CA2230606A1 - Process for preparing peroxides - Google Patents

Abstract

A process for the faster manufacturing of hydrocarbon, fluorocarbon and chlorocarbon acyl peroxides is disclosed wherein a hydroxide, a peroxide and an acyl halide are reacted under continuous vigorous agitation conditions so as to bring the reaction to substantial completion in less than one minute.

Description

Trl~,F
PROCESS FOR PREPARING PERO~Y~DES
1 ) OF Tl~F Il~VFl~TION
The present invention relates to a near ;. ~ .us process for the 5 1 1 1~.... r~ . c of hydrocarbon, _uorocarbon and chlOl~ca.~.l acyl ~Ai~S. This process uses simple e~ and is particularly safe for the th~rrn~lly least stable pero~ides. Acyl~c.vAidesareusefulas ;1~ forvinylpoly~ andin organic ~ylllllesis.
l~l~N~CAI, RAc~Gl~ouND OF T~F, Il~ lYl~ON
Per~uoroacyl ~luAidcs are cu.. ll-~ollly made by stirnng ~uCu~C hy~LOAidC
and hy~ ,l peroAide (or metal peroAides such as Na202) with an acyl rhlor~rl~
or flnc~ritle dissolved in an organic solvent. Reaction t~nes typicauy range from about 1 minute to about 60 Il~ S
Per_uoroacyl ~ ,Aides are ~n~-~lly thought of as too hy~ ly~ally nnct~hle to be useful as initi~t~ r.~ in the ~se~lce of water. Thus, as the il~t~ y of stirring or emlll~ifir~ti~ n is ;. .~ ed in the ;~yl~ is of a perfluoroacyl peroAide, a point may be reached a~ which peroAide yields dc~ ,asc as a result of c-. .1 l ll~l ;. .g hydrolysis ~ ns~ Jlllc5eAposure~ in p~rti~ r, is known to have an ~c~ g effect on the hydrolysis of organic cu~ oullds.
US 4,075,236 passes acid ~llnri-l~, hydlo~c~ idc, and a~uc~us alkali metal hydro ~ide througlh a series of two l l l~ lly stirred ~ .. . vessels to produce ~.UA~ t~.~ o~ the fonnula Ry~[(C=O)OO]nRI~ S~u.;lul~,s R and R' are left quite general in the clairns but it should be noted that all specific ~ ~ A- ~ ~p]l' S
contain only carbon and hy~ . ~Ith--llgh it appears that this patent daims 25 ~C~ AI~ ~1 reaction rates, the ~mrl.os are written so that eAact reaction times are not rrl~nti- n~ " however, that the 10 gaUon reactor size .l;c~ ed in c...~ with imr~ or design, at column 4 line 20, applies to both .~aclu~ thenthe reaction times in the C~ Amrles would be about ~ lr UlUI~S)~
DD 128663 passes acid hAli-lçs, hy~Lu~.uAide, and &~lue~3us aUcali through 30 a series of ~ ol~ or l!o~ a single cascade reactor with intense mi~ing. This produces ~oA~ t~,.s of the formula R(C=O)OOR', where the R and R' groups are again e~mrl i*~-d by compounds CUI I I A ;~ Ig just carbon and hvdrogen. Short reaction times of 0.5 to 15 mil~ul.S are achi~v~d by using elevated reaction Le~ e~ s (40 to 95~C) in conjunction with a series of vesseIs to control the 35 heat of reaction.
In U.S. 5,021,516, in which HFPO oligomer peroAides were made from HPPO ~ligorn~r acid finori-l~s, c~ubul,~lcs~ and 30% hydrogen pero~ide, l~,a ~io n times were 10 . . .;~ s to 7 hours.

2 PCTrUS96/13976 SIIMM~RY OF TF~F Tlw~ oN
The present invention relates to a process for making ~e, ,~ides, pero~y-,a.7L,Vlldt~o7 ~ u~_ot~o a~ld ~u~yd~ids by o~ Lill~; the le~ lx, to Vi~,Ul~JUsagitation.
S One embodi~ment is a continuous process for ~luduc~lg perfiuoroaeyl ùAides c~ ;xi . ~g the steps of led~ lg an aeyl halide with an ~ cu~
hyd~ide mi~ed with a pero~ide sP1Pct~p-l from t'ne g~oup c..... ~ of hy~Lu~,e,pero~ide and alkyl hy~llu~u~ ide wl.~.~. the i.ll~lU~ C consists of c...~.1.;.~;..~
specifie reactor cv~ x with specific mPthoclx of vigorous ~ so as to 10 ~ a good yield of pero~ide product i~n less than one minute, ~lcf~ly lessthan thirty se~.. lc An a~ eo11c metal pero~ide could also be used as a l. i&_~lL
with acyl halide in the process.
In a ~lcfel~d emboflim~nt of this e~ ~tinnouc process, le~ with continuous ~git~ti~n sta~ts in an slgit~tic-n (reaction) vessel and c~,..1 ;....~ s, with 15 ~git~tion, in transfer lines. Dc~l.,aO...g the ~cc~-m1-1~ti- n of pero~ide in the ;ul~ vessel reduces the threat of e~plosion and the ~c........ .1~ .n of to~ic ~ludu ;IS. Recent 1. ,~ s such as the methyl iso~y&klte disaster in Bhopal, Indid have sn~-ste~ that it is de~ .~b1~ to run 1la~ 7~1uuo rez~rtinn~ partially in the lc~;Li~ vessel and partiaUy in the LI~lOf~,r lines so as to ~le~ e the ~crnm~ tinn ofhA>_.clo.. o~O.Jl.;,1~ cs Insuchco.. 1;.. ous~luc~cosesthe.1;.. .~.~.5;0n.~ oft~ ,i,f~,~
lines are i ll~olL~ and should be set up so as to cn..1;~ c~ 1y mi~c the ~ ~";~ and organic phases, as .1~ sc - ;1 ~ell 'oelow. In a ~l~,f~ ,d ~-lbo.l;.. - -.1 at least 10% of pero~ide finrmsltinn occurs in the ll~loLL line.
The process can also be carried on in a similar ll~lllel with a batch proccss 25 by ~,~b3ecLillg the batch process to VigOlUUS :ig;1;~ The batch process achieves si~ y shorter ~e~Lioll times in cn..~ ;co,~ to the prior art. The batch process is particularly useful for the l,.~ of small lab-scale 4~ .s The batch process cn. ~ s cu. .1 ~ an acyl halide with either a metal pc~u~i~ or with an &lueuus hydro~ide mi~ed with a ~VAid~ selected from the 30 gr~p C~ g of 1IYdLU~C.1 pero~ide and an an alkyl hy~ u~u~idc while c. ntimlon~ly ~;1i.1;..~ the l~c1;~ , using vigorous s~it:ltinn means, until ~bol~7~ 1 c~....pl~ L;nn of the reaction. The reaction is Oul,O~ ially coll~l~,le in less than one rninute.
Jet, static, and ultrasonic agitators are ~fell~,d for the co~tinuous process 35 and ultrasonic a~,it~tolO are plcfc~lGd for the batch process. Reaction times for con~pleti< n of the processes are ~u cf~,,dbly about 0.0l secolldO to about 30 s~c~

~;

~Rll~ ' n~.~CR~PllON OF T~, nR~VVI1~G
Figure 1 is a s~ des~ ,Lion of a flow Iti~Lvn system for the cnntim1on~ n1tr~onic embodiment of the process.
Figure 2 is a graph of the ~IA~ b~ n percent yield HoFPOdP and 5 tubing leng~, in inches.
Figure 3 is a s~ ;c des~ L,on of a flow l~,a.;Livll system for the c~ v..~ jet/1~ o..ic embodiment of the process.
n~TATl~n n~.~CR~PIlON OFT~ nR~w~s In Figure l, syringe pump (3) is loaded with r luev~-c ~rv~d~ and syringe 10 pump (2) is loaded with a~lucous hydro~ide. Syringe purnp (l) is loaded with acyl halide, pl~r~,.. hly dissolved in an organic solvent such as Freon E2.. The streams from the three pumps (l), (2) and (3) are v~l ;o.~11y chilled by ;.. ~ . in wet ice and joined into a sinBe stream at fitting (4). T- ---~ ly after e~iting fitting (4) the liquid stream is run into reactor cavity (5) which can also be chilled with ice.
15 Ull ., ~~o..i~ hom (6) provides intense mi~cing in the reactor cavity. The ~actor is run for several . . . ;. .~ s to flush out the lines. Waste flows into waste c....~,.;.... (7) The reactor strearn is then diverted to sample poly cn..l,.;... r (8) where effluent is coll~cte~l for product ~ ySiS.
n~TA~ n nF~cRTpI~IoN OF TR~ n~lV~TION
The l~,aCIivn can be carried out either in a batch process or cn.-l ;.. ~,.. ~, process.
In a batch process the d ~ ' - and ;,. .. l,. ;~i . .~ result is the 1 ~lu~ in ;Lvn time. The le~ in the batch reactor are c-...ln- t~ cl .s;..~ v u ~1y andcnmim1ously aghntrA The reaction occurs in less than one ...;....~ f~,l~ly 0.0l to about 30 seco~ The ~,a_liOl~ as ~iLi~d, for e~1e, in r~ e 23, results in an . .. ~r ~l~C~t~ flly stable a lueous fli~ion of ~.v~~. Such a ~li~inn, if ,,,VlV~ y used as an ;. ,;I ;-1.. for poly.. ;~_1;.. ~, can result in an "C ~ s fli~prr~i~ n of ~olyll~. which may be useful in coating apr1irati<~n~
As mrntinn.ocl above, in the contin-7Ons process, the .c~r1;~ ; are contr ~tçd 30 in the reactor but the process may shift a portion of the l~a.;Loll from the reactor to the ~r~L lines where the ~,&~;Livn iS ~b ~ lly c.. l~L t~,fl The ~
ofthe Llall~ lines are critical to ~ulOIIIvtG col~l;....-..~e of the rGrftif)n Reaction in the tranfer lines will co.~l ;....e only so long as the organic and aqueous phases remain ;..1;... ~If-ly mi~ed This can only occur where the 1~1V~.11e;1~ of the fluid 35 through the line allows enough turbulance to counter the natural lend~ -y of the organic and a l~,eoui, phases to 5~ala~t;. In F~z~mrle l l(B) fle s . ;l~ed below, for plfe~ a stream of KOH/H2O2 is c- .. .1 ~ tl with a stream of HFPO dimer acid fl11~ rifl.q in a Hoke~D T fitting that serves as a jet mi~er. The r~af tif n is catried -from the jet mi~er tbrough st~inless steel tubing with an internal ~ ...f ~ r of about 0.085 inch. Table 1 and Figure 2 show how the yield of HPFO di~mer ~ dG
varies with the length of the tubing. In this e-~r~mrle, 38 inches is the c~
length. The ~ .I s must stay in the line or reaction vessel long enough to 5 ~y~luacll c( m~ tion but not so long that the cc~ vces~. ~e~ yield.
See, for e~mrle, the ~1iccllccinn in U.S. Patent 5,399,643, colurnn 1, lines 25-28, which is i~lcvl~vlal~d herein by r~r~ e In an embodiment, such as dc..~,li~d by F.~n ple 11(13), a few y~-.L~,I~.
can be used to ~l.omnnctr~t~ the 1~ of the i~ lv.,~ L rl~im~l herein. The 10 1/8" Hoke~ fitting is about the size of a man's thumb. The line lu~illg from the eactor that lC~ S the ~5 ~ and in which the reaction is C~~
several lengths of ~ l l; Within tbis very smaU l~ l ;.... volume (about 3A ml.)enough pero~ide is g~,.lG~ ~I to run a co.nlll~ l scale fluolu~olylll~,r plant on a c-..--...~..;ial basis. It is clear that if the 1/8" Hoke~l9 fitting could be scaled up to a 15 volume of one gallon, the rate of fluid flow t'nrough the reactor would be about 87,000 gaUons per hou;. When c~.lll~ared to the rates taught in the art, F~mrle l l(B) makes HFPOdP at 57,000 lbs/gal/ hr as CC.l.~ d to 36/lbs/gaVhr for US 2,792,423 (making nu ~ ull ~.Grul idGs) and 1,300,000gms/ga~hr as cc.lll~al~d to 80-120gms/gaVhr, of active o~ygen cnnt~nt, for US 4,075,236 20 (ma'~ng ~ LV~ idGS). Thus, the ~.duclivilr of tne process of r~ S about 1000 tirnes the produe~.vily of the nearest art.
O~1;11~U~ OW rate through the lines is a filnrtirn of a ~lU~ ll,~. of ~
and ~ ;r~lv~ S Thesev ~'~' sincludethedegreeofinitialturbulence in the fluid that enters the tranfer line (i.e., tube), fluid dc .~;L;. -s and v;~co~;l ;FS, the 25 kin~ti~s of ~,.vAiJ~ ! the kin~ti~ s of pero~cide ~3~ ,.A~ n the 5 bs~ e or p~5,c~ of surface active by-products during hydrolysis, the lJl~s ~-~e of added ~. . . r~ or phase Llal.sLr ca~al~ " .ea.;~i~ll cc.. -~-- .l . ,. I ;. .. ~c, the identity of the cu~ t~ " c, the identity of the organic phase, the rate of LLal~rcl of organic and "1~7~ 1C species l.~ .l the ~~ us and organic phases and the t ..l~ -v~.c.
Mi~ing ,.. Il.o~lc useful in the proccss are high energy and high shear "-~ s ;... l..~ g ultr~onic, jet and static mi~ing methods (jet and static methods are l~lcf~ d for CO1-IJ1~UUU~ ~lùCC,SSeS, nltl~sonic methods are ~ cll~,d for batch uccsses)~ and ~ lur/lulol~. (m~rh~nit~l mi~ers). The mi~ers can be used singly or in series, in batch or cnntinn~-us processes. Various lllclhods of mi~ing wiU35 wod~ as long as they have enough energy and shear and are c~ cte.l rapidly.
mi~ing m~tho-1~ herein in~ 7 but are not linlited to, pif ~:oelc~ ' or ,1. ;C tive tr~nc~lnt~rs cuu~led to l.,sull~l horns and to liquid ~o~ d W O 97/08142 PCT~US96/13976 sonic and ultrasonic homo~ ~ , known to those skiUed in the art as methods of mi~ing and homo~ ~ liquids.
In order for agitation to reduce reaction time to less than one minute, ~l.,f~,.ably less than thirty ScCOl~dS in the present process, Reynolds nuul~ . should S fall from about 1000 to about ~0,000 for tne jet mi~ers and static mi~ers, and those se~ of transfer line within which pero~ide ft~rm~tion is cornrletç~l Reynolds ..... ~ are clim~.n.ci- nlecc ~ .s that .. 1.,.. ~. ~c-;,.e fluid flow in pipes or a fluid ;. .g with other solid bodies, c~k~ tecl from average fluid velocity, the of the body or pipe, and the kinetic viscosity of tne fluid. For e~S-mrlf-., 10 in the case of fluid flow through a pipe, the Reynolds number (al~ cvldtcd as RN) is calculated from the f~llowing formula:

RN 2~

15 in which V is the average fluid velocity in the pipe, R is the inside radius of the pipe, and N is the kinetic viscosity.
r*cn tc.ll~,alul~s are -lOQC to 40QC. OQC to 25QC are ~lcfc~l~d. A
de,,uable feature of the current process is that it runs at :~ulr~ ially ~q~ ~ .hl~
,5 and in some embo&ents cooling of the transfer lines can be c.. 1,1~ t .ly avoided (See P~mrle 1 lC). The reason the present process can make highly nnct~b]ç pero~ides at alllbi_nl ~< --~p~ ---c where prior art pluc~.,ses can not is the relative speed with which the process makes ~uAide, and can deliver it for use or St~r~g~.
Acyl halides her~in are sçlect~d from the group co~ .g of R(C--O)X and 25 RO(C=O)X in which X is -Cl, -F, -Br or ~ lc~.ably -Cl or -F;
R is selectecl ~.m the group co~ ;..g of:
(i) CnF~a~Hz~ wh~,lc~ + y + z = 2n + 1, n is 1 to 8, linear or b.,~ r-carbon bearing -(C=O)X ~lc~ably plilllaly;
(ii) G(CF2)w[ClF(CF3)CF2~tOCF(CF3)CF21y[0CF(CF30 where w is O to 8;
:~ is O or l;
y is O to 7;
z is O to 1 ; and w +~+y+z > 1 35 and G is a flllc)rin~ or a ~ub~liluled carbon group that is not highly l~,liVe toward water, hydro~ide, or hydrogen pero~ide and having one or more r. ..~ l groups such as, but not limited to, -F, -COOCH3, -S02F, H, -Br, -CFBrCF2Br (~ = O), -Cl, -I, -CN, -OC6F5; and .

W O 97/08142 PCTrUS96/13976 (ïu) an ~vll~Lic, hydro, chloro or perfluoro carbon coLllpvu~d having the same r..~ groups listed directly above.
Acid halide groups with higher mol.-c~ r weight alkyl groups are ~lcÇc.lcd in the process for hydrolytic stability when making aqueous dispersions of fluoro-S carbon pero~ides.
The bases useful herein are strong bases in water. They include metal or tetralkyl ~mmonillm hydro~ide or water soluble e4uivalc.-L,. The bases used are .l.,fc~1y > 0.1 molar in water. 1 to 5 molar base (for e~mpl- 15-23 wt % in the case of KOH) is most lulG~cll~,d. For fluoloc~ul,v.l processes KOH and LiOH are 10 most ~7lcr~,llcd. NaOH, CsOH and R'4N+OH-, where R' is CH3 and C2Hs- are hlr. Forl-y~Lvcalbv -s, useful bases are R4N+, where R' is CH3, C2H5-, or CH3CH2CH2CH2- and KOH, LiOH, NaOH, and CsOH.
Pl~rcll~d peroAide rea~t~nt~ are the H2~2 and t-BuOOH. The pero~ides are ~lcL,~ ly water soluble.
Solvents useful in the re~ti(~n are those that are u l.ea;~ with ~.vAides and that readily snlnhili7~- the ~v~idc. Solvents can be, for e~ , gases suchas h- .~flnG, v~lv~ylene if the reaction rmi~ture is held under .,. . rr;, ~ , to a liquid phase. No organic solvent is needed if the pero~ide is stable enough to be h~n~ safdy pure (See T~ ples 53 and 60). Fluol'vc~'vv 20 solvents are, in general, more useful with flll... ;,- ,t. ~1 perw~ides and hy~llvcdLbv solvents with Lydlvc~lJull peroAides. In gen~ , solvents are s~l~-cterl from a group of orgar~ic fluids .-i.~.,-- * ~ ;,. .1 by their ;.. - -- ~ c~ toward the l~ their ability to dissolve at least 1% of the pero~ide by weight and their ~cc~ .;1;ly in c- ~- ~--- ~- ~- ~ end use a~ ions for the ~e vAid~s. Suitable solvems are, for 25 ~mrle~ s~-~Dctecl from the group c-.. .~;;~l ;. .~ of Freon~9 El, Freon~19 E2, Freon~l9 113, fluorocarbons, chlorofluoloc~ubons, hy~Lvrluorocarbons, Lyvlvnuoluc lvloc~bulls, he~ane, cy~lol ~- ~ ~- . and mineral spirits.
S~ ce yield in some cases. Suitable ;,--. r~ include flll~; . .,.t~.~l s~ t ~tc and hydrocarbon s- . . r~ A l~ f~ .1 group c- . ~ of 30 ,--.... ;.. ~c.lluvluocl ~o~ and sodiurn dodecyl sulfate.
Ratios of n,~ Ix vary based on the structmes of the .~
U.S. 2,792,423, which is incvl~vl.,t~,d herein by IGrcl~n~c" claims a ratio of 1 to 2 equivalents of base to acid halide. With HFPOdP an 8 fold e~cess of base over acid halide is useful. The V~JLilIIUIIl equivalents of base to equivalents of acid J
fl~l~ n-l~ for this process falls within a 1:1 to 1:10 ratio. In the case of HFPOdP, .v,.ide is made in greater than a 50-90% yield using from 1 to 10 equivalents ofbase per e.lUiVi~l~,.ll of acid halide. It is e~l.e~k d ~at ~1;.,.;..;.~1.;.~, yields can be made outside these bounds. For the particular e.~ - - -l and cnnrlition.c d~ s~-- ;l .e-l W O 97/08142 PCT~US96/13976 in F~qmrle 1 l(B) the ple;r~ ,d ratio of base to acid fluoride was about 4 to 9 with the most ~lcre~l~,d ratio about 6.5.
Aqueous ~ ;v~ ~ of perfluorodiacyl peroAides are . rnong the useful products of the present process. It has been observed that the higher molo ~qr 5 weight peroAide used in a ~ ivn the longer the peroAide s~ vi~ ~s before . .;r;- nl l~y~uly~,~, results. Thus, for many a~lu~,OUS apFlic~tinn!~, it is that the higher molecular weight peroAides wiU be the most useful.
In ~ . . .n ~_. y, using ~,c ~ d mi~ing methods, Ll~livi~ ally or in series, in batch or comin~lo l~ ~vcessf s~ a rapid method as been developed that is uniquely 10 suited to production and delivery of highly lln~tqltle ~luAidf ~,llu~,Lul~,S by ViV~llg pro~hl~*- n and dcliv~,.y of the pf ~VAidf in less than one minute and in many cases less than 30 seco~ This is ~ g, p~rti~llqrly in the case of per~uoro acyl peroAides since the high i~lt~ ,iLy mi~ing and ~,Llvl4;1y basic cnn-litic-n~ ~vould be r~l.e~,t~l to ~ e~b-le ~ vAide loss to hydrolysis. Particular 15 adva.~l~s to the c~ vu. process herein are the fact that the process shifts a~i~..;r;~ portion of the ~?~ti~-n to the ~ ,r~,r lines. The processes ~lplu~, the safety of the .., .... r~ of the peroAides by ;... ~ , produ-iLivi~y while avoiding large ~.,.~Aide Jll~ f -. .l u. ;~5, and in some cases ~voi.li~lg the need for r~mger.q-ti~n The cor-tin-loll~ ~ivCCSSCS ~lc S ~ ;l -c~ herein [as shown in F~qrnrlf~ 1 l(c)] pqrti~ qrly when used in the ~bs~F --ne of heating or cooling fqriliti~s, can be c~ .h~.lLly used as a ~v~L~ble process.
Some of the novel ~c.vAides produced by the present process are of the following structure:
{ G(CF2)WtCF(CF3)cF2l~[ocF(cF3)cF2ly[focF(cF3)]~:(c=o)o- ~ 2 w is O to 8;
A iS O or l;
yisOto7;
z is O to l; and W+A+y+Z> i.
Specific C.l~O~ ; rl:~im~cl herein are of the above ~tLlu~,lu~ whf ~ G, W, -A, y and z are as in~ qt~-l below:
G = CH300C-wis 1 to4 ~ is O
yisOto7 zis 1 G = BrCF2CFBr-W is O
is O
y is 0 to 7 z=1 G = C6FsO-w is O
~ is O
yisOto7 ziS 1 G is I-w is 2 to 8 ~ is 0 yis0to7 z is 0 or 1 (z is 1 when y >0) Amv~g the specific novel cv~ ds r~ mr~l herein that faIl within the 20 above (lefinitinn.~ are the following:
tcH3o(c=o)cF2cF2ocF(cF3)cF2ocF(cF3)(c=o)o]2~
[CH30(C--O)CF2CF20CF(CF3)(C--0)0]2, [ICF2CF2(CO)0]2 [C6F5OCF(CF3)CF2OCF(CF3)(CO)O]2 and tBrCF2CFBrOCF2CF(CF3)0CF2CF2(CO)0]2.
r~ lNmONS
The following acrv ~yllls and l~ S are employed in the ~l~.S.~ n~, HFPOCOF. HPPO Dimer Acid Fln- rifle: CF3CF2CF20CF(CF3)COF
~7POdP. HFPO Dimer Pero~ide: [CF3CF2CF20CF(CF3)(C=0)0]-2 Fre-.n(~ CF3CF2CF20CFHCF3 Freon(~ P-7 CF3CF2CF20CF2CF(CF3)0CFHCF3 Freon 113: CFC12CF2Cl 4P, Perfluv.vbulylyl Pero~ide: [CF3CF2CF2(C=0)0]-2 ~, Pefluo,v~,v~ivllyl Pero~ide, [CF3CF2(C=0)0]-2 ~,Iso~uLy-yl Pcro~ide, [(CH3)2CH(C=O)O]-2 F.t~C, bis-(2-E~ylhe~syl)pero~ydica-bv .aLe [CH3CH2CH2CH2CH(C2H5)CH20(C=0)0]-2 FC-143: C7F15COONH4 -E~E. MethylPerfiluoro[8-(fluvlvrv.ll~yl)-5-methyl4~7-~ios~ .n..~ e], CH30(C=O)CF2CF20CF(CF3)CF20CF(CF3)COF

D~EP, Diacyl peroAide from DAE:
[CH30(C=O)CF2CF20CF(CF3)CF20CF(CF3)(C=0)0]2 S 1!~ E. Methyl perfluoro [5-(fluvr ~Çvllllyl)4~hrs~no~te], CH30(C-O)CF2CF2OCF(CF3)COF

l~P, D,a.;yl~.uA,de from MAE, ~CH30(C--O)ClF2CF20CF(CF3)(C=0)012 SF, FSO2CF2CF2OCF(CF3)CF2OCF(CF3)COF
10 ~, D;a~yl~uAidc from SF
[FS02CF2CF20CF(CF3)CFzOCF(CF3)(C0)0]2 1~, 7-H -Perfluv~vh~ vyl Chloride, H(CF2)6COCl 7~P, Dia.;yl~e.uAide frvm 7HCl, H(CF2)6(CO)OO(CO)(CF2)6COCl ~, BrCF2CFBrOCF2CF(CF3)0CF2CF2COCI
~, Dia.;yl~,rvAidc from BrCl, [BrCF2CFBrOCF2CF(CF3)0CF2CF2(CO)0]2 IF, 3-Iodu~lnuvlu~,v~iollylrl~-ri-lç, ICF2CF2COF
IP, [ICF2CF2(C0)0]2 P~ VAycoF~ C6F50CF(CF3)CF20CF(CF3)COF
P11~I1VAYP~ [C6F50CF(CF3)CF20CF(CF3)(CO)0]2 t-Bu~ cetS-~, CH3(C=O)OOC(CH3)3 5Cl, H(CF2)4CH2O(C=O)Cl 5PDC, Pe~uAydic~lJon~te from 5Cl: H(CF2)4CH20(C=O)OO(C=O)CH2(CF2)4H
5COF, a llliA~UlC of CI;3CF(COF)CF2CF2CF3 and CF3CF2CF(COF)CF2CF3 5P,the peroAide llUAIUlG from 5COF, [CF3CF2(CF3CF2CF2)CF(CO)O-]2 8nd [(CF3CF2)2CF(cO)O-]2 li.XAl~P~
The following F~mrles were c~ ndl~te.d in the ~ ..s dci,clibed by Figure 1, where llulllb~ d sL,uelui~s are as in~lic~~-~d above.
~XAl~:PT,~ 1 3û C.. ~ v~ ~ Reactor P~ ,.. of HFPO Dimer P~,~VAide in Freonq ) E2 Mi~ing T Followed by Ul~-~s.:...i-- MiAer The three l~ , aqueous H202, aulu~ us KOH, and HPFOCOF were ~im~ u~ly pumped ~hough a rni~ing T to an ultrasonic cell and then to a product collector, all in about l to 30 seconds. A detailed dcscli~lion of the 35 process is given below with reference to the numbered parts in the a1t~rhed reactor (Figure 1). It should be further noted that the s~l.. ,.l ;~ is not meant to be 1i...~ , For ~ Jlr, gear pumps could be used in the place of syringe pumps or the ~~ n~ ,tiOll cavity pictured as the reactor could just as well be a jet W O 97/08142 PCT~US96113976 mi~er, a static mi~er, an ulL~ ic whistle (holllo~ ), a m~ l mi~er (~L~ ulol), or absent alLo~,_Lllel (see E~ample 9 below). The essence of our Cnntim7C~ process is the rapid transit of precisely metered flows of le~
through one or more zones of intense mi~ing.
S Syringe pump (3), loaded with 15% by weight aqueous H202, was startcd up at 0.55 mVmin. Syringe pump (2), loaded with 24% by weight ~ueous KOH, was started up at 1.00 mVmin. Syringe purnp (1), loaded with 4% by weight HPPO dimer acid fl~ rirl~- in Freon~ E2, was started up at 3.30 ml/min. The strcarns from the three pumps, chilled with wet ice, were joined into a single stream 10 at a 1/8th" union cross Hoke~ fitting (4), a design feature possibly having an ...~;..h .~~ effiect on yield as can be seen from ~ ,lr 9 below. The ratio of at this point was 4 mdes ~f H2~2: 8 moles of KOH: 1 mole of HFPOCOF, making for an 8 fold e~cess of both H2~2 and KOH over HFPOCOF
in terms of reaction ,loi~ h,)/. T.. -P.l;~t~ Iy after e~iting union cross (4) at 0~C, the liquid stream was run via an 0.035" I.D. line of--0.1 ml volumc into the bottom cup of the 1.6 ml ~ ir reactor cavity (S) also chilled with wet icc.
The power source to the 3/8" .1;;.. ~ ~ ul~l~soluc hom (6) was tumcd on ~.vidi 18 to 20 watts of power to the ultrasonic cavity. Product e~ited as a stream at the top of the 11~ - ~ iC cavity at 2~C with an average rlc ~ e time in the cavity of 20 20 sec~. ..1s The reactor was run for several . ~ . ;. .. ~t~ ~; to flush out the lines and achieve steady u~.~ lg cor diti~n~ this Ç~ being run into waste poly c...-l~;.... (7). The~ streamwasthendi~,.t~,;ltos~mrlingc-~ (8) where 97 ml of çl 11. .. - .I were cc~ ctPd for product analysis. The organic layer was s~ t~,d and washed twice with 75 ml of 5% a~ o~ sodium ~ic~ dt~,. This gave 63 ml of 0.093 M HE~POdP in Freon~19 E2 for a 93% yidd based on star~ing HPPOCOF.
Using the c~ s- -- ;1 ~cl in r~ F 1 above, >70% yields of HPPOdP were obt~inPd using a va~iety of re~id~nre ~nes, ~loi~ . ;- ratios, c~ s, solvents, and mpthr~ls of mi~ing, typical results being y,~ C~;i in Table 1 below. In several of these runs the ultrasonic cavity (S) and ultrasonichom (6) pictured in the reactor s-~ 1 ;c were replaced by a 27 el~ f Kenics static mi~er having an intemal volume of 1.4 ml. Other static mi7~er designs should r...,.,;. ,- equallywell.

W O 97/08142 PCT~US96/13976 T,~RT.T~. 1 C ~ _ Reactor P~ of HFPOdP in Freon~D El and Freon~ E2 WT 9~o CON 'F~ATIONS RP~Tr~F~rP
MOLAR
E~c ~I2~ ~ T~FPOCOF ~ATIOS1 FRP~ Nl!9 ~5 ~ ~Pp~p2 15% 24%4% 4/8/1 E220 sec Ultrasonic 93.3%
2 10% 18%6~o2.5/5/1 E211.5 sec Ultrasonic 91.S%3 - 3 5% 12%8% l/Vl E220 sec l~ asonic 89.1%4 4 5% 24%4% l/vl E23 sec Ultrasonic 87.8%

5 30% 30%8.1%1.2/2.1/1 El 16.2 sec Ultr~sonic 82.9%

6 30% 30%16.3% 1.2/2.1/1 El 28.2 sec Vltt~sonic 82.3%

7 lS% 24%4% 1/2/1 E22.6 sec Static 73.3%

8 15% 24%4% l/Vl E21.3 sec Static 74.6%
lRdative molar ratios of H2021K0H/HFPOCOF in starting reactallt mi~
2Yield of HFPOdP based on startillg quantity of HFPOCOF
3Yield = 91.5 + 7.5%, average of 5 runs (85.1, 87.6, 88.0, 103.9, 92.8) 4Yield = 89.1 + 2.1%, average of 3 runs (89.5, 91.7, 86.21) 5Based on thel.6 ml volume of . ' ~ - cavity, ignoting the volume of the mi~ing T
(~0.2 ml) and the volume of the line to the cavity ~.~Pl,~, 9 Ccntim-Qus Reactor ~ Lion of HFPO Dirner Pero~ide in Freon0 E2 T.. l.. ~ ing in Mi~ing T Alone The sarne c~ as in r~ ,lr 1 was used e~cept that no power was supplied to the llltr~c~ni~ hom. The ll~ co.~ir cavity thus became a dead volumepl~OVi~ lg re~ en~e time for further reaction after ~ e~cited the union cross fitting serving as a mi~ing T.
SyIinge pump (3), loaded with 5% by weight ~l"e~J~ ~c H2~27 was started up at 6.07 mlJrnin. SyrLnge purnp (2), loaded with 24% by weight ~qn~-ous KOH, was started up at 3.47 mVmin. Syringe pump (1), loaded with 8% by weight HFPO dimer acid fln- ri-le in Freon0 E2, was started up at 22.87 rnl/rnin. The streams from the three pumps, chilled with wet ice, were joined into a single stream at a l/8th" union cross Hoke~ fitting (4) just after having passed ~ough a ~-.;.. :-.. -. con:~tricti<~n of 0.09" inch. The ratio of ~ .S at this point was 4 moles of H2~2: 8 moles of KOH: 1 mole of HFPOCOF, making for an 8 fold e~cess of both H2~2 and KOH over HFPOCOF in terms of reaction stoic~llion~.-try.T-----' ly after e~iting union cross (4) at 0~C, the liquid stream was run via an 0.035" I.D. line of ~0.1 ml volume directly into the bottom of a 1.6 rnl vessel (the 20 n~ -- ~ ~ reactor cavity with no power to the horn) chilled in wet ice. Product W O 97/08142 PCT~US96/13976 e~ited at the top of the cavity at 2~C with an average ~ c time in the reactor cavity of 3 seconds. The continllous reactor was run for several . . . ;. ~ s to flush out the lines and achieve steady ~.~L~lg C~ n~7iticm~ this ro~ o- being run intowaste poly c~ (7). The l~a.;lan~ stream was then ~ ,.t,cd to s~
S u""~ .r (~) where 97 rnl of effluent were collected for~l~nlu~;L .u~l~,~. The organic layer was ~ 1 and washed twice with 75 ml of 5% A~lJeu~ sodium b~a.l~l.~Le. This gave ~5 ml of 0. 118 M HFPOdP in Freon~ E2 for a 58.5% yield based on staT~ing HFPOCOF.
~.X~P-,~, 10 C~ tin~ous Reactor ~rL~ I;on of H~PO Dimer Perw~ide in Freon~9 E2 ~ ing T Followed by Static Mi~ing The sarne c.l! - ;p. . .~ - .l as in r ~ le 1 was used e~cept that two 27 dc~ t Kenics mi~ers, each with an intemal volume of 1.4 ml, were ~71~ t~ ~ for the ultrasonic reactor.
Sy~inge pUTnp (3), loaded with 5% by weight a luG~uS H202, was staTted up at 37.5 rr~hnin. Synnge pump (2), loaded with 24% by weight ~ uc KOH, was started up at 21.4 mll~nin. Syringe pump (1), loaded with 4% by weight HE~PO diTner acid f~ 7nfle in Freon~lD E2, was staTted up at 70.7 ml/rnin. The stroams from the three pu~s, chilled with wet ice, were joined into a single stream at a 1/8th" union cross Hoke~l9 fitting (4). Turbulent rnil~ing in the Hokel~9 fitting alone may initiate some l'e5~Cti~n as can be seen from F.~z n~le 9 above. The ratio of ~r ~ at this point was 4 moles Of H2~2: 8 moles of KOH: 1 mob of HFPOCOF, making for an 8 fold e~cess of botn H2~2 and ROH over HFPOCOF
in terms of ~ ,Li~ stoi~hiom~try. T....~ di .t~ Iy after e~citing union cross (4) at 0~C, the liquid stream was run via an 0.035" I.D. line of ~0.1 ml volume to two Kerucs mi~ers in series, each having an intemal volume of lA ml and c....l;.;..;..g 27 mi~ing elcll~.lL~ in a 3/16" O. D. X 7.5" long sl~ ss steel tube. Product e~ited the two static mi~ers at 6~C with an average recirl~-nr e tirne in the two static mi~ers of 1.3 sec~ The reactor was run for several .. .;...~t ,5 to fLush out the 30 lines and achieve steady ~...il~ c~nf1itir n~, this foreshot being run into waste poly c~ (7). The l.,a~ stream was then diverted to s~mplin~
c...~lAi."r (~) where 130 ml of effluent were collected for~r~luc~ analysis. Theorganic layer was s~alalcd and washed twice with 75 ml of 5% aqueous sodium b,c~lnJllaLc. This gave 70 ml of 0.087 M HFPOdP in Freon~9 E2 for a 90.5% yield 35 based on starting HFPOCOF.
When flow rates of all the reagent st~cams were dc ~ ,d 2X ~l~clca5ill~
l~-s~ re time in the static rni~ers to 2.6 s~c~..rl~, ~e yield of pero~ide dcc ca3~d to 86.9%.

F,~PT,~.11 Con*ntlous Reactor P~ ;on of HFPO Dirner Pero~ide in Freon~ E2 A. Jet ~~ r Alone~ ooc: The same e~ w~ used as in F~mr1~P 1 e~cept that the H2~2 and KOH streams were col-,l,...ed prior to m~ing with the 5 organic strearn and union cros~; (4) and ultr~onic reactor (5) have been replaced by a jet mL~er. The jet mi~er was a l/8" Hoke~9 T with an internal .li~ t- ~ of 0.094" and an internal length of 0.76", making for an intemal volurne of 0.086 ml.
In these runs the organuc phase w~ pumped straight through the 1/8" Hoke~9 T at 66.5 mvrninute. The cc..~lb~d ~ s KOH/H202 ph~e was pumped into the l/8" Hoke~l9 T at 63.1 ml/rnin via an 0.044" I.D. tube set 90~ to the organic flow.
Rr-1--~ ;u~ the .~ .ter of the tubing . l~t~ the side ir~let of the T to 0.044"
provides the orifice ~limpnsi~n~ required for jet mi~ing at the flow rates givenabove and for the v ;~co~;l ;es/ l~ s of the fluids.
Syringe pump (3), loaded with 15% by weight aqueous H2O2, was started up at 22.8 ml/rnin. Syringe purnp (2), loaded with 24% by weight ~qneo-1~ KOH, was started up at 40.3 ml/min. Using a 27 c~ Kenics static mi~er, the KOH
and H2~2 strearns were cvlnb ..ed to a single ~ ueou~ strearn flowing at 63.1 rnUrnin. Syringe purnp (1), loaded with 8% by weight HFPO dirner acid flnori~l~P in Freon0 E2, was started up at 66.5 rnl/rnin. Using 0.044" I.D. tubing, 20 the col.;~ 'l"- ous strearn was passed into the side arm of the Hoke~ T at 63. l rnl~min i. ~ g;~ ~ into the organic stream moving straight through the Hoke0 T at 66.5 rnl~nin. The ratio of l~.~CIA~ at this point was 4 moles of H2~2:
8 moles of KOH: l mole of HFPOCOF, making for an 8 fold e~cess of both H2~2 and KOH over HFPOCOF in terms of reaction stoi~ u y . The liquid stream e~iting the Hoke~D T at--7~C was run via an 0.085" I.D. line of ~3.3 rnl volume to the cc llPctic-n bottle. The 0.085" ~ m~ter of the e~it line is such that the turbulent flow may persist after the jet mi~er, although this was not c.~..l ;....~d by Average l~;d~ tirne in the reactor was 0.04 seco.lds cu..~ g just the volurne of the Hoke(l9 T serving as the jet mi~er or 1.6 s~ c~, ,.1~ co. ,.~ . ;, ,g 30 the volume of the jet mr~er and the e~cit lines together. The cu-.Li..uous reactor was run for l.S ...;. ~'~t~,S to flush out the lines and achieve steady ~e~.-l ;-.g cu~
this Çvl~_sll~t being run iUltO waste poly co.~ ,r (7). The le~ strearn was thent~,d to s -~ .1;..g c~ ~le~ (8) where 65 to 64 ml of organic phase were c~ 11PctP~l over the ne~t minute for ~u-~oses of product analysis. The organic layer 35 was .5~ h,d and washed twice with 75 rnl of 5% aqueous sodiurn bica,l,vll~u.
Over the course of two s~e runs this gave s--1nti~ n~ 0.173 M and 0.167 M in ~ HFPOdP in Freon0 E2 for yields of 89 and 84% ~ ly based on starting HFPOCOF.

Ma~ng ~65 rnl of 0.167 M HFPOdP in a volurne of 0.086 rnl in a rninute colle.,~ollds to a productivity of 43,000 lbs of HFPOdP/gallon/hr. The fastest co~ ~ble prior art we are aware of, US 2,792,423, reports making 7.2 g of 4P/100 rnl/min for a pro~luc;~iv ily of 36 lbs of 4P/gallor~hr. Putting this in the S l:mgnage of US Patent 4,075,236, a productivity of 36 lbs of 4P/galJhr is equal to ~1200 g of active o~ygen content/gal/hr and a produ~LiviLy of 43,000 lbs of HFPOdP/gaVhr is equal to 950,000 g of active o~ygen content/gal/hr.
B. JetMi~erAl~nP.26~C: Thesamee.r.;l.. 1 wasusedasinE~ample 1 except that the H2~2 and KOH streams were c- .."l,;l-r d prior to rni~ing with the 10 organic strcarn, union cross (4) and ultrasonic reactor (5) have been reF~ by a jet rni~er, and the 0~C ice bath has been replaced by a 26~C water bath. The jetmi~er was a 1/8" Hoke0 T with an internal ~l; .. ~ . of 0.094" and an intemal length of 0.76", making for an intemal volurne of 0.086 rnl. In these runs the organic phase was purnped straight through the 1/8" Hoke~l9 T at 125 rnl/minute.The c~ h;.~ d aqueous KOH/H2O2 phase was ~ .,d into the 1/8" Hoke~19 T at 85.0 rnlhnin via an 0.038" I.D. tube set 90~ to the organic flow. ~edll~ing the n .. t~ r of the tubing . . lt ~ ;- .g the side inlet of the T to 0.038 " provides the orifice .1;.. l~;~,~C required for jet m3~ing at the flow rates given above and for the v;~cv~ 5 of the fluids.
Syringe purnp (3), loaded with 12.9% by weight aqueous H2O2, was started up at 31.2 rnVmin. Syringe purnp (2), loaded with 18.3% by weight c~,.. g KOH, was started up at 53.8 mlhnin. Using a 27 el .. l Kenics sta~ic mi~er, the KOH and H2~2 strearns were co.-.l,;..ed to a single aqueous stream flowing at 85.0 ~/rnin. Synnge purnp (1), loaded with 5% by weight HFPO dimer acid n~ 1F in Freon~l9 E2, was started up at 125.0 ml/min. Using 0.038" I.D.
tubing, the cc~ rl aqueous stream was p~sed into the side a~ of the Hoke~l9 T
at 85.0 rnl/min imringing into the organic stream moving straight through the Hoke~l9 T at 125.0 mlknin. The ratio of re~ nts at this point wa~s 4 moles of H2~2: 6.5 moles of KOH: 1 mole of HFPOCOF, making for a largc e~cess of both H2~2 and ROH over HFPOCOF in terms of l~a~,~-on stoic~ . . .- l . y . The liquid stream e~iting the Hoke~9 T at 31~C was run via an 0.085" I.D. line of ~3.3 ml volume to the coll~ction bottle. The 0.085" ~La~ ,t~,L of the e~it line is such that the hlrbulent flow may persist after the jet mi~er, although this was not c~ r~ ~--- d by t~ Average residence time in the reactor was 0.025 seconds c~n~ rin~ iust the volume of the Hoke~19 T serving as the jet mi~er or 0.97 se~ lds cr n.si~lerin~ the volume of the jet mi~er and the e~it lines tog~th~
The continuous reactor was run for l .S . . . ;. .- - ~s to flush out the lines and achieve steady o~ . ,-l ;..g con-lition~, this r~.~h~" being run into waste poly c....l~ ;... l (7).

W O 97/08142 PCT~US96/13976 The ~ ;L~lL stream w~ then dive~L~d to s~n~p1ing c~ r (8) where 121 rnl of organic ph~e were collected over the next minute for ~ul~oses of product a.laly~is. The organic layer w~ ~ ed and washed twice with 75 rnl of 5%
~qll~ous sodium bi~,aLl,u-~le. Ioclnm~tric titr~tinn found 0.118M HFPOdP in S Freon~9 E2 for a yield of 95.5~ based on starting HFPOCOF.
Ms~king 121 rnl of 0.118 M HFPOdP in a volurne of 0.086 rnl in a rninute co~ ,pullds to a pro-lu ;livily of 57,000 lbs of HFPOdP/gallon/hr. The f~test cc.lll~le prior art we are aware of, US 2,792,423, reports malcing 7.2 g of 4P/100 rnl/rnin for a produuLivily of 36 lbs/gaUon/hr. Putting this in the l~n~lage ûf US Patent 4,075,236, a productivity of 36 lbs of 4P/gaUhr is equal to ~1200 gof active o~ygen content/gaVhr and a productivity of 57,000 lbs of HFPOdP/gal~hris equal to 1,300,000 g of active o~ygen content/gal~hr.
~ F~nrle 1 lB, the re~ction ~ ule is carried from the jet rni~cer to the product crll~cti. n vessel via a 38" length of ~ rss steel tubing with an intemal 15 .l;--... te~ of 0.085 inch. ThetablebelowshowshowtheyieldofHPFOdirner pero~ide varies with tne length of tubing. The sarne data is ~ d in Figure 2 wL~ tubing length is ~elated to yield of HFPOdP.
O~I i111UU~ r~.1;. .. length for this particular reactor co. .r;~ and r..l;... ~1;~...~t~risabout 38inches. Withshorter~ r..~ s~thefe~ iv 20 does not have time to go to co 1l~ nn and yields can be ~i~,.;r;. ~ y lower.
Clearly, much of the HFPO dirner pero~ide is being fomled in the ~ f~ . l ;. .~.following the jet rni~er. For ~uullJoses of ~r~du~ivi~y c~lr~ ti~n~ it is, ~ .l.~$, more ac~ ale then to cc,..~ .J the true reactor volume to be the c~....h;... -1 volume of the jet mi~er and its f~llowing ~ r~ . 1 ;nf, about 3.3 rnl in this ;. .~ ~ ~l-e Using 3.3 ml as the reactor volume in this ~ ,le makes our ~lutlu~;livily about 1400 lbs of HFPOdP/gaVhour or 31,000 g of active o~ygen cc.lll~nl/gaUon/hr, still large ;....~ ~c~ s over prior art.
F.ffect of Tr~n~fer ~ .in~. T ~n~pth ~ n HFPO-IP Yield Tnhiru T ~n~h yi.~l~l of }~FPO-lP
6" 33.3%
16" 72.2%
27" 72.1%
38" 90.3% + 4.6%
45" 80.4%
51" 77.8%
64" 39.1%

C. JetMi~serAl--n~ A.~ te.. ll~.n~ e; PartB ;.. ~ t~Iyabovewas repeated with no l~ P~ G control on the ~-------,----,~ feeds, the jet mi2ser, or s~i line to the c~ cti- n vessel. Ambient t~ , aLul~ the day of the run was ~4~C and product e~iting the reactor had a tc ~ G of 30~C, 121 m~ of organic5 phase being coll-octecl. The organic layer was sep~ and washed twice with 75 rnl of 5% ~ olle sodium bi~a l,undle, Iodom~tlic I ;l ~ ~ found 0.103 M
H~'POdP in Freon'l9 E2 for a yield of 83.8% based on star~ing HFPOCOF.
~ king 121 ml of 0.103 M HFPOdP in a volurne of 0.086 rnl in a mi~te c~. ~c sl,~ to a produ~ilivily of 48,000 lbs of ~FPOdP/gaUonfl~r. The fastest c.. ~ Ie prior art we are aware of, US 2,792,423, reports m~king 7.2 g of 4P/100 ~/min for a produ.,livily of 36 lbs/gaUon/,hr. Putting this in dle 1--~~ g of US Patent 4,075,236, a productivity of 36 lbs of 4P/gal~r is aIual to ~1200 gof active o2~ygen content/gaVhr and a productivity of 48,000 lbs of HFPOdP/gal/hr is cqual to 1,100,000 g of active o~ygen conta~t/gal~r.
F,~AMPI ,F'. t ~
ntinnous Rcâctor ~ n of HFPO Dimer Pelo,.ide in Freon~l9 E2 Jet Mi~er in Series with Static M~er HFPOdP was ~ ~.,d in e~actly the same ~ under the same C~ R as in T~~i.. plr1 lA e~cept that the line leaving the jet rni~er was 20 replaced by two Kenics static mi~ers in series, each having an intemal volume of 1.4 ml and c....~i.;..:..~ 27 rmi~g ~1~,.",,.,l,~; in a 3/16" O.D. X 7.5" long '-steel tube. Yields over the course of two runs, 86.5 and 86.8%, were ecs ..l ;:~11y the same within ~ error as those o~ d with the jet rni~er alone in r ~- 11Aabove.
~X~
~r.ntin~ol~ Reactor E:~c~ n of 3P in Freon~l9 El ~ ing T Followed by Ul~ ol~ir Mi~er or Static Mi~er A sollltion cu..l~;.-;..~ ~8 wt % pe~ ulu~lo~iullyl ~hlori~le dissolved in Freon~l9 El was first ~ cd. A 1 liter :jl,.;..lr..~5 steel cylinder c .. .l;.i. .;..g 747 ml (1142 g) of Freon~D El was chilled on dry ice, ev ~ ~ ~t~ , and 98.1 g of uuluplu~ic~llyl t hk~ le ~ tillf~,-l in. .As;~ aU the ~ nuulv~'~;~ yl ~hl~rid.~ to have dissolved in the Freon~9 El with a density on its part of 1.3 glml at room ~ p~ -.c, the final solution should then have a net volume of 747 rnl of El + 75.5 rnl CzF5COCI =--822 rnl and a net weight of 1142 g El + 98.1 g C2F5COCI = 1240 g, giving ~0.67 M C2F5COCl in Freon~ E1. At the room t~ t~ , the ~ei~c in such cylinders has been found to be about S psig.
Using PV = nRT this implies that e~h 100 ml of vapor space in the cylinder will contah at most 1 g of CzF5COCl lost from the Freon~9 E1 phase. For ~ul~oses of W O 97/08142 PCT~US96/13976 yield r~lr~ ti- nc this loss of C2F5COCl to the vapor phase was ignored, although it should be kept in mind that this loss of C2F5COCl to the vapor phase wiU makeyields appear a bit lower than they actually are.
Syringe pump (3), loaded with 5% by weight aqueous H202, was started S up at 16.64 mVmin. Syringe pump (2), loaded with 24% by weight s~lu~o~ls KOH
c~ g 0.6 g of FC-L43 ~... r~c,~.~1 (perfluc,l.,o~;L~"oic acid :~mm-~ninm salt)/100 ml, was started up at 9.46 ml/min. Syringe pump (1), loaded with ~8%
by weight C2FsCOCl in Freon~9 El, was started up at 37.8 rnl/min. The streams from the three pumps, chilled with wet ice, were joined into a single stream at a 10 1/8th" union cross Hoke~l9 fitting (4). The ratio of ~ n..l~ at this point was 1 mole of H2~2: 2 moles of KOH: 1 mole of HFPOCOF, making for a 2 fold e~cess of both H2~2 and KOH over HFPOCOF in terms of reaction stoirhir)rnrfry.
T....... ~.l; ,h ly after e~iting union cross (4), the liquid strearn was run via an 0.035"
I.D. line of ~0.1 rnl volume into the bottom cup of the 1.6 rnl ultrasonic reactor 15 cavhy (~;) also chilled with wet ice. The power source to the 3/8" fli~m~fer llltr ~sonir horn (6) was ~urned on, providing 18 to 20 watts of power to the ultrasonic cavity. Product e~ited as a stream at the top of the ..1~. ~co.~ic cavity at 18-19~C (having entered at 8~C) with an average lc~;-le.---e time in the cavity of 1.5 secc-n~1c. The reactor was run for several ...;....t~ s to flush out the lines and 20 achieve steady ~aLulg conditionc, this foreshot being run into waste poly C~ n;..- - (7). The ç~a~i~;~lL strearn was then diverted to s~mpling C~ ln;~ ( (8) where 69 ml of cLau~ were collc Cte~ for product allalyi,is. The organic layer was s~ip ;l and washed twice with 5% a~ue~us sodiurn bicall.vllale, and i<!~l-.. ..1. ;- _lly titrated as 0.164 M 3P in Freon~l9 El for a 43% yield based on an s~nm~cl starang C2F5COCl co~ n' ;on of 0.67M.
A similar run in the z~l s~ --- f- of FC-143 ;,... r~. IP. .- gave a 23% yield of 3P.
A third run using no ~--- rP- IP. .l and instead of an ~lltr~orlic rni~er a 1.3 sec ~,s;de-.-,., time through two 1.4 ml, 27 cl~.. ~ Kenics mi~ers gave a 18% yield of 3P.
FXAMPT F. 14 Ulllasol~ic Batch P~ HFPO dimer Pero~ide in Freon~l9 El An ice-chilled, 150 rnl beaker (barely large enough) was loaded with 3.01 g of KOH pellets (0.0458 mole) dissolved in 10 ml of water, 78 ml of Freon'l9 El, and 4.70 ml of 30% aqueous hydrogen pero~ide (0.0458 mole). A ~ . . horn 35 ~tt~hf-d to a 150 watt Branson ultrasonic power source was ~ ed roughly halr~ ~ down into the 2" deep pool of rç~ct~ntc A solution of 7.4 ml of HFPO
dimer acid flnori-l~. (0.0352 mole) in 13 rnl of Freon~ El was added. The ulL.~"~ic source was turned on for 30 secon~ Lavvlllg roughly 38 watts of power. The layers were ehen s~l.~aLed and washed twice with 50 ml of ~5%
~q~leol-~ sodium bi~,all,ulldle, layer sep~r~h~n~ taking 20, 20, and 10 sec~
~ ,s~c~;~i.~,ly. This gave 68 rnl of 0.228 M HFPO dimer pero~de (82% yield basedon star~ng HFPOCOF). TaWe 1 below ~ . ;,r s results for F.Y~mple 14 here and similar runs, F~mpl~s 15 to 19, in Freon~9 El and E2 solvents.

T,~RT.F, I
nic Batch P~ ;v~ HFPOdP in Freonll9 El and E2 KOH and H2~2 in E~cess MO~.~R R~TIOS
TOTAL YD~D
E~ HFPOCOF KOH H2~2 FREON~9 H20 RXW TIME ~PO<~
14 1 1.3 1.3 91 ml E1 10 ml 30 sec 82%
1 1.7 1.7 91 mlEl 28 ml 10sec 6S%
lv 1 1.7 1.7 91 ml El 28 ml 30 sec 68%
17 1 5.1 5.1 91 ml El 75 ml 10 sec 60%
18 1 17 17 182 ml E2 170 ml 10 sec 74%
19 1 17 17 182 ml E2 170 ml 30 sec 78%
li',X-Al~'lPI,~ 20 Ultrasonic Batch Pl~ala~ ll HFPO dimer Pero~ide in Freon0 El HFPOCOF in E.scess An ice chilled beaker was loaded with 1.64 g of KOH pellets (0.024 mole) ol-_d in 10 ml of water, 50 ml of Freon~9 El, 1.22 ml of 30% aqueous ll~,d,~g_.~ pero~ide (0.012 mole), and 5.5 ml of HFPOCOF (0.026 mole). A
hom ~ d to a 40 Khz, 150 wate Dukane nl~rasonil~ power source was i.. ~.~e<l roughly halfway down into the pool of r~ The ulllasu.lic source was eurned on for 30 secon-l~ ae full power. The layers were ehen sc p~ . ~t~ d and washed twice with 50 ml of ~5% aqueous sodium bi ,dLbullaLc;. This gave 50 rnl of 0.156 M HFPO dimer ~r~ide in Freon~ El (60% yield based on starting HFPOCOF, 65% yield based on starting H202). When using 5.0 rnl of HPPOCOP
(0.024 mole) instead of the 5.5 ml ,cl?o~t~d above, the yield of HFPOdP was 58%
based on both the H~POCOF and the H202.

Ultrasonic Batch Pl~lion 4P in Freon~ E1 A. T.ithillm Hvdro~ide as Base: A 250 rnl beaker was loaded with 2.5 g of lithium l.~d~ ide monohydrate (60 mmoles), 28 rnl of water, 0.1 g of ~mmonillm pcrflu~ Jo~ le, and 90 ml of Freon~ El with ice bath cooling. Once solution was c~ ~, 6.13 ml of 30% llydlog~,.l pero~side (60 mmoles) were added with stimng. The solution was chilled again to ~0~C and then 5.2 rnl of ~lflu~,lv-buty~ (35 mmoles) were added. An ultrasonic horn cv...~~ ~k ~1 to a 40 Khz, 150 watt Dukane power source was started up at ~50% of ,n~,. ;,.......
power and lowered into the reaction mi~ture After S seconds, ~ -,..co.~ wa~sstopped, the organic layer was s~ut~ d and w~hed twice with 50 ml of 5%
aqueous NaHC03. The organic layer had a total volume of 90 rnl and w~ found S to be 0.156 M in 4P for an overall yield of 80% b~ed on starting ~Inuolvl~u~ylyl ~-hlorifl~-. Producing 90 rnl of 0.156 M 4P solution in 5 sec~n-ls CO~ D~V .ds to a produ.;LiviLy of 400 lbs of 4P/gal/hr. The best prior art applicants are aware of for 4P, or for that matter any other process for any acyl pero~ide, is 36 lb/gal~hr (US Patent 2,792,423).
B. potAc~ .HYdr ~ q R~q~: A250mlbeakerw~loadedwith 3.96 g of 85% KOH pellets (60 rnrnoles), 28 ml of water, and 90 ml of Freon~l9 El with ice bath cooling. Once solution w~ co...l~l. t~-7 6.13 m1 of 30% l~yd~'v~
pero~ide (60 mm~-lçs) ~ere added with stirring. The solution was chilled again to ~0~C and then 5.2 ml of perfluv.vl~ulylyl .~hl- ricle (35 ~' ~~) were added. An ultrasonic hom co.. e~;t~ ~1 to a 40 Khz, 150 watt Dukane power souree w~ searted up at ~75% of .. -~ .............. ..~- power and lowered into the lea~ l ll~lulc,. After 15 sec. ..-.l~ ., c. .. - i~-l ion w~ stopped, the organic layer was s. ~ ~1 andwashed twice with 50 ml of 5% aqueous NaHCO3. The organic layer had a total volume of 88 rnl and was found to be 0.150 M in 4P for an overall yield of 76%
20 based on starting perfl~ vbulylyl chloride.
C. TcL~lvl;...~ nillm Hy~ as B~q~ A beaker w~ loaded with 15.5 rnl of 20% a~u~us i ~ -"".-";"", hydro~ide (22 .."..olPs) and 78 rnl of Fseon0 El with ice batll cooling. Once solution w~ cc~ te, 2.35 rnl of 30%
Ly~/gen pero~ide (23 rnmnl ~ ~) were added with stirnng. The svluLwll w~ ehilledagain to ~0~C and then 2.8 ml of ~uvlvl~ulylyl r~l~ri~le (19 "l :s) were added. An ~-11 ., ~ horn cc~ tecl to a 40 Khz, 150 watt Dukane power source w~ started up at--100% of .. ~ ;.-.. - power and lowered into the reaetion mi~tnre~ er 15 SCCO~ C~UItraSOn;r~t;Cm w~ stopped, the organic layer was S~1J -~ FA and washed twice with 50 ml of 5% aqueous NaHC03. The orgaruc 30 layer had a total volume of 70 ml and w~ found to be 0.088 M in 4P for an overaU yield of 65% b~ed on starting ~ . Il~o~vbu~ylyl chloride.
D. TeL.~ . .. l I .vl~ll~llolliuln Hvdro~cide ~ Base: A beaker w~ loaded with 7.9 rnl of 25% aqueous tetramethyl~ . hydro~ide (22 mmoles) and 78 rnl of Freon0 El with ice bath cooling. Once solution was complete, 2.35 ml of 30%
- 35 l1~ Vg~ v~,ide (23 ........ - le5) were added with stirring. The solution w~ chilled again to--0~C and then 2.8 ml of perfluvlvlJulylyl chloride (19 mmol~s) were added. An ~ ~~ ' horn co- ~i~CI~ ~1 to a 40 Khz, 150 watt Dukane power source was started up at ~100% of .. ~ ;.. power and lowered into the l~,~ iOII

W O 97/08142 PCT~US96/13976 mi~t~7~. Ager lS secnnflc~ ultrasr,nir~tinn was stopped7 the organic layer was se~.~u~t~.d and washed twice with 50 rnl of 5% aqueous NaHC03. The organic layer had a total volume of 75 rnl and was found to be 0.075 M in 4P for an oveIall yield of 59% based on starting ~e~nuolu~uLylylrh1or~
E. C~cinm lIy~ -A if ~ C R~c-~ A 250 ml beaker was loaded with 10.4 ml of 50% aqueous cesimn hydro~ide (60 mmnl~c), 20 rnl of water, and 90 ml of Freon~l9 El with ice bath cooling Once solution was cu. . .~ , 6.13 ml of 30%
hydrogen pero~ide (60 mmnles) were added with stir~ng- The so~ .. was chilled ag~un to ~0~C and then 5.2 rnl of perfluolo~uLylyl chlnr~ (35 . .-.- .r,lP s) were added. An ~ o~-ic hom C~J~ e~ 1 to a 40 Khz, 150 watt Dukane power source was sta~ted up at--75% of .. -~ ;.. power and lowered mto the .~
After 15 seconds, ultr~oni~tion w~ stopped, the organic layer was s~r-i~ t~l andwashed twice with 50 rnl of 5% aqueous NaHC03. The organic layer had a total volume of 88 rnl and was found to be 0.112 M in 4P for an overaU yield of 56%
15 based on starting p~ l)ulylyl chlori~
F. Soflillm HYdro~ as ~c~- A 250 rnl beaker w~ loaded with 2.4 g of NaOH pellets (60 .. nles), 28 rnl of water, and 90 ml of Freon~l9 El with icc bath cooling. Once solution was cn...~ e7 6.13 ml of 30% hy~llo~ pero~ide (60 .~ lrs) were added with stirring. The solution was chilled again to ~0~C andthen 5.2 ml of ~ uol~Jbulylyl ~hlnri~e (35 mmnl~s) were added. An .. .l~ . A50. ~ i~
hom co~ r ~ t~ .l to a 40 Khz, 150 watt Dukane power source was started up at ~75% of . . . ~ . . power and lowered into the res~ion IllU'.IUl~.. After 15 secu...l~., nltr~c....i~ ~- w~ stopped, the organic layer was su f~ d and washed twice with 50 ,nl of 5% n l"eo~. NaHC03. The organie layer had a total volume of 88 ml and was found to be 0.075 M in 4P for a7n overall yield of 37%
based on starting ~f . I~ ol~u~ylyl ~hlnri~.7 F.X~l~P~.~. ?-~
7J7.~ ....;c Bateh PlG~,-.,.I;on 3P in Freon0 E1 A beaker w~ loaded witn 3.96 g of 85% KOH pellets (60 ......... ~ s), 0.1 g of ~ - n.-OIvo~ , 28 rnl of water, and 40 ml of Freon~ 7 1 with ice bath cooling. Onee solntion was comrl~!te, 6.13 ,nl of 30% I-y~LoOe.~ ,Aide (60 ~les) were added with stirring The solution was chilled again to ~0~C and then 87.5 g of a Freon~g El solution C~ ;.;..;..g 6.88 g of perflu~ r~,~;v.lyl rhl~n j~7.~ (3~ r~moles) were added. An ultrasonic hom co~ c~ ;d to a 40 Khz, 150 watt Dukane power source was started up at ~75% of .. ~ ;... power and lowe ed into the reaetiC-n ~ c;. After 30 secc,.lds, Illt~.~7~c ni~ 7tion was sto~ped.
The organie layer s~ ~1 imm~ t~1y and was washed twiee with 50 ml of 5%
&~lue~Ju~ NaHC03. The organie layer had a total volume of 85 ml and was found W O 97/08142 PCT~US96/13976 to be 0.136 M in 3P by iodometric tih~ti~n for an overall yield of 61% b~ed on starting perftuo~ yl chloride.
~X~ P~,~, 7~
UlLI~ollic Batch Pl~a~ n of HFPOdP as Aqueous Di~r~ n S A beaker was loaded with 1.78 g of 85% KOH peltete (27 '~ f~lPS~)~ 0.l g of ~ V~ perfluoroo.,~ , and 60 ml of water. Once solution wae c~ L,te the mi~ture wae chilted on ice and 1.4 ml of 30% hy~llu~ pero~ide added (14 mm<~les) and then 5.0 ml of HFPO dimer acid ll...J. ;<le (23.8 ~ - ~"s).
An nltrpeonic hom c- .... ~ct ~ ~l to a 40 Khz, 150 watt Dukane power source wae10 started up at ~100% of ~ ;-.----.. power and lowered into the r~action l~
After 30 sec~ the power wae tumed off to the ultraeonic hom and the white stirred gently while cooling on ice for a 1 minute l;~ lillg period.
Fif~y rnt of Freon~ 113 were added to the reaction ~ ur~.* The reaction mi~cturewas then l-~u~r~,lled to a 5~pa" ~ y funnel and shaken vigolously co,, ' Lld~ r~ of HFPOdP to the Freon(l9 113 layer. After w~l ,lg twice with 50 mt of 5% aqueous NaHCO3, the organic layer had a total volume of 50 mt and was found to be 0.150 M in HFPOdP by io-l-.mP-tnc titr~tion~ giving a 63% yield of HFPOdP based on star~ing HFPOCOF and accvulllillg for 53% of the starting H202. E2r-tllming to the original water layer, its volume was found to be 60 mt and 20 io~ ti~tion found it to be o~l lo M in pero~ide~ accvullLi l~, for the 1~ ."a;.,;"~ 47% of the starting H202. The pero~ide in the aqueous phase is likely either u..,~ Led hy~L~Jg~.l peroxide, the pero~y acid sodiurn satt CF3CF2CF20CF(CF3)(C=O)OO~Na~, or a mi~ctnre of the two. Indeed in a simitar run, ~ :.l;r;~ io.~ of the ~ '~"~ phase by sulfuric acid foltowed by ,,~ with 25 Freon~9 113 gave a 9% yield of pero~ide ~ cl to be the peracid CF3CF2CF20CF(CF3)(C=O)OOH . The resutts of F~mrle 23 here as welt as of r ~ ' - 24~.LO ~ the effects of power level, rip~ning time, and s... ra. ~a.~t c.. ~ t;"" are ;,--.. ~. ;,- -1 below.

These d~ s are not stable without continuous agitation~ al~ough it is ""I ;r l, h rl that true em~ n~ coutd be obtained by e~ploring other s~ ~ r~
ilJly aided by the a~l-liti.-n of smalt amounts of ftuorocarbon ftuids such as hP~fln.J.~ ylene o~ide based oli~5c,lll~,.~. In order to avoid the ha~.al~t~lus phase sep~ti~n of apu~e pero~ide layer, Freon(lD 113 was added as soon as P,,'i' ' ~ was stopped.

W O 97/08142 PCT~US96/13976 TA~T.F 2 Ultrasonic Batch P~alion of HFPOdP as Aqueous Dic~.
Acculd;..~ to Method of F~mrJe 23 Ultr~onic Power D~ n*
E~.FC-143 % Tirne Ri~nin~ Tirne Yield HFPOdP
Vary Power to Ultrasonic Hom 23 0.1 g100 30 sec1 rnin 63%
24 0.1 g80 30 sec1 rnin 62%
0.1 g60 30 sec 1 min 54%
26 0.1 g40 30 sec 1 min 47%
27 0.1 g20 30 sec 1 min 12%
Vary Ripening Period 28 0.1 g75 30 sec none 48%
29 0.1 g75 30 sec1 rnin 55%
0.1 g75 30 sec 3 min 48%
31 0.1 g75 30sec Srnin 54%
32 0.1 g75 30 sec10 min 50%
Vary Level of FC-143 S--rf~rt~nt 35 none 100 30 sec1 rnin 30%
36 0.05 g 100 30sec 1 rnin 42%
23 0.1 g 100 30sec 1 rnin 63%
37 0.2 g 100 30 sec1 rnin 55%
Vary Period of Ul~o.-i~ ;on and E~i~ning Period 38 0.1 g 75% lS sec none 55%
39 0.1 g 75% 30 sec 3 rmin 62%
40 0.1 g 100% 30min none 49%
*These dispersions are not stable without conl;.~ous s~its~ti<~n In order to avoid the l.aLal~lvu;, phase set~ ic.n of a pure pero~ide layer, Freon~9 113 was addedas soon as ~gitz-tion was stopped Clearly it is iUll~ llalll that enough power be delivered to the nl~ .J.~
horn and that enough s~lrf~t~nt be present. It is ;..t . ,~ ;..g that one can achieve modest 30% cvllvcil~ions of HFPO dimer acid fluoride to pero~ide in the absence S of both organic solvent and emulsifying agent. Varying the ripening period from 0 to 10 ...;..~ ,s or r~h~n~ing tne period of ultrasonic mi~ing from 15 sec to 30 ~.; ~- t s had orly rninor effects on yield: thus, ~ perci~ ns of HFPOdP are ly stable hydrolytically, HFPOdP in water ~ .. . possibly having a hydrolytic half-life on the order of hours.

W O 97/08142 PCTrUS96/13976 ~itiation of 'l ~: Poly...~ . with A~ueous HFPOdP
A500mlpoly..~ .;, Al ;n.~ kettleloadedwith 1 gof A~ J~r'~
nuv~ooct~no~t~- ~. . . r,.- ~ and 100 ml of water was pre~ hill~d on ice.
In P.~mple 40, above, a 5 rnl sample of ~l; ~ was wiLlldlawll 5 30 ~ec~ into the period of nl~ ;r;- ~ti- n This di~ e~;~ivll sample w~
;,,,,,..uli; lely added to the polyrnerkettle. The kettle was sealed, l~r 7' ~ ~t,~
s~ c~l with argon and cv;,~ tr~l~ and then filled to 67 psi with tetrafluv~ Lllylene from a 1 liter ~;yLI~le~ c~ ;.;.-i..g ~33 g of r~ n~ u~,LLyl~c.
This ~ lUl~ was heate:d to 37-41~C. After 2 to 3 hours, the poly~ ic ~lu~lu-;l 10 was ~c,co~,~.cd by filtr~tion, washed 3X with 1:1 ~-,. ll.~.~(~l water, and dried under ~uulll giving 11.8 g of poly(tetrafluo,u~ ylu.,c).
~XAl~qPI ~ 41 Batch Stator/Rotor E'lc~>~ualion of HFPOdP in Freon~ El The COllt~ S of a Waring blender cup loaded with 3.95 g of ~v~
15 hydro~ide pellets (0.06 mole) ~ solvcd in 28 ml of water, 78 rnl of Freon0 El, and 6.13 ml of 30% hy~l~v~en pero~ide (0.06 mole) were stirred for 10 scc~ ls Then 7.4 ml of HFPO dimer acid fluoride (0.0352 mole) dissolved in 13 rnl of Freon0 E2 was poured in ~ a single charge. The mi~ture was stirred at "LO"
speed for 30 sec~ ls~ The organic layer was scp~ ~' and washed twice with 20 50 ml of ~5% sodiurn l,ic~uL,ullale, giving 77 ml of 0.168 M HFPOdP soluLull (74% yield). The yield of HFPOdP .Icc~ sed to 55% with a 10 secollds reaction time and to 47% with a 5 secolld reaction time.
~,X~ IPI,F' 42 Batch Stator/Rotor P~ ~a~ion of HFPOdP in Freon0 E2 The collt~,nL~ of a Waring blender cup loaded with 3.96 g of ~OI~;ull-hy~Lo~idc pellets (0.06 mole) dissolved in 28 ml of water, 169 rnl of Freon~D E2, and 6.14 ml of 30% hyd~ .l pero~ide (0.06 mole) were stirred for 10 s~c~
Then 3.7 ml of HFPO dimer acid fhlori~le (0.0176 mole) dissolved in 13 rnl of Freon~l9 El w~ poured in as a single charge. The n~Alule was stirred at "LO"
30 speed for 15 sec~ The organic layer was sc~ d and washed twice with 50 ml of ~5% sodium l~iCdl~ull~l~, giving 184 ml of 0.037 M HFPOdP solution (78% yield). The yield of HFPOdP dc_l~ased to 74% with a 5 second reaction ~ time.
~.~rUPI ~ 43 Ultrasonic Batch Pl~ al~lion of DAEP in Freon~9 El An ice chilled beaker was loaded with 1.5 g of KOH pellets (0.022 mole) dissol~ed in S ml of water, 78 ml of Freon~l9 E1, 2.35 ml of 30% a~l-,cuu~ Lyd uAide (0.023 mde), and 5.5 ml of DAE (--0.019 mole). A 1;~ .. hom W O 97108142 PCT~US96/13976 attached to a 40 Khz, 150 watt Dukane .. l~ .. ic power source was i.. ~erl ;vughly halfway down into the pool of re~ct~nfs The ul~ ,v..ic source was turnedon for 30 secvl~s at full power. The layers were then S~ t~ and washed twice with 50 ml of ~5% aqueous sodium bi.~.l,o~ . This gave 75 ,nl of 0.086 M
S DAEP in Frcon~19 El (68% yield based on starting DAE). A DAEP s~ tinn passed over Drierite~ l.yd.vu;, c~l~inm sulfate), was I . ~ r~ . . ed to a 20~C water bath, and s ,- ~ es witndrawn over time for ioclom~triC 1 ;l ~ At 20~C, DAEP
dcco..l~osed with a rate of ~9 X 10-5 sec~l (a half-life of 2.2 hours).
Use of DAEP to Initiate Poly..-~ - ;~..l;....
A500mlpoly.. ~ nket~ewasloadedwith 100mlofFreon~9 113 and chilled on wet ice. Once cold, 5.0 rnl of a solution 0.086 M in DAEP ~IvAidc wasadded. The kettle was flushed S X witn argon, once with S psi of TFE, and then fed with TFE from a one liter cylinder c~ ;..;..g ~33 g of ~ . Over the ne~t two hours TFE ~rt~7~7~uc in the kettle cLv~cd from 47 to 32 psi with h ...~
going through a ~ ~ ~, ~ ;- ~ ~- ~- ~ ~ of 44~C. Solid ~oly~ L was isolated by ft~ tion, w~hed 2 X with Freon~' 113, and d~ied giving 13.6 g. At 372~C the melt inde~
with a 15 kg weight w~ 0.08 glmin. Tnfr~tp~l analysis of cold pressed films found what was likdy a -COOCH3 band at 1790 cm~l.
7l~ 1UPT ~ 44~
Ultrasonic Batch P~ of MA~P in Freon~9 El An ice chilled beaker was loaded with 3.0 g of KOH pellets (0.044 mole) dissolved in 5 rnl of water, 78 ml of Freon~l9 E1, 4.7 ml of 30% .q~ lly~
perw~ide (0.046 mole), and 3.9 ml of MAE (~0.019 mole). A ~ hom h~A to a 40 Khz, 150 watt Du~ane U~ 7~.7lliC power source wa~s ;- ~ ~- ~ ~. . ~cd roughly halfway down into tne pool of .~ . The nl~onic source was tumed on for 30 seco~--ls at full power. The layers were then i,e~A~ A and washed twice with 50 ml of ~5% a~lu~uus sodium bical'~ ull.~t~,. This gave 73 ml of 0.089 M
MAEP in Freon0 El (68% yield based on starting MAE). This solution was ,.".. ~r~ d to a 20~C water bath and 5.0 ml 5~mpl~s p. . ;~li~lly willl~Lavvll for io~u~ ;cl;~ At 20~C .~he ~lu~idc dcco...,uosed at a rate of ~8 X 10-5 sec~l (a half-life of ~2.4 nrs). If prior to tne kinetic ,..c~ ..l s the MAEP solution is passed tnrough Drierite~l9 (a~hy~L~.us calcium sulfate) to ;emove the last bits of water, then the rate comes out ~9 X 10-5 seC-l7 w~ ~r;"~
error.
~.X~I~P~ ~. 44~
Ultrasonic Batch P~ ali~ of SFP
An ice chilled 'oeaker was loaded with 1.5 g of KOH pellets (0.022 mole) and 0.05 g of ~nm--nillm ~IU~ OeI~O~Ie dissolved hl 5 ml of water, 78 ml of _ _ _ W O 97/08142 PCTrUS96/13976 Freon~lD E1, 2.35 rnl of 30% aqueous hydrogen pero~ide (0.023 mole), and 6.5 of SF (~0.019 mole). A ~ horn Ast~rhto~l to a 40 Khz, 150 watt Dukane ultrasonic power source was ;.. ~ . ~ed roughly halfway down into the pool of,c~ . The ultrasonic source was turned on for 30 secu lds at full power. The S layers were then sc~t~ d and washed twice with 50 ml Of ~5% ~ sodium bic~l/ullalc. This gave 75 ml of 0.083 M SFP in Freon~l9 E1 (~66% yield based onstarting SF). A SFP solution passed over Drierite0 (~Ihy~Luu;, c~lr-i-lm sulfate), was ~ r~ d to a 20~C water bath, and s~mp'-s will~Law-- over t~ne for io~lomf~,tric, titration. ~t 20~C, SFP dccoll-~osed with a rate of ~7 X 10-5 sec~l (a 10 half-life of 2.9 hours3.
ln the Al'~S'~ e of :~mmrnillm perfluoroo~ .,le yield was 54%. In the absence of ~ O~ . perfluorooctanoate and using Freon~9 113 as solvent, yield was 61%.
Use of SFP to Initiate Poly . . .~ I ;u. .
A 500 ml poly. . .~ l ;on kettle was loaded with 100 ml of Freon~l9 113 and chilled on wet ice. Once cold, 5.0 ml of a solution 0.06 M in SFP ~UA dc was added. The kettle was flushed 5 X with argon, once with 5 psi of 1 ~ ~:, and then fedwith l~;fromaonelitercylinder-7riginAllyc...lls.;..;..g~33gofTFE. Over the ne~t 1 to 2 hours l ~P ~Ul~,S;~U~C in the kettle dropped from 49 to 30 psi with 20 ~ .l... e going through a .~ ;.. of 52~C. Solid polymer was i~ ted by filtr~ti~n, washed 3 X with Freon~Z9 113, and dried giving 14.0 g. At 372~C a melt indeA ~ ;.... -..I with a 15 kg weight gave no flow. Cold pressed polymer flms showed IR bands at 1470 cm~l c. l~ l with the ~l~,se.l~ of -S02F groups.
Use of SFP to Fnn~tiQn~li7~ ~.,.k.e.~t.
A 200 ml r.b. flask at -25~C was chaL~,_d with 140 ml of 0.07 M SFP
(9.8 mm- l~s SFP) in Frcon~ E l . Using a V~.iUUIII pump, the solution was Gv~,uul t~,d down to ~40 ml while controlling its t~,,ll~la~uie to -15 to -30~C.VVhile at -30~C, 10 ml of dcuAy~ cl l~..Le.~ (112 mm~les) was added by syringe. The IllL~IUI~ was ~ lly warmed to room le-..l ~ (~25~C) and 30 stirred for several days under llihU~ l. l V~Ul~LIing down gave 9.97 g of yellow fltud. GC/MS showed two major peaks, 30 area percent (C6H5)CF(CF3)OCF2CF(CF3)OCF2CF2SO2F (e~act mass 541.943674 versus a cs~ t~l mass of 541.9669039) and a second nni~ ed m~t~n~l Radical Dirner from SFP
A beaker was loaded with 3 g of ~OH (44 mmoles) pellets dissolved in 10 ml water, 4.7 ml of 30% aqueous hydrogen pero~ide (46 mm~ les), and 150 rnl of Freon~D 113 with ice bath cooling. Tmm~ t~ly upon ~1rlition of 13 ml of SF
(~38 mm- l~.s), the Illi~lulG was u~ so~ ..t~ A for 30 seconds using the ...-~.;........

W O 97/08142 PCT~US96/13976 power output from a ~ . horn ~tt~ch~d to a 40 Khz, 150 watt Dukane l-1tr7sonic power source. The organic layer was s~ d, washed 2 X 50 ml of 5% ~ 1 o:_s sodium l~ic~l,<,l~ale and dried by passing through 25 g of D~ e~
)/~UU;~ c~1cil-m sulfate pellets). Four such runs were c.. 1~ d and found to S be 0.081 M in SFP by ioclnmr~ric thr~tic)n The ~ ueL was then allowed to stand for four days at room t~ ...pl ~ ~-...c in a r.b. flask under a ~osili.,c l"~ -e of ug~ ll. As a safety ~C~ the ll~il'.~UlC was reflu~ed for 6 hours, washed with acidic ~queous KI and then ~UCOU5 sodium ~. ~l l ile. The organic phase wasdned and then stripped down to 50.82 g of milky fluid on a rotary GV~4Ul.. Lt~.
Vacuum ~ till:-tit~n gave 25.47 g of water white oil bo 05 = 83~C. Gas ~ LLol~ldLu~lly found a major 74% C(jlll~V~ that by ~ 1-. . - . ;c~l inni7 ~ )n lllass ~pc. L~vSCOpy showed a parent - F peak at 910.86 (Calc 910.8571) cc.~ ondi to the radical dimer shown below:
tFS02CF2CF20CF(CF)3CF20CF(CF)3-] -2 ~XAl~IP~ ~ 4~
Ultrasonic Batch Plc~ n of 7HP
An ice chilled beaker was loaded with l.S g of KOH pellets (0.022 mole) dissolved in 5 ml of water, 78 ml of Freon~ El, 2.35 ml of 30% a l~.cous hydrogen pero~ide (0.023 mole), and 4.1 ml of 7HCl (~0.020 mole). A ~ ;~ ,.. ~ ;.... ~ hom ~ dtoa40Khz, 150wattDukane.. ll.;c.. ~ powersourcewas ;.. ~
roughly hal~w~.y down into the pool of r~ .;, The ~ co..;C source was tumed on for 30 sec~ n~ls at filll power. The layers were then s~ t ,d. This gave 75 rnl of ~0.10 M pero~ide in Freon~D El for a 77% yield.
Use of 7HP to Initiate Poly...~ n A 500 rnl poly.. ; ~ n ketde was loaded with 100 rnl of Freon~l9 113 and chilled on wet ice. Once cold,5.0 rnl of the 0.1 M pero~ide 5~1n~inn ~ d in r~ qJle 45 was added. The kettle was flushed 5 X with argon, once with S psi of TFE, and then fed with TFE from a one liter cylinder ori~in~lly ccJ~ ~33 g of ;. Over the ne~t ~2 hours TFE ~>lc~ur~ in the kettle d.o~?ed from 60 to 30 psi with ~ , going through a .. ~ ~ ;.. of 58~C. Solid polyrner was isolated by 1 ;1~ , washed 3 X with Freon~9 113, and dried giving 14.2 g. At 372~C thernelt inde~c with a 15 kg weight gave a flow of 0.04 ghnin. The polymer had an infrared adsol~lioll at 3006 cm-l cl~ns;~ .IL with the presence of -CF2H end groups.
~Al~Pll F 46 Ultrasonic Batch Pl~ p:.i.., ;t)n of BrP
An ice chilled beaker was loaded with 1.5 g of KOH pellets (0.022 mole) dissolved in S rnl of water, 78 rnl of Freon<~9 El, 2.35 rnl of 30% aqueous L~

W O 97/08142 PCT~US96/13976 peroxide (0.023 mole), and 9.2 ml of 75% pure BrCl (~0.023 mole). A l;lz ~;.....horn attached to a 40 Khz, 150 watt Dukane nl~ n.~ic power source was i" " .. ~etl roughly halfway down into thc pool of ,~ ~<~ The n~ 3.~........... ic source w~ turned on for 30 seCc?n~l~ at full power. The layers were then Sep~.~t~ tl and S washed twice with 50 ml of ~5% ~ eu~ .~ sodium l;~ica bvl~t~,. This gave 74 ml of 0.10 M BrP in Freon~9 El (~64% yield bascd on starting BrCl).
Use of BrP to Initiate Poly A 500 ml poly. . ~ ;on kettle was loaded with 100 rnl of Freon~l9 113 and 5.0 ml of a solution ~0.1 M in BrP pero~ide. The kettle was flushed 5 X with 10 argon, once with 5 psi of l ~l ;, and then fed with TFE from a one liter cylinder ~riginzlly CO~ r. ~33 g of l ~. Over the next ~2 hours l~- p.~si,~e in the kettle dropped from 54 to 25 psi with l~ c going ~l..vu~;l. a .~ .. of 54~C. Solid ~oly~was isolated by filtr ti~n, washed 3 X with Freon~19 113, and dried giving 20.6 g. At 372~C the melt index with a 15 kg weight gave a flow of 0.09 g/min.
~ rPT,~ 47 Ultrasonic Batch Pl5~ ;on of IP
An ice chilled beaker was loaded with 1.5 g of KOH pellets (0.022 mole) dissolved in 5 ml of wal:er, 78 ml of Freon~9 El, 2.35 ml of 30% Z.~ lly~llu,~n pero~ide (0.023 mole), and 1.5 ml (3.04 g = 11 ~7 - s) of IF. A t;~- .;.. .horn s~tt~rh~l to a 40 Khz, 150 watt Dukane . .1 ~ . . ir power source was ;.- ~. . .- . ~1 roughly h~lÇway down into the pool of l~ ;, The ~ ...;C source was turned on for 30 seconds at full power. The layers were then ;,~ t ti and washed twice with 50 ml of ~5% ~queo -~ sodiurn bic~ . This gave 72 ml of 0.022 M IP in 25 Freon~lD El (~29% yidd based on starting IF). Even though this was stored in a -15~C refrigerator in~ hle iodine crystals were soon visually ~t,~
Iodine ;,. .h;,~ cl peroxides appear quite lm~ts~hle and should be hs-m11~.fl with cZlltion- On~mzn-l, contin-l--us methods of making ~.v~,ide may be a p:lrtif ~ rly safe and e~ ve way to make and use iodinated ~ vAides.
UseofIPtoInitiate Poly.. - -;,-~;.. ~
A500ml~oly...- -;,~ n kettlewasloadedwith 100rr~ofFreong~ 113 and 20 ml of a solution ~0.022 M in IP pero~ide. The kettle was flushed 10 X with argon, twice with 5 psi of 1 F 1~:, and then fed w*h l ~: from a one liter cylinder (~rigin~lly cv~ ;-~;U g ~33 g of TFE. Over the next ~2 hours ~ s~ le in the kettle dropped from 58 to 38 psi with l~ a~ul~ going through a .. ~ ....... ...of 40~C. Solid polymer was isolated by filt~ti~n, washed 3 X with Freon~9 113, and dried giving 9.5 g. At 372~C the melt inde~ with a 5 kg weight gave a flow of 2 g/min, the extrudate being a strong pink from l~,leased iodine.

CA 02230606 l998-02-26 W O 97/08142 PCT~US96/13976 FX ~IPT,~,48 Ull., ~ ic Batch F'~ n of PL~l<~yP
An ice chilled beaker was loaded with 1.5 g of KOH pellets (0.022 mole) dissolved in 5 ml of water, 0.05 g ~.. u.. ;.. p~.rfln~uo~ t~ 2.35 ml of 30%S ~ co~s hydrogen ~.o"i~ (0.023 mole), and 5.29 g of Pl~llu~yCOP
~10.7 mm. les) dissolved in 100 ml of Freon0 El. A l ;~ ;... horn /~ .1 to a 40 Khz, 150 watt Dukane ull.~sv.~ic power source was ;..... ~ ,d roughly l~ .ydown into the pool of n,~cl ~ ~I.c The ulll~ol~ic source was turned on for 30 SeC~J~ at full power. The layers were then s~ tu~1 and washed twice with 10 50 ml of ~5% a.lue~ sodium bicOlbol~dle. The organic phase was washed through 25 g of Dl;~ e~ (~ihy~uus c~ m sulfate) with an aA~;l ;OI ~l 25 rnl of Frcon0 El. This gave 115 ml of 0.05 M Pl~.~o~yP in Freon0 El (~100% yidd b~ed on star~ ng Pl~ J~yCOF).
Use of Pll~,nc,~yP to Initiate Puly. ~ .. .;, nl ;. n A 500 rnl ~oly.... -. ;,.~l ;.. .kettle was loaded with 100 rnl of Freon~l9 113 and 20 ml of ~0.05 M in Pl.e.lu~yP in Freon~ E1. The l~ettle was flushed 10 X with argon, twice l,vith 5 psi of TFE, and then fed with l ~ from a one liter ~ yli~lle~
. ri~ins~lly c.~ ;..;..g -33 g of TFE. The r ,~;liùm ,..;,~ JIl~f ~ from 20 to 59~C going through a .. -~ ;........ ~ ,Ul~ of 60 psi finally ending up at 31 psi at 20 28~C ~4 hours later. Solid polymer was isolated by filtr~ n, washed 3 X with Freon~l9 113~ and dried, 15.42 g. At 372~C with a 15 kg weight in a melt inde~c no flow was obse~ed.
F'X~l~Pl,F', 49 ~ o..i~ Batch ~ t;~n 1~3P in He~ane A. Usir~ ~ h~ c h~co- A beaker was loaded with 27.8 ml of 40% aqueous tetra-n-buty1,.. ~.. i.. l~d~v~dG (40 mmoles), 4.1 ml of 30% l~yd~o~ pero~ide (40 mm~ s~, and 100 rnl he~anc with chiUing and stirring- After adding 3.15 ml of i~ouulylyl rhlor~ (30 .. ~-lf s) thc m~ture was ~1~."q~" ic ,t~ t1 at ... ~ .. power for 30 3eco.--1c using an ..ll~qo~ic horn 30 c.. ~.. r~.t~ ~I to a 40 Khz, 150 watt Dukane power source. The organic layer was sc~ 1 and washed twice with 50 ml of 5% ~l~ ~O~ NaHC03 giving 95 ml of organic phase that was found to be 0.074 M in pero~ide by i~kln.~ titrs~tion foran overall yield of 47% based on starting ~sobulylyl chloride.
B. U~ y&v~ as b~ce: Dissolve 3.96 g of 85% KOH
(60 mm(-~ As) in 6.13 ml of 30% hyvlvO_.l pero~ide (60 mm-l~s) with ice bath cooling in a beaker. Add 90 ml of he~ane and 3.15 ml of lsobuly,yl ~hlo i~,o (30 ~ ). An n1t~cc-nic hom c~ l to a 40 Khz, 150 watt Dulcane power source was started up at ~75% of . . .~ power and lowered into the ,er mi~cture. After 1 rminute, ultrasrni~tion was stopped. The organic layer s~ t d .l;, t ly and was washed twice with 50 mI of 5% aqueous NaHC03. The organic layer had a total volume of 90 ml and was found to bc 0.049 M in IBP by i{~,df mPtric titr~tion for an overaU yield of 29% based on star~ing iso~ yl~l 5 f h1~rifl.o.
~X~l~IPT,h'. 50 ~tr~o~ic Batch P~ n 5PDC in Freon~D E1 A beaker was loaded with 0.92 g of lithium hy~h~ide ~ nolly~ t., (22mm~1es)and0.05g.~ .. ~-~;.. -.pPrflllnroo~ .. o~ issol~lcdinlomlof 10 water. Once solution was c- mplete and the mi~ture had been chilled with an ice b~~ ~~ ydr~ge~ 2 ~ h.~3, ,~-r,~c,f-~
2.5 rnl of 94% pure 5Cl (~13 m~noles) were added. An nltr~ ni~: hom c~ t~ ~1to a 40 Khz, 150 watt Dukane power source was started up at--75% of . . ~h ~
power and lowered into the reaction mi2~ture. After 25 seconds, ulL.A~ n was stoppe~l The organic layer was washed twice with 50 ml of 5% aqueous NaHC03. The organic layer had a total volume of 73 rnl and was found to be 0.034 M in SPDC by iod- m-otri~ dliOll for an overaU yield of 38% based on starting SCl.
~,XAl~IPI ~ 51 ContimlQus Reactor ~aldtion of EtHDC in Freon~ El ~ ing T FoUowed by Ultr~)nit~ er Syringe pump (3), loaded with 15% by weight ~l"- ~--5 H2Oz, was started up at 0.659 îr~min. Syringe pump (2), loaded with 24% by weight &~ u~ KOH, was started up at 0.744 ml/min. Syringe pump (1), loaded with 0.255 M
2-ethylhe~yl chlc~.uÇvl.l.. ,t~, in Freon0 El, was started up at 3.396 ml/min. The streams from the three pumps, held at 20~C by a ~>UlI.~JUll~ 3 water bath, were joined into a single stream at a l/8th" union cross Hoke~l9 fitting (4). The ratio of .~ - at this point was 3.5 moles ~f H2~2: 4-5 moles of KOH: 1 mole of 2 CIIIY~ Y1 ~h1G.~r~ e7 making for an 7 fold e~cess ~f H2~2 and a 4.5 fold 30 e~cess of KOHover 2 elhyll~ yl cl~lo.~r~ intenns of r~ ti~r~ U;~.2~ y~
T~ , after e~iting union cross (4), the liquid stream was run via an 0.035"LD. line of ~0.1 ml volume into the bottom cup of the 1.6 ml ultrasonic reactor cavity (S) also chilled with wet ice. The power source to the 3/8" <liz-m~t~r nll.~e~ hom (6) was ~ned on at 50% of full scale, providing ~15 watts of 35 power to the ultrasonic cavity. Product e~ited as a stream at the top of the cavity at 22~C with an average ,~ e time in the cavity- of 20 ses~n~ The reactor was run for several ...;.~ ,s to flush out the lines and achieve steady ~ lg co.-.l;l;o~-.s, this foreshot being Iun into waste poly W O 97/08142 PCT~US96/13976 c~ r (7). The çca*~ stream was then diverted to s~mplin~ c-,..l;.i.~f ,~ (8) already c~ ,g 100 ml of 2% a~l.,eous NaHCO3. The organic layer was d and washed twice with 75 ml of 5% ~ql~eou~ sodium lJ:c~L.u,lalc. This gave 38 rnl of 0.126 M ETHDC in Freon<B El for a 91.8% yield based on starting 2-eLllyl~ yl chl~J~fv~ e ar~d in a ~nr1i~te run 39 ml of 0.121 M E~HDC in Freon~9 E1 for a 90.4% yield.
ulp flow rates so as to obtain a 15 second r ~ rnre tirne gave a yield of 76% and il~ ,asillg flow rates still further to obtain a 10 second r~!Q;d~nre time gave a 66% yield of EI~HDC. ~ h~ g the ..~ C~ iC mi~er with a jet mi~er of 0.086 ~ volurne and lLU~llillg at 190 ml total flow/m~n (a nnminz~ e of 0.027 sec) reduced ETHDC yield still fur~er to 9.3%.
~,~PI ~. 52 Ultrasonic Batch ~ of t-BuLyl~ t, in He~ane A. lPc~ .. hydro~ide as b~c~o- A beaker was loaded with 1.23 g--85%
KOH pellets (19 m m-les) dissolved up to 15.9 ml with water, 5.4 ml of 70%
e~,--c t-butylllydl~p~ ide (40 mm--les), lOO n~ of he~ane, and 1.4 n~ of aoe~yl rhlon~l~o (20 mm~l ~s) with ice bath cooling. An ~ Ov~ ~ ~c hom c~-- ,,~r~ d to a 40 Khz, 150 watt Dukane power source was sta ted up at .~ ;.. power and lowered into the ~action ll~l~C. After 25 sccn---l~, ulhi~,~ - w~ st~ppe~
tne organic layer ~ t' CI and was,hed twice with 50 m1 of 5% aqueous NaHC03.
The organic layer had a total volume of 100 ml and was found to be 0.082 M in ~v~ic by ivf~ ' ;(.. . for an overall yield of 41% based on starting acetyl ~'1.1 ;.1.~., B. Usirtp 10% e~cess ~olq~ ;.lr. q~ I~qc~ ~qn~1 t~tra- l-bu~yl-25 ~....... n.. ~ r~h1or~tf~ qc phqf~ ~n~:f~r (~.~ IA77y;~l; A beaker was loaded with 1.23 g ~~5% KOH pellets (19 mmnl~s) and 0.1 g of tetra-n-butyl ,-~..,~I---;---.
h, dissolved up to 15.9 ml with water, 5.4 ml of 70% e~L~,~,u~
t-butyll"~v~lv~dc (40 .. ~l~s), 100 ml of he~ane, and 1.4 ml of acetyl chloride(20 mm~ ) with ice bath cooling. An ~ .. c.~ horn c....... ~r~l to a 40 Khz, 150 watt Dukane power source was started up at .~ ;.. power and lowered into the ~qc~ n mi~lwe. After 25 seco~ s~ ultrqconirqtion was stopped, the organic layer se~ d and washed twice with 50 nl of 5% aqueous Na'HC03.
The organic layer had a total volume of 100 ml and was found to be 0.11 M in ~v~de by iodometric titrqtion for an overall yield of 55% based on stalting acetyl 35 chloride.

W O 97/08142 PCT~US96/13976 li'X~ lPI,~. 53 UlLIasumc Batch l;~'e~c~ion ofl3cl~uyl ~hlnrirl~
A. Givin~ ylPeA~ Major F~v~u~l. A beaker was loaded with 1.23 g ~85% KOH pellets (19 mm~'-s) dissolved up to lS.9 ml with water, 2.04 ml of 30% ~ ou<: hydrogen ~rUAi~ (20 . -l-s), 50 ml of Freon~l9 El, and 2.32 ml of benzoyl ~hlnrirl~ (20 mmnl-~s) with ice bath cooling. An ~ horn ~_c~ r~l to a 40 Khz, 150 watt Dukane power source was started up at ... -~ ;.. - power and lowered into the reaction mi~SIlre After 25 sea~
AI ;~n was stopped, the solids f~tered off, washed with water, and sucked 10 d~y on the filter. This gave 1.52 g of white solid, a 63% yield of benzoyl p~lu~idc based on star~ing benzoyl chloride. The fltrate was brought to pH--1 by ehe 1iti~n of C~ t' A sulfilric acid and then ~ cl three times with methylene ~hl~rirlto. The three methylene ~ hlorirle e~tracts had a c~.h ~-~d volume of 80 ml, 10.0 m1 of which took 0.2 ml of 0.1 N thi--s -lf~t~ in i~l~ln....J. ;-- ~lVAi~
15 titr~tion The failure to find ~ -..o....'-; of pero~ide in the .l.~ c .hlori~ eAtractS ;...~ that little if any benzoyl hy-lL~vAidG
[C6H5(C=O)OOH] was formed during the N.__~ioll.
B. Givin~ Pc.l,~.~oic Acid (Benzoyl IIy~l.u~ l~ ~ r Product A beaker was loaded wi~ 50 ml of 30% hy~v~ l ~.VAi~, 3.42 ml of 20 45% ~ e~J--c KOH (40 m m~'~s), and 0.2 rnl of 40% ~ tsulyl .~
I,yd-vAide, the ...;,~I...c being made c~lltiollcly with ice bath cooling and stirring.
An ~ c~...;c horn cn...,~t~ d to a 40 Khz, 150 watt Dukane power source was started up at .. -~ ;................ .power and lowered into the lr~ Then 2.32 ml of benzoyl rhlnFi~l~ (20 .. nlc s) were added and after lS secn~ of 25 Illl-_~o~ it~ti~n the hom was tumed off. Trace solids, 0.08 g, were f~tered of~
The filtrate was ;~ t~ ly added c ~l-tic n~ly to an ice cold llUAIUUC of S ml of d sulfuric acid + 45 ml of water + 25 m1 of lll~lLyl~ c . I .lc.. ;~ The ,lLyl~ hlori~ layer was S~ t~ d and cn---h;--~l with two ~ se~lu~ 25 ml .-~tr -~tc. This gave 75 ml of ~~ Lyl~ e rhlori~l~ solution, 5.0 rnl of which took 9.1 ml of 0.1 N ~ .... l r~l~ in iodnm.otri~ l ;l . . l ;.~I-, a 34% yield based on starling benzoyl chloride ~c~.. i.. ~ the product to be [C6H5(C=O)OOH].
F'.~P~ li 54 Ulllc.s~)~-ic Batch P.~ ~al~lion of Trichlvlvace~yl Pero~ide A. P~ Hvdro~ as Base. A beaker was loaded with 1.98 g of ~ 35 KOH pellets (30 .. u... oles), 28 ml of water, 90 ml of Freon'l9 113, 3.07 ml of 30%
~-r e~ Ly~ug~ll perv~;ide, and finally 1.9 ml of trichlvlvace~yl f hlnri~
(17 mmnles) with ice bath cooling. An ulL,~cv.~~ horn c...~ t A to a 40 Khz, 150 watt Dukane power source was immto~ ly started up at 75% ..~

W O 97/08142 PCT~US96/13976 pOWOE andlowered into the .~c~;~.. n~lult;. After 15 scc.~ , the ulllasv,~ic hom was turned off and the reaction 11~lLLL~, poured into a prerhill~-l 250 nil r.~ r flask sitting on d~y ice. The lcaL;Iio~ was first allowed to freeze solid and then warmed with swirling. As soon as the Freon(l9 113 ph~e had5 thawed entirely to 87 ml of fluid, a 5.0 rnl sample was drawn and found to take 6.0 ml of 0.1 N sodium thi~ sl-lfat~ in iod~Tnrfric l ;l ~ ;~f;~ (yield 61% based on starting trichlc,~oa~ yl chloride). In a control ~ . . .l, the ~yll~SiS was c~-~f~ d e~cept that the trichluluace~yl f~h1on~1e was omitted from the r~
In this case pero~ide LiLIaliOll of the freshly thawed Freon~9 113 layer took but 10 0.1 ml of 0.1 N sodium thi-~s~llf~fe It should be noted that trichl~ ,ac~lyl ~Ai~ is :,u rr;~: ~ .lly u~ e that a typical work up -- s~ the organic phase and wa~l~lg twice wif~ 5% a~lu~,ou~ sodium l,;~aLL...l.alc gave but a 9%
yield in an ~JlL~,~Wi;~e i~ ;f~1 :!iyllllL.,i~i.~.
The method d~ s. . ;l~cl here avoids the ~.coll~ ly low t~ S, 15 1 hour r~a~n times, and l~ .k~ ti~n ~ ues of prior work with trichlOrv~c~.yl pe v~ide (See US Patent 2,580,373) s~ likely to apply to vther ~ h~ e p u~ides. Being able to make a highly ~~~,~ 1P ~c~v~ide and deliver it to a reactor all in a matter of sec~ n~l~ makes the use of such ~lvA-clc s much more ~l~L.cl on an ;...~ . ;Al scale. The reported half-life of trichlolva~ly 20 perv~ide is ~0.6 hr at 2~C.
B. T ithi-lm L~ Base. A heaker was loaded with 1.25 g of lithium l~ydLvAide l~lv~ohy~' - (30 _1 - ~), 28 ml of water, 90 ml of Freon<l9 113, 3.07 ml of 30% a~l eou~ hyL~lOg~ l p~nl~le, and finally 1.9 ml of trichlvr~ ~ yl rhlorille (17 mm<~les) with ice bath cooling. An ulllai~Ol~ic horn c.. ~ l to a 40 Khz, 150 watt Dukane power source was ;........ ~.J;, h .IY started up at 75% .. -~ ;.. ..power and lowered into the re~çtion ........ li~luL~. After 15 sec~
the ~-11~ ~s~-;c horn was turned off and the reaction l-l,~ c poured into a preçhill~l 250 rr~ r.l.... - ~el flask sit~ing on dry ice. The ~ ,lion .. ~lu.. , was first allowed to freeze solid and then wamlc~d with swirling. As soon as the Freon~l9 113 phase 30 had thawed entirely to 88 rnl of fluid, a 5.0 ml sample was drawn and found to take 6.2 rnl of 0.1 N sodium thiosulfate in io~lo...~ l. ;c tiLIaliull (yield 64% based on starting trichlv.vacelyl cl.l~,;~).
J~X~l~PI ~. 5~;
Ultrasonic Batch Pl~alalioll of SP
A beaker was loaded with 2.5 g of lithium hydlvl~ide lllvllOhy~Lal~
(60 '--), 28 ml of water, 0.1 g of ~mm~ninm ~,lnuolvo~ ~,o;-l~, and 90 ml of Freon~l9 El with ice bath cooling. Once solution was cvll~l~ , 6.13 ml of 30%
Ly~Lvg~ ~lvl~idc (60 mmcl -s) were added with stirring. The sol..~ was çhilled W O 97/08142 PCT~US96/13976 again to ~0~C and then 3.S rnl of 5COF (18.7 mmoles) were added. An ultrasonic hom c. nn-cted to a 40 Khz, 150 watt Dukane power source was started up at ~50% of .. ,.~ .. power and lowered into the rç~-ti~n n-u~ . After 15 seco- .-1~, ultr~c.nir~tion was stopped, the ? ., ~ ~ui~ phase frozen on dry ice, and 5 then 5.0 rnl of the organic layer withdrawn for ~ u~de titr~tion The organic layer was found to be 0.01 M in 5P for an yield of ~10%.
~X~l\IPT ~ 56 Ultrasonic Batch F'~ ;on of CF3CF2CF20CF(CF3) (C=O)OOC(CH3)3 F~mple of mi~ed fluorocarbon/hydrocarbon ~LIu~;lul~
A bea~er was loaded with 1.23 g ~u~ ;u.~ hydro~ide pdlets (19 mrnoles), 15 mL water, 100 ml Freon~l9 113, 0.1 g of t~ al~ulyl~ perchlorate, and 5.4 rnL of 70% aqueous t-butylhy~u~elu~ide with ice bath cooling. With c.. ~;.. e~t ice bath cooling, 4.0 rnL of HFPOCOF (20 .. -lles) were added. An ulL.~.s.,~uc horn cu. ~ -t~ cl to a 40 khz, 150 watt Dukane power source was ;~ r .l ;~ ly started up at 75% .. u. ~ ;.. power and lowered into the reactionrni~ture. After 25 seconds the hom was tumed off. The organic phase was 5C~ ;,t~,.1 and washed t~,vice with 50 mL of 5% ~ ous sodiurn l~icaul~OI~. Thisgave 100mLoforganicphase,5.0rnLofwhichtook7.2rnLof0.1 Nlt.io;,..l r~t~
in i- t- m~tric titr~ti~ n for pero~cide (36% yield). This pero~ide was found to have 20 a half-life of ~5 hours al: 0~C.
T~X~l~PT li. S7 Ull-~ ....;c Batch P~ aLion of Dichluru~etyl Pero~ide A beaker was loaded with 1.25 g of lithiurn hy~llu~ide l~ollohyJl~Lt;
(30 mm~ s)~ 28 mL of water, 90 mL of Freon~ 1 13, 3.07 mL of 30% aqueous hy~Lvgen pero~cide (30 .~ lrs)~ and finaUy 1.64 rnL of dichlorva~e-yl chloride (17 mm~les) with ice bath cooling An ~~ll . ,.co~ hom co~ t~ ~l to a 40 Khz, 150 watt Dukane power source was ;............ I;; lely started up at 75% m;~power and lowered into the l~;a~;Liull mi~ture After 15 sec~nflc~ the ultrasonichom was tumed off and the reaction mi~ture poured into a pre~hill-~l 250 mL
30 ~ flask sitting on dry ice. The reaction mi~ture was first aUowed to free~ solid and then w~llled with swirling. As soon as the Freon~l9 113 phase has thawed entirely to 90 mL of fluid, a 5.0 mL sample was drawn and found to take 4.5 mL of 0.1 N sodiurn thiosulfate in iodometric titration (yield 48% based on starting dichluluace yl chloride).

W O 97/08142 PCT~US96/13976 F,~ ~nJP7.~.~8 U1L~SO1IiC Batch Prep. of HFPOdP
F~z~mpl~ of low water content r~z/ction ~ ule r . _. . .pl~ of work-up not reqniring phase .,~-S A beaker loaded with 78 mL Freon~D E1, 1.0 mL 30% hy~31vge.l peL~
(10 mm~les)~ and 2.1 mL of HFPOCOF (10 mm(~l ) was chilled in an ice bath.
Finally 0.86 mL of 45 wt % ~ue~u~; KOH was added (10 mm- ~s). An ~ .11. ~c. .-~ ;c horn co...-~c~ l to a 40 khz, 150 watt Dukane power source was ;....~ t,;ly started up at .. -~ ;.. - power and lowered into the reaction mi~ture. After 10 25 secc-n~ls the horn was turned of~ This gave a white, opaque organic phase with a few dn,~lt,t~ of water visibly fl~ting up top. Using an adlliti~n~l 10 mL of Freon~ E1, the reaction mi~ture was washed through a ~ ogr~phy column ~c,cl.a~g~,d with 50 g of D.i~.it~,0 (a~Uydlvu~ CaS04 peUets) and 25 mL of Freon~19El. Aperiodof2to3 ...;....1. swasaUowedforthereacton~ ulcto 15 wash through the D~ c~l9. This gave 100 mL of slightly hazy Freon~ El s~ -, 5.0 mL of which took 3.5 mL of 0.1 N thi~ lf~tr in iocl~. ~ .. h ;~' pero~ide liL.aliOll (70% yield based on stamng HFPOCOF). Even after wasl;u.g this organicphase twice with 50 mL of 5% ~ln~ons sodium bic&LL.~,llalc, it stiU took 3.5 mL of 0.1 N thi~sll1f~te in pero~ide ~ ~at ....~ . t~ cl or e~ccess hy~:Lug~,.
20 pero~ide did not make it past the Dli~ c0.
~ many other G ~ ~plf ~ herein the ~ u~ ~l is worked up by ~ y s -r ,, the water phase from the organic phase that c~ the pero~ide. The organic layer that has been ~ 1 can then be washed and, if llly sQII~l ;n~ is n~ e~l~ passed through a bed of ~lc~: c~..I r- ,....pl~ 58 here - s an ~ work-up giving d~y ul~Lialv~ snllltinn free of hy~llug_.l ~o~idc c~...l-...i.-; ~1 if the original pero~ide ~ylllhcsis is run with ...;..;...~1 water as in the e~ here, the ~r.~ 1 separation and w~l~ g steps can be omitted entirely, the l~Uvll mi~ture going instead directly through a ~ ssi~nt bed. Thiscould be of particular value in col~uLul~ small, pv~ ~,v~dc &~
30 units.

W O 97/08142 PCT~US96/13976 ~,~l~IPI,~. 5g IntrQsonicBatch E~v~ t;-n of { CF3CF2CF2OtCF(CF3)CF2O]6.3CF(CF3)(c=O) } -2 ~Qrnrle of high MW perv~cide r.;.. ,l,lF of low water content r~Q~tion ~
F.~Qmrl~ of work-up not 1~l..;....g ph_se sep~ ;o~-F~Qmple using an oli~olll_ ic rni~
A beaker loaded with 78 mL Freon~l9 E1, 2.04 mL 30% I-y~l~vg~ v.
(20 mmf les) _nd 8.5 rnL CF3C2CF20tCF(CF3)CF20]6 3CF(CF3)(C=O)F
(~10 mmnl~s~ 1300) was chilled in an ice bath. FinQlly 1.00 rnL of 45 wt % ~u~,vus KOH (12 mmnles) was added. An nlf ~v.~;c horn c-... F~,t~ cl to a 40 khz, 150 watt Dukane power source was ;.~ ly started up at ~
power and lowered into the ~~ Urt. After 15 sec~ S the horn was tumed off. This gave a viscous white Pmnl~i- n Washing the ~mlllci- n through 50 g of Drierite~l9 (~Lyvlvlls CaS04 pellets) in a CL~ Y column with an 2.-1~1;1 i. .. ,~l 25 mL of Freon~19 El gave 80 mL of hazy Freon~ El sol~lti.~n, s~o mL of whichtook3.9mLof0.1 Nthif~slllf~f~inio~l.. I.;- ~IVAid~ , a62%
yield of {CF3CF2CF20tCF(CF3)CF20]6 3CF(CF3)(C=O)~-2. This p,vced..,c keeps the volume of wal:er smaU so f~at passage through a drying agent is 1~
20 and avoids a layer sepQrQtion step that would be made .l;rr;~ by the ~lll~i/~n (see below).
As an ~mple of how ~liffi~.lt layer s~ .c can bec~ne when rnaking mf lPclllQr weight pero~ides, a beaker was loaded with 3.01 g KOH pellets dissvl~_d in 10 mL water, 78 mL Freon~19 El, and 4.70 mL 30% lly~ll'oge.
25 ~v, idc with ice bath cooling. After adding 8.5 mL of '~F3CF2CF20tCF(CF3)CF20]6 3C(CF3)(C=O)F, the reaction l,~ u.e was ulh at ..... --~ ;.. - power with the same 150 watt hom as above for 25 s~r-....L~ This gave a viscous, milky-white,-single-phase eml~ n F~_~lg this ~mll~ n on dry ice gave ~el~ O a bit of a layer se~ i.... Bringing oh~ly 30 acidic with sulfuric acid ~.ovuc~d a rnilky upper layer, a hazy rniddle layer, and a ge,l~;ll~,uO looking lower layer. P~O. . " ,~1 .ly soaps such as CF3CF2CF2O[CF(CF3)CF2O]6 3CF(CF3)(C=O)0-K+ are formed as by-products of the ~.v~de ~y~llhesis and may become an ~u~l~,as.llg problem with i~ ,asillg -~ - lQr weight.

~X ~ M P~,~,60 Benzoyl E'e~Aidc a. Jet + Ultrasnn~ inF at 0~C. The thrce ~ required to make benzoyl pero~ide were brought together under c~n-7iti--n.~ of intense mi~ing in a 5 flow system. See Figure 3.
Three ISCO syringe pumps were simlllt~n~ously started up, the first (20) d~ L..lg 40 ml/ nin of benzoyl ~hlc~ris~e, the second (22) dcL~_Li..g 148 rnl/lr~in of 15% aqueous hy~llug~n~LOAlde~ and the third (24) delivering 162 mlJinin of 20% ~ eolls KOH. ~ terms of the b~l~nl~e~l eq~l~*~.n for making benzoyl 10 pero~ide, this put H2~2 in a 4X stoi~ l .;n..~h;~. e~ccess and KOH in a 2X
rt~ ' - e~ccess over benzoyl ~hlot;~l~ The KOH and H2~2 streams were Cul~ ed in a 27 e~ 1 Kenics mi~er (intcrnal volume 1.4 ml, 3/16" OD X 7.5"
long) ;~ t~ ly before being run straight through a 1/8" Hoke~l9 T fitting. In the c~e of this aqueous strearn flow rates were high enough that no orifice was needed, the 0.08" inside .a;~ . .r of the 1/8" ~ steel tubing lU~ g to the Hokc~9 fitting being ~ e The benzoyl ~1.1." ;.1,~ stream was brought in the side arm of the Hoke~D T fitting using an 0.02" orifice to ;. .~ .~s. ~c its velocity.
Fluid ~ S on the inlet side of the Hoke~l9 T fitting were 38 psi. The high velocity benzoyl chlori~e strearn i...l,;..~;..g into the S-r ~vus stream created intense 20 ~ called jet mi~cing (26) within the Hoke/~9 T fit~ing.
T.... ~ ; .t~ ly upon leaving the Jet rmi~er/Hoke(~9 firdng (26), thc ~
ture was ;,,II)jc~,t~ d to an ~ ;l ;"~ -1 about 13 sec~n~1s of intcnse s.~ in order to in~l~ase re~-~tion tirne, re~ nce tirne in the jet sni~er/Hoke~l9 fit~ing bcing only 0.01-0.02 sec. The l.,..,.~r~ .1in~ COI~ CI ;- ~~ the jet rni~er/Hoke~9 fitting to the 25 product coll~-ction vessel was çis- ~ .1 ;AIly wl~ed around an ~lltr~sr~nic hom (30) as ~1~ 3~ ~d 'oelow.
The jet ~ /IIoke~9 fitting (26) was ;.. -~ d in a 0~C water bath along with a 20 KHz lkw ultrasonic hom (30) ~ ~-~ 1" in ~1;~.,.. h by 2~" long. A
190" length of 0.1875" ID Teflon~l9 PFA tubing c~ tne jet mi~cr/Hoke'l9 30 fitting to the product c~ ctif-n flask w~ first p~sed about 1/2" below the tip of the ~-11 . ~~...;c hom, then sri-~l. cl for 166" of its leng~ around the hom, and finally, for its last 24", run out to the product collection flask. At tne flow rate in this the l~ e time in the jet rni~er + the 190" of ~ f~ . lin~ was 15 .3~c~ lc. The ,.,...~r~ ~lin-o was run to an F.1~ el flask c~,..l~;..;u~ 200 ml of 35 ice cold toluene ~ .olv~l, the ben7oyl pero~ide. Optionally, the toluene can be omittcd and the ~.~u~l pero~ide filtered off as solid. Product pulsed out of ther. .1;... as roughly inch long cylinders of self ~1h.~nt white paste ~:lh . ",.I ;..~
with fluid. As the lx,.lLuyl pero~ide pulsed out the end of the ~ r~ ~lin~ .~e y following the jet miAer sirnilarly pulsed from about 10-35 psi. Product was collected for 33 sec~ lx An ~ 50 ml of toluene was added to the product collection flask 1~- ;- .~; ~ ~~, all the benzoyl peroAide into snlntinn~ Washing the toluene 2X with 150 m1 of 10% a~ucous KOH gave 272 ml of toluene solution 0.354 M in benzoyl ~l-JAide by iorl- mr.triC titration (102% yield).
b. Jet ~i~ one. The e~ was set up and r~n much as in part (a) above eAcept flow rates and water bath l~ alu~s were varied. More irnportantly the ultrasonic hom was never turned on. Results are i,.. ._. ;,~ .1 in the table below.
Water Bath R-~s;f~ e Tr~ l .. e T rne Yield ~.,.. ,............................. ~rt 0~C 15 sec 54% Solid in oil 0~C 11 sec 49% Solidinoil 26~C 15 sec 66% Solid Paste, pulsed 46~C 15 sec 78% Solid Paste, pulsed Using an optically clear ,.,...~r~ .1;.., made of Teflon~9 PFA, visual ;...~pc~
showed that much of the benzoyl peroAide had formed as a solid by the time that the ~i~liu~ Al~llc e~i~ted the Hoke~9 LlLing/jcl miAer. Minim~l ~*~l ;. . .;, .l ;. .. ~
15 il~,v~,~d yields c~ ly. illcleas~g bath t~ p. .~ c to 26~C and then to 46~C .ll."~,ased yields to 66% and 78% l~ .,ly.
F.~Al~IPI F. 61 Cnntin.l-)us Reactor p~p-- ~ of HFPO Dimer Pe. uAi~l~ in Freon~ E2 Use of Gear Pumps Instead of Syringe Pumps ~ the c~ o~ eA~ll~l~s above syringe pumps were used to deliver the IA.~I X The eA~ elll here shows that the use of gear pumps gives a similar result. Using the same d~aLalus and set up as ~ l for E~ample 1 lB all three syringe pumps were rep~aced by gear purnps ..._.... r~ rl by Micro Pump. For the solution of HFPO dimer acid flllori~e in Freon(l9 E2 the pump was e~lui~d 25 with a ~185-000 head while for tne KOH and H2~2 solntinnx both pulnps were c~lui~d with #187-000 heads. All three gear pumps used a Model ~7144-00 drive motor from Cole Parrner 1 U111~-1I C<Jl~ ly as controllers. Running much as d~r,~il>ed for E~ample 1 lB gave an 84.3% yield of HFPO dimer peroAide. The run is d~ e~ in greater detail below.
- 30 The same e~ D~~ was used as in EAample 1 eAcept that the syringe pumps were replaced by gear pumps, the H2~2 and KOH streams were cn...h,;..~
prior to rni~ing with the organic stream, union cross (4) and ulL-~ollic reactor (~;) have been replaced by a jet rniAer, and the 0~C ice bath has been l~laced by a 26~C water bath. The jet mi~er was a 1/8" Hoke~9 T with an internal .~ of 0.094" and an internal length of 0.76", making for an intemal volume of 0.086 rnl.
In these runs the organic phase was pumped straight through the lt8" Hoke'l9 T at 125 mVminute. I'he c~,l..bi,~d ~r7eo~c KOH/H202 phase was pumped into the 1/8" Hoke~ T at 84 ml/min via an 0.038" I.D. tube set 90~ to the organic flow.
R~h-ring the ~ t. . of the tubing ~ --t - ;--~ the side inlet of the T to 0.038"provides the orifice ~im~n ci~ n.c required for jet rmi~ing at the flow rates given above and for the v; ,co ~ s/~ s of the fluids.
Gear pump (3), loaded with 12.9% by weight ~ ou~ H2O2, was started up at 31 ml/min. Gear pump (2), loaded with 18.3% by weight flqlleOll'C KOH, wasstarted up at 53 ml/min. Using a 27 ~ Kenics static mi~er, the KOH and H2~2 streams were co,nL,.--~,d to a single a~ ,ous stream flowing at 84 mlknin.
Gear pump (1), loaded with 5% by weight HPPO dimer acid n,.... ;flC in Freon~9 E2, was startcd up at 125 ml/min. Using 0.038" I.D. tubing, the ~ ~ .cd a~cous stream was passed into the side arrn of the Hoke~19 T at 85.0 mlhnin g;~2 into the organic stream moving straight through the Hoke~l9 T at 125 ml/min. The ratio of ~ at this point was 4 moles Of H2~2: 6.5 moles of KOH: 1 mole of HFPOCOF, making for a large e~cess of both H2~2 and KOH
over HFPOCOF in terms of ~ st i~hi~m~ty, The liquid stream c~iting the Hokeq9 T at 28~C was run via an 0.085" I.D. line of ~3.3 rnl volume to the c~ll.~ctinn bottle. The 0.085" .1; ~ of the e~it line is such that the tnrb~ ontflow may persist after the jet mi~er, ~lth- ugh this was not c~ r.. t~ by Avcrage reci~1~n~e time in the reactor was 0.025 s~ c c.-n~ nng just thc volume of the Hoke'l9 T serving as the jet rni~cr or 0.97 sec~
25 c~ le- ;~8 the volume of the jet mi~er and the e~it lines togr,th~r. The Cu~ u~uS
reactor was run for 2 ~ t~ S to flush out the lines and achieve steady ~E~
c....~l;l;-...~, this roL~ ol being run into waste poly c~ - r (7). The reactantstrearn was then diverted to s~mrling c~ c. (8) where 123 ml of organic phase were coll~~-~e~i over the ne~t rninute for ~ oses of product analysis. The organic 30 laycr was s~ r~ and washed twice with 75 rnl of 5% z l~Jev~ sodium bica.l,o..dle. Iod~m~trit~ titr~*on found 0.102 M HE7POdP in Freon~19 E2 for a yield of 84.3% based on starting HFPOCOF.
~rPl lh 62 ~..ntinllous ~eactor P1~ ~Q ~ ~1 ion of HFPO Dimer Pero~ide in Freon~!9 E2 Jet Mi~er The samc e~ - ~ ~l was used as in F~mrle 1 lB e~cept that all flow rates were ;..- ~c~e~l and the I ~ ~ r~ after the jet m~er was r l~n~rl from a tube ~3' long by 0.085" in inside d;~ b r to a tube 22.5" lon~ by 0.155" in intemal W O 97/08142 PCT~US96/13976 . . The jet mi~er was as before a 1/8" Hoke~9 T with an internal ~ . of 0.094" and an internal length of 0.76", making for an internal volume of 0.086 ml.
In these runs the organic phase was pumped straight through the 1/8" Hoke0 T at 250.08 mVminute. The cn...h;..~ aqueous KOH/H202 phasc was purnped into the 1/8" Hoke~l9 T at 168.48 mlhnin via an 0.044" I.D. tube set 90~ to the organic flow.
Rr~l. .. ;. ~g the ~ of the tubing ~ . ;. .g the side inlet of the T to 0.038"
provides the orifice ~limrn~ions required for jet mi~ing at the flow rates givenabove and for the v i~co~ s of the fluids.
Syringe pump (3), loaded with 12.9% by weight aqllroll~ H2~27 was started up at 62.3 ~/min. Syringe pump (2), loaded with 18.3% by weight ~u~ us KOH, was started up at 106 mV~nin. Using a 27 rl~ -1 Kenics static mi~er, the KOH and H2~2 streams were co~ cd to a single ~uCOuS stream flowing at 168 mlJ~nin. Syringe pump (1), loaded with 5% by weight HFPO dimer acid n... ;de in Freon0 E2, was started up at 250 ml/min. Using 0.044" I.D.
15 tubing, the cvlllL..~.cd ~ .,ev~ stream was passed into the side arm of the Hoke~9 T
at 168.48 mlJmin impinging into the organic stream moving straight through the Hoke~ T ae 250.08 mlhnin. The ratio of 1~ n...-~; at this point was 4 moles of H2~2: 6.5 moles of KOH: 1 mole of HFPOCOF, making for a large e~cess of both H2~2 and KOH o~rer HFPOCOF in terms of l~a~ .;v... l.y. Fluid 20 yl~ ulc just before the jet mi~er was 7.4 psi and 1.8 psi just after. The liquid stream e~iting the Hoke< 9 T at--30~C was run via an 0.155" I.D. line of ~7 ml volume to the c. llectinn bottle. The 0.155" ~ of thc a~it line is such thatthe turbulent flow may persist after the jet mi~ter, ~lthml~h this was not col-l;....~d by ~ l Average . ~ ~;<k . .. c time in the reactor w~ 0.012 sec~
25 co..~ ~ ;..g just the volume of the Hoke~l9 T serving as the jet mi~er or 1.02 s~ c.~. ,.1~ C~JI ..~;.1~ ~ ;- .g the volume of the jet rni~er and the e~it lines tc,~lh~
The c~ ntimlous reactor was run for 1 rninute to flush out the lines and achievesteady ~Jy- -~ g con-litinn~, this r~,l.,.,Lo~ being run into waste poly c.,..l;.;... - (7).
The ~ ..l stream w~ then diverted to s~-..l.l;..~ c~...~ .;.- r (8) where ~123 ml of 30 organic phase were cnll~cte~l over 30 secn.~ for y~ ose~s of yroduct ~u~aly~is.
The organic layer was S~ ~t~ -1 and washed twice with 75 rnl of 5% ~ c~
sodium bic~l~ullale. The organic phase titrated 0.115 M in HFPOdP in Freon~9 E2 for a yield of 94% based on starting HFPOCOF.
Mal~ng ~123 ml of 0.1 lS M HFPOdP in a volume of 0.086 rnl in 30 s~co.. -l~ c~ yol~ds to a productivity of 108,000 lbs of HFPOdP/gallorv~r.
The fastest co,..~ lr prior art we are aware of, U.S. 2,792,423, reports ma'King 7.2 g of 4P/100 mllmin for a produ~;livily of 36 lbs of 4P/gallon/hr. Putting this in the 1A ~ ~ of ~ll;. I,,,,. .~l 3, a productivity of 36 lbs of 4P/gal/hr is equal to ~1200 g of active o~ygen content/gaVbr and a productivity of 108,000 lbs of HFPOdP/gal/hr is equal to 2,380,000 g of active o~ygen content/gaVhr.
Ull.lc~ .,l ;. . .~fing our produ~-LviLy by treating both the jet mi~er and the transfer line as the reactor, we col~3e~vuLivclye~ AIe this ~ lr makes HFPOdP at >1300 LBS/ga11onfl~r and equivalent active o~ygen content at >29,000 g/gal/hr.
h'~A~rPI,~. 63 ntinllt n~ Reactor P ~ ;tJ~ of HFPO Dilner Pero~idc in Freona9 E2 Mini Jet Mi~er The e~ dcs~;libcd in ~ A. . .lAc 62 mal~es enough HFPO ditner perw~ide to run a large cu... ~ cial-scalc poly.. ~ ;nn facility. For ~u,~.,~,s of lUlll~illg small ~ Lu~le p o~,es~~,s or of col-pling our pero~ide ~ r to a poly.. ;,s.l ;on pilot plant, the volume of product produced in E.sample 62 can be scaled down at le~t 20 to 200X.
Much the same e-l- ~ ;1.. . ~- -- ~I was used ~ in r~ 1c 1 lB e~cept that all flow 15 rates were de~ a~ed, orifice rli~m~.fe~ were fle~ eA in the jet mi~er, and the , . ,...~,r~. 1 ;.... foll.owing ~e jet mi~er was made much ~ ,1 in intemal ~1; ..f h--.
The jet mi~er w~ a 1/16" Holce~19 T with an intem~ of 0.047" and an internal length of 0.75". In these runs the orgatlic ph~e was ~u~l~d straight through the 1/16" Hoke'l9 T at 1.25 to 12.5 ml/minute using orifices from 0.01 to 0.035"infliAm~ter Theeo...... ~ sKO~VH202ph~ewas~u~l~cdinto the 1/16" Hoke0 T at 1.78 to 17.8 mlknin via an 0.044" I.D. tube set 90~ to the organic flow ~d using orifices 0.007 to 0.02" in ~ In aU runs here, the jetmi~er and its feed liIles were ;.. - ~~e~ in a 26~C water bath. A typicat run is J~ s~ 1 in detail below.
Syringe pump (3), loaded with 12.9% by weight aqueous H202, w~
started up at 0.388 ml/m~n. Syringe pump (2), loaded with 18.3% by weight A~lu<~ KOH, was started up at 0.663 n~min. Using a 27 r~ Kenics static mi~er, the KOH and H2~2 streams were CO~ d to a single aqueous stream flowing at 1.05 rnlknin. Syringe pump (1), loaded with 5% by weight HFPO dimer acid fhlnrifl~ in Freon~l9 E2, was started up at 1.565 ml/min. Using 0.J44" I.D.tubing, the c-,...l,:..P~ ~U~,OUS stream was passed into the side ann of the Hoke~19 T
at 1.05 ml/min via an orifice 0.01" in f~ t~ . . The aqueous stream then i~ ~S
into the organic strearn as the organic stream flows straight through the Hoke~9 T
at 1.565 mVmin, the organic stream having been injected into the Hoke~9 T via anorifice 0.02" in ~iAm~ter. The ratio of ~ at this point was 4 moles of H2~2:
6.5 moles of KOH: 1 mole of HFPOCOF, making for a large e~cess of both H2~2 and KOH over HFPOCOF in terms of .~ ;.t~ y. Fluid ~l~,i,;,~c;
before the jet m~er was--30 psi and ~25 psi just after. The li~uid stream was run W O 97/08142 PCT~US96/13976 via an 0.01" I.D. line 11" long to the product collection area. The 0.01" .li;.... ,t~
of the el~it line is such that the turbulent flow may persist after the jet rni~er, ~lth~ h this was not c.-l.r;....~ ~1 by ~ ..f ~I The c~ u~;, reactor was run for 20 rninutes to flush out the lines and achieve steady ~/~cld~i-~; c~..-.1;1;-~o.~, this S f~ ot being run into waste t~....l;.i.~f.l (7). The ~ ;t~ll stream was then diverted to s~mpling ci..~ (8) where 30 ml of organic phase were c~ ct~d over 20 minutes at al~ .lL 1~ . . . e for ~ oses of product audlysis~ The organ~clayer was s~ .t~ -1 and washed twice with 75 ~r~ of 5% aqueous sodium bicdllJ~ dlc. The organic phase titrated 0.083 M in HFPOdP in Freon~9 E2 for a 10 yield of 67.5% based on star~ing HFPOCOF. This yield is ~-~bdl~ly a lower limit le. ;..g that no effort was made to protect the product against thermal deco~ osilivn during the 20 minute coll~ction period. This is Run #4 in the table below.
DlMPNSION TN r~F~
FLOW RATP-C: ~1!~ .lM~u~I~E O~'P~: TRAN.'~:FP.R T .lNP.
RUN COFl KOH2 H2023 TOTAL AQ ORG. LENGTH LD. YlE~D
#1 12.5 5.3 3.11 21 0.02 0.035 44 0.035 75.8%
#2 6.25 2.65 1.55 10.5 0.01 0.02 16.5 0.02 76.1%
~3 3.13 1.325 0.775 5.25 0.01 0.02 11 0.01 6S.996 #4 1.565 0.663 0.388 2.63 0.01 0.02 11 0.01 67.S9~o #5 1.25 0.53 0.311 2.1 0.01 0.02 11 0.01 66.09 15% by weight ~O dime~ acid ~UOIide iU Freon~l9 E2 218.3% by weight aqueous KOH
15 312.9% by weight aqueous H2~2 h'XAl~lPI .~, 64 Continuous Reactor P~cp~ ;on of HFPO Dimer Pero~ide in Freon0 E2 Mini U1L~ niC Mi~er Three ISCO punnps were used to deliver the sarne le~t streams as in 20 F~mr1~ 63 above to a 39 inch length of l/8" I.D. by 3/l6" O.D. Teflon0 PFA
tubing. A wide ulLl~svnic horn with tip flim~n~ ns of l/4" by 3" was used to provide high iu~n~ y rni~ing within a 3" sc~ of the Teflon~l9 PFA tubing. This was ~comr1ichf~d by ;~ g the hom in a water bath~ placing a hemi-~;y~
cavity l/8" under tne hom. and l~ g the Teflon~l9 PFA tubing ~rough the 25 cavity. The cavity focuses the sound by reflection to i ~clcase the ulL.~svu--d il~t~.~iLy in the tube as it passes under the hom. This set up also avoids contact ,en the hom and collo~i~e sc-hlti- n~ and pemlits larger line ~ te~ than in W O 97/08142 PCT~US96/13976 n~ 63. Largc line ~ t . .~ are particularly i.ll~u.~ if solid ~lUdU~ 7 such as benzoyl pe:ro~ide a~e to be g~ t ,d Syringe pump (3), loaded with 12.9% by weight A~l, eou, H2~2~ was started up at 0.388 rnl/m~n. Syringe purnp (2), loaded with 18.3% by weight a~ ,uus KOH, was started up at 0.663 mlhnin Syringe pump (1), loaded with 5%
by weight HPPO dimer acid fllln~1fe in Freon~9 E2, was started up at 1.565 mUmin.
The two a lucuus streams were first co...~ d in a 27 ~]~ Kenics rni~er in a 3/16" O. D. by 7.5" length of Dl~ s;~ steel tubing. The c.. h;.. ~ l a.l~cv~C stream was then co- ~ .h;. .~d with the HFPO dimer acid flnnrifle stream in a 1/8" Hoke~2D T
fitting that emptircl into a 39" long piece of 1/8" I.D. by 3/16" O.D. Teflon(l9 PFA
tub~ng. The ratio of ~ n -l~ at this point was 4 moles ~f H2~2: 6.5 moles of KOH: I mole of HFPOCOF, making for a large e~cess of both H2~2 and KOH
over HFPOCOF in terms of reaction ;,loi~ ~.in.~.. h,~, The TeflonqD PFA tubing was passed through a focusing cavity under an l~ltr~nni~ hom dcli~l ...g ~101 watts with the tip of the hom and the focusing cavity ;.. ~ eAin a 26~C water bath. As the tubing passed under the hom it's c~ t~ went from clear to white and back again to two visible phases when the c~ t' ~~1~ came out from under the hom. Theliquid stream was lun via the l ~ ;. .;. ~ 30" of the Teflon~9 PFA tubing to theproduct coll~ctinn area. The cn~.l i.. ~c reactor was run for 20 .. ~;'''lt~ S to flush 20 out the lines and achieve steady O~l~thlg cnn-litinnc, this f~l~,sl-ot being run into waste c~...ln;.. ~ (7). The n,n~,l&ll stream was then di~,.hd to s~ .l;..g co..ln;..c.
($) where 30 ~ of organic phase were cnlle~ted over 20 ~;---it' ~ with ice bath cooling to 0~C for ,uu~osei~ of pLUdUCI analysis. The organic layer was se~ t~
and washed twice with 75 ~ of 5% ~ - o..c sodium l>ic~bu~ t~,. The organic phase titrated 0.102 M in HFPOdP in Frcon~ E2 for a yield of 85.3% based on starting HPPOCOF.
~.XAlUP~ l~. 65 C~ v~ ~ Reactor E~ . ~ ;.... of HPPO Dimer Fe~u~de in Freon'~9 E2 Mini UlL,~ol~ic Mi~er Cùl~ luuus Ph~e Sep~rs~tit~n Three ISCO pumps were used to deliver ~e same ~ .I streams as in r.~ .plr 63 above to a 39 inch length of 1/8" I.D. by 3/16" O.D. Teflon~9 PPA
tubing. A wide ultrasonic hom with tip .1;...~ ~.x;~nx of 1/4" by 3" was used toprovide high .ut~sily îni~ing within a 3" se~ l of the Teflon~l9 PFA tubing~ This was a~c~ by ;.... ~ the hom in a water bath, placing a hemi cylin-lrir~l 35 cavity 1/8" under the hom, and lUlll~illg the Teflon~9 PFA tubing ll~luu2 h the cavity. The cavity focuses the sound by lGllr~l;.... to ;~ G the ull~Asv~
~t~ y in the tube as it passes under the hom. This set up also a~oids contact h,t~._c,n the horn and COllu~ , Snlllti~n~ and permits larger line ~ . .,x than in W O 97/08142 PCTrUS96/13976 F.-5~ ,1f 63. Large line .1;~ . . are particularly iln~o~ if solid ~7l~ u~ i such as benzoyl ~e.~,~dc ar~ to be ~ t~ Cl Syringe pump (3), loaded with 12.9% by weight ~ s H202, was started up at 0.388 mllmin. Syringe pump (2), loaded with 18.3% by weight A~ eous KOH, was started up at 0.663 mlhnin. Syringe pump (1), loaded with 5%
by weight HFPO dimer aeid n.. ;.k in Freon~l9 E2, was started up at 1.565 ml/mirL
The three streams werc combilled using a 1/8" union cross fitting that ~ d into a 39" long piece of 1/8" I.D by 3/16" O.D. Teflon0 PFA tubing. The ratio of , at this point was 4 moles ~f H2~2: 6.5 moles of ROH: 1 mole of 10 HFPOCOF, making for a large e~cess of both H2~2 and KOH over HFPOCOF in terms of l~ ,t~ ,,,,. hy. The Teflon~19 PFA tubing was passed through a focusing eavity under _n ~ ~1 ~ ~ iC hom J~ lillg 110 watts with the tip of the hom and the r ,. u~ cavity i. ~ 1 in a 26~C water bath As the tubing passed under the hom its c~,. .t~ ; went from clear to white and baek again to two visible 15 phases when the c- ~-~t~ ; came out from under the hom. The liquid stream wasrun via the ~ ;U g 30" of Teflon~l9 PFA tubing to a glass s~,l ~ vessel, from which the lower organic layer CO~ 8 pero~ide was pur{~ped off at 1.5 mlknilL
and sent to the product coll~ction area. The reaetor and S~F were run c....l;...~ ly for 20 ~ t ~ to flush out the lines and achieve steady ~ 1;-.g 20 c-~-~-l;l ;-.c, the ~~ pA ~ ~ ~t.S~ organic phase being p ln~ed to waste c~ - d 5~ r (7). The ~p ~ organic phase was then di~,,~d to sA-~ g C~ - . (8) where 91 n~
of the ,,~ t~ d organic phase were c~ cterl over 60 . . .~ t_s with ehilling to 0~C
for ~ul~G~es of produc~: _nalysis. The organic layer was washed twice with 75 of 5% aqueous sodium bic~l,~.ale. The organic phase titrated 0.106 M in 25 HFPOdP in Freon~lD E2 for a yield of 87.4% based on star~ing HFPOCOF.
F.XAMP- ~. 66 ~ Batch P1~ ' ;on of tCF3CF20CF(CF3)(C=0)0]2 An ice chiUed beaker was loaded with 3.96 g of 85% KOH pellets (0.059 mole) di~solved in 28 ml of water, 70 ml of Freon~ El, 6.13 ml of 30%
30 ~uc~Ju;~hyd{'~g~ .ide (0.060 mole), and 44 g of 33.4 wt %
CF3CF2OCF(CF3)(C=O)F (-0.052 mole) dissolved in Freon~1D El. A I ;~
horn attached to a 40 Khz,150 watt Dukane ulll~vl.ic power source was ;... ~e.l ~u~,l.ly half vay down into the pool of l. a~ . The ultrasonic source was turned on for 30secollds at full power. The layers were then s~alAt~ d and - 35 the organic layer washed twice with 50 rnl of 5% aqueous sodiurn l);c~hlJolldtc.
This gavc 90 ml of 0.11 M pero~ide in Freon~9 El for a 38% yield of tCF3CF20CF(CF3)(C=0)0]2--W O 97/08142 PCT~US96/13976 FX~ M P~,~,67 .,..;c Batch ~l~p,~ on of tCF3OCF(CF3)(C=O)O]2 An ice chilled beaker was loaded with 3 96 g of 85% KOH pellets (0.059 mole) dissol~l~,d in 28 ml of water, 70 ml of Freon~ 1, 6.13 ml of 30%
5lq~l~ollC hydlo~ pero~ide (0.060 mole~, and 42.5 g of 26.3 wt %
CF30CF(CF3)(C=O)F (~0.048 mole) dissolved in Freon~ El. A ~ .. horn ~lsarh~-~l to a40Khz, l50wattDukane ~ c~ ;cpowersourcewas ;....~ c~l roughly halfway down into the pool of l~e~ The ~ source was ~rned on for 30 seconds at full power. The layers were then s~ d and the organic 10 layer washed twice with 50 ml of 5% ~u~ u,, sodium l~ic~l,u.,~te. This gave 92 ml of 0.10 M ~. ~idc in Freon0 El for a 38% yield of [CF30CF(CF3)(C=0)O]2-

Claims (57)

What is claimed is:

1. In a continuous process for producing acyl peroxides wherein the reaction occurs in a reaction vessel which is connected to a transfer line, comprising the steps of contacting an aqueous hydroxide and a peroxide with an acyl halide wherein the improvement consists of combining specific reaction vessel configurations with means for supplying vigorous agitation so as to decrease reaction time to less than one minute and wherein the acyl halide is selected from the group consisting of R(C=O)X and RO(C=O)X in which X is -Cl, -F, -Br and -I
and R is selected from the group consisting of (i) -CnFXClyHz, wherein x + y + z = 2n + 1, n is 1 to 8, linear or branched, and the carbon bearing -(C=O)X is preferably primary or secondary;
(ii) G(CF2)w[CF(CF3)CF2]x[OCF(CF3)CF2]y[OCF(CF3)]z-where w is 0 to 8 x is 0 or 1;
y is 0 to 7;
z is 0 to 1; and w +x+y+z ~ 1 and G is F or a substituted carbon group that is not highly reactive toward water, hydroxide, or hydrogen peroxide and having one or more functional groups selected from the group consisting of-F, -COOCH3, -SO2F, X -Br, -CFBrCF2Br (x = 0), -Cl, -I, -CN, -OC6F5; or (iii) an aromatic, hydro, chloro or perfluoro carbon compound having one or more functional groups selected from the group consisting of -F, -COOCH3, -SO2F, H, -Br, -CFBrCF2Br (x = 0), -Cl, -I, -CN, -OC6F5.

2. The process of Claim 1 wherein the means for supplying vigorous agitation is selected from the group consisting of jet, static, ultrasonic and mechanical mixers, arranged singly or in series.

3. The process of Claim 2 wherein the agitation provided by the means for supplying vigorous agitation is at least partially maintained by flow through both the reactor and the transfer line.

4. The process.of Claim 1 wherein the reactor and connected transfer line are closely coupled without dead spots so as to provide continuous agitation and reaction throughout both the reactor and transfer line.

5. The process of Claim 3 wherein agitation initiated by the mixers causes at least 10% of peroxide formation to occur in the transfer line.

6. The process of Claim 1 wherein the reaction time is 0.01 to 30 seconds.

7. The process of Claim 1 wherein the peroxide is selected from the group consisting of hydrogen peroxide, alkyl hydroperoxide and metal peroxide.

8. The process of Claim 7 wherein the peroxide is selected from the group consisting of H2O2 and t-BuOOH.

9. The process of Claim 1 conducted within a temperature range of -10°C to 40°C.

10. The process of Claim 1 conducted within a temperature range of 0°C to 25°C.

11. The process of Claim 1 conducted at ambient temperature and without temperature control.

12. The process of Claim 1 wherein X in the acyl halide is -Cl or -F.

13. The process of Claim 1 wherein the concentration of the aqueous hydroxide is greater than 0.1 molar in water.

14. The process of Claim 13 wherein the concentration of the aqueous hydroxide is 1 to 5 molar.

15. The process of Claim 14 wherein the aqueous hydroxide is selected from metal or tetralkyl ammonium hydroxide.

16. The process of Claim 15 wherein the aqueous hydroxide is selected from the group consisting of KOH, LiOH, NaOH, CsOH and R'4NOH wherein R' is CH3, C2H5-.

17. The process of Claim 1 wherein the aqueous hydroxide is R'4NOH
wherein R' is CH3CH2CH2CH2 and R is hydrocarbyl.

18. The process of Claim 1 conducted in the presence of a solvent.

19. The process of Claim 18 wherein the solvent is selected from a group of organic fluids characterized by their inertness toward the reactants and their ability to dissolve at least 1% of the peroxide by weight.

20. The process of Claim 19 wherein the solvent is selected from the group consisting of Freon~ E1, Freon~ E2, Freon ~ 113, fluorocarbons, chlorofluorocarbons, hydrofluorocarbons, hydrofluorochlorocarbons, hexane, cyclohexane and mineral spirits.

21. The process of Claim 1 conducted in the presence of a surfactant.

22. The process of Claim 21 wherein the surfactant is selected from the group consisting of fluorinated surfactants and hydrocarbon surfactants.

23. The process of Claim 21 wherein the surfactant is selected from the group consisting of ammonium perfluorooctanoate and sodium dodecyl sulfate.

24. In a batch process for producing acyl peroxides comprising the steps of contacting an aqueous hydroxide with a peroxide and an acyl halide wherein the improvement comprises continuously agitating the reactants, using agitation means, until substantial completion of the reaction, wherein the reaction is substantially complete in less than one minute and wherein the acyl halide isselected from the group consisting of R(C=O)X and RO(C=O)X in which X is -Cl, -F, -Br and -I and R is selected from the group consisting of CnFxClyHz, wherein x + y +z = 2n + 1, n is 1 to 8, linear or branched, and the carbon bearing -(C=O)X is preferably primary;
G(CF2)w[CF(CF3)CF2]x[OCF(CF3)CF2]y[OCF(CF3)]z-where w is 0 to 8 x is 0 or 1 y is 0 to 7 z is 0 to 1; and w +x+y+z ~ 1 and G is F or a substituted carbon group that is not highly reactive toward water, hydroxide, or hydrogen peroxide and having one or more functional groups selected from the group consisting of-F, -COOCH3, -SO2F, H, -Br, -CFBrCF2Br (x = 0), -Cl, -I, -CN, -OC6F5; or R is an aromatic, hydro, chloro or perfluoro carbon compound having one or more functional groups selected from the group consisting of -F, -COOCH3, -SO2F, H, -Br, -CFBrCF2Br (x = 0), -Cl, -I, -CN, -OC6F5.

25. The process of Claim 24 wherein the agitation means is selected from the group consisting of ultrasonic and mechanical mixers.

26. The process of Claim 24 wherein the reaction time is 0.01 to 30 seconds.

27. The process of Claim 24 wherein the peroxide is selected from the group consisting of hydrogen peroxide, alkyl hydroperoxide and metal peroxide.

28. The process of Claim 27 wherein the peroxide is selected from the group consisting of H2O2 and t-BuOOH.

29. The process of Claim 24 conducted within a temperature range of -10°C to 40°C.

30. The process of Claim 29 conducted within a temperature range of 0°C to 25°C.

31. The process of Claim 24 wherein X in the acyl halide is-Cl or -F.

32. The process of Claim 24 wherein the concentration of the aqueous hydroxide is greater than 0.1 molar in water.

33. The process of Claim 32 wherein the concentration of the aqueous hydroxide is 1 to 5 molar.-

34. The process of Claim 32 wherein the aqueous hydroxide is selected from metal or tetralkyl ammonium hydroxide.

35. The process of Claim 34 wherein the aqueous hydroxide is selected from the group consisting of KOH, LiOH, NaOH, CsOH and R'4NOH wherein R' is CH3, C2H5- or CH3CH2CH2CH2-.

36. The process of Claim 35 wherein the aqueous hydroxide is R'4NOH
wherein R' is CH3CH2CH2CH2- and R is hydrocarbyl.

37. The process of Claim 24 conducted in the presence of a solvent.

38. The process of Claim 37 wherein the solvent is selected from the group consisting of Freon R E1, Freon R E2, Freon R 113, fluorocarbons, chlorofluorocarbons, hydrofluorocarbons, hydrofluorochlorocarbons, hexane, cyclohexane and mineral spirits.

39. The process of Claim 24 conducted in the presence of a surfactant.

40. The process of Claim 39 wherein the surfactant is selected from the group consisting of fluorinated surfactants and hydrocarbon surfactants.

41. The process of Claim 39 wherein the surfactant is selected from the group consisting of ammonium perfluorooctanoate and sodium dodecyl sulfate.

42. The product of the process of Claim 1.

43. The product of Claim 42 as a dispersion in water.

44. A peroxide produced by the present process of the following structure:
{G(CF2)w[CF(CF3)CF2]x[OCF(CF3)CF2]y[OCF(CF3)]z(C=OO)O-}2 wherein:
w is 0 to 8 x is 0 or 1 y is 0 to 7 z is 0 to 1; and w+x+y+z> 1 G is selected from the group consisting of CH3OOC, BrCF2CFBr, C6F5O-, and -I.

45. A peroxide as described by Claim 44 wherein:
G is CH3OOC-w is 1 to 4 x is 0 y is 0-7 Z = 1.

46. A peroxide as described by Claim 44 wherein:
G is BrCF2CFBr-w is 0 x is 0 y is 0 to 7 z = 1.

47. A peroxide as described by Claim 44 wherein:
G = C6F5O-w is 0 x is 0 y is 0 to 7 z is 1.

48. A peroxide as described by Claim 44 wherein:
G is I-w is 2 to 8 x is 0 y is 0 to 7 z is 0 or 1.

49. A composition of the formula:
H(CF2)4CH2O(C=O)OO(C=O)CH2(CF2)4H.

50. A peroxide mixture comprising [CF3CF2(CF3CF2CF2)CF(CO)O-]2 and [(CF3CF2)2CF(CO)0-]2

51. A composition of the formula:
CF3CF2CF2OCF(CF3)(C=O)OOC(CH3)3.

52. A composition as described by Claim 45 selected from the group consisting of [CH3O(C=O)CF2CF2OCF(CF3)CF2OCF(CF3)(C=O)O]2, and [CH3O(C=O)CF2CF2OCF(CF3)(C=O)O]2.

53. A compound as described by Claim 48 of the formula [ICF2CF2(CO)O]2.

54. A compound as described by Claim 47 of the formula [C6F5OCF(CF3)CF2OCF(CF3)(CO)O]2.

55. A compound as described by Claim 46 of the formula [BrCF2CFBrOCF2CF(CF3)OCF2CF2(CO)O]2.

56. The product of the process of Claim 24.

57. The product of the process of Claim 56 as a dispersion in water.