CATAI,YTIC PREPARATION OF ORTHO-TOI,UIDINE BY C. 0. HENKE AND 0.W. BROWN
Introduction I n several previous papers we have given the results of some studies on the catalytic reduction of nitrobenzene to ani1ine.l I n this paper we shall give the results of some studies on the reduction of ortho-nitro-toluene to ortho-toluidine. The chemical reactions involved are very similar and one would expect the catalytic preparation of the two amines t o be likewise very similar. Although similar in many ways the catalytic preparation of the ortho- toluidine has brought out many interesting differences from the catalytic preparation of aniline. Experimental Details The apparatus used was the same as that used in our previous work on aniline I n all the experiments in this paper the catalyst tube was a ‘/zinch iron pipe. The orthonitro-toluene used was purified by shaking with sodium carbonate solution, distilling with steam and then subjecting it to two simple distillations. The method of work was the same as in our previous papers. The 2 cc pipette, under the conditions under which it was used, delivered 2.243 grams of ortho-nitro-toluene. The analysis of the product was likewise the same except that, after washing the product into the graduated flask, 25 cc concentrated HC1 were added before diluting to the mark in order t o make sure of the complete solution of the ortho-toluidine. Nickel as Catalyst The nickel nitrate used was “Baker’s Analyzed” c.p. material. According to the maker it contained 0.75y0 Co, 0.002% Fe, 0.001% C1 and 0.002% SOs. The nitrate after 1 Brown
and Henke: Jour. Phys. Chem., 26,161,272, 715 (1922).
Catalytic Preparation of Ortho- Toluidine
53
the addition of nitric acid was ignited in a small porcelain evaporating dish in an electrically heated muffle. The temperature was raised to 450" C. Seventeen grams of the resulting oxide were put in the furnace and reduced and heated in hydrogen to 384" C. In the first experiment which was carried out a t 190 " C, with the hydrogen flow a t 14 liters per hour and the ortho-nitro-toluene flow a t 14.5 grams per hour the material yield of ortho-toluidine was 68.4y0 of theory. This low yield was not due to incomplete reduction but to the fact that the reduction was carried too far. The product was colorless and the odor of ammonia was pronounced. After using this catalyst in five experiments with differ'ent rates of flow of hydrogen the seventh experiment, which was carried out under the same conditions as the first experiment, gave an 85.3% material yield of ortho-toluidine. The increase in material yield of ortho-toluidine from 68.4 to 85.3y0 was not due to an increase in activity of the catalyst but rather to a decrease in its activity. The product in all experiments was colorless and ammonia was present as shown by the odor. The nickel was too active and reduced a part of the nitrobenzene too far yielding cyclo compounds and ammonia. After this initial decrease in "activity, resulting in an increase in material yield
TABLE I Catalyst-nickel Temperature of catalyst-190' C Rate of flow of ortho-nitro-toluene-13.5 Hydrogen in liters per hour
*5 9.5 14 38 47 63
Excess of hydrogen
-24' 50 110 335 A10
8.50
grams per hour Material yield of orthotoluidine in yoof theory
79.0 83.5 85.3 52.3 78.3 72.9
In this experiment the rate of flow of hydrogen was 24y0below the amount theoretically required.
C. 0.Henke and 0.W . Brown
a54
Ortho-nitro-toluene in gms per hour
27
13.5 4
Excess of hydrogen in %
10 110 A00
Material yield of ortho' of theory toluidine in %
86:s 85.3 84.6
From the results of Table I1 i t will be seen that within the limits studied the lower the rates of flow of ortho-nitrbtoluene the lower the yield of ortho-toluidine. However the difference in material yield within these limits is small. At the slow rates the product was colorless, while at the highest rate (27 grams per hour) the product was slightly colored indicating that a part of the ortho-nitro-toluene was incompletely reduced. I n our first paper on the catalytic preparation of aniline we have shown that the oxidation and reduction of the nickel catalyst decreased its activity. An attempt was made to increase the yield of ortho-toluidine by decreasing to a certain
Catalytic Preparation of 0 )thQ-Toluidize
55
extent the activity of the nickel catalyst so that it would not carry the reduction too far. The nickel was oxidized in a current of air and heated to 500" C. It was then reduced and heated in hydrogen to 438" C. The average of four experiments carried out at 190" C, with a hydrogen flow of 14 liters per hour and the flow of ortho-nitro-toluene a t 13.5 grams per hour, was 86.5y0 compared with 85.3% when used under the same conditions previous to the oxidation and reduction. The catalyst was oxidized again, heated in a current of air to 515 O C and reduced and heated in hydrogen t o 450" C. Two experiments carried out under the same conditions as after the previous oxidation and reduction gave 86.1% and 85.8% yields of ortho-toluidine, which are in good agreement with the previous experiments. The product, after oxidation and reduction of the catalyst, was slightly colored. I n our previous work we secured over 95% material yields of aniline. These experiments on ortho-toluidine were carried out at the same temperature where the highest yields of aniline were secured. Also the catalyst was prepared under conditions which would give the best yields of aniline. However the yields of ortho-toluidine with the best rates of flow of hydrogen and ortho-nitro-toluene, are about 8% lower than the best yields of aniline secured. The low yields of orthotoluidine are due to the reduction being carried too far, the product being but slightly colored only in a few experiments. This would seem to indicate that ortho-nitro-toluene is more easily reduced than nitrobenzene and that the reduction of the ortho-nitro-toluene is not so easily stopped a t the amine stage. Lead as Catalyst A lead catalyst was prepared by reducing 20 grams of orange mineral (orange mineral is a red lead having an apparent density of 7.8 and made from sublimed litharge). During the reduction the temperature did not rise above 315" C. The results obtained with this catalyst are given in Table 111. The experiments of Table I11 are given in the order in which they were carried out. The first three are under the
C. 0.Henke and 0.W . Brown
56
-Temperature of catalyst degrees C
310 310
310 327 327
TABLE 111 Catalyst-lead from orange mineral
1
Hydrogen in
1
Ortho-nitrotoluene in gms pel;lt;;'osur per hour
14 . 14 14 7 14
1
Excess of hydrogen in%
3.s 3.8
640
3.8 3.0 3.8
640 300 640
I
640
Material yield of ortho-toluidine in % of theory
80.1 00.2 93.2 93.2 91.0
same conditions. The results show that the catalyst increases in activity with use. I n the third experiment the yield of ortho-toluidine is 93.2y0. A catalyst prepared from this same red lead was found in a previous paper1 to give a 92.7y0 yield of aniline when used under practically the same conditiohs. Thus the yield of the ortho-toluidine is slightly higher than the yield of aniline. When the temperature was increased t o 327" C the yield decreased to 91% in the fifth experiment, indicating that about 310" C is the best temperature for the reduction with this catalyst. Another catalyst was prepared from an amorphous red lead having an apparent density of 23.0. Forty-nine grams of this red lead were put in the furnace and reduced a t 310" C for 11/4hours. It was then used at 310" C with nitrobenzene until the yields became constant. With 14 litres hydrogen and 3.8 grams nitrobenzene per hour the yield of aniline at 310" C was 93.4%. When ortho-nitro-toluene was used instead of nitrobenzene with the same rates of flow and the same temperature the yield of ortho-toluidine with this same catalyst was 94.60j0. With this catalyst as with the previous lead catalyst the ortho-toluidine yield is slightly higher than the aniline yield indicating that ortho-nitro-toluene is more easily reduced than nitrobenzene. In our previous paper we found that (with an iron catalyst tube) a catalyst from an amorphous red lead, of an apparent density of 16.5, was the best of the 1
Henke and Brown: Jour. Phys. Chem., 26, 325 (1922).
Catalytic Preparation of Ortho- Toluidine
57
lead catalysts for producing aniline, over 96% material yields of aniline being secured. We would expect a catalyst from this red lead to give still higher yields of ortho-toluidine, but unfortunately we had no more of this material and could not t r y it. In our previous paper a catalyst prepared from an amorphous white lead gave very low yields of aniline when used in a n iron tube, the yield decreasing after a few experiments to 59%. However in a glass tube this catalyst gave over 94% total material yields of azobenzene and aniline. A catalyst was prepared from 24 grams of this same amorphous white lead which had been prepared by electrolysis as described in ,our previous paper. The white lead was reduced and heated in hydrogen to 327" C. Experiments carried out with this catalyst at 327" C and with the hydrogen flow at 14 litres per hour and the ortho-nitro-toluene flow a t 3.8 grams per hour gave yields of but 65 to 50%. This catalyst then, when used in an iron catalyst tube, is as poor a catalyst for the preparation of ortho-toluidine as for the preparation of aniline. A mixture of 1 gram Fe203and 40 grams of the amorphous red lead when reduced and used as a catalyst gave practically same yields of ortho-toluidine as the amorphous red lead alone.
Discussion of Results with Lead and Nickel Catalysts Low material yields were secured with both of these catalysts, the highest yields being secured with lead. The nickel however can be used at a very much greater rate than the lead. The reason for the low yields is not the same in the two cases. Nickel is too active and carries the reduction too far while the lead although it gives a higher yield than the nickel, produces azotoluene along with the toluidine, the azotoluene formation cutting down the yield of the toluidine. The azotoluene is only slightly volatile and condensed for the most part in the end of the catalyst tube where it solidified to a glassy red solid. I n our studies on the catalytic preparation of aniline
58
C. 0. Henke and 0. W . Brown
we found that nickel carried the reduction of a part of the nitrobenzene too far while the lead would reduce a part of the nitro, . benzene only to the azobenzene stage. However the nickel did not reduce as much of the nitrobenzene too far as i t did of the ortho-nitro-toluene. Thus with nickel over 95% yields of aniline were secured while only about 87y0 yields of orthotoluidine were secured. The ortho-nitro-toluene is more easily reduced than the nitrobenzene. This same thing is indicated by lead, the ortho-toluidine yields being slightly higher than the aniline yields with the same catalyst.
Silver as Catalyst Silver falls between nickel and lead in that i t does not carry the reduction past the amine stage and likewise does not produce azo compounds. It then should be an excellent catalyst for the reduction of the ortho-nitro-toluene t o orthotoluidine. The silver catalyst was prepared from the carbonate. The carbonate was prepared by dissolving 34 grams AgN03 in 200 cc water and precipitating with 140 cc of ammonium carbonate solution containing 14 grams of the solid salt. The precipitated silver carbonate was washed, filtered and then dried a t about 100' C. Twenty-two grams of the dried carbonate were put in the furnace and reduced and heated in hydrogen to 312' C. It was then used in the experiments listed in Table IV. The temperature used was the best temperature for the production of aniline. The experiments of Table IV are listed in the order in which they were carried out. The yield at first is excellent, being almost quantitative. The product also had a good appearance being practically colorless. The yield however, after use of the catalyst at higher rates of ortho-nitro-toluene, in the last experiment, in which the rate of ortho-nitro-toluene was only a little higher than in the first, decreased to 76.770. Another silver catalyst was then prepared like the catalyst used in Table IV. The results with this catalyst are given in Table V.
Catalyiic Preparation of Ortho- Toluidine
Ortho-nitro-toluene in grams per hour
Material yield of orthotoluidine in yo of theory
Excess of hydrogen in %
3.4 3.4 4.5 4.5 12 8 5.4 3.8
59
720 720 540 540 130 260 440
99.3 99.3 98.9 97.7 79.0 78.6 74.8 76.7
A40
TABLE V Catalyst-silver Rate of flow of hydrogen-14 Experiment number
litres per hour
Temperature Ortho-nitro- Excess of Material yield of of catalyst toluene hydrogen ortho-toluidine degrees C gmsperhour in % in Yoof theory
--
38d5 to 44d5 inc. 45d.5 . 46d5 47d5 48d5 49d5 50d5 51d5 li2d5 53d5 fi4d5 55d5 5Gd5 57d5 58d5 69d5
300 300 300 300 300 300 300 300 320 320 320 320 320 320 340 330
~
*
4.5 4.8 3.8 3.2 4.5 3.2 4.8 6.4 7.5 13.4 4 .5 3 . 2. 3.2 3.2
'
640 830 540 480 640 780 .540 780 480 350 300 110 540 780 780 780
98.5 98.5 97.0 94.7 95.1 97.0 97.4 97.0 97.0 95.9 91.7 75.6 90.2 92.5 91.0 85.0
60
C. 0. Henke und 0. W . Brown
Experiments 38d5 to 44d5 inclusive were all carried out under duplicate conditions and gave a yield of 98,5y0. Likewise the eighth experiment gave a yield of 98.5y0. Thus during these first eight experiments the activity of the silver catalyst remained fairly constant. Referring back t o Table IV we see that the eighth experiment which was carried out undernearly the same conditions, gave on1y.a 76.7y0 yield of ortho-toluidine. The only difference between the first eight experiments of Table V and the first eight of Table IV is that in some of the first eight experiments of Table IV, a higher rate of ortho-nitro-toluene was used than in any of the first eight of Table V. This indicated that a high rate of flow of orthonitro-toluene tends to decrease the activity of the silver catalyst. This same behavior is indicated by experiments 46d5 to 51d5. Experiments 46d5 and 47d5 were carried out a t slightly higher rates of ortho-nitro-toluene than experiments 38d5 to 45d5. After use a t these slightly higher rates the yield did not reach 98y0 at 300" C. Upon increasing the temperature to 320 O Ca 98.1y0yield was secured in experiment 52d5. However, after being used at the higher rates a t 320 O C the yield decreased as is shown by experiment 57d5 in which the yield was 92.5%. An increase in temperature to 340" C in experiment 58d5 was not beneficial. This higher temperature seemed to cause a still further decrease in activity for in experiment 59d5 the yield was only 85% compared to 92.5y0 in experiment 57d5 which was carried out under the same conditions just prior to use at 340" C. Several attempts were made to secure a silver catalyst in such a condition that its activity would not decrease with use, when used a t the higher rates. Powdered AL03 mixed with Ag2C03 was not so good a catalyst as Ag2C03 alone. Likewise a mixture of CaC03 and Ag2C03made by precipitating the two together from a solution of the mixed nitrates by Na2C03was no better. Asbestos was soaked in a silver nitrate solution and dropped into a solution of sodium carbonate, but gave poorer results than the silver carbonate alone. Silver on pumice was much worse.
Catalytic Preparatim of Ortho- Toluidine
61
All the attempts a t increasing the life of the silver catalyst gave negative results. As a result, although nearly theoretical yields of ortho-toluidine can be secured with silver as catalyst, the process would not be feasible due to the short life of the silver catalyst.
Copper as Catalyst The catalytic activity of copper is very similar to that of silver. Copper produces very little if any azobenzene and likewise reduces none or practically none of the nitrobenzene farther than.the aniline stage. Copper then should, like silver, give high yields of ortho-toluidine. In our previous papers we have given the results of many experiments with copper as catalyst and have pointed out many things affecting its activity. For producing aniline, copper reduced from the precipitated oxide is much better than that from the oxide obtained by ignition of the nitrate. I n Table VI are given the results of some experiments with a copper catalyst prepared by reducing the precipitated oxide. The catalyst is the same catalyst that was used in our study of its activity in the reduction of nitrobenzene, the results of which have been given in a previous paper.l It had been prepared by precipitation with NaOH from the pure nitrate. The pure nitrate had been prepared by refining copper twice by electrolysis as described in our previous paper. Before being used in the experiments listed in the table it had been used in 40 experiments under various conditions with nitrobenzene. One of these conditions (higher temperatures) is detrimental to the activity of the catalyst. As a result, its activity is a little less than it would have been had it not been used a t the higher temperatures (300' C). The experiments are listed in the table in the order in which they were carried out. I n three of the experiments listed nitrobenzene was used instead of ortho-nitro-toluene so that the aniline yield may be compared with the ortho-toluidine yield.
-
1
Brown and Henke: Jour. Phys. Chem., 26, 715 (1922).
C. 0. Henke and 0. W . Brown
62
TABLE VI from precipitated oxide Temperature of catalyst-260' C Catalyst-copper
Nitro compound in gram: per hour
Hydrogen in litres per hour
0.G
3.5 3.5 7 7 14 7 7 7 14 3.5
5.8 5.8 5.8 5.4 8.7 9.0 10.2 5.8 5.8
,
Excess of hydrogen in %
Material yield of amine in .yoof theory
- 31
85.6 aniline 96.4 ortho-toluidine 97.2 ortho-toluidine 97.5 ortho-toluidine 95.6 aniline 92.6 aniline 94.1 ortho-toluidine 79.0 ortho-toluidine 94.3 ortho-toluidine 96.7 ortho-toluidine
20 140 140 380 60
50 40 390 20
1 In this experiment the rate of flow of hydrogen was 373 less than the theoretical amount needed.
The results of the second and tenth experiments are in good agreement and show that the activity of the catalyst did not change appreciably during this period of time. Oncomparing the tenth, second, third, fourth and ninth experirn'hts it will be seen that the lower rates of hydrogen are better than the higher rate, 7 litres per hour being best, as shown by the third and fourth experiments. The third, fourth, seventh and eighth experiments show that the lower the rate of ortho-nitrotoluene the higher the yield of ortho-toluidine. Comparing the ortho-toluidine yields with the aniline yields it will be noted that with the rate of hydrogen a t 14 liters per hour in the fifth and ninth experiments the aniline yield was higher than the ortho-toluidine yield. However in all the other experiments (under nearly comparative conditions) the orthotoluidine yield was higher than the aniline yield. From this one would expect that with a copper catalyst, whose activity had not been decreased by use at too high a temperature, the ortho-toluidine yield would a t least be equal to if not higher than the aniline yield. The yields of aniline under different conditions have been given in the paper preferred to above.
Catalytic Preparation of Ortho- Toluidine
Hydrogen in litres per hour
Excess of hydrogen in %
Material yield of amine in yo of theory
14 14 3.5 7 14 14
640 580 85 270 640 ,580
97.2 ortho-toluidine 90.6 aniline 98.3 ortho-toluidine 97.9 ortho-toluidine 96.4 ortho-toluidine 91.9 aniline
63
Upon examining the results of Table VI1 it will be noted that the yields of ortho-toluidine are much higher than the yields of aniline. The yields of aniline are low, only 90.6% and 91.9% while the yields of ortho-toluidine are high, even up to 98.3%. The fact that the catalyst for the ignited nitrate gives low aniline yields and good ortho-toluidine yields, the difference between the yields secured and the theoretical not being due t o intermediate products or t o the carrying of the reduction too far, indicates that the ortho-nitro-toluene is more easily reduced than the nitrobenzene. Nickel gave low yields of ortho-toluidine, not because of incomplete reduction, but
64
C. 0. Henke and 0. W . Brown
because it carried the reduction too far, so that not only is the ortho-nitro-toluene more easily reduced to the amine stage but also it is more easily reduced past the amine stage than the nitrobenzene. Likewise with lead as catalyst slightly higher yields of ortho-toluidine than of aniline are secured with the same lead catalyst.
Summary 1. Nickel is too active and reduces a large part of the ortho-nitro-toluene farther than the toluidine stage. Attempts t o decrease the amount reduced too far were for the most part unsuccessful. The highest yield secured with nickel as catalyst was 86.8Oj,. 2. Lead catalysts give slightly higher yields of orthotoluidine than of aniline. Thus a lead catalyst from an amorphous red lead gave a 94.6y0 yield of ortho-toluidine while the aniline yield with the same catalyst under the same conditions was only 93.4y0. 3. Silver, when first used, gives 99% material yields of ortho-toluidine, but its activity decreases with use especially a t the higher rates of ortho-nitro-toluene. Attempts to prepare a silver catalyst with such properties that its activity would not decrease were unsuccessful. 4. Copper gives higher yields of ortho-toluidine than nickel over 97% yields being secured with a catalyst whose activity had been decreased by use a t too high a temperature. A copper catalyst prepared from an ignited nitrate gives higher ortho-toluidine yields than aniline yields. The activity of the copper catalyst does not decrease with use, as a silver catalyst does. For producing ortho-toluidine copper is superior to nickel because a much higher yield of the toluidine can be obtained. 5 . Ortho-nitro-toluene is more easily reduced than nitrobenzene. Laboratory of Physical Chemistry Indiana University Bloomington