Title Page
 Table of Contents
 List of Tables
 The reaction of silver salts of...
 The reaction of silver salts of...
 Preparation of 2-trifluorometh...
 The use of the grignard reagent...
 Biographical data

Title: preparation of certain fluoroolefins containing the trifluoromethyl group.
Full Citation
Permanent Link: http://ufdc.ufl.edu/UF00091633/00001
 Material Information
Title: preparation of certain fluoroolefins containing the trifluoromethyl group.
Series Title: preparation of certain fluoroolefins containing the trifluoromethyl group.
Physical Description: Book
Creator: Taylor, Robert Edward,
 Record Information
Bibliographic ID: UF00091633
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: alephbibnum - 000549785
oclc - 13297656


This item has the following downloads:

Binder1 ( PDF )

Table of Contents
    Title Page
        Page i
    Table of Contents
        Page ii
    List of Tables
        Page iii
        Page iv
        Page 1
        Page 2
        Page 3
        Page 4
        Page 5
        Page 6
        Page 7
        Page 8
        Page 9
    The reaction of silver salts of perfluorocarboxylic acids with olefins
        Page 10
    The reaction of silver salts of perfluoroacarboxylic acids with olefins
        Page 11
        Page 12
        Page 13
        Page 14
        Page 15
        Page 16
        Page 17
        Page 18
        Page 19
        Page 20
        Page 21
        Page 22
        Page 23
        Page 24
        Page 25
        Page 26
    Preparation of 2-trifluoromethylbutadiene-1,3
        Page 27
        Page 28
        Page 29
        Page 30
        Page 31
        Page 32
        Page 33
        Page 34
        Page 35
        Page 36
        Page 37
        Page 38
        Page 39
        Page 40
        Page 41
        Page 42
        Page 43
        Page 44
        Page 45
        Page 46
        Page 47
        Page 48
        Page 49
        Page 50
        Page 51
        Page 52
    The use of the grignard reagent in the preparation of some trifluoromethyl styrenes
        Page 53
        Page 54
        Page 55
        Page 56
        Page 57
        Page 58
        Page 59
        Page 60
        Page 61
        Page 62
        Page 63
        Page 64
        Page 65
        Page 66
        Page 67
        Page 68
        Page 69
        Page 70
        Page 71
        Page 72
        Page 73
    Biographical data
        Page 74
        Page 75
Full Text

ACKNOWLEDGEMENTS ................ iv
INTRODUCTION ...................... 1
Discussion I ................... 10
Experimental I................... 20
Discussion II. ................. 27
Experimental II.................. 40
STYRENES...................... 53
Discussion III................... 53
Experimental III.................. 60
SUMMARY......................... 69
BIBLIOGRAPHY...................... 71
BIOGRAPHICAL DATA................... 74

I Hydrocarbon Esters of Ferfluorocarboxylic Acids ..... 26 II Results of Pyrolysis Studies of Trifluoroacetone
Cyanohydrin Acetate.................. 51
HI 2-Trifluoromethylbutadiene-l, 3 and Intermediates. .... 52
IV Substituted Styrenes and Benzyl Alcohols......... 68

The author wishes to express his gratitude to the co-directors of this research, Dr. Paul Tarrant and Dr, George B. Butler, without whose guidance this investigation would have been impossible. Appreciation is also extended to the other members of the author's supervisory committee for their aid and encouragement.
The author wishes to thank Miss Mary Louise Van Natta for her excellent cooperation in obtaining infrared spectra of the compounds prepared. The author also wishes to thank Smith, Kline and French. Inc., and the Office of The Quartermaster General for the financial support of this work,
A great deal of credit for this work goes to the author's wife for without her patience and deep understanding this work would have been impossible.

In the past decade the developments in the study of organic fluorine compounds have grown so rapidly that the literature dealing with such compounds has become quite extensive. The reason for this interest is due to the stability which fluorine often contributes to organic molecules. For example, polytetrafluoroethylene and poly-chlorotrifluoroethylene, manufactured under the trade names Teflon and Kel-F, respectively, are resistant to attack by all corrosive agents except molten alkali metals and elemental fluorine. Teflon and Kel-F have a high electrical resistance, low dielectric loss factor and zero water absorption. The working temperature of Teflon is from -100" to +500* F. while the range for Kel-F is from -100 to ?350 F.
The objective of the work to be presented here was to prepare monomers containing the trifluoromethyi group which upon polymerization would give elastomers that would remain rubberiike at low temperatures. The need for a rubberlike polymer possessing good low temperature properties has become critical due to increased activity in the colder regions of the earth. The present day elastomers

are inadequate to meet the requirements for these cold areas.
Since ethylene, propylene, butadiene, isoprene and styrene are used in the elastomers now available, it was decided to study the synthesis of some of these compounds with fluorine or trifluoromethyl groups substituted for hydrogen. Polymers of the fluorocarbons, such as Teflon, are very resistant to hydrocarbon solvents, possess excellent electrical properties and have a broad usable temperature range but are not rubberlike. Therefore, it was thought that a polymer prepared by homo- or co-polymerization of monomers containing fewer fluorine atoms than the fluorocarbons might retain the elasto-meric properties of the parent hydrocarbon and yet possess the desirable properties of the polyfluorocarbons.
Until recently the introduction of fluorine into a molecule could only be achieved by the action of fluorinating agents such as antimony trifluoride, hydrogen fluoride or fluorine itself. Fluorin-ations employing such reagents usually require halogenated intermediates which are often difficult to obtain. Furthermore the yields are frequently low due to undesired side reactions.
Haszeldine (14) has shown recently that, under the influence of ultraviolet light, trifluoromethyl iodide would add to certain olefins, frequently with the formation of dimers, trimers and higher telemers. For example, trifluoromethyl iodide adds to chlorotrifluoroethylene

as follows:
CF3I -...... CF3- + I-
CF3- + CF2*CFC1 .....> CF3CF2CFC1*
CF3CF2CFC1- + CF2= CFG1 ----> CF3CF2CFC1CF2CFC1-
CF3CF2CFC1CF2CFC1- + I" ------> CF3GF2CFClCF2CFClt
This procedure offered a new approach to the problem of introducing fluorine atoms into an organic molecule. Lovelace (20) and Tarrant and Lovelace (28) found that in the presence of peroxides, dibromo-difluoromethane and bromochlorodifiuoromethane would add to certain olefins to give compounds of the type CF2XCH2GHXR. This development offers a convenient method for the introduction of CF2Br- and CFjjCl- groups into the molecule. However, a satisfactory method for introducing the trifluoromethyl group was still lacking, since the preparation of trifluoromethyl iodide is tedious and involves a reaction between iodine and silver trifluoroacetate.
Kirschenbaum, Streng and Hauptchein (19) reported recently that thermal decomposition of silver salts of perfluorocarboxylic acids gave high yields of a saturated fluorocarbon resulting from coupling of the fluorocarbon radical of the acid. The mechanism of the reaction was shown to proceed through a perfluoroacid anhydride. Decomposition of the acid anhydride gave the fluorocarbon, carbon monoxide and carbon dioxide.

as will be discussed in Section 1 it was postulated that per* fluoroalkyl free radicals are formed by the thermal decomposition o* silver salts of perfiuorocarboxylic acids. It was thought that by decomposing the salt in the presence of an olefin, the perfluoroalkyl free radical would attack the carbon-carbon double bond of the olefin, thus offering a satisfactory method of introducing perfluoroalkyl groups into an organic molecule. However, the reaction did not give the secondary radical, R|GH2CHR. which could have reacted in one of several ways to be described later. Since the reaction of these silver salts and olefins did not present a convenient method of introducing a perfluoroalkyl group into a molecule, the reaction was, not investigated .to the fullest extent.
The monomer, E-trifluoromethylbutadiene-1, 3, was believed to hold a great deal of promise to give an excellent low temperature rubber. Warner (29) attempted its synthesis by following the detailed directions given in the patent of Hill and Town (18); the reactions claimed are:
CHSCMgBr CF3COCH3 -----.......>
CH3C(CF3)(OH)CSCH .......> CH3C(CF3)(OH)CH CH2
CH2=C(CF3)C=CH .......> CH2= C(CF3)CH= CH2

However, this work could not be reproduced and it was later learned that the work described in the patent had never been performed (10).
Since the commencement of this work Henne (16) has reported the synthesis of several trifluoromethylbutadienes, including 2-trifluoromethylbutadiene- 1 3. The Henne procedure for the preparation of 2-trifluoromethylbutadiene-1, 3 gives mixtures of products in several steps, resulting in low yields. This dissertation describes two methods of preparing 2-trifluoromethylbutadiene -1, 3 in good yields using trifluoroacetone as the starting material in both cases. The procedures to be described do not involve the formation of mixtures, thus eliminating the necessity of careful purification of the intermediates.
Interest has been shown in fluorine-containing styrenes and the preparation of several substituted styrenes comprises the third portion of this dissertation. McBee and Sanford (21) recently described a method of preparing ring substituted styrenes by the reaction of the appropriate arylmagnesium halide with acetaldehyde. Dehydration of the resulting carbinol gave the desired substituted styrene. It seemed of interest to prepare a series of alpha-trifluoromethyl-styrenes of the type:

where n = 0, I, or 2. These syntheses were accomplished by adding the appropriate Grignard reagent to trifluoroacetone followed by dehydration of the intermediate to the desired alpha-trifluoromethyl-styrene as discussed in Section III.

All temperatures reported in this dissertation are uncorrected and are given in degrees Centigrade. In distillations carried out at reduced pressure, pressures above 10 mm. were measured with a Zimmerle gauge which could be read to +0.2 mm.; pressures below 10 mm. were measured with a McLeod gauge which could be read to -0.1 mm. Refractive indices were determined with an Abbe refrac-tometer at the temperatures indicated. Densities were measured in one of two pycnometers, having volumes of 0. 6954 ml. and 1. 5863 ml. Molar refractions were calculated by the Lorenz-Lorentz equation.
Infrared spectra were obtained on a Perkin-Elmer model 21 double-beam Infrared Spectrophotometer.
Analyses were run by Peninsular ChemResearch, Inc., Gainesville, Florida and by Clark Microanalytical Laboratory, Urbana, Illinois.

1, I, i -Trifluoroacetone, prepared by the Claisen condensation of ethyl trifiuoroacetate with ethyl acetate as described later or purchased from Peninsular GhemRe search, Inc., Gainesville, Florida,
Octene-1, obtained from Matheson, Coleman and Bell, Inc.
Heptene-1, prepared by dehydration of n-heptyl alcohol obtained from Eastman Organic Chemicals.
Hexene-1, obtained from Phillips 66.
Butene-2, obtained from the Matheson Co., Inc.
Methyl bromide, obtained from the Matheson Co., Inc.
Sodium borohydride, obtained from Metal Hydrides, Inc.
Lithium aluminum hydride, obtained from Metal Hydrides,
1,1,2-Trifluoroethylene, obtained from Mr. M. R. Lilyquist of this laboratory.
m-Bromobenzotrifluoride, obtained from Halogen Chemicals,
1,1-Dichloro-2, 2-difluoroethylene, Genetron 170, obtained
from General Chemical Division of Allied Chemical and Dye Corp.

3, 5-Bis(Trifluoromethyl)bromobenzene, prepared by broraination of 3, 5-bis(trifluoromethyl)benzene obtained from Halogen Chemicals, Inc.
1-Chloro-l, 2, 2-trifluoroethylene, obtained from E. x. du Pont de Nemours and Co.
Malonic acid, obtained from the Matheson Co., Inc.

Discussion I
It has been shown by Kirshenbaum, Streng and Hauptchein (19) that the silver salts of perfluorocarboxylic acids can be decomposed thermally to give excellent yields of a saturated fluorocarbon containing twice as many carbon atoms as were present in the fluorocarbon radical of the acid. For example, silver trifluoroacetate and silver heptafluorobutyrate gave hexafluoroethane and n-perfluoro-hexane, respectively. These investigators have postulated that this reaction proceeds by the reaction mechanism:
2 CF3CF2CF2COOAg .....> (CF3CF2CF2CO)20 + AgzO
(CF3CF2CF2CO)20 > n-C6F14 + G02 + GO Ag20 + CO --------> 2 Ag + C02
2 CF3CF2CF2COOAg----- > n-C6F14 + 2 COz 4- 2 Ag
In contrast to this reaction, it should be noted that pyrolysis
of alkali and alkaline earth salts of perfluorocarboxylic acids, i. e.

potassium, sodium, barium, etc., give good yields of perfluoro olefins (13), (15), For example, pyrolysis of sodium perfluorobutyrate gave an almost quantitative yield of perfluoropropylene:
CF3GF2CF2COONa --> CF3CF=:CF2 -f CQ2 + NaF
The highly polar bond of the electrovalent sodium salt would be expected to favor heterolytic thermal decomposition. Silver perfluorobutyrate, however, possesses a considerable amount of covalent character as evidenced by its solubility in organic solvents. Due to this covalent character, silver perfluorobutyrate would be expected to decompose homolyticaiiy.
The formation of the intermediate acid anhydride may be explained by the transient formation of the free radicals, C3F^GOO* and Ag*. The activated silver atom can then attack a second molecule of silver perfluorobutyrate to rupture the carbon-oxygen single bond giving silver oxide and the transient C3F7CO and C3F7COO* radical.. Combination of the acyl and acyloxy radicals give the acid anhydride. All of these steps must occur almost simultaneously in the condensed phase. The formation of the acid anhydride may be shown schematically as:

C3F7COO ;---- Ag
/ j .......> (C3F7CO)20 + Ag20
P !
C3F?C ------ OAg
It has been shown by Szwarc and Murawski (27) that the thermal decomposition of acetic anhydride is a unimolecular reaction forming ketene and acetic acid and does not involve free radicals. In contrast, the thermal decomposition of perfluorobutyric anhydride, in the presence of silver oxide, give perfluorohexane. According to Kirshenbaum, et al., the different nature of these two reactions strongly suggests a free radical mechanism for the decomposition of perfluorobutyric anhydride. Consequently, perfluorohexane may be formed by a ternary collision of transient perfluoropropy 1 radicals resulting from the thermal decomposition of the anhydride:
(C3F7CO)20 -------> C3F7C-0' + C3F7G
G3F7C-0* -------> C3F7- + C02
C3F7C- -------> C3F7' + GO
Z C3F7........> -C6Fi4
As Kirshenbaum, et^at,, state, "it is recognized that the perfluoroacyl and perfluoroacyloxy radicals are decomposed easily under these pyrolytic conditions into perfluoropropyl radicals, carbon monoxide and carbon dioxide. The instability of the perfluoroacyloxy radical

is illustrated by the fact that peroxides such as di(tetrafluoropropionyl peroxide catalyzed the polymerization of ethylene, even at room temperature (3).
From the above discussion it may be readily seen that thermal decomposition of silver salts of perfluorocarboxylic acids such as CF-jCOOAg and C^F^COOAg would be expected to give the free radicals, CF3 and C3F<7* respectively. Previous work in this laboratory had established the fact that radicals such as CC13- CF^Bv, CF2C1*, and CFjjClCFCl* react efficiently with olefins to give new radicals from which alkanes or alkenes were obtained. It therefore seemed of considerable interest to study the thermal decomposition of silver salts of perfluorocarboxylic acids in the presence of olefins with the expectation that coupling of the fluorocarbon radicals could be prevented by the formation of a new radical as illustrated below:
CF3COOAg --------> CF3C-0- + Ag
CF3C-O .......> CF3' + COz
CF3- + CH2s CHR .......> CF3CH2CHR
The resulting new radical might couple with another radical, abstract a hydrogen atom from a second molecule of olefin to give a 1, 1,1-tri-fluoroalkane or lose a hydrogen atom to give a 1,1, 1-trifluoroalkene. Since Campbell, Knoblach, and Campbell (4) had prepared

1.1, l-trifluorooctene-2 by another route and had proved its structure beyond a doubt, this compound was to be the first goal of this new synthetic procedure. If thermal decomposition of the silver salt of trifluoroacetic acid proceeded as postulated to give CF3' radicals, it was conceivable that reaction with heptene-1 might result in the formation of 1,1, l-trifluorooctene-2. The reactions may be illustrated as: .
CF3COOAg .......> GF3' + CO., + Ag*
CF3 + CH2*CHC5HU ......-> CF3CH2GHC5Hn
CF3GH2CHC5H11 ......-> CF3CH=CHG5HU + H*
A considerable amount of material was obtained from the reaction but its physical properties did not agree with those reported in the literature (4) for the fluoroalkene. The boiling point of the material obtained was found to be about 20* higher than the boiling point reported for 1,1, l-trifluorooctene-2 indicating a material of higher molecular weight. The possibility of a product formed by coupling of the secondary radical, CF3GH2CHC5H^j[, was considered. However, a material formed by coupling would have a considerably higher boiling point than that of the reaction product. Elemental analysis did not agree with the value calculated for the coupled product, Cl6H28F6. Conization of the reaction product, according to the procedure of Hill,

et al., (17) gave only recovered starting material and produced no acid fraction characteristic of a carbon-carbon double bond.
Due to an infrared absorption peak in the 5. 57-5. 60 micron region, as compared to 5. 56 micron for the carbonyl absorption of ethyl trifluoroacetate, it seemed quite possible that the product might be an ester. Elemental analysis and molar refraction of the reaction product agreed closely with values calculated for CF3COOC.7HJ5. A sample of the product obtained from the reaction of silver trifluoroacetate and butene-2 was subjected to alkaline hydrolysis and tri-fluoroacetic acid and butanol-2 were identified in the hydrolysis mixture, indicating that the product obtained from the above reaction was sec-butyl trifluoroacetate. As a confirmation, a known sample of sec-butyl trifluoroacetate was prepared from trifluoroacetyl chloride and butanol-2. The physical properties and infrared spectra of the known sec-butyl trifluoroacetate were identical with those obtained from the reaction product of C^COOAg and butene-2.
To determine the applicability of this reaction to other silver salts of perfluorocarboxylic acids, silver perfluorobutyrate and octene-1 were reacted since the resulting n-octyl perfluorobutyrate had been prepared and the physical properties reported (22). The physical data obtained were in close agreement with those reported.
The results of these and other reactions of silver salts and

hydrocarbon olefins will be found in Table I. It will be noted that there is considerable deviation between observed and calculated molar refractions. This is due to the fact that a constant value of 1;100 was used for the atomic refraction of fluorine. This is open to criticism on the grounds that the atomic refraction of fluorine is not a true constant since it appears to vary from 0. 8 to 1.5 with no particular correlation between molecular structure and atomic refraction.
Ester formation during the thermal decomposition of silver salts in the presence of an olefin was quite unexpected. Perfluoroacyl and perfluoroacyloxy radicals have previously been considered extremely unstable. The results of the research reported here indicates that these previous assumptions of the stability of perfluoroacyloxy radicals are subject to criticism. To obtain ester formation it is necessary for the transient perfluoroacyloxy radical to exist as such and to add to the carbon-carbon double bond of the olefin. Since an ester is formed it is quite evident that the above requirements are met, thus indicating a certain degree of stability for the perfluoroacyloxy radical. It has been shown above how the perfluoroacyloxy radical may be formed by decomposition of R^COOAg into Rf COO* and Ag*, where Rf is a perfluoroalkyl group.
Perfluoroacyloxy radicals may also arise from decomposition of the acid anhydride. One perfluoroacyl radical would be produced

lor each perfluoroacyloxy radical. It appears likely that a perfluoroacyl radical could either attack the olefin to give a ketone or decompose to carbon monoxide and a perfluoroalkyl radical. This new radical would probably couple to form a perfluoroalkene. However, neither a ketone nor a perfluoroalkane was ever isolated from the reaction product. This fact would indicate that the perfluoroacyloxy radicals are the result of a hemolytic split of silver trifluoroacetate.
In contrast to the expected decomposition of the perfluoroacyloxy radical into an alkyl radical and CO^ it seems likely that the second step of the mechanism constitutes the formation of a new free radical:
RfCOO- + CH2-= CHR ......-> RfCOOCH2CHR
There are four possible routes open to the new radical: 1.) a hydrogen atom may be abstracted to give a saturated ester; 2.) a hydrogen atom may be expelled to give an unsaturated ester; 3.) a second perfluoroacyloxy radical may add to give an ester of a glycol; or 4.) the secondary radical may dimerize to give a diester having twice the number of carbon atoms which were present in the radical.
Due to the fact that only a saturated ester of the perfluorocarboxylic acid was formed it appears certain that the mechanism involves the abstraction of a hydrogen atom from some unknown source:

RfCOOCH2GHR + H* --.....> RfCOOCH2CH2R
The only source of the abstracted hydrogen is the hydrocarbon olefin. Abstraction of a hydrogen from the olefin would yield a new hydrocarbon radical which could dimerize or expel a second hydrogen atom to form either a diene or an alkyne:
*RCHCH2ChLcH2 + H- RCHs CHCH= CH2 + H-
Repeated attempts to isolate any of these postulated hydrocarbon olefins were unsuccessful.
It is quite probable that the quantity of material formed by the loss of a hydrogen atom was extremely small, as a consequence of the small runs, and as a result the isolation of the unknown material was not achieved.
Attempts to prepare esters of trifluoroacetic acid and hepta-fluorobutyric acid by direct esterification of the acid with the appropriate alcohol were unsuccessful. Only sec-butyl trifluoroacetate and sec-butyl-heptafluorobutyrate could be prepared from the acid chlorides and alcohols. This lack of reactivity is probably due to the

low boiling points of tbe acid chlorides resulting in a short period of contact with the alcohol.

Experimental I
Procedure: Unless otherwise stated the conditions for the reaction of silver trifluoroacetate and silver heptafluorobutyrate with olefins were as described here. The silver salt and olefin were sealed in a 300 ml. rocking autoclave and heated at 250" C. for four hours, coolt and opened. In the case of butene-2 the autoclave was cooled in Dry Ice prior to loading and unloading.
Reactions, of Silver Trifluoroacetate:
a. Addition to Octene-1: The autoclave was charged with
1, 34 moles (150 g.) of octene-1 and 0. 25 mole (55.2 g.) of silver
trifluoroacetate and reacted as described. Upon opening the reaction
vessel, 142 g. of liquid was obtained which was steam distilled. The
organic layer was separated, dried and fractionated to give 18 g.
(31. 8%) of octyl trifluoroacetate, bp. 57* / ll mm., 62 /15 mm,,
n2i 1.3742, d2.4 1.0221. D 4
Calculated for CF3COOC8Hl7: C, 53.2j H, 7.57; MRD, 50.05 Found: C, 53.4; H, 7.84; MRD, 50. 57,
b. Addition to Heptene-1: One and five-one hundredths moles (103 g.) of heptene-1 and 0. 25 mole (55. 2 g.) of silver trifluoroacetate were reacted, steam distilled, the organic layer separated, dried

and fractionated to give 10 g. (18. 9%) of heptyl trifluoroacetate,
bp. 145-8* C., n" 1.3672, d23 1.022.
u 4
Calculated for CF3COOC7Hl5: C, 50. 7; H, 7.13; MR^, 45. 43 Found: C, 50.9; H, 7.27; MRD, 46.64.
c. Addition to Hexene-1: Sixty-two hundredths mole (52 g.) of hexene-1 and 0. 25 mole (55. 2 g,) of silver trifluoroacetate were reacted, steam distilled, .the organic layer separated and dried. Fractionation gave 3.5 g, (7. 08%) of hexyl trifluoroacetate,
bp. 124-5* C, n2* 1.3568, d24 1.0109. Analysis:
Calculated for CF3COOC0H13: C, 48.4; H, 6.61; MRD, 40.80 Found: C, 48.0; H, 6.18; MRD, 42.93.
d. Addition to Butene-2: Two moles (110 g.) of butene-2 and 0. 25 mole (55. 2 g. ) of silver trifluoroacetate were reacted, cooled in Dry Ice, the liquid transferred to a flask and distilled through a Vacuum jacketed column with a head cooled by Dry Ice to give 38 g. (89. 4%) of sec-butyl trifluoroacetate, bp. 90. 0-93. 2* with a fraction of28g., bp. 93.0-93.2* C., n22 1.3391, d2| 1.0532.
Calculated for CF3COOC4H9: C, 50.08; H, 5.30; F, 26.84; MRj), 31.57

Found: C. 49.98; H, 7.07} F, 26. 37; UKD, 33.78.
(1) Saponification of sec-Butyl Trifluoroacetate: A solution of one mole (40 g.) of sodium hydroxide in 100 ml. of water was prepared and added to 10 g. of sec-butyl trifluoroacetate in a 250 ml,, three neck flask equipped with a stirrer and reflux condenser. After refluxing for two hours, during which time the initial two layers disappeared and then reformed, the organic layer was hydro* distilled at 80-85 C. The distillate was salted out to give 4 g. of organic material which was dried and distilled to give:
Cut bp. wt. n2p
I 88-92* C. 0.6 g.
n 92-100 1.3 1.3966
HI 100-105 1.6
No satisfactory boiling point could be obtained on this distillation due to superheating of the vapor. Infrared spectra of Cut II was identical with that of a known sample of sec-butyl alcohol.
The aqueous residue from the original hydrodistillation was made acid to litmus and extracted exhaustively with ether, the extracts dried and treated with an ether solution of piperazine to give an insoluble precipitate, mp. 244. 5*r with decomposition. The known

piperazine salt of trifluoroacetic acid melted at 250. 0-0. 5* C A mixed melting point gave 248. 0-8. 5* C.
The following reactions were attempted using the conditions given below.
e. Attempted Reaction with Octene-1 at 150* C.: One and seven hundredths moles (120 g.) of octene-1 and 0, 25 mole (55. 2 g, j of silver trifluoroacetate were heated at 150* C, for five hours in the 300 ml. autoclave. The resulting liquid was steam tUstilled, the organic layer separated and dried. Fractionation gave only 80 g. of recovered octene-1.
f. Reaction with Heptene-1 in the Presence of Benzoyl Peroxide: A 500 ml. three neck flask equipped with a stirrer and reflux condenser was charged with 0.125 mole (27. 6 g.) of silver trifluoroacetate, 0. 633 mole (62 g.) of heptene-1 and 5 g. of benzoyl peroxide and heated on a steam bath for six hours. Distillation gave 47 g. of recovered heptene-1, bp. 80-95* At a pot temperature of 140* C. the head temperature fell to 67* with the evolution of HF, Six and one-half grams of unidentifiable material was obtained boiling at 67-9* C. n2* 1.3146, d24 1,2169.
g. Attempted Reaction with Heptene-1 Under Ultraviolet Light: A quartz flask was charged With 0. 48 mole (47 g.) of heptene-1

and 0.125 mole (27. 6 g.) of silver trifluoroacetate and heated on a steam bath under intense ultraviolet illumination. Distillation gave only 32 g, of recovered heptene-1.
2. Reactions of Silver Heptafluorobutyrate;
a. Addition to Octene-l: Twenty-five hundredths of a mole (82 g. ) of silver heptafluorobutyrate and 0. 50 mole (56 g.) of octene-1 were reacted at 225* C. for three hours to give a highly carbonized material, After steam distilling, separating and drying the organic layer, distillation gave 19 g. of recovered octene-1, bp. 120-5* and 38 g. of octyl heptafluorobutyrate, bp. 50* /2mm., n2^ 1. 3568,
d2| 1.1726. Constants reported (22) for C3H7COOC8H17: bp. 108*/27 mm., n2J 1. 3582, d2j| 1.185. Analysis,
Calculated for C3F7COOCgHi7. C, 44.25} H, 5.24; MRD, 59.29 Found: C, 44.7; H, 5.27; MRD, 60.92.
b. Addition to Butene-2: Twenty-two hundredths of a mole (72 g,) of silver heptafluorobutyrate and 1.2 moles (77 g,) of butene-2 were reacted and the cold material transferred to a distilling pot under a vacuum jacketed column and Dry Ice and acetone head. Butene-2 (50 g.) was removed at 0* C. to 10* C. and the residue transferred
to a heated distilling column. Further distillation gave 36 g. of

sec-butyl heptafluorobutyrate, bp. 123. 0-124. 5 C. with a center cut taken at 124.0-124.5* C., n2* 1.3213, &Z\ 1.3082. Properties reported (22) for C3F7COOC4H9: bp. 126* C., n% 1.3213, d2| 1,284. Analysis:
Calculated for C3F7COOC4Hg: C, 32.57; H, 3.51; MRD, 40.81 Found: C, 32.79; H, 3.33; MRD, 41.11,
A known sample of sec-butyl heptafluorobutyrate was prepared from heptafluorobutyryl chloride and butanol-2 which had the constants, bp. 124-6* C, nD 1. 3208. Infrared spectra obtained from the reaction product of G3F7COOAg and butene-2 and from the reaction product of C3F7C0C1 and butanol-2 were identical.

Boiling MRD %C %H % Msc.
Formula Point t* C d4
__C_ Calc. Fd. Calc. Fd. Calc. Fd. Calc. Fd.
CF3COOC{CH3)HC2H5 93. 0-3. 2 22 1. 3391 1. 0532 3L 57 33. 78 50.08 49.98 5.30 7.07 26.89 26.37
CF3COOC6H13 124-5 24 1.3568 1.0109 40.80 42.93 48.4 48.0 6.61 6.18
CF3COOC7Hl5 145-8 23 1.3672 1.0222 45.43 46.64 50.7 50.9 7. 13 7.27
CF3COOC8H17 57/11 mm. 24 1. 3742 1. 0221 50. 05 50. 57 53. 2 53. 4 7. 57 7. 84
62/15 mm.
C3F7COOC(CH3)HC2H5 124. 0-4. 5 21 I. 3213 1. 3082 40.81 41. 11 32. 57 32. 79 3. 51 3* 33
C3F7COOC8Hi7 50/2 mm. 22 1. 3568 1; 1726 59.29 60.92 44; 3 44. 7 5. 24 5.27

Discussion H
One of the monomers which seemed to hold a considerable amount of promise as a rubberlike polymer at low temperature was 2-trifluoromethylbutadiene-l, 3. A study of the preparation of this diene comprises the second portion of this dissertation.
Previous attempts by Warner (29) following the detailed patent of Hill and Town (18) failed to give the desired monomer. The patent described a method of preparation involving the addition of ethynyl-magnesium bromide to trifluoroacetone to yield 2-trifluoromethyl-3-butynol-2. The initial step is complicated by the fact that an equilibrium exists between ethynylmagnesium bromide, acetylene and acetylene-bis-(magnesium bromide):
2 CH^CMgBr .......> BrMgCSCMgBr + CH=CH
The presence of this equilibrium lead to the formation of a symmetrical
glycol, 2, 5-bi8(trifluoromethyi)-3-hexyne-2, 5-diol. A possibility
also existed that the first product of the addition of ethynylmagnesium

bromide across the carbonyl double bond of irifluoroacetone could undergo a tautomeric shift allowing the salt to react as a new Grignard reagentt
3H3-C-C=CH .......> CH3-C-C3CMgBr
ii i
This new Grignard reagent may then react with a second molecule of trifluoroacetone to form the glycol. It was shown by Warner that the predominance of the acetylenic glycol in the reaction product was due more to tautomerism in the addition product than to the disproportion-ation of the, ethynylmagne sium bromide. Since the synthetic scheme claimed in the patent was in error in the first step* attempts to synthesize 2*trifluoromethylbutadiene-l, 3 by this procedure failed. Henne (16) has recently reported the same observation, not only with trifluoroacetone and ethynylmagne sium bromide, but also with acetone and ethynylmagne sium bromide.
Since the initiation of this work Henne (16) has reported the preparation of 2-trifluoromethylbutadiene-l, 3 by a rather involved procedure. Trifluoroacetone was treated with ethylmagnesium bromide, one-third of the ketone being reduced to 1,1,1-trifluoropropanol-2 while the remainder condensed with the Grignard reagent to yield 2-tri-fluoromethylbutanol-2. Pyrolysis of the acetate of this carbinol gave

a mixture of olefins which was not separated but was allylically brominated with N-bromoeuccinimide to a mixture of allyl bromides in equilibrium:
CH2= C(CF3)CHBrCH3 <......> CH2BrC{CF3)= CHCH3 <......>
CH3C(CF3)a CHCH2Br <------> CH3C(CF3)BrCHe CH2
These ally! bromides were not separated but treated with triethyl-amine to give 2-trifluoromethylbutadiene -1, 3. Due to the mixtures obtained, the overall yield is less than 5% of theoretical.
Owing to the availability of trifluoroacetone, this material was chosen as the starting point for two possible synthetic procedures to prepare 2-trifluoromethylbutadiene-l, 3, Darral, Smith, Stacy and Tatlow (6) have reported the preparation of trifluoroacetone cyano-hydrin from trifluoroacetone and hydrogen cyanide. Trifluoroacetone cyanohydrin has two reactive groups, the hydroxyl and the nitrile, which offer two possible points of attack. Two routes may be developed for the synthesis of the desired 2-trifluoromethylbutadiene-l, 3 as illustrated below:

CF3COCH3 .....* CH3C(CF3)(OH)CN
^Jeat | 2 H
\ CH3MgBr -2H20
CF3 j,
GH2= C-COCH3 CH2a C(CF3)CKr: CH2
v2 H ^/^-HZ0
The addition of methylmagnesium bromide to the nitrile group of trifluoroacetone cyanohydrin, followed by hydrolysis of the salt, yielded 2-trifluoromethyl*2hydroxybutanone-3, This is a common reaction of nitriles involving the formation of an imino salt intermediate:
which is easily hydrolyzed to a carbonyl group. It appears that the presence of the trifluoromethyl group on the same carbon as the nitrile has little or no effect on the above reaction.
Reduction of the carbonyl group to a hydroxyl gave 2-trifluoro-methylbutanediol-2, 3. A Meerwein-Fonndorf-Verley reduction (23)

using aluminum isopropoxide in isopropanol yielded the glycol in only 23. 9% of theoretical. A reduction of this type is dependent upon the equilibrium:
Al OCH(CH3)2 3
which may be driven to the right by the constant removal of acetone. In practice this is accomplished by fractional distillation until a sample of distillate give a negative phenylhydrazine test for acetone. After removal of acetone is complete, the aluminum complex is decomposed with acid to yield the alcohol. Due to the low yield of 2-trifluoromethyl-butadiol-2, 3, a better reducing agent was sought.
Reduction with lithium aluminum hydride (24) increased the yield to 65. 7% of theoretical but required great care and anhydrous conditions. Sodium borohydride (5) has been employed in the reductions of acid chlorides, aldehydes and ketones and does not require the anhydrous conditions necessary for lithium aluminum hydride. For this reason reduction of 2-trifluoromethyl-2-hydroxy-butanone-3 was performed with sodium borohydride to give the glycol in 55.1% yield. The use of sodium borohydride also eliminated the danger accompanying any reduction employing lithium aluminum hydride in diethyl ether.
Dehydration of 2-trifluoromethylbutanediol*2, 3 was achieved by use of phosphoric oxide. Initial attempts to dehydrate the glycol

with phosphoric oxide resulted in very low yields of 2-trifluoromethylbutadiene-l, 3, and a great deal of carbonization. It was found that higher yields of the diene could be obtained by adding the glycol to phosphoric oxide with vigorous stirring at room temperature until an even paste was formed. By careful heating with constant stirring, the paste could be decomposed to 2-trifluoromethylbutadiene-l, 3 in 36. 9% yield. There appears to be an intermediate complex formed involving the glycol and phosphoric oxide which decomposes into the olefin at elevated temperatures.
rearrangement exists during the dehydrationx>i 2 trifluoromethyl-butanediol-1, 3 with phosphoric oxide. One or more of three compounds could possibly result:
It should be noted here that an opportunity for a pinacol
The above three materials may result by the reaction scheme outlined below (1):

CH3**C""GH-CH3 + OH
CH-i-C-CH-CHc> -*--
3 6h 3"h20
CF, I 1CH ^ ~ C ~ C **H
- H
- H
CH3-C-C-CH. 1 1 **
CH -C-CH-CH3 5 +
CF3 4 1
i 1
- H+

In unsymmetricai pinacols theoretical migrational aptitudes often do not agree with experimental data and it is extremely difficult to predict the major product of such a reaction. However, since none of these materials could be identified in the reaction mixture it is apparent that a reaction of this type did not take place to any great extent during the preparation of 2-trifluoromethylbutadiene-l, 3.
The above procedure resulted in good yields in every step except the last, involving the dehydration of a glycol. It should be noted that carbinols in which the hydroxyl group is on a carbon which also carries a trifluoromethyl group are often difficult to dehydrate. For example, the dehydration of trifluoromethyldimethylcarbinol with phosphoric oxide to give 2-trifluoromethylpropene requires a temperature of 160 C. For this reason it would be desirable to eliminate such a dehydration step in the synthesis of 2-trifluoromethylbutadiene-l, 3. In order to eliminate this difficult step a study was made of the thermal dehydration of trifluoroacetone cyanohydrin to 2-trifluoro-methylacrylonitriie. It was expected that the latter would be useful in the following reactions which would lead to 2-trifluoromethylbutadiene-l, 3;

GF3 ^*^*3 05*2
CH2= C-CHOH-CH3 -----> CH2rs C-CH* CH2
Attempts to dehydrate the eyanohydrin with thionyi chloride by following the directions of Dickey (7) failed. It had been claimed (8) that the dehydration of this compound could be achieved by pyrolytic decomposition of the acetate.
and acetyl chloride in 83. 8% yield. Preliminary experiments indicated that the acetate did not pyrolyze appreciably at temperatures below 450 C. while considerable carbonization was found to result at 550-600* C. Therefore, a temperature of 500 $* c. was chosen as the operating temperature of the pyrolysis tube which consisted of a 1 x IE inch pyrex tube packed with borosilicate beads. Optimum operating conditions were found to be 10 liters per hour of dry nitrogen, 120 drops per minute of acetate and a furnace temperature of 500 C. However, since crude gas and liquid metering devices were used the results were not consistent. Duplication of the results was very difficult. A considerable amount of low boiling material was obtained from each run, and this was believed to be a mixture of
The acetate was prepared from trifluoroacetone eyanohydrin

trifluoroacetone and hydrogen cyanide. The data for several typical runs will be found in Table HI. It will be noted that the highest yield of 2*trifluoromethylacrylonitrile obtained was 46. 6% of theoretical.
One attempt was made to add methylmagne sium bromide to the carbon-nitrogen triple bond of 2-trifluoromethylacrylonitrile. The reaction was performed as described for the preparation of 2-trifluoromethyl-2-hydroxy-butanone-3. However, the material obtained from the reaction was not the desired ketone. Infrared spectra indicated the presence of C-F, -C=N and possibly a C=C double bond. Elemental analysis gave the following results: 49.7% C, 5. 99% H and 9.13% N. The data obtained agreed most closely with material possessing the emperical formula C^HgF3N2. which could have a structure CH3CHeC{CF3)(CH3)CN. The required analysis are 47. 6% C, 5. 17% H and 9. 26% N. However, since it is difficult to account for the formation of such a compound in the reactions used, it appears likely that the material isolated was impure.
It has been shown (c. f, M. S. Kharasch and O. Reinmuth, Grignard Reactions of Nonmetallic Substances, pp. 782-3, 1st ed., Prentice-Hall, Inc.) that somex,/9-unsaturated nitriles undergo 1, 4-addition of a Grignard reagent to the conjugated system. For example, addition of a Grignard reagent to < -phenylcinnam onitrile results in a saturated nitrile:

c6h5chs c(c6h5)cn----1?----> c6h5chrch The 1, 4-addition may be due to steric hindrance resulting from the alpha substituent.
: For the case of methylmagnesium bromide and 2-trifluoromethyl-acrylonitrile a reaction of this type would produce 2-trifluoromethyl-butanenitrile, ch3ch2ch(cf3)cn. Unfortunately the reaction product could not be identified. 6
3-Trifluoromethylcrotonic acid, prepared previously by Schwarz (26) offers a convenient starting point for a series of reactions which would lead to 2-trifluoromethylbutadiene-l, 3:
Malonic 1
cf3coch3 ch3-c-gh2cooh ->
- co2 OH 2 4
cf3 cf3
i i
gh3-g= ch-cooh *-L"-4 ch3-cs chch2oh
. 3
socl2 lialh4/ p25
ch3-c=chcocl ch2sc-chsch2
3-TrifluorOmethyl-3*hydroxy-butyric acid was prepared by condensing trifluoroacetone and malonic acid in the presence of pyridine. The intermediate dicarboxylic acid was decarboxylated by heating the

mixture to 130* C. Fractionation gave the hydroxy acid in good yield. 2-Trifluoromethylcrotonic acid was prepared by dehydrating 3-trifluoromethyl -3 -hydroxy -butyric acid with 50% sulfuric acid. It will be noted that this intermediate hydroxy acid possesses the undesirable structure, RC(CF3)(OH)R'. However, the dehydration of a beta hydroxy acid proceeds quite easily by refluxing with sulfuric acid. Because of the presence of the carboxyl group it was possible to eliminate the difficulty experienced in the dehydrations discussed above.
3-Trifluoromethylcrotonic acid could be reduced directly to 3-trifluoromethyl2-butene-l-ol with lithium aluminum hydride, however, this reduction gave the alcohol in only 38.4% yield. It was found that reduction of 3-trifluoromethylcrotonyl chloride, prepared in 69.8% yield by refluxing 3-trifluoromethylcrotonic acid with thionyl chloride, gave a 70% yield of 3-trifluoromethyl-2-butene-l-ol.
3-Trifluoromethyl-2 -butene-1 -ol was readily dehydrated to give 2 -trifluoromethylbutadiene-1, 3 when treated with phosphoric oxide. It should be noted that this dehydration involves removal of hydrogen and hydroxyl groups from the 1,4 position. It was found in this case, as in the dehydration of 2-trifluoromethylbutanediol-2, 3, that the initial formation of a paste was essential for satisfactory yields of diene.
The physical properties of the 2-trifluoromethylbutadiene-1, 3,

prepared by the procedures described above agree closely with the values reported by Henne (16).
Homopolymerization of 2-trifluoromethylbutadiene-l* 3 at 65* C. using benzoyl peroxide as the catalyst gave a clear, very viscous oil. Homopolymerization using boron trifluoride as the catalyst gave a slightly yellow viscous oil which contained small particles of solid. A sample of diene was observed to become viscous after standing for several weeks at room temperature. This latter observation indicates the tendency to polymerize even in the absence of a catalyst. These polymerizations produced only short chains as evidenced by the formation of an oil and not a solid. It is possible that longer chain formations could be produced by use of more drastic conditions.
This work has produced two independent procedures for the production of the monomer, 2-trifluoromethylbutadiene-l, 3 from a readily available starting material, trifluoroacetone. Satisfactory yields were obtained throughout.

Experimental 'It ' Preparation of Trifluoroacetone:
Seven moles of sodium ethoxide was prepared from seven moles (16'1 g,) of sodium sand and 7. 5 moles (345 g.) of ethanol in one liter of ether contained in a five liter, three neck flask equipped with a stirrer, addition funnel and ice-water cooled reflux condenser. Seven moles (994 g.) of ethyl trifluoroacetate was added as rapidly as safety permitted to the cold ether slurry of sodium ethoxide followed by the rapid addition of ethyl acetate. After standing for several days, the ether and excess esters were distilled under aspirator vacuum, the steam bath being used to remove the last traces of volatile material. The resulting solid residue was dissolved in two liters of 40% sulfuric acid and the mixture refluxed. T rifluoroacetone was led via a rubber tube to a cold trap immersed in ice water connected to a trap in Dry Ice and acetone. Fractionation gave 613 g. (78. 2%) of trifluoroacetone, bp, 20-22* C.
2. Preparation of Trifluoroacetone Cyanohydrin:
A solution of 5 moles (345 g.) of potassium cyanide in one liter of water was treated at 0-10* C. with 3.03 moles (336 g.) of trifluoroacetone in apparatus described above. Two liters of 6 N. sulfuric acid was added at a temperature below 15* C. After standing overnight, the heavy organic layer was separated and the aqueous

layer extracted several times with 100 ml. portions of diethyl ether. The combined organic layer and extracts were dried over anhydrous sodium sulfate and distilled to remove ether and give 400 ml. of crude trifluoroacetone eyanohydrin, bp. 40-70* /100 mm. The crude material was dissolved in an equal volume of dry ether and again dried over sodium sulfate. After removal of the solvent, the residue was fractionated under reduced pressure to give 298 g, {70, 8%) of the desired trifluoroacetone eyanohydrin, bp, 65-70* /43 mm.
3. Preparation of 2-TrifluorpmethyI-2-hydroxy-butanone-3:
A two liter, three neck flask equipped with a stirrer, addition funnel, gas inlet tube and ice water reflux condenser was charged with two moles (48. 6 g,) of magnesium turnings and flame dried under an atmosphere of dry nitrogen. Seventy-five ml. of dry ether was added and the flow of methyl bromide begun through the gas-inlet tube until a reaction commenced. The flow of methyl bromide was stopped and the flask cooled with ice-water to control the reaction. After the initial reaction had subsided 425 ml. of ether was added and the flow of methyl bromide was resumed and continued until all of the magnesium had reacted. One mole (139 g.) of trifluoroacetone eyanohydrin was added to the cold methylmagnesium bromide solution. After standing overnight, the mixture was hydrolyzed by the careful addition of one liter of 10% sulfuric acid, the ether layer separated, the water layer

extracted and the combined extracts dried over sodium sulfate. Fractionation removed the solvent and gave 110 g. (70, 5%) of 2-trifluoromethyl -2 -hydroxy-butanone -3, bp. 114-8* C, Constants for pure 2- trifluoromethyl-2-hydroxy-butanone-3 are; bp. 118. 0-8. 4* C,, n2 1.3581, d2| 1.2701. Analysis:
Calculated for CsH7F302: C, 38. 47; H, 4.52; MRD, 26.53 Found: C, 38.3; H, 5.00; MRD, 26.99.
4, Preparation of 2-Trifluoromethylbtttanediol-2, 3:
a. Reduction with aluminum isopropoxide: A solution of 0, 685 mole (140 g.) of aluminum isopropoxide in 250 ml. of dry isopropyi alcohol was prepared in a one liter, three neck flask equipped with a thermometer, magnetic stirrer, addition funnel and 24 inch distilling column with a variable take-off head. To this solution was added 0. 554 mole (83 g.) of 2-trifluoromethyl-2-hydroxy-butanone-3 and an azeotrope of acetone and isopropyi alcohol was distilled between 60 and 70* C, until a negative test for acetone was obtained. Finally, 50 ml. of isopropyi alcohol was distilled during which time the residue solidified. Six-hundred ml. of 6 N, sulfuric acid was added and the mixture stirred until solution was complete. The resulting solution was extracted with ether, the extracts dried and distilled to remove solvent and isopropyi alcohol, Vacuum distillation of the residue gave 21 g.

(23. 95%) of 2triflttoromethylbutanediol-2, 3, bp. 58* /5. 5 mm.,
n2^ 1.3866, d2| 1. 3282.
Calculated for C5HqF3C>2: C, 38. 0; H, 5. 74; MRD, 28. 04 Found: C, 38.16; H, 5.84; MRD, 28.00.
b. Reduction with lithium aluminum hydride: A slurry of
0.42 mole (16 g,) of lithium aluminum hydride in 700 ml. of ether was prepared in a five liter flask equipped as described previously in the preparation of trifluoroacetone, A solution of 0, 25 mole (39 g.) of 2-trifluoromethyl-2-hydroxy-butanone-3 in 300 ml, of ether was added to the cold slurry at such a rate as to maintain a temperature of 15* C After standing for one hour, 0. 75 mole (66 g,) of ethyl acetate was carefully added, followed by one liter of 10% sulfuric acid. The ether was separated, the water layer extracted with ether and all extracts dried over sodium sulfate. Fractionation gave 26 g. (65. 7%) of 2-tri-fluoromethylbutanediol-2,3, bp. 41-3*/2 mm,
c. Reduction with a water solution of sodium borohydride: A solution of 0. 1 mole (3. 79 g.) of sodium borohydride was prepared in 50 ml. of water in a 250 ml, flask equipped with a stirrer, addition funnel and reflux condenser. Two-tenths of a mole (31, 2 g.) of 2-tri-fluoromethyl-2-hydroxy-butanone-3 was added dropwise to the cooled solution over a period of 45 minutes. After two hours, 0.2 mole (8 g,)

of sodium hydroxide was added with cooling and the mixture exhaustively extracted with ether. The extract was dried and fractionated to give 5 g. (15. 8%) of 2-trifluoromethylbutanediol-2, 3, bp. 49-51* /2. 2 mm.
d, Reduction with dry sodium borohydride: A 500 ml, flask equipped as above was charged with 0, 25 mole(9.46 g. ) of dry sodium borohydride and treated with 0. 75 mole (117 g.) of 2-trifluoromethyl-2-hydroxy-butanone-3 with external cooling. After stirring for three tours one mole (40 g.) of sodium hydroxide in 250 ml. of water was added with cooling and allowed to stand overnight. The resulting clear solution was exhaustively extracted with ether, the extract dried and fractionated to give 66 g. (55. 1%) of 2-trifluoromethylbutanediol-2, 3, bp. 68-72* /10 mm.
5, Preparation of 3-Hydroxy-3-trifluoromethylbutyric Acid:
A solution of 2, 3 moles (240 g,) of malonic acid in 370 ml, of pyridine was prepared in a two liter flask equipped as previously described with a gas inlet tube and cooled in ice-water. Two moles (224 g,) of trifluoroacetone was bubbled into the cold solution during which time the solution became quite pasty. The mixture was heated slowly to 90* C and maintained at that temperature for 14 hours and finally heated to 130* c, for eight hours or until the evolution of carbon dioxide ceased. Distillation removed pyridine and gave 212 g.

(61. 6%) of 3-hydroxy-3-trifluoromethylbutyric acid, bp. 100-115* /8 mm.
6, Preparation of 3-Trifluoromethylcrotonic Acid:
A two liter flask equipped with a stirrer and water separator topped with a reflux condenser was charged with 1. 23 moles (212 g.) of 3hydroxy-3~trifiuoromethyibutyric acid and 500 ml. of 50% sulfuric acid and refiuxed. As the unsaturated acid was formed, it was azeo-troped and removed via the water separator. After 14 hours no more of the unsaturated acid was removed. The crude acid was dried over sodium sulfate and distilled to give 134 g. (70.8%) of 3-tr ifluoromethyl-crotonic acid, bp. 160-6*.
7. Preparation of 3-Trifluoromethylcrotonyi Chloride;
One-tenth mole (11. 9 g.) of freshly distilled thionyl chloride and 0.1 mole (15.4 g.) of 3-trifluoromethylcrotonic acid were refluxed with stirring in an appropriate flask until the evolution of hydrogen chloride stopped. The residue was distilled to give 12 g. (69. 8%) of Z-trifluoromethylcrotonyl chloride, bp. 99.5*100.5* C, n 1.3916, d2| 1.3037. Analysis:
Calculated for C5H4CIF3O: CI, 20.6; MRD, 31.47 Found: CI, 20.2; MRD, 29.71.

8. Preparation of 3-Trifluoromethyl-2-buten-l-ol:
a. Reduction of 3-trifluoromethylcrotonic acid: A slurry of 0. 447 mole (17 g. ) of lithium aluminum hydride in one liter of dry ether was prepared in a five liter flask equipped as previously described, cooled in an ice bath and treated with 0. 5 mole (78 g,) of 3-trifluoro-methylcrotonic acid at such a rate as to produce an internal temperature of 5-10 C. After an hour, 150 ml. of water was added followed by 1. 5 liters of 10% sulfuric acid. The ether was separated, the water extracted with ether and the extracts dried thoroughly over sodium sulfate. Fractionation gave 24. 0 g. (38.4%) of 3-trifluoromethyl-2-buten-i-ol, bp. 144-5* C., n22 1.3761, d22 1,2021.
Calculated for C5H7F30: C, 42.8; H, 5.04} MRD, 26. 36 Found: C, 42.61; H, 5.44; MR^, 26.75,
b. Reduction of 3-trifluoromethylcrotonyl chloride: The procedure was the same as described above for the reduction of 3-tri-fluoromethylcrotonic acid using 0. 58 mole (22 g.) of lithium aluminum hydride and 0. 477 mole (82 g.) of 3-trifluoromethylcrotonyl chloride
to give 47 g. (70, 0%) of 3-trifluoromethyl-2-buten-l-ol, bp. 141-6* C., n2 1.3762, d2l 1.202.

9. Preparation of 2-Trifluoromethyibutadiene-l, 3;
a. By dehydration of 2-trifluoromethylbutanedlol-2, 3:
A 500 ml. three neck flask equipped with a six-inch column and ice water cooled variable take-off head, stirrer and addition funnel was charged with 0.1 mole (14. 2 g.) of phosphoric oxide. One-tenth mole (15.8 g.) of 2-trifluoromethylbutanediol-2, 3 was added drop wise and slowly heated. A relatively high temperature(200-250* C J was required before decomposition of the intermediate began, Thirteen grams of crude product was obtained up to a head temperature of 76* C, Fractionation of the crude material gave 4. 5 g. (36. 9%) of very volatile 2-trifluorornethylbutadiene-1, 3, bp, 35. 0-5. 5* C n2 1.3431, d2| 1.037. Analysis:
Calculated for C5H5F3: C, 49. 3; H, 4. 13; MRD, 24. 36 Found: C, 49.51; H, 4.34; MRQ, 24.87,
b. By dehydration of 3-triflttoromethyl-2-buten-l-ol: Thirty two-hundredths of a mole (45 g.) of 2 trifluoromethyl-2 -buten-l-oi was added slowly with external cooling to 0. 3 mole (42. 6 g, J of phosphoric oxide in a 500 ml. flask equipped as above in part 9a and stirred without heating until a uniform paste was formed. Heat was then applied and 23 g. of crude product obtained. Fractionation gave 22 g. (60. %%) of 2-trifluoromethylbutadiene-1, 3, bp. 36* C.,

n | 1. 3434. Infrared spectra of this sample was identical with that of the sample prepared by the dehydration of 2-trifluoromethylbutadiol-2, 3.
10f Preparation of Trifluoroacetone Cyanohydrin Acetate:
A two liter, three neck flask equipped with a stirrer, thermometer and reflux condenser was charged with 2. 61 moles (363 g.) of trifluoroacetone eyanohydrin and 3.00 moles (234 g.) of acetyl chloride and refluxed until the evolution of hydrogen chloride ceased, about 5-6 hours. The crude reaction product was distilled directly from the reaction flask, dried and fractionated to give 369 g. (83. 8%) of the desired acetate, bp, 85 /72 mm.
11. Preparation of Trifluoromethylacrylonltrile;
The equipment used for pyrolysis of trifluoroacetone eyanohydrin acetate consisted of the following: A flowmeter for measuring the volume of nitrogen and a dropping funnel for the acetate were connected to one end of a 1 x 12 inch pyrex tube packed with borosilicate beads. The pyrex tube was heated by a combustion furnace and the internal temperature automatically controlled by a thermocouple and pyrometer, Fyrolyzed material was lead directly from the furnace into an ice-water cooled receiver equipped with a reflux condenser. A tube connected the top of the reflux condenser to tvo cold traps in Dry Ice and acetone.
A typical reaction was carried out as follows: 45 g. of acetate

was added in 57 minutes. Nitrogen was added at the rate of 10 liters per hour. The crude product (37. 5 g. J was fractionated to give 14. 0 g. (46. 6%) of 2r trifluoromethylacrylonitrile, bp. 73-85 C.
Several typical runs are recorded in Table II. Trifluoromethylacrylonitrile has the following physical properties: bp. 78*9* C,, n2^'4 1. 3261. d2!- 4 1.1753. MRD, calculated for C4H2F3N, 20. 79} MR^ found, 19.96. Elemental analyses were Trot obtained since this compound has been reported previously (8).
12. Attempted Addition of Mcthylmagnosium Bromide to Trifluoromethylacrylonitrile:
Two moles of methylmagnesium bromide was prepared in a 2 liter, three neck flask from 2. 00 moles (48. 6 g.) of magnesium and enough methyl bromide to react completely with it in one liter of dry ether. To this cold solution was added 1, 75 moles (212 g.) of trifluoromethylacrylonitrile dissolved in 200 ml. of dry ether. After refluxing for four hours, during which time a wax-like precipitate formed, 10% sulfuric acid was added and the precipitate dissolved. The ether solution was separated and the water layer extracted with additional ether. After drying and removal of the ether, fractionation gave the following:

Redistillation of cut III gave 15 g. of material, bp. 127.0-7.5 C.,
n2? 1.4108, d2J 0.9350.
Found: C, 49.7; H, 5.99; N, 9.13.
Cut_bp. v/t.
I 75-80 26 g.
U 80-124 10 g.
III 124-126 17 g.
IV Residue

wv Acetate Time Rate of N2 liters : wt. Fractionation % Yield
Run of Reaction Crude Product ucuc-r CutH Cut HI OF* t
per hour b. p. wt. b.p. wt. b. p. wt. CH2aC*CN
1 45 gm. 4? min. 11 37 gm.; 73 -78.5 11.4^ 80-120 8.5 Residue 37.9
2 45 gm. 36 min. 17 40.5 75 -80 4 (2) 35-140 6.8 lio-152 22*2 13. 3 conv. 26. 2 yield
3 45 gm. 285 min.. 10 26.6 75 -105 11.5 105-120 10.5 Residue 38. 3
% 45 gm. 42 min. 10 32.0 75 -80 7. 1 80-144 4.2 Residue 23.6
5 45 gm. 57 min.. 10 37. 5 73 -85 14.0 175-140 9.0 Residue 46*6
(i) n2^4 s 1.3261. d2!'4 s 1. 1753, MRq calc'd for c4h2F3]sr> 20.79, Founds 19. 96.
Product is CH2* C-CM
(2) Mainly CH-COOH b.p. 118*
(3} Mainly CH3-(j:-OOCH3, b.p. 150* CN
(4) Washed with water, NaoGO, sotation, dried and distilled.

Boiling t. ^ MRD f&c *H *Mic Formula Point ^..................... "-;-1;;" ............1 '"-
* C Gale. Fd. Calc. Fd. Calc. Fd. Calc. Fd.
CH3-9C-CH3 118.0-8.4 25 1.3581 1.2701 26.53 26.99 38.47 38. 3 4.52 5.00 OH CH3-CCH-CH- 58/5.5 25 1,3866 1. 3282 28,04 28,00 38.0 38.165,74 5.34 OH OH
CH3-C= CHCOG1 99.5-100.5 23 1.3916 1. 3037 31.47 29.71 20. 6 20.2
CH3-C=GH-CH2OH 144-5 22 1.3761 1.2021 26. 36 26.75 42, 8 42,4 5.04 5.44 CF3
CH2=6-CH=CH2 35.0-5.5 22 1.3431 1. 037 24. 36 24.87 49, 3 49.5 4.13 4,34
CH3MgBr+ CH2=C-CN 127,0-7^5 27 1,4108 0,935 49.7 5.99 9.13

Discussion HI
The third portion of this dissertation contains a discussion of methods of the preparation of certain substituted styrenes. Some trifluoromethyl sub stituted styrenes have been reported previously (25) (2) (21). The synthesis of these monomers was generally carried out by the dehydration of substituted alpha-methylbenzyl alcohols which were prepared by means of the Grignard reagent:
\S >MgBv + CH3CHO .....> \S ^C-CH3
(CF3)n (GF3)n OMgBr
(CF3)n OH' (CF3)n
n a 1 or 2
The Grignard reaction was carried out with bromo, bromochloro and bromodichloro derivatives of (trifluoromethyl}benzene and bis-(trifluoromethyl )benzene. In each case in which a Grignard reagent was
formed the magnesium reacted with the bromine substituent and not

with chlorine or fluorine. The Grignard reagents were condensed with acetaldehyde and the salt hydroiyzed to the substituted alpha-methyls-benzyl alcohol. Dehydration of these alcohols was accomplished with a suspension of phosphoric oxide in dry benzene at room temperature. It was found that the presence of a trifluoromethyl group ortho to the bromine atom slightly retarded the formation of the Grignard reagent. Otherwise, the trifluoromethyl group appeared to have little, if any. effect on the Grignard reagent.
It has been claimed in a patent by Dickey and Stanin(i 1) that it is possible to prepare styrenes substituted in the alpha position of the vinyl group by the reaction of a Grignard reagent with a ketone followed by dehydration of the intermediate carbinol. For example, alpha-difluoromethyl- and alpha-trifluoromethylstyrenes may be prepared by reacting difluoroacetone or trifluoroacetone with a phenyl -magnesium bromide, hydrolyzing the salt and dehydrating the carbinol6 to the alpha-difluoromethyl- and alpha-trifluoromethylstyrenes. A list of eight dehydrating agents was given, among which were phosphoric oxide (under 1 mm. pressure) and sodium acid sulfate. As will be shown later, dehydration with potassium acid sulfate was not accomplished nor was dehydration accomplished with phosphoric oxide under vacuum. It was later learned that "The directions in the patent are entirely speculative" (9) indicating that the work had never been

Since considerable interest has been indicated in fluorine-and trifluoromethyl-containing styrenes it seemed of value to prepare a series of alpha-trifluoromethylstyrenes containing trifluoromethyl groups on the ring. Also, due to the indication that the work described above had never been performed, it seemed advisable to first attempt the preparation of alpha-trifluoromethylstyrene.
The reaction of phenylmagnesium bromide and trifluoroacetone gave alpha-trifluoromethyl-alpha-methylbenzyl alcohol as claimed by Dickey and Stanin. However, attempted;dehydrations with potassium acid phosphate, sulfuric acid and phosphoric oxide under vacuum were unsuccessful. Dehydration was finally accomplished by distilling the carbinol from phosphoric oxide at atmospheric pressure. The styrene was isolated by fractionation of the crude distillate. Physical properties, infrared spectra and elemental analyses served to characterize the styrene.
3-Trifluoromethyl-alpha-trifluoromethyl8tyrene and 3,5-bis-(trifluoromethyl)-alpha-trifluoromethylstyrene were prepared by the procedure as described for alpha-trifluoromethylstyrene, employing the appropriate Grignard reagent and trifluoroacetone. Dehydration of the intermediate carbinols was easily accomplished by distilling from phosphoric oxide. The substituted-styrenes were characterized by

physical properties, infrared spectra and elemental analyses.
It is of interest to note that the boiling points of the three intermediate carbinols are not in line for materials of increasing molecular weight in a homologous series. This may be readily seen in the table below:
Boiling % Absorption
Carbinol Point Yield 2. 79y 2.89*/
C6H5C(CF3>(OH)CH3 62-66/4.5 mm. 74.7 --- s
3-(CF3)C^4C(CF3)(OH)CH3 87.5-88.0/4 mm. 67.8 w m-s 3,5-{CF3j2C&H3G(CF3)(OH)CH3 59.0-60.0/3 mm., 40.2 m m
e ss strong} m-s ss medium-strong; m s medium;
W as .weak,.:
Evidence has been presented (12) to show that an unassociated hydroxyl group exhibits an absorption band at approximately 2.75//, while an associated hydroxyl absorbs near 3. 00//. As is readily seen in the above table, the absorption peak at 2. 79// increases in strength as the aromatic ring becomes increasingly electronegative due to the presence of trifluoromethyl groups. On the other hand, the absorption band at 2. 89/ constantly decreases in intensity. If the 2. 79/ and 2. 89/ bands are assigned to unassociated and associated hydroxyl groups, respectively, it becomes apparent that the degree of association is being constantly decreased as the ring becomes more electronegative. It is known that

Association of this type would decrease the boiling point of the material as the degree of association increased. : The second possible type of association is of the inter-molecular type:
\; ch
Association of this type would increase the boiling point by increasing the effective molecular weight of the carbinol. As the degree of associ* ation decreased the boiling point would decrease. Since this is the observed result, it becomes apparent that this is the actual type of
tne presence of an electronegative group on the ring formed by hydrogen bonding decreases the tendency for that molecule to associate. Since the trifluoromethyl group is extremely electronegative it would be expected that its pretence would lessen the association as actually appears to be the case.
Two types of association are possible in this case, The first is of the intra-molecuiar type:

association occurring in the above carbinolo.
Warner (29) has reported that styrenes may be prepared from an aryl Grignard reagent and certain fluoroolefins. For example
This new synthetic technique offers a one-step reaction for the preparation of certain substituted styrenes. Since several new arylhalides and fluoroolefins were available It seemed worthwhile to study the reactions of a few of these.
If phenyimagnesium bromide could be made to react with 1,1,2-trifluoroethylene one would obtain alpha, beta-difluorostyrene in one step, The reaction was Erst attempted at ice-water temperature. However, the olefin was not absorbed. Recycling the olefin for several hours resulted in only a very little being absorbed. Hydrolysis of the salt and fractionation produced only benzene. It was assumed that the lack of reactivity may be due to the extremely short contact time resulting from the volatility of 1,1, 2-trifluoroethylene. The reaction was repeated by cooling the Grignard solution in Dry-Ice and acetone. 1,1 <' 2 -Trifluoroethylene was added and the mixture allowed to stand for several days* Hydrolysis and fractionation gave only benzene. It
C^HgBr + Mg C^HgMgBr
C6H5" 4- CF-^CCl-.----CgHgGl
C6H5~ + *MgBr
? G^HgCF-siCClg*
C6H5CF*CC12 + F*

was not practical to attempt this reaction under pressure after Warner (29) reported a violent explosion during the attempted addition of ethyl~ magnesium bromide to 1,1-dichloro-2, 2-difiuoroethylene

Experimental III 1, Reactions ax PaenylmagneBium Bromide:
a. With Trifluoroacetone: A one liter, three neck flask equip* ped with a stirrer, pressure equalized addition funnel and ice-water reflux condenser topped with a drying tube was charged with 0. 5 mole (12.2 g.) of magnesium turnings and flame dried under an atmosphere of dry nitrogen. The nitrogen was supplied from a cylinder, dried by passing through concentrated sulfuric acid and a tube containing Drierite and admitted to the system through the top of the addition funnel. Dry ether (400 ml. } was added to cover the magnesium and a small portion of bromobenaene added to initiate the reaction. Cooling was often required to control the reaction once it had commenced. After the initial reaction had subsided, the remainder of 0. 5 mole (78, 5 g.) of bromobensene was added at such a rate as to cause gentle boiling of the solvent. An hour was allowed after completion of the addition to insure complete reaction. The Grignard reagent was then treated with 0, 6 mole (67g.} of trifluoroacetone via a gas inlet which replaced the addition funnel and the mixture left overnight. Hydrol* ysis was accomplished by poxiring the ether solution onto lOO^mli. of hydrochloric acid and 1 kg. of cracked ice. The ether layer was separated and the water extracted several times with small portions of ether. The combined extracts were dried and fractionated to give

71 g. (74. 7%) of C6H5C(CF3)(OH)CH3, bp- 62-6* /4.5 mm., n 1.4656,
d2J* 1.2511.
Calculated for C6H5C(CF3HOH)CH3: C, 56.8; H, 4.78; MRD, 41.70 Found: C, 57.01} H, 4.87; MRD, 42.07.
b, With 1,1,2-Trifluoroethylene:
(1) At Ice Water Temperature: One half mole of phenylmagnesium bromide was prepared in 500 ml. of dry ether as described above. For this reaction a flask was used which had a stopcock sealed onto the bottom. The ether solution was transferred to a second flask via the bottom tube and treated with 0. 5 mole (41 g.) of 1,1,2-trifluoroethylene at ice water temperature. During the two hours of stirring most of the 1,1, 2-trifluoroethylene was recovered in a cold trap protecting the system. Hydrolysis was performed by pouring onto
1 kg. of cracked ice and 100 mi. of hydrochloric acid, the ether separated, and the water layer extracted. Fractionation of the dried extracts gave only benzene and a very small amount of the unidentified high boiling material.
(2) At Dry Ice Temperature: One half mole of phenylmagnesium bromide was prepared in 500 ml. of ether and filtered into
a second reaction flask as described previously. After cooling in Dry

Ice. 0. 5 mole (41 g.) of 1,1, 2-trifluoro ethylene was added and the mixture allowed to stand for several days at this low temperature. The solution was allowed to warm slowly to room temperature and most of the olefin recovered. Hydrolysis was performed as usual and the dried extracts fractionated to give only benzene and a very small amount of unidentifiable material,
2. Reactions of m-(Trifiuoromethyl)phenylmagnesium Bromide:
a. With 1, l-Dichloro-2, 2-difluoroethylene: The following reaction was repeated according to the work described by Warner (29) to determine the reactivity of m-(trifluoromethyi)phenylmagnesium bromide.
One half mole of m-(trifluoromethyl)phenylmagnesium bromide was prepared from 0, 5 mole (112. 5 g.) of m-bromobenzotrifluoride and 0, 5 mole (12. 2 g. ) of magnesium in 300 ml, of dry ether by the usual procedure in equipment as described in part lb of this section and filered into a second reaction flask equipped with a gas inlet, stirrer and ice-water condenser where it was treated with 0.5 mole (66, 5 g,) of 1, l-dichloro-2. 2-difluoroethylene. Hydrolysis was carried out after three hours, the ether separated, the water extracted with ether and the extracts placed to dry over calcium chloride. Fractionation gave 50 g. (38. 7%) of m-CF3C6H4CFs CC12, bp. 65-7* /3-4 mm. Warner reported a yield of 8%.

b, With Trifluoroacetone: The Grignard reagent was prepared from 0. 325 mole (73 g.) of m-bromobenzotrifluoride and 0. 325 mole
(7. 9 g.) of magnesium in 400 ml. of dry ether in an appropriate flask and treated with 0.325 mole (36.4 g.) of trifluoroacetone. After working up by the usual procedure, the dried ether extracts were fractionated to give 56. 4 g. (67. 3%) of m-CF3C6H4C(CF3)(OH)CH3, bp. 87.5-88.0/4 mm., n2* 1.4148, d2| 1.4267. Analysis:
Calculated for m-CF3C6H4C(CF3)(OM)CH3: C, 46.6; H, 3.13; MRD, 46.32
Found: C, 46.71; H, 3.04; MRD, 45.30.
c. With Formaldehyde: The Grignard reagent was prepared from 0. 5 mole (112 g.) of m-bromobenzotrifluoride and 0. 5 mole
(12. 2 g.) of magnesium in 500 ml. of dry ether. Formaldehyde, formed by the pyrolysis of 0. 56 mole (50 g.) of paraformaldehyde, was passed into the ether solution. Hydrolysis was carried out after standing for three hours by the usual procedure, the ether layer separated, the water extracted with ether and all extracts dried. Fractionation gave 29 g. (33.0%) of m-CF3C6H4CH2OH, bp. 66-70* /l. 5-2 mm., n2^ 1.4606, d24 1.2949. Analysis:
Calculated for m-CF3C6H4CH2OH; C, 54. 6; H, 4. 01; MRp, 37.08

Founds C, 54.48; H, 3.89; MRD 37.28.
3. Reactions of 3,5-Ms(Trlfluorornethyl)p&enylmagneeium Bromide:
A. With Trifluoroacetone: The Grignard reagent was prepared in the usual fashion from 0. 645 mole (189 g.) of a mixture of 3,5-bis-(trifluoromethyUbromobenzene and a very small amount of 2,5-bis(tri-fluoromethyl)bromobenzene and 0. 7 mole (17 g.) of magnesium in 400 ml. of dry ether. The ether solution was treated~with 0. 7 mole (78.5 g.) of trifluoroacetone and worked up in the usual manner after standing overnight. The dried extracts were fractionated to give 82 g. (40. 2%) of 3, 5-(CF3)2C6H3C(CF3)(OH)CH3, bp. 59-61* /3 mm. with a center cut taken with the following constants: bp. 59-60* /3 mm., n2^ 1.3966. d2| 1,5118. Analysis:
Calculated for (CF3)2C6H3C(CF3)(OH)CH3: C, 40. 6j H 2,17} MRD, 50.94
Found: C, 40.43; H, 2.51; MRD, 51.91.
4. preparation of the Substituted Styrenes:
The equipment used for the following three reactions was the same, consisting of a 500 ml. three neck flask equipped with a stirrer, addition funnel and short fractionation column topped with a variable take-off distilling head. A cold trap in Dry Ice and acetone was used to protect the system. The procedure followed was the same in ail

three cases as described in the first preparation.
a. Preparation of of-Trifluoromethylstyrene: The flask as described above was charged with 0.1 mole (15,0 g,) of phosphoric oxide and cooled in ice-water. One tenth mole (19 g, ) of -methyl- ot-trifluoromethylbenzyi alcohol was added and the mixture stirred until an even paste formed. The ice bath was then replaced by a heating mantle and the mixture slowly heated until the desired product distilled
at 148-157* C. Fractionation of this crude distillate gave 11. 6 g. (66.9%) of C6H5C(CF3)sCH2, bp. 148.0-151.0* G. with 9. 1 g. taken bp. 148.0-8,5* C, n2^ 1.4603, d24J 1.167. Analysis:
Calculated for C6H5C(CF3)= CH2: C, 62. 7; H, 4. 09; MRD> 41. 39 Found: C, 62.57; H, 4.16; MRD 40.42.
b. Preparation of 3-Trifluoromethyl- oc-trifluoromethylstyrene: The above procedure was repeated using 0. 2 mole (28.4 g.) of phosphoric oxide and 0. 194 mole (50 g.) of 3 *Trifluoromethyl *-methyl- <*-tri-fluoromethylbenzyl alcohol to give crude material boiling up to 151* C. Fractionation gave 23 g. (47.7%) of 3-(CF3)C6H4C(CF3)sCH2,
bp. 151-9* C. with a center cut taken at 157-8* C., n2^ 1.4151, d22 1,346.

Cut bp, wt. n*g
I 157-161 11 g. 1.3916
II 161-172 7.5 g. 1. 3921 in 172-186 10.0 g. 1.3938
IV 186-187.5 12 g. 1. 3954
V 187.5-188.5 25 g. 1,3961
Cuts III, IV and V are recovered carbinol. Infrared data indicated Guts I and n to be the desired styrene. Guts I and II represent a 9. 3% conversion.
Calculated for 3-(CF3)C6H4C{CF3)=CH2: C, 50.01; H# 2.52; MRD, 46.00
Found: C, 49.89; H, 2.89; MR^, 44.68.
c. Preparation of 3, 5-bis(Trifluoromethyl)- -trifluoromethyl styrene: The procedure of part 4a was repeated in the following two sections to give the results described below.
(1) The flask was charged with 0. 2 mole (28. 4 g.) of phosphoric oxide and 0. 237 mole (77 g.) of 3, 5-bis(trifluoromethyl)-0(-methyl- (-trifluoromethylbenayl alcohol added as described. The mixture was heated until material distilled between 170-180* C, Redistillation gave:

(2) The procedure was again repeated using 0, 1 mole (14. 2 g.} of phosphoric oxide and 0.144 mole (47 g.) of 3, 5-bis(tri-fluoromethyl)-Of-methylOf-trifluoromethylbenayl alcohol to give the crude distillate which gave the following upon fractionation:
Cut bp, wt. n22 d2|
I 158 -161.5 2,2g. 1.3922 1.457
II 161.5-165 1. 1 1.3921
III 165 -178 6.2 1.3926 1.473
Cuts I and II from part 1 above and the above three fractions were combined and fractionated to give 19. 8 g. of 3, 5-(CF3)2C.*H3-C(CF3)=CH2 bp, 56-63* /15 mm. with a 14. 8 g, center cut taken with the following constants, bp. 59. 5-60, 0* /15 mm., n2^ 1. 3921, d2| 1,456. Analysis:
Calculated for 3, 5-(CF3)2C6H3C(CF3)= CH2: C, 42.8j H, 1.63; MRD, 51.63
Found: C, 42.41; H, 1.92; MRn, 50.43.

Formula Boiling Point t C nD d4 MR D % C ft H % Misc,
C Calc. Fd. Calc, Fd. Calc. Fd. Calc. Fd.
^? Js*3 C6H5CCH3 OH 62-6/4.5 22 1. 4656 1. 2511 41.70 42, 07 56.8 57.01 4. 78 4.87
148,0-8.5 21 1.4603 1. 167 41,39 40.42 62.7 62.58 4.09 4.16
CF3 3-(CF3)C6H4 9F3 3-(CF3)C6H4C= CH2 157-8 22 1.4151 1.346 46.00 44. 68 50.0 49.89 2. 52 2.89
CF3 3,5(CF3J CACCH3 OH 59-60/3 22 1.3966 1,5118 50.94 51.91 40.6 40.43 2. 17 2.51

9F3 3,5{CF3)2C6H3C= CH2 59.5-60.0/15 22 1.3921 1.456 51. 63 50. 43 42.8 42. 41 1. 63 1.92
3-(CF3)C6H4CH2OH 66-70/1,5-2 21 1.460 4 1.2949 37. 08 37. 23 54.6 54.48 4.01 3.89

Hydrocarbon esters of perfluorocarboxylic acids have been prepared by the pyrolysie of a silver salt of a perfluorocarboxylic acid in the presence of an olefin. For example: pyrolysis of silver trifiuoro-acetate in the presence of butene-2 gave sec-butyl trifluoroacetate. This is a new and unexpected reaction of silver salts of perfluorocarboxylic acids. The results of this work indicate that the stability of the perfluoroacyloxy radical is considerably greater than previously supposed.
A study has been made of the preparation of 2-trifluoromethylbutadiene-!, 3 from trifluoroacetone. Two diverse methods have been found. The first procedure involves the preparation of a cyanohydrin which is in turn converted to a keto-alcohol by reaction with methyl-magnesium bromide. Subsequent reduction of the carbonyi group and dehydration yielded the desired diene. The second procedure requires the preparation of 3-trifluoromethylcrotonic acid which may be reduced to a primary alcohol and dehydrated to 2-trifluoromethylbutadiene-1, 3.
Several trifluoromethyl-containing styrenes have been prepared by reaction of an aryl Grignard reagent and trifluoroacetone
followed by dehydration of the intermediate alcohol. The effec^ of

trifluoromethyl groups on hydrogen bonding and the effect of hydrogen bonding on boiling points and infrared spectra is discussed.
The following new compounds have been prepared and characterised: 2-trifluoromethyi-2-hydroxy-butanone-3, 2-trifluoromethyl -butanediol-2 3, 3-trifluoromethyl-2-buten-l-ol, 3-trifluoromethyl-of-methyl-o^-trifluoromethylbenzyl alcohol, 3-trifluoromethyl- of-trifluoro-methylstyrene, 3, 5-bis(trifluoromethyl)-Of-methyl- *trifluoromethyl-benzyl alcohol, 3,5-bis(trifluoromethyl)-oc-trifluoromethylstyrene and 3-trifluoromethylbenzyl alcohol.

1. Alexander, E. R.Ionic Organic Reactions. 1st ed., pp. 45-9,
John Wiley and Sons, Inc.
2. Bachman, G. B, and L, L, Lewis. J. Am. Chem. Soc. 69, 2022
3. Bullitt, O. H. Jr. U. S. Patent 2, 559, 630 (1951)
4. Campbell, K. N., J. O, Knoblach, and B. K. Campbell, J. Am.
Chem. Soc. 72, 4380 (1950)
5. Chaikin, S. W. and W. G. Brown. J. Am. Chem. Soc. 71, 122-5
6. Darrall, R. A., F. Smith, M. Stacey and J. C. Tatlow. J. Chem.
Soc. 2329-32 (1951)
7. Dickey, J, B, U. S. Patent 2,472,812 (1949)
8. Dickey, J. B. U. S. Patent 2,541,466 (1951)
9. Dickey, J. B. Tennessee Eastman Company, Private Commu-
10. Dickey, J. B. Eastman Kodak Co., Private Communication to
D. A. Warner.
11. Dickey, J, B. and T. E. Stanin. U. S. Patent 2,475,423 (1949)

12. Ferguson, L. N. Electron Structures of Organic Molecules.
lsted., pp. 251-3, Prentice-Hall.
13. Hals, L, J,, T. S. Reid and G. H. Smith. J. Am. Chem. Soc,
73, 4054 (1951)
14. Haseldine, R. N. J. Chem. Soc. 2856 (1949)
15. Haszeldine, R. N. J. Chem. Soc. 4259 (1952)
16. Henne, A. L. and P. E. Hinlcamp. J. Am. Chem. Soc. 76,
5147-8 (1954)
17. Hill, C. M., L. Haynes, D. E. Simmons and M. E. Hill.
J. Am. Chem. Soc. 75, 5408 (1953)
18. Hill, H. M. and E. B. Town, (to Eastman Kodak Co.). U. S.
Patent 2,490, 753 (1949)
19. Kirshenbaum* A. D, A. G. Streng and M. Hauptchen.
J. Am. Chem. Soc. 75, 3141-5 (1953)
20. Lovelace, A. M. Ph. D. Dissertation, University of Florida
21. McBce, E. T. and R. A. Sanford. J. Am. Chem. Soc. 72,
4053-5 (1950)
22. Minnesota Mining and Manufacturing Co,, Technical Bulletin
on Heptafluorobutyric Acid, page 5.
23. Organic Enactions, Vol. II, pp. 178-223, John Wiley and Sons, Inc.
24. Organic Reactions. Vol. VI, pp. 469-509, John Wiley and Sons, Inc.

25. Renoil, M. W. J. Am. Chem. Soc 68, 1159 (1946)
26. Schwarz, M. Florida State University, Private Communication.
27. Sawarc, M. and J, Murawski. Trans. Faraday Soc. 47, 269
28. Tarrant, P. and A. M. Lovelace. -J, Am, Chem. Soc. 76,
3466-8 (1954)
29. Warner, D. A. Ph. D. Dissertation, University o Florida (1953)

Robert Edward Taylor was born on October 24, 1929* in Eaton Rapids, Michigan. He attended the public school of Belle Glade, Florida, graduating from Belle Glade High School in May, 1947,
In September, 1947, he entered Florida Southern College and received a B. S. degree in June, 1951. While an undergraduate he worked three years as a laboratory assistant at Florida Southern College.
He entered graduate work at the University of Florida in June, 1951, and received the M. S. degree in August, 1952.
While working toward the Ph. D, degree he has been a research assistant under a grant from Smith, Kline and French, Inc. and on a project sponsored by The Office of the Quartermaster General under Contracts DA44-109-qm-522 and DA44-109 -qm-1469.
The author is a member of Omicron Delta Kappa, Gamma Sigma Epsilon, Sigma Xi and the American Chemical Society.

This dissertation was prepared tinder the direction of the
Chairman of the candidate's Supervisory Committee and has been approved by all members of the committee. It was submitted to the Dean of the College of Arts and Sciences and to the Graduate Council and was approved as partial fulfilment of the requirements for the degree of Doctor of Philosophy.
January 29, 1955
Dean, College of Arts and Sciences
Dean, Graduate School

i. 'ffppjcfl.Y_JL_C^gy _, as copyright holder for the
aforementioned dissertation, hereby grant specific and limited archive and distribution rights to the board of trustees of the university of florida and its agents. i authorize the university of florida to digitize and distribute the dissertation described above for nonprofit, educational purposes via the internet or successive technologies.
this is a non-exclusive grant of permissions for specific off-line and on-line uses for an indefinite term. off-line uses shall be limited to those specifically allowed by "fair use" as prescribed by the terms of united states copyright legislation (cf, title 17, u.s. code) as well as to the maintenance and preservation of a digital archive copy. digitization allows the university of florida or its scanning vendor to generate image- and text-based versions as appropriate and to provide and enhance access using search software.
this grant of permissions prohibits use of the digitized versions for commercial use or profit. signature of copyright holder
printed or typed name of copyright holder/licensee f
personal information blurred
date of signature
please print, sign and return to: cathleen martyniak uf dissertation project preservation department university of florida libraries p.o. box 117008 gainesville, fl 32611-7008
in reference to the following dissertation: author: taylor, robert
title: the preparation of certain fluoroolefins containing the trifluoromethyl
group, (record number: 549785) publication date: 1955

University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs