The Oxidation Methanol With Air Over Iron, Molybdenum, and Iron-Molybdenum Oxides
The Oxidation Methanol With Air Over Iron, Molybdenum, and Iron-Molybdenum Oxides
The Oxidation Methanol With Air Over Iron, Molybdenum, and Iron-Molybdenum Oxides
53
Summary
By the use of allylmagnesium bromide the P-bromo ether synthesis of
olefins has been extended to the preparation of 1,4-diolefins. 1,4-Penta-
diene, 1,4-hexadiene and 1,4-heptadiene and the corresponding a-allyl-P-
bromoalkyl ethers have been described.
The rule that unsaturated compounds with a double bond in the alpha
position boil lower than the corresponding saturated hydrocarbons has been
confirmed. It has been further shown that in any given family the con-
tinuous chain diolefins form a definite series with gradually increasing
boiling points depending upon the relative position of the double bonds.
The 1,2,4,5-tetrabromoparaffins corresponding to the diolefins described
have been prepared and found to show evidences of isomerism analogous to
other polybromo aliphatic derivatives.
COLUMBUS, OHIO
[A COMMUNICATION
FROM THE LABORATORY CHEMISTRY
OF ORGANIC OF THE UNIVERSITY
OF WISCONSIN]
The oxidation of methanol with air over various metal and oxide catalysts
a t 300 to 700’ results in the formation of varying amounts of formaldehyde,
carbon monoxide and carbon dioxide. Sabatier and more recently Lowder-
milk and Day have given a review of the literature on this oxidation to
which little need be added a t this time.’
In 1926 a catalyst containing equal amounts of iron and molybdenum
was found to be very efficient for the oxidation of methanol to formalde-
hyde.2 The investigation described in this paper was initiated in order
to correlate the characteristics of the iron-molybdenum catalyst with those
of the catalyst containing only iron or molybdenum, with the hope of
throwing some light upon the mechanism of the oxidation.
Catalysts. Molybdenum Oxide.-Five grams of ammonium molybdate and 4.5 g.
of malic acid were dissolved in water and made ammoniacal with ammonium hydroxide.
Sabatier-Reid, “Catalysis in Organic Chemistry,” Van Nostrand Company, New
York, 1922, pp. 90--97; Lowdermilk and Day, THIS JOURNAL, 52, 3535 (1930).
Homer Adkins and Virgil E. Meharg, in an application for a U. S. Patent (1927)
assigned to the Bakelite Corporation. The authors of the present paper are indebted
to Mr. Meharg and the Bakelite Corporation for the design of the very efficient washing
tower and for the design and construction of the catalyst chamber, which were used in
this investigation. The senior author is also grateful to Mr. Meharg for many pleasant
associations and very valuable and willing cooperation while the former was temporarily
employed by the Bakelite Corporation during the summer of 1926.
April, 1931 CATALYTIC OXIDATIOK OF METHANOL WITH AIR 1513
To this solution was added 60 cc. of three-sixteenths inch steel ball bearings, previously
pitted with dilute hydrochloric acid and washed with distilled water until free of all acid.
The solution was slowly evaporated almost to dryness with stirring. The steel balls,
covered with the paste, were then poured onto a watch glass in layers of one thickness.
The catalyst balls were further dried by placing in an oven a t 60" for several hours.
Iron Oxide.-Three grams of iron reduced by hydrogen was added to a water
solution of 6 g. of malic acid. The iron was almost completely dissolved by gently
boiling the solution for three or four hours and the undissolved iron removed by decanta-
tion. The solution of ferrous malate was made ammoniacal by the addition of ammo-
nium hydroxide. Sixty cubic centimeters of pitted steel ball bearings was added and the
preparation of catalyst continued as described above
Iron-Molybdenum Oxide.-One and one-fourth grams of iron reduced by hydrogen
was added to a water solution of 4.5 g. of malic acid. The iron was almost completely
dissolved by gently boiling the solution for three or four hours and the undissolved iron
removed by decantation. To this solution of ferrous malate was added with stirring a
water solution of 3 g. of ammonium molybdate The combined solution was made
ammoniacal by the addition of ammonium hydroxide, steel balls were added and the
catalyst prepared as described above.
The catalysts thus prepared were placed in the aluminum reaction tube and de-
composed in a slowly moving current of air for an hour or more a t 360-375". They
were then ready for use.
Apparatus
The apparatus used in the experimental work is shown in Fig. 1. The air, which
was supplied by a compressed air line A, was dried by passing it through a bottle B of
e
LTy
-
r
i
tube was made of a flattened piece of aluminum tubing 75 cm. in length and 50 X 6 mm.
(1. D.) in cross section. It was filled in the bottom and inlet side with glass rod and
balls. A 15-cm. depth of catalyst was placed in the outlet side. The catalyst tube was
immersed in an electrically heated bath (E) of sodium and potassium nitrates. The
temperature of the bath was controlled by means of a Leeds and Northrup potentiome-
ter controller. The products of the reaction were passed through a water-cooled coil
condenser and then through a bubbling tower G 60 cm. in length, down which water
was passed a t the rate of about 100 ml. per hour from the reservoir H. All the formalde-
hyde and methanol in the gases were collected in the wash tower and flowed back
through the condenser into bottle F. I is an exit tube for gaseous products leading to an
Orsat apparatus.
Examination of Products.-The water-insoluble gaseous products were analyzed
in an Orsat apparatus. The carbon dioxide was determined by absorption in potassium
hydroxide solutions. The remaining oxygen was removed by yellow stick phosphorus.
The carbon monoxide was removed by bubbling through an acid solution of cuprous
chloride. The water-soluble products collected in bottle F were analyzed for formalde-
hyde by the method of Haywood and Smith.8 The unreacted methanol was determined
by oxidation with chromic acid.'
The percentage efficiency as used in this paper refers to the percentage of oxygen
used in the conversion of methanol t o formaldehyde and was calculated from gas analysis
data in the following manner
% Eficiency =
% oxygen used - (1.5 X %
' carbon dioxide + % CO)
% oxygen used
The percentage efficiency so calculated agrees quite closely with the percentage effi-
ciency calculated from the weight of formaldehyde obtained, the amount of methanol
passed over the catalyst and the weight of methanol which was found unchanged in the
distillate.
The Formation of Methylal.-Under certain conditions of operation formaldehyde
reacts with methanol in the catalyst chamber with the formation of the methyl acetal
of formaldehyde. The compound is only partially condensed with the other liquid
products and some escapes with the gaseous products of the reaction. This acetal is
not at all readily hydrolyzed. The method of Lowdermilk and Day6 for the deter-
mination of unchanged methanol may give quite inaccurate results because methylal
would appear with methanol in their procedure for analysis. The presence of methylal
among the reaction products may be suspected if there is a 0.5 to 2% reduction in the
volume of the effluent gas when bubbled through sulfuric acid, or if there is a discrepancy
between the efficiency as calculated from the gas analysis and from the formaldehyde
and methanol in the distillate. Methylal was formed in considerable quantities in one
series of experiments, apparently due to the accumulation of alumina on the walls of
the catalyst chamber. No further difficulty was experienced after the chamber had
been thoroughly cleaned with a sodium hydroxide solution and with hydrochloric acid.
Methylal was also formed when rather completely oxidized aluminum turnings were
used as a support for the catalysts.
General Procedure.-The catalyst was put in place and decomposed and methanol
and air passed over it continually (except for the removal of samples) for several days or
weeks. Two or three times during each twenty-four hours the passage of the reactants
was stopped and the amount of methanol which had been passed over the catalyst de-
136 liters of air were passed over the catalyst in an hour, 2.31 g. of the
methanol was converted into formaldehyde, as compared to the conversion
of 2.41 g. of methanol when 6.67 g. of methanol in 50 liters of'air were
passed over the catalyst.
4. The percentage conversion of methanol to formaldehyde rose
with decrease in the rate of passage of the methanol over the catalyst.
When 20.0, 10.2, 6.67 and 2.57 g. of methanol were passed over the catalyst
with 50 liters of air, the yields were 26.9,33.3, 36.2 and 54.2%, respectively.
5 . The conversions of methanol to formaldehyde a t 373, 390 and 400"
were approximately the same. However, a t 361' lower conversions were
obtained. When approximately 9.5 g. of methanol in 93 liters of air was
passed over the catalyst a t 361, 373, 390 and 400') yields of 23.8, 31.5,
32.1 and 32.4%, respectively, were obtained when the catalyst had
reached its "steady state."
The characteristics of an iron oxide catalyst, as concluded from a study
of its behavior when methanol mixed with air was passed over it, were
found to be as follows.
1. The catalyst was very active, the methanol being oxidized almost
wholly to carbon dioxide. When the amount of methanol passed over
the catalyst was varied from 11.2 to 33 g. per hour (373') 93 liters of air
per hour), there were only traces of formaldehyde and this only with a
catalyst that had been in operation for several days and with more meth-
anol passing over the catalyst than could be oxidized to carbon dioxide
by the available oxygen.
2. The activity of the catalyst became less as the temperature of the
reaction was lowered and increased as the temperature was raised. At 310°,
with 20 g. of methanol and 93 liters of air, 9.4% of the methanol passed
over the catalyst was converted to formaldehyde; 81.0% remained un-
reacted, the remaining 9.6% having been converted into carbon dioxide.
At 340' approximately 18% of the methanol was converted into formalde-
hyde, with about 68% remaining unreacted; the remaining 14% was
converted into carbon dioxide. At 400 or 443', no formaldehyde was
formed, 4.92% of the methanol remained unreacted, the remainder having
been lost as oxides of carbon.
3. An iron oxide catalyst that has been used a t a low temperature
(340') when raised to 373' was less active toward oxidizing methanol
and produced more formaldehyde than it formerly did a t 373'. It gradu-
ally recovered its original activity, giving results identical within experi-
mental error with those obtained previously a t 373'. Previous to the
lowering of the temperature, the results obtained a t 373' showed the
formation of 1.3% of formaldehyde with very little methanol (5.1%)
remaining unreacted. After raising the temperature again to 373'
the results obtained for the first two experiments showed conversions
April, 1931 CATALYTIC OXIDATION OF METHANOL WITH AIR 1517
of methanol. The catalyst was quite stable and may be used continually
for months.
At least in the case of molybdenum oxide there is evidence that the
reaction between methanol and oxygen takes place when both reactants
are on the surface of the catalyst.
Evidence has been presented which indicates that the catalyst surface
of these oxide catalysts is continually renewed during use; therefore
the relationship between an effective catalyst and the reactants must
not only be such that the former brings about the reaction of the latter,
but the reactants must maintain the catalyst in the state of oxidation
which is favorable to the desired reaction.
MADISON,
WISCONSIN
[A COMMUNICATION
FROM THE LABORATORYOF ORGANIC
CHEMISTRY
OF THE UNIVERSITY
OF WISCONSIN]
It is well known that the reaction of alkyl halides and magnesium results
in the formation of Grignard reagents and of hydrocarbons, as indicated
in equations I and 11.
RX + Mg = RMgX (1)
2RX + Mg = Rz +
MgXz (11)
There has been in progress for some time in this Laboratory a study of the
factors which determine the ratio of these two competitive reactions.
During the course of this investigation some observations have been made
in regard to the effect of copper upon the ratio of the reaction products
which are of such practical importance that it seems desirable to publish
them in advance of a more comprehensive paper, especially in view of the
results of Gilman, Peterson] Schultz and Heck,' who reported that mag-
nesium in certain copper-magnesium alloys reacted more rapidly with
alkyl halides than did pure magnesium. Upon the basis of these findings
it seemed that the use of the alloy might be advantageous in the prepara-
tion of Grignard reagents and this practice has been followed by a number
of investigators.
The determination of the effect of various experimental conditions upon
the ratio of the Grignard and Wurtz reactions was made in the following
manner. The apparatus used was essentially of the type and size de-
scribed and used by Gilman and Meyers for the preparation of ethyl-
1 Gilman, Peterson and Schulze, Rec. tras. chim., 47, 22 (1928); Gilman and Heck,
Bull. SOC. chim., 45,250 (1929).