Catalytic Synthesis of Acetonitrile by Ammonolysis of Acetic Acid S.I

Eurasian ChemTech Journal 3 (2001) 173-178

Catalytic Synthesis of Acetonitrile by Ammonolysis of Acetic Acid S.I. Galanov*1,2, O.I. Sidorova*2, A.K.Golovko2, V.D. Philimonov3,

L.N. Kurina1,2, E.A. Rozhdestvenskiy1 1Tomsk State University, 36, Lenina Ave., 634050, Tomsk, Russia 2Institute of Petroleum Chemistry, 3, Academichesky Ave., 634021,Tomsk, Russia 3Tomsk Polytechnical University, 30, Lenina Ave., 634043, Tomsk, Russia

Abstract The influence of principal parameters (reagent ratio, reaction temperature, temperature gradients along a catalyst layer) on the yield of the desired product was studied in the reaction of acetonitrile synthesis from acetic acid over γ-alumina. Thus, the increase in ammonia:acetic acid ratio leads to the increase in acetonitrile selectivity and yield. In this work it has been demonstrated that initial temperatures of 360- 380°C are optimum to effectively carry out the process of acetonitrile synthesis. The increase in reaction temperature allows one to increase the yield of acetonitrile, but at elevated temperatures the catalyst carbidization and contamination of the desired product were observed. The additives to the reaction mixture of the substances that decrease the rate of compaction products (CP) formation and participate in the desired product formation are very effective for decreasing the catalyst carbidization. The effect of the composition of a reaction mixture on a catalyst lifetime is considered. The addition of ethyl acetate to acetic acid promotes a greater carbidization as compared to pure acetic acid. The application of a mixture of acetic acid with acetic anhydride at similar acetonitrile yield decreases the catalyst carbidization.

Introduction the temperatures of 440-480°C. But the reaction carried out at this temperatures, along with high energy Acetonitrile is widely used in organic synthesis as expenses for reactor heating, leads to thermal pyroly- an intermediate reagent, solvent, azeotropizer [1,2]. sis of acid (or products of synthesis) followed by Acetonitrile is commercially produced as a by-prod- catalyst carbidization and contamination of the de- uct of acrylonitrile synthesis. Therefore the investi- sired product. At elevated temperatures, when steel gation and development of inexpensive and commerci- reactors are used, hydrocyanic acid may be formed. ally convenient methods for acetonitrile production The reduction of the temperature of the synthetic pro- represent an urgent problem. Developed are the cess up to 350-380°C is promising because it is pos- methods for acetonitrile synthesis from ammonia and sible to use water steam to heat the reactor instead of acetic acid [3,4], alcohols [5], paraffins and olefins flue gases [4]. [1,6]. Nevertheless, the syntheses from alcohols [5], Thus, the optimization of acetonitrile synthesis paraffins and olefins [1,6] are characterized by an should be carried out by the following parameters: insufficient selectivity and low process efficiency. - Enhancement of acetonitrile efficiency per a unit Acetonitrile synthesis from acetic acid is more of catalyst volume; promising due to less expense for isolation and puri- - Reduction of the amount of intermediate products fication of the desired product. In the past acetoni- (ammonia acetate and acetamide) and by-products that trile synthesis from acetic acid and ammonia was determine the energy expenses for rectification and carried out in the tubular reactors with space velocity purification of acetonitrile; -1 of acid feed of 0.3 h [3] and 0.0076-0.06 mol/s·l - Reduction of the reactor temperature that deter- [4], acetonitrile yield above 80 % was observed at mines the energy expenses of the production and the *corresponding authors. E-mail: [email protected] period of the catalyst life between regenerations.

 2001 al-Farabi Kazakh St. Nat. University 174 Catalytic Synthesis of Acetonitrile by Ammonolysis of Acetic Acid

Experimental perature difference between the external reactor wall

and the middle of the layer was 0.5-1.0°C. NH3:

The study was made in a flow steel (12X18H10T) CH3COOH ratio was changed within 1.5-4.0, the reactor with an upward reagent flow at a preheating reaction temperature was 350-450°C, acetic acid load of the reaction mixture up to the reaction tempera- was 0.49-1.02 g/cm3·h. The reduced contact time (s) ture. To enable the temperature measurements in the was calculated as the relationship between the catalyst catalyst layer, the reactor was made with an internal volume and the space velocity of the gas-vapour mix- thermocouple pin with moving platinum - platino- ture at the reaction temperature. The catalyst was γ- 2 rhodium (PP) thermocouple passing through the Al2O3 (Sspec = 149 m /g, total pore volume is 0.68 3 catalyst layer; thermal EMF was measured by a volt- cm /g). The synthesized products (H2O, NH3, , meter. To reduce coaxial temperature gradients, the CH3COONH4, CH3CN, CH3CONH2, CH3COCH3 CH O and HCN) were analyzed by gas chromato- reactor was placed into a heated massive unit made 2 of steel [7], the unit temperature was regulated by a graphy on a column 3 mm in diameter, 2 m long, SDA high-precision temperature controller (BPT-3) with Separon as a sorbent, the detector was a catarometer, a PP thermocouple; the precision of the temperature carrier gas was helium. maintenance was 0.5°C. The initial temperature in the reactor (metal unit) was regulated by a BPT-3 Results and discussion temperature controller. During blank tests the temperature gradient was 6-8°Ñ along the 185-mm high The reaction of acetonitrile formation from ace- catalyst layer with the graining of 1-2 mm. The tem- tic acid and ammonia occurs at several stages [3, 4]:

I II III CH COOH + NH CH COONH 3 3 3 4 CH3CONH2 CH3CN + 2H2O (1)

4α 3 = − The reaction of acetamide formation is weakly kg/m ·s; q ( T c T ) is the heat amount input to ∆Í= 25.3 êJ/mol, the next stage III of d exothermal: – a unit of the reaction volume per a time unit in a acetonitrile formation is endothermal: ∆Í = 127.9 êJ/ reactor with the constant temperature of the metal mol [8]. Endothermal process is stable under all the unit Òñ; Ò is the temperature in the reactor; d is the circumstances (in distinction to exothermal one), that diameter of the pipe containing a catalyst; α is an is why at a constant heat exchange only one station- effective factor of heat transfer from the heat-trans- ary state takes place. Temperature drop leads to the fer agent to the catalyst. decrease in reaction rate that stops the further tem- The changes in the temperature profiles along the perature reduction, in the similar way the system catalyst layer shown in Figure and the data given in turns back into its original state at the temperature Table 1 allow determining an optimum acetic acid rise. That is why at the establishment of a station- load (contact time) at the temperature of 360°Ñ (Fig. ary reaction mode the temperature gradients (tem- 1 b, c). The conversion and selectivity reached the perature reduction) are observed along the catalyst maximum values when the temperature gradients in layer (see Fig. 1). The shape and distribution of tem- the final section of the catalyst layer are not observed perature gradients along the catalyst layer may be (Fig. 1, Table 1). This evidences a complete reaction one of the parameters to control the reactor opera- process. The reduction in the acid load (increase in tion. In this case the heat balance of the reactor is contact time) from 1.02 to 0.58-0.67 g/cm3·h at the described by the following equation: increase of acetonitrile yield (Table 1) leads to a de- dT crease in efficiency from 0.57 to 0.4-0.46 g of aceto- = − 3 Yf w hiri q (2) nitrile per cm of a catalyst·hour, in this case a side dz reaction of acetic acid decomposition is observed fol- where Yf is the heat capacity of the volume unit, J/ lowed by acetone formation. The acetone selectivity 3 3 m ·Ê; w is the linear rate of the flow, m/s; hi is the is 0.6 % at the acid load of 0.58 g/cm (Table 1). At 3 thermochemical factor, J/kg; ri is the reaction rate, the acid load of 1.02 g/cm ·h and the temperature of

Eurasian ChemTech Journal 3 (2001) 173-178 S.I. Galanov et al. 175

Flow direction Layer height, cm [3, 4], in this case water influences the equilibrium 04 9 1419shift both on the stage II and III. Based on the de-

5 4 pendence of temperature gradients (temperature de- 3 15 crease) along the catalyst layer for low ammonia:acid 2 = 350°Ñ (Fig 25 ratio of 1.5 and 2.0 and Tin . 1a) one 1 may see that due to the reversibility of the reaction 35 II the temperature gradient along the catalyst layer 45 is minimum, the yield of the desired product (aceto- 55 î Ò Ñ à nitrile) is low. On increase in partial ammonia pres- sure (Fig. 1a) the temperature gradient at the final Flow direction Layer height, cm section of the catalyst layer decreases, at the in- 0 4 91419 creased yield of the desired product (Table 1). The 5 8 7 increase in initial temperature in the reactor by 10°Ñ 15 6 provokes the increase in temperature gradients in the 5 25 middle of the catalyst layer (Fig. 1 a, b) at the in- 35 crease in acetonitrile selectivity and yield (Table 1). 45 This effect may be explained by the increase in the

55 î rate of the reaction III (Eq. (1)) and the shift of the Ò Ñ b equilibrium of the reaction II towards acetamide for- Flow direction Layer height, cm mation. On increase in acid load and ammonia:acid 04 91419 ratio (enhancement of the space velocity of reactants,

5 11 reduction in contact time), at a comparable acetoni- 9 12 trile yield (experiment ¹ 8 as compared to experi- 15 10 13 ments ¹ 9, 10, 11) a temperature drop by 10°Ñ is 25 observed in the final section of the catalyst layer, 35 connected with a large mass transfer and cooling of 45 the catalyst layer by the reagent flow. The rise in 55 î initial temperature of the reactor from 360 to 380°Ñ Ò Ñ c at the same acid load significantly accelerates the Fig. 1. Profiles of the temperature drop along the catalyst layer. a, b: the initial reactor temperatures are 350 and stage III (Diagram 1) and leads to the increase in the 360°C, respectively, acid load is 1.02 g/cm3·h; c: for the absolute value of the temperature gradient from curves 9, 10, 11 the initial reactor temperature is 360°C, ∆Ò=25-35°Ñ (Òlayer = 325°Ñ) to ∆Ò = 45-55°Ñ 3 acid loads are respectively 0.49, 0.58, 0.67 g/cm ·h; for (Òlayer = 325°Ñ). Correspondingly, (Fig. 1b, curves the curves 12 and 13 the initial reactor temperature is ¹ 5-8, and c, curves ¹ 12, 13), maximum tempera- 380°C, acid load is 1.02 g/cm3·h. ture drop occurs in the initial section of the layer. Temperature gradients are not observed in the final 350-380°Ñ acetone is not detected in the reaction prod- section. Thus, the increase in selectivity and yield of ucts. acetonitrile may be reached by increasing the reaction One of the parameters influencing the process of temperature (Table 1). acetic acid ammonolysis may be varying the partial In addition to energy expenses for reactor heating, pressures of the reactants. The study of the influence the temperature selection is conditioned by cracking of NH3:CH3COOH ratio on the process parameters processes: the contamination of the desired product demonstrated that at equal acid loads the increase in by the products of thermal pyrolysis and catalyst ratio leads to the increase in acetonitrile selectivity carbidization followed by the reduction in catalyst and yield despite of the reduction in contact time activity. Thus, due to catalyst carbidization, at T = (Table 1). This is connected to the reversibility of the 440°Ñ acetonitrile yield changed from 98.5 to 91.5 % reaction II (Eq.(1)), the increase in the partial pres- within 5 hours of operation. The reduction in reaction sure of ammonia shifts the equilibrium towards ac- temperature allows one to significantly decrease the etamide formation followed by acetonitrile formation amount of CP formed (Table 2) and to extend the (reaction III). The decrease in acetonitrile yield at the catalyst life between regenerations. The maximum dilution of acetic acid by water was noted elsewhere carbidization of catalyst in the initial section of the

Eurasian ChemTech Journal 3 (2001) 173-178 176 Catalytic Synthesis of Acetonitrile by Ammonolysis of Acetic Acid

Table 1 Influence of the reactant ratio, acetic acid load and temperature on process parameters (*- temperature gradients of experiments show in Figure, with corresponding numbers) *No of Q, NH :CH COOH T ,C° 3 3 τ , s KS, % ,S% ,B% , % expie r men t in gm/c 3·h riato rec CHCH3CONHCONH22 CHCH3CNCN CHCH3CNCN 10325 15.0 15. 15.7 888. 327. 612. 55. 20325 10.0 20. 11.5 942. 265. 774. 68. 30325 10.0 30. 10.1 992. 113. 826. 79. 40325 10.0 49. 02.8 982. 120. 829. 82. 50326 15.0 15. 16.7 848. 163. 876. 76. 60326 10.0 20. 16.5 829. 181. 858. 79. 70326 10.0 30. 19.1 962. 64. 983. 86. 80326 10.0 49. 00.8 190 31. 916. 96. 90396 00.4 23. 30.1 1-0 949. 399. 100 386 00.5 28. 22.6 9-9. 979. 498. 101 376 00.6 20. 22.3 9-9. 959. 598. 102 328 15.0 15. 15.7 959. 45. 905. 95. 103 328 10.0 25. 15. 909. 30. 907. 97. τ red is the reduced contact time at the reaction temperature, s Q is the acetic acid load, g/cm3·h K - acetic acid conversion, %

SCH3CONH2, SCH3CN is respectively acetamide and acetonitrile selectivities (mol %)

BCH3CN is acetonitrile yield (mol %) Table 2 Influence of the temperature and reaction mixture composition on the formation of compaction products (CP), changes 2 in specific surface and average catalyst efficiencies per hour of a sample (initial specific surface Sspec=149.0 m /g)

2 NTo in,m°C Rixtureseaction citinompo oP%SC spec, m /ag cta

104H00-46 C 3C4OOH 03.1 123 0.53

203H50-40 C 3C4OOH 01.0 154 0.57

303H50-40 95 wt.% C 3COOH + 5 wt.% (CH3CO)2O208.0 133 0.62

403H50-40 95 wt.% C 3COOH + 5 wt.% CH3COOC2H5 00.73 162 0.22 % CP is the percentage of the compaction products formed per catalyst weight per 1 cm3 of the acid input. 3 acat is the average catalyst efficiencies per hour, g of acetonitrile per cm of catalyst·h catalyst layer downstream of the reaction mixture flow products formed (their content is 8-10 % of catalyst was observed, the final section of the layer being weight), the rest consists, obviously, of oxygen and carbided the least and the last. Probably, it is connected nitrogen. One of the methods for decreasing the with the accumulation of water formed in the reaction catalyst carbidization is the introduction of water gases and with the decrease in the concentration of vapours to the reaction mixture, nevertheless, as noted acetic acid, the main CP source. Hydrogen-carbon in [3, 4], the application of diluted acetic acid results analysis of the carbided samples demonstrated that in the reduction in acetonitrile yield and catalyst effi- C:H ratio in compaction products is ~2. Hydrogen ciency. That is why justified is adding to the reaction and carbon make only 30-40 % of the compaction mixture of the substances that would decrease the rate

Eurasian ChemTech Journal 3 (2001) 173-178 S.I. Galanov et al. 177 of CP formation and participate in the reaction that When using an acid-anhydride mixture, an increase results in the formation of the desired product, e.g., in the conversion is observed but acetonitrile yield acetamide, ethylacetate or acetic anhydride. The ad- remains the same as at pure acetic acid application dition of ethyl acetate to acetic acid promotes a greater due to the decrease in acetonitrile selectivity (Tables carbidization (Table 2) as compared to pure acetic 1, 3). This is connected with the fact that at the intro- acid. The application of a mixture of acetic acid with duction of acetic anhydride into the reaction mixture, 5 wt % of acetic anhydride at similar acetonitrile yield along with the reactions shown in Eq.(1), the reactions decreases the catalyst carbidization (Tables 2, 3). of the acetic anhydride are added:

+NH3 CH CONH + CH COOH (CH3CO)2O + NH3 3 2 3 2CH3CONH2 + H2O 2CH3CN + 2H2O (3) I II III

Table 3 Influence of the ratio of reagents and temperature on the catalytic parameters of process. -1 (95 wt.% ÑÍ ÑÎÎÍ and 5 wt.% (ÑÍ ÑÎ) Î) Space velocity of mixture - 1.02 h 3 3 2

NTo in,H°C N 3:CH3C%OOH ratio KS, CHCH3CONHCONH22,S% CHCH3CNCN,B% C CHH3CNCN, % 10356 15. 922. 186. 853. 77. 20306 20. 965. 122. 887. 82. 30358 12. 918. 47. 995. 93. 40308 25. 979. 33. 986. 95.

Conclusions one to extend the installation operation without catalyst reloading, as the regeneration significantly Thus, to effectively carry out the process of ac- provokes the catalyst destruction. etonitrile synthesis, NH :CH COOH ratio of 2-3 and 3 3 τ τ τ initial temperatures of 360-380°Ñ are optimum (ap- acat = (acat· 1)/( 1+ 2) (4) plication of a mixture of acetic acid with acetic an- τ h dride). At the beginning the rise in the initial reactor where acat is the efficiency of an active catalyst; 1, y τ temperature to 440°Ñ allows one to improve the ac- 2 are respectively the duration of the periods of activity and activity recovery [9]. etonitrile efficiency by 30 % (as compared to the initial temperature of 380°Ñ) from 0.680 to 0.886 g of acetonitrile per cm3 of catalyst·h, then due to the References catalyst carbidization it decreases. Under these conditions the average efficiency is 0.806 g of aceto- 1. Paushkin Ya.M, Osipova L.V. Uspekhi Khi- nitrile per cm3 of catalyst·h. In this case the life be- mii.1959. Vol. XXVIII, n 3. p. 237-263. tween regenerations of a catalyst operated at Tin = 2. Pereushanu V., Korobya M., Muska G. Produc- 440°Ñ reduces, as compared to the time of operation tion and Application of Hydrocarbons.// Transl. at Tin = 380°Ñ. Regeneration time, taking into account from Romanian. V.G. Lipovich, G.L.Avrekh a smooth temperature rise and a stepwise increase in (Ed). Moscow: Khimiya. 1987. p. 288. oxygen content in oxygen-vapour mixtures intended 3. Kovalenko T.I., Kovalenko V.I., Malook G.P., to prevent catalyst overheating and destruction, is 2 Dubchenko V.N. Zhournal Prikladnoy Khimii. hours. The reduction of the reaction temperature and 1972. N 8. P. 1824-1827. addition of acetic anhydride to acetic acid enables to 4. Khcheyan Kh.E., Shatalova A.N., Nikitin A.K. increase average catalyst efficiencies per hour acat Improvement of the Processes of Main Organic (Table 2) calculated using Eq.(4) due to the absence Synthesis: in Book of Scientific Papers of of resinous compounds formed at a high-temperature VNIIOS . M., TSNIITEneftekhim, 1984. p. 24- pyrolysis of acid in the reaction products, that allows 31.

Eurasian ChemTech Journal 3 (2001) 173-178 178 Catalytic Synthesis of Acetonitrile by Ammonolysis of Acetic Acid

5. Vodolazhskiy S.V., Yakushkin M.I., Devekki 8. Rabinovich V.A., Khavich Z.Ya. Brief Chemical A.V., Khvorov A.P. Zhournal Organicheskoy Handbook. Moscow, Khimiya. 1978. 392 p. Khimii. 1988. Vol. 24 N 4. P. 699. 9. Khaylov V.S., Brandt B.B. Introduction in the 6. USSR Patent N 738657, MPC B 01 J 23/24. Technology of Main Organic Synthesis. Moscow, Aliyev S.M., Sokolovskiy V.D., Boreskov G.K. Khimiya. 1969. 560 p. 7. Ioffe I.I., Pismen L.M. Engineering Chemistry of Heterogeneous Catalyst. Moscow, Khimiya. 1965. 246 p. Received 31 July 2001.

Eurasian ChemTech Journal 3 (2001) 173-178