ON THE MANNICH REACTION
JACK EUGENE FERNANDEZ
A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE
DEGREE OF DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
UNIVERSITY OF FLORIDA
3 1262 08552 2893111111
3 1262 08552 2893
TABL' OF COIITEITF
I. IITRODUCTIOIT. . . . . . . . . 1
II. EXPERII TIITAL . . . . . . . . 6
A. General Considerations . . . . . 6
B. Description of Materials . . . ... 7
C. Rate Studies on the Reaction of 2,4,6-
Trinitrotoluone with Tetracthylmcthyl-
enediamino . . . . . . . . 7
D. Rate Studies on the Reaction of 2,4,6-
Trinitrotoluene with Totraethylmethyl-
enediamine . .. . . . . . 12
E. Rate Studies on the Reaction of 2,4,6-
Trinitrotoluene with Tetraothylmethyl-
enodiamine . .. . . . . . . 1
F. Attempted Rate Studies on the Mannich
Reaction of p-Bromoacetophenone with
Tetraethylmethylenediamine . . . . 15
G. Attempted Preparation of Diethyl-
amino-p-bromoacetophonone. . . . . 17
H. A Study of the Yields Obtained in the
Reaction of Acetone with Formaldehyde
and Various Secondary Amincs . . . 19
I. Attempted Mannich Reaction of Triphenyl-
nethane with Tetraethylmethylenediamine. 20
J. Attempted Mlannich Reaction of
Methylethylactophenone. . . . . . .21
K. A Study of the Reaction of Diethyl-
anino with Formaldehyde. . . . . .25
L. Attempted Preparation of
S Hydratropic Acid . . . . . ... .32
III. DISCUSSION OF RESULTS . . . . . . .36
A. Rate Studies on the Reaction of
Tetraethylmethylenediamine . . .. .36
B. A Study of the Yields Obtained
in the Reaction of Acetone with
Formaldehyde and Various Secondary
Amines . . . . . . . . . .37
C. The Storic Factor in the Mannich
Reaction . . . . . . . . .39
D. A Study of the Reaction of Diethyl-
anine with Formaldehydo. . . . . .41
IV. SUR-IARY. . . . . . . . . . . 41
V. BIBLIOGRAPHY. . . . . . . . . .47
VI. ACKIIOILERDGLNT ... . . . . .. .48
VII. BIOGRAPHICAL IIOTES. . . . . . . . .49
The Mannich reaction may be defined as the cordensa-
tion of ammonia, or a primary or a secondary aminc, with
formaldehyde and a compound containing a hydrogen atom of
pronounced reactivity due to proximity to an electron
withdrawing group. The original work by Mannich and his
coworkers was concerned primarily with aldehydes and
ketonos as the active hydrogen compounds but subsequent
work has broadened the scope of the reaction to include
carboxylic acids, esters, keto acids, acetylenes, nitro-
alkanes, nitro phenols, etc. The amino is usually employ-
ed as the hydrochloride, although in many cases the free
amine can be used to better advantage. The formaldehyde
is used either as the 37^ aqueous solution or as para-
formaldehyde whenever an organic solvent is cc'loyed.
From the above discussion it can readily be seen that
the Mannich reaction is one of great scope and utility.
Perhaps one of the most important applications of the
Mannich reaction in synthesis is in the production of
othylenic compounds by decomposition of the corresponding
Mannich bases. Thus unsaturated aldehydes, ketones, etc.,
may be prepared. The :annich bases ro al.o used as
The mechanism of the Mannich reaction has been the
object of much study in recent years and many views have
been brought forth, some of which are quite contradictor/.
Since the study of any reaction mechanism must always in-
clude a study of the intermediates involved, this portion
of the problem has probably received the most attention
and has been the object of the nost controversy. Bodendorf
and Koralewski (2) state that neither the methylol of the
amine nor that of the active hydrogen compound completely
satisfy the conditions of an intermediate. They also state
that the mothylol derivative of dinethylamine and pi::-
eridino could not be r.urified by distillation since they
readily decompose on heating. They have shown further
that these compounds were unstable in the presence of acid_
and would not form derivatives such as acetates or ben-
Johnson (3) and Scnlkus (4) have demonstrated the re-
action of aliphatic amines with formaldehyde and both
primary and secondary nitro paraffins. They have shown
that the same product is obtained by the use of either of
the following reaction sequences:
R2lTI / HCHO R2TCT20H
R2ICiiHO2H z HC (Ci 3)2I2 2: C.2C ( '1)2 1102 1120
or HC:0 ,/ :C(CH3c )2170 IIOCH2C(CI ) 202
HOCiI2C(CIT )2II02 R2 '2 C 2(C'3 )2 102 / H20
Johns-n indicates that the work < scrib d does not give
sufficient evidence for establishing tl.e mechanism of the
reaction. oe, however, suggests that :hen the n-thylol
of the nitro paraffin is used the initial stop is a decom-
position of the nothylol into formaldehyde and nitro
paraffin followed by reaction of the anino with form-
aldehydo. Here again the arinoraethylol worre neither
isolated nor identified.
Alexander and Underhill (5) studied the reaction cf
ethyllmalonic acid with formaldehyde and diomthylamine.
These authors concluded that (a) the reaction follows
third order kinetics in acid solution, first order with
respect to each of the three components; (b) the rate of
the reaction shows a critical dependence on pH, it passes
through a maximum at about pH 3.8; (c) under the conditions
of the experiment no reaction takes place between ethyl-
malonic acid and formaldehyde; and (d) smooth third order
curves are obtained for the reaction only if the amino and
formaldehyde are mixed and allowed to stand for 12 hours
before adding the ethylmalonic acid, but if the formalde-
hyde and amine are replaced by dimethylaminomethanol, the
reagents may all b? nixed at once. They postulate that
the reaction is initiated by a reversible condensation of
dinothylamine an] formaldehyde to give dirnethylaninomethan-
ol. A hydrogen bonded adduct of the aminomethylol and acid
HA could then attack the ethyinalonic acid, probably in
the enol form, to produce, by the elimination of water
and the conjugate base A", a protonated molecule of dicethyl-
aminomethylethylmalonic acid. Reaction of this species
with A-" would give the free Marnich base.
This mechanism does not soon plausible from several
points of view'. The formation of the salt of dimethyl-
acino, for example, is no':here taken into account. It
s:-ems unlikely that free dinethylmaine or dimethylamino-
methanol could exist in acid solution. Moreover the meth-
ylol compound is probably a poor choice for an inter-
mediate since it has been shown by the author (7) that
the principle product in the reaction of secondary amines
with formaldehyde is the mothylenedianine and not the
aminomnthylol. The author has also shown that the re-
action of amines with formaldehyde is quite rapid compared
to the Mannich reaction proper, hence the reaction of
anine with formaldehyde uould not be a rate determining
step as was postulated by Alexandor and Underhill. Lie-
berman and Wagner (6) were the first to propose the ncth-
ylenedianinc as an intermediate. That this compound is
a better choice for an intermediate was shoim by the
author's previous work (7) and will be further substan-
tiated lnter in this dissertation. Liob2rman and Wagner
state that the ncchanisn of the Hannich reaction prob.aly
involves the fornntion of the cation1 IpCC+ from the
nothylnodiomino or aminomothylol, and -ubsoquent co'-
bination of this ion vith the anion formed by thie removal
of a proton from the active hydrogen compound. Those
authors state that the formation of the cation is induced
by added acid or by the acidity of the active hydrogen
compound or both. Experimental results supported the
following inferences: (1) excessive acid interferes
with the primary condensation of amine and carbonrl con-
pound and depresses the ionization tendency of the active
hydrogen compound, and (2) excessive alkocli decreases
or prevents the formation of the cation, and therefore
obstructs or stops the reaction.
This mechanism is howovor inconsistent with the form-
ation of methylonediamino if this can be considered an
example of the Mannich reaction in which one molecule of
amine is the active hydrogen compound, sinae the removal
of a proton fr-n an amine to forn an amide ion requires a
tremendous amount of energy especially in acid solution.
In all other respects the mechanism seems to be consistent
with the experimental facts.
Since all the previous work on the mechanism of the
Mannich reaction has dealt with the reaction in aqueous
solution and with the acid-base relationships, ti'is work
was undertaken to study the reaction in nonaqueous and
nonpolar solvents. A study of the steric factors was also
undertaken since in previous work this problem has been
almost completely ignored.
A. General Considerations
In distilling the secondary amines used in this re-
search, a 20 inch column packed with Berl Saddles was used.
In all the other distillations a Claisen distilling appa-
ratus was used. Whenever the distillations were carried
out under reduced pressure, a Zimmerli gauge was used.
Temperatures are all uncorrected, and in keeping with
scientific works, they are all recorded in degrees centi-
B. Description of Materials
The following is a tabulation of the chemicals used
in this study containing the source, and, or, method of
2,4,6-Trinitrotoluono, Eastman Organic Chemicals, used as
Triphenylmethane, Eastman Organic Chemicals, used as re-
Sec-Butyl alcohol, Eastman Organic Chemicals, used as re-
Trioxane, DuPont, used as received.
Trioxymethyleno, Eastman Kodak Co., used as received.
Hydratropic aldohyde, Givaudan-Delawanna, Inc., used as
Karl Fischer reagent, Fisher Scientific Co.
Standard water in methanol solution, Fisher Scientific Co.
Diethylamine, Carbide and Carbon Chemicals Co., used as
Dibutylamine, redistilled, B.P. 159-600.
Diallylanine, redistilled, B.P. 1090.
N-Phenylpiperazine, rodistilled, B.P.
Morpholino, redistilled, B.P. 128.50.
Piperidino, redistilled, B.P. 1060.
Dibenzylamine, Eastman Kodak Co., used as received.
C. Rate Studies on the Reaction of 2,4,6-Trinitrotoluene
with Tetra ethylethylenediamine
(1) A 0.1000 molar TIT solution was prepared by dis-
solving exactly 22.714 g. of TIIT in enough toluene (dried
over calcium chloride) to make a total volume of one liter
at 230. A 0.1000 molar totraethylmethylonediamine solution
was prepared by dissolving exactly 7.9291 g. of the methyl-
onodianine (prepared according to the method described by
Lewis (8)) in sufficient toluene to make a total volume of
Fifty milliliters of each of these solutions, measured
from a 50 ml. volumetric flas:, wore mixed in a 100 ml. vol-
uretric flask and enough toluene added to bring the volume
up to exactly 100 ml. The solution was well shaken noting
the time and then poured into a 300 ml. 3 necked standard
taper flask fitted with a mechanical stirrer, thermometer,
and micro reflux condenser. The entire flask was immersed
in a water bath maintained at 23 0.10.
Samples of ca. 3 ml. were withdrawn at noted time in-
tervals and the reaction stopped in each by washing with
two 4 ml. portions of approximately 4 N hydrochloric acid
followed by one 4 ml. portion of distilled water. The
samples were then placed in glass stoppered weighing bottles
over calcium chloride and kept for analysis.
The optical density of each sample was determined di-
rectly from the infrared spectrogram.
In order to determine the extinction coeficient a in
the Bougeur-Beer equation (A = abc, whore A is optical den-
sity, a is the extinction cooficient, b is the cell thick-
ness, and c is concentration) several dilutions were made
from the previously described 0.1000 molar TNT solution
and the optical density determined for each. Table I lists
the values of optical density obtained from the 7.44 micron
band for each concentration of TNT used. A graph of optical
density vs. concentration is given on the next page. The
concentration determinations in this experiment were made
directly from the graph and not by substituting in the
equation itself. In this way, small deviations from lin-
earity were taken into account and a more consistent set
of values uero obtained. The true concentrations of TNT
0.03 0.04 0.05
were determined by plotting the values of CTrI (concen-
tration of TIIT) vs. time obtained from the infrared
spectrograms and extrapolating to zero time. The differ-
ence between this value and 0.0500 was subtracted from
each value to correct for the error which resulted from
the washing process.
CTUT A a
0.1000 1.32 66.0
0.0500 0.778 77.8
0.0400 0.647 80.9
0.0300 0.503 83.8
0.0200 0.342" 85.5
0.0100 0.169 84.5
0.00500 0.087 87.0
Average a 83.3 omittingg the 66.0 value).
In Table II are reproduced the data derived from this
experiment along with the values of the specific velocity
constants which were calculated from these data. The
equations used ore: kI .03log( C ), k2 1( )
t C x t x)
and k3 a (C--x)2--2) whoro k2, 13 are the first,
second, and third order specific velocity constants; t
time; C z concentration of THT and MDA (totraothylmothyl-
onediamino); and x a the amount of Mannich base formed.
SI I I I I
0 2 4 6 8 10 12
Second order curve
t ve. x
14 16 18
Figure II is a graph of the second order function x
plotted against time.
Sample Ho. Time (hrs.) (C x) kI k2 k3
1 ,0.5 0.0473 0.1012 2.28 39.0
2 1.0 0.0465 0.0690 1.51 26.0
3 2.08 0.0416 0.042 1.94 35.3
4 3.08 0.0394 0.0725 1.75 32.5
5 4.0 0.0376 0.0645 1.65 31.4
6 .5.42 0.0348 0.0620 1.61 31.7
7 6.42 0.0321 0.0641 1.74 35.5
8 7.42 0.0322 0.0546 1.50 30.4
9 .19 0.0271 0.0691 2.06 45.5
10 19.42 0.0193 0.0447 1.64 42.6
11 43.92 0.0110 0.0304 1.61 54.1
Average k2 1.39 liter/mole hour (determined graphically).
D. In this experiment the procedure of C. was repeated
with the exception that the initial concentration of MDA
was changed. This variation was introduced in order to
determine the order of the reaction with respect to each
component. If the reaction is second order with respect
to TI~ and zero order with respect to MDA, or vice versa,
the equation k2a )( ) would be obeyed. If the
reaction is 'first order with respect to each component, the
equation k2b 2.303-log10CMDA(CTT -x) would be obeyed.
t(CT1T1 CMDA) CTiT(CMDA -x)
The initial concentrations were CT0i = 0.0500 molar and
CMDA = 0.0250 molar.
Exactly 50.00 ml. of 0.1000 molar TMT solution, meas-
ured from a 50 ml. volumetric flask, was mixed with 25.00
nl. of MDA, measured from a 25 ml. volumetric flask. The
time was noted and the resulting solution was diluted to
exactly 100 ml. in a volumetric flask. The entire operation
was conducted at a constant temperature of 230. The solu-
tion was then poured into the reaction vessel and treated
as in C. Table III contains the data collected from this
Sample No. Time (hrs.) CTNT x CMA x k2a k2b
1 1 0.0482 0.0232 3.103 1.531
2 2 0.0479 0.0229 1.833 0.900
3 4 0.0462 0.0212 1.793 0.862
4 6 0.0420 0.0170 3.14 1.407
5 18 0.0317 0.0067 6.07 1.91
6 20 0.0318 0.0069 5,25 1.676
7 25 0.0300 0.0050 6.30 1.758
8 40 0.0280 0.0030 7.33 1.541
Average k2b = 1.69 liter/mole hour.
As in the previous determination, constant values for
C rNT -(7C1~,-x J
r I I I
k were obtained only from the second order rate equation.
The fact that the data conformed to the equation for k2b
and not to that for k2a indicates that the reaction is
second order and first order with respect to each com-
ponent. The graph of the k2b function against time is
given in Figure 111.
E. This experiment was merely a rerun of part C. and
was made as a check on the results obtained in that ex-
periment. The toluene, in this case, was redistilled
just before use. Both the 0.1000 molar TIIT and the
0.1000 molar MD1' solutions were prepared using the re-
distilled toluene. Fifty milliliters of each solution
were mixed in a 100 ml. volumetric flask and enough
toluene added to bring the volume to exactly 100 ml.
Samples uore then withdrawn and treated as in C. The
data are reproduced in Table IV. The second order curve
is shown in Figure IV.
F. Attempted Rate Studies on the Mannich Reaction of
p-Bronoacetophenono with Tetraothylmothylenediamino
In this study a 0.1000 molar p-bromoacotophonone
solution was prepared as described above for TIT. A
trial run employing 100 ml. of a solution which vas
initially 0.0500 molar in p-bromoacetophonone and 0.0500
molor in tetrnethylmethylenediamine was made in order to
determine the approximate rate of the reaction. Tuo
Second Order Curve
t vs. xC
L L A_1- L
20 25 30
samples vcre withdrawn and treated as described above
for T!Q at the and of 0.5 hours and at the end of 3.0
hours. The infrared spectrograms showed no appreci-
able difference in concentration of the ketone which
indicated that no reaction had occurred.
Before any further work was attempted on rate deter-
ninations, it was decided to attempt to prepare the
lannich base of p-bromoacetophenone since this com-
pound has not been reported in the literature.
Sample No. Time (hrs) ki k2 k
1 1 0.0782 1.37 57
2 3 0.0587 1.33 58
3 5 0.0622 1.53 73
4 7 0.0631 1.71 89
5 11 0.0536 1.51 85
6 23 o0.406 1.60 123
7 28 0.0365 1.56 160
8 49 0.0266 1.48 166
Average k2 x 1.56 litcr/molo hour.
G. Attempted Preparation of Diothylamino-p-bronoacoto-
phenone, 3.2 g. (0.02 mole) of tetract.ylmethylenedi-
amino, and 5 ml. of toluene was heated on a stean bath for
5.5 hours. The mixture was extracted with 4 N hydro-
chloric acid and the resulting solution neutralized
irth aqueous sodium hrdroxide. The oil layer whicl
separated was removed and the aqueous layer extracted
with toluene. The toluene extracts were added to the
original oil layer and the resulting solution dried
over Drierite and evaporated to remove the toluene and
diethylamine. Three grams of starting material was
(2) A mixture of 3.5 g. of the ketone, 3.2 g. of the
methylenediamine, and 5 ml. of 95% ethanol was warmed
on a steam bath for 5.5 hours. At this time 50 ml. of
ice water was added to precipitate the original ketone
and any Mannich base which might have formed. The pre-
cipitate was washed three times with 1:1 hydrochloric
acid and then with water leaving 2.8 g. of starting
material. No product was recovered.
A rerun of this experiment also failed to yield any
(3) A solution of 37.8 g. (0.19 mole) of p-bromoaccto-
phonone, 14.6 g. (0.2 molo) of diethylamine, 8.4 g. (0.28
mole) of paraformaldohydo, 14 ml. of con. hydrochloric
acid, in 80 ml. of 95% ethanol was refluxod for 10 hours
on a stonm bath. The ethanol was distilled off and the
residue washed with dilute hydrochloric to extract and
Iannich base pronent. Neutralization of this solution
with sodium hydroxide yielded no product. The oil remain-
ing after the acid washing described above Wias treated
with sodium hydroxide solution to decompose the Nannich
base hydrochloride which right have been insoluble in
water. Fractionation of this oil gave only a small
amount of starting material B.P. 96-1000 at 2.5 mm. and
mostly a black tarry residue.
A rerun of the above using 0.2 mole of amine hydro-
chloride in place of amine and con. hydrochloric acid,
after heating for 18 hours an-] treating as above, result-
el in 5 g. of material B.P. 98-90 at 2.5 am. 2,4-dinitro-
phenylhydrazone: M.P. 225-80; oxime: M.P. 122-5. This
material was p-bronoacetophenone. Only a black tarry
residue remained in the distilling pot.
H. A Study of the Yields Obtained in the Reactions of
Acetone with Formaldehyde and Various Sccondary Amines.
A mixture of 30 ml. (0.41 mole) of acetone, 8.4 g.
(0.28 mole) of parafornaldchyde, 5 ml. of 95" ethanol, and
0.19 mole of the secondary amine was allowed to stand at
room temperature for one week with frequent shaking. At
the end of this time 55 g. of powdered sodium bisulfite
was added with vigorous stirring and cooling. After stand-
ing in the ice bath for several minutes the mixttro was
filtered and the solid bisulfite addition compound decon-
posed by dissolving in dilute sodium carbonate solution.
The oil layer was removed and the aqueous layer extracted
four timze with cther. The co.-ie.d other extracts and
oil Inyor urao dried over an! y-'rous calcium zulfato, placed
under an aspirator, cndl finally unler a vacuun of 2 4 mn.
to re~ovo the otf.cr and unreacted naotone. The resulting
material :a::; taken to be the total yiold of Mannich base.
Table V is a tabulation of the amines usC. tozothor
ith the yields of Mannich base obtained in each case.
Amino Yiold of Mannich bane(c.) S Yield
Diothylnaine 0.6 2.2
Dibutylaline 10.5 27.8
Diallylamino 0.5 1.6
N-Phonylpipcrazsri 7.1 16.1
Morpholino 0.1 0.3
Flperidine 1.2 1.1
Dibonzylainoo 2.9 5.7
I. Attemteod l'annich reaction of TriphonylLlthanc with
To 103 C. (0.11 mole) of triher'elr.othano dissolved
in 30) ml. of dioxano uas oaded 67 G. (0.43 mole) of totra-
ct!'ylm-tl.ylonecdiaino in cno portion. Tho apparoat s con-
sisted of a on liter 3 noc;ed FT flnsa: fitted lith a
therno=-t-r, "tirr.r, andl reflux condeonor tho top of
wI ich wa- connected dou.nwarl by a ia .s tub2) into a large
test tube immersed in an acetone-Dry Ice trap designed to
collect any diethylamine which formed. The reaction
vessel was armed to 85 5o and maintained at that tem-
perature for 4 days. At the end of this time no reac-
tion had occurred, i.e., no diethylamine had collected
in the ico trap. One milliliter of 20 sodium hydroxide
was then added to determine the effect of a basic catal-
yst. Heating and stirring were then continued for an
additional 4 days. At this time no distillate had coll-
ected in the ice trap. Mlost of the dioxane n.as then re-
moved by distillation under diminished pressure, after
which the mixture was poured into ice-water. After one
recrystallization from ethanol, the precipitated solid
was filtered off and dried in a vacuum dessicator over
con. sulfuric acid, N.P. 87-900. The recovery of tri-
phcnylmethane was nearly quantitative.
J. Attempted Ilannich Reaction of Methylothylacetophen-
The mothylethylacetophonone used in third experiment
was prepared as follows by the reaction sequence:
(a) C3CH2CIIOCCH3 2 1OCl2 C:3CH2C1Ci- / S02 / HC1
(b) CHI CI' C CHc / 1e dry CH CHCICH(CH )MC
32 2 3ther
CiH CH2CI(CH )MgCl / CO2 C CH 3C!2CI1(CII3 )COO(eCl
CH3CI2CHI(CIF3)C00M1;C 1 .2 > Cyi3CH2CH(CH3)COOH /
(c) CH3 CH2 C(CH3 )COOH SCC12 CH 3CH2CH(CH )COC1
SO 2 HC1
CCH3CH(C2CH(C3 )COC1 / C6H---6 3 CH3CHI2CH(CH )COC6H6 /
(a) To a solution of 296 g. (4 noles) of soc-butyl
alcohol in 348 g. (4.4 nolos) of pyridine in a 2 1. 3
necked flask fitted with a thermometer, reflux condenser,
addition funnel, and mechanical stirrer, was added, while
cooling, 619 g. (5.2 noles) of thionyl chloride at such
a rate that the temperature never rose above 200. When
the addition :as complete, the ice bath was removed and
the mixture allowed to warm up to room temperature. The
flask was then warmed to 50-600 and kept at that tempera-
ture for one hour. The mixture was poured into ice water
and the oil separated. Distillation of this material
yielded 89 g. of sec-butylchloride, B.P. 66-80, (24' of
theory based on the alcohol).
(b) To 13.4 g. (0.55 mole) of magnesium turnings in 50
ml. of anhydrous cther was added a crystal of iodine and
3 g. of sec-butylchloride. The reaction was allowed to
proceed of its omn accord for 10 minutes, after which 50
ml. of other was added. The remainder of the 46 c. (0.5
mole) of the butylchloride in 300 ml. of other was added
dropwise. When the addition was comploto the mixture was
rcfluxcd for one hour. The flask was then cooled in an
ice bath and poured slowly into a beaker of Dry Ice sus-
pended in 100 ml. of other. The reaction was allowed to
proceed for 10-15 minutes and then 250 ml. of 1 : 1
aqueous sulfuric acid was added slowly to hydrolyze the
product. The ether layer was separated and the aqueous
layer extracted with two 50 ml. portions of ether. The
combined ether solutions wore washed with 150 ml. of 25'
sodium hydroxide to produce the sodium sa.t. The dissolved
ether was removed from the aqueous solution by warming on
a steam bath. The acid was then liberated by the addition
of con. hydrochloric acid and separated. Distillation of
this material yielded 29.7 g. of mcthylethylacetic acid,
B.F. 165-720, (58.3' of theory based on the butyl chloride).
(c) The acid chloride of the above material was prepared
by reacting 53.6 g. (0.53 mole) of the acid with 89.4 g.
(0.75 mole) of thionyl chloride for one half hour followed
by rofluxing for two hours and then distilling off the
excess thio-yl chloride. The resulting acid chloride was
added dropwise to a cooled, well stirred mixture of 160 g.
(1.2 moles) of aluminum chloride in 800 ml. (9 moles) of
benzene. Aft:r addition was complete the mixture was
stirred for an additional 2 hours at room temperature and
thon pou-od into 500 g. of ice and 125 ml. of con. hydro-
chloric acid. The aqueous layer was extracted with bon-
zone and the combined benzene solutions was ed with
sodium bicarbonate solution and then with water. Distill-
ation yielded 69.4 g. of methylethylacetophcnone, B.P. 106-
100 at 10 em. (81.7. of theory based on methylethylacetic
In the attempted preparation of the Mannich base of
mothylethylacetophenone, three modifications were used:
(1) no catalyst added, (2) acid catalyst added, (3) basic
(1) Into a 500 ml. 3 necked flask fitted iuth a thermo-
meter, stirrer, and reflux condenser was placed 68 g.
(0.42 mole) of methylethylacetophenone, 79 g. (0.5 mole)
of t-traethylmcthylonediamino, and 75 ml. of 95% ethanol.
The mixture was heated at reflux temperature for 16 hours.
The contents of the flask were then transferred to a
Claisen distillation apparatus and nost of the ethanol re-
moved (72-870). The residue was washed once with water and
then frattionated to yield 52.4 g. of ncthylethylacotophen-
one, B.P. 105-80 at 10 mm., (77& recovery of starting
material). Ho hiCher boiling material uas obtained.
(2) A mixture of 30.8 g. (0.19 mole) of the aceto-
phenone, 21 g. (0.19 mole) of diothylamino hydrochlorido,
8.4 g. (0.28 mole) of paraformaldchyde, and 25 ml. of
951 ethanol uas heated at reflux temperature for 24 hours
in tho apparatus described above. The mixture was then
treated with a solution of 15 g. of sodiun hydroxide in
75 ml. water to decompose the amine hydrochlorides. The
oil layer was removed and the aqueous layer extracted
several times with ether. The combined extracts and oil
layer was distilled to yield 23.5 g. of the original
ketone 7.P. 104-70 at 10 mm., (76.41 recovery of starting
material). Here again, no higher boiling material was
(3) The procedure of (1) above was repeated with the
exception that 0.5 ml. of 20% sodium hydroxide was added
at the beginning of the reflux period. This experiment
resulted in 27.3 g. of the original ketone, B.P. 104-60,
(885 recovery). No higher boiling material was obtained.
K. A Study of the Reaction of Diethylamine with Formal-
In this series of experiments the reaction between
diethylanine and formaldehyde was studied by measuring
the amount of water formed during the course of the re-
action -t21ini / HCIIO --- t21:ICH20H O A .t21TCH2IT t2 / 1120.
Exactly 1.50 molar trioxone solution was prepared by
dissolving 22.5 g. of trioxane in enough methanol (distil-
led from magnesium methoxide) to make a volume of 500 ml.
Similarly, 1.50 molar dicthylamino solution was prepared
by dissolvin- 54.75 g. of redistilled diethylamino (B.P.
56.50) in enough methanol to make a volume of 500 il. A
blank was determined for each of these solutions by titrat-
ing 25.00 ml. of each with Karl Fischer roagent which had
been standardized against standard water in r.othnnol sol-
ution (1 ml. 1 ng. water) immediately before use. It
should be noted here that in all the titrations which vare
performed, the Karl Fischer reagent was standardized just
The titrations were performed conductometrically
using the apparatus lhowun in Figure V where (A) are two
50 al. burets, (B) is a CaC12 tube, (C) is the titration
cell, (D) are two platinum electrodes, (E) is a magnetic
stirring bar, (G) is a galvanometer, (H) is a 1.5 volt
dry cell battery, and (R) is a variable resistance. The
end point was detected by titrating until the cglvanometer
needle showed a deflection which did not drift back to
the zero point.
(a) Tuonty five milliliters of each of the above solutions
wuro pipetted into the titration cell (a 250 ml. three
necked standard taper flask) and stoppered. The mixty're
was stirred for 30 minutes at room temperature by means
of a magnetic stirrer and then titrated with Karl Ficchcr
reagent. The data obtained are as follows:
Mg. uvater found. . . . . . . . .15.23
Mg. lat2r in oriCinal solutions (diethylamine) . 9.39
(fornmldohydo) . 2.77
MS. water formed in the reaction . . . . . 3.07
Millinoles water formed in the reaction. . . .. 0.177
Theoretical yield of water . . . . .. .18.75
Percent yield of water . . . . . . . 0.94
(b) Two determinations were made as follows. To 50.00
ml. of 1.50 molar diethylanino solution was added 25.00 ml.
of the 1.50 molar formaldehyde solution. The mixture was
stirred for 30 minutes and then titrated with Karl Fischer
reagent. When the end point had been reached, an addition-
al 25.00 ml. of the formaldehyde solution was added. The
resulting solution was stirred for 15 minutes and then
titrated. The first titration indicated that 0.0094
millimole of water had formed in the reaction. The second
titration indicated that 0.096 milliroles of water had
formed in the reaction.
(c) A mixture of 25.00 ml. of each of the original solu-
tions was allowed to stand at room temperature for 48 hours
and then titrated with Karl Fischer reagent.
Found: 0.46 millimolos of water had formed, (2.45% com-
(d) In order to determine the effect of the solvent in
this reaction it wns decided to carry out a series of ex-
periments similar to the ones described above using dioxane
as solvent. The diethylamine and formaldehyde solutions
were prepared exactly as doscribel above with the ex-
ception that dioxane was used instead of methanol. The
dioxone had been previously dried over calcium chloride.
A solution of 25.00 ml. of each of the 1.50 nolar
solutions was allowed to stand with frequent stirring
for 20 hours at which time it was titrated with Karl
Fischer reagent as described above.
Mg. water found. . . . . . . . . 124.42
:g. water in original solutions (diethylacine) . 63.51
(formaldehyde) . .5748
Mg. water formed in the reaction . . . . . 40
Millimoles water formed in the reaction. . .. . 0.19
Theoretical yield of water (millimoles). . .. 18.75
Percent yield of water . . . . . . .. 1.01
(e) Twenty five milliliters of each of the above solu-
tions were mixed and allowed to stand with frequent stirr-
ing for 72 hours. The solution was titrated with Karl
Fischer reagent and the fc]lowing data collected.
Mg. water found. . . . . . . . . 116.36
Mg. water formed in the reaction . . . . .-4.66
Millimoles water formed in the reaction. . . .-0.024
Percent yield of water . . . . . . . .
(f) Since the above experiments all resulted in only a
very scall degree of reaction, it is obvious that the rate
controlling step is that of the delolynerization of the
paraformaldehyde. Represented schematically:
HCHO / (C2HI5)2IH -- Products.
It was therefore decided to perform a series of experi-
ments using monomeric formaldehyde solution instead of
the solution of the trimer.
The formaldehyde solution was prepared by passing
formaldehyde gas ( Generated by heating 13 g. of trioxy-
methylene) into a weighed quantity of anhydrous methanol
in a 250 ml. volumetric flask. The gain in weight of the
flask was 12.00 g. (0.400 moles), so that the solution on
dilution to 250 ml. was 1.60 molar in formaldehyde.
A mixture of 25.00 ml. of this solution and 25.00 nl.
of the 1.50 molar mothanolic solution of dicthylanine was
stirred for one hour at room temperature and then titrated
with Karl Fischer reasnot as above.
Mg. water found. . .. .. . . . . . 216.2
Mg. water in original solutions.(diethylaminc) . 9.39
formaldehydee) . 27.77
Mg. water formed in the reaction . . . 179.0
Millimoles water formed in the reaction. .. .. 9.94
Theoretical yield of water . . . . . . 18.75
Percent yield of water ... . . 53.0
In the above eoperimont the end point faded very
rapidly and continued to do so until the theoretical
amount of water had reacted with the Karl Fischor reagent.
At this point 20.3 nillimolos of water had formed. The
first end point was taken as that which gave the first
large deflection of the galvanometer needle.
(g) Twenty five milliliters of each of the above solu-
tions were mixed and allowed to stand with stirring for
10 minutes. This mixture was titrated with Karl Fischer
reagent to determine the extent of the reaction in this
time interval. It vas observed that the first drop of
reagent caused a very large deflection of the galvanometor
needle. This end point faded very rapidly and every
subsequent addition produced the same effect until 50 ml.
had been added. At this point the titration was stopped.
The position of equilibrium could therefore not be estab-
lished in this experiment.
(h) Twenty five milliliters of each of the above solu-
tions was mixed and allowed to stand with stirring for
24 hours. At the end of this time the mixture was tit-
rated with Karl Fischer reagent to give the following
Mg. water found . . . . . . . . .201.4
Mg. water in original solutions (diethylamine). . 9.39
(formaldehyde). . 27.77
Ig. wat-r formed in the reaction. . . . . .164.2
Millimoles wter forme! in the reaction . . . 9.12
T':eoretical yield of winter. . . . . . . 18.75
Percent yield of water . . . . . . .48.64
In this experiment the end point (point of equilibrium)
was determined as in (f).
L. Attempted Preparation of Hydratropic Acid
(a) Permanganate Oxidation of the Aldehyde
Ninety nine grams (0.74 mole) of hydratropic aldehyde
was dissolved in 20-30 ml. of acetone. One hundred grams
of potassium permanganate was dissolved in water and added
dropwisc, with stirring, to the acetone solution. After
all the permanganate solution had been added, 600 r'. of
3N HC1 was ndded with stirring to decompose the MnO2. The
organic layer v;as removed u;ith a separatory funnel and
filtered through a sinterod glass crucible. It was then
treated with a saturated solution of sodium bisulfite to
remove any unroacted aldehyde. This mixture was again
filtered and the organic layer washed twice .:ith dilute
HC1 and then three times with water. This material was
then separated and dried in a vacuum desiccator over
CaC12 at ca. 1 am. overnight. Yield: 37.5 g. of a sub-
stance which was insoluble in sodium carbonate.
A rerun of the above experimer.t in which the tomper-
atrre was never allowed to exceed 500 resulted in a
material which was also insoluble in sodium carbonate
(b) Filver Oxide Oxidation of ITydratropic Aldehyde
To a cooled suspension of 108 g. of silver oxide in
100 ml. of dilute sodium hydroxide solution was added
slowly 120 g. of hydratropic aldehydo. After refluxing
the mixture for one hour, it was filtered and the filtrate
acidified with dilute sulfuric acid. The oil layer was
separated and the water layer extracted with ether. The
combined ether extracts were added to the oil layer and
the resulting solution dried overnight over Drierite.
After removing the ether under reduced pressure, the prod-
uct was distilled under reduced pressure to yield 18 g.
of material B.P. 160-30 at 24-5 mm. It was noted that
considerable decomposition occurred during distillation
so that it varn decided not to purify further runs by this
Several subsequent runs of this reaction using various
modifications of the conditions failed to pro-uce ar:
hydratropic acid and this method a'.a abando- 1. It was
believed that hydratropic acid might be prepared Via the
nitrile so the following experiments were performed.
(c) Alpha-chlorocthylbenzon2 vas prepared 4y passing dry
hydrogen c'lorido into 10) g. of freshly distilled styrene
(B.P. 400 at 13 nm.) cooled in an aceto e-Dry Ice bath
until the gain in ;eight of the reaction vessel and its
contents had reached 50 g. After the mixture had been
allowed to varm up to roo~ temperature it was distilled
to yield oc-chloroethylbenzene, B.P. 60-710 at 10 mm.
To a solution of 34.5 g. (0.53 mole) of KCli in a
mixture of 153 ml. of ethanol and 61 ml. of water at the
boiling point was added dropwise 63.2 g. of the chloro-
ethylbenzene. After addition was complete the mixture
was allowed to reflux for an additional hour. After
cooling, the mixture was separated and the water layer
extracted several times with ether. The combined extracts
were added to the organic layer, heated with charcoal,
filtered, and dried over Drierite. The ether was removed
under diminished pressure and the dark brown liquid
which remained was used without further purification.
The liquid obtained above was hydrolyzed by heating
under reflux with a mixture of 100 ml. of water and 100
ml. of con. sulfuric acid for two hours, then pouring
into a beaker of ice water. The oil uas separated, washed
once with water, dried, and distilled to yield 10 g. of a
clear liquid, B.P. 140-5 at 3-3.5 am. This material was
found to be insoluble in sodium bicarbonate solution.
Many subsequent trials of this experiment, using also
the bromo-derivative, failed to yield any hydratropic acid.
(d) To 14.6 g. (0.6 mole) of magnesium turnings in 100 ml.
of anhydrous other in a 3 liter flask ws added a few
milliliters of (-bromoethylbenzcne and a crystal of iodine
to initiate the reaction. The reaction was allowed to
proceed of its own accord for 13 minutes after which tino
the remainder of the 111 g. (0.6 mole) of the bronocthyl-
benzene was added through a separatory funnel while cool-
ing the flask in ice water. The reaction mixture urns
refluxed gently for 1.5 hours and then pieces of Dry Ice
were aided in excess while stirring. After about 15 min-
utes, dilute sulfuric acid was added to hydrolyze the
product and dissolve the magnesium. The organic layer was
separated and the aqueous layer extracted several times
with ether. The combined other solution was treated with
dilute sodium hydroxide and the resulting water solution
neutralized with hydrochloric acid to liberate any hydra-
tropic acid. No product resulted from this treatment.
Evaporation of the ether solution, however, loft a mater-
ial. which was partially liquid and partially solid. Some
physical constants of these products are given here.
Liquid fraction: B.P. 136-4o0 at 10 mm.
Solid fraction: M.P. 125.5-127.5P, molecular eight 212.
These products are believed to be isomeric diphenyl-
butanes. The literature (10) Gives for 2,3-diphenyl-
butane a B.P. of 1400 at 10 em., and for 2,2-diphenylbutane,
a M.P. of 127-80. The molecular weight of the diphenyl-
butanes is 210.3.
III DISC"S "IO~ OF R .CULTS
A. Rate Studios on the Reaction of 2,4,6-Trinitrotoluene
with Tetraethylmethylened iamine
The determination of the order of this reaction was
made by assuming each rate law and plotting the appropriate
function based on that particular equation against time.
The equation which resulted in a straight line was taken
as the one which satisfied this reaction. In this way,
assurance could be had that the data did not fit the equs-
tions which were discarded.
The rate studies on the reaction between TNT and tetra-
othylmethylenediamine which wore described in the experi-
mental section of this dissertation all led to the con-
clusion that the reaction obeys the second order rate law.
The results obtained by the use of unequal concentrations
of the starting natarials indicate that the reaction is
first order with respect to each component. The values of
the rate constants determined from the slopes of Figures
II, III, and IV respectively, are tabulated belou.
Experiment Initial Concentration Fecond Order
Ho. TiT IDA K liter/mole hour.
1 0.0500 0.0500 1.39
2 0.0500 0.0250 1.69
3 0.0500 0.0500 1.56
The mechanism of this reaction can then be represented
simply ns the combination of one molecule of T!'T with one
of MDA to form an activated complex uhich can break down to
form a molecule of Mannich base and a molecule of the
secondary amine. This can be represented schem-tically as
N02 ;(C2Hq)2 1102 ------N(C 2 )2
"20 /r- 2NO 2
1102 1 I(C25 5)2 ;02 iI(C 2 2
-- !02 7 I2CH2CNH21(C2H5)2 / HIT(C2H5)2.
The prediction of an activated complex as shown here is
explained on the basis of the results obtained from the
experiments on triphenylmethano and nothylethylacetophenone
which will be discussci later in this section, (cf. pp.
B. A Study of the Yields Obtained in the Reactions of
Acetone with Formaldehyde and Various Secondary Amines
The percent yields of Hannich base obtained from the
reaction of acetone with formaldehyde and some secondary
amines are tabulated in Table V of the experimental sec-
tion. These values are reproduced here alon- ith the
ionization constants of the amines in order of ,ecr-asing
Amine Fercent Yield Kb
II-Phenylpiperazine 16.1 ..........
Piperidine 4.1 3.40x10-l'
Diethylamine 2.2 2.26x10-
Dibutylanine 27.8 9.94x10-5
Diallylamine 1.6 3.12x10-6
M-orpholine 0.3 3.02x10-6
Dibenzylamino 5.7 1.48xl0-6
With the exception of !-phenylpipor zine whose ioniza-
tion constant could not be found, and dibutylamine and di-
bcnzylamino which do not conform, a correlation of percent
yield with Kb is quite obvious. That is, the yield of
Mannich base appears to fall off as amines of lower basic-
ities are used. This observation is completely in line
with the author's previous :ork (cf. reference 7) i-rich
indicated that the formation of methylonediamine was favored
by low basicity of the amino. This was explained by the
fact that the cnCrry requirement in removing a hydrogen
atom front an amine is louer the rcaklr the amine is as a
base. The phenorncnon un-lr dicc- scion hore can be cx-
plained by the same line of reasoning. Since the reaction
of a methylonediamino with acetor.- is dependent on the
rupture of the C-It bone, tlo reaction is on'anced by
instability of the nethylenedianine. The Mannich reaction
of acetone with secondary anines is therefore aided by a
high amine basicity since highly basic anines would, in
the course of the reaction, produce methylenediamines of
lower stability and therefore higher reactivity toward the
C. The Steric Factor in the Mannich Reaction
Thoe conditions under which attempts were made to pre-
pare the Mannich base of triphenylmethane are described on
pages 20-21. It was found that no reaction occurred on
vigorous :arming over a period of 8 days with or without a
basic catalyst. The conditions under which the prepara-
tion of the Mannich base of methylethylacetophenone was
attempted are described on pagcs 24-25. This reaction also
failed to proceed regardless of the nature of the catalyst
used or the length of time of heating.
It is believed that the failure of these reactions to
procei as expected can be explained on the basis of steric
considerations since presumably the hydrogen atoms involved
are labile enough for reaction to occur. Furthermore, on
the basis of this argument, the intermediate can be predicted.
Fishcr-Iershfolder-Taylor models were made of the above
compounds. By removing a hydrogen atom from the methylene
carbon of a tetraothylmothylenediamino model, an attempt
was made to join this group to the models of t.e above con-
pounds to give respectively
(C2 ) 2 --(C )2 (C2H ) 2H--"-- (C2"5)2
CT 6"5- 6n and C- C--ig
These r.odels uere found impossible to construct. If the
assumption is made that the nethylenediamine is the nec-
essary intermediate, the failure of these compounds to
react can be explained on a steric basis.
A nodel of diethylaminomethanol was constructed and
a hydrogen atom removed from the methylene carbon atom.
Adducts of this group with the above compounds represented
(C2 I5 )2-C -OH (C2H )21 -C:--oI
C 65-C-C -615 and CH3-- --C-25
were readily constructed which shows that if steric hin-
is the crucial factor in these reactions the nothylol is
not the intermediate. The true intermediate is then, based
solely on steric considerations, the mothylonediamino.
That methylothylocetophenone should react under the
conditions of the IMannich reaction were it not for steric
hindrance is suggested by the fact that n-butyrophonone does
react under these conditions cf. reference 11). The model
of the adduct of tetraothylnethyl-nedianinc and n-butyro-
phenone was found to be quite readily assembled. Photo-
graphs of the models of triphenylmothane and mnthylethyl-
acetophenono together with that of tetraethylmethylene-
dianine are shown on the following pages. Also shorn are
photographs of the models of the tetraethylmathylonedi-
anine-diphenylacetonitrile adduct, diphenylacetic acid,
and phonylcyclohexylacetonitrile. The Mannich base of
diphenylacotonitrile with dimethylamino has been prepared
(cf. reference 9) but these authors state that those of
diphonylacetic acid and phonylcyclohexylacetonitrile could
not be prepared. It is worthy of consideration that the
atomic models of the tetraethylmothylonediamine-diphenyl-
ac2tic acid and tetracthylaothylonediamine-phenylcyclohexyl-
acetonitrile adducts could not be assembled, since this
fact is in accord with the steric requirements postulated
Although the nasertion that the methylenediamine is
the necessary intermediate in these exao.les of the Iannich
reaction is of course not definitely proven by the above
facts, the evidence does point very strongly in that direc-
D. A Study of t' Reaction of Diethylamine with Formal-
A summary of the data obtained from this series of
experiments is Given in Table VI.
. .-........... .
Diphenylacetic Acid and
nitrile and Tetraoethyl-
__jiih ^^^^*i ^
f ... .............. ...........
D r-I CM \0 lN r- *
0 0 O CM 4- 0 0
(-1 * *
P r i
GI-H O 0 4- H 0
S 3* * *
S 0 00 0 0 C 0
S l tN r co 0 o
*C 0 H ON CM O O0 0 C
0 * *
oO r-1 C-O '0 0
O \ V\ O O O O O
43 *I 0 0
> *Cr O O CO O C r-1 C
S* o *0
E-l r-l r l r-l r- E r-l
43 0 0 0 S
+ tr\ O lp\ tr\ 1f\ tri tf1
H re rl r4 r 0 r rl rl 3
o 0 80 0
0 z sIP\ t br\ z \ 1R C
C )C M C3 C C1 C C1 Oc
4 0 0 O
0 0 *r 0
0 H 0c cr tr\ '0 N
The information which has been derived from these
experiments is based on the following reaction which has
been proposed for the reaction of a secondary amine with
formalJehyde (cf. reference 7 ).
(C2H5)2t1H / HCHO (C2H )211CH20H
i-(C2H )2NCH2i(C2H )2 / iHC-0 / -H20.
By titrating the water formed during the course of the
reaction the amount of methylenediamine is known since the
molar quantities of wator and methylenediamine are equal.
The values determined in Experiments 6 and 7 Given in
Table VI are of course only approximate since the end points
were not sharp. These values therefore indicate mnrely
the approximate equilibrium position with respect to the
totraethylmethylcnediamine. INo information can be derived
from these experinmnts concerning the part played by the
aninomethylol. It is knorn that approximately 50 of
methylcnediamine had fornmd at equilibrium, but the rela-
tive quantities of the oth-r materials ware not known.
A point of importance which has been established by
thase experiments other than the position of equilibrium
with respect to tetracthylamthylenadiamine ir the fact that
the reaction proceeds quite rapidly to a point of very
dynamic equilibrium. It was impossible? to determin- the
velocity of the reaction by this method but it can be
assumed that it is quite rapid since the deterir.ations
made after 1 and 24 hours respectively resulted in essen-
tially the game value for the extent of the reaction. From
this, the assumption that the rate determining step in the
Mannich reaction is the reaction of the amine-formaldehyde
product with the active hydrogen compound is completely
justified. The phrase "amino-formaldehyde product" is
used here because it has not been definitely established
that the aminomithylol cannot serve as the intermediate,
although part C. of this section gives strong evidence
against this possibility.
In the case of experiments 1, 2, 3, 4, 5 of Table
VI the rate determining stop was depolymcrization of
trioxane to produce monomeric formaldehyde.
A study has boon made of the rate of the reaction of
2,4,6-trinitrotoluene with tetraethylmcthylcnediamino in
toluene solution. These studies led to the conclusion
that the reaction follows second order kinetics, first
order with respect to each component. A mechanism has
been postulated for this reaction which involves the com-
bination of one molecule of each component to form an
activated addition complex which can decompose into the
:annich base and diethylamine. The postulate of an acti-
vated addition complex between the trinitrotoluene and
methylonedianine is substantiated by a study of the steric
factors which enter into the reaction. These studies
based on the theory of steric hindrance point to the meth-
ylenediamine as the necessary intermediate. Evidence has
also been found which leads to the conclusion that the
lMannich reaction is enhanced by high amine basicity.
The reaction of diethylamine with formaldehyde has
been studied and found to lead to a 50 formation of the
methylonedianino a' equilibrium. Although the rate of this
reaction was not determined, it was found to be quite high.
Equilibrium is probably attained in a matter of minutes.
This conclusion justifies the assumption that tlo rate deter-
mining stop in the I'annich reaction is the reaction of the
mnthylcncdiamine with the active hydrocon compound.
Several futile attempts uere made to prepare hydra-
tropic acid. The methods which were used are described.
(1) Blicke, Organic Reactions, Vol. I, 303.
(2) Bodendorf and Koralewski, Arch. Pharn., 271, 101-16
(3) Johnson, J, An. Chen. Soc., 68, 12-18(1946).
(4) Senkus, J. An. Che-. oc.,, $8, 10-12(1946).
(5) Alexander and Underhill, J. Am. Chem. Soc., 21, 4014-
(6) Liebornan and :.agncr, J. Or,. Chcm., 14, 1001-12(1949).
(7) Fernandez, M. S. Thesis, University of Florida, 1952.
(8) Lewis, Ph.D. Dissertation, University of Florida, 1952.
(9) Zaugg, Horrom, and Vernsten, J. An. Chen. Toc. 5, 288
(10)Feilbrcn, Dictionnr:- of Orranic Compounds, Vol. II,
(11)Ruddy and Buckley, J. An. Cheo. Soc., 2,, 718-21(195C).
The author wishes to express his sincerest appre-
ciation to all the members of his supervisory committee,
and especially to Dr. George B. Butler who supervised
this research project. His friendship, guidance, and
constant inspiration made this work possible.
The author is also indebted to all the other men-
bers of the staff and to his follow students who aided
him greatly throughout the course of his research.
Thanks are also given to the author's wife for
her patience and encouragement.
Jack 3. Fornnndoz was born in Tanpa, Florida on May
18, 1930. He entered the University of Florida in Sep-
tember, 1947 whero he held the positions of student assis-
tant and Research Corporation assistant as an undergrad-
He received the degree of Bachelor of Science in
Chemistry in Juno, 1951 and in the same year entered the
Graduate School of the University of Florida where he has
hold a U. S. Ilavy research assistantship ar.n an Army
Ordnance research assistantship. He receiv-d the degree
of pastor of Science in August, 1952.
He is a member of the Beta Alpha Chapter of Gna
Sigma Epsilon, Sigma Xi, and the American Chemical
This dissertation was prepared under the direction of
the candidate's supervisory committee and has been
approved by all members of the committee. It was submit-
ted to the Dean of the College of Arts and Sciences and to
the Graduate Council and was approved as partial fulfill-
ment of the requirements for the degree of Doctor of
August 9, 1954
Dean, College of Arts and Sciences
Dean, Graduate School
k/ .J ," 1