Jackson S.D., Hargreaves J.S.J. Metal Oxide Catalysis

Подождите немного. Документ загружается.

20.5 Ammoxidation of Unconventional Molecules 801

The elements claimed are the same as those reported in References [120, 121] .

The preparation procedure employed is known to lead to the formation of

VOPO

, rather than (VO)

. The presence of Sb, however, may lead to a modi-

ﬁ cation of the structural features. Indeed, the authors claim the presence of

vanadyl pyrophosphate as the major phase present in catalysts, with a minor

amount of vanadium phosphate. The atomic ratio between the components of the

γ - alumina - supported active phase was V/Sb/P 1/1.9/1.18. The reaction conditions

were 425 ° C (at which the best yields were reported), and a feed ratio of reactant/

air/ammonia of 0.6 – 1.0/4.2/1.5. The following results were claimed under these

conditions:

• from cyclohexanol, conversion 78%, selectivity to adiponitrile 75%, selectivity to

hexanenitrile 18%. The by - products were CO, CO

and cracking products,

• from cyclohexanone, conversion 60%, selectivity to adiponitrile 48%, selectivity

to hexanenitrile 30%. The by - products were CO, CO

and cracking products.

This performance is outstanding, but it has never been successfully reproduced

by other teams active in this research ﬁ eld.

The reaction of cyclohexanol ammoxidation has been investigated by Chen and

Lee, with a V

catalyst [125] . The feed mixture comprised cyclohexanol/oxygen/

ammonia in the ratio 1.2/9/15. A large excess of ammonia was thus used. Cyclo-

hexanone, adiponitrile, adipic acid, benzene and CO

formed. The best yield to

adiponitrile was 4%, with a conversion of approximately 52% (as inferred from

the sum of the yields), at W / F = 0.45 g s cm

− 3

and 365 ° C. At a higher W / F value,

the maximum yield to cyclohexanone was reached (19%). All these products under-

went consecutive reaction to benzene and CO

. In a pulse reactor, the yield to

adiponitrile was 6.3%, at 96% reactant conversion. Cyclohexanone was also used

in the pulse reactor, but the predominant product was CO

. The authors proposed

that the reaction pathway includes the dehydrogenation of cyclohexanol to cyclo-

hexanone, its oxidative cleavage to adipic acid and subsequent transformation to

the dinitrile. The formation of benzene occurs via dehydration to cyclohexene,

followed by dehydrogenation. The ammoxidation of cyclohexene exclusively yielded

benzene.

The ammoxidation of cyclohexanol or cyclohexanone is also reported in one

patent [126] , using a cyclohexanone/air/ammonia/H

O ratio of 1/10/2.5/10 (the

presence of added steam is noteworthy), 450 ° C, and a contact time of 1 s. Several

catalysts were used, as reported in Table 20.6 . The products obtained were aniline,

phenol and adiponitrile.

The nature of the products formed is rather unusual. The formation of aniline

implies a reaction between the ketone and ammonia to yield the cyclohexanonei-

mine. The latter then rearranges with aromatization, to yield aniline. Aromatiza-

tion occurs with cyclohexanone, leading to the formation of phenol. Therefore, the

formation of the aromatic ring is quicker than the opening of the aliphatic cycle.

This is the key point of the reaction involving cyclic reactants: the competition

between the parallel reactions of ring opening and aromatization controls the

selectivity.

802 20 Catalytic Ammoxidation of Hydrocarbons on Mixed Oxides

20.5.3

The Ammoxidation of Cyclohexane and n - Hexane

Some papers and patents report on the ammoxidation of cyclohexane in the gas

phase [124, 127 – 129] or in the liquid phase [130] . In Reference [124] , which was

discussed in the previous section, the ammoxidation of cyclohexane is reported to

give 50% adiponitrile selectivity and 35% hexanenitrile selectivity for a conversion

close to 70%, at 425 ° C.

Osipova and coworkers [127] used a catalyst based on Ti/Sb mixed oxide, cal-

cined at 750 ° C. Under these conditions, titania is in the rutile form, and can host

up to 7 mol% Sb

in solid solution. Higher Sb content led to the development

of also an equimolar Ti/Sb compound (TiSbO

), while in systems with more than

50 mol% Sb

, the additional formation of α - Sb

was found. A catalyst with the

composition 70% Sb

– 30%TiO

gave the best performance.

Catalytic tests were run in a pulse reactor, at 400 ° C, with a cyclohexane/

oxygen/ammonia feed composition in mol% of 3/6/4 (the balance being He).

The main products obtained were adiponitrile ( ADN ) and benzene, with an

overall selectivity of more than 90% (the cyclohexane conversion is not reported).

The rates of benzene and ADN formation are plotted in Figure 20.10 as functions

of the Sb

content of catalysts. It is shown that the overall formation of benzene

considerably decreased on increasing the amount of Sb in catalysts. The forma-

tion of ADN decreased, but the decrease was less pronounced than that of

Table 20.6 A summary of catalyst performance reported in [126] for

cyclohexanone ammoxidation.

Catalyst Comment Yield aniline

(%)

Yield phenol

(%)

Yield ADN

(%)

K/Ni/Co/Fe/Bi/P/

Mo/O - silica

Typical catalyst for propene

ammoxidation

3.1 11.6 tr

V/W/Mo/O - silica Typical catalyst for

electrophilic oxidation

10.2 19.2 0.6

P/Mo/O - silica Heteropolycompound 8.9 10.0 –

W/O - silica 5.4 7.8 –

Fe/V/Sb/O - silica

Catalyst for propane and

propene ammoxidation

0.1 18.9 0.3

Sb/Mo/O Catalyst for allylic

(amm)oxidation

– 3.4 –

ADN: adiponitrile

20.5 Ammoxidation of Unconventional Molecules 803

benzene. Moreover, the formation of benzene was much greater than that of

ADN for low Sb contents, while the two rates became comparable for intermedi-

ate and large values of Sb content; therefore, the selectivity to ADN increased on

increasing the Sb content in catalyst.

Osipova and coworkers [127] also found that under non - steady conditions (those

utilized in the pulse reactor), a considerable fraction of ADN remained adsorbed

on the catalyst. A rough estimation for the catalyst with 30% Sb

led to the con-

clusion that the overall amount of ADN produced in the single pulse approximately

corresponded to 1/3 of the reacted cyclohexane. In particular, ADN bound strongly

to the catalyst surface and made heavy (polymeric compounds), at least at 300 ° C

and in the presence of ammonia. Adsorbed ammonia contributed to the polymer-

ization of ADN at the catalyst surface.

Simon and Germain [128] investigated several catalytic systems and found that

cyclohexane conversion is high (70%) even in the absence of catalyst at 460 ° C (feed

cyclohexane/oxygen/ammonia 1/3.6/1.5) yielding mainly cyclohexene, benzene

and CO

, with minor amounts of other compounds.

The presence of a catalyst led to the formation of C

dinitriles (maleonitrile,

fumaronitrile, succinonitrile), C

dinitriles (glutaronitrile) and C

dinitriles (muco-

nonitrile, adiponitrile), but the yield of these compounds was very low. In the best

case, with a V/Mo/O catalyst (atomic ratio V/Mo 4/1; phase: V

), the yield to

maleonitrile was 1.9% and 0.8% to fumaronitrile, 17% to benzene, 23% to CO

with traces of mucononitrile, at a conversion of 57% at 460 ° C. With the same cata-

lyst, the initial selectivity (extrapolated at zero conversion) to C

nitriles was approx

5% (negligible to other nitriles), while the predominant primary products were

benzene and carbon oxides. For temperatures lower than 420 ° C the predominant

product was cyclohexene, while at higher temperatures benzene and CO

prevailed

(Figure 20.11 ).

Figure 20.10 Rates of adiponitrile (ADN) ( 䉬 ) and benzene ( 䊏 )

formation as functions of the content of Sb

in catalysts

for cyclohexane ammoxidation. Elaborated from [127a] .

804 20 Catalytic Ammoxidation of Hydrocarbons on Mixed Oxides

The performance was very similar with a Ti/Mo/O catalyst (atomic ratio Ti/Mo

1.14/1; phases TiO

and MoO

) and with a Bi/Mo/O catalyst (atomic ratio Bi/Mo

0.9/1.) Other catalysts (Sn/Mo/O, Sb/Mo/O, Sn/Sb/Fe/O) were more selective

than V/Mo/O to either benzene or CO

The mechanism of reaction included the rapid oxy - dehydrogenation of cyclohex-

ane to cyclohexene and benzene. The latter was then transformed to successive

products (see the section on benzene ammoxidation). The opening of the ring may

occur either for cyclohexene or for benzene itself.

Table 20.7 summarizes the performance of catalysts described in a patent [129a]

which reports the formation of saturated dinitriles (C

, C

and C

). All catalysts

described contain Sb oxide as the main component. In References [129b, 129c] ,

the same catalysts are described as those reported in Reference [129a] , but with

the addition of a halogenated compound in the feed stream.

The addition of halogenated compounds increased the conversion of cyclohex-

ane, while the selectivities remained substantially unchanged. This indicates the

presence of a radical - based mechanism for the activation of cyclohexane. In fact,

it is known that the radical chain reaction is enhanced in the presence of halogens,

owing to their tendency to give rise to X • species.

The only example reported in the literature concerning n - hexane ammoxidation

is, once again, Reference [124] . In this paper, for a conversion of approx 12%,

a selectivity to adiponitrile close to 40% and to hexanenitrile of approx 30%

are reported, at 425 ° C and with a molar feed ratio hexane/air/ammonia of

0.6 – 1.0/4.2/1.5. To our knowledge, these interesting results have never been

conﬁ rmed.

Figure 20.11 Effect of temperature on catalytic performance

of V/Mo/O catalyst in cyclohexane ammoxidation [128] .

20.5 Ammoxidation of Unconventional Molecules 805

20.5.4

The Ammoxidation of Benzene

The ammoxidation of benzene is described in some papers and patents [124, 128,

131, 132] . Whilst benzene has been reported to remain unconverted in the pres-

ence of ammonia and oxygen with a V

catalyst [124] , it was efﬁ ciently trans-

formed to nitriles with catalysts based on mixed molybdates [128] .

Figure 20.12 reports the performance of the best catalyst, based on V/Mo/O

(V/Mo atomic ratio 4/1) at 465 ° C, with a feed composition of benzene/oxygen/

ammonia/nitrogen 1/4.4/1.6/17.6 (molar ratios). Among the various systems

investigated, those giving the best performance were based on V/Mo/O and Ti/

Mo/O which gave quite similar performance. Other systems were active but not

selective to nitriles (i.e. Sn/Mo/O), or were neither active nor selective (i.e. Bi/Mo/

O, Sb/Mo/O, Sn/Sb/Fe/O). The C

unsaturated dinitriles (mucononitrile) included

the three isomers: cis - cis dicyanobutadiene, cis - trans dicyanobutadiene and trans -

trans dicyanobutadiene.

The data apparently indicate that the C

dinitrile was a primary product, and the

overall selectivity at low benzene conversion (less than 7%) was around 20%; the

selectivity to mucononitriles (the three isomers) rapidly declined when the benzene

conversion was increased above 7 – 8%, and ﬁ nally became nil. The C

unsaturated

dinitriles, maleonitrile and fumaronitrile, reached a maximum selectivity of around

20% and 10% respectively, at conversions lower than 10%. It is not clear whether

they were primary or secondary products, because of the errors made in yield and

selectivity calculation for low benzene conversions (and also errors made in data

extrapolation from ﬁ gures). It is worth noting that the C balance was much lower

than 100%. Therefore, it is likely that some products were not detected.

Table 20.7 A summary of catalytic performance reported in [129a] for

cyclohexane ammoxidation.

Catalyst

T ( ° C), τ (s),

feed

Cyclohexane

conversion (%)

Selectivity

to ADN (%)

Selectivity

to GN (%)

Selectivity

to SN (%)

Sb/Sn/O 2/1 450, 1.95,

19.3 23.7 14.2 3.5

“ 550, 1.95,

27.2 14.1 10.6 2.2

“ 450, 3.9,

21.4 22.9 13.7 3.9

Sb/Sn/O 4/1 445, 2.0,

12.4 18.7 18.3 5.1

Sb/Sn/O 1/2 470, 2.0,

14.3 14.1 15.7 1.3

Sb/Sn/O 3/1 430, 1.5,

19.5 21.7 12.4 5.1

Sb/U/O 2/1 435, 2.0,

21.3 24.3 10.3 4.0

Sb/Ti/O 2/1 440, 2.0,

24.2 19.7 13.5 5.9

a feed composition (mol%) cyclohexane/oxygen/ammonia 5/10/6.6 (balance N

b feed composition (mol%) cyclohexane/oxygen/ammonia 5/10/10 (balance N

c feed composition (mol%) cyclohexane/oxygen/ammonia 8.4/10/14 (balance N

d feed composition (mol%) cyclohexane/oxygen/ammonia 7/14.7/14 (balance N

GN: glutaronitrile; SN: succinonitrile; ADN: adiponitrile.

806 20 Catalytic Ammoxidation of Hydrocarbons on Mixed Oxides

The C

dinitriles were more stable than the C

dinitriles. Although their selectiv-

ity decreased with increasing benzene conversion, they were still present at high

benzene conversion. CO

formed mainly by decomposition of the nitriles, but a

primary contribution (benzene combustion) cannot be excluded. The authors also

mention a slight increase of the selectivity to C

dinitriles on increasing the reac-

tion temperature (giving 5% more when raising from 440 to 500 ° C).

The authors made a hypothesis about the mechanism. They assumed that

benzene undergoes H • abstraction to generate a mono - radical which then reacts

with an NH

fragment to generate aniline (however, the latter is not found among

the reaction products) and di - radicals. The ortho di - radical is transformed to o -

quinone diimine, the para to p - quinone diimine. The opening of the ring in the

former case generated the C

dinitriles, and in the latter case maleonitrile + 2CO

The meta di - radical is supposed to give CO

. In the case of benzene oxidation, the

same authors reported that only p - quinone and maleic anhydride were found as

the products (oxidation of the para di - radical), while o - quinone and muconic acid

(oxidation of the ortho di - radical) were not found.

Unger [131] investigated several catalysts for the ammoxidation of benzene

(Table 20.8 ). The report shows that the performance of catalysts was not much

different from that reported in Reference [128] . However, in the only example

where the conversion was reported, the selectivity to mucononitrile was 17% (for

a yield of 7%), at 40% conversion. In this case, therefore, the compound can be

saved from consecutive degradation. It is likely that one key factor in saving muco-

nonitrile from consecutive degradation is the reactor conﬁ guration and catalytic

bed arrangement. This is also clearly demonstrated in Reference [128] .

Reference [132] also describes catalysts and methods for the ammoxidation of

benzene to mucononitrile. The performance seems to be greatly affected by the

type of reactor used and the reactor conditions.

Figure 20.12 Catalytic performance of V/Mo/O catalyst in benzene ammoxidation [128] .

20.5 Ammoxidation of Unconventional Molecules 807

Catalysts based on Fe/Sb/O, Sn/Sb/O and U/Sb/O were fairly active and selec-

tive to C

unsaturated dinitriles, but non - selective to mucononitrile. Catalysts made

of Bi/Mo/P/O, alumina - supported Mo/O and alumina - supported W/O had low

activities.

20.5.5

Ammoxidation of C

Hydrocarbons

Although industrial interest in the synthesis of acetonitrile directly from C

hydro-

carbons is currently limited, with acetonitrile being mainly produced as a by -

product in acrylonitrile production, there are a number of indications regarding

the future need of direct production of acetonitrile by C

hydrocarbon (ethane, in

particular) ammoxidation. In fact, acetonitrile is used as a solvent and also as an

intermediate in the production of many chemicals, ranging from pesticides to

perfumes. Production trends for acetonitrile generally follow those of acrylonitrile,

but the growth rate for acetonitrile use is higher than that of acrylonitrile. The four

Table 20.8 Summary of results reported in [131] for benzene ammoxidation.

Catalyst Benzene

conversion

(%)

T ( ° C), τ (s),

feed comp

Selectivity to

mucononitrile

(%)

Selectivity

to maleic

anhydride

(%)

Selectivity to

maleonitrile

(%)

Bi/Mo/P/O - BPO

15 – 20 450, 1.4,

8 0

1 8 0

Co/Mo/O - SiC ng 450, 1.4,

9 5

5 0

Co/Mo/O - Al

ng 450, 1.5,

7 5

2 0 0

V/P/O - Al

ng 450, 1.5,

7 0

(mainly

cis - cis)

30 0

V/P/O - Al

ng 450, – ,

7 8

1 2 1 0

Bi/Mo/P/O - BPO

ng 450, – ,

9 9

0 0

V/Mo/O - BPO

40 450, – ,

1 7

n g n g

a Feed: benzene/oxygen/ammonia/nitrogen 30/120/70/200.

b Feed: benzene/oxygen/ammonia/nitrogen 30/60/60/250.

c Feed: benzene/oxygen/ammonia/nitrogen 30/70/70/200.

d Feed: benzene/oxygen/ammonia/nitrogen 15/90/110/425.

e Feed: benzene/oxygen/ammonia/nitrogen 17/60/80/450.

f Feed: benzene/oxygen/ammonia/nitrogen 15/50/50/200.

g the selectivity is expressed only in reference to the condensable (high - boiling) products.

h this is a true selectivity.

ng = not given.

808 20 Catalytic Ammoxidation of Hydrocarbons on Mixed Oxides

main producers of acetonitrile in the United States are: Ineos, DuPont, J.T. Baker

Chemical and Sterling Chemicals.

Acetonitrile can be produced by catalytic ammoxidation of ethane and propane

over Nb - Sb mixed oxides supported on alumina, with selectivities to acetonitrile

of about 50 – 55% at alkane conversions of around 30% [133] . In both cases, CO

forms in approximately a 1 : 1 molar ratio with acetonitrile, owing to a parallel

reaction from a common intermediate. When feeding n - butane, the selectivity to

acetonitrile halves. Bondareva and coworkers [134] also studied ethane ammoxida-

tion over similar types of catalyst (V/Mo/Nb/O).

A different type of catalyst is constituted by Co ion - exchanged zeolites and

mesoporous materials (MCM - 49). Co - ZSM - 5 was found to be selective in ethane

ammoxidation [135] . A good correlation between the acidity and the catalytic activ-

ity was observed. The strength of ammonia bonding to the catalyst appears to have

a crucial effect on the activity of Co - ZSM - 5. Li and Armor [136] reported that

dealuminated zeolite was active for the ammoxidation of ethane to acetonitrile.

Pan and coworkers [137] instead studied ion - exchanged Co - Na - MCM - 49 and Co -

H - MCM - 49 materials for the same reaction, reporting that the presence of

ammonia in the feed considerably improved the selectivity and total yield of eth-

ylene and acetonitrile.

It should be mentioned ﬁ nally that acetonitrile could be also prepared by the

catalytic ammoxidation of bioethanol over vanadium - alumino - phosphate ( VAPO )

catalysts [138] , which is an alternative starting from biomass - derived raw

materials.

20.5.6

Conclusions on the Ammoxidation of Unconventional Molecules

Analysis of the literature published in the ﬁ eld of catalytic, gas - phase ammo-

xidation of ‘ unconventional ’ molecules allows the following conclusions to be

drawn:

1. The mechanism of catalytic ammoxidation generally accepted in literature for

activated methyl groups (allylic position in oleﬁ ns, or side position in

alkylaromatics) includes the generation of a

−CH

radical as the ﬁ rst, rate -

determining step. The data analyzed in the present report conﬁ rm that the

generation of a radical species is also the key step for the ammoxidation of

other molecules, such as butenes, cyclohexane, cyclohexanol and n - hexane. The

ﬁ rst step is less evident in the case of butadiene or benzene (and also in the

case of ethylene ammoxidation to acetonitrile). From benzene, the formation

of the Ph · radical is hypothesized in literature. By analogy, the formation of a

–

CH · species can be postulated from butadiene. Catalysts able to perform

this type of attack are a function of the type of molecule used, but in general

V oxide is one key component of the catalysts.

2. Once the radical species has been generated, this is followed by π - allyl complex

formation in the case of substrates in which the double bond is in the α - position

with respect to the radical. Insertion of N into the substrate generally occurs

20.5 Ammoxidation of Unconventional Molecules 809

by insertion of the NH

2 −

species from an M

NH species (isoelectronic with

the M

O species); M ions known to generate this moiety are Mo

and Sb

However, it seems that the mechanism may indeed include the insertion of an

species. The formation of the latter does not require the presence of metal

ions able to generate the M

NH species. This may explain why with some

molecules investigated, the performance of catalysts that do not include either

Mo or Sb is at least comparable (and often better) than that of catalysts including

these elements. So, the key - point is the generation of radical species that can

either react to generate the nitrile precursors or react to yield the by - products.

It is worth mentioning that the same hypothesis of a radical mechanism and

reaction with NH

has also been made for ammoxidation of alkylaromatics

[97c] .

3. An alternative mechanism may include ﬁ rst an O - insertion step, followed

by the transformation of the oxidized compound into the cyano - containing

compound. An example might be the oxidation of butadiene to maleic anhydride

followed by hydrolysis to the acid, the formation of the diamide and the

oxydehydrogenation to the dinitrile.

4. There are several contradictions in literature concerning the nature of the

products obtained. Table 20.9 tries to summarize the literature information,

with an indication of the best catalyst reported and of the best yield to nitriles.

The main discrepancy concerns the nature of nitriles obtained. In general, it

is possible to infer that starting from more reactive C

compounds (e.g.

cyclohexane), which may also undergo a relevant number of transformations,

Table 20.9 A summary of literature data on ‘ unconventional ’ ammoxidation reactions.

Reactant Major products Best catalyst Best

yield

(%)

Butadiene maleonitrile + fumaronitrile (crotonitrile) V/W/Cr/P/O - TiO

6 7

n - Butane maleonitrile + fumaronitrile (crotonitrile) V/W/Cr/P/O - TiO

2 6

Cyclohexanol adiponitrile + hexanenitrile (aniline, phenol) V/Sb/P/O - Al

7 2

Cyclohexanone adiponitrile + hexanenitrile(aniline, phenol) V/Sb/P/O - Al

4 7

Cyclohexane adiponitrile + glutaronitrile + succinonitrile

+ maleonitrile + fumaronitrile (benzene,

cyclohexene)

Ti/Sb/O 10

n - Hexane adiponitrile + hexanenitrile V/Sb/P/O - Al

Benzene mucononitrile + maleonitrile + fumaronitrile V/Mo/O 9

a Reference [124] is very controversial.

810 20 Catalytic Ammoxidation of Hydrocarbons on Mixed Oxides

the formation of both saturated and unsaturated nitriles (mono and di) with

the number of C atoms ranging from 4 to 6 is possible. This is a further

indication of the radical nature of the mechanism of the reaction. No author

has reported the formation of C

, C

or C

nitriles (cyanhydric acid, acetonitrile

and acrylonitrile), but we believe that this is highly likely. When the reaction is

instead carried out over benzene, the number of nitriles that can form is

lower.

20.6

Use of Other Oxidants for Ammoxidation Reactions

There is very limited literature data on the possible use of other oxidants in ammo-

xidation reactions, because air (O

) is the preferable source from an economic

point of view. However, it may be interesting to cite direct propane ammoxidation

with N

O and O

over steam - activated Fe - silicalite zeolite [139] . Yields of acryloni-

trile and acetonitrile below 5% were obtained using N

O or O

as the oxidant. Co -

feeding N

O and O

boosts the performance of Fe - silicalite compared to the

individual oxidants, leading to acetonitrile yields of 14% and acrylonitrile yields of

11% (propane conversions of 40% and product selectivities of 25 – 30%). The ben-

eﬁ cial effect of O

on the propane ammoxidation with N

O contrasts with other

O - mediated selective oxidations over iron - containing zeolites (e.g. hydroxylation

of benzene and oxidative dehydrogenation of propane), where a small amount of

in the feed dramatically reduces the selectivity to the desired product. It is

shown that the productivity of acrylonitrile, and especially acetonitrile, expressed

as mol(product) h

− 1

kg(cat)

− 1

, is signiﬁ cantly higher over Fe - silicalite than over

active propane ammoxidation catalysts reported in the literature.

20.7

Conclusions

Catalytic vapor - phase ammoxidation on mixed oxides is an important class of

industrial processes. Propene ammoxidation to acrylonitrile is a well established

process for the synthesis of this widely used monomer and intermediate. Over the

40 years since its commercial introduction, the yield to acrylonitrile has nearly

doubled to over 80% with the fourth generation of catalysts. This is due to the

intensive research effort and understanding of the several factors underpinning

catalytic activity. Commercial catalysts contain over 20 elements, the presence of

all of which is necessary to optimize the catalytic behavior.

A new current challenge is the shift from propene to propane as feedstock. Since

the early 1990s it has been necessary to develop new generations of multi - compo-

nent catalysts to improve the performance. However, both technical and economic

conditions (the price differential between propane and propene) now exist for the

commercial introduction of direct propane to acrylonitrile processes. Initial exam-