Patented Apr. 8, 1924.

1489,741
UNITED STATEs PATENT OFFICE.
AUGUSTU's E. CRAVER, OF CLIFFSIDE, NEW JERSEY, ASSIGNOR TO THE BARRETT
COMPANY, A. CORPORATION OF NEW JERSEY.
PROCESS OF PRODUCING PHTHALIC ANHYDRIDE,
No Drawing. Application filed February 14, 1922. Serial No. 536,514.
To all whom it may concern: the oxidation thereof may proceed according
Be it known that I, AUGUSTUs E. CRAVER, to the following equations.
a citizen of the United States, residing at (1) CH+ 3O = CHO,--HO. 55
Cliffside, in the county of Bergen and State (2) Alpha naphthoquinone.
5 of New Jersey, have invented certain new
and useful Improvements in Processes of
Producing Phthalic Anhydride, of which . (3) CH--12OPhthalic =
anhydride.
CHO,--4CO,--2HO
the following is a specification. - a .. r Benzoquinone. 60
This invention relates to improvements (4) CH+ 18O =Maleic CHO,-6CO,--3HO
10 in catalysts, particularly to catalysts which anhydride.
are capable of causing oxidation reactions, (5) CH+24O = 10CO, +4.H.O.
and to processes of oxidation wherein such In my work, when optimum oxidizing
catalysts are employed. It relates more conditions are employed (which conditions 65
specifically to the selective oxidation of in
15 naphthalene and has particular reference to as many cases represented the practical limit
the production of phthalic acid anhydride and proportiontemperature,
to catalyst
of air to
time of contact,
naphthalene) in
by the vapor phase oxidation of naphthalene order to obtain the maximum efficiency from
in the presence of an oxygen-containing gas the various catalysts employed, the forma 70
and a catalyst. . tion of naphthoquinone and benzoquinone is
20 I am aware that the catalytic vapor phase found to be quite small and in the majority
oxidation of naphthalene to phthalic anhy. of experiments was absolutely negligible.
dride has been described before, but the foll Therefore, only equations 2, 4 and 5 are to
lowing disclosure of my invention will make considered in appraising the efficiency of 5
clear the improvements which I have ef be a catalyst to be used for the commercial
25 fected in the composition of the catalyst production of phthalic anhydride from
over the previous processes. naphthalene.
In the U. S. Patent No. 1,285,117 a proc Depending on the catalyst employed, I
ess is described which claims the use of have found the production of phthalic anhy 80
vanadium oxides as the catalytic material. dride and maleic anhydride and the com
30 Mention is made in the specification of said plete.combustion of the naphthalene, the lat
patent to the effect that other materials may terrepresented by equation (5)--see above
be mixed with the oxides of vanadium, such occurring in varying proportions.
as magnesium oxide, alkaline earth oxides, It is obvious that reactions (4) and (5) 5
other metallic oxides, asbestos, or inert ma are to be avoided if possible, or at least re
35 terials, but no information is stated concern duced to a minimum, as not only does the
ing the composition of these mixtures, the production of maleic acid or the complete
optimum proportions of ingredients, the op combustion of naphthalene represent a loss
timum running conditions, or indicating of naphthalene, but it also complicates the
that these mixed catalysts differ from the recovery of phthalic anhydride and greatly
40 oxides of vanadium in their catalytic effect, increases the exotherm of the reaction, so
or that they differ very widely among them that if the heat is not eliminated it will con
selves. W siderably augment the temperature of the
In the British Patent No. 170,022 a proc catalyst and cause a still greater violence of 95
ess is described in which fused vanadium the oxidation, which, depending on the cata
45 oxide is used as the catalyst for the oxida lyst employed, will give either more maleic
tion of naphthalene to phthalic anhydride. acid or complete combustion or both. To
Also, in U.S. Patent No. 1,284,888 a proc those acquainted with processes of catalytic
ess is described for the manufacture of oxidation, it is a well known fact that satis 0.
phthalic anhydride, naphthoquinone and factory results can only be realized by main
50 benzoic acid, which claims the use of molyb taining a uniform temperature of the cata
denum oxides as the catalytic material. lyst and especially is this true if a hydrocar
In the catalytic oxidation of naphthalene, bon air mixture be employed, in which the
 s 1489,741

amount of available, oxygen is considerably anhydride and a time of contact of the air
in excess of the theoretical requirement, naphthalene mixture with the catalyst of
which condition is highly desirable. This 0.44 second gave phthalic anhydride, maleic
enormous increase in the exotherm, if the
5 oxidation proceeds beyond the phthalic an
acid and a measurable amount of the prod
hydride stage and the consequently in ucts of complete combustion, i. e., carbon di 70
oxide, and water, these products existing in
creased difficulty of controlling the catalyst the proportions of 4.3:1.0:1.5 respectively.
temperatuite at the desired and favorable Thus, I have found that when vanadium
temperature, can be better understood when
O the heat of reaction of equations (2), (4)
oxide is used under optimum running con
ditions from a commercial standpoint about 75
and (5) are compared. For example, the 40% of the naphthalene that is passed
heat liberated when one pound of naphtha through the catalyst undergoes complete
lene is converted completel into phthalic. combustion and conversion into maleic acid
anhydride is only 6300 British thermal and represents therefore a 40% loss of naph 80
5
units, while if one pound of naphthalene is thalene. The remaining 60% of naphtha
converted into maleic anhydride it is 12,700 lene is converted into phthalic anhydride ex
d when it undergoes complete combustion, cept for a small amount of unchanged
17,300. naphthalene and naphthoquinone.
Therefore, a catalyst which will cause
20 major portion of the reaction, to progress ditions
the . I have further found that varying the con 85.
of the oxidation does not seem to
only to the phthalic anhydride stage is materially change this ratio of oxidation
highly desirable as the heat liberated dur products although, of course, the percentage
ing
making
the the
oxidation
control
will
of
be catalyst
the
quite small, thus yield may change.
relatively
25
The phthalic anhydride production with
simple, and in addition
of naphthalene. . . .
will avoid the waste all the metallic oxides enumerated above,
other than vanadium oxide, was extremely .
I have discovered that when naphthalene is
vaporized in any convenient manner and is small, being almost negligible, and was found
incapable of any material augmentation by
mixed with air in suitable proportions, and varying the conditions. In order to obtain 95
30 finally passed over various oxides of indi
an idea of the activities of these oxides I will
vidual metals at elevated temperatures, the give the results of experiments made with
catalytic action of the oxides of the various molybdenum oxide and uranium oxide, these
metals on naphthalene is widely different. two oxides being the most efficient phthalic
Among the oxides which I have so investi anhydride forming catalysts in their respec 00
35 gated are those of vanadium, molybdenum,
tive classes. With molybdenum oxide at
uranium, chromium, tungsten, iron, lead, 550 C., a yield of about 6.0% of theory of
aluminum, bismuth, and tantalum. These phthalic anhydride was obtained with a com
oxides were investigated with the object of plete combustion of about 10%, the remain
obtaining their relative efficiencies for the der of the naphthalene passing through un 05
40 commercial production of phthalic anhy
changed, except for a very small amount of
dride, and the practical limits of catalyst naphthoquinone, the formation of maleic
temperature, ratio of available oxygen to acid being negligible.
naphthalene, and time of contact had to be Uranium oxide, however, at 450° C. gave
employed with some of the catalysts which a yield
45 gave low phthalic anhydride productions. It
of phthalic anhydride of 5% of the 0
WaS f that only vanadium oxide gave an the remainder ory, along with a 25% complete combustion,
hourly production and a yield of phthalic through unchanged of the naphthalene passing
except for a small pro
anhydride which could be considered as com duction of maleic acid and naphthaoquinone.
50
mercially practicable. The remaining me Thus of the single metallic oxides investi
tallic oxides arranged themselves, in two gated, only vanadium oxide was found to s
classes as follows:- give a phthalic anhydride production of a
Class I, characterized by relatively low magnitude great enough to be considered
Rhtie anhydride production. and relative from a commercial standpoint, but the ac
y low complete combustion, includes the companying large maleic acid formation and 20
55 oxides of molybdenum tungsten, tantalum,
complete combustion are serious disadvan
aluminum and lead. . . . . . . . tages, as they represent a loss of naphthalene
Class II, characterized by relatively low and cause an unduly excessive heat of re
phthalic anhydride production and relative action, thus making it much more difficult to
60
ly high complete combustion, includes the maintain a uniform catalyst temperature, 25
oxides of uranium, bismuth, iron and chro which is absolutely essential for the success
mium. .. . . .. .. of the catalytic oxidation of naphthalene.
Wanadium oxide at about 400° C. catalyst Furthermore, the production of large accom
temperature and with a ratio of air to naph panying amounts of maleic acid increases
thalene possessing about three times the theo the difficulty of obtaining pure phthalican
65
retical oxygen requirement to form phthalic hydride from the reaction products and also O
 1489,741 83

complicates the condensing system for con over the result that would be obtained by
densing the reaction products, owing to the using vanadium oxide alone as the catalyst.
highly corrosive nature of maleic acid. In an investigation dealing with the sta
I have found, however, that very greatly bility of phthalic anhydride, in which a mix 70
augmented productions of phthalic anhy ture of the latter and air was passed through
dride with relatively very small maleic acid a vanadium oxide catalyst under conditions
roductions and complete combustions can approximating as closely as possible those
e realized under conditions which are high existing during the catalysis I have found
ly satisfactory commercially, if proper mix that phthalic anhydride undergoes consid
tures of the oxides of various metals are em erable conversion into maleic acid and com 75
ployed in preference to the oxides of the in plete combustion. I have found, for ex
dividual metals, and this constitutes the new ample, that if a mixture of 8 parts of air
and useful improvements in the composition. and 10 part by weight of phthalic anhy
of catalysts for the commercial production dride be passed through a vanadium oxide
5 of phthalic anhydride, of which this is a catalyst at 4.25° C. at a speed equivalent to 80
specification. a time of contact of the reacting mixture
I have investigated a large number of with the catalyst of 0.4 second, about 53%
catalysts consisting of mixtures of various of the charged phthalic anhydride undergoes
metallic oxides, including in the latter the complete combustion and conversion into
Oxides of vanadium, molybdenum, uranium, maleic acid. These results account for the 85
copper, cobalt, chromium, sodium, alumi large amount of maleic acid and complete
num, manganese, tungsten, tantalum, tita combustion occurring when vanadium oxide
nium, cadmium, lead, iron and bismuth. is used as the catalyst. .
I have found, for example, that small theAnother very efficient mixed catalyst for 90
amounts of manganese oxide, when admixed sistsproduction of phthalic anhydride con
of a mixture of 35% molybdenum oxide
with vanadium oxide, very materially re and 65% vanadium oxide. With this mix
duce the maleic acid formation and com
plete combustion of the naphthalene which ture somewhat less than 25% of the naph
occurs when vanadium oxide is used alone. thalene passing through the catalyst under 95
For example, when a catalyst consisting of goes complete combustion and conversion
approximately 85% of vanadium oxide and into maleic acid. The remainder of the
15% of manganese oxide, which propor naphthalene is converted into phthalic an
tions were found to be most satisfactory, hydride except for a very small amount of
Was employed, only 20% of the naphthalene naphthalene passing through unaltered. 100
passed through the system was found to have The proportions of molybdenum oxide to
undergone complete combustion and conver vanadium oxide may be varied from almost
sion into maleic acid, the production of equal quantities of the two to about three
or four times as much vanadium oxide as
which latter was relatively quite small, while molybdenum oxide without departing from 05
the remaining 80% was converted into
4) phthalic anhydride along with a very small the scope of the invention.
production of alpha-naphthoquinone and Other mixed catalysts, which have been
with a very small amount of naphthalene than found to be more efficient and satisfactory
passing through unaltered. Thus, by em vanadium oxide are vanadium oxide
ploying this mixed catalyst a very much uranium oxide mixtures with the vanadium O
5, smaller maleic acid production, a somewhat oxide largely predominating and also vana
smaller complete combustion, and a rela dium oxide tungsten oxide mixtures con
tively much larger phthalic anhydride pro taining about 20% of tungsten oxide.
duction were obtained compared with those I have also investigated a number of
resulting when a catalyst consisting of ternary mixtures and have found that even 5
s Vanadium oxide alone is employed. I have somewhat better results can be achieved.
found, however, that a somewhat increased For example, the replacing of 5% of
catalyst temperature must be employed with molybdenum oxide in the previously men
this mixed catalyst over that used for vana tioned 65% vanadium oxide 35% molybde
num oxide mixture by means of manga 120
dium oxide, thus making it appear as though
55 the catalytic action of vanadium oxide has nese oxide or copper oxide, reduces still fur
been considerably depressed. The stopping ther the relative amounts of maleic acid
of the bulk of the oxidation of naphthalene naphthalene.and complete combustion of the
production
at the phthalic anhydride stage is perhaps
Thus, by employing mixtures of the oxides
also due to the fact that this mixed catalyst
60 has decidedly less oxidizing action on theof the metals mentioned previously and by 25
formed phthalic anhydride than has vana adjusting the relative proportions of the
dium oxide. The relative amounts of the component metallic oxides present, I have
oxides may be varied somewhat, say from been able to obtain yields of phthalic an
about 10% to 30% of manganese oxide while hydride far in excess of those procurable
from these oxides when used singly, and in 3)
35 still obtaining considerable improvement
 a2 is 1489,741
addition these yields were obtained under
running conditions which are entirely satis
or other suitable powdered, granulated or
factory from the standpoint of the commer fibrous material, which is chemically inert
cial production of phthalic anhydride. Fur and acts merely as a mechanical distributor,
may then be added to the prepared solution
thermore, by the use of these mixed cata or water suspension of the metallic com 70
lysts, I have been able to reduce the maleic pounds and the whole evaporated to dryness
acid formation to a very small, practically while being stirred, after which it is ignited
negligible amount, which when considered in air or other gases or in the presence of
O
along with the greatly reduced complete the naphthalene oxygen-containing gas mix
combustion of the naphthalene, is seen to ture to be used in the catalysis and at the 75
cause a material decrease in the exotherm of same temperature as that used in the oxida
the reaction, thus enabling a more accurate tion.
and easy control of the catalyst tempera My invention will be further explained in
ture. The amount of naphthalene lost has
15 thus been considerably reduced by employ connection with the following example,
which is given for illustrative purposes. 80
ing these mixed catalysts. Furthermore, the It is intended not to limit the procedure
greatly reduced maleic acid formation facili to the exact details given, as the process can
tates the separation and purification of the be varied over widelimits, both in the choice
phthalic anhydride. of the conditions and also in the composi
Investigation of other mixed catalysts tion of the catalyst, without departing from 85
consisting of the metallic oxides previously
mentioned indicated that the oxides more theAspirit and ofscope of the invention.
mixture approximately
basic than those occurring in the fifth and air to 10 part by weight of naphthalene 10 parts of
sixth groups, such as those of sodium, cop por is passed through a catalyst consisting va
25
per, lead, cobalt, aluminum, cadmium, etc., of crushed pumice impregnated with a mix 90
when mixed with vanadium oxide, tended to ture of 65% vanadium oxide and 35% of
increase the relative amounts of maleic acid molybdenum oxide, held at a temperature of
formation and complete combustion and to 450° C. and maintaining
decrease the amount of phthalic anhydride tween the reacting gases aand
30 formed.
time of contact be: .
the catalyst of
about 0.36 second. It is preferable
It has been impossible to arrive at any troduce the catalyst or the carrier contain to in 95
definite explanation of the surprising influ ing the same into tubes, containers or other
ence which one oxide exerts on another confined space, through which the reacting
oxide when both are intimately mixed with gases are passed. The products of the re
35 each other.
Although several mixed oxide catalysts of action, together with the very small amount 00
unchanged naphthalene are then con
have thus been described; it is the intention densed and separated by any of the well
not to be limited to these specific examples, known methods, either physical or chemical
as many other mixtures containing the in nature.
40 oxides previously mentioned, which I have
investigated, can be prepared In choosing ofIfthefractional
reaction
condensation and separation 105
products be employed, it is
these oxides to constitute the mixed cata possible to pass the exit gases containing the
lyst, it is essential, however, that they all small amount of unchanged
should catalyze the oxidation of naphtha after being supplemented withnaphthalene, additional
45 lene to SEE anhydride. d naphthalene and enriched with oxygen or 0
In the preparation of these mixed cata oxygen-containing gas, immediately into an
lysts, it is preferable that the component other converter similar to the first and al
metallic oxides be in the most minutely di ranged in series with the latter.
vided condition possible and intimate con It will be evident that the various condi
50 tact with each other in order to secure the
maximum advantage of the mixed catalyst. tions of the reaction, as stated above, are 15
The preparation of the catalyst can be conditionsofmay
capable wide variation. Among these
accomplished by starting with a solution ture and pressurebe atmentioned which
the tempera
the reaction is
containing the salts of the metals, the ox
55 ides of which metals are desired in the fin carried out, as for example, the reaction
ished catalyst and which metallic salts on or may be carried out at atmospheric pressure 20
ignition leave only the oxides of the metals, timeat ofincreased or diminished pressure; the
or it is also possible to use a water suspen the catalyst; theof relative
contact the reacting gases with
proportions of
sion of the oxides or hydroxides of the met naphthalene to oxygen - containing
60 als. However, I have achieved the most which may consist of air, oxygen, or ozone, 125
gas,
satisfactory results by employing a solution or a mixture of any or all ythese or other
of the complex organic acid compounds of gases which contain free oxygen. These
the metals, as described in my copending conditions are all more or less dependent on
application, Serial No. 513,111. The
rier, which may consist of crushed pumice
car each other. Furthermore, much variation
exists in the choice of the mixed catalyst in 3.
 i489,74 5
the relative amounts of oxides present, 3. The process of producing phthalic an 30
which in turn will necessitate variation or hydride, which comprises passing naphtha
adjustment of the other conditions enumer lene in the vapor phase and an oxygen-con
ated above in order to utilize to the greatest taining gas into contact with a catalyst com
advantage the improvement in the composi prising an oxide of vanadium and an oxide
tion of the catalyst which I have effected. of molybdenum in substantially the propor
Instead of employing very pure naphtha tions of 65 to 35 by weight.
lene I have found that the gratifying re 4. The process of producing phthalic an
sults, which have been described herein for hydride, which comprises passing naphtha.
10 the mixed catalysts, can also be obtained
with relatively impure naphthalene. lene in the vapor phase and an oxygen-con
taining gas into contact with a catalyst com 40
In the specification the words “maleic an prising an oxide of vanadium and an oxide
hydride' and “maleic acid' and also the of molydenum in substantially the propor
words “phthalic acid” and “phthalic an tions of 65 to 35 by weight at a temperature
15 hydride” are considered synonymous. of about 450° C.
I claim:- 5. The process of producing phthalic an
1. The process of producing phthalic an hydride, which comprises passing naphtha
hydride, which comprises passing naphtha lene in the vapor phase and an oxygen-con
lene in the vapor phase and an oxygen-con taining gas into contact with a catalyst com
20 taining gas into contact with a catalyst com prising oxides of vanadium, tungsten and
prising an oxide of vanadium and an oxide molybdenum. 50
of another metal of the fifth and sixth peri 6. The process of producing phthalic an
odic groups, hydride, which comprises passing naphtha
2. The process of producing phthalic
25 hydride, which comprises passing naphtha
an lene in the vapor phase and an oxygen-con
lene in the vapor phase and an oxygen-con prising gas
taining into contact with a catalyst com
taining gas into contact with a catalyst com of a metal of theof6thvanadium
an oxide and an oxide 55
periodic group.
prising an oxide of vanadium and an oxide In testimony whereof I affix my signature.
of molybdenum. AUGUSTUS E. CRAVER.