|UFDC Home||myUFDC Home | Help|
|Table of Contents|
|List of Tables|
|List of Figures|
|Table of Contents|
Table of Contents
List of Tables
List of Figures
THE EFFECT OF METAL IONS ON THE
AUTOXIDATION OF BENZALDEHYDE
EDWARD NATHAN KRESGE
A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE
DEGREE OF DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
ACIaOWLE rPG.. I ITTS
The author wishes to express his sincere appreciation to his advisor
and teacher, Dr. W. T. Lippincott, for his empathy and friendship and for
his inspiration and direction of the research.
The author also wishes to express his appreciation to the members of
the Department of Chemistry and to his friends for their encouragement and
TABLE OF CONTENTS
LIST OF TABLES v
LIST OF FIGURES vi
I. Aims and Scope 1
II Autoxidation of Benzaldehyde 4
EXPERIMENTAL PROCEDURE 18
I, General Statement of Experimental Procedure 18
IH Purification of Benzaldehyde 19
I. Preparation and Purification of Other Chemicals 21
IV. Oxidation Apparatus 25
V. Typical Autoxidation Run 29
VI. Analytical Methods 33
I, Manganese Systems 35
I. The Catalytic Decomposition of Perbenzoic Acid 70
II Cobalt Systems 75
IV, Cerium Systems 81
TABLE OF CONTENTS (Continued)
I, Manganese System 88
HL Cobalt System 99
m c Cerium System 102
BIOGRAPHICAL SKETCH 113
LIST OF TABLES
1. Activation Energies for the Autoxidation of Benzaldehyde 7
2. Effect of Manganese (II) Concentration on the Oxidation o1
3. Effect of Manganese (III) Concentration on the Oxidation of
4. Effect of Perbenzoic Acid on the Oxidation Rate at Low
Catalyst Concentrations 45
5. Effect of Initial Benzaldehyde Concentration on the Oxida-
tion Rate 49
6. Effect of Oxygen Pressure on the Oxidation Rate 52
7. Effect of Water on the Manganese Catalyzed Autoxidation
of Benzaldehyde 53
8. Effect of Benzoic Acid on the Rate of Oxidation 56
9. Change in the oxidation Rate with Temperature 61
10. Effect of Manganese (II) with Acetic Anhydride 64
11. Variations in Bonzaldehyde and Perbenzoic Acid Concen-
trations During the Oxidation 65
12. Effect of Cobalt Concentration on the Rate of Benzaldehyde
13. Effect of Benzaldehyde Concentration on Oxidation Rate 79
14. Effect of Cerium on the Oxidation of Benzaldehyde 83
LIST OF FIGURES
1. Oxidation Apparatus 26
2. Calibration of Recorder 28
3. Typical Autoxidation Run 31
4. Determination of the Induction Period 32
5. The Effect of Manganese (II) Ion on the Oxidation Rate of
6. Induction Periods for Manganese Catalyzed Autoxidation
of Benzaldehyde 39
7. The Effect of Manganese (HI) Ion on the Extent of Reaction 40
8. The Effect of Manganese (II) Ion at Low Concentrations on
the Rate of Oxidation 42
9. Plot Showing the One-Half Order Dependency on Perben-
zoic Acid 47
10. The Effect of the Initial Benzaldehyde Concentration on
the Oxidation Rate 50
11. Plot of log -dO2/dt versus log [BzH] 51
12. The Effect of Water on the Extent of Reaction 54
13. Perbenzoic Acid Concentrations During the Autoxidation
at Various Catalyst Concentrations 58
14. Increase in the Peracid Concentration During the Induction
15. Exponential Build-Up of Peracid During the Induction Period 60
LIST OF FIGURES (Continued)
16. Arrhenius Plot for Manganese-Catalyzed Autoxidation 62
17. Autoxidation of Benzaldehyde 66
18. Decomposition of Perbenzoic Acid by Manganese (II)
19. Decomposition of Perbenzoic Acid by Manganese (III)
20. Comparison of Manganese (I) and Manganese (III)
Catalyzed Decomposition of Perbenzoic Acid 74
21. The Effect of the Cobalt Concentration on the Benzalde-
hyde Oxidation Rate 77
22. Variation of the Oxidation Rate with Benzaldehyde Con-
23. Typical Cerium Catalyzed Run 82
24. The Effect of Cerium Ions on the Induction Period 85
25. The Effect of Cerium Ions on the Post Induction Oxida-
tion Rate 86
I. Aims and Scope
Transition metal ions have been used as catalysts for many types of
free radical reactions. The nature of their role in these reactions has not
been fully established, although considerable effort has been expended to
elucidate the detailed mechanistic involvement of the metal ions in the reac-
tion sequence. The autoxidation of alcohols, phenols, ethers, amines, enols
and hetols is catalyzed by heavy metal ions, but the most extensively studied
reactions are the autoxidations of hydrocarbons and aldehydes.*
The autoxidation of aldehydes proceeds with ease, even at room
temperature, and undoubtedly proceeds by the chain reaction sequence:
RCHO + In' -- InH + RCO- (1)
RCO. + 02 ----RCOOO (2)
RCOOO + RCHO --*-RCO + CO + OOH (3)
where In* is an initiator radical. The peracid that is produced by this reac-
tion then reacts with unreacted aldehyde to yield the final product, an acid, as
indicated by the equation
RCOOOH + RCHO ---2 RCOOH (4)
*References and a summary of this work are given in Part H of the
To be effective as a catalyst the metal ion must have two readily ac-
cessible oxidation states differing by one unit. Examples are cobalt (II) and
(III), manganese (I) and (II), copper (I) and (II) and corresponding pairs of
ions from cerium, vanadium and iron. It has been suggested that the metal
ion may be interacting with the aldehyde to form an initiating radical accord-
ing to the equation
M+3 + RCHO M+2 + H+ + RCO- (5)
or with the peracid formed in the system via the following reactions:
M+2 + RCOOOH M(OH)+2 + RCO" (6)
M(OH)+2 + RCOOOH -- M+ + HOH + RCOOO (7)
Hence, if a metal in the lower oxidation state is added to a solution containing
peracid, the metal ion will be oxidized, then reduced, then oxidized again, etc.
Each of these reactions will generate a free radical that is capable of removing
a hydrogen atom from a molecule of aldehyde thus initiating the chain. Analo-
gous steps are postulated for the autoxidation of hydrocarbons.
It is reported that mixtures of ions of different metals are more effi-
cient catalysts than the individual ions taken separately (1).
The aim of this work is to study the autoxidation of benzaldehyde in an
effort to elucidate in more detail the mechanism of metal ion catalysis. Hav-
ing done this it might be possible to extend the work to explain the synergism
between two different metal ions in an autoxidation.
This work is limited to an investigation of the effects of manganese,
cobalt and cerium salts in the autoxidation of benzaldehyde in glacial acetic
The information reported herein was obtained by 1.) non-kinetics
studies which include identification of the reaction products and the isola-
tion and characterization of intermediates, and 2.) kinetics studies on the
rate of oxygen uptake, the rate of substrate loss and the rate of formation
and loss of the intermediate peracid as various parameters are altered.
Further kinetics studies were performed to establish the nature of the in-
termediate stages of the reaction in the absence of benzaldehyde using
manganese ions as a catalyst.
II. Autoxidation of Benzaldehyde
In 1832 Liebig (2) observed that benzaldehyde was converted to benzoic
acid on standing in the air and that the reaction was accelerated by light.
Other workers (3, 4) found that autoxidizing benzaldehyde was a strong oxidiz-
ing agent and that, when acetic anhydride was used as a solvent, twice as much
oxygen was consumed (4), yielding acetic acid and benzoylacetyl peroxide as
products. The rate of oxidation is decreased by an extensive number of inhi-
bitors and is increased by radical initiators and by traces of metals. Blckstrim
(5, 6) postulated that a chain mechanism was involved in the autoxidation of both
heptaldehyde and benzaldehyde after discovering quantum yields from 10,000 to
15, 000 for the reactions. Christianson (7) had previously suggested a chain
mechanism for these reactions because of the observed inhibitions.
Subsequently, this reaction and others have been studied by a large num-
ber of workers so that some detailed mechanisms of aldehyde autoxidation have
A. Non-kinetics studies
The first product of the benzaldehyde autoxidation is the peracid which
reacts heterolytically with unreacted aldehyde to give the final acid product in-
dicated by the equation
C6H5CO3H + C6HsCHO C6H5CO2H5C )
The rate of this reaction is slow (8, 9) so that in an autoxidation there will be
an increase in the concentration of the peracid as the reaction proceeds (10,
A complex, stable at -300, was isolated using peracetic acid and
acetaldehyde (13). It decomposed to acetic acid upon heating with manganese
(II) ion and was differentiated from peracetic acid by using a 10 per cent
potassium iodide solution. Iodine was liberated more rapidly by the peracid
than by the complex. Bawn and Williamson (14) proposed that a complex be-
tween two molecules of aldehyde and one molecule of peracid was also formed
in acetaldehyde autoxidation. Evidence from their studies indicated that the
decomposition of the complex gave one molecule of acetic acid and one mole-
cule of aldehyde.
A complex of this nature has not been isolated from benzaldehyde-
perbenzoic acid mixtures (15, 16), although the kinetics data of Wittig and
Pieper (16) suggested that such a complex formed rapidly and decomposed
slowly at room temperature. Two possible structures for the complex were
O0 0 OH
CgH5C .CHC6H5 and C6IH5COOC6 H5
*Two titrations were used to find the concentration of the complex; the
first, to find the peracid concentration and the second, to find the peracid and
A structure analogous to compound (II) has been proposed as an intermediate
in the Baeyer Villiger reaction by Criegee (17) and supported by Doering and
Other products observed under various reaction conditions are carbon
dioxide (13) and a small amount of uncharacterized yellow solid (10, 19). With
inhibitors added, a variety of products were found, presumably resulting from
radical attack on the inhibitor (20, 21).
B. Kinetics studies
A variety of kinetics studies have been reported. The results of some
of these are summarized in Table 1. (22, 23) and in the following paragraphs.
Order in aldehyde. The dependence of the rate on the benzaldehyde
concentration was found to be second order in photoxidations (24), second order
in thermal oxidations at high oxygen pressure, first order in benzoyl peroxide
initiated reactions (12) and three-halves order in the presence of cobalt ions
Order in oxygen. Almquist and Branch (25) in 1932 found that the ther-
mal oxidation is first order in oxygen, but for the same reaction in benzene
solution there was not a simple oxygen dependency. However, as the benzalde-
hyde concentration was increased the reaction approached first order in oxygen
The order changes with both temperature and pressure. At 160 the
thermal oxidation is zero order, but as the temperature is increased a first-
order relation is established (27).
ACTIVATION ENERGIES FOR THE AUTOXIDATION OF BENZALDEHYDE
Initiation Oxygen order Experimental Reference
Photo Chem. 0 1.8 27
Thermal 0 7.6 27
Thermal 0 13.6 12
Thermal 0-1 17.7 26
Photochemical 5. 24
Cobalt 0 14.7 8
Benzoyl peroxide 0 17.2 12
The oxidation rate was found to comply with the generalized expression
Rate = a ) 9)
which gives orders between zero and unity depending on the values of the para-
meters a, b, and c.
The cobalt ion-catalyzed oxidation was found to be zero order in oxygen
from 550 to 950 mm. pressure (8). The photoWxidation in decane solution has
also been reported to be zero order in oxygen.
Activation energy. Studies of the activation energy have been compli-
cated by the change of order in oxygen with temperature. Both the activation
energy of the benzaldehyde autoxidation and the effect of temperature on the
oxygen concentration in solution are reflected in the rate-temperature relation-
ship. If the oxygen order is shown to be zero, then the observed activation
energy from an Arrhenius plot will be equal to the real activation energy.
However, if the rate is dependent on the change of oxygen solubility with
temperature the apparent activation energy will consist of a chemical rate term
and an equilibrium term so that a correction must be made to obtain the activa-
tion energy from an Arrhenius plot. Table 1 summarizes the activation energies
reported by various workers.
Initiation. The autoxidation of benzaldehyde has been initiated in many
ways including radical sources, ozone, metal catalysts and photolysis. For
the thermally induced reaction there is considerable disagreement as to the
possible initiating step and some authors feel that the "uncatalyzed" reaction is
in fact initiated by traces of metal ions or peroxides. The metal ion catalyzed
initiation will be discussed in more detail.
Chain length. In a radical chain reaction, under steady-state conditions,
the number of chains initiated must equal the number of chains terminated.
BEckstrim and Beatty(19), using anthracene as an inhibitor in benzalde-
hyde autoxidations were able to get some measure of the chain lengths by assum-
ing that each molecule of anthracene oxidized during the reaction corresponded
to one chain termination. The number of molecules of benzaldehyde oxidized as
measured by the total oxygen consumption divided by the number of chains
terminated gave the average chain length. The chain lengths calculated are
very large. This may be accounted for by the fact that anthracene is a rather
ineffective inhibitor and other termination processes are occurring simultane-
ously. The results, however inaccurate, undoubtedly show that the autoxidation
is of chain character.
Briner and Papazian (28) determined the chain length for ozone-initiated
benzaldehyde oxidations in carbon tetrachloride solution to be as high as 5,000
using air and as high as 50,000 using pure oxygen.
Inhibition. Bafckstrom (5) showed that in the photo-initiated autoxidation
of benzaldehyde the order in light intensity depended upon the kind of inhibitor
used. Using anthracene as an inhibitor, the order in light intensity was found
to be 0.65, while for others the following values were obtained: benzyl alcohol,
0.5; diphenylamine and hydroquinone, 0.9. Pure benzaldehyde gives an order of
0.5 in light intensity.
Waters (29) studied the retardation of benzoyl peroxide-initiated benzalde-
hyde autoxidation. His kinetics results indicate thatp-cresol and m-2-xylenol
act primarily as chain-transfer agents and in view of the observed dependencies
of rate on the concentrations of benzaldehyde, benzoyl peroxide and inhibitor,
it was concluded that the main chain ending process was a dimerization, or
possibly a disproportionation, of the phenolic inhibitor radicals:
2 C7H70* -- inert products (10)
With the m-2-xylenol system, it was established that the dimer (C8H9O)2 is
further oxidized to 3, 5, 3', 5'-tetramethyl-4,4'-diphenoquinone. The di-
quinone is still a retarder of the autoxidation and it was inferred from further
kinetics studies that it acts as a chaining stopping agent by combining with
Later kinetics studies by Waters (20, 21, 30) of the benzoyl peroxide
catalyzed autoxidation, retarded by a number of polycyclic hydrocarbons, in-
dicated that chain termination is initially effected by combination of benzoyl-
peroxy radicals with the hydrocarbon.
The peracid-aldehyde reaction. The reaction by which the peracid goes
to the acid has been found to be first order with respect to both the peracid and
aldehyde (13) in the case of acetaldehyde. It was observed that the intermedia-
tory peracetic acid reacts with acetaldehyde to yield a peroxide compound
which was decomposed by manganese (MII) catalyst to give two molecules of
acetic acid. Bawn (15) has indicated that no complex is formed between per-
benzoic acid and benzaldehyde. However, there are some kinetics data that
suggest the rapid formation of a complex which decomposes slowly to benzoic
There has been more extensive work in the peracid oxidation of ketones
than of aldehydes. It is now generally accepted that the reaction is ionic in
character and proceeds by the following steps:
0 O O-H
R-C-R' + R-C-OOH - R-C-R'
R-0-C-R+ R-C-OH (11)
This scheme accounts for 1.) the isolation of hydroxyhydroperoxides which can
be decomposed by heating, 2.) the fact that the migratory aptitude of R and R'
is promoted by their electron releasing ability, 3.) the observation that in the
oxidation of acetophenone with perbenzoic acid the rate determining step is the
acid catalyzed addition of the peracid to the carbonyl group (31) and 4.) the
radioactive isotope data (18).
Benzaldehyde reacts with hydrogen peroxide in ether to give the correspond-
ing acid and 0.7 per cent phenol (32). Benzaldehydes substituted with electron
pushing groups when reacted with hydrogen peroxide under the same conditions
give higher yields of phenols. For example, 2, 4-dimethoxybenzaldehyde gives
26 per cent phenolic products and 74 per cent substituted benzoic acid. Phenolic
products produced as a side reaction could be responsible for the self-inhibition
of an autoxidation.
C. The reaction mechanism
A mechanistic rationalization of the kinetics and non-kinetics data for
aldehyde autoxidation has been formulated as more and more facts were deter-
mined. The first substantial break-through was showing that the reaction
proceeded via a free radical mechanism. Then the question as to the particular
nature of the individual steps became the subject of considerable effort and, at
times, considerable conjecture.
If the individual steps are to be resolved from idnetics data, it is
necessary to know the concentrations of the various radicals involved.
Until the present time, no method has proven powerful enough to measure
the very small radical concentration in an autoxidation with sufficient
accuracy to be helpful. Electron paramagnetic resonance has the calculated
theoretical sensitivity to detect 2 x 10"14 moles of free radicals (338 34), but
it has been largely limited to radical species that are stable in solution or
captured in the solid state. However, spectra have been observed of free
radicals involved in vinyl polymerizations (35) and in highly inhibited autoxi-
If it is assumed that the radical concentration is constant, i.e., a
steady state has been reached in which the rate of radical production equals the
rate of disappearance, it is then possible to write differential equations for the
rate of radical production and disappearance and set them equal to zero. In
this manner, it is possible to determine the ratios of the rate constants of
the separate reactions in the total sequence and to predict the order in sub-
strate and catalyst. If the result agrees with the experimental rate law, then
the mechanism proposed is a possible one. There is, of course, the chance
that the results coincide fortuitously. Thus, this type of agreement is neces-
sary but not sufficient for a proof of mechanism.
The initiation reaction. Photochemically induced reactions (37) are
primarily initiated by homolytic cleavage between the alkyl or aryl group and
the carbonyl group:
RCHO R- + HCO-
Cleavage between the carbonyl and hydrogen probably does not occur, since
attempts to isolate acetyl iodide from the products of acetaldehyde-iodide
photolysis have been unsuccessful (38). This is reasonable in light of the
higher bond dissociation energy of the C-H bond as compared with the C-C
Mulcahy and Watt (26) have studied the kinetics of uncatalyzed autoxi-
dation of benzaldehyde in benzene solution. The initial rate of oxidation is
given by the relation:
d O2 ] FRH]2 [O2 [
dt b[RHj +c[ 02] (1
The reciprocals of the constants b and c vary with temperature according to
the Arrhenius equation. The authors were able to conclude that:
1.) The mechanism of the thermally initiated
reaction is different from the various
catalyzed, photochemical and photosensi-
tized oxidation previously studied (14, 15,
2.) More work on the kinetics of uncatalyzed
thermal oxidations is desirable.
They proposed the following mechanism:
RH Q* initiation (14)
Q* + 0--- Q*02 (15)
Q*o + RH --- ROOH + Q* (16)
Q2 -- x (17)
Q + RH- -X termination (18)
where Q and QO2* are unspecified chain carriers. This mechanism is similar
to that suggested by George and Robertson (42) to explain the kinetics of tetralin
oxidation at 1100.
The major paths by which metal salts catalyze autoxidation appear to be
reaction with the peroxide intermediate (43, 44, 45, 46) and/or direct initiation
involving the aldehyde. With acetaldehyde the autoxidation rate is proportional
to aldehyde and cobalt concentrations but independent of the oxygen pressure (14).
Thus, the kinetics suggest:
Co+2 + CHI3CO3H Co+3 + CH3CO2 + OH- (19)
Co+3 + CHI3COt3 -H Co+2 + CH3CO3' + H+ (20)
as primary processes.
For benzaldehyde in acetic acid Bawn found the rate law to be:
d  k [RCHOj 3/2 [Con] 1/2 (
This rate law may be rationalized in terms of a direct initiation reaction.
Initiation of this type was first proposed by Haber and Willstttter (47)
for the iron (II) ion catalyzed autoxidation of acetaldehyde:
CH3CHO + Fe3+-- CH3CO' + H+ + Fe2+ (22)
Uri (48, 49) has suggested a third type of initiation in the cobalt ion catalyzed
oxidation of methyl stearate in which the cobalt ion acts as an oxygen carrier
capable of initiating chains:
Co+2 + 02 CoO2* (23)
CoO2.+2 + Co+2(RHI) --- Co+2 + Co+3R- + HO2* (24)
HO2. + RH -- HO2H + R (25)
Able (50, 51) suggested the same scheme for the lead catalyzed oxidation of
The propagation reaction. It is generally agreed that the chains are
RCO- + 02 RC03. (26)
RC03' + RCHO ----RCO3H + RCO. (27)
The termination reaction. There are three possible termination reac-
tions in uninhibited aldehyde autoxidation:
2RCO ---RCOCOR (28)
RCO' + RCO-" RCOO2COR (29)
2RCO3 --- (RC3)2 (30)
In an oxidation that is zero order in oxygen it is reasonable that the concentra-
tion of RCOg3 is large in respect to the last reaction.
Inhibitors may be added to the reaction mixture to react with one of the
chain-carrying radicals to shorten the average chain length thus decreasing the
reaction rate. There are two possibilities with phenolic inhibitors:
RCO- + R'OH RCHO + R'O' (31)
RCO3. + R'OH ---RCO3H + R'O (32)
If R'O" is consumed by a radical-radical reaction, the order in the initiating
reaction will be one. Intermediate orders would result from a competition
between the various termination reactions.
There is some evidence for the self-inhibition of liquid-phase photolysis
initiated autoxidation of benzaldehyde (24, 52). It was suggested that a com-
pound corresponding to the formula
OH OH OH
I I I
was responsible. In the cobalt catalyzed benzaldehyde oxidation with no solvent,
water and phenols formed during the reaction retarded the process (53). The
introduction of strong initiator decreased the retarding effect to a certain
Termination reactions that involve metal ions have been suggested (54,
55, 56) in the ferrous-ion-peroxide-initiated polymerization of acrylonitrile
and similar monomers. There appears to be competition between the termina-
RO- + Fe2+-- RO" + Fe3+ (33)
and the propagation reaction:
RO' + CH2 =CHR' -- ROCH2-CHR (34)
Bamford et al., (57) have shown that ferric chloride is an efficient in-
hibitor of vinyl polymerization and Hammond (39) has found that ferric chloride
inhibits the autoxidation of tetralin in chlorobenzene.
Metal ions reacting with radicals can not be excluded as a possible
termination reaction in aldehyde oxidations (56, 58).
I. General Statement of Experimental Procedure
The rate of autoxidation of bonzaldehyde in glacial acetic acid under the
influence of metal-ion catalysts was studied in the apparatus described in the
following section. The oxidation rate was followed by observing the uptake of
oxygen as a function of time. From these data, the length of induction periods,
the rate of the reaction and the extent of the reaction were determined as func-
tions of various parameters.
In some runs the rate of loss of benzaldehyde was followed. In other
runs the rate of the formation of perbenzoic acid was found. The products of
the autoxidation were identified for selected runs.
II. Purification of Benzaldehyde
The general procedure for the purification of benzaldehyde for use in
kinetics experiments has been to perform repeated vacumn distillations under
an inert gas such as hydrogen or nitrogen (26, 59).
Benzaldehyde (Merck) was allowed to stand one week over anhydrous
potassium carbonate. Purified nitrogen was kept over the benzaldehyde to
prevent oxidation. The benzaldehyde was filtered and transferred to the dis-
tillation flask together with fresh anhydrous potassium carbonate.
The distillation was performed using a one meter column equipped
with a variable take-off head controlled by an electronic repeat-cycle timer.
The externally heated column was packed with single-turn glass helices. The
distillation flask was fitted with a nitrogen bleed capillary. A "fraction-
cutter" was also used so that the entire distillation could be run without
exposing the system to the atmosphere. After maintaining total reflux for
one hour 3,000 ml. of benzaldehyde was collected at a reflux ratio of six to
one. A 2,300 ml. center-cut was collected and redistilled at a reflux ratio of
ten to one. A center-cut of 1,800 ml. was collected from the second distilla-
tion and subjected to both infrared and gas phase chromatographic analysis.
The samples for infrared were run in 0.025 mm. cells on a Perkin-
Elmer Model 21 Infrared Spectrophotometer and there was no peak at 2.9
microns indicating that no appreciable amounts of water, benzoic acid or
perbenzoic acid were present.
The benzaldehyde was stored in a blackened pyrex flask fitted with a
delivery tube and a gas inlet. Removing benzaldohyde from the storage flash:
was accomplished by applying a positive nitrogen pressure through the gas
All glass apparatus was used in both the preparation and storage of
the benzaldehyde. Dow-Corning high vacuum grease was used on all ground
II. Preparation and Purification of Other Chemicals
A. Acetic acid
Baker reagent grade glacial acetic acid was distilled at atmospheric
pressure using the column and distillation head described under the purifica-
tion of benzaldehyde. Prior to the start of the distillation one per cent of
acetic anhydride was added to the distillation flask. After total reflux for two
hours 100 ml. of forerun was discarded, the center-cut was collected at a
reflux ratio of ten to one and 400 ml. was left in the distillation flask. The
system was kept under purified nitrogen throughout the distillation.
B. Manganese (II) acetate
Hydrated manganese (II) acetate, reagent grade, was obtained from
Fisher Scientific Company. A weighed amount of the compound was dried
over phosphorus (V) oxide at 1030 and 1.5 mm. of Hg for 72 hours. The
weight loss indicated that the product was anhydrous.
C. Manganese (III) acetate
Manganese (III) acetate was prepared by adding 80 g. of acetic anhydride
to 20 g. of 50 per cent manganese (II) nitrate in a magnetically stirred 2000 ml.,
three neck flask fitted with a reflux condenser, addition funnel and thermometer.
The manganese (II) nitrate solution was heated to 1450 before the anhydride was
added drop-wise. A temperature of 120 to 1250 was maintained without further
heating by regulating the addition rate of the anhydride, When the evolution of
nitrogen dioxide stopped the reaction mixture was allowed to cool to room tem-
perature and then it was placed in an ice bath for one hour. The solid manganese
(II) acetate was then filtered from the acetic acid and excess acetic anhydride.
The product was washed six times with acetic anhydride followed by eight wash-
ings with dry ether. The ether was removed from the light tan crystals under
reduced pressure at room temperature.
The equation for the overall reaction between manganese (II) nitrate and
acetic anhydride is:
2 Mn(NO3)2(H20)6 + 15 (CH3CO)20
2 Mn(CH3COO)3 + 4 NO2 + 1/2 02 + 24 CH3OOH (35)
The method used was modified from the procedure used by Chretien and
D. Cobalt (II) naphthenate
Cobalt (II) naphthenate was obtained from the Noudex Corporation as
"Noudex Cobalt 6". It was found to contain 5.40 per cent cobalt as cobalt (II)
naphthenate in naphthenic acid (61). A solution 3.00 x 10-1 M cobalt in redistil-
led glacial acetic acid was prepared and analyzed spectrophotometrically. The
concentration of this solution was found to be 3.00 x 10-1 M cobalt.
*The writer is indebted to Mr. K. Hickey for the analysis of the cobalt
E. Cerium (IV) naphthenate
Cerium (IV) naphthenate was obtained from the Noudex Corporation as
"Noudex Cerium 6". Stock solutions were made up in glacial acetic acid.
F. Oxygen and nitrogen
Cylinder Oxygen (U. S. P.) was obtained from Linde Company. Passing
the oxygen through a long tube packed with anhydrous calcium chloride removed
possible traces of water.
Gas Phase Chromatography of the oxygen showed only one peak. One
cylinder was used for all of the autoxidation runs. No attempt was made to re-
move possible traces of ozone. However, no ozone was observed on the vapor
Prepurified nitrogen was obtained from Airco Company and passed
over hot copper to remove traces of oxygen before use as an inert atmosphere.
Oxygen was undetectable in the purified nitrogen by vapor phase chromatography.
G. Benzoic acid
Benzoic acid (U. S. P.) was obtained from Baker Chemical Company. It
was recrystalized three times from distilled water and dried for five hours at
room temperature under reduced pressure (m. p. 122.10).
H. Perbenzoic acid
Perbenzoic acid was prepared by the method of Braun (62) using the
modifications suggested by Kolthoff (63). An 85 per cent yield was obtained.
The white crystalline solid was dried for several hours under 10 mm. of Hg at
300 and stored at -5.
Prior to using the perbenzoic acid a sample was titrated to determine
the amount of active oxygen so that a correction could be made for decomposi-
I. Other chemicals
All other chemicals used were C. P. grade or better.
IV. Oxidation Apparatus
The oxidation apparatus used in this investigation is shown in Figure
1. The reaction flask was maintained at constant oxygen pressure by feeding
the gas from a supply tank when oxidation of the substrate occurred, A mer-
cury manometer controlled a relay that actuated a solenoid valve bleeding
oxygen from the supply tank to the reaction flask. The reaction flask pressure
could then be maintained to + O~ 5 mm, of Hg,
The pressure drop in the supply tank was used to measure the oxygen
uptake. This pressure was recorded by a Taylor "Fulscope" Recorder which
consists of a set of bellows that plot pressure as a function of time on a cir-
cular chart. The chart records the per cent of oxygen used,
The relationship between per cent oxygen and the actual moles of
oxygen was determined by two methods. One method consisted of determining
the volume of oxygen bled from the tank at various per cent readings by collect-
ing the gas over water. The second method utilized was to absorb oxygen from
:the tank into a solution of alkaline pyrogallol. The weight increase of the
pyrogallol solution was then determined.
The results of both determinations show that the pressure recorder is
linear and that one per cent on the recorder chart corresponds to 1.40 mM of
oxygen (Figure 2).
Figure 1. Oxidation apparatus
Figure 1. Oxidation Apparatus (continued)
List of Parts
1. Taylor "Fulscope" Recorder model 455M
2. Azco solenoic valve no. 6867E
3. Bellows type needle valve
4. Four mm. glass tubing
5. Oxygen inlet for flushing the system
6. Bleed valve relay
7. Vent valve
8. Thermometer and thermoregulator
9. Sample removal valve
10. Bath temperature relay
11. Stirrer pump
12. Calrod heater, 250 watt
13. Copper tubing
14. Inlet valve from oxygen cylinder
15. Pressure regulator manometer
16. Insulated constant temperature baths maintained within + 0.070
17. Oxygen supply tank
18. Pyrex reaction flask
19. Teflon magnetic stirrer
20. Waco stirrer motor, 300 rpm
0 by volume
* by weight
2 4 6 8 10 12
Figure 2. Calibration of recorder
V. Typical Autoxidation Run
A. Sample preparation
In general, 25 ml. of benzaldehyde was added to 200 ml. of glacial
acetic acid containing the metal catalyst. The total volume was then brought
to 250 ml. by adding more acetic acid. The sample was transferred to the
reaction flask which was connected to the system and immersed in the constant
temperature bath with the sample vent open. The system was flushed with oxy-
gen for two minutes at 5.0 liters per minute, after which the recorder was
started. One minute after starting the recorder the sample vent was closed.
Covering the bath prevented light from reaching the reaction flask.
Solid additives were weighed out on a Mettler balance (sensitivity
0.02 mg.) and dissolved in 200 ml. of acetic acid. Additive solutions were
pipetted into 200 ml. of acetic acid prior to adding the substrate.
All equipment was thoroughly cleaned before use to avoid contamination
from preceding runs in the following way:
After allowing the reaction flask to stand in concen-
trated sodium hydroxide solution for 24 hours it was
rinsed with tap water. This was followed by rinsing
six times with distilled water and six times with re-
distilled acetone. Purified nitrogen was used to dry
the flask and other equipment. Pipettes were
cleaned with three charges of acetone, three
distilled water rinses and six rinses with re-
B. Sample removal
Withdrawing samples after the run had started to determine the concentra-
tion of benzaldehyde and benzoic acid was accomplished by inserting a pipette
down the sample vent into the reaction mixture. The pipette was designed so
that it was small enough to fit through the sample valve. Samples were removed
only when the system was under one atmosphere.
C. Interpretation of autoxidation data
The data from the recorder chart were transferred to rectangular coor-
dinates. For example, the results of a typical run with 0.988 M benzaldehyde
and 8.4 x 10-4 M cobalt (II) ion at 500 are shown in Figure 3. The initial rate
was determined from this curve by plotting log(-dO2/dt) as a function of log [BzH]
and extrapolating to the initial benzaldehyde concentration, where -dO2/dt is the
oxidation rate and [BzH] is the concentration of benzaldehyde.
In a number of cases induction periods were observed. The length of the
induction period was found by the method of intercepts (61). A plot of a run
0.988 M benzaldehyde and 4.0 x 10"4 M cerium (IV) ion at 500 is shown in
Figure 4. The induction period, ti, was found to be 77 minutes.
.-.. ........._.. .__ i_ I
15 20 25 30 35
time (min. )
Figure 3. Typical autoxidation run ( 0.988 M BzH, 8.4 x 10-4 M Co(II),
500, 1 atm. 02 )
" 20 -
20 40 60 80 100 120 140
time ( min. )
Figure 4. Determination of the induction period ( 0.988 M BzH, 4.0 x 10-4 Ce(IV), 500, 1 atm. 02 )
VI. Analytical Methods
A. Peracid determination
Perbenzoic acid was estimated by titrating the iodine liberated from
the reaction between the peracid and potassium iodide (95) using the follow-
ing method: A five ml. sample was run into 25 ml. of 1.0 N sulfuric acid
containing 2.0 g. of potassium iodide and the liberated iodine was titrated
after three minutes with 0.984 N thiosulphate solution. A starch indicator
was used to show the end point.
B. Benzaldehyde determination
The benzaldehyde concentration was found using a Perkin-Elmer gas
phase chromatograph. Samples of constant volume (4 microliters) were in-
troduced into the column with a ten microliter syringe equipped with a device
to insure the delivery of equal sample volumes (Chaney adaptor). The area
under the benzaldehyde peak was found to be proportional to the concentra-
tion of the aldehyde and was used as a measure of the concentration for the
C. Product identification
Soluble products. The solution from an autoxidation was filtered to
remove any solid precipitate. The solution was then subjected to gas phase
chromatography. Retention times were used to determine what compounds
had been eluted from the column. The peak heights were used to estimate
the concentration of the products.
The reaction mixture was also subjected to paper chromatographic
analysis. The excess acetic acid was stripped off at reduced pressure and
a small amount of the solid residue was chromatographed using butanol-
pyridine-water (10:2:1) as the eluting solvent. The compounds were visualiz-
ed on the chromatographic paper by spraying sections of the paper with one
per cent ferric chloride, starch-iodide and ammoniacal silver nitrate solu-
Insoluble products. A yellow solid which precipitated from the man-
ganese ion catalyzed autoxidation of benzaldehyde was subjected to extensive
physical and chemical tests.
The results of the product identification will be discussed in the Re-
I. Manganese Systems
The results in this section show the effects of benzaldehyde, manganese,
water and acetic anhydride concentrations and oxygen pressure on the autoxida-
tion. Both manganese (II) and (mI) ions were used to catalyze the reaction. The
dependency of the rate on temperature was determined. Experiments were also
performed to determine the effect of the oxidation products on the reaction.
The metal ions were added by introducing a known volume of stock
catalyst solution in acetic acid or in some cases weighing out the salt as a solid.
The rate of oxygen uptake was determined in the oxidation apparatus and the
initial rate and the length of the induction period were then found by the method
described in the Experimental Section. Unless otherwise stated the reaction
temperature was 500.
Several general observations concerning the course of the reaction can
1. The color of the metal ions was undiscernible in the concentrations
used. However, just prior to the uptake of oxygen the solution turned brown.
This effect was most pronounced in the runs with induction periods. The pro-
duction of the brown color invariably preceded the start of the oxidation. The
brown color results when perbenzoic acid and manganese ions are in the system.
2. Oxygen uptake occurs smoothly in the presence of manganese ions
but stops abruptly at a time dependent upon the manganese ion concentration
and certain other factors.
3. At low concentrations manganese ions act as catalysts. However,
at higher concentrations this catalytic effect disappears and a retarding effect
A. The effect of manganese (I) ion concentration
A series of autoxidation runs at constant benzaldehyde concentration
and total volume was performed to establish the effect of manganese (II) ion on
the reaction (Table 2). An increase in the oxidation rate was observed as the
manganese ion concentration was increased from 2 x 10-5 M to about 2 x 10"4
M. A further increase in the manganese ion concentration resulted in a de-
creased autoxidation rate. These effects are shown in Figure 5.
An induction period, proportional to the manganese ion concentration,
was observed at manganese concentrations greater than 5 x 10"4 M (Figure 6).
The uptake of oxygen was observed to stop before all of the benzalde-
hyde was consumed. The limited uptake of oxygen, called the extent of
reaction, is expressed quantitatively as the number of moles of oxygen consumed
in the reaction. The extent of the reaction increased and then fell off as the man-
ganese ion concentration was increased as shown in Figure 7.
B. The effect of manganese (II) ion at low concentrations
From the data obtained using a wide spread of catalyst concentrations,
it is obvious that the effect of manganese ion is not a simple process, but possi-
bly a combination of both catalysis and inhibition.
EFFECT OF MANGANESE (II) ION CONCENTRATION
ON THE OXIDATION OF BENZALDEHYDE
( [BzH] o= 0.988, 500o 1 atm. 02)
[Mn (II)] d02/dta Extent of reaction ti
Mx 105 mM 02/min. mM 02 min.
a initial post induction oxidation rate
without acetic anhydrid
without acetic anhydride
2 4 6 8 10
[Mn(II)] x 104
Figure 5. The effect of manganese (II) ion on the oxidation rate of
benzaldehyde ( 0.988 MBzH, 500, 1 atm. 02 )
0.6 1.0 1.5 2.0
[Mn(II)] x 103
Figure 6. Induction periods for manganese catalyzed autoxidation of
benzaldehyde ( 0.988 MBzH, 500, 1 atm. 02 )
2 4 6 8 10
[Mn (II)] x 10
Figure 7. The effect of manganese (II) ion on the extent of reaction
( 0.988 MBzH, 500, 1 atm. 02 )
At manganese ion concentrations of less than 2 x 10-4 M the metal
ion catalyzes the autoxidation, A plot of the autoxidation rate against the
square root of the catalyst concentration was linear from the lowest man-
ganese (II) ion concentration, 2 x 10-5 M, to 2 x 104 M manganese ion, as
shown in Figure 8. This indicates that the manganese-initiated reaction is
one-half order in metal ion,
C. The effect of manganese (II) ion concentration
A series of experiments was performed using manganese (I) ions
rather than manganese (II) ions as a catalyst. The dependency on manganese
(I) was found by holding the concentration of benzaldehyde and the total vol-
ume constant while the catalyst concentration was varied over the range,
8 x 10-6 to 4 x 10-3 M. The results are summarized in Table 3.
In general, the autoxidation rate exhibited a similar dependency on
manganese (III) ions as it did on manganese (II). At low concentrations of
the metal ion (less than 4 x 10-4 M) the rate increased as the catalyst concen-
tration was increased, but it then leveled off and finally, at higher manganese
(III) concentrations became inversely proportional to the concentration.
The autoxidation rate using manganese (II) ion was significantly
greater, under the same conditions, then the rate using manganese (II) as a
catalyst. Thus, manganese (I) appears to be more effective in initiating the
autoxidation of benzaldehyde than manganese (II).
The effect of manganese (III) ion on the extent of reaction paralleled
that of the manganese (II) system. The extent of reaction at first increased
then decreased as the manganese (III) concentration was increased (Table 3).
1 2 3 4 5
[Mn (II)] 2x 105
Figure 8. The effect of manganese (II) ion concentration on the rate
of oxidation ( 0.988 MBzH, 500, 1 atm. 02 )
EFFECT OF MANGANESE (II) ON THE AUTOXIDATION OF BENZALDEHYDE
( [BzH] o=0.988, 500, 1 atm. 02)
[Mn(m)] -dO2/dta ti Extent of reaction
Mx 105 mM O2/min. min. mM 02
0.8 0.379 0 75.6
2.5 0.516 0 90.3
45.8 0.798 5 121.8
200 0.369 69 112
400 0.229 306 90
initial post induction rate
For equal catalyst concentrations, the extent of reaction was greater for man-
ganese (II) than for manganese (II).
The contrast in the extent of reaction between the two catalysts may be a
ramification of the relative efficiencies of initiation and/or inhibition by man-
ganese (II) and (m) ions.
Induction periods were observed at high manganese (I), ion concen-
trations (Table 3). They are somewhat shorter than the induction periods
observed in the manganese (I) systems.
In summary, it has been shown that:
1. Manganese ions display both inhibition and initiation of the autoxida-
tion of benzaldehyde with the latter predominating at low manganese ion
concentrations and both processes occurring as the metal ion concentration
2. At high metal ion concentrations the length of the induction period is
proportional to the manganese ion concentration, with manganese (TI) giving
longer induction periods than manganese (III).
3. Both manganese (II) and (III) behave in a similar manner, but man-
ganese (III) is the more effective catalyst.
D. The effect of perbenzoic acid
The dependency of the manganese ion catalyzed autoxidation of benzalde-
hyde on the concentration of perbenzoic acid was studied by adding known amounts
of the solid peracid at the start of the run and following the rate of oxygen uptake.
The results are shown in Table 4 and indicate that for low manganese ion con-
EFFECT OF PERBENZOIC ACID ON THE OXIDATION RATE
AT LOW CATALYST CONCENTRATIONS
(PzH% = 0.988, [Mn(n)]= 4.0 x 10-5, 500, 1 atm. 02)
Mx 104 mMOl2/min.
centrations (4.0 x 10-5 M) small amounts (less than 5 x 10-4 M BzOOH) of
perbenzoic acid will initiate the oxidation. However, this initiation reaches
a maximum at about 5 x 104 M perbenzoic acid and further increases in the
peracid concentration result in a decrease in the reaction rate.
A plot of -dO2/dt versus [BzOOH] 2 is linear over the concentra-
tion range between 1 x 104 and 5 x 10-4 M (Figure 9). This indicates that
there is a one-half order dependency on the initial concentration of perbenzoic
The following experiments were performed to examine the effect of
both high metal ion and perbenzoic acid concentrations. A reaction mixture
containing 2.0 x 10-3 M manganese (II) and 1.2 x 10-2 M perbenzoic acid had
an oxidation rate of 0.13 mM 02/min. and no induction period. An equivalent
run without peracid had an oxidation rate of 0.31 mM O2/min. and a 73 min.
induction period. This indicates that an initially high peracid concentration
combined with a high metal ion concentration results in a severe retardation
of the oxidation. A peracid determination on the above reaction shows that the
added peracid was quickly decomposed and that there was no appreciable build-
up of peracid during the reaction.
E. The effect of benzaldehyde concentration
The rate of oxygen uptake was followed for reaction mixtures varying
in benzaldehyde concentration, but all at 500, and containing 2.0 x 10-5 M
manganese (II) acetate.
I - I
1.0 1.5 2.0 2.5
[BzOOH 'x 102
Figure 9. Plot showing the dependency on perbenzoic acid
( 0.988 MBzH, 500, 1 atm. 02 )
The oxidation rate increased with the benzaldehyde concentration as
shown in Table 5 and Figure 10. A plot of log(-dO2/dt) versus log [BzH]
(Figure 11) was not linear suggesting that the order of the reaction in ben-
zaldehyde changes with concentration, being about 1.3 order at low benzalde-
hyde concentrations and 0.2 at higher concentrations.
F. The effect of oxygen pressure
The dependency of the autoxidation rate on the oxygen pressure was
examined by carrying out a series of runs at constant aldehyde and catalyst
concentrations at 500.
The results of these experiments show that within experimental error
the autoxidation rate was independent of oxygen pressure (Table 6).
G. The effect of water
A series of experiments to determine how added water would alter the
reaction was performed. Both the benzaldehyde and manganese (II) ion con-
centrations were held constant while various amounts of water were added to
the reaction mixture,
The autoxidation.rate was found to be inversely proportional to the
water concentration (Table 7). The rate falls off sharply as small amounts of
water are added; this effect then levels off so that a continued increase in the
water concentration results in only small changes in the rate. However, add-
ing water increases the extent of the reaction (Figure 12). Although no
quantitative determination was made it was observed that the amount of yellow
solid formed during the reaction decreased with increasing concentrations of
EFFECT OF INITIAL BENZALDEHYDE CONCENTRATION
ON THE OXIDATION RATE
( [Mn(I)] = 2.0 x 105, 500, 1 atm. 02)
[BzH] o -dO2/dt mM 02/min.
rate after an induction period of 366 min.
I I I I I
2 4 6 8 10
[BzHil x 101
Figure 10. The effect of initial benzaldehyde concentration on the
oxidation rate ( 2.0 x 10-5 M Mn(II), 500, 1 atm. 02 )
0.6 0.7 0.8 0.9 1.0
log ( [BzH] x 101 )
Figure 11. Plot of log dO2/dt versus log [BzH]
EFFECT OF OXYGEN PRESSURE ON THE OXIDATION RATE
( [BzH] =0.988, [Mn(II)] =4.0ox105, 500, 1 atm. 02)
EFFECT OF WATER ON MANGANESE CATALYZED
AUTOXIDATION OF BENZALDEHYDE
( [BzH] =0.988, [Mn(II)] =4.0x10-5, 500, 1 atm. 02)
H2 ] -dO2/dt Extent of reaction
M x103 mM O/mln. mM 02
- -- ^' -- 2
a [Mn(I)] =2.0 x 10-5
2.5 5.0 7.5 10.0
[H20l x 102
Figure 12. The effect of water on the extent of reaction ( 0. 988 M BzH,
4. 0 x 10-5 M Mn(II), 500, 1 atm. 02 )
water. No yellow solid was formed when the reaction mixture was 2.0 x 10-5 M
manganese (II) ion and 1.11 x 10-1 M water.
H. The effect of oxidation products
The effect of solid product. The solid product from an autoxidation run
was washed six times with glacial acetic acid and added to a run containing 0.988
M manganese (II) acetate. Both the rate of oxidation and the extent of reaction
were identical with that of a run containing equal substrate and catalyst concen-
The effect of soluble products. A run containing 0.988 M benzaldehyde
and 2.0 x 10-5 M manganese (I) ion, was allowed to proceed for 24 hours. At
this time a 10 ml. sample from this run was withdrawn and filtered and added to
a second run of equal substrate and catalyst concentration. The second run ex-
hibited a long induction period after which some oxygen was slowly absorbed by
the system. The maximum rate of oxidation for this run was 1.1 x 10-2 mM
O2/min. as compared with a rate of 3.5 x 10-1 mM O2/min. for a run of identi-
cal benzaldehyde and catalyst concentration.
The effect of benzoic acid. The effect of benzoic acid on the manganese
(II) ion catalyzed reaction was measured by adding known concentrations of the
acid to the reaction mixture. The results are shown in Table 8. These data
indicate that the oxidation rate was unchanged for different initial benzoic acid
I. The effect of catalyst on the peracid concentration
The concentration of perbenzoic acid was followed iodometrically during
the autoxidation of benzaldehyde at a number of catalyst concentrations. The re-
EFFECT OF BENZOIC ACID ON THE RATE OF OXIDATION
( [BzH ]o 0.988, [Mn(II) =4.0 x 105,. 500, 1 atm. 02)
M x 102
mM 0 mmn.
suits as shown in Figure 13, indicate that the peracid level increases steadily,
reaching a maximum, after which it slowly decreases. As the catalyst concen-
tration was increased the maximum concentration of perbenzoic acid obtained
during the run was lowered.
J. Peracid concentration during induction period
The build-up of perbenzoic acid was followed for a run that exhibited an
induction period due to the high concentration of manganese (II) ion. The reac-
tion mixture was 0.988 M benzaldehyde and 2.0 x 10-3 M manganese (II). The
peracid level was observed to increase slowly during the initial part of the ini-
tiation period, followed by a sharp increase at the start of rapid oxygen uptake
and then a slow decrease in concentration (Figure 14). The perbenzoic acid
build-up during the induction period appears to be exponential as shown in
K. The effect of temperature
Experiments carried out at different temperatures with reaction mixtures
containing 0.988 M benzaldehyde and 2.0 x 10-5 M manganese (I), show how the
autoxidation rate was changed with temperature (Table 9).
The Arrhenius plot of these data was found to be non-linear as shown in
Figure 16. Non-linearity can result if there are two competing reactions of
different activation energies occurring. The rate fell off at higher temperatures
suggesting that the inhibition reaction has a higher overall activation energy than
the initiation reaction.
4 Q 2.0 x 10-5
0 4.0 x 0-5
@ 2.0 x 10-3
_ r It
100 200 300 400 500 900 1000
time (min. )
Figure 13. Peracid concentration during the autoxidation at various
catalyst concentrations (0.988 M BzH, 500, 1 atm. 02 )
I I I I I I I I
40 80 120 160 200 240 280 320
Figure 14. Increase in peracid concentration during the induction
period ( 0.988 MBzH, 2.0 x 10-3 M Mn(II), 500, 1 atm. 02)
o. 6 0.
I I I I
20 40 60 80
time (min. )
Figure 15. Exponential build-up of peracid during the induction
period ( 0.988 MBzH, 2.0 x 10-3 M Mn(II), 50o, 1 atm. 02 )
CHANGE IN OXIDATION RATE WITH TEMPERATURE
( [BzH] =0.988, [Mn(II)] =2.0 x 105 1 atm. 02)
Temperature dO/dt mM O2/min.
3.0 3.1 3.2 3.3 3.4
1/T x 103
Figure 16. Arrhenius plot for manganese-catalyzed benzaldehyde
autoxidation ( 0.988 M BzH, 2.0 x 10-5 M Mn(II), 1 atm. 02 )
L. The effect of acetic anhydride
If acetic anhydride is added to the autoxidation run the effect of chang-
ing the manganese (II) ion concentration is altered as shown in Table 10. In
this experiment 1.0 ml. of redistilled acetic anhydride was added to the reac-
tion mixture before the start of the run and the rate of oxygen uptake was
followed at 500.
The effect of the change in the metal ion concentration on the oxidation
rate with and without added acetic anhydride is shown in Figure 5. The results
indicate that added acetic anhydride had little effect on rate of oxygen uptake
when the manganese (II) ion concentration was less than 2.0 x 10"4 M. How-
ever, at higher catalyst concentrations, the oxidation rate for runs with added
acetic anhydride was considerably higher than the rate without the anhydride.
M. Loss in benzaldehyde
The course of the reaction was followed by the loss in benzaldehyde, the
oxygen absorbed and the peracid formed. The analytical data for a typical run
are shown in Table 11.
The results show that oxygen uptake and loss of benzaldehyde stopped at
the same time; however, the concentration of perbenzoic acid was at a maximum
at the time the uptake of oxygen stopped, after which it slowly decreased as shown
in FIgure 17.
A change in the perbenzoic acid concentration resulting from a reaction
with benzaldehyde would not be reflected by a like change in the aldehyde concen-
tration, due to the insensitivity of the gas phase chromatographic determination
of the benzaldehyde.
EFFECT OF MANGANESE (I) WITH ADDED ACETIC ANHYDRIDEa
( [BzH] =0.988, 500, 1 atm. o.)
[Mn(n)] dO2/dt Extent of reaction
Mx 105 mM O0/min. mM O
4.0 0.349 44.6
4.0 0.349 51.6
8.0 0.419 99.0
20 0.509 139.5
40 0.465 120.0
100 c 0.409
al.0 ml. of acetic anhydride per 250 ml. of solution
no acetic anhydride added
Induction period of 6.0 min.
VARIATIONS IN BENZALDEHYDE,PERBENZOIC ACID CONCENTRATIONS
DURING THE OXIDATION
( [BzH] =0.988, [Mn(II)] =4.0x10-5, 500, 1 atm. 02)
Reaction time [BzH ] [BzOCH] 02 absorbed dO2/dt
minutes M M x 103 mM mM 02/min.
0 0.99 0.0 0.0 0.372
65 0.81 2.36 21.3 0.280
120 0.70 4.83 36.8 0.245
175 0.63 10.00 51.8 0.230
55 0.59 6.15 58.8 0.0
300 0.59 3.38 58.8 0.0
360 0.59 1.78 58.8 0.0
60 120 180 240 300 360
time ( min. )
1 atm. 02 )
of benzaldehyde ( 0.988 MBzH, 4.0 x 10-5 M
The same general results were obtained over the entire range of
catalyst concentrations (2 x 10-5 to 2 x 10-3 M Mn(I) ).
N. Some observations concerning the extent of reaction
Since the autoxidation stops before all of the benzaldehyde is consumed
the following experiments were performed to establish the factors involved in
1. The addition of fresh manganese (I) ion had no effect on the reac-
tion after it had stopped.
2. The addition of benzoyl peroxide and 2, 2'-azobisisobutyronitrile to
a reaction that contained unreacted benzaldehyde promoted a small uptake of
3. The addition of cobalt (II) naphthenate had a pronounced effect.
When the solution was made 1.2 x 10-4 M in cobalt a rapid uptake of oxygen was
observed after a 30 minute induction period. The oxidation rate after the addi-
tion of the cobalt solution was 0.20 miM 02/min. The benzaldehyde concentra-
tion at the point the cobalt (II) naphthenate was added was found to be 0.38 M.
The autoxidation rate of an independent run 1.2 x 10"4 M in cobalt at a benzalde-
hyde concentration of 0.38 M was found to be 0,256 mM 02/min.
The existence of the induction period indicates that when the cobalt (I)
naphthenate was added radicals were being produced but that an inhibitor was
preventing the initiation of the chain sequence. However, after exhausting the
inhibitor from the system the radicals introduced by the cobalt catalyst were
able to propagate the reaction chain at a rate nearly equal to that of an identical
reaction mixture without the inhibitor.
0. Product analysis of manganese-catalyzed benzaldehyde autoxidation
A number of runs were analyzed to determine the products of the reac-
tion. The products found include benzoic acid, perbenzoic acid, a trace of
water and a yellow solid.
Gas phase chromotography of the runs invariably gave four peaks. The
peaks were identified as oxygen, water, acetic acid and unreacted benzaldehyde.
Neither benzoic acid nor perbenzoic acid were eluted from the column.
The gas phase chromatographic analysis showed that water was pro-
duced in trace amounts which increase as the reaction proceeds. The final
water concentration estimated from peak areas was about 1 x 10-3 M for a run
containing 0.988 M benzaldehyde and 2.0 x 10-5 Mmangaganese ion and 5 x 10"4
M for a run 0.988 M benzaldehyde and 4.0 x 10-5 M manganese ion. Both runs
were at 500 and 1 atm. oxygen.
Paper chromatographic separations using pyridine-butanol-water as a
solvent gave three products, unreacted benzaldehyde (visualized with Tollen's
reagent, Rf = 0.13), benzoic acid (Rf = 0.96) and a trace of a phenolic or enolic
product (visualized with one per cent ferric chloride solution, Rf = 0.81).
The insoluble product from the manganese ion catalyzed autoxidation of
benzaldehyde had a melting point greater than 3500. Elemental analysis* showed
the compound to contain 70.66 per cent carbon, 3.96 per cent hydrogen and 25.38
per cent oxygen (calculated for CyH502: C = 69.42, H = 4.16 and O = 26.42 per
Analysis were performed by Galbraith Microanalytical Laboratories,
No ash was obtained on heating. The solid was soluble in 5.0 per cent sodium
carbonate solution and in dimethylformamide and insoluble in acid and in cer-
tain polar and non-polar solvents.
The infrared spectra of the substance showed absorption at 1025, 1060,
1200, 1350, 1450, 1590, 1680, 1730, 3070 and 3460 cm-1.
When the solid was reacted with hot 50 per cent nitric acid benzoic acid
was isolated from the solution. Reaction with methyl magnesium bromide gave
a gas and a second solid, which upon hydrolysis gave a product having no carbonyl
absorption band (1680 cm-1) in the infrared. The compound would not react with
potassium iodide under acid conditions, thus ruling out the possibility of it baing
Since acetic acid appears to be relatively inert towards the attack of
benzoyloxy radicals (64) it is unlikely that the solid was produced by such attack
on the solvent. There is, however, much evidence that indicates radical attack
on the aromatic ring occurs with ease (65, 66, 67).
On the basis of these observations the following structure is suggested
for the solid product:
O C-OH O
II. The Catalytic Decomposition of Perbenzoic Acid
The decomposition rate of perbenzoic acid catalyzed by manganese (II)
and (III) ions was studied in acetic acid in the absence of benzaldehyde. Stock
solutions in glacial acetic acid of perbenzoic acid and manganese (II) and (I)
acetate were used to make solutions of different concentrations for the kinetics
The rate of decomposition was followed by withdrawing 5. 0 ml. samples
at fixed time intervals and determining the peracid concentration iodometrically.
The samples were shaken in a 500 water bath. No attempt was made to exclude
oxygen from the system.
On adding either manganese (II) or (III) ions to the solution of perbenzoic
acid a light brown color formed with an intensity proportional to the metal ion
concentration. The same color remained during the entire reaction. A solid
precipitated from the mixture which had the same characteristics as the solid
formed in the manganese catalyzed autoxidation of benzaldehyde. Qualitative
examination of the reaction mixture after 90 per cent of the peracid decomposed
indicated that the amount of solid formed is proportional to the metal ion concen-
tration. No solid was observed in the uncatalyzed decomposition.
Paper chromatographic analysis of the soluble decomposition products
indicated that the only product formed was benzoic acid. The phenolic product
observed in the autoxidation reactions was not found in this case.
A. Order in perbenzoic acid
The logarithm of the measured perbenzoic acid concentration, log
BzOOH plotted against time gave a straight line as shown in Figures
18 and 19. This shows the decomposition to be first order in perbenzoic acid
for both the manganese (II) and (III) system.
B. Order in manganese ion
The rate of perbenzoic acid decomposition was followed for a series
of manganese (II) and (II) ion concentrations to establish the order in metal
ion. A plot of d log [BzOOH ] /dt versus catalyst concentration was linear as
shown in Figure 20. The decomposition of the peracid is thus first order in
The decomposition rate may be expressed by:
-d [BzOOH] /dt = k [BzOOH] [Mn] (36)
where [Mn ] is the concentration of manganese (I) or (III). The value of the
rate constant, kj, for the manganese catalyzed decomposition of perbenzoic
acid at 500 is 5.9 1./mole/sec, for the manganese (II) system and 4.6 1./mole/sec.
for the manganese (Hi) system.
A small quantity of gas was evolved during the decomposition. Gas phase
chromatography indicated that the gas evolved was carbon dioxide. The amount
of carbon dioxide was found to be independent of the presence of oxygen over the
system. About one per cent of the perbenzoic acid decomposed to give carbon
It should be pointed out that the iodometric method used in following the
concentration of the peracid would not distinguish perbenzoic acid from any ben-
zoyl peroxide formed during the reaction.
O. OM Mn(II)
2.28 x 10-6 M
2.28 x 10-5 M
3.43 x 10-5 M
20 40 60 80 100 12
time (min. )
Figure 18. Decomposition of perbenzoic acid by manganese (II) acetate
( 1.09 x 10-2 Mperbenzoic acid, 500 )
.14 x 10-5 M
_ 0. 0 M Mn(III)
2.28 x 10-6 M
2.28 x 10-5 M
3.43 x 105 M
20 40 60
80 100 120
time ( min. )
Figure 19. Decomposition of perbenzoic acid by manganese (III) acetate
( 1. 09 x 10-2 M perbenzoic acid, 500 )
1.14 x 10-5 M
10 20 30 40
[Mn] x 106
Figure 20. Comparison of manganese (II) and manganese (III)
catalyzed decomposition of perbenzoic acid
m. Cobalt Systems
The results in this section show the effect of cobalt catalyst and
benzaldehyde concentration on the autoxidation at 500. In the concentration
range used the color of the cobalt catalyst was indiscernible. However, after
the oxidation was in process the reaction mixture turned a faint tan color.
The uptake of oxygen was nearly quantitative if the reaction was allowed
to approach completion. In no case was the difference between the amount of
oxygen used and the amount of benzaldehyde consumed over 2. 0 per cent.
The only products detected were benzoic acid and the intermediate
perbenzoic acid. Starting with 0.245 moles of benzaldehyde, 0.240 moles of
benzoic acid were recovered.
A. The effect of cobalt concentration
The dependence of the oxidation rate on the cobalt (1I) concentration was
examined by carrying out a series of runs at constant benzaldehyde concentra-
tion but varying the cobalt (II) ion concentration from 5.0 x 10-5 to 1.2 x 10-3
A plot of the oxidation rate against the square root of the cobalt concen-
tration, [ Co 1/2, was linear indicating a one-half order dependency on the
cobalt (II) concentration. This is shown in Figure 21.
EFFECT OF COBALT CONCENTRATION ON THE RATE OF
( [BzH] o =0.988, 500, 1 atm. O2)
M x 104 mM 02/min.
I I I i
1 2 3 4
[Co()] x 10o2
Figure 21. The effect of the cobalt concentration on the benzaldehyde
oxidation rate (0.988 M BzH, 500, 1 atm. 02 )
No induction periods or retardation effects were observed in these studies.
B. The effect of benzaldehyde concentration
A series of experiments at constant catalyst concentration (3.0 x 10-5
M cobalt) in which the bonzaldehyde concentration was varied between 0.1 and
1.0 M was performed to determine the order in aldehyde (Table 13).
A plot of the oxidation rate against the three-halves power of the benzalde-
hyde concentration was found to be linear as shown in Figure 22.
The rate law for tte cobalt catalyzed benzaldehyde autoxidation becomes:
rate = k [Co ] 1/2 [BzH ] 3/2 (37)
where [Co ] is the concentration of cobalt (II) ion and [ BzH ] is the concentra-
tion of benzaldehyde.
This is exactly the relationship found by Bawn (8) at 250.
EFFECT OF BENZALDEHYDE CONCENTRATION ON OXIDATION RATE
([o(n)]= 3.0 x 10-5, 50, 1 atm. 02)
[BH] o -dO2/dt
I i I I I
2 4 6 8 10
[B.zH]/2 x 101
Figure 22. Variation of the oxidation rate with benzaldehyde concentration
( 3.0 x 10-5 MCo(III), 500, 1 atm. 2 )
IV. Cerium Systems
The effect of cerium salts on the autoxidation of benzaldehyde in acetic
acid was studied at 50. Cerium (IV) naphthenate and cerium (II) nitrate were
dissolved in acetic acid and these solutions were used in preparing the reaction
The solution was faintly yellow and turned to a straw color after oxida-
tion started. The only products isolated were perbenzoic acid, benzoic acid
and unreacted benzaldehyde. Figure 23 shows the course of a typical run. All
the benzaldehyde did not react. The extent of the reaction was found to be de-
pendent on the cerium concentration. This is shown in Table 14. Addition of
cobalt naphthenate (1.5 x 10-5 M) to the reaction after it had stopped caused a
rapid uptake of oxygen resulting in the complete oxidation of the benzaldehyde.
A. Induction periods
The effect of cerium ion concentration. A series of experiments in which
the concentration of Cerium (II) and (IV) was changed with the other parameters
held constant, showed the effect of the cerium concentration on the length of the
induction period. The results are shown in Table 14. The length of the induction
period was proportional to the cerium ion concentration up to 6 x 104 M, but the
length of the induction periods dropped off at higher concentrations, as shown in
60 120 180 240 300 360 420 480
time (min. )
Figure 23. Typical cerium catalyzed run showing both inibition and
retardation (0.988 M BzH, 8.0 x 10- M Ce(IV), 500, 1 atm. 02 )
EFFECT OF CERIUM ON THE OXIDATION OF BENZALDEHYDE
( [BzH = 0.988, 500, 1 atm. 02)
[Ce ] Oxidation d0/dt' ti Extent of reaction
M x 104 mM-O/min. mn. mM 02
-- 2 w.^--
_ ~~ ____
The effect of the oxidation state of cerium. The effect of both the con-
centration of cerium (II) and cerium (IV) on the induction period is shown in
Figure 24. Both oxidation states gave nearly equal induction periods at equi-
B. Rate of oxidation
The effect of the cerium (iI) and (IV) concentrations on the post induc-
tion period oxidation rate of benzaldehyde is shown in Table 14. and plotted in
Figure 25. Low concentration of cerium ions gave a very slight increase in the
oxidation rate, but by and large, the concentrations of cerium (III) and (IV) ions
had little effect on the post induction period oxidation rate of benzaldehyde.
C. Extent of reaction
The effect of the cerium ion concentration on the extent of the reaction
is shown in Table 14. The extent of reaction was decreased as the concentra-
tion of cerium ion was increased for both oxidation states of the metal ion.
D. The effect of added initiator
The effect of a radical initiator on the cerium ion catalyzed system was
examined by adding 4.3 x 10-3 M 2,2, '-azobisisobutyronitrile to a reaction mix-
ture containing 0.988 M benzaldehyde and 4.0 x 10-4 M cerium (HI) and following
the oxidation at 50. The induction period was found to be 48 minutes for this
run compared to 73 minutes for an equivalent run without the added initiator.
The post induction period oxidation rate for the runs with and without initiator
were nearly the same.
2 4 6 8 10
[Ce] x 104
Figure 24. The effect of cerium ions on the induction period
(0.988 M BzH, 500, 1 atm. 02)
0.4 0 Ce(III)
0.3 0 O
Figure 25. The effect of cerium ions on the post induction oxidation
rate (0.988 M BzH, 500, 1 atm. 02 )
The experimental results of this research are discussed on the
basis of current free radical theory. Particular attention is devoted to
the consideration of the role of manganese ions in benzaldehyde autoxi-
dation and to the manganese ion catalyzed decomposition of perbenzoic
acid. The effect of cobalt ions and cerium ions on the autoxidation are
considered and the different catalyst systems are compared.
I. Manganese System
The results of this research have shown that the manganese catalyzed
autoxidation of benzaldehyde is an extremely complex phenomena. Fortunately
the number of chemical species involved in this process is relatively small so
that a reasonable mechanism can be developed from the information obtained
in this work.
Prior to assembling the various bits of information into a mechanistic
picture, perhaps it would be advantageous to briefly summarize a few of the
important observed relationships. This summary can then be used as a guide
in developing the mechanism.
1. The reaction is first-order in benzaldehyde in the concentration
range 2 to 5 x 10-1 M and approaches zero-order as the benzaldehyde con-
centration is increased to 9 x 10-1 M.
2. The reaction is one-half order in either manganese (II) or (111) ions
at concentrations in the order of 1 x 10-5 M. However, as the metal ion con-
centration is further increased, the rate of oxygen uptake reaches a maximum
and then at still higher concentrations (1 x 10-3 M) decreases sharply. Induc-
tion periods appear at metal ion concentrations about 2 x 10-3 M.
3. The rate of oxidation is greater with manganese (III) than with
4. The reaction is one-half order in perbenzoic acid over the concen-
tration range 1 to 5 x 10 M. At high initial concentrations, perbenzoic acid
inhibits oxygen uptake.
5. In the absence of benzaldehyde, perbenzoic acid is decomposed more
rapidly by manganese (II) than by manganese (III) ions. The rate law for the de-
rate = k* [Mn ] [BzOOH] (36)
6. During an autoxidation the rate of benzaldehyde disappearance
parallels the rate of oxygen uptake and the rate of appearance of perbenzoic
acid. However, there is an abrupt halt of oxygen uptake and benzaldehyde con-
sumption which occurs when the perbenzoic acid concentration reaches a certain
point. After this the peracid concentration falls off sharply with no further oxy-
It is apparent from these results that the path of the reaction is altered
easily and frequently.
A. Kinetics of manganese catalyzed benzaldehyde autoxidation
Since manganese ions play the central role in this process it might be well
to spell out this role. It has been pointed out by others (14, 39, 43, 45) and veri-
fied here that manganese (I) and (III) ions can decompose perbenzoic acid to
produce radicals. These radicals are capable of propagating the chain reaction.
Low catalyst concentration. At concentrations where a half-order depend-
ence of the rate on manganese ion concentration is observed, the chain sequence
Mnlu + BzOOH -- BzOO* or BzO* (38)
BzOO- + BzH Bz- + BzOOH (39)
Bz* + 02 --BzOO (40)
2 BzOO* ----inert products (41)
The rate law derived from this reaction sequence using the steady-
state approximation is:
dO2 kb 'p12
dt 71/2 BZH] [BzOOH] 1/2 [ Mn] 1/2 (42)
Results of this work are in complete agreement with this equation.
Thus a one-half order dependence on manganese ion, a first-order depend-
ence on benzaldehyde (at low benzaldehyde concentrations), a one-half order
dependence on perbenzoic acid and a zero-order oxygen dependence have been
observed. It thus appears that the reaction sequence given above adequately
explains the behavior of the system in this concentration range.
The value for the combined constants from equation 42., kp kbl/2 kt1/2,
for the manganese catalyzed autoxidation of benzaldehyde is 4.9.1/mole/sec.
High catalyst concentration. At higher manganese ion concentrations
the oxidation rate is decreased and induction periods are observed. From
this qualitative evidence it appears that manganese ions are able to play a
part in a radical termination process. This is not a unique idea; it has been
proposed as the scheme by which metal ions inhibit various radical chain reac-
tions including both autoxidations and polymerizations (54, 55, 56, 57,.68).
A reasonable chain sequence involving the metal ion in the termination
reaction is as follows:
Mn+n + BzOOH ---BzOO- or BzO. (43)
BzOO. + BzH ---Bz. + BzOOH (44)
Bz. + 02. BzOO. (45)
Mn+n + BzOO* ---Pinert products (46)
Derivation of the rate expression for the above reaction sequence is
accomplished with the aid of the steady-state assumption* and is:
1, [~EH] -kb [ Mn] (47)
-dO2/dt=-- [BzH] [BzOOH] e
This equation may now be examined to see how well it fits the experi-
mental observations. The form of the equation is exactly that of the peak and
post peak portions of the curve in Figure 5. where the initial rate is plotted
against the manganese ion concentration at constant benzaldehyde concentra-
tion. It is in the range of manganese ion concentration represented by the
"falling-off" portion of the curve that inhibition by manganese ions is apparently
occurring. Recognizing that dO2/dt expresses the initial oxidation rate, so
that the benzaldehyde concentration is virtually constant, Equation 47. predicts
that the oxidation rate will be nearly independent of the manganese ion concen-
tration at low concentrations. In this range a change in the manganese concen-
tration will have little effect on the value of the exponential term. This
See Appendix I.
corresponds to the "leveling-off" portion of the oxidation rate versus curve
(Figure 5). However as the manganese ion concentration is increased to
higher values it will have a significant effect on the exponential term and
- dO2/dt will be sharply decreased as is observed.
This is good evidence for the termination by manganese ions. Since
manganese (nI) and (II) ions exhibit the ability to decrease the rate of oxida-
tion it is likely that both are capable of terminating chains unless a mixture
of these ions is formed during the course of the reaction as postulated by the
Specific rate constants. Since Equation 48 apparently describes the
phenomenon occurring during the reaction, some information concerning the
rate constants kp, 1t, and kb can be obtained from the data. The values for
kb, obtained from the slope of a plot of In rate versus [ Mn ] for a series of
experiments summarized in Tables 2 and 3 are:
Ib Mn(II) =1.0 x 103 1./mole/sec.
kb Mn(II) = 5.1 x 102 1./mole/sec.
Similarly values for [BzH ] kb kp/ktt can be obtained from the intercepts of
these plots*. From these values the ratios of kp to kty are found to be:
Mn)= 8.0 x 10-3 tMn() = 1.85 x 10-2
Assuming Ic has the value reported by Ingles and Melville (7) ktt becomes:
A value for the initial perbenzoic acid concentration must be assumed
here. The value chosen is 1 x 10-5 M. This was selected because 1 x 10-4 M
was shown to increase the rate and iodometric analysis indicates that the con-
centration was not over 5 x 10-5 M.
ktMn(II) =4.1 x 105 1./mole/sec.
kt'Mn(I) = 1.7 x 105 1./mole/sec.
Thus the manganese ion termination rate constants have been estimated.
Comparison of rate constants. An interesting comparison can be made
between the kb values obtained in the manner described above and in those ob-
tained by following the rate of decomposition of perbenzoic acid in the presence
of manganese ion but in the absence of benzaldehyde (see Results, Section II).
The values obtained in the latter case are:
kbMn(II) =5.9 1./mole/sec.
kMn() = 4.6 1./mole/see.
These values are considerably lower than those reported above. It will be re-
called that the decomposition process which occurs in the absence of benzalde-
hyde is a non-chain reaction involving only initiation and termination of the
radicals, there being no species present for propagation. Furthermore the
analytical method used to follow the decomposition rate will not differentiate
between the peracid and benzoyl peroxide which is a highly probable product
conceivably formed by the reaction:
2 BzO. -- (BzO)2
As a result of this the kb values are useful only in considering relative rates.
Thus, a comparison of the ratio of I~ Mn(I) to kb Mn (mI)
with that of kb Mn(II) to kb Mn(II) reveals that the relative decomposition rates
are nearly identical. This suggests that the radical products from the manganese
(II) ion decomposition are the same as those from the manganese (II) decomposi-