the amines is usually accompanied by simultaneous dehalogenation which thus lowers the yield of the desired product and promotes the formation of corrosive halogen acids:

A number of different approaches have been made to overcome this problem, including the modification of catalyst behaviour by the addition of bases (1,2,3), solvents (4) or other process additives (5,6,7,8) to the reaction mixture. It can be concluded from the literature therefore that in a typical halonitroaromatic hydrogenation, dehalogenation is minimised by the selection of appropriate reaction temperature and pressure, and by operating the reaction system under non-acidic conditions so that if any halogen acids are formed they are immediately neutralised. A common disadvantage of some of the methods of avoiding dehalogenation is that they necessarily involve costly extra processing steps before, during or after a reaction. Recent work in our laboratory, reported here, has resulted in the development of a modified platinum on charcoal catalyst which is [i] sufficiently active for the hydrogenation of halonitroaromatics, [ii] does not promote dehalogenation and [iii] does not involve complicated operational procedures.

This new catalyst, designated Type 56, is now commercially available from Johnson Matthey, and the metal loading normally employed is of the order of 1 per cent of the support weight.

Reaction Conditions

For a typical halonitroaromatic hydrogenation to take place, in a non-solvent system, temperatures of approximately 100°C and pressures between 1140 and 1480kPa are required. Furthermore, if the reaction is to proceed at a reasonable rate, sufficient agitation of the reaction system is necessary to overcome gas-liquid mass transfer restrictions. In order to satisfy these requirements and operate under conditions which relate to commercial processes, two types of high pressure autoclave were employed:

  • A Baskerville mini-glandless autoclave in which the reaction mixture is stirred.

  • A modified Cook autoclave which provides agitation by rocking.

The products of the reaction form a two phase system consisting of an organic and an aqueous layer. However an ionic organic halide formed by dehalogenation tends to reside in the aqueous layer. Consequently, to obtain a true analysis of the reaction mixture, the dehalogenated product was shifted to the organic layer by addition of sodium bicarbonate. The organic layer was then separated and diluted with diethyl ether and the catalyst removed by filtration. The resulting solution was analysed by gas-liquid chromatography and the selectivity of a catalyst for the completed reaction was defined as the following percentage:

Effect of Agitation

The results given in Table I demonstrate that for both of the autoclaves employed, with obviously different agitation characteristics, the new platinum on charcoal catalyst gives selectivities greater than or equal to 99 per cent. In comparison, a conventional carbon supported platinum catalyst displays poor selectivity with both the Baskerville and the Cook autoclaves.

Table I

Comparison of the Hydrogenation of 2-Chloronitrobenzene in the Cook and Baskerville Autoclaves

Temperature: 90°C Pressure: 1140 kPa
Catalyst platinum/carbon Catalyst loading: weight per cent of substrate Reaction rate mol./min./g Pt Conversion per cent Selectivity to 2-chloroaniline per cent
Baskerville stirred autoclave, 1000 rpm
Conventional 0.7 0.9 100 66.0
Modified 0.5 1.26 100 99.5
Cook hydrogenator, 250 cycles per minute
Conventional 0.5 1.63 100 80.0
Modified 0.5 1.58 100 99.0

The results of a series of experiments using the Cook rocking autoclave, where the rate of agitation of the mixture was varied, are given in Table II.

Table II

Effect of Agitation Rate on the Hydrogenation of 2,5-Dichloronitrobenzene in the Cook Autoclave

Catalyst Loading: 0.5 weight per cent of Substrate Amount of Substrate: 0.5 moles

Temperature: 110°C Pressure: 1140 kPa
Agitation rate cycles per minute Reaction rate mol./min./g Pt Product distribution, per cent
2,5-DCA 0-CA m-CA A
200 1.19 99.66 0.08 0.21 0.04
250 2.56 99.20 0.18 0.35 0.27
350 2.78 99.16 0.19 0.53 0.12

Here, increased agitation is seen to increase the overall dehalogenation. Between 200 and 250 cycles per minute a rapid increase in the rate of hydrogenation occurs. A further increase in the rate of agitation has little effect on the rate, possibly indicating that the reaction is now under chemical kinetic control. This behaviour is similar to that observed previously by Acres and Cooper during the hydrogenation of nitrobenzene over charcoal-supported palladium metal catalysts (9).

Effect of Pressure

The effect on the selectivity of the new catalyst when operating the reaction at an elevated pressure of 2170kPa is illustrated in Table III. It is clear that increased dehalogenation occurs under such conditions, and therefore low pressures of the order of 1137kPa are recommended for use with this catalyst. However, it is worth noting that even at high pressures the catalyst remains superior to a conventional noble metal catalyst operating at the lower pressure normally employed.

Table III

Effect of Modified Catalyst on the Hydrogenation of 2-Chloronitrobenzene in the Baskerville autoclave

Catalyst Loading: 0.63 weight per cent of Substrate

Temperature: 90°C Conversion of Nitrobody 100 per cent
Catalyst platinum/carbon Pressure kPa Stirring rate rpm Reaction rate mol./min./g Pt Selectivity to 2-chloroaniline per cent
Conventional 1137 1000 1.00 66.0
Modified 1137 1000 1.01 99.5
Modified 2170 1000 2.52 98.0
Modified 2170 500 0.63 92.0

2.5-DCA 2.5-dichloroaniline

0-CA ortho-chloroaniline

m-CA meta-chloroaniline

A aniline

Catalyst Loading

The variation in selectivity obtained with different catalyst loadings is given in Table IV. It is apparent that dehalogenation increases as greater amounts of the catalyst are used, which may be a consequence of increasing the number of sites active for haloamine hydrogenolysis.

Table IV

Effect of Catalyst Loading on the Hydrogenation of 2,5-Dichloronitrobenzene in the Cook Autoclave

Temperature: 90°C Pressure: 1140 kPa Agitation Rate: 250 cpm
Catalyst loading: weight per cent of substrate Reaction rate 103 mol./min. Product distribution, per cent
2,5-DCA 2,5-DCNB o-CA m-CA A
0.25★ 2.23 50.89 49.08 0.03 — —
0.50 6.25 99.32 0 0.24 — 0.43
1.0 8.93 97.23 0 0.51 0.88 1.15
1.5 10.00 96.77 0 0.79 0.37 1.67
2.0 11.63 95.26 0 0.76 2.07 1.92

• This run was not taken to completion

2.5-DCA 2.5-dichloroaniline

2.5-DCNB 2,5-dichloronitrobenzene

o-CA ortho-chloroaniline

m-CA meta-chloroaniline

A aniline

A typical hydrogenation trace obtained with 2,5-dichloronitrobenzene is illustrated in Figure 1. The reaction proceeded continuously until the completion of region A. If no dehalogenation occurred no further hydrogen uptake was observed in region B. However, if catalyst loadings greater than 0.5 per cent were employed the hydrogen pressure drop did not abruptly cease at the completion of A, but continued after the reaction had entered B.

Fig. 1

During the hydrogenation of 0.5 moles of 2,5-dichloronitrobenzene hydrogen is stored in a reservoir. As the reaction proceeds, and the nitrobody is converted, the decline in hydrogen pressure within the reservoir is recorded. The reaction temperature is also simultaneously monitored and the end of a reaction is indicated by a sharp fall in temperature of about 10 to 15°C

Effect of Temperature

As the reaction temperature is increased, while all other reaction conditions remain constant, there is an increase in the amount of dehalogenation. The results obtained for the hydrogenation of 2,5-dichloronitrobenzene are presented in Figure 2. An apparent activation energy of 13.1kJ/mol was observed at temperatures above 110°C, which indicates that the reaction suffers from mass transfer restrictions. Below 110°C the effect of temperature on the rate of reaction is more pronounced and the apparent activation energy was 39.4kJ/mol.

Fig. 2

The hydrogenation of 0.5 moles of 2,5-dichloronitrobenzene as represented here by an Arrhenius plot. Using a catalyst loading of 0.5 per cent of its substrate weight, it is evident that at temperatures above 110°C (reciprocal temperature 2.68 × 10−3K−1) the reaction rate begins to suffer from mass-transfer restrictions thus causing the Arrhenius plot to curve

Alternative Substrates

So far only those results obtained with the modified platinum catalyst for the hydrogenation of both 2-chloronitrobenzene and 2,5-dichloronitrobenzene have been reported. However, the performance of the catalyst has also been assessed with a number of alternative haloaromatic substrates.

Chloro Substituted Compounds

The results obtained with several other chloro substituted compounds are summarised in Table V. The results show that the catalyst can hydrogenate a wide variety of substrates selectively to the desired products. It is likely that the selectivities quoted are lower than can actually be attained as the operating conditions were not optimised for each substrate.

Table V

Hydrogenation of a Variety of Chloro Substituted Compounds

Substrate Autoclave Catalyst loading: weight per cent of substrate Temp. °C Pressure kPa Agitation rate cpm Reaction rate mol./min./g Pt Conversion per cent Selectivity to desired products per cent
4-chloronitrobenzene Baskerville 0.63 90 1205 1,000 1.01 100 98.8
3,4-dichloronitrobenzene Baskerville 0.63 90 1205 1,000 1.65 100 97.2
2-chloro-6-nitrotoluene Cook 0.63 80 1140 350 1.68 99.78 97.5
2-chloro-4-nitrotoluene (Batch 1) Cook 1.88 80 1140 350 0.27 99.95 99.6
2-chloro-4-nitrotoluene (Batch 2) Cook 0.63 80 1140 350 1.85 99.98 98.6

Fluoro Substituted Compounds

Aromatic fluoro-amines offer great potential in the pharmaceutical industry. Since the strength of the carbon-fluorine bond is greater than that for the other halogens, the replacement of fluorine by hydrogen during a catalytic hydrogenation would not be expected to occur readily. However, hydrogenolyses of carbon-fluorine bonds have often been reported (10) and even fluorine species attached to an aromatic system can be replaced (11). Therefore, in view of their commercial importance, the performance of the modified catalyst with such compounds was evaluated.

The rate of hydrogenation of the fluoronitrobenzenes was lower than that observed with other halonitroaromatics and catalyst loadings of between 2 and 4 per cent were necessary with both 2- and 4-fluoronitrobenzene. However, 2,4-difluoronitrobenzene was hydrogenated rapidly with a loading of only 1 per cent at a reaction temperature of 110°C. Table VI describes the hydrogenation of 2- and 4-fluoronitrobenzene and also 2,4-difluoronitrobenzene. It is important to note that, as expected, very little dehalogenation took place.

Table VI

The Hydrogenation of Fluoronitroaromatics in the Cook Autoclave Amount of Substrate: 0.5 moles

Temperature: 110°C Agitation Rate: 250 cpm Pressure: 1140 kPa Conversion of Nitrobody 100 per cent
Substrate Catalyst loading: weight per cent of substrate Reaction rate mol./min./g Pt Product distribution per cent Selectivity to desired products per cent
Dehalogenated product Fluoroanilines
2-fluoronitrobenzene 4 0.59 0.21 99.79 99.79
4-fluoronitrobenzene 4 0.36 0.15 99.85 99.85
2,4-difluoronitrobenzene 1 2.50 0.16 99.84 99.84

Bromo Substituted Compounds

A brief examination was also made of the hydrogenation of 2-bromonitrobenzene on the Baskerville autoclave and a summary of the results obtained are presented in Table VII.

Table VII

Hydrogenation of 2-Bromonitrobenzene in the Baskerville Autoclave Catalyst Loading: 0.5 weight per cent of Substrate

Amount of Substrate: 1 mole

Temperature: 150°C

Pressure: 1137 kPa

Stirring Rate: 1000 rpm

Reaction rate mol./min./g Pt Selectivity to 2-bromoaniline per cent
0.95 97.6

The selectivity is lower with the bromo-substituted compound than with the corresponding chloro- or fluoro-substituted compounds, in agreement with the relative strengths of the various carbon-halogen bonds.

Discussion

A new Johnson Matthey Type 56 platinum on carbon catalyst has been shown to be excellent for the hydrogenation of halonitroaromatics. As expected, the selectivity of the catalyst seems to depend on the halogen substituent and generally decreases in the order fluorine > chlorine > bromine. There are several advantages of using the new catalyst compared with other patented processes. The disposal of process wastes is relatively simple and inexpensive compared to the hydrochloric acid reduction route. The catalyst seems capable of operating satisfactorily for a wide variety of conditions and although the results have referred to solvent free systems, the catalyst is also effective in the presence of solvents. This new catalyst is often found to be more active than conventional noble metal catalysts, thus permitting low catalyst: substrate ratios. Finally no process additives are necessary and since the catalyst is highly selective, further purification stages after hydrogenation may be avoided.

References

  1. 1
    W. Bohm, British Patent 1,483,330; 1977
  2. 2
    J. R. Kosak, U.S. Patent 3,546,297; 1970
  3. 3
    Y. Hirai and K. Miyata, U.S. Patent 4,070,401; 1978
  4. 4
    J. P. Luvisi, U.S. Patent 3,150,185; 1964
  5. 5
    L. Spiegler, U.S. Patent 3,073,865; 1963 and U.S. Patent 3,328,465; 1967
  6. 6
    R. J. Gait, British Patent 1,498,722; 1978
  7. 7
    J. R. Kosak, U.S. Patent 4,020,107; 1977
  8. 8
    J. D. Hildrith and D. G. Haslam, European Patent Appl. 0,000,805; 1978
  9. 9
    G. J. K. Acres and B. J. Cooper, J. Appl. Chem. Biotechnol., 1972, 22, ( 6 ), 769
  10. 10
    M. Hudlicky, J. Fluorine Chem. 1979, 14, ( 3 ), 189
  11. 11
    F. J. Mettille and D. J. Burton, Fluorine Chem. Rev., 1967, 1, ( 2 ), 315
 
Read more of this issue

Find an article

ArticleSearch