THE PREPARATION AND REACTIONS OF
SOME ALIPHATIC FLUOROETHERS
JOHN ADAMS YOUNG
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
The term nfluoroether" is employed in this paper to denote any
aliphatic ether in which one or more of the alkyl hydrogen has been re-
placed by halogen, at least one of the substituents being fluorine,
Thoroughly correct usage would no doubt restrict this term to ethers in
which fluorine is the only substituent, but since many of the ethers here
described are actually fluorochloroethers, fluorobromoethers or fluoro-
chlorobromoethers a slight irregularity in the interests of simplicity
may be condoned.
Nomenclature recommended by the Internation Union of Chemistry
at Liege for ethers is both cumbersome and obtuse; therefore, the com-
pounds in this paper are named as derivatives of dialkyl ethers wherever
possible. It is much easier to grasp the structural relationship of
CIC12CF2 OC2t and CC12= CHOC2S when those are named respectively
o ,oj-difluoro-(33-dichlorodiethyl ether and( ~-fdichlorovinyl ethyl ether
than when the same compounds are named, according to the I. U. C. system.
2,2-dichloro-,1l-difluoroethoxyethane and 1,l-dichloro-2-.ethoxyethene.
Because of the availability of ethyl alcohol and the fact that
many of its compounds boil readily at atmospheric pressure, this alkyl
group appears in most of the ethers prepared. In all but a few cases,
the fluorinated portion of the ether molecule also consists of a two.
carbon residue. The terms olefinn" and "saturated compound" used in conw
nection with the preparation of various ethers refer to derivatives of
ethylene and ethane respectively unless otherwise specified.
The results of this investigation have been divided into the three
main parts listed in the Table of Contents. Some of the reactions will
be found in both Chapter II and Chapter III) although this led to some
slight duplication, it was felt that this course was preferable to the
alternative of large, heterogeneous collections of rather unrelated data.
Where a reaction is mentioned only in passing, reference is made to a more
complete discussion. Fluorination of polychloro ethers not containing
fluorine has been tabulated under Chapter III rather than under a separate
heading since these results are closely allied to the other material in
this section. General conclusions are stated both in the appropriate dis-
cussions and in the General Summary.
All teneoratures are Civen in degrees centigrade. Tho molar re-
fraction (1%) values are calculated frcmn experimental data using the
Lorenren orent equation; theoretical values arc calculated by use of the
additive values for elements and bonding Given in Lange t Handbook of
Chemistry, 6th ed., published in 1946.
TABLE OF CONTENTS
PREFACE . . . . . . . . .
. . ii
LIST OF TABLES . . . . . . . . . vi
INTRODUCTION . . . . .
PREPARATION OF INTERMEDIATES . .
0 4 44 5 4 4 S 3.C~
# 4 . 0 0 0 0
A. Discussion. . . .. . .* . . . . 5
B. Preparation of Olefins.
a a 0 0. 0
C. Preparation of Saturated Compounds' .
D. Preparation of Intermediate Ethers.
S. . . 13
PREPARATION OF FLUOROETHERS. . . . . . . . 17
A. Discussion.. . . . . . .
B. Experiental. .. . . . . ...
. . . 17
1. Ethers prepared from olefins. . . . .
2. Ethers prepared from saturated compounds. ..
3. Etherification of CF3CH2Br and the CHF2- series
C. Conclusions. . . . . . . 6
REACTIONS OF FLUOROETHERS . . . . . . 53
A. Discussion. . . . . . . . . 53
1. Hydrolysis of fluoroethers . . . . . .53
2. Formation of polyethoxy compounds . . . 61
3. Reaction of fluoroethers with aluminum chloride 62
i. Thermal decomposition of fluoroethers . . . 65
5. Reaction of fluoroethers with antbimony fluoride 66
B. Experimental. . . . . . . . . 75
Hydrolysis of fluoroethers. . . . . .
Preparation of other derivatives of chloro-
fluoroacetic acid . . . . . . .
Reaction of fluoroethers with aluminum chloride
Thermal decomposition of fluoroethers . . .
Reaction of fluoroethers with antimony fluoride
TA3BI OF COTENTMS (6ontiaued)
C. Conicusions. *
GErnTAL SOim'lU ,...
BIB IOf OGORK Y. ,
# b,* 0 *. . 4 & a
4 4 0
A' 0 9 4-C~-
.* . 90
* 0* 937
a 9 i $5'
* * 96
* 4 4 *97
LIST OF TABLES
0 ~~ 0 4 0. 0 0904 *440 0094 0
. 0 0 9
* 0 0 4
"J* a 0 0 0 0 0 0
14. Etherification of
15. Etherification of
16. Etherification of
17. Etherification of
18. Physical Constants
19. Esters Prepared by
20. . .
21. . . .0
22. . . . .
23. . . 0 *. .*
24#. * 0 0
I O 0 9 0 0 0 0 0 . 4 * 0 0 0
rd From Oefine ~. *
rod From Saturated Compounds .. .
CF 3CH2r . . . . . . . . .
CHF2CHc2l. .* 0 * 0 a 0 0 * *
SHydrolysis of o ~ a-Difluoroethers.. .
.40*. .. ....... . 0 0
The chanistry of the fluoroethers diffcra profoundly from that of
the dialkyl ethers. In most 'clemcntary organic texts ethers are S-narily
dispoed' of in a very short chapter, two or three preparations and reac-
tions being presented. Even'in references on physical organic theory l
discussion of the etherification and hydrolysis mechanisms is very brief.
Reactions of. aliphatic ethers lead in. general only to the allyl halides
or alcohols used as starting rater ials, so that their chemistry is often
a closed circle. As synthetic intermediates,' ethers are rarely used.
Coapare with this behavior some instances from the field of fluoro-
others. Etherification of CliF2C7C12 can givie nine possible products, six
of which have been found actually to occur, Reaction of fluoroethors may
lead directly to alkyl halides, .aeyl halides, simple esters, orthoesters,
acetals or vinylethers, some of which are difficult to prepare by any other
means. The very mechanism of formation. of many fluoroethers is not laiown
with certainty, while even less work has been done on the theory of their
reactions. The presence of several fluorine atoms, alone or with other
halogens, usually confers great stability on a molecule, It has been sur-
prising to find that the fluoroethers are unstable in .aay cases and at
least1as reactive as the corresponding chloro- or bromo-compomuds*
Fluoroethers were first prepared and invostigatod by Fred Swarts,
the Belgian chemist whose monumental work over a span of half a century
laid the foundation for modern organic fluorine chemlstry7 Between 1920
and 1935 comparatively little was done in this field, but under the impotus
of the fluorine program, necessitated by the atom bomb development, new
strikes were taken. Since the war two new processes, addition of alcohols
to fluoro~lefins and electrolytic fluorination, have brought the field of
fluoroethcrs to its present state. Fluoroethers have been proposed or
utilized as heat transfer agents, dielectrics and lubricants, Although
their properties are in some cases excellent, their present high cost re-
duccs their industrial value. One outcroirth of the work herein described
is the use of fluoroothers as intermediates in a convenient synthesis of
the more difficulty prepared haloacetic acids,
The present project had its inception in the observation that
CIHClCF20C 2~t, when allowed to stand with concentrated sulfuric acid and
subsequently diluted with water, gave a good yield of an organic product
later identified as ethyl chlorofluoroacetate. Since this chlorofluoroaceto
group had heretofore been prepared only by .the use of the extremely poison-
ous methyl fluoroacetate, this reaction constituted a valuable new method
of synthesizing derivatives of chlorofluoroacotic acid, and accordingly
seven new compounds containing this group were prepared from the ester by
The ether, CHFC1CF2OC2I2H, was made by the addition of ethyl alcohol
to CF2g= CFC1. Addition of alcohols to olefins had proved itself an ac-
cepted synthesis of fluorocthero; however, an unsymmetrical olefin such as
CF2gCFCl can theoretically add alcohol in two different ways to Give two
different products, C!IFClCF2GC2IH and CIIF2C FCOC2015, and fairly scanty
evidence had been offered to shoTr the actual orientation of the adduct.
Hydrolysis to the corresponding ester with such facility offered an excel-
lent proof of structure for fluoroethers made in this manner. "ork was
extended therefore to other olefins to test the general utility of the re-
action both as a proof of structure and as a synthetic method.
A supplementary method of ether synthesis, used previously by
Starts 2728 and l:c'ee and Bolt, was also developed. In the addition
method by which CIlFClCF2OC2H5 was made, the olefin was bubbled through a
solution of sodiiun othylate in ethyl alcohol, this being the Coneral method
of preparing ethyl fluoroethers, If an appropriate saturated compound were
to be substituted for the olefin, such a reaction nedium would be capable
of abstracting HIX giving an olefin which would then add to the alcohol to
give tih desired ether. This procedure was successful j and maxy of the
difficulties involved in storage and handling of compounds boiling bclow
room teapecrature wore eliminated; furthermore, equal or better yields of
others were generally obtained.
Uoe of saturated reactants rather than olcfins led, however, to
further complications. It has been stated that an olefin could conceivably
add alcohol to give two different products. Since saturated compounds hav-
ing both lalogen and rvhdrogen on each carbon, or two different halogens on
the same carbon, could presumably split out 1oX to give two different olefins,
the number of possible ethers increases rapidly, The equations given below
illustrate this point:
SCFg= C'2 -- CIF2CF2OC2H5
CF CCII F2C!:CLOC2.U
CCF2 2ClF2OC2 1
~ *^ cnhcicpzocaite
The reaction between CIIF2CiHCI and alcoholic potassium lydrorido
could rive three olofins and consequently six simple ethers, not to mention
the secondary reactions usually present. It was, therefore, of prime toi
portance to determine whether or not either the dehaloccnation or the addi-
tion was polydirectional. In solving this problem, we did indeed find
that thee use of saturated compounds often led to very complex suites of
Study of the reactions of these ethers had already resulted in
the useful iydrolysis reaction. Further investigation chowed that the
ethers underwent a good many reactions, nost of which gave unexpected
products. In spite of the apparent diversity of such reactions as acid
hydrolysis, formation of esters and acetals, reaction with aluminum chloride,
thermal decormosition, and fluorination, it has been found possible with the
aid of nodern orCanic theories to express all those as variations of one
The foregoing has been written merely to give some idea of the
scope of the paper and the reasons behind the directions in hich the work
grew. It is by no neans a diCest or ciuxnarx of the problems encountered.
Complete discussion of these problema is civen in the sections so entitled
of Chanters II and III, and a review of the conclusions reached is pre-
sented in the General Sumary.
With few exceptions the intermediates used in this investigation
were prepared by the standard methods of fluorine chemistry; halogenation
fluorination, dehalogenation, dehydrohalogenation. A brief outline of
the synthesis is given for each compound listed.
Occasionally, special treatment was necessary. Since most of the
olefins added alcohols readily, alcohol could not be used as a solvent for
dehydrohalogenation ihen the olefin itself was desired since yields would
have been greatly reduced by formation of the corresponding ether.
COF2 CIHBr was prepared in fairly good yield by using no solvent at all.
CF3 CCI" CC1CP3 was synthesized through a novel procedure originated
by Henne and Trott3 in which CC12 CC1CC1= CC12 is reacted with SbF3Cl2.
Pentavalent antimoony is a good chlorinating agent and converts the diolefin
into CC1,CC1= CC1CC13 by 1-4 addition; the terminal allylie 0013 groups
are then capable of complete halogen exchange with the antimony fluoride,
leading to the desired olefin.
Dichlorovinyl ethyl ether has been known for a long time; however,
its production directly from CHIFCHGI2 is new and interesting.
The method of fluorinating CI1C CITC2 used by Huakins 1was im-
proved with respect to both efficiency of fluorination and actual conver-
sion. Whereas his crude product never gave better than a 30% conversion to
CIIF2CHC12 and contained at least twenty mole percent CHFCICHC12, adoption
of a partial reflux method permitted us to obtain conversions and yields
of CHF20CH12 respectively of 57% and 87%. The crude product contained
less than 10, of the rono-fluoro co..p)ound, CTCI1C;!C1., and even ravo as
iuch as 10Q conversion to the tri-fluoro comound, CTF2CIFCl.
A moderate number of fluorinrated intermediates are nT available
co-morcially, A list of those applicable to this wcrk follows, the trade
dcoignation also being cilven.
Chemical Formula Trade Uame
1I CIF2CII2C1 Genetron 110
2o4 GFela01,FC2 Freon 113
3: CiF2CTIC2 arnd CF2CICIc2C1
(nixod) Genetron 120
I4. ClIF2CF2Cl Freon 124-A
5 CFgiC1C Genetron 101
6, CIF20CCI, and CF201CHC012
(n6ixFd)l Gr-cntron 130 and Gcnctron 121
B, Preparation of Olefins
CF2ClCFC2g 0- CPF2 CFC1
Four hundred cc, isopropanol and 4 moles (262 gan) zinc dust fere
placed in a three-litcr, three-necked flask fitted vith a dropping funnel,
stirrer, and ice-cooled roflutr condenser. From t te top of the condenser,
tubing was led through a bubble counter into a Dry Ice-acetone trap pro-
tected by a calcium chloride tube. The mixture was heated to boiling and
a mall amount of CF2CFC: (Freon 113) added through the h funnel. lihon it
became cvident by the evolution of gas that the reaction had started, the
main bulk of the Freon 113 (hl6 gi,, 3 moles) was added at such a rate
that only very slow reflux resulted, No external heating was necessary,
At the ccnclsion of the reaction, the crude olefin in the trap was remove
fro- the cooling bath and allowed to distill slowly into a second trap,
The yield was 348 (p. or 75% theoretical.
CF2ClC:H2Er- ,r ,uCFg CHBr
Appro:Jmately 2 moles (100 ron) technical potassium hydroxide
was powdered and placed in a three-nc:ed, 500 cc. flask fitted with
dropping funnel, stirrer and reflux condenser. One-half mole (80 gMi.)
CF2CCIlPDr was added slowly and allowed to react with stirring for tree
hours, the olofin distilling out as fcr.ed into a Dry Ice-acetone trap,
The crude olefin was purified by distillation below room temperature as
described above. The yield was 46 Cm. or 64% theoretical.
0CC1 CC1 CF
0C S GCC1CC01== 3 010-.1 1ccl3C= cccc1 --l> CFccCCics cccc
One hund-rcd fifty-three c(. (0.85 nole) anhydrous SbF3 was pondered,
placed in d pressure vessel and cllorirnted for two hours at 1350 in a
rocking autoclave. The vessel was cooled, vented and 38 gn. (0.21 mole)
SbF3 and 130 7r. (0.5 mole) hexacklorobutadiene added. The vessel vas
again scaled and slha-en at 150 for three hours, then cooled below room
temperature, opened and the contents poured on ice and hydrochloric acid.
After washinr with dilute acid and water, the crude product was fractionated,
Giving 80 g. of CF3CCIc- CCCF-3, b.p., 66-70.** The recorded boiling point
is 66C. The yield was 50% theoretical,
1, 1-Dic!loro-2-luoroe the ne:
One role CHF2CHC12 (135 gm.) was added over two hours to a solution
of two noles (112 gCm.) potassium lydroxide in 300 cc. 95% ethanol held at
100 in a two-liter flask fitted with stirrer and dropping funnel. The re-
action was run at 10-156 for five additional hours. The crude product
was isolated by pouring into a larco volume of ice and watcr, washed and
dried. Fractionation gave 69 CI?= CCg00 boiling at 36.40*, or a yield
Material B,.P n
Product.*-r** 36-40 1.3975
GIIF= CC1g.... 37 1.4016
Tids ~as obtained from Cd W. Huskins, who obtained it by pyrolysis
of CF2CCIICCl2. Details of the procedure may be found elsewhere.
C. Saturated Copounds
CC20 1 CHR-- CC2LDE2&B"e---> C3 22
Six hundred gi. (6.2 moles) of vinylidene chloride was placed in
a two-liter, thrce-necked flask equipped with a stirrer and dropping fun-
nel and cooled in an ice-salt mixture. Three hundred twenty cc. (6.2 molest
of bromine was added with stirring over three hours, the temperature not
being allowed to exceed 1$. The crude product, CC12BrCH2Br, was washed
with water, dilute sodium hydroxide, again with water and then dried, after
which treatment it was entirely suitable for fluorination. A few batcher
were slightly colored after the bromination and were distilled in vaeuo
with no loss. Yield, 1523 gai. or 96% of theory.
Three moles (537 eg.) of powdered anhydrous antimony trifluoride
was placed in a pressure vessel and chlorinated with no external heat for
four hours. The pressure head was then replaced with an iron condenser
arranged for reflux. Tubing was led from the condenser exit to the bottom
of a large beaker filled with crushed ice and dilute hydrochloric acid.
Two moles (514 GC.) CC12BrCH2Dr was then added through the condenser and
the vessel immersed in a heating bath held at about 70.* The bath tem-
perature was kept at 70-800 for one hour, then increased to 1200 over two
hours. The condenser was then drained and distillation permitted for a
half-hour while the bath temperature was raised to 160-1800. The organic
layer was separated, washed with dilute sodium hydroxide and water, dried,
and fractionated. Conversion to the trifluoro compound, boiling at 25-
300 C, ranged from 25-55%, and the yield was from 50-86%. The only by-
product was the difluoro compound, none of the mono-fluoro being obtained
under those conditions. The reported boiling point of the trifluoro com-
pound, CF3CIC1r, is 26.50,
CC12= 0C4g CC12BrCH2Br CF2C0CH2Br
Bromination of the vinylidene chloride was carried out as de-
scribed above. The bromo compound was fluorinated as follows:
Three moles (537 gm.) powdered anhydrous antimony trifluoride was
chlorinated in a one-liter, three-necked flask for one-half hour at atmos-
pheric pressure. Reflux condenser, stirrer and liquid temperature ther-
mometer were then attached, two moles ($51 gfin) CC12BrCH2Br added, and the
flask heated rapidly to 800. After two hours the reflux condenser was re-
placed by a downward condenser and the mixture distilled to a liquid tem-
perature of 1300 The crude product was washed with dilute hydrochloric
acid, water, then dried and distilled, giving 10 gm. CF3CH2Br boiling at
26-30, and 19t gn. CF2ClCH2Br boiling at 6--700. This was a 95% yield
and 54% conversion for the difluoro compound, The reported boiling point
for CF20Clf2Br was 66C. 20
CC12~ CH2. D2 10 CCG1BrCH2Br Sb CFC12CITBr
Bromination of the vinylidene chloride was carried out as de-
scribed above. Fluorination was essentially the procedure used for the
difluoro compound, with the following exceptions the antimony trifluoride
was not chlorinated, catalytic amounts of SbCl5 (5-10%) being added in-
steadj the reaction mixture was allowed to reflux for two hours distil-
lation was carried to a vapor temperature of about 1e. Yield of the
purified compound, boiling at 105--1100, was about 88%. The literature
gives a boiling point of 1108.0 20
l,1-Dichloro-2,2-difluoroethano and l-Chloro-1,2, 2-trifluoroethane:
CIF2CHcl2 and CHPF2CHFC1
CHC12C I1-- CCHF2CI)Cl2 and CHFP2CI Fl
Four and four-tenths moles (790 gm.-) anti:ony trifluoride was
chlorinated in a pressure vessel rith no external heat for one hour*
The pressure head was removed and replaced with a sealed metal stirrer
and iron condenser set for refluxA Four moles (675 gmn,) sym-tetra-
chloroethane was then added through the condenser, tubing attached
leading from the top of the condenser to the bottom of a beaker filled
with ice and hydrochloric acidj the vessel dropped into a heating bath
previously warmed to about 100, and the bath heated fairly rapidly to
140o, at whichh point the reaction usually began. The bath temperature
was kept at IO--1500 for about two hours. The condenser was then
drained and distillation allowed for twenty or thirty minutes, After
the usual treatment, the organic layer was distilled, giving LI gm of
material boiling from 1--300 and 307 im. boiling from 5--700 These
fractions correspond to C!F2CIIFC1, whose reported boiling point is 170,
and to CIIF2CHC12, whose boiling point is 600 .. These quantities rep-
resent conversions and yields respectively for CIIFCHFC1 of 9 and 13%;
for CIIF2CHCl2 of 57 and 87%. Other batches gave generally similar re-
suits, although the use of impure CHC12CHC12 gave greatly lowered yields
of the more highly fluorinated products. None of the monofluoro dom-
pound, boiling at 102 20as ever isolated.
a2-Dichloro-, 1-dIifluoroethane s
This coi-.wmuid was obtiL:nvd frC C. '. r-uclins and -n cTI de by the
clJlorlnat on and ubsoequent fluoriniation of vinylidcnc chloride. Dotall
of Cte 2roccd':rc my beo ound olcowLore.
1,1, 2-Tribrano-2, 2-dfluoroothcanos
CCl CH" +CCrl rr::2Cr----3* 2C^2r----r
c" C rIrT:p~ er- c'C'-4' CF "r--"
.'- ',,- r Tra prc2arCd as dOcrItCd previously and lrominated ? f
the follV-:1r.' rocolc.-;r'
'.to nI.iroM cc. carCon tctrua.U.orie unan placed in a ,0jo.cc'., Tirc-
.ne-:c. ;'lc rit t.d WrtV LrC:L--piLns funirtl, tirrCr, *.eo rfl-x c.docr,
and ras c*It tub. rlo, ..o- (98 '":., 0 :61. oln) was addeJd over one i'our
at a rate suice..lr t to :-anta:-i a si,- ;:t: red color an C06 c. (0.6 :.lo)
7 C CHB r waec lbbled in .ir~-altar:cotnl'", he crzude .?rodu: t was T;ched
;rltt dilute codi.':: i;.'-Jrn;d-idc solution, nith water, and dried. Tho CC1 s uas
rcnovcd by .:l.otl. ilnjtlo after Which 122 of c;rC: llr at
10--13 i2 la.lt32, was ol-tained corrcsoirndlr.; to a y71ld of c4.
?Vclical ata fro= the litcrat.un e arot b.,. 1,3.o n. 1.L7-'l9320
2n.' c:!-nC -. Co:u = 1 :,;ac c I l.:.c; ..re: :. C Z-rC:Ln. reli -. -r-tion
Kan carri:.d o .."'. a11 cn.cr..ell a,-;o cr iCCe;2r :.2-. l'T y-old of 1., ;i-act
was 77V, b.p. S110, l 1.463.- Literature values are 118,# and 1.46U
D. Intermediate Ethers
S3P -Dichlorovinyl ethyl others
Cl2= CIAFc"d C012= CHOC2 0 or 1g2C0C~H12 CC3 CHOCVco
Preparation of this compound in a pure state required the addition
of alcohol t tthe olefin under pressure; however, when the ether was to be
used only for further chlorination, a sufficiently pure product was ob-
tained by direct reaction of CHF2CICl2 with alcoholic potassium lhyroxide
at higher temperatures than usual and atmospheric pressure.
SSeventy-three grams (0.6 noles) CHF- CC12 made as described
previously was added to a solution of 56 gas. (1 mole) potassium hydroxide
in 1O0 nl. 95% ethanol in a pressure vessel and rocked for three hours at
120". Distillation of the washed and dried product gave 67 amo. boiling
from 14o0 to 16*, or 75% of theory.
Two hundred and seventy grams CHF2CEC1 (2 moles) was added with
sti.rriric to a solution of 300 gns. (about 5 moles) technical Grade potas-
sium hydroxide in 800 ml. 95% ethanol over three hours at reflux tcVpera-
ture* No heat was necessary. The reaction was run one hour longer. Dis-
tillation of the washed and dried product cave 38 mas. boiling from $55
to 1 06, l6. gOas. at 120 to 1o0*, and 14 gis. higher-boiling 3ateriale.
This is a 59% yield, although the product contained scBe ethyl chloro-
acetate, as shoan by the following preparation and discussed further in
Material _2_ Chlorine
Product from C0F= 001C 1.ht7 1.203 31.92 o102l5.
Product from CIIF2CIC12 1.4510 ** ,** *,*
CC12= CHOC2H5 ., 1.203(20) 3159 50.31
OC -Dibramo- (3 (3 dichloroethyl e1tyl ether
CHF2CIIG12 .1u'coolII-- CHOC CC0012 CH1Br00C2
Tro hundred and eishty-two grans of CC12 CIIOC211z (2 nolec)
was broinated without solvent at 0-10* until a por-anent red color
appeared. The raberial was washed, dried and distilled, givinS 27 gps.
boiling at 30-35~/c.2ri. and 378 c s*. boiling at 65-70O/c, ila This is
a 72, yield, assuiming tha thc vinyl ether was pure. The lcar-boiling
natcrial wans probably ethyl chloroacetate present a aan iEpurity in the
Virr-I ether used for brcinaition. (See Chapter II for further details)
o ,, 4 -Tetrachloroethyl ethyl ether;
CC012 CiHOG2001 CHCl0 HClOC21
cc 2= 08002"5 cal canc 5
One hundred thirty-seven crams (1 nolo) dichlorovinryl ethyle their
vws dissolved in 200 ml* carbon tetrachloride and lnept at -5 to +$* while
a slow stream of chlorine Tas bubbled in* After tlree hours the wei ht
had increased 63 sgis., 'iiclh was 93% of the stoichimetric amoul:m. After
washing and drying, the carbon tetrachloride was removed. Vac~na frac-
tionation at 14-16 ta. Cave a slight forerun of seven grams, then 167 gis.
boiling at 75-78, for a yiold of C81 theory.1
Ifaterial B.P4 n1 25 d2.
Product ....... 780/1.-16 m.: 1, 730 1.421 11.22
cc13CH coc2115 79/16 . 1.423(18) 41.8
o toe( ,( a(3 -Pentachloroctbyl ethylV ethcrs
CCl= rcai~cirs CC1iclocal^ I CCLa lc~lolIiOC05 C2Cl -2^
0 "1 ,"02" Aco2 CO1,CIClOC2S c co2 C 2C150e2U5
The tetrachloroethyl ethyl ether prepared above was delhydroallo-"
Conated to trichlorovinyl ethyl ether as follows
One hundred eighteen grams (0.55 moles) CC13CHiC102H was addod,
over one-half hour to a solution of 56 min. (1 mole) potassium hydroxido
in 200 rl. of 99 ethanol. About 70 rnl otihanol tras added throughout
the reaction to maintain _Pluidity, the temperature being kept at 5-10*
during the addition and for one hour loner, with continued stirring.
Distillation of the 1wached and dried product cava 75 g.1s .boilirg at
160-170", or 65% of theory.
Material n.P2 d 25
rrcduhct 1 160-170 l.6G 1.323
ccL,= iio j___ 160 1.332 (20*)
The trichlorovinyl ether so obtained was further chlorinated
by dicsolving it in 200 ml, carbon tctrachloride and bubbling a slor
stream of chlor-in throu,:h the solution at -56 to +5 until about 95%
of theoretical increase in weight had been gained., After washing, dry-
ing, awir removal of the solvent, distillation in vacuo gave 81 gip
material boiling- at 63-66*/.-6ma.,. or 81i of tleor-y.
Iatcrial DP. d I I $ Chlorinc
Product... 63-66/5-6 n 1.l076 1.o56 17.U1 71.66
C2Cl0OC21 235 .., 1.507 46687 71.96
+ r.. . . .- II-: : t : + ..... ..m 1 lr
The Preparation of Aliphatic Fluoroethers
In G~ncral, aliphatic others are made by one of two methods;
alllation of an alcohol
ROI + ROH 0i i ROR 4 10
or by exchange of an altkoy roup for halogen ("Wiilliason reaction)
RONa + Rt --- ROR' HlaX
Tli first method has not been employed in fluorine chemistry
because of the scarcity of fuoro-alcohols. The cccond is not much more
helpful for tw-o reasons. Fluorine is itself a halogen, albeit an unusual
one, and nay c.:c-Imnoc with the al:or-id. ion. An apparent emrn~le of this
reaction is the fori--tion of lcatyl or'holbra"o.oacotate froErcz -difluoro-
( -bronootrhyl ethyl other.
M21 F2crC2 0C(OC-H3
In addition, the usefulness of the i-illianson method depends on
the reactivity of the ahlogen atan. Because of the strong electron-
attracting inductive effect of fluorine atons, not only will the electron
density around the halogen, and consequently its chances of being emitted
as halide ion, be decreased, but the same inductive effect will loosen
the bonds of the neighboring hydrogen atoms and increase the tendency to-
ward loss of 2 and olefin formation. Consequently neither of the dialJrl
ether preparations can be successfully extended to fluaroethern as a gen-
bral method, A few e:cceptional preparations by the Williamson method will
be noted later.,
In 194, twTo patents worce Lrantcd on the addition of organic
moloculcs across the double bhn of hca-i-ily Ihalooenatod olcfins contain-
ing at leas tvwo fluorine atoms. In this May luoroa-ideos wre prepared
from ~aines and fluoroethers fro= alc:hols.
CF2= CF2 + MPm2 C:IF2CO!.I2
Cr2S Cr O n0 N y a C, rCio20o
Thei latter rethod has since been i-proved by Parl Vail, Lea and
Lacher, 2~vi:o shoswd that the anhyidouc alcohol, sodium alciohilate, and
a.utocla3 of thl patent cpccificatiorn could be replaced by operation at
atmospheric pressure using merely a curated soIution of potassium by-r
droxidc in theo dceirod alcohol. This eothod is nor in conoral ise,
The ad'dtion: reaction is inboi-ctiu;; in that its cch n;' is the
rovers of usual olcfin behavior, In a typical olefin addition, as vith
ctlyolcrn and TC1, the attac:in:; frae.:cnt is the eloctroplhilic lIrci-oen
ion, the reactivity of the olcfin is cnhanccd Irj rreatcr clectro:n density
around one of the carbon ,atom, and t.ho reaction is acid-catalyzed. The
ne chanism is usually formulated as
flU H H E H ii
C0i = CB, --+ SC-^ C -Ji-> *C--GH -c-- > CIC- Co
1H H H CI H H
'iith fluoro-olefins, the reaction is roared as nuclooplilic,
the negative ion adding first; it is baso-catCal3-ed, and increases in case
with decreasing electron density around the olofinic carbon. The mechan-
inm can be shown an
r F v or* FF i* F F
CF2 = CF2 --- *C--C" ---+ ROC--" -- OC- CH
F F FF F F
With any unsynmetrical olefin, this addition reaction could lead
to two different products, depending on the orientation of the H and OR
groups in the adduct. Trifluoroclioroethylone, CF2= CFC1, for example
could give CHFC1CF20R and CIFgCFC10G. From theoretical considerations,
Park et l 21found the former to be the probable configuration, and
Hanford and Rigby, obtaining CiH2ClCOOH on heating C2I72ClOC2H5 Vith
ailica gel, assumed that additions to olefins of the type CF2 CXO gpve
o, -difluoroethors. The first authors also state that C!HF2CHFClOC2H5
"would permit the easy elimination of HC1, with the formation of
CF2= CFOC2H5." Ue have found that theo(ot--difluoroethers can lose
IhF also in this manner, but that the final product is not likely to be
the predicted vinyl other, and this last conclusion may consequently be
umnTrranted. At any rate, the configuration to be e:.~ctcd in the re-
Eultir. ethers had not been conclusively proven.
RirZb and aznford used a hih-temoorature hydrolysis of
CHi2ClCF20C2II5 With silica gel to determine structure, and Sarts 27,28
oxidized his ethers to the corresporA-inr substituted acetic acids irith
fuming nitric acid. Although these reactions rcre plausible proofs of
structure, neith-er allowed the isolation of the product in any consider-
able yield. W7e have found that such drastic conditions are not necessary
and, moreover, that well-defined products can be easily obtained in good
yield from even the most stable of the ethers merely by treating them with
sulfuric acid attemperatures ranging from 10 to 50'. Two at -fluorine atoms
are removed and the resulting product is an ester of a substituted acetic
CIHP2CF20C2Ol.---& CHF2COOC2H1' + 21F
This reaction constituteb a very convenient proof of etructurc
and was used to determine the configuration of many of the ethers ob.
tained in this research. It is in soe cases a valuable synthetic ap-
proach. The' reaction is further discussed in Chapter III.
Other work in this laboratory7 has shorn that addition of
mines to fluoro-olefine in the presence of tater gave both possible
aides, the addition being predominantly in one direction at low teopera-
turea but increasing in randomness with increasing temperature. Some de-
gree of disorder might then be expected in the ethers. V7e have found,
however, that in all cases under the extremes of temperature and pressure
used, the addition of alcohols takes placo in one direction only and only
one ether can be isolated, the configuration being such that the most
heavily fluorinated carbon of the olefin is attached to the alkoxy croup.
Thus, even in CC3.= CWI, in which the effect of the itro chlorine night
be expected to outweigh that of the sinSle fluorine and make the carbon
of the CC12* group the more electron-deficient, the resulting simple
ether is C10012MC1F0C2U%.
The ethers resulting from the olefin addition method of pre-
paration are listed in the follroing table.
The preparation of fluoroethers by the addition of alcohols to
fluoro-olefins is easily carried out with no special equipment and gives
good yields. Tho main disadvantage of the method is the necessity of iso..
rating the olefin for reaction. Most of the olefins are quite low-boiling
and their isolation and storage is inconveieint and sometimes impossible
with only a Dry Ice-acetone bath, whose effective toaperature is about
FLUROlT]IERS PiEPARMD FEOM OLEFI;S
efin Ethor BP, ('C)
Lit. Fo'id Lit, Found
CF2 CFCI cOFC1CF20CH-3 6/630 70 1.3338b 133o
OF2 CFC1 CHFC1CF20C2H5 82/630 88 1.3479b 1.3421
OF2= CFC1 CiFC1CF20C3Y7 102/6310109 13575b 1.3536
CF C12 CIIC12CF2OC210 120 63/100 1.3919 1.3922
CF2 = C:1 CH2C1CF20CC25 91-93 32/13 0 1.3513
CIuF= CC12 c 0lg COCHgI 3114 123 .. 1.575
aTaken at atnocpheric pressure (uncorrected) unless otherwise
i, Tli maeasurcemnt ,eas made at 20*,
CThis cam.ound was prepared by HI C. trown n on aval Research
d, 2m ,Ml alogen Yield
Lit. Found Cale. Found Calc. Found (fTheory)
rzol3**rr*rr*sar*lerr~r- ---- --- ------- -r ..I .~....~:.... .:,.__.., :_~~. ~~_
-70, as coolant, For e~raple, the use of CF2 V'CF2, boiling point -78*,
is quite icpracticali If the pure olefin is desired, the only practical
preparation available is delalogenation -ith zinc, since delydrohalogenr-
ation, the other customary method of preparation, is usually done in
alcoholic potassium 1l-drod.dc solution and the olefin formed often reacts
1ith the alcohol, giving greatly lorTered yieldsS The nocessar;- inter-
nediates for the dehalogcnation are in sawe cases difficult to obtain o~
synthesoiz, w:eroeas the corresponding intermediates for dehVdrohalogen-
ation are generally more easily available., It seemed logical thon to
utilise a supplementary method of ether preparation first used by Swarts
and later I I'cDee and Bolt8 In this method an appropriate saturated
compound is reacted with excess of alcoholic potassiua lydroxide. Appar-
ently an olefin is first formed which adds alcohol to give the ether; the
addition reaction undeasirable in olefin preparation is thereby put to a
profitable use, This synthesis is valuable whenever the olefin itself is
not the desired production
Since CHFP2CF2C1 is one of the more umreactive of the fluorinated
allky halides, this compound was chosen to test the applicability of the
reaction, and gave Cood yields of CIFF2Cr2OC2Hr- when run under pressure at
about 00"*:. In general, the reaction T-ith allcoxide ion and alcohol could
be carried out as easily as the olefin addition; the only other case in
which an autoclave was necessary was when CHFI20FCl1 was used. The method
tas in fact referred to addition whenever possible since it eliminated
one step and gave equal or better yields. By this neans the f.ei0;
etherc were prepared.
FLUOROTIIERS PREPARED FROI' SATUTRTED COMPOMTDS
Saturated Ether B.P. (CC)
Compound Lit. Found Lit. -4Found
. Cl- c. c/
CH2ClCHgC I... CH2C1CF20C2H5 91-93 90 ... 1.3723
CFaC0C.CC ,1 CliCC2F2QCO2H 120 120 1.3969b 1.3890
CF2CCIIF2..i CHF2F200C2H5 -6 ... 1.3010
CF2c1CI2Br... RCI2 CF20C2II 114-115 114 . 1.3970
CF2jrIIrBr2'2 C'Br22CF20oC4H 67/2 66/30 ... 1.4433
CF2BrCClBr.. "CHG3rCF02 *OC* 53/30 ... 1.150
tat atospheric presc~ e uncorrectedd) unloc otherwise
~his ncauccan_ ras r.a.de at 20', CCorrocted from a 17"valuo,
On analysis, this ccaound tool: 98,0% of the amount of silver
nitrate calculated for an ether containing two halogen atmfns other than
S% Halogen I field
Lit. Found ale Found Cale Found (% Theory)
L. ~ A.^~..~. A & .1
From the foregoing, it can be seen that there are two general
methods of preparation for fuoroethors; addition of alcohols across
fluoroolefins, and reaction of alcoholic potassium lydrorlde or sodium
alcoholate with suit a able fluorinated allyi halide, Sine conditions
for the latter are suspiciously similar to those for the classical Tilliam
son yntheois, harmeer, there may be some question as to whether the re-
action actually proceeds byr the addition of the alcohol to the olefin
formed in situ,- or Thethcr the alcoholate reacts directly with the halide
in a Williamson type reaction, Both courses would lead to the s-.c final
cF2C1CI2C1 RONa i CI1CGFOP I; aCl (";illianson)
CFlCHC2,Ii--. C~~g + ROH-- OW CFIC + Na~t (Olefina-
That such a suspicion may bo valid is shown by the formation of
fluorooMbl-rs whose configurations are contrary to those shown to be normal
for olefin addition, for instance the(3-ftluoroethers. There are several
eammples of this type of product. Starts isolated a small amount of
CHIF20dOC,20I25f ro. aHTgBr, and Benning and Park obtained a poor yield
of GoCfO2OC? 0 from CGFCH2Cl. To these wo would like to add a third case,
one which w1 feel has been incorrectly interpretated ever since Swarft pera
formcd the e;oerinent almost forty years ago,
In 1911, S ars, on fluorinating CBalB3r, obtained a small
quantity of riaterial which boiled at 26' and had the composition C 2 3H
corresponding to CF CiE% r or CF2ErC.l 2 He found the material to be ex-
trcmel3 unreactive, butr on hating in a sealed tube at 1O0* with sodiua
methylate obtained about three cce of product, presumably an ether, boil-
ing at about Ii?% After treating this with fuming nitric acid, Swarte
isolated "a small amount of sodium fluoroacetate." -He therefore claimed
that the confiCuration of the other was CIFC2fgCH3 andconsequently that
the bromo compound was CF2DrCHz2P, We believe Swarts was in error for
the following reasons
1) Difficulty of fluorination increases in the order CXA- (Clil2.,
(HrX It is unlikely that the lone broamine would react before the CBrr
group was completely fluorinated*
2) Fluorination of CCl2BrC I r has given good yields of indubitable
CF3CII2Br (D.P, 26*) but no isoneric CF2BrCIIF has ever been reported,
Fluorination takes place in the steps C01F CHgBr2j CF20lCH01 2r, CF3CI112Dr
all of which have been isolated* Since bwomine is more easily exchanged
for fluorine than chlorine, Swartst original CBr3CI2Br would be even less
likely than CC12ErCl.2Br to givo the supposed CF2IbC2F *
3) The extreme inertness of the fluorinated molecule toward mercuric
oxide and sodium metIylate indicates a CF. group. In contrast to this
behavior is that shaon by the compound Swarto assumed he had, CF2'rC1EF2 ,
17c have found that the analogous Ce7102CI01 gives an 80% yield of
C!2C1CF20C2i.5 u6der v.ry nodcrate conditions with alcoholic potassium
hydroxide.' If Sarts had actually started with 2CF2BrCH2, it should have
reacted readily with sodium metbylate.
4) Denning and Park, -who prepared CF3CHI20C2L5 from CF3CHg2Cl and
sodium metlylate in 19L3, found drastic conditions necessary for this
reaction and obtained only a 25% yield after 55 hours at 130* in an
autoclave. The low order of reactivity encountered in this trifluoro-
etylW halide resembles that described by cnarts and ascribed by him to
the reactive CF2 CHX grouping.
For the above reasons we believe that Swarts actually obtained
eF 3CO03 and, wrorming with very small quantities may have mistaken
sodiun trifluoroacetate for the monofluoroacetateo.
Swrarts reaction, if we are correct, and certainly the other two
exa nples cited, are examples of genuine Iilliamson ether syntheses. Theso
are the only instances of this type recorded; however, they do shor- that
the reaction does occur in some cases. It therefore seemed advisable ;to
study further the relation between the Wlliamansn reaction on one hand
and olefin formation and addition on the other, in compounds where there
night be competition between the two reactions.
The relative tendency toward the two alternative reactions was
first studied in the pair of cornounds C F2CH2C1 and CF2ClCHI It can
readily be seen that these two isemers will give different products on
etherification if the reaction is of the %Wilianson type:
cHF2C201c + OR* CH20CtOIR + Ce"
CFC1Cn0 Oar--- cH3CF2o +* c
but the seae .products if the reaction is olofin formation and addition,
CF2CHC1U3 CB2 22
Results on this set of compounds proved entirely unexpected. Both
were very resistant toward sodium ethoxide or alcoholic potassium hydroxide,
and only partial reaction was obtained even after forty-eight hours at 100'
in the autoclave, with a UI1 ratio of base to halide. CF2CICIf gave no
halogonatod product at all; instead, a 20% yield of ethl7 acetate vra the
only material isolated. It uas identified by the following prclprtie',
Material B.P. 2
Produ ct 70-77 1 369 0909 21,90
c1I3COo2H 177 1.3708 0.901 (20') 22033
The presence of this seemingly unrelated compound can be explained
as follows. delzydrohalogenation of CF2C1CH3gave CO p2i which added
alcohol to give CI3CF20C2I15 This ether, on the basis of modern organic
theory, should be very reactive because of the marked asymeWEtry of charge
on the two left-hand carbon atonno, and should hrdrolyse readily to give
the corresponding ester, ethrl acetato.
,2-Chloolloro diluoroethano was also unreaofivo but gave tho sam
product, o.tll acetate, in small. yiold. Since ethyl acetate ias formed
in both cases, the evidence is very strong that the olefin mechanism was
followed by both compounds, as this mechanism is the 27nly one capable of
explaining the formation of an identical product from the two isoneric
Under a very wide range of temperature and concentration fto
CHr H c was isolated. This would indicate that olaifin formation,
with subsequerrb addition of alcohol, operated preferentially over the
Wiliamson type ether formation, even in the simple halides whore tendency
toward the latter reaction should be at its greatest.
Further attempts to stud4 the Williamson reaction in formation of
fluoroethers were made witt the pair of compounds CHFg2CH2C and CF3CF-2Ir,
These were chozcn to dCmontrat- the relative effects of the C7- .and
CF3- croupsz on a1-71 halido reactivity in the villiancon reaction. A
better choice would have been CF3CII2P and C.T2C213Br oCr CF3CITCl. and
CIIF2I2C1, .cli inating azy com-plicating effects of the change in the pri-
mary halo-en atom, bat unfortunately the natchin' compounds were not
Formation of ethyl acetate frac CHfF2CGH0 has already been nen-
tioned. Careful e. mination of the residue resulting from fractionation
of the otherification products shored an appreciable amount of a nuch
higher boiling component, which was finally identified from the following
Ilaterial Bf .P
Product ....... 116-7 1.h132 1.031 37.0 22,8
c 2CCI(oc2nS )2 157 .*, 1.026 (15) 3841 23.1
Proesnce of this compound and of ethyl acetate showed that dely-
drchlo-::nation liad taken place in both possible directions. Elimination
of 1C1 had civen C:I2- CF2, resultinS cvcntually in etiyl acetate, Thile
cl.inlnction of IF had given CI[F= CHIC1, Addition of alcohol to the latter
olcfin gave a probable intermediate CICH2lCIIFOC212 alcoholysis of which
would result in the acotal obtained.
CF2,CiF2= H2 -V' CCI3CF20C2H---- CTH3COOC2f5 I
CF= CjH1c.-* C12C21CFOC2jIo.i-* CH 2C1ic(noc25)2 EI
PRcaction II was unexpected since the C:IF2- group is supposedly
quite table. Certainly the Ih!drogen on the carbon aton bearing the two
fluor~ano sO aold caMe off mor readily as a proton than that on the car*
boa bearing only one chlorine in a supplementary manner, the chlorine
should come off more readily as halide ion than the fluorine.
Side reactions should be conveniently lacking in the reaction
of CF3CH Br. The CF3- group is a notably stable one, moreover
CF3oCH20C2H had boon obtained in small yield by Benning and Park and
CF3CH20CH3 probably by Snarts. Although the trifluoromethyl Croup
should :shovf a greater inactivating effect on an alryl halide than a di-
fluorometlyl, the change from chlorine to bromine should world: in the
opposite direction, so that CF3CH2Br should have very approximately the
same order of reactivity as CH2CHI2C1 in a Williamson type reaction.
It should have fewer side reactions since olefin formation is possible
in only one direction and the CF3- group is supposedly more stable to-
nard alcoholic potassiua hydroxide than the CIHF2-
Unfortunately the behavior of CF3CH2Br did not conform to these
predictions. No simple ether was ever obtained, under a iide ranke of
experimental conditions including those given in the patent specifi-
cations., This compound was also exceedingly unreactive and was usually
recovered unchanged in good part. The only material isolated in early
runs was a high-boiling liquid, extremely lachrymatory and totally unlike
the expectOd fluoroetler. This material was identified as the orthoestor
of bromoacetic acid from the following data as listed in Table 10,3
This orthooster could only have come from the corresponding
simple ether CH2BrCF20C2f, by replacement ofau-fluorine by alkoal
groups. Consequently we searched for this intermediate other in the
reaction products and in one run succeeded in obtaining a 20% yield of
material with the characteristics as listed in Table JI
_L'atcrial B. d2 _
Product........ 55-60/slight vacuum 1*3970 1.513 30.6
CH2I2rC?20OC2H 114 1.3958 1..86 30.0
This product is undoubtedly the broriodifluorocther,. CH2DrCF2OC2H5.
Formation of these compounds shows even more strongly the ex-
trome reluctance of the fluorinated ayl~J halides to form an ether by
a simple displacement reaction with allodxide ion, and the proponder-
anco of the olefin-addition reaction over the 7illiamlson. Even the
CF3- group is breached before the molecule will relinquish the brcAuido
to any degree as bromide ion. Qualitative tests on the inorganic salt
precipitated during the reaction, however, ohovwd that some brcmine
was present in the ionic state, so evidently some trace of the William-
con reaction must have taken place. Since the latter reaction took
place to no measurable extent, we were unable to determine the effects
of the F3- and CFS2- groups on this reaction.
The third set of compounds which we used to study the relation-
ship between the Williamson and olofin-addition reactions in fluorinated
allyl halideo consisted of a series of five balogenated ethanas con-
taining the CIIFP- configuration. The members of one gvoup of this series
were CHF2C lCl, CIIF2CIICh, and CHF2CC013 This group should show the
progressive offoct of substituting chlorine for hydroon on the carbon
alpha to the CII2-, The second group in this series included CIIF2OCl1,
CilF2C!TC1, and CHF2CF2Cl, and was chosen to show the progressive effect
of substituting fluorine for hydrogen on the carbon adjacent to the
-CGI2 group, when that carbon also holds a chlorine atom. Those choices
cave a series of five cCmpoundc containing; the 'C;[2- group and showing
a gradual chan-e in the architecture of the noloculeo
CEF2CCl3- 011CF2CHC12- ClT'2C12CCl-4 C1F2CIFCl 1 CiiF2CF2C1
I2 34 S
The compounds on either end of this series, ClIF2CCIl and
CIEF2CF2CL, have been shown to give Good yields of the ethers presumably
formed by olofin addition, 91 while the middle compound, CIHE GiC1l
according to analogy with CIIF2CII2B, should undorco a Williamson r*
action. The other two compounds ,night therefore exhibit tendencies in
The reactions of the third and fifth members of this series have
already been discussed. 2-Chloro-l,l-difluoroethaane CIF2C0II2C, did not
form any of the i'iamson ether under the conditions c=plo'ed, but led
only to ethyl acetate and the clloroacetal th bi-directional de3rlro-
halogenation and alcohol addition. l-Chloro-l,l,2,2-tetrafluoroetlhne,
C01i2C2C01, gavIo ood yields of CIFl2CFg2OGCn with no side reactions.
l,l,l-Trichloro-2,2-difluoroctlhne, CHC'2C001 p is difficult to prepare,
and sir6c Gowiland9 had shorm that itt gave good yields of CHC12CF202OCI,
probably ty olefin addition, it was not further studied in.:this research.
It is evident fraC the reactions discussed heretofore that
GICF2CICI2 could give rlse to a great nu=bor of products,, depending on
ihFich of the possible olefins simple others, acetals*, ortho esters, or
enters were formed. The folloTMing charb slihcs the possibilities con-
sidered, leading to as may as nine products in the reaction mixture.
Actually one of the main products, the dichlorovinyl ether, tVas not even
included in this forecast.
C12=2 CHF.P 900002
CHF2CHCk CHlFCCl2OC2fln a?
2== ,,;IC1-*5 CH2010C2o2C2 1 5
In spite of the data on the two previous 'capounds, showing that
a C.F2- croup or evon a C 3 group can split out HP in the presence of
alkaline reagents, it waS believed that the probable course of reaction
would be the lower one in the preceding diagra, the reason being that
a proton should be more easily removed from the CIIF- group, the in-
ductive effect of/ the fluorine being much more intense than that of the
These predictions were incorrect. Although CHF21HC12 reacted
readily with alcoholic potassium hrydroxide, at low or moderate tempora-
turea the main product was found to be the olefin CIIF= CC2, which
could be easi% isolated. Evidently abstraction of 1F proceeds Trith
greater ease than that of IC1. Some corroboration of this fact is. found
in the work of Tarrant and hiuskins, who found that CFGlCH10 save
much better yield-s of CFoF2= CHC than did CF2C:ICg12, when both compounds
were subjected to pyrolysis. Inthis case also, the proton comes from
the carbon having apparently the greater elcctron density, urtlhernore,
as schanG previously, reaction of alcoholic potassitu lykrdro i e with
CF2C1CILC41 cive an excellent yield of the ethber CLC1GF2OC2F H ,while
in tlh case of CflF2C0!C~ the main product is C= CCI.2.
Since a Goodly naount of caF- C012 was obtained in rost casts,
it 1i e-rident tha this olefin does not add to alcohols very readily*
Mhen it does, it -ould seen probable that the simleo ether forrnid would
be C~L2FCC12OC,-e as the two chloriners should sholr a croater incuctrive
effect thai tthe loe fluorirc.. IIcracvcr- no cther of this structure aor'
aWy compound possibly derived from it, Io nonofluoroacetates, .e ,
w a ev~o': found. At ioT to moderate reaction temperatures, CH= CC12
gave an ether ihich boiled at 220* and gaxe analyses approaching those
calculated from tho formula CHC121CHFOC2H. Swarts had previously
claimed the preparabion of this ether from the action of coJiui ethylato
on CIF2CIHC12 and CUCJ.H CUFC1,. and had also found it difficult to purify-.
In spito of poor analytical result, wre believe that adequate proof of
the presence of this compound is furnished by the fbolSo*ng evidence
1) The boiling point was identical with that of CHC12CF2OC2H.
3L:cent in usual circumstances, the substitution of fluDrine for Irdro-
cen has little effect on the boiling point,
2) The presence of CIHF= CC2 in the fraction mixturo has been
3) -fl-ther reaction of the ether wTitfl alcoholic potassim hydroxide
Gave the vinyl cther CC2. C 2IC .. This would indicate that one nole-
cule of i had been abstracted, and that the original structure was as
Previous work had sh=on that more than one oletin was formed
wherever' tiis was possible, and therefore the higheriboiing portions
of the crude reaction product vire thought to be poly-ethoxy co pounds
derived from CiF- 00CC12 and CF2 CHCL Isolation of the dichloroviyl
ether proved that such was not altogether the case; moreover, a small:
amount of rmatcrial was obtairnd which was' indicative of Cl2CC10OC2H5,
probably adulterated with GCCLgS CHOCG2 I as the chlorine perccntaEe rwa
U material B.P. D % CI
Produc&t..,,. 0- 141o.010 1.156 31.-
cH2ccooC2 H5 u14 1,4196 1.159 28.9
Further evidence of the presence of this ester is found in the
hirh-ter~eratu re praration of dichlorovinyl ether directly from
CHF2CI',2, as driven in Chapter The boiling points of ethyl chloro-
acetate and the dichlorovinyl othor differ only by two decces or less,
and consequently the twio cannot be separated by distillation IIHrrveri
the dichlorovinyl ether made in this manner from CIF2CHCL02 as then
*Further discussion of this reaction will be found in Chapter III
further chlorinated to the totraciloroether, with an accompanying large
rise in boiling point. Amy ethyl clloroacetato, wrhn present with the
virnl ether, could remain unaffected and should then be capable of
easy separation by distillation from the hiiher-boiling ether, On vacuum
distillation of the crude tetrac loroether thus prepared, a substantial
forerun was obtained which on rodistillation under atmospheric pressure
boiled at 144* and chawed a refractive index of 1.4210, These figures
are close to the accepted values for ethyl chloroacetate given above
and different from those for C12= CHOC2I5 given in Chapter I. This
compomud can hardly be anything other than the CH2C1COOC2H5 whose pre-
sence was postulated in the preceding paragraph.
Presence of the etlhl chloroacetate indicated that a small per-
centage of the reaction had followed the course originally expected,
Giving successively CI01* CF2, H2CCF20o CICCOOC2, and possibly
CH2CC1(OCo2H)3. The intermediate fractions were therefore intensively
examined, and a small amount of compound whose boiling point corres-
ponded to the simple ether CUI2C1CF2OC2I was obtained. On acid hbdrolysis
this portion gave ethyl chloroacetate, the normal product of such an other
on lwIdrolysis. A very small quantity of high-boiling material vas usually
present, but could not be purified well enough to permit identification
as the orthochloroacotate. The relative amount of CF2= CIC1 derivatives
were albray very inferior to those derived from CIF= CC12.
Evidently, the treatment of CIF2CIC12 with basic reagnts abstracts
HF rather than UCI, leading to the rather unreactive olefin CIF= CC012
the simple ether CIC12C'-OC2H5r, and the dichlorovinyl other CC12= IIOC225.
Smaller -amounL of cCpounds derived from the olefin CF2 CHlo0 are
found. These conclusions agree with those of Swartes
Analysis of the reaction produces in tehe theriication of
CHF2CITC12 was rendored especially difficult both by the large number of
products and the proximity of their boiling Points. As the chart below
shOws, we could expect on distillation under ideal conditions one co-
pound every twenAy degrees or so fr.o 35 to 160o" Variation of reaction
conditions naturally tended to increase the yield of some of these pro-
ducts at the expense of others, but only rarely was any particular frac-
tion completely missing. Added to these difficulties was the lack of
stability of some of the compounds# especially if traces of water were
STMMfIIAIO DU P sOUCTS OF CIIFp2ICl2
Ditillation Temperature (*C)
The last member of the CHF2- series studied was CiIF2CIFC1. This
exocted to show the nost complex set of reactions of all: since three
olefins are possible on dehydrolmlo.onation, namel C CF2 CIiF CF2= GI10
and CFC1= CHF. Although this compound was much nore inert than CG-'C'dCl12
it actually turned out to be considerably simpler in its reactions than
expected. Because of its locr boiling point, satisfactory reaction vas
achieved oily in an autothlarew When reacted for three hours at 120'
under pressure, a 53% yield of CH2ClCF2002H5 was obtained, along with :
anall amount of higher boiling component tentatively identified as the
orthochloroester. The physical constants of the corounds are given i
Material B.P# ^ d 01
Produc-tZb.X* 89-93 1.3681 1.168 2718 #.40
CH2ClCF2OC2H! 90 1.3723 1.176 27.8 ,..
Product I.,1i 95/30-5oCn 4ll77-1.,21 ^..,L .* 19.$
CU2ClC(3C2H4}3 75/13r 1.199 1,4W. 1 U.6
A snall amount of the suspected CI1C1CF2OC2II$ on hydrolysic gave
a 72% yield of ethyl chloroacetate. This behavior, along with the physi-
cal properties tabulated above, proves that the ether was indeed
CH2C1CF2OC2H5 although the related olefin, CiF2 CHC1, is probably the
least likely to be expected of the throe possible olefinsa Although
there is a superficial resemblance botrreen the reaction of CIIF2CHC12 and
CF2CHFC1 in that EF is split out in both cases, the fundamental diffor-
once is that the proton comes from the COP2- group in the latter and does
not in the former,
Strancoly enough, these wiroe the orniy ccnounds isolated on
distillation of the CIIzCIC Cl ethcrificaLion prcdacts. No derivatives
of tho other 01efins were indicated in any way, and the high yield of,
CF2m C2!V1 derivaivvoE rules out any appreciable quantities of other
The overall purpose of stidy of the CI'-. scrics 'as to cot'ab-
lish eame relation botreen the differences in reaction with alcoholic
potassit~ i li-ydoL:do and the c&ango ian structure of the nolccule. Our.
conclusions cannot beas eep as eep as im would lite, because of the in-
consistent behavior of the different rcnbers of the cerios.
First of alla none of the compounds showed any marled tendency
toward the Williamson tyc of ether formation, and it can be safely
stated that all the members react in the dehydrohalooenation-addition
Starting with CIF2HCC1, substitution of either chlorine or
fluorine alpha to the CIF2- Croup increases base of reaction. Although
at first glance such substitution lowers the polarity of the molecule
and would therefore be expected to docrcca the ease with which a polar
molecule could be abstracted, the actual effect is probably to decrease
the electron density around the carbon of the Cf 2 group and thus
facilitate tis loss of lbdrogen as hydrogen ion.
No general rule is apparent as to the direction of delydro-
haloconation. SMw4k 9 stated in 1901, "ie cannot seen to establish
any rule determining the nature of the halogen Which enters into reaction
in the transformation of the polyhalogenatod derivatives under the in-
fluence of Alkalli, and present knowledge does not seen to offer much
iLmrovemc'nt in this rtate of affairs. In coneral, those Io-or4Lds wtich
have IhCroon on only one carbon seae to react nr.c= r~ y, in that chlorine
and b;Cc-;mTc tiZc lost teforc luorino. Thus, all tihe C2.:CIC: and
C72xCIi2 coMpoundc lictcd in Table 7?Cav the cxpoctcd dfluorocthlcrz
Tchn b-ydro2n is present on both carbons, hoiCiCvr, there seems to be no
rirme or reason to the course of reaction:. ith C:i'2C:IC12, AL is ab-
csractd a.nd the (-1 adds to the carbon atcA: holding cne. fluorine, Itlth
CHF2CIrTC1, iF 'is abstracti-d and the O'". adds to the caz-bc= holding t:o
fluori.cs, twith C;r2C0l2C1 hulro is no cloas.-cut reaction. and oi.;lir 1 01
or IPcan bo abLbracLed. Those idiosyncrasies cannot be c plaind Iy
the sirle inductive effects of fluorine and chlorine.
1. Ethers prepared frcom olefins
The follovlng procedure was used as a general method for -
addition of alcohols to fluoroolefine.
A solution of 75$ is. technical crade potassium hydroxide in
400 nml of the desired alcohol was placed in a 2-liter, round-bottomed
flask equipped with stirrer, thoracicter, and gas 'inlet tube and con-
nected to a Dry Ice-acetone tail trap to condense any unreacted olefin.
Approximately two nolcs of olefin were bubbled through the stirred
solution at a temperature of about 10., the addition usually taking
about two hours. Any unreacted olefin in the tail trap was then re-
cycled when desired. The solution was poured into tree volumes of
cold water, the organic layer separated and washed twice with water,
dried with calcium chloride, and distilled through a 30 inch column
packed with class helices.
This general method was followed in synthesis of the follaowin
ethers. Properties of these compounds are listed in Table 6
-Chloro- ,o~ (, -trifluoroothyl motlyl ether
Three hundred and fourteen grams (2,7 noles) of CF2= CFC1 was
reacted with ethanol as described above. Distillation gave 205 gms.
CIHFCCF2OCi3 boiling at 69-71, corresponding to 71% of theory. A
69.5-70.5* fraction was taken for analysis.
S-Chloro-t ,o( .-trifluoroe ;-.ctl etyl cthor
Trro hundred and seventy-seven craras (2.4 moles) of CF2r* CFC1
was reacted with ethanol as described above. Distillation gave 327 gas.
CiIFC1CF20C2He boiling at 87-90*, or 85% of theory. An 88,0-8850 frae-
tion was taken for analysis.
S-Chlloro-l (ot( 1 Z-trifluorootlyl n-oropyl ether
Two hundred and tlirty-five graan (2 moles) CF2u CFCI was
reacted with a Saturated solution of potassium hydroxide in n-prowl
alcohol. Distillation gave 275 's, of CHFCOCF20 7 H boilinC at 1091*
corresponding to 78% of theory.
9 -Dichloro-o(, C -difluoroethyl othyl ether
Reaction of 97 gias (0*73 moles) of CF2' Cd12 with ethanol at
o40* ave 71 gas. of CIHC12CF20C2HI5 boiling at 63L/100 m. on distillation,
or 59% of theory* This preparation was carried out by Ho C. Droam on
Naval Research Project TN8onr03o
S*-Chloro-. oL o -dlifluoroetlyl methyl ether
Fifty-six grams (0,5 mole) of CF2= CHO1 tras reacted rrith rethanoi
at 20-30, Distil.lation gave 12 gas. of H1 C1CF 00C boiling at 32 /30Mn.
TIis corresponds to 11% of theory, This ethor was also prepared by
Vi C. Brown on 118onr503.
2. Ethers prepared froa saturated conro:mds.
The follouling general procedure was used for thcce compounds
Thichl would react normally rith alcoholic potassium Iydroxidu at room
About 60 ms. technical grade potassiuz hly]rodido- -as dissolved
in 200 al. of 95 ethanol and cooled to 10' in a 2-liter, round-bottomed
flank equipped with stirrar, thermometer, and dropping funnel. One-Imlf
mole of saturated compound was then added at such a rate that a temper-
ature of about 10' was maintained. Cooling was applied when needed and
the addition of further quantities of ethanol was sometimes necessary to
maintain sufficient fluidity for efficient stirring. The usual treat-
ment of dilution, washing, drying, and distillation through a packed
column was then followed. Physical properties of the ethers made in
this manner are listed in Table and a brief description of their
_~-Chloro-o( d-d-Uluorooet l othyl other,
One hundred and tlirty-five grams (1 nole) of CFClC0 C1 was
reacted as described above. Dintillation gcave 120 Cas. of CHCE2012002H5
boiling at 88-90', or 33,1 of theory. This yield holdd it compared with
the nreparatior. of CT2Cl0CFT20C 3 7b the addition of methanol to CF2= CHC1,
in which only an 11% yield was obtained.
( 4-Dlichloro-ot -difluoroethyl etl~ ether.
One hundred and seventy grars (1 mole) of CF2ClC1C12 'was reacted
as described above except that the reaction te noraturo vas 20-250i Dis-
tillation gave 8 grais. CI2C12CFg20C25 boling at 120-122%0 corrosponding
to83% of theory. This yield was also Greatly superior to that obtained
by the olefin-addition nethod.
-R5amo-oot-diflnuoroet l1 et!y1 ether.
ITinety ra;m (0.5 mole) of CF2ClCH:IWr was reacted as described
above. 3Disallation gave 7$ Ps. CIIgacOFgO 2h1 boiling at 113-136,
or CO% of thooy.
,0A -fDiio no-, at -difluoroeti1l otcl1 cthere
SOne hit-1-cd and twenty-two a~a= .(O.4.:ole) of CFgBrCiIrg was
reacted ac described above, Distillation gave 8 zs. C1Er2CCFOC215
toiling at 6-66/29-30m., corresponding to 79% of theory.
- Dromo- -chloro-ot, aL-difluoroethyl tliyl ether.
One liundred and forty-five grams (0.6 nole) cr2ErCIiCi3r Vas
reacted as described abovo. DistillAtion gave 1014 C-=. CilrClCl22OC25I
boiling at i 3/30ansa or 66% of theory.
o9 .o, A f -Tetrafluoroethyl ethyl ether,
Fifty-six graas (1 mole) of potassium lhyrozide was dissolved
in I30 li. of 95% ethanol, the solution added to a pressure vessel, and
cooled thoroughly in a Dry Ice-acetono bath, c Z y-eih-t i rars (0; 0 ole)
of CICFF2CF1 vas uddcd, land the vessel called and rocked for three hours
at 120' The custcary treatment gave 70 yield of C-ICIOCqF5i, Loil-
ing at 57-53% with n' former or residue.
. T'hri tnn o iP (TnvA +1m rI m- t
Since all these comounds except CF2CC:I3 vcre reacted zany times
under u='i, conditions, the data can boet be examined by sizca-rizinig the
reaction conditions and appropriate yields in the following tables The
reaction of CF2ClC3 i with sodimu ethylatc took place as described beio~.
A solution of 12 Spa. (0.5 mole) sodium in 80 1l. of absolute
ethanol was placed in a bomb and cooled in a Dry.T ce-acotone bath; Fifty
panp (05; maole) of CF2C12Cz3 as added in liquid form the vessel sealed,
and rocked .48 hours at 105". The bomb was cooled and vented and the
crude product treated in the customary manner. Distillation gave nine
grams of material boiling at 70-77'" with no residues This material was
subsequently identified as ethyl "acetate The yield of, ester was 21% of
thoary*, lNo liquid halogen-containin products awe obtained;
A) Aliphatic fluoroethers have been prepared by the addition of
alcohols to fluoroolefins in the prcsonco of sodium alcoholate o' potas-
suium hydroxide, and by reaction of alcoholic potassium hydroxide with
suitable flutmrinated polyyhalogen coMpotmda.
ETIRIFICAICYNO OF CF C:'ty r
Reaction Conditions Products ($ Conversion)
Hous Te .(c Ee/Haldo/Solvent Reagent Unreacted 112 ErOC20C2H5 CHI2BrC(OC215 )3
3 40 31// KonH 56 .*
S 60 2/2/12 i:aoc2g 27 .
12 75 1/2/2 IIaoc2715 h7 ..-. d
24 75 1/1/3 iTaoc2i3 12 10
24 100 2/2/12 IIaC0H3 12 i -. fot isolated
t8 100 1/1/1A. UnOCI015 Inot isolated 25 19
h8 100 1/1/1.5 IIaOCC2 .4.. 25
8 150 3/2/12 NaCC2Hy 27 .... 10
24 150 2/2/12 :TaOCr2I5 126 5. 11
24t 10 1/2/12 !:accIi3 1 4 ;: ot isolated
run in autoclave.
Absolute cbiharol except runs 5
d Considerable deco-p osition.
and 10 (dibuctyl ether)
ITiERIFICATION: OF CIiF2CTI2n
All reactions run in autoclave
Absolute ethanol ezcopt run (95, ethanol) and run $ (dibutyl ether).
CCarrected for EtoiQ dmexLric excess. %ossibly the intcrmed Iate CIClCiHFOC2H5.
nr.vc-tion Condiiticnc Products (; Conversion)
1 0 01
jp .r C1 4 (
iM : P4 0 Q 3 8 C>
- I .~ IA I ~ ~ I I
i"T.i~s compound was over positiv-ely identifica.
Probably contaniinated with C1pC01COC2gH.
dThis reaction mwas run in an autoclavo.
In absolute ethanol.
ETiE1RIFICATICv I OF CF2C:FC1
Hours Tempa ('C) Ratio Reagent
17 10-20 1/1/6 XOH No reaction..*.. ,*.-*--...- ..--.-
a 60o 2/1/6 INaOC2H C 10% CH2SCFG20o2HS ....,,.,;. .
6 50 1/16 KOH No reaction..........................*
3 120a 2/1/3 KOH Trace olefin or unreacted, 53%
C112C1CF2OC2i5, 5% Cr2CC(lC(ocg5)3*....*,
Tlnhis reaction was run in an autoclave.
2) The configuration of the resulting others is such that the ethor
orygen is attached to the most heavily fluorinated carbon of the olefin.
Thle iconeric ethers having the opposite orientation are not formed,
3) Configuration of the others wero. proven by their reaction with
sulfuric acid to givo esters of substituted acetic acids,
I) The formation of ethers front saturated poPyhalogen compounds
probably takes place by dehydrohalogenation and addition of alcohol
across the double bond so created, rather than by a simple displacement
reaction of halogen by alkosy.
$) Where both the displacement reaction and the dethdrohalogenation-
addition reaction are possible, the latter will be favored*
6) No general rule can be drawn concerning the direction of dehldro-
halogonation in fluorinatod po2yialogen compounds. HF is sometimes ab-
stracted more easily than IHX The CHF2- group does not show particular
reositance tcward the action of alcoholic potassium hydroxide.
7) The fluoroether and fluorobramo compounds reported by SSarts
as CH2FCFFOCH, and CF2BrCH2 are believedto be CGfCH2OH3 and CFPCH2Pr.
8) Substitution of either chlorine or fluorine alpha to a CHP2-
group increases the ease of dehydrohalogenation, but no statement can
be nade concerning the effect of such substitution on the direction of
Reactions of Aliphatic Fluorothers
It has been stated previously that the reaction of CIFC1CFO2Ce2l
vrth sulfuric acid to give a good yield of FOICOCOOC21H tras the factor
which led to the present investigation* The reaction itself has proved
valuable in three ways; as a proof of structure for the o)k co-difluoro-
ethersa as method of preparation of esters of other substituted acetic
acids) and as a means of preparing the chlorofluoroaceto group, Ovhi'c
heretofore had not been easily obtained.
Swartso prepared bromofluoroacety1 bxroide by the oxidation of
CBr2E CFP, and ethyl iodoluoroacetate by the treatment of ethyl bromno~
fluoroacetate with potassium iodide) but was unable to obtain CHIFCiCOOH
derivatives either by oxidation of CC12 CHF or by fluorination of
CHC 2COOC2H5 with antimorC trifluoaide, although fluorination oith potas-
siud fluoride has since been used successfully to obtain CH2FCONH2 and
other nonofluoro compounds. Gryakiewizs-rochimoski et al
synthesized CFC1COOCH3 by chlorination of the extremely toxic methyl
fluoroacetate, but this preparation, in addition to its hazardous nature
gave lcw yields. The hydrolysis of (3-chloro- olad t -trifluoro ethers
io therefore the first satisfactory method of synthesis to be developed
for compounds containing the cllorofluoroaceto group.
AccordingZl, seven new compounds containing this group were pre-
pared. The esters were made by. ydrolysis of the corresponding chloro-
fluoroethryl ethers except for the n-butyl ester, which was obtained by
the transesterification of retiyl chloroiluoroacetate-. The other de-
rivatives were all made by conventional orthods. Properties of these
compounds are listed in Table 18.
The use of the hydrolysis reaction as a proof of structure has
been discussed earlier. Many fluoroethers were prepared and their
structures elucidated by this method so far as is Inown, no compli-
cations were observed in any of the hydro3yses. Throughout the rork,
it has been found that the reaction was free of side reactions, easy
to control .and capable of cood yields. One limit to its applicability
is that tri-substituted acetic acids cannot be prepared frm olefins
since one hydrogen is always present after the addition of an alcohol
to an olefin. The data concerning the ethers and esters obtained from
them by hydrolysis are shown in Table 19.9
Several possible mechanisms for the hydrolysis of oc ,~o( diflorow
ethers can be proposed. Acid ~ydrolsis of diall3l ether s considered
as takin the following course; Initial proton attack on the ether
oxygen, elimination of RsH, and salvation of the resulting carbonium ion,
forming a second alcohol molecule and regenerating the hrydogen ion.
ReO*R'..-- R^-S-- R 'On +
4 b---+ noH + +
The large inductive effect of the halogens would probablyj modify
an exactly similar reaction mecmnis for fr th uoroother hydrolysis.
Because of this effect, the ether oxygen might stay with the fluorinated
end of the ether molecule and the resulting alcohol would lose HF to form
PHYSICAL COJSTAiNS OF CHLOOFLUOROACETIC ACID AIfD TERTIATIVE$
3 .P.. D Chlorine, a
Coancound *C d5 Calcd. Found D
CHC1FCOOH ,......., 162 1.4085 1.532 31.52 31.35 13.14
CIClFcoci ..,,.,., 69,5 1.3992 1.L63 54,15 $5%.17 21.57
CniCCOig2 ......... 72/1 m, 1.4535 1.510 31.79 31.52 19.97
CIIG1cOCIT .. .,,. 66 1.3627 14267 37.92 37.93 16.t0
CIC1FCoGc3 .:., ,,. 116 1.3903 1.323 28.02- 27.93 22,69
CHClFCOOC2U *** 128 1.3927 1.225 25.22 12.91 27.38
CiICFC0OOC3 (n).... 1J7 1.3994 1,170 22.94 22.85 31.98
CIC1FCooC}Li (n) 65 -166 l.067 1.12 i 21.03 20.82 36.90
olccular refraction as calculated from Lorcntb-Lorentz equationau
''olocular weijits Found 161; Calcd. 168.6,
Cionic chlorine: Found 26.85; Calcd. 27.03.
ESTERS PREPARED BI HYDROLYSIS OF o( -DIFLUOROETIiSRS
Yield DB.P (OC)a
Ether Ester Theory) Lit. Found
cl egoc 5 caplcooc2 00 144 NJ$
CI iClCF20o C 3 CICICOOC23 38 3 11
CFCC1CFOOCI~ C:IFC1CQOCU^ 3 flO-11 n6
CIIFC1CF2OC2HTI C7ClCOOCG2II 90 128
CHFClCF20OC3H C-FC1COOC37II(n) 66 ,. 17
CiF2CF2OC2H ~ C.HFc2cOOc2H 60 99 99-100
CHCl2CF2OC2H G C1COOC2 I2 H 82 158 157-5
CHBrClCFg2o2H CIIBrC1COOC2Ho 87 174-76
CH2BrCF2OC2 CI2BrCOO0C2 70 168 160-65
C:IBr2CF2cCI CMBrCOOC2H5 81 192 95-9Q/30
at atmospheric pressure (uncorrecbed) unless othelrise
b !easurcd at 20. :casured at 17*.
dCorrected fra a value taken at 12,5IS
On analysis, a sample of this material took 99.8% of the
theoretical quantity of silver nitrate for titration.
fi 5 Foantd it ^oum a Gale,; Foud C-lca Fo-
- ~ ~ .' --~--1_ U---- 0 a- ^I_ U -. -_~_~_ -
an acid fluoride. The latter could then react with the ethyl alcohol
formed by/ colvation of the carbooniim ion to Cive the ester as the final
CHFC1-C--CpII5--- a CaFC1C-OH + (Cr25)+
C( clcFao ---2,iCC0c o fl
(ca o cengn a*
CHIC1COF + C2IIOHII--- C,1FClCOOC20 I + Fp
Since fluorine forms hydrogen bonds at least as e-si2' as docc
oWgoen, it is entirely possible that the initial attack: takes place on
the a -fluorine rather than on the ether onyen. The following sequence
of reaction might then results
CC1 --C2 l--. IFClC -o-C2H---k CI1 C1= 0 (C215)+
The reactions of these fra rents could then follCto the cane
pattern and tive the same end result as choun in the preceding sclhee.
Both the above mechanism sugccst the splitting, of the other
lnolecule into acidic and alcoholic components and the rccombin.ation of
these elements into a final ester. Since the presence of rlaUy carbon-
itu ions is a notorious cause of side reacticns, it secms unusual that
such a nochanism could lead to yields of ester sometimes higher than
80% of the theoretical. The ethyl carbonium ion generally assumed to
be formed in the reaction between ethanol and sulfuric acid leads to
ethylene or, by reaction with another alcohol molecule, to dicthyl other,
but no dictl-yl cther -;a ever isolated in the hVdrolysio of fluoracthers,
and fuirt ernore the gases evolved ~-r rc complctcly abccrbcd in dilute
sodium lydTromids, vriereas ary cthylcno -..'o'ild pass through this solution.
If the reaction does not proceed by elimination of an allhyl car-
bonirm ions it irould seen a reasonable possibility that the etho~j
linkag is never tbozeoen during the ydrolysis. Some cvidenco indicates
tlia such nay be the case. If the acid fluoride is present as a dis-
crete molecule in the reaction naixure, it should react readily i ith
ary alcohol and if nore than one alcohol is present, the final product
should ie a rixturo of estcrs. Tilen the iV-rolysis was actually carried
out in the presence of a large amount of ietlmnol, however only the
ethiyl ester was formed. The yield ;ras rciuccd, probably because of di-
lution of the acid.
-From the evidence at hand, the follorrinC scheme soems to be a
reasonable mechanism. Proton attack takes place on the oC -fluorine, IF
is lost, and the resulting complex is stabilized by solvation in the
sulfuric acid. On dilution, the couple= reacts Iwith water to form the
'- + Ho O0
Ci1FC1-C-OC2H--- (CnFCl-C1OC211) lSO4 --+ C1F12-C-0C2H_ + HgSO4
CIIFC1-C-002HS--- CHFCl-C-OC2eH 4 H
Sucl a mechanism would account for the absence of by-products
and the lack of ayr methyl foster from hydrolysis in the presence of
methanol, since the etho.V linkage was apparently not broken at any time.
In vlew of the suppocod stability of the -CF2- i*'oup, it is
surprising that the hydrolysis reaction takes place at all, It is
difficult to compare the oa ,Cl --dcifluoroethors with the analogous
( cd -dichloroethers since there are few literature references to the
latter.. Sumnerbell et al found that 1,l,2,2-tetrachlorodioxane was
o.sily hlidrol2zed to o~lic acid, and although ,o( ,(3 ,# -penta-
chloroethyl ethyl ether is stable enough to be distilled at atmospheric
pressure and a temperature of 210* without excessive decomposition, it
i lIydrolysed by concentrated sulfuric acid, The chlorine atoms in
(, -dichiloroothyl ether are notable resistant to hydrolysis but treat-
ment of o4 -monochloroethers with ater or acid usutaly given HC1 and the
The relate reai reractivity of Ct o -substituted fluoroethers toward
hydrolysis corresponds with that predicted from theory. If the initial
proton attack is directed at the o( -fluorlne, this attach: should be
facilitated by increased electron density around the ot -carbon atom,
and conversely any electron-attracting ( -cubstituents should decrease
the reactivity. Such a line of reasorJnin would lead to the following
order of hydroytic reactivity for ot k difluoroethoros
t CHCF2-) C"2AC2-) CIX2CF2-) GC3"2
and for 6-substituentss
H) Br) Cl> F
Experimentally, the ethers showed the following properties.
CH3CF20C2HS was apparently vory easily hydrolyzed; CH2Br-I CH2C1-,
CHBr2- and CHIO1-difluoroothers were of approximately the same stability,
CIhFC1CF20C2H5 Twas less reactive; CHFgCF22002 was fairly difficult to
hydrolye, and C4F9OC4Fp, a perfluoroetlier, could not be hydrolyzed
even w hon refluxed with concentrated sulfuric acid at 110.i. This
series of reactivities corresponds completely with the predicted order.
Some of the increasing inertness, horwer,. may be due to the greatly
reduced solubility of the ether in acid when a large number of halo -
cons are present., A general statement concerning the stability of the
fluoroethers might bet Those substituted in the -position only are
stable,* those substituted in both theO( and -* positions are stable
to a degree determined by the exact number and position of substituent
atoms, and those substituted only in the oa position are quite ro-
activo and unstable. As stated in Chapter II, we were unable to iso-
late the o( -substituted ether, CH3CF20C2HY, in the reaction of CH3CF2CI
with sodium ethoxide although it almost certainly was formed, presumably
because of its ready lydrolysis to ethyl acetate.
An intensive effort was made to study the kinetics of the hy-
drolysis reaction so that the mechanism might be clarified,, Little
success was attained, chiefly because no homogenous reaction Mediun
could be found having cufficiont acidity to promote the reaction at a
measurable speed, but same interesting results were obtained. It was
found that not only culturic acid, but also CH3S03H, CC13COOH, CF COOH,
alone or in incrt solvents such as benzene; in short, any fairly strong
acid compatible Twith the ethers to some extent, could promote hydrolysis.
According to the descriptions in the literature of these com-
pounds by Sararts 27and Benning and Park. 2
2. Formation of polyetho?- compounds
A reaction related to both hydrolysis and etherification is the
formation of polyothoxy compound, sone of which were identified in
Chapter IIf hen the fluoroether was made from an olofin or a reactive
saturated compound, only trace amounts of these polyethoyl compounds
were formed, but with ir.neasing severity of reaction conditions they
were formed in larger amounts. Substantial quantities of the ortho-
bfamoacetate, the chloroacetal, and the orthochloroacetate were obtained
respectively fra CF3C1.4r CiF2CH2C1, and CIF2IHFCl Since H11 C, Brown
has succeeded in obtaining the orthoestors from the simple @-*halo.
ofda -difluoroethors by reacting the latter with alcoholic potassium
hydfoxide under pressure at higher temperatures, formation of the poly-
ethoxy compounds must be considered as a reaction of fluoroethors.
Tho mechanism of this reaction is in doubt* Just as was the
case in simple ether formation, the reaction can proceed either by re-
placement of halogon by an alkoxy group or bt abstraction of HF and
addition of alcohol across the double bond so formed. If the latter
hypothesis is correct, it would seen probable that the intermediate
vinyl ethers would sometimes be encountered. Actually, only one vinyl
ether, CC12= CHOC211, was isolated, possibly being formed as shown
QKOHI EtIOl KO
CIlFgCHC32 C---9 S1 CCIF 22m, C"iCGI2CIFOC2 --- CCol= CIoclip
These contradictory results are somea-Mat allied with those of ITehor
and Fleece, who found that treatment of CC13CIIGOC05 and CIlC CI2IClOC21H
Theeoo ethers are probable intermediates but wcre not isolated.
with zine in alcohol, both of which reactions should give the same type
product, actually gave in the first case the vinyl ether and in the
second, the acetal. Both reactions could be made entirely specific.
CCl3C1HmClOC2UH--- CCl2= CHOC2II5
Tleher and Fleece attested to explain their results tb assuminC
that in the latter case the vinyl ether produced initial was so polar-
ized, because of abstraction of a polar molecule, as to undergo further
addition, while in the former case this polarization vas lacking. It is
difficult to see how this explanation, even if correct, could account
for vinyl ether and acctal for-tation from fluoro-comrounds, since in our
case HF was abstracted from both molecules. The only difference between
the two fluoro-compounds cited lies in the mnmbcr of halogens present on
the ( -carbon of the intermediate ethers. At ary rate, the evidence is
too meager to permit drawing any conclusions as to whether tlh polyethoxy
compounds are formed bct simple replacement of halogen 1y alkoxy or through
a vinyl ether intermediate.
3. Reactions of Fluoroethers with aluminum chloride.
It is Imown that aluminum chloride will replace fluorine atoms
in organic fluorine compounds by chlorine atoms, and it was thought that
in this way fluoroethers having chlorine atoms on the Ot -carbon might
be prepared. ">hen small amounts of CIIFCF20C025 and anhydrous aluminin
chloride wore mi:ed in a preliminary experiment, a vigorous reaction en-
sued and continued until the flask seemed entirely dry, although the
Temerature never reached the boiling point of the ether, A residue of
apparently unchanged aluminum chloride was left. Most of the gas evolverc
Was neither condensed nor dissolved when bubbled through ice water. FurE
their aiall-scale experiments showed the presence of an acidic component
in the evolved gases, and accordingly a larger scale run was made so that
the comoxunds formed might be identified, The reaction products were
allowed to distil out through an ice-cooled condenser into an ice-cooled
receiver connected to a Dry-Ice-acetone trap. About equal quantities of
material collected in the first receiver and in the trap. From the
material in the receiver, three fractions were obtained on distillation.
Identification of two of these is given in Table 20.
Material BAP BI d2 M %01
Fraction 1 ,. 36-38 1,3370 1.401 17,00 .31.30
CiCOlC0"F *.....*, *.g ,.o .. 1612 30.97
Iractioin 2 ... 67-71 1.3970 1.075 21.36 ..
CcrlC ool -..,.. 69.5 1 3992 1*.68 21.09
his compound had nox been prepared proviousy. '
The unidentified fraction had a boiling poinb of about 180, was
considerably heavier than water, and was unreactive toward water and
alcohol. The properties indicate the possibility of sone volatile
halogen compound, possibly CHC12, which boils about .*,
The material in the Dry Ice trap was subjected to low-temerature
distillation. Although vigorous reflux occurred at a temperature of -20. ,
no constant reflux tenpcrature was observed, as it is very difficult to
obtain Cood results without equipment specially designed for low-tompora.
ture operation. A substantial amount of product was collected at a
teu.perature of 12-150 This material Cave a molecular woicht of about
100 in a Victor e-iey-r determination. It is believed that this material
was ethryl chloride, although these figures are not entirely compatible
ioth that cipound, w~lch boils at 12" and has a molecular weight of 6bS
Reasons for this choice will. be seen later*
: evidently, then, reaction of CHIFClP0OC2Hl with aluminum iclorido'
resulted in formation of the acid fluoride, the acid chloride, and vari-
ous volatile, unreactive compounds, probably containing halogen*
'A Eimilar reaction was run on CHclGCF20C2H5. The acid cilorido
was easily identified, as shown below.
L materiall B.P.
Fraction l ,... 107., 1*.500 1.523 26.3
CBCl2COC1,..,, 107-8l f .** 26A.
rB analogy with the previous run of CHFClCF20C21!, the acid
fluoride, whose boiling point is 7f*, should have also been obtained.
A large fraction was collected between 60* and 70o, but there ima no
constant temperature. It was noticed 'tha this material, although vio-
lent in its reaction with water, did not dissolve completely. On react-
ion of the 60-70* fraction with ethanol a good deal of the expected
ethyl dichloroacetate was obtained, but also present was a small amount
of product boiling between 50' and 60', and having a refractive index
of 1.l348. The respective values for chloroform are 61* and 1.425.16
The 50-60* fraction gave a strong carbylamine test wirth alkali and
aniline, which is very indicative of chloroform.
We can therefore state that the reaction of CIC12CF2OC2f5 with
aluminum chloride gave Ci1Cl1COCI, CIC13, and probably CIC12COF, along
with lower boiling inert compounds, probably similar to those formed from
_4h Thermal decomposition of Fluoroethersi
TwVicc during; the investigation it ras noticed that distillation
of a fluoroether, under condition: which had previously led to no un-
toward effects, resulted in exonsive decomposition. Copious acidic
fuaes were generated and the distillate was certainly altered from the
original ether, being much more volatile and reacting violently with
both water and alcohol* Reaction of the distillate with ethanol gave
the characteristic odor associated with esters. 'It seems very lil:el
that in both eases the product resulting fronm the decompositionw ws the
acid fluoride, Production of this low-boiling compound would account
for the greater volatility of the distillate and its reactivity toward
the liydroxyl groups of water and alcohol. Some light on the nature of
an even lower-boiling material forced concomitantir is given by references
in the literature* Thus, CH2ClCHC1OC02H decomposed on standing to give
e l chloride and uncharacterised chlorie-cntnig proucts17 w
eIThy'l chiloride and uncharacterized chlorine-containing products, while
:CH2CIClOO23g dOn pyrolysio gave dichloroacetaldelyde and ethyl chlorido;
By analogy with these trheral reactions, formation of an acyi fluoride
Would also result in formation of an alkyl fluoride
CHC12Ci lOC125l----l CHC12CHO + C2OICl
CHFCICF2HOC7 --- CHC12COF *+ C03y
It is difficult to see why such a decomposition should take place,
of the hundreds of times these or other fluoroethers were distilled, .when
conditions, as far as is known were no different from those maintained
ini ary other routine distillation The reaction must be completely
differoai fromn acid Iydrolsis, since the acid fluorides formed are Very
volatile and, if the hydrolysis of C1flpCF2OC2H proceeded by this nechan-
imn, the ethyl fluoride would leave as a gasj whereupon subsequent di-
lution and reaction of the acid fluoride with water would give as a pio-
duct not the ester but the free acid, vhich is water-soluble BesiBdes$
as mentioned before, the Gases evolved on acid lydrolysis consisted en-
tiroly~of HF and were completely soluble in aqueous almali vhile the
alkyl fluorides would be insoluble.
Reaction of Fluoroethers wTith SbF~
The preparation of fluoroethers by the fluorination of" chloro-
ethers would be a valuable addition to the synthetic methods presented
in Chapter II. Before starting this phase of the investigation$ success
seemed reasonably certain since Booth and Durchfield had demonstrated
that fluoridation of CCO1fl3 took place very readily and without compli-
cations to give excellent yields of di- and trifluoroeethyl methyl ether,
Experiment showed, however, that refluxing of 51CIFCOC20H5 *HYd
CG201CF20C2H5 with antimony trifluoride gave no reaction at all. Uith
chlorinated antimony trifluoride, we 7erc able to isolate only a E.-all
amount of product from the latter ether, posciblyr CH2CICOOC2H5. The re-
luctance of these ethers toward farther l.uorination was ascribed to the
presence of the two of -fluorines,, and accordingly CCl3CIIC1OC2H5 was next
submitted to fluorination. This comound gave off a volatile water-
insoluble gas, but the main product was a material which boiled at about
100*, was solid at room tenpcrature, and was very soluble in water*
This was definitely not a fluoroether, and was identified as chloral
hydrate as shonsm belcw.
eatorial B.*P IT.P. % Cl
Product* t t 4 **t 4*t 0,, 95-98 50-56 62,5
Chloral Hydrate ..., 96 c.52 63.0
*The product gave a strong carbylamine test with aniline and
An e:.'zinauton of the literature shocTed that the tetrachloro-
ether is usually trade from chloral alcoholate and PCl5, and can be
hydrolyzed back to alcohol, HC1, and chloral by heating with acid61
Fluorination of CC1GBrCHBrOCgHS E l also attempted. A lo.-
boiling compound mas isolated which ras at first thought to be a genuine
fluorination product, but the material was later identified as ethyl bro-
arterialal B.P*. d MR
Product *,,.... 3$5-4 1.200 l loo 19,67
cgTIIr h.r 38 1.231 1.431 19.30
The last fluorination attempted was that of C2C15OC2H5. On react-
ion with antimony trifluoride this compound cave two reactive fractions
whose boiling points corresponded with CC13COF and CC13COCL0 Both fractions
gave ethyl trichloroacetate then reacted -rilh ethanol. The usual volatilon
non-reactive gas was also encounterred*
Although all these reactions with antimnon trifluoride were un-
successful as fluorinations, they rwre very important in the theoretical
interpretation of fluoroetler reactions, and served as the chief link in
tying together the three apparently dissimilar reactions of chlorination
with aluminum chloride, fluorination with antimony trifluoride, and tlicral
decomposition. The relations betnmen the three typos of reaction may be
seen ,y a brief review of the data.
1) On fluorination, CC12BrC-3rOCC2H gave C2gIsr, CC13CiClO0C2I5 gaM
CC13CiiO, and 02C010C21 gave CC13COF and CCl0COC1.
2) The fluorination of CCb0C0020C2S duplicated both the aluminum
chloride and thermal decor.0ocition reactions in that acyl halides were
3) Tluorination of CC13CIIC1OC215 to cive chloral resembled the liter-
ature references in the formation of the free aldehde. This similarity
lent stren-th to the assxi-tion that an alkyl halide is formed as one
product, since the references state that such was the case with
CIICHClOC2H5 and CHC12CHlOC2Hi both of which compounds closely rooemble
4) In the fluorination of CC12DrCHBrOC2H5, the ally1l halide C2ItSBr
was actually isolated and identified*
'i'hen put in the form of equations with blank spaces left for the
products vhich were not concretely identified, theso data give the follow-
ing set of reactions
CIFClCFgOCC thermal decomPosiOn
C.I.FlCF2OC II ...5.C- .. .
CGCl12ErCHBrOC2-- + C25gBr
CC13CIC1lOC2H5 SbF3100 +
C0!C0CHClOCH0 thermal decomosition CIC1CHO + C2H50C
CH31 ClOC2H15 decomposition- +--- C+21'I5
It is not hard to see that there exists a strild.n similarity
amon- there three types of reactions. If the original ether has two
a -haloons, the result is a molecule of acyl halide and a molecule of
alkyl halide if it has only one oriZinal o -Ihalogen, the result is one
molecule of aldehyde and one molecule of alklr halide* The general
equations maybe written as
RCCORP. --. RCOX + RX
EKXR o -- RCHO + R'
The alde yde or acyl halide and the alIr1 halide are the primary
products. Lore dccp-seated reactions with aluminmu chloride or antimony
tricbleoride could lead then to halogen interchange which would account
for the presence of both atcl chloride and acyl fluoride, or to the other
halogon-containinf products which seem to occur sometimes
In support of this general relationship between the three varied
types of reaction is the fact that a mechanism for the general reactions
postulated can be found which is applicable to all three types. This
mechanism will be developed with the help of modern theory in the nemx
In a discussion of the thermal decomposition# it must be postu-
lated that So-ae chemical roagent is present, as distillation temperatures
of 100* are not likely to set off the free radical node of deco-nosition
found in pyrolysis. at much higher tcpcratureos, All the fluoroethers
except CIFC1CF0OC2II and CITF2CF2OC21 have been observed to decompose
slowly on standing with the formation of small amounts of 17. We may
assume then that a trace of HF was present in the sample of Ci~C10F2O0CH7
ihich decomposed on distiltion4 We have therefore three different re-
agents-4F, A101 3 and SbFWy all of which lead to the same products. All
these reagents are alike, however, in that they can hold additional
r.native ions to fcrm fairly stable corrllexcz. II has a considerable
tendency .to for-:.- the bifluoridc ion, h the ion A1C14; is responsible
for thU 2manifold Friedel-Crafts reaction, and antiron;- i=as a like tendency
a s i evidenced by the for.=tion. of the ions SbCl6 and SbS All these
reaents .ay therefore Le attracted to the a o-hloen of the ethers, as
is demonstrated blow for HIF
CHF"l0C12 -2----* C:IFClC-Q-C3H1
The C-4 bond is thereby loosened :the process being assisted lby
the effect of the unshared electrons on the ether o:ygei, and, in this
particular case, by.the inductive effect of the isopropyl group also.
The net effect is to loosen the C-F and the O-allyl bonds.
Probably one bond is broken first ar.d the other folloTs suit.
CHFCl-O-C30H0----, C1HFGl --C H + CHFGC 1 0 +* (Cb )
If ions are released, they interact to form the al8kl halide and recener-
ate the catalytic agent, in this case HF. The-net result is
(0oH37) + H2-F---2 C3Hy? + 1'
the formation of one molecule of alkyl halide, one of acyl halide, and
catalyst regeneration. The reaction is an acid-catalysed reaction, where
the term "acid" refers to an electron-acceptor in the sense of the Lewis
acid-base theory. The same line of attack may be followed by aluminum
chloride and antimony trifluoridoe though possibly not in exactly the same
zmnner -1 It should be rec?.le: that ar-ng the rcac nts cvoldLnM the hyI
drolyis of luoro'othzr were srvcral other Ihdrogen acid-,- while boron
trifluoride alsoa gave somr inrdlcation of catalyingS the reaction. The
lact i., good case of catalysis by a ca -und, like ai inira chloride
and antimony trifluoridrl, efhich is an acid only by the Lemis definition,.
Th.e .ain dfc1eronce botw.eciI this nechanis and that of acid
,,arol;Tis is that in the latter case, the ether lir:aco is apparently
not 'bro!en. This fact could be accounted for if the activity of the
positively c:iarged complex were decreased before th allylf carbonium
ion were emitted. In acid hydr-olysis the co. plex may be solvated by
culf'ric acid; in the other reactions, th, direct for-rntion of small
aiouhts of esters ray be due to trace amo'mts of water reacting with the
co olcx. The additional atcm of oxygen in the ester must have come from
sonewhere, and the HF present in the ethers or in not entirely anrydrous
antimony trifluoride attacks Slass to form water as a product, If the
complex ion -rcre to react with either sulfuric acid or water, the charge
would be dispelled, hIdrogen ion formed, and the ester produced on fur-
ther dilution. If there is no such polar compound as sulfuric acid or
water present, the charge on the complex cannot be lowered except by
release of a positively charged carbonium ion.
The h ydrolysis, fluorination, reaction vrith aluminum chloride,
and thermal decomposition m~a be fundle-ntally alil:, however, in that
the initial step in any one may be an cloctrophilic attack on the
o- -haloCon, resulting in loss of nXI and forming a positively charged
complex which -y or may not loso that charge by emission of an all;y
carbonium ion. If the charge is lost, the products are the acyl and
alkyl halides; if not, theebhoq group remains attached and the final
product is an ester.
It remains to be e:?lained why none :of the ethers tested could
be fluorinated, especially the pentachloroothyl ether, although tri-
chloromothyl methyl ether reacts so well. The best explanation of these
divergent results seems to be based on the recently advanced theory of
II-perconjugation, the term used to describe an ionic form of a
methyl fgoup, has become an established part of modern organic theory.
The supplementary concept of "negative hyperconjugation" as used to
describe a resonating, ionic form of a trifluoromothyl group, has been
postulated recently with evidence for the existence of such an effect.2
An exter ion of this effect to the trichlorometlyl group seems adequate
to explain the lack of fluorination.
'I ,prconjugation ftacative 17ypr conjugation
It its Oeil-known that a CCL.O- group 'is more easily fluorinated
than a CIC12- or C12Cl group; liorwever, it should be the most difficulty
as any two chlorine atons decrease the electron density around the third
and lessen its chances of leaving as chloride ion*. A negative hypercon-
jugative effect would account for the relative reactivity of this group
since the chlorines would exist partially in an ionic form and could
therefore be easily exchanged for fluorine.
The most easily luoricated COl3- groups of all are those
attached to a conjugated or allylic system, because the mesororic
effect can then come into play to aid the reaction. Examples of this
clas3 arc bensotrichloride and 3,3,3-trichloropropene.
=V 1 +Q7 3 1 .0
Actually, a CI31- group alpha to an cthor linkage is somaewh-ire
between tlhe ordinatl and the conjugate class Although not a truly
conjugate Eyste, the OWgIl is capable of a very pronounced inducto-
rnoric effect because of the unshared electrons o on athe oggen and the
normal 1iyperconjugative effect of the *3O group.
The cegativc Iyperconjugative effect of the COl3- group is
thereby [gratly enhanced, and it is no wonder that this group is almost
as easily fluorinated as benzotrifluoride,
iith a C2Cl group rather than a CCl3-, such stabilization of
an ionic form is imposoiblo since another carbon has been interposed
between the CC13- group and the ether oxygen, and consequently there is
no unusual drive toward fluorination.
01 01 '
Cr 4 0-
The effect of the alUkoy Troup- is transmitted most casi!y to
the Ot -halogen, nakin: the latter the initial site for reaction and
leading eventually to allyl chloride-acyl chloride fomation by the
nechanio discussed earlier, The p-b1hlorines are not fluorinated and
fission takes place.
1. ydrolysis of Fluoroethers.
The followin- general method of acid hydrolysis was developed
as the most suitable for the r majority of the ethers which were prepared
One mole of ether was placed in a flash fitted with dropping
funnel, stirrer, and thermometer, and cooled to about 0. Two rolcs of
96% sulfuric acid were then added at such a rate that the te-peraturc
did not exceed 10, i'ihen about half the acid had been added, the original
two-phase system usually became homogeneous. The reaction was continued
at the saie te.-erature for about an hour after the addition of acid was
complete. Although HF fumes were emitted, damaeo to glassware was not
severe at the low temperatures employed. The reaction mixture was then
poured on about 200 cn. of crushed ice, mixed, and the organic layer
separated. The crude ester was washed twice with water dried, and dis-
tilled through a column packed with glass helices. PiFyical constants
of the esters are given in Table IS.
One hundred and forty-nine Cra-s (1.0 mole) of CIFClCFOCiI20 3 as
reacted as described above. Distillation through a 30 inch colzn i ave
67 gm, of C:IFCICOOCH3, boiling at 116,0-116.5, or 53% of theory.
CHFClCF20oc2IJ I-2"' cimrccooc2n5
Three hlmdrcd and twenty-two sram. (2,0 noles) of CHFC1CF200C2I
ras reacted as described above. Distillation through a 30 inch column
gave 252 gn. of CHFCOOC2H5 boiling at 128-130%, or 90% of theory. A
sample boiling at 128* was taken for analysis.
CIIFCICF2gOC3Hy1 *2 CIFC1COOC3H7
One hundred and ninety grams (1.1 mole) of CFCIfCF200C3'7 Va
reacted as described above. Distillation through a 30 inch column gave
110 ca. of CGVciCOOC3I7 boiling at 147,9 or 66 of theory.
One hundred and forty-four gram of CI2C1CF20C21 (1,0 mole) was
reacted in the usual manner. Distillation through a 15 inch colman gave
98 ga. of C12C1COOC20H5 boiling at 143-160, or G0% of theory, A sample
boiling at 11 vas taken for analysis.
CHT012CF2002U 5 C SCk, C~1OOC
One hundred and seventy-eight grams (1.0 mole) of Cle CF 0CH
was reacted by the General nothod. D stillation through a six inch
column gave 127 gm. of CHCl0COOC102 bollinr jl 7-158% or 82% of
C2hNAtCF2OC2U 23 .. kCOOCot
Scventy-five grams (0O4 mole) of CHIIBrCF2OCaHI reacted with
sulfuric acid to give 47 gao, of CH2ErCOOC2H5 boiling at 160-165" on
distillation through a short column. This corresponds to 70% of theory.,
CHBr2CF20C021I$ -g2 Cir2COOGC2H
Sixty-two Crams (0.23 mole) of CHBr2CF2OC25U was treated as
described above, Distillation .of the product through a 1$ inch colinn
gave 46 i'n. of CHtrC000C2fl boiling at 9-8/3o30 n,, or 8V14 of theory.
CHl0BrF2o050C- --7 CIClBrCGOOC21
Fifty-one grams (0.23 mole) of CiClIrCF200C0H was reacted as
described above. Distillation of the product through a si: inch column
cave to0 n. of CiClEBrCOOC20 g boilin at 17-1761 or 87, of theory.
CI2CF20C21 an- L- ci2Go200021
IIydrolysis of C2IFgCF200pj at 0-10* as described for the other
others gave poor yields of the difluoroacetalte. The following procedure
In a flasl equipped w.th theercomoter, dropping funnel, and reflux
condenser were placed 29 ca. (0.2 nole) of CTIFCF20C01, 10 gr. of sulfuric
acid, and 5-10 g of powdered class. The mix-ture 'an stirred and heated
alowly until the liquid t c.eeraturo had climbed about twenty degrees above
the original 55-577 The reaction nixJuro was then treated in the usual
mannerr In the presence of the powdered class there was little r no
damage to Elassrare eyen at the higher temperatures used, Distillation
of the product through a 12 inch column cave 12$ g. of C0i22C00CX2f boiling
at. 99-100', or 60;o of theory.
Attempted hydrolysis of perfluoro di-n-butyl ethers
Fifty Lrams (0.1 mole) of (C4F9)20 obtained from the Minnesota
Mining and Ianufacturing Coqpaky was rcflused at 102* liquid temperature
vith 22 gi. (0.25 mole) of sulfuric acid. Aftfr a total of about thirty
hours reflux, the reaction nmizto re ac poured into water, separated,
and the organic layer -n:ashnd and dried. Distillation lhrouOt a short
column gavc 50 gia of ether, or 100% recovery. INo HP fumes wore ever
observed throuzlout the attcnpted reaction, and coalescqnce of the two
layers never tool: place.
Tydrolysis of CHFClCF20C2IIL in the presence of methanol:
One hundred and sixty-one crams of CIC10CF2OC2iI5 (1.0 mole) and
32 gi. (1.0 mole) of methanol were cooled to 10" and treated with acid
accord1nh to the general method, Dictillation of the product through a
30 inch column ave 64 gwa of recovered ether and 50 ga* C1HFC1CC002n5
boiling at 127-130', There was no trace of the nethyl ester, which
boils at n16-117* Conversion to the cthyl ester was 35% of theory.
Since the methanol g.eatly decreased the yield in this exoari-
ient, a second run was made in which the methanol f.a added to the mixture
of CiVClCFaC010U and sulfuric acid only after the disappearanc of the
second .-pase, hnen z eact is. *.as virtually cacrplot. Nlo sign of the rethyl
coter rwa obtained on distillation of the crude product.
Attempted study of the 3inetics of hydrolyaist
: The best method of followingthe course of the reaction was thought
to be the measurement of the Hi geaeratod, and the most promising results
iero obtained in an apparatus whereby measured azo'unts of fluoroether and
culfuric acid were subjected to controlled stirring and heating, the lF
converted to SiF4 by contact with powdered glass, the 5SI4 distilled into
standard sodium hydroxide solution, and the excess base back-titrates
with standard acid. Since the reaction mixture was initially hotero-
geneous, however$ this procedure was not adequate for obtaininC any
valid reaction rate measurements. The following results were secured.
Lthor m HF Generated
CHFClCr1FCH , . .. 134t, 1.34
CHFClCF10OC1Hb . ,, ,* 17, 1.65
CIIFC1CFOC3H74 (iso), . 4 .4 c, 3.75
Those results show a trend which would be predicted by the
inductive effects of the al!:yl radicals methyl, ethyl, and isopropyl.
They cannot be used as actual reaction rate masurcments, htoever, be-
cause of the two-phase difficulty just mentioned. Ho solvent of suf-
ficient acidity to promote the hydrolysis at a measurable rate Tas over
found, although small scale experiments, in which a small amount of
various acids was added to the ethers and the -ixture warmed, showed
that HF was generated slowly with CC13COOH and CF3C0001, both these acids
being miscible with the ethers, and Ci3SO3H, whiicli is no more soluble in
the ethers than sulfuric acid. It night be of interest to note that
the reaction of CH201CF202gOC5 with CF3C00O 7as very slo- until a trace
of water was added, when the reaction gained speed immensely. This
migrt indicate that the true catalytic agent for the hydrolysis, in the
case of hydroCen acids, is actually the lydrogen ion, rather than the
whole molecule. Such behavior should be observed if the hydrolysis is
a true acid-catalyzed reaction.
2. Preparation of Other Dorivatives of Chlorofluoroacetic Acid.
cirFClcooc:i- C-L22H HPO2L- 0'CiFclco H (n)
This ester was obtained by transesterification of the uethyl
star. About 0.$ mole (62 gn.) of CFC1COOCH3 was refluxed for two
hours with one mole (75 g9.) of n-butanol and 3 ml. of 850 phosphoric
acid. After removal of the ethanoll and unreacted butanol, 59 ca. of
butyl chlorofluoroacetate boiling at 155-156' was obtained, T is
corresponds to 70% of theory,
The aide ras obtained in 886; field by the armonolysis of ethyl
chlorofluoroacetate. One nole (141 c-.) of eoter was cooled to O in
a flask equippe1. with an efficient rtirrcr, dropping funnel and ther-
nmcoter. About 200 3l. of concentrated a~nonia solution was added over
a period of one-half hour, After a c~ort time a vigorous reaction re-
sulted and extcrn-zl cooling was applied to keep the tenrerature below
10%. Dhen the addition was complete, the water and alcohol wore removed
under reduced pressure so that the te:pe-rature of the mixture did not
exceed 60'. Vacuum distillation of the residue Gave 90 [o. of amide
boiling at 72-77* at 1 mu. pressure.
Chlorofluoroacetamide (0* nole) was slowly added to a 200 ml4,
3-neck flask containing phosphorus pento:ide (0.25 mole), Distillation
of the product gave 35 ~ai of material boiling at 50-70*. Redistil-
lation Gave 30 gaO, of caorofluoroacetonitrile boiling at 66'" The
crude product was not washed as it reacts rapidly with sodium hydroxide
and to some otent with water. Yield was about 6Z of theory.
One mole (140.5 gin.) of ethyl chlorofluoroacetate was added to
a 10% solution of sodium hydroxide ahile the temperature was main-
taincd below 15*. The reaction mixture was allowed to warm to room
temperature and was stirred thereafter for three hours. The i.xtbure
was concentrated to a pasty mass under reduced pressure; toluene was
then added and the mixture distilled to remove additional water. About
100 ml. of 85% phosphoric acid was added and the mixture heated gently
and stirred. Phosphorus pentoidde (50 eas) was then added and distil-
lation of the cllorofluoroacotic acid begun at atmospheric pressure with
a free lame. The product was rcdictilled through a 24 itch column
packed -with glass helices. A 50% yield of chlorofluoroacetic acid
boiling at 162-6fi- was obtained.
Sulfuric acid was used in one exeriment but the yield of organic
acid was low since considerable charring took place during its distillation.
Chlorofluoroacctic acid (047 nole, $3 on.) was added over a
period of one-half hour to phomphoruzs pentachloridc (0~51 mole, 105 oim)
in a flask maintained at 25-30 and containin- a reflux condenser cooled
by ice water and a stirrer. The reaction was continue :'. at room ternera-
ture for an additional half hour after the addition of the acid, The
product wad distilled twice through a short colTmn pacer1 ii~th glass
helices and 50 gan of chlorofluoroacetyl chloride boiling at 69-70' was
obtained; this corresponds to an 81l yield,
. .The reaction of fluoroethere with alminiz chloride&
(S-C=0oro-$ ,o -trifluorocth:-l ethyl cthort
T;o h un~rcc and tenty-ore srar'as (I44 moles) of CHCICF20C2IIH
was added to 185 gi. (1 h moles) of anlrdrous aluminum chloride -lith
stirring over a period of two hours The reaction products wcrc alloTod
to distil out as formed through an ice-cooled condenoer into an ice-
cooled receiver connected to a Dry Ice-acetone trap* The reaction flask
was th.-i. heated ( cntly until no nore liquid distilled1 About 80 l*4 of
product collected in the iccd receiver Diotillation through a 15 inci
column yielded the following fractions: 15 gmn, 16-23*; 4 ga* 23-26';
21 Ga., 36-30; h gna 38-68"; 26 ca., 69-75; 10 c- residue. TLh 36--38
fraction and the 69-75* fraction were identified recpcctivcr l as
CIFCICOF and CiFClCOCl. (See Table 20). The 16-23* fraction mws not
The crude product in the Dry Ice trap: ;'as subjected to lo;:-
temperature distilltiion, using: a Dr- Icc-acctoie cooled co;Cndenser .a
fraction't-ig: colim., with a Dry Icc-acctone cooled head. To sa-tis-
factory seisaration was obtained. The original reflu.: ec:erat-,uc rac
about 019 to -20, which vas not as low as should be c.:ccsed cr
et-!7l fluoride, D.P. -320t A cut as taken at 12-315 as rcprcenting
cthyl chloride, B.Pe 12. although a nolcculua weight doterjrminaticn
gave hirhi result's. (See section A-3 Lof tius carter)
-DichloroC.o(,o -difluoroethyl et,-l ether:
r'ne role (179 in.) of C12Cl0CFOC .HI was added to 0.66 mole
(C0 Dg.) of arihydrous cauin'n chloride in an apparatus essentially
tsitlar too hat described in the nroceding reaction, Tho crude nro-
duct rm-s fractionated and gave the follorring reru-lts: 15 -1. 6.-70o,
5 Ca. 70-1050, s5 cg. 106-100, niTe 105-1030 fraction Twa identified
as C ,ClCC001, as show in Table 21. Since the 61-70 fraction was not
completely soluble in Tater, it was obvious that this was not entirely
the acid fluoride, C!~01 , B.P, 71t % Conscqucnt1y, thia material
vwas added to ethanol, diluted 'w ith water and the organic layer Tashed
and driel. Distillation of the all naount of reaction product C-ve
about two ar7is bDiling at 60-70*, and oi.ht grars boiling at 156-1508
The lower-boilinc fraction iad an odor rese.bling chloroform, gave a
strong carbylamine test with sodium Ihdroxide and aniline, and showed
a refractive index of 1.34t8,. The boiling point and refractive index
for chloroform are given respectively as 61' and 1,IA2$;
The 356-158* reaction product was easily identified as ethyl
dichloroacetate. Refractive index and density were respectively 1*4358
and 1.277, while literature values are given as boiling point I38~,
refractive index 1.t361, and density 1.282, This is good evidence that
most of the original 60-70* fraction consisted of CHGlCO2OF
4. Tho thermal decomposition of fluoroethers.
A sample of CHl01CF2003H7 (iso) which had been prepared about
a week previously was redistilled from an ordinary distilling flask.
Large amounts of acidic fumes were generated whose odor resembled that
of acetyl chloride. About 75% of the charge was lost, but 20 ml. of
material was obtained at a vapor temperature of approxzimately 50o the
boiling point of the original ether being 100** The distilled material
persisted in evaporating rapidly at room temperature, and the snall
amount left was finally added to alcohol, producing a violent reaction,
and having an ester-like odor when diluted. Because of the small amounts
of material involved, the products could not be identified. This ob-
servation is in accord with that of Brouni who observed the formation of
acid halides during the distillation of some of his fluoroethers,
J. The reaction of fluoroethers with antimony trifluoride.
.-Chloro oL(1o-difluoroethyl ethyl ethers
Thirty-six crams (0.25 nole) CH1L1CFg)C gIwas nixed with 15 gin.
(0.08 mole) of antimony trifluoride and refluxed for three hours. The
liquid teloerature showed no change over this time. The use of 15 ECm.
of antinony trifluoride previously chlorinated at atmospheric pressure
for 15 minutes resulted in a violent reaction, but the liquid temperature,
which should fall with replacement of chlorine by fluorine, rose slowly
from 608 to 78" and HF was emitted. iUashing, drying, and distillation
of t!h crude product gave on2y four craas of material boiling at l43-l16"0;
The diotilate had an ester-like odor, The boiling point of ethyl chloro-
acetate is lh1h4
O, -Dibrono- -dichloroethyl etihyl ether:
Three hundred and Gcventy-five grams (1.25 moles) of CCklrCHIrOC23l
was added to 328 gm. (10% excess) of antimony trifluoride in the large iron
vessel customarily used for fluorination of CiC12CHC12 and CC13CIC12, stirred
and heated, A violent reaction started at about 100-110 and large quantities
of gas were generated. Because of leaks around the stirrer seal, only 20 ga,
of product Tia obtained, being distilled out through a water-cooled reflu:
condenser and collected under ice water. This material, after washing and
drying, was distilled to cive 20 in.. boiling at 35-45*i It was identified
as ethyl bromide, as shonn in Table 23.
ao,, ( -Tetrachloroethyl cthyl other:
Sixty-three grams of CCl3CHClOC2H5 (0o3 mole) and 72 on. (OQ4 molo)
of antimony trifluoride were heated at 60-70' in a flask connected with a
reflux condenser, an ice-cooled trap, and a Dry Ico-acetone tail trap.
Affor one hour, the reflux condenser was replaced by a downward condenser
and distillation started. The crude product was redistilled, giving 16 gm.
of material boiling at 93-90., solidifying at room temperature, and freely
soluble in water. This material was identified as chloral hydrate.
(See Table 22)
0(1*t 9 tI-Pentachloroeothl ethy1l ethert
Fifty-six grams (0.23 mole) of C2C15C2caH was heated in an oil
bath at 100' with 79 g*m (0.46 mole) of antimony trifluoride, in a flask
connected to an air-cooled reflux condenser, an ice-cooled trap, and a
Dry Ice trap. The temperature was gradually raised to 150* over one and
one-half hours while about 15 an, of material distilled over into the
first trap. Redistillation gave 10 gn. boiling at 72-80* and six grams
boiling from 90-12w5. Both fractions were very reactive toward water or
alcohol. The two nore combined and reacted with excess ethanol, the pro-
duct separated by dilution, washed, and dried. On distillation eiiht
grams of ethyl trichloroacetatc, boiling at 164-168%, was obtained. Fur-
ther distillation of the residue from the fluorination gave 12 cin. of
material which on redistillation also boiled at 165-168* and was also
identified as ethyl trichloroacetate, as shown in Table 23. The boiling
point of CC13COC1 is listed as 11U .16 and the boiling point of the
corresponding acid fluoride should be about thirty degrees lower, Pre-
sunmbl both the chloride and the fluoride wre formed in thit case,
of esters of the corresponding substituted acetic acids.
2. In the formation of sim.le o ~ct -difluroethers from olefins or
saturated compounds, secondary reactions nay lead to polyethoay com-
pounds such as acetals and orthoesters.
3, Although ot.#d -difluoroothers react vigorously with anIydrous aluminum
chloride,. simple replacement of fluorine by chlorine is not possible with-
out fission of the ethers Among the reaction products to be e ected are
acid fluorides, acid chlorides, al1l 1ai.des, and various other halogen
l4 Fluoroethers cannot be prepared by fluorin.tion of the corresponding
halocthero ocept in the case of a haloer-ated rethyl rrou, alpha to the
ether ozy.en. A reason for this behaviour has been advanced.
5I Th~cmal stability and resistance toward acid hydrolysis of the
o ,oat-dif2uoroethers increases in the order
EeCI 3C Y:C' 2 (CXg2C32r (Gx
6 Increase of stabilit- due to ,-substituents increases in the order
I ( (BrC (F1'
7. The perfluaroether CYF9OCXF9 could not be bhdrolyzed&
8. Eight new co-pounds containing the chlorofluoroaceto group have been
prepared and characterized.
9, A nc -,anin-i lia-s bccrn -rcIorc4l for ,caction ofC fluorocthcrvz with
cc1Turlc acid, almcrinim c1h1oride, ntn.zony trii2uoride, ard for their
1. Aliphatic fluoroethers have been prepared by the addition of alcono-
to fluoroolefins in the presence of sodium alcoholate or potassium ly-*
dro;cde, and by reaction of alcoholic potassiuln lhdroid.de with suitable
fluorirntcd polyhalogen compouds.
2. The configuration of the resulting ethers is such that the ether
o-xgen is attached to the most heavily fluorinated carbon of the olefin.
The icsomric ethers having the opposite orientation are not formed.
3. Configuration of the ethers was proved by their reaction with sl-
furic acid to give esters of substituted acetic acids.
4. The formation of ethers from saturated polyhalogen compounds probably
takes place by dohydrolhalogenation and addition of alcohol across the
double bond so created, rather than by a simple displacement reaction of
halogen by alkco:ide ion.
$5 No general rule can be drawn concerning the direction of dehydro-
halogenation in fluorinated polyhalogen compounds. HF is sometimes ab-
stracted more easily than H01 or HBr. The CF3 and WF2- groups, in
reaction with alcoholic potassium hydroyxde, do not chart the relative
inertness usually e:.ibited by these groups.
6, Substitution of chlorine or fluorine alpha to a CIH2p- group increases
the ease of delydrohalogenation, but the effect of such substitution on
the direction of de bydrohalogenaLtion cannot; in goncral, be predicted.
7. The fluoroothor and fluorobromo compounds reported by Snarts as
C1FCF22C003 and CFgBrCF? are believedto be CF3CI20C and CF3CH2br on
the basis of data observed in this research.
8. In the fonr.ation of astle of, ol-di fluoroothers fro olcfins or
saturated copoaunds, sccondkay reactions ray lead to polyethosy con-
pounds such as acetals and orthoosters.
9. o( -Difluorocthlers can be hydrolyzcd by acids to 0ive good yields
of the corrpcson'ding csters of cubstitutcd acetic aclds,
10. Thcr; l stability and resistance tcr'ard acid hydrolysis of the
c(,Co-difluorolth:rz incrcac-s in the order
C13CF2- KC`72- CF cr2" (-3CF 2-
Incrnc.e of elability due to (S-estfstitutcnts increases in the order
U Br C1 F
11i Althoudhi < io( -difluoroethcrs react vigorously -Tith anhydrous
aluminum chloridep sinlo replacement of fluorine by chlorine is not
posciblo r.ithout fission of the ether. Anon the reaction products to
be expected are acid fluoridcs, acid chloride, a.ly haliodes and vari-
ous other halogen co-pounds.
12. Fluoroethers iave hot been prepared by the fluorination of the
corresor.dinc haloethers tith SF. except in the case of a halogenated
nethyl group alpha to the ether o-ygen. The recently advanced concept
of negative hypcrconjugation has been e:rpanded to account for this
13. Eight new compounds containing; the chlorofluoroaceto group have
been prepared and characterized.
14. A caomon nochaniml has been proposed for reaction of fluoroethers
with sulfuric acid, aluminum chloride, antimony trifluoridei and for their
This chc nicz postulates an olectrophiilic ,.tact-: at th:c allon-fluorins-
,and a cubsequont carbomi=zz ion typ reaction.
1 BEeilctein, 1Handbook1 of Organic Chemiatry, ith Ed., Berlin:
Julius Cprincor, 1918.
2. DEnnring and Park, Ui Si Patent 2,336,921;, 1913.
3i Deyerstedt and Jc lvain, J. An. Chea eoc., 2, 1273 (1937)
4i Dooth and Burchfield, J* An. Chem. Soc., 0, 2070 (1935)
$5 Brorm, Addition of Organic Yoleculec to Fluorooleftin, unpublished
Ph.D. dissertation, University of Florida, 1950.
64, rch, 2cr., 4l, !45 (1878)
74 Fried, The Preparation of Certain Alpha-IIalogenated Acetamides,
unpublished Master' thesis, University of Florida, 1949i
8; Geuther and Laatsch, Aann., 218 36 (1883)
9. GOcland, rritish Pat. '23,49j- 1940,
10; Grysnicewics-Trochimowski, Sporzynlki, and inrcj Reci tray. clhim.,
s66, )`4 (1907)
11I Hamenett, Physical Organic Chemistry, lot Ed., new Yorkl ecGraw-Hill
Eook Co., 19i0.
12 4 Ianford and r Piby, ,U S. Partcn 2,1402,274j 1946.
134 le^nne and Trott, J4 An., Ch~ci. Soc, 0, 1820 (1947)
14 1 IoffiEa, J4 Ore. Chcm., I 10 (199)
15. Huslkir, Frep-ration and P.eactions of Certain Fluorochloroettlylenes,
unpublished Itasters thesis, Universit- of Florida, 199.
16, Lange, lHandbook of Chemistry, 6th Ed., Sanduslyi IIandboo!z Publishers,
174 Lieben, Ann., I, 121 (1859)
18. TicEee and Bolt, Ind. En.. Chem., 9, 19 12 (1947)
19. Ihcher and Fleece, Jo Aum Chem. Soc., 48, 2416 (1926)
20. Organic Reactions, Vol II, ITore York; John Wiley and Sons, 19!J.
21. Park, Vail, Lea, and Lacher; J. Am. Chem, Soc., 0, 1550 (1948)
22. Paterno and 1Eazzara, Ber. 6, 1202 (1873)
23. Remick, Electronic Interpretations of Organic Chemistry, let Ed.,
New York: John Wiley and Sons, 1943.
24. Rieby and Schroeder, U. S. Patent 2,409,315; 1946.
25. Roberts, Webb, and c61cC1ll; J. Am. Chem. Soc., g2, 1408 (1950)
26. Sum-erbcll, Uthoefcr, and Lappin; J. Am. Chem. Soc., 1352 (1947)
27. Starts, Bull acad. roy. Belg., 1901, 383.
28., '....., Ibid., 1911, 563.
29. M!ie. Cour. acad. roy. Zolc., 1901, 61.
30. Tarrant, personal communication to the author.
31. Young and Tarrant, J. Am. Chem. Soc., 72 1860 (1950)