The acrylonitrile process

Citation for published version (APA):van der Baan, H. (1980). The acrylonitrile process. In R. Prins, & G. C. A. Schuit (Eds.), Chemistry and chemicalengineering of catalytic processes : NATO Advanced Study Institute, 1979, Noordwijkerhout, The Netherlands:proceedings (pp. 523-533). (NATO ASI Series, Series E: Applied Sciences; Vol. 39). Alphen aan den Rijn:Sijthoff & Noordhoff.

Document status and date:Published: 01/01/1980

Document Version:Publisher’s PDF, also known as Version of Record (includes final page, issue and volume numbers)

Please check the document version of this publication:

• A submitted manuscript is the version of the article upon submission and before peer-review. There can beimportant differences between the submitted version and the official published version of record. Peopleinterested in the research are advised to contact the author for the final version of the publication, or visit theDOI to the publisher's website.• The final author version and the galley proof are versions of the publication after peer review.• The final published version features the final layout of the paper including the volume, issue and pagenumbers.Link to publication

General rightsCopyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright ownersand it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights.

• Users may download and print one copy of any publication from the public portal for the purpose of private study or research. • You may not further distribute the material or use it for any profit-making activity or commercial gain • You may freely distribute the URL identifying the publication in the public portal.

If the publication is distributed under the terms of Article 25fa of the Dutch Copyright Act, indicated by the “Taverne” license above, pleasefollow below link for the End User Agreement:www.tue.nl/taverne

Take down policyIf you believe that this document breaches copyright please contact us at:openaccess@tue.nlproviding details and we will investigate your claim.

Download date: 14. Apr. 2020

523

THE ACRYLONITRILE PROCESS

H.S. van der Baan

Laboratory for Chemical Technology, Eindhoven University of Technology, The Netherlands

1. The reaction model

A major event in the history of oxidation catalysis was the dis­covery of bismuth molybdate (Idol, 1959) (Bi 203 + Mo03) as a selective catalyst for the partial oxidation of-propene and also, in a one step operation, for the ammoxidation of propene. Other catalysts are oxides of uranium and antimony or of iron and antimony. The latter catalysts may be promoted by potassium or phosphor. The latest development are the bismuth molybdate multi­component catalysts that contain also one or more of the elements cobalt, iron, nickel, potassium and phosphor. The catalysts that have been studied most are the bismuth molybdates. These are active within the composition range Bi2Mo06 to Bi2Mo30 12 •

Under normal reaction conditions i.e. at temperatures above 675 K the catalysts behave according to the Mars-van Krevelen model: the oxygen from the catalyst is used in the oxidation or ammoxi­dation reaction, after which the reduced catalyst is reoxidized. This follows conclusively from the work of Keulks (1970), Wragg et al. (1971) and others, who have shown that in reactions with

160 in the catalyst and 1802 in the gasphase it takes some time

before 180 is found in the reaction products, acrolein and water. The stoichiometric reaction equation for the catalytic ammoxida­,tion of propene:

( 1)

S24 H.S. van der Baan

shows that the propene molecule must loose three hydrogen atoms, and that the dehydrogenated intermediates must be bonded to the catalyst in such a way that they are protected against direct oxi~ation by gaseous oxygen. A model of the whole reaction sequence is given by Lankhuijzen (1979) :

2 3 4 5 t (C 3H6)a ~ (C 3HS)a ~ (C3H4)a t (C3H40)a t

6 + ~

7 +

8 t +

Fig. 1. Reaction scheme for the formation of acrolein and acry­lonitrile.

~ (!) N

cO .,..., .., ~ .... ''';

Reaction step no. ::l 0 () ..c: ::l ::l .., ,..0 0 b() ..c: 00 OJ ::l () '''; .;& .... ,..0 00 cO (!) ~ cO cO .... OJ .-< cO

::>:: ::r:: 0 ~ [f) H

1 H a stract~on {st + b . 2 + C

3HS formation

Mo Bi Bi Mo Mo04

Ho

{nd + . 3 2. H abst:act~on C3H4 format~on on.

Bi Mo Bi Mo Mo04 Mo

4{formationof with ° from-

C3H4O Bi Mo Bi Mo Ho04

Mo

Table 1. Location of active ensembles in the oxidation df propene; Matsuura and Schuit (1970)(1971), Haber (1973), Otsubo et al. (197S), Peacock et al. (1969), Sleight (in press), Lankhuijzen (1979)

The acrylonitrile process

Much attention has been devoted to the question on what active surface ensembles the reaction steps of figure I take place. A number of the results are given in table 1.

5Z5

As far as the oxygen is concerned most authors expect it to enter the catalyst at an ensemble connected with Bi. Of the four main reactions that take place:

C3H6 + NH3 + H 0z -r C3H3N + 3 HZO (I)

C3H6 + °2 -r C3H4O + HZO (2)

C3H4O + NH3 + ! °2 -r C3H3N + 2 HZO (3)

2 NH3 + I! 0z -r NZ + 3 H2O (4)

the reaction orders are shown in table 2.

Reaction order with respect to C3

H6 NH3 C3H

4O °z

* (I) 1-0 0 0 * (2) 1-0 0

** (3) 0 1-0 0 ** (4) 0 0

Table 2. Reaction orders in the ammoxidation of propylene. * Depending on temperature and reactant concentration

(Lankhuij zen 1979). ** actually: slightly negative.

The zero order in oxygen indicates complete coverage of the re­oxidation ensembles, i.e. the oxygen supply to the reaction sites is not influenced by variations in the gas phase oxygen concen­tration. The zero order in ammonia for the acrylonitrile formation (both from propene and from acrolein) ind{cates that th~ reactin.g·· nitrogen surface species have completely covered the sites avail­able to them. As the nitrogen formation is first order in ammonia there also must exist a nitrogen containing species on other sites, where the degree of coverage is rather low. Lankhuijzen assumes that the latter sites are the same as those used by the intermediates derived from propene thus causing the slightly negative order in ammonia for the acrylonitrile formation.

526 H.S. van der Baan

Insert instead:

The competition between ammonia and propene for the same sites is also found in the somewhat negative order exerted by propene on the nitrogen formation, as shown in table 2. That on the other hand, also ammonia and oxygen have common features follows from Lankhuij zen! s 'observation that in pulse experiments where a pro­pene-helium mixture is pulsed over the catalyst the conversion per pulse is increased in exactly the same way by addition of oxygen as by addition of the same quantity of ammonia. In this connection we recall the observation of Sancier 5t al who found that in pulse experiments thy8production of 0 1 acro­lein was enhanced by the addition of 02 • This all suggests that propene and "first order" ammonia absorb on one type of site, related to Mo and that oxygen and "zero order" ammonia are connected with Bi.

The acrylonitrile concentration has a small but noticeable nega­tive influence on the reaction rate, i.e. on the propene conver­sion. The main byproducts of the reaction are H

20, CO2/CO, CH

3CHO,

N2

, CH3

CN and HCN. The latter two have commercial value and are

generally recovered from the product stream.

2. The acrylonitrile process

For the choice of the most appropriate reactor the following data are of interest: -Most of the reactants and products are flammable and have ex­plosive properties over wide ranges when mixed with air, as shown in table 3.

Ignition temperature Explosive range K % by vol. In air

C3H6 770 2 -II

NH3 924 15 -'28

C3H4

O 551 2.8-31

C3H3N 754 3.1-17

Table 3. Ignition temperature and explosive range for the main reaction components of the acrylonitrile synthesis (Sax, 1975).

-The reaction produces about 20 MJ/kg acrylonitrile. -The catalyst is stable and can be used for years without much

The acrylonitrile process

products

propene

ammonia

transport air

off gas

r:eoxidized catalyst I

reduced catalyst

air for oxidation

527

Fig. 2. Two reactor design for ammoxidation of propene. Complete separation of air and propene plus ammonia.

loss of activity, if the temperature is kept within reasonable limits, say below 800 K.

-Acrylonitrile is not completely stable under reaction conditions.

The high heat of reaction requires a reactor with an efficient cooling system, especially as the combination of the inflammable reaction mixture and hot spots would lead to dangerous situations. Although initially fixed bed tubular reactors of the teat ex­changer type were used, fluid bed processes developed at an early stage. As fluid bed reactors are very efficient in -equalizing temperature differences, the possibility of hot spots is elimi­nated in such reactors, provided that they are properly designed i. e. that no quanTities-of stagnant catalyst can build up in corners or on internal beams. A fluid bed requires a catalyst that is mechanically stronger than the pure oxides, Admixture of a Si02 carrier, however, decreases the propene selectivity somewhat,

Initially these Lluid beds were also designed with the object of separating the oxygen containing stream from the combustible feed components. This is e.g. shown in figures 2 and 3. The sys­tem of figure 2 has a separate reoxidation reactor, while in the

528

propene

anunonia

transport

air

air for oxidation

conversion section

oxidizing section

transport air exit

H.S. van der Baan

Fig. 3. One reactor design for ammoxidation of propene. The air is to a large extent separated from the propene plus ammonia (Calahan, 1969).

liquid propene

liquid anunonia

glycol/water evaporators

to refregeration

HP steam

products

air blower

boiler feed water

Fig. 4. One t"eactor de..sign RiJ:hout &e.p_a.ra:tiun-o_f a-i-r andp.:w-pene plus ammonia (Pujado et al., 1977).

The acrylonitrile process 529

reactor of figure 3 the reoxidation takes place in the lower part of the fluid bed, so that air with a much reduced oxygen content is contacted with the propene-ammonia mixture higher up in the· bed. Also stearn has been added to the feed to reduce the explosiveness of the mixture of reactants. In the modern reactors however air, propene and ammonia enter the reactor together as sho~ in fLgure 4, in accordance with the experience gained in other oxidation reactors, e.g. the fluidized bed oxidation of a-xylene to phtalic anhydride. To withdraw the considerable heat of reaction, the reactors are all equiped with internal coolers, often arranged in the form of horizontally spiralized tubes, located at regular intervals along the height of the fluid bed. This form of the coolers has the added advantage that the back mixing in the dense phase of the fluid bed is somewhat reduced. This has a benificiary effect on. the acrylonitrile yield as the further conversion of acrylonitrile is somewhat reduced.

3. A mathematical model of the conversion in the fluid bed.

3.1. Of the models available, such as those of van Deemter (1961), Kunii and Levenspiel (1965), Partridge and Rowe (1969) or of Fryer and Potter (1972) we will use the model of Partridge and Rowe of 1969, because it is simple and contains two assumptions that reflect properties of a fluid bed with a number of horizon­tal cooling grids, viz: 1. The gas flow is plug flow upward through the dense phase.

Because the grids suppress backmixing the backmixing models of van Deemter and Fryer and Potter will be less appropriate;

2. The volume fraction of bubbles is constant over the bed height. The bubbles are fully segregated and do not coalesce. The grids are acting as redistribution trays, on which on all trays bubbles are formed with the same diameter. The mean bubble diameter can be set at 1.5 to 2 times the distance be­tween the elements of the grid, say db ~ 0.06-0.1 m. This leads to a linear bubble rise velocity of 0.5-0.7 mls (see equation R (4)*).

Further assumptions of the Partridge and Rowe model are: 3. The (empty) bubbles are s-u-r-rounded by a "cloud" and a "wakeJL--­

moving with the bubble (see R figures 12 and 6 respectively). The whole ensemble is called the cloud. The outer part is called the dense phase-cloud phase overlap or shortly the overlap. Thus:

cloud = bubble + overlap (5)

* This notation indicates equation (4) of the chapter of P.N. Rowe on basic fluidization.

530 B.S. van der Baan

4. The gas in the whole cloud is ideally mixed; 5. There is only resistance to mass transport between the dense

phase and the overlap; 6. The volume fraction of the bubble phase 0 (volume of bubbles

per unit volume of reactor) is constant over the bed height; 7. The linear gas velocity in the dense phase is equal to the

minimum fluidization velocity; 8. The reactions take place in the dense phase and in the overlap

phase. In these two phases the catalyst fraction is the same.

3.2. The mathematical expressions for the Partridge and Rowe fluid bed model.

We will indic~te2linear velocities (m/s) by V and superficial velocities (m 1m s) by V •

s

3.2. I. The ratio of cloud volume to bubble volume.

We define a:

a =

~n which Vah linear velocity of the bubbles and Villf velocity t min. fluidization or

a V s,b mf

V s,mf

From equation 5 we have

v c

Partridge and Rowe derived:

or w{th (8)

linear

(6)

(7)

(8)

(9)

v = V a + 0.17 c b a - I

(10)

3.2.2. The gas transport v~a the dense phase and the cloud phase.

For the superficial feed velocity V we can write s,f

V s,f V d + V s, s,c ( 11 )

The acrylonitrile process

Further

Thus

v s,d

v = V' - vmf (1-0) S,c s,f

531

(12)

(13)

3.2.3. The overall mass transport coefficient between dense phase and overlap k follows from

o

k d o c

Sh = -D-

The Sherwood number follows from the correlation:

Sh = 2 + 0.69 Re! Sc l/3

with

Re

and

Sc

which is valid for 30 < Re < 2000.

(14)

( 15)

(16)

(17)

For the relative bubble velocity V Partridge and Rowe have used r

(18)

Vb a~d Vmf can be found with figure R 9 and equation R (2) res­pect~vely.

From equation (10) we have

d,~" a,V":cOi 17 . (J9)_

This allows the computation of k . The effective transport co­efficient is E k , because a fragtion (1 -E) of the exchange surface is occup~ed by catalyst.

3.2.4. The reaction rate r can be expressed as

r = k Cpropene = k p ~P (molls kg cat) (20)

This gives for the rate ~n the dense phase:

532 H.S. van der Baan

3 rd = r Pc (1 - s) mol/s m dense phase (21 )

where Pc = the catalyst particle density and in the cloud phase:

Vb 3 rc = r ~c (1 - s) V- (mol/s m cloud phase) (22)

c

3.2.5. The differential equations.

As we can assume that the process is isothermal, and as we neglect the pressure drop over the bed no further data are required. The reactor can then be modelled according to figure 5.and from this figure we can derive for the dense phase:

L

x + d x -c -c

z + dz I

- -----1

o

I.........".... z -----....,1

x -c

mass transport

(x - xc) - r =-d - d

r----

1----­~

o (23)

Fig. 5. Model of a fluid bed reactor for the ammoxidation of propene. x Represents the specific propene concentration (mol/kg) .-

and for the cloud phase:

v S,c

d x s k TI d 2 -c 0 c

-d-z- + --"""V,-----c

o (24)

which can be solved by straight numerical integration from the initial values x = xd = x , the propene inlet concentration. With a const~nt ~e~ectivi~y o~68 percent the acrylonitrile con­centration is everywhere:

The acrylonitrile process 533

xA = 0.68 (x - x) - -0

(25)

and for the outlet concentration of acrylonitrile we have finally

x = -A,e

Literature

v x + V x s,d-A,d,e. s,c -A,c,e

Vsf

I. Calahan, J.L., Nilberger, E.C., USP 3, 472, 892 (1969) 2. Haber, J., Z. Chem. 13,241 (1973) 3. Idol, J.D. (Standard-oil Co.) U.S. Pat. 2 904 580 (1959) 4. Keulks, G.W., J. Catal. 19, 281 (1970) 5. Kunii, D., Levenspiel, O~ Ind. Eng. Chem. Fund. 7, 446 (1968) 6. Lankhuijzen, S.P., The acrolein and acrylonitrile-synthesis

over a bismuth molybdate catalyst, Thesis, Eindhoven 1979 7. Levenspiel, 0., 'Chemical Reaction Engineering' 2nd ed.,

John Wiley and Sons Inc., New York (1972) 8. Matsuura, I., Schuit, G.C.A., J. Catal. 20, 19 (1971) and

J. Catal. 25, 314 (1972) --9. Murray, J.~, J. Fluid. Mech. 21, 465 (1965) and 22, 57 (1965)

10. Otsubo, T., Miura, H., Morikaw~ Y., Shirasaki, T., J. Catal. 36, 240 (1975)

I I. Partridge, B.A., Ronn, D.N., Trans. Inst. Chem. Engrs. 44, 335-348 (1969)

12. Peacock, J.M., Sharp, M.J., Parker, A.J., Ashmore, Ph., Hockey, J.A., J. Catal. 15, 379; 398 (1969)

13. Pujado, P.R., Vora, B.V.-,-Krueding, A.P., Hydrocarbon Pro­cessing ,(5), 169 (1977)

14. Sancier, K.M., Wentreck, P.R., Wise, H., J. Catal. 39, 141 (1975) --

15. Sax, N.I., 'Dangerous properties of Industrial materials", 4th ed. v. Nortrand Reinhold CY, New York (1975)

16. Schuit, G.C.A. in B.C. Gates, J.R. Katzer and G.C.A. Schuit: 'Chemistry of Catalytic Processes'; McGraw Hill, New York, chapter 4, p. 325 (1979)

17. Sleight, A.W., in J.J. Benton and R.L. Garten (eds,), 'Ad­vanced materials in Catalysis', Acad. Press, New York (in press)· .. - _. ..- -

18. van Deemter, J.J .• Chem. Eng. Sci. 13, 143 (1961) 19. Wragg, R.D., Ashmore, P.G., Hockey,-Y.A., J. Catal. ~, 49

(1971)