Document K6ybJ2O7O0qk79ZOV6rpKNqo

'V-. ^^/<*fc.:S `AjmzrSi? 7HfKfif ft: CONFIDENTIAL SYNTHESIS AND APPLICATION STUDIES ON HYDROLYZED AROCLORS AND DERIVATIVES r**ic^i' 1rr RD-58-Pin. No. 1115(j) .J December 30, 1957 RicE RESEARCH Dayton By: J. Schwendeman, J. K. Craver, J. Dazzi, J. R. LeBlanc and J. A. Herblg fhts f(,port ^ jnforma!ion Job Nos. contained therein the property 9^11, 9411-03-3, and 9416-03-3 of the MONSANTO CHEMICAL COMPANY DISTRIBUTION Organic Division _ (a)' " (b) W. B. Hicks (c ) F. M. Murdock (a^ M. C. Throdahl/L. E. Klein (e) F. B. Zienty/J. H. Lum Plastics Division (f) iT W. Mohrman `(g) Research and Development Files . Research and Engineering Division | h | T~. K. Craver/J . Schwer. deman ij'1 E. W. Gluesenkamp/F. C. Meyer (k) J. R. LeElanc/D. B. Sharp (l) M. Kosmir. fm) H. K. Nas n) I. 0. Sal i Extra (0) Extra r(q)\ Extra ?! Ex t ra Ex t ra Extra St. St. St. St. St. Loui 5 Loul 3 Loui 3 Loui S Loui 3 Spr ingfi eld St. Loui s Day- tTT Day t 0 r Day ton Dav - p, n Day ton St . Loui s Day ton DSW 621253 STLCOPCB4095283 STLCOPCB4095284  *. 4 : < < The alkaline ^hyijspISrior Aroclorlhas ipt^^-ded in Inxe^st new family of Mohjarito-baaed chemical;; j^tennedlates i f thelppiy^. . chlorinated blspnetfojis. These can be used to prepare epoxy-fypel resins (by reaotiSn with epichlorohydrin). Our Interest in these resins was aroused by the possibility they offer for (l) new !; markets for Aroclors; (2) possible patent protection; and C3) tk improved flammability, adhesion, and dielectric properties, which might be conferred by the high chlorine content. With this in mind, a detailed study of the hydrolysis, glycidylation reaction and application of the epoxy resins in various outlets was under taken . The major current use of epoxies is in the preparation of sur face coatings. Because of this, much of our work was directed toward the preparation of drying-oil fatty acid esters and then application and evaluation these esters in coatings. We r. pea that such an epoxy resin might fulfill the requirements of a high temperature, thermosetting and flameproof metal-metal adhesive. This report describes efforts to develop such an adhesive. The scope of study was broadened briefly to include lasting, petting, and laminating applications of the unmodified Aroolor epoxies. Cther stuiies were aimed at using hydrolyzed Aroclors as curing agents i.n p i i ye u'.u j, and the epoxy e'hers as vinyl stabilizers and as molifiersAfor alkyd resins folx e. p*/e,S %t IJh|| ffhf 11 ^ wfW DSW 621255 'i? 'sy -j if f * f f * " l f ? ? i T tif'V% STLCOPCB4095285  - i > - ' --< i.. 'v lethanojllffi? aucaxiner hydrolysis of Aroclor 1268 under pressure? ^ ^as irtveaw.gat^d^l jThe; process studied 5ls similar to the Organic ' | .Ghemicals^ivi8ienhydrolysis of tetraohlorobenzene to 2,4,5 - trlohlorophenb||and of hexachlorobenzene to pentachlorophenol (Ref. 1 & 12)1 ^Variables studied were time, temperature, excess NaOH and water concentration. The product was isolated as dihydrolyzed Aroclor 1268 (dihydroxypolychlorobiphenyl). r, Qlycidylation of dihydrolyzed Aroclor 1268 was accomplished most satisfactorily by adaptation of the tetrachlorobisphenol A glycldylation procedure developed by Organic Chemicals Division Research (Ref. 2). The glycidyl ethers are hard, brittle, amber colored solids, capable of the typical epoxy reactions. We evaluated these Aroclor epoxy resins as surface coating resins, studying in detail their esterification with drying oil- fatty acid. By comparison with conventional epoxies, these new resins react very rapidly. The resulting esters were evaluated both as varnishes and as paints. Particular atten tion was given to the properties which would make these resins attractive as surface coatings. At the same time, detailed studies were made in efforts to develop suitable formulations for adhesive, potting, and laminating applications. Noteworthy characteristics are the high heat distortion values and ad hesion to glass. The Aroclor epoxies were examined and found promising also as stabilizers for^polyvinyl chloride, and curing agents for high acid value alkyds. Exploratory studies on styrene-soluble Aroclor epoxy fumaric acid polyesters were made. This work was carried out during the period May 25, 1955 to December 1, 1957. MH | H rf * f f - i : l f c \ f \ >. * DSW 621256 STLCOPCB4095286 t WA M* W^-S- | - | Complete^phenolio Efcnctionality in hydrolyzed Aroclor^.2.,j; oan be obtained by 3te- methanolio hydrolysis process, Hjfdrblyzed Aroclor 1268 having neutral equivalents equal to theory Was* obtained reproduoitCJy and consistently. The diglyoidyl ether oan be prepared reproduciblv in good yield and high oxlrane oxygen content (91.2J< of theory) from this hydrolyzed Aroclor Z2S8. Epon 828 is 83% theory. The dlglycidyl etherof hydrolyzed Aroclor 1268 reacts very rapidly with drying cLl fatty acids to produce eaters having good color, low viscosity, and enhanced tolerance for hydrocarbon solvents. Reaction ~Smperatures may be 60C lower than with con ventional epoxies. These esters, when usd as coatings, show satisfactory drying times, excellent adhesion to metal, good water resistance, high flexibility and touginess. Pigmented samples sho* excellent weather stability and flameprooflng The alkali resistance jf these coatings is relatively poor and the resistance to weaiierlng in unpigmented films is not as good as conventional epoxy esters. The Aroclor-epoxies cai be readily cured with most conventional epoxy curing agents. Best results property-wise are obtained by using high melting curing agent3 which can be dry blended with the resin. These cured resins have fair adhesion to glass and ceramics, but very poor adhesion to metal. The rapid reaction of Aroclor-epoxy resins with dibasic acids could lead to plasticisers, polyester resins and non-drying alkyds. The Aroclor-ecoxy can be used to cure hlgh-acid-value alkyds to give interesting new surface coatings. , The dihydroxy material itself can be used to cure regular epoxy resins, yielding hard, brittle, nonflammable castings. The Aroclor 1268-epoxy is comparable to Paraplex d-60 and 4 Epon 828 as a heat arc light stabilizer fdr plasticizedvpoly-f yinyl* chloride. '' ' 1 .$ DSW 621257 STLCOPCB4095287 0 the|0rg'anioj. IJvisidn^aooept ^lyz_e_c_ JP^fasi iew; produo tfan|i that' the application dat p re 8fe nfett^iejbp isedtofdevelop uses within the.surface coating Indus try As industry Interest develops. It would be advisable to under take further prooess studies on the hydrolysis step, aimed at lower costs and"better color of the diphenol, since cost and color may be the ehlef objections to this product. Application studies aimed at broadening the uses of the dlphenol should be instituted. Potential markets exist in polycarbonates, epoxy resin curing agents, and polyesters, in addition to the surface coating UBes outlined here. Further applied research is indicated on the Aroelor-epoxy-alkyds, on. the glass-glass adhesives, and on flameproof coatings. The hydrolyzed Aroclors should be promoted as flameproofing epoxy curing agents - comparable to HET anhydride - and as modifiers generally for epoxy resins. For additional suggestions on the use of hydrolyzed Aroclors please see References 3, 4 and 5. 4 i ? - - i *v M # - i ' -f : r i 4 * I I i >it t DSW 621258 STLCOPCB4095288 i m HYDROLYSIS to' bAe't _ , . .. .. , ff The chemistry;- of |the` hydrolysis reaction .appears : - ill #411 :; 4 ; I I v- - ' i. + 4 NaOH -> NaO (448) ONa + 2 NaCl 2 H2O 2. NaO <g) ----^J^ONa + 2 HC1 -> HO<^V-^--V^OH + 2 NaCl (411.2) The molecular weight, of the Aroclor was calculated from the chlorine analysis (6?.=$ Cl) and from this the theoretical neutral equivalent was determlr.-' d a.-- 25.6 . The Initial r.ydr lysis product was substantially 90^ dlpnencllc. Application stu-divs particularly In the flei 1 of adhesives and rotting compounds indicate! the desirability of a ICC# diphenol. Synthesis work war directed to this goal and by the use of ex cess NaOH in metr.anol solution, essentially 1 ' '% dihydroiyzed material was tbtuir.e :. (Ref. 1) We have shown thVa rj w < :reasir.g the excess NaCH in the hydrolysis "f Aroclor 1263 s 'icantly Increased the phenolic content of the product i?igu -- Increasing the reaction temperature from 190C to 24oC did not Increase the phenolic content as much as anticipated. The effect of Increased temperature Is summarized below: EFFECT OF TEMPERATURE ON HYDROLYSIS OF AROCLOR 1266 2056 excess NaOH 4ojfc excess NaCH 190^ 219 N.E.** 211 240 216 N.E. 21C f - *f *0ne hour held Derlod In methanol n'e- ' 205-6` f til i Increasing'; the reaction time at 190C.frora 1 hour to 3; hours, ess< Napy (guns jl Jet I) dec&cgsse DSW 621259 * W*<f #') STLCOPCB4095289 STLCOPCB4095290 STLCOPCB4095291 Dimethyl ether, identified as a by-product it\ other methanolic NaOH hydrolysis reactions, appears to be the main by-produot. About 30 g of a low boiler was trapped during Tenting of Run 17 (Experimental Section). Formation of dimethyl ether would ac count for the pressure build-up during the reaction hold period (see Figure 2). At present, it is assumed that methoxlde ion attacks the methanol by an S^2 mechanism displacing the hydroxyl group as hydroxyl ion. :H-^0e + CH,0H CH^OCHx + OH The methanol recovery for Run i6 (Table II) was 96*, (92% of material containing 83* methanol and 17% water). In Run l*i. assuming the same ~atlo of methanol to water, methanol recovery was 88*. Recycle of methanol was rot attempted. Hydrolysis ir. water alone failed at 2 dcC, owing to insolubility of the Arcelor. H'wever, about 9a% (about 86* d!phenol) hydrolysis took place in 50* me than cl-water during I ncurs at lpOC (Run r) using 20* excess base. These same renditions should be attempted using 80* excess NaOH. Attempts to hydrolyze Aroclor 1268 in a 5-gallon vibrator-stirre: Blaw-Kncx autoclave resulted very little hydrolysis 1 <b * <3 attributed to inadequate agitation resulting from too snort a stroke or. the vibrating discs. Scale-up from 1.5-1 Iter to 2-gallcn went smoothly except for Run lb in which tot' large a batch was attempted, i Abnormally high pressures built up due to a small vapor space. Insuffi cient agitation resulted In incomplete hydrolysis. The reaction appears to be slightly exothermic when lpC' first reached. Hence, some cooling is always necessary the reaction at 1?C-195C (see Figure 2). f\ 0 [ Acidification was carried out above 9o C since below this tem perature the product precipitates as a hydrate (Ref. 2) Acidi fication was dene cautiously since foaming was encountered at the beginning of the acidification ir. almost all batches The neutral equivalent was determined by the method of D. K. Lynch, St. Lduis Research, and was estimated by ;E. M. Hubbard to be t 1* accurate. Based or. this figure, material analyzing 205.6 to 207-6 N.E. was assumed >ntain theory diphenol (205.6 calculated) fS^drdx1y1. I onteerryvttfdete ti 1 \o if wfife srriql a jjstaV lar&] rtyfat athdd, .31 aJuea rtanf mT i*ti, lride su >weuc DSW 621262 STLCOPCB4095292 STLCOPCB4095293 STLCOPCB4095294 i t '': %r 1 ft _ -. 'I 'p Si _ fr^erent difficult?! f ^of the OH analysis, the N.E. determination was used as the de- * |eiding measure of purity. ;| v ^ ' " II " ' ** ' - ' . t .All products melted at 60-68C, the range being attributed to -r the presence of isomers. . BIOLOGICAL EVALUATION Table 5 of the Appendix contains a brief summary on the biological evaluation data on hyarolyzed Arociors 1242, 1248, 1254, 1262, 1268, 1270 and 1271. Only the last two are difunctional compounds, the others are mixtures of mono- and dihydroxy Arociors (difunctional hydrolyzed Aroclor 1268 has not been tested as yet) The lower chlorinated Arociors undergo incomplete dlhydrolysis. Biological data failed to show special new leads worthy of further investigation. GLYG IDYLATION "he chemistry of diglyc. idyl at 1 on :s as fellows; HC x "Cl 6.5 (4ll) OH 2 c:ch2chch2 catalytic NaOH CH.O'* HCH2C \ cleY och2chch2 -t- 2 NaOH C Hp -CHCHpO (" V \/ X-y \_._}-0CH2CHQH2 + 2 NaCl - 2 H2( "Cl 6.5 * 4 ; i ?he* adaptation of the jSt. Louis .tetrachloro ,blsphenol A glycidyl-j. ! r; I t atibn procedure (Ref I 2) *gives glycldyl1 ethers of dlhyarolyzed t Aroclor 1268 which are high in epoxy content and very low in chlorine.j T^ie epoxy content |5*?0-5v56^ oxirane Lvaleni L-.jL JL . DSW 621265 STLCOPCB4095295 The digiyoidyl ether is a lighIbiljer "resin aelting Products with lower epoxy content have higher me1tingfpoints| perhaps due to the presence of "polymeric11 units. The potting and encapsulating trade would prefer a liquid glycidyl ether? but this cannot be obtained from dihydrolyzed Aroclor 1268. The lower-chlorinated hydrolyzed Aroclors might yield liquid glycidyl ethers, but these hydrolyzed Aroclors are not completely dlfunctlonal. The surface coating applications (which consume the greatest quantity of epoxy resins) on the other hand, pre fer the high melting resin such as this. No measure was made of a shorter dehydrohalogenation step, nor was the use of higher temperature, toluene, or chlorobenzene as reaction solvents tested. At this early stage of investigation our main goal was to prepare reproducibly high epoxy-content diglycidyl ethers, free of alkali, for Applied Research. APPLICATION STUDIES Esterification The structure of conventional epoxy resins is: CrU OH " C HgC HC H2 - ofVc-'\och2chch2- 0 CH^ ) -c - CHv A och2ch6h2 / Since "n" is an average, it may have fractional values. This structure can react with carboxylic acids to give esters in two ways (Ref. 6): 1) Reaction of oxirane oxygen R-COOH + CH2-CH--- -- OH R-COOCEp-CH----- The ester group may be on either the d or p carbon atom. _ No ? waiter i3 liberated but a hydroxyl group ls^fo^Taed| . 2) Reaction of carboxylic acid with hydroxyl. This1 Is a t ` conventional esterification with loss of water. i.Thej | HydroxyliB|may be ieither ithqs,e ` ' 4 |i) I! DSW 621266 mm r STLCOPCB4095296 *In addition;* the following may occur in varying degreea depend ing; on reaction conditions: i. , s f ;t :: ' i ' 5) Ether formation --CH- + CHo-CH-- OI H VV / T"?H 6-CH2-CH- 4) Hydration of epoxy by water a HOH + CH2-CH-- ?H ho-ch2-ch-- It has been shown that in the uncatalyzed reaction at 200 of caprylic acid with an epoxy resin, the first three reactions occur in the approximate order of 2:1:1. Reaction (4) is sup pressed by the removal of water by azeotroplng with xylene. Our study of the esterification of the diglycldyl ethers of hydrolyzed Aroclors is divided into several well defined areas: 1) Esterification of oxirane oxygen only of diglycldyl ethers based on 90# dihydrolyzed Aroclors. 2) Esterification of oxirane oxygen only of ethers based on 100# dihydrolyzed Aroclor. 3) The preparation of more completely esterlfled epoxies, i.e., oxirane oxygen plus some hydroxyl esterification. Our Initial work was on diglycldyl ethers derived from hydrolyzed Aroclor 1268, from the St. Louis pilot plant, which was 90# difunctional. The diglycldyl ethers derived from this phenol were found to be unsatisfactory for casting purposes because of their lack of 100# dlfunctionallty. However, they did make satisfactory coating resins. The presence of 10# morofunctlonal material apparently does not detract from this material's film forming characteristics. We used soya fatty acid as a standard unsaturated acid, but made some esters from tall oil to observe color and drying times. _l I' " A : * * *. t>. . veryTearly in our work, we noted a marked variance in the rates of esterification of he oxirane oxygen of the several ethers; as shown dn the following Table IIL and graphics Ly iin ^Figure ,3 " rthll 1 it- & urn DSW 621267 STLCOPCB4095297 STLCOPCB4095298 Ft TABLE III SOYA PATTY ACID ESTERS OF QLYCg>YL KTHKR3 (Oxirane Oxygen Only) Ester No. Derivation * Reaction Time Reaction Epoxy Hrs at 200C Rate 354868-1 354872 354873 354874 354875 354877 354881 Aroclor 1270 Aroclor 1262 Epon 1004 (Bisphenol A) Aroclor 1268 Tetrachloroblsphenol A Tetraohlorobiphenol Epon 828 (Bisphenol A) 3.83 5-39 1.68 5-10 6.44 6.17 9.4 0.5 2.5 3-5 3-5 6.0 11.8 8.7 Rapid Slow Slow Slow V.Slow V.Slow V.Slow The extremely fast reaction of 354868-1 was very Interesting and exciting when compared to the reaction time for Epon 1004. As further data accumulated. It was found that the fast reaction was not limited to derivatives of Aroclor 1270, but also occurred in those of 1268 and 1262. The only significant difference that was found between fast and slow reacting diglycldyl ethers was one re lated to % epoxy and consequent molecular size or "n" value. This correlation Is shown In Table IV and In Plgure 4. \} I TABLE TV SOYA PATTY ACID ESTERS OP DIOLYCIDYL BT^ntWR OP HYDROLYZED AROCLOR 1268 (90* D1PTJNCTIONAL) Ester No, Mol, Vt. % (Based on Epoxy % Epoxy) Hydrolyzable Reaction Time Reaction Chlorine Hrs at 200 Rate 359068 5.16 620 354874 5.10 628 359060 359061 359062 4.93 648 4.53 ,706 $5.90 | 1820 359094-1 3.37 950 359094-2 3.24 , 988 359056 4 .52 i 35107911life! | 0.35 0.62 0.48 0.015 0.20 0.8 V.Slow Slow Slow Past Past V.Past V.Past ,Past , * , yinsop.M; 9 tprodupf A ! IH DSW 621269 ^* 11- ' ji ^ ; ' f 3 " ' ? t T - ' rr STLCOPCB4095299 STLCOPCB4095300 * f' || . I=f. -% - . .. . . *T: Jro this, and other data. It appears the esterification of the oxlrane oxygen?of4tfce?<llglyoidyl ethers of 90fl dihydrolyzed Aroclor Is slow when, "n" Is less than " mates 0.?. The upper ^1 init for but it appears to be less than 5.4. The observation of molecular size as a determinant in the rate of esterification is supported by the following series of esters made from the diglyoldyl ethers of tetraohlorobisphenol A (Table V and Figures 5 A o) received from Organic Research (Ref, 7), TABLE V SOYA FATTY ACID ESTERS OF OilRAKE OIYQEN OF DIQLYCIDYL ETHERS OF TETRAOHLOROBISPHENOL A (Figures 5 3c 6) St. Louis X W,B, Reference Epoxy Reaction Time Hours at 2C0 A84890 A8490O A84983 A84982 A84981 6.4 3 36 1.98 1 06 0.6l 0 1.2 2.7 6.0 11.0 3.8 14 10 2,0 62 Slow Past V.Fast Fast Very slow The commercial epoxy resins further support our concep 1004 (n - 6) esterlfles more rapidly than Epon 828 (n Figure 3- Ve observed that hydrolyzable chlorine In gross amounts Is detri mental to a fast reaction. In preparation No. 359064, "n" was 0.82, and the esterification should have been rapid. However, hydrolyzable chlorine was 3.51J* and the reaction was slow. The series of esters reported in Table IV shows that hydrolyzable chlorine content is not too critical as ohlorlde contents range from 0.00 to 0.89^ and the distribution of values Is random In the series. At an early stage ln^our wo^k,.It was felt that thevanalysis for . epoxy content was open ffco question and that our seemingly slow reactions were due to attempts to react more fatty acid than was warranted by the actual* oxirane content. Accordingly,. two ^ex periment lamoMtfef L< DSW 621271 IT mvmnpn STLCOPCB4095301 of Fatty Acid Unreacted STLCOPCB4095302 Time at 2 0 0 (m inutes) STLCOPCB4095304 5 -1 f EgaL _^ _ Aroblora ^iVe^^uS'faSj^ne^e^e^^n^ ImpWtant* IdvintS^s"___ . , conventional epoxies (or those"based on tetraohlorobisphenoi A). Reduction of time^in the kettle should be of economic importance to the paint vehicle manufacturer, as shown in Appendix B. As a result of our success with the 90# dlfunctlonal compounds, additional successful efforts were made to increase the di functionality of the hydrolyzed Aroclor. Our later esters were all based on dlglycidyl ethers of substantially 100# dihydroxy polychlorinated biphenyls. The bulk of the samples have had an "n" value of ca. 0.09. The reaction rates are rapid, but are significantly slower than when the 90# difunctional material was used. We have no explanation for this. A number of esterifi cations are tabulated in Table VI TABLE V] NEr 55c % Spoxy 5 65 5 52 5 52 u u Q- 0 - ;6 Mol Wt 566 580 58c 71C See L OGC 0 C C-? c CO c 1p neactlon Time Hours at 200 i OC a h2 3 25 6.25 0-0 5 C-0.5 Reac:i Rate Past Slow Slew V Sioi V Fas V Pas Comments Ca(0H)2 aide: Ca(OH'2 ante Thus, the evidence we have on the \C0% bifunctional resins tends ir. general 10 support our incept - tnat molecular size infiuerve: tne rate of reaction Tr.ere are some exceptions Experiments 567968 and 6? using a resin with an "n" value of C . *3. should have given :ast reactions. They were slow even when lime was added to catalyze the reaction. Tr.e method of synthesis of this particular resin was different than the method U3ed for otner resins An attempt was made to Increase the "n" value by reacting a dlglycidyl ether o:' "n" 0 value with aihydrolyzed Aroclor 1268.- Each molecule of the iiphenol should react with two of dlglycidyl ether to give a molecule with "n" - 2. The resultant "n" of the resin was raised to 0,35. However, this method of synthesiB should not have-beerf the reason for the slow reaction since Experiments 367955 and 56 used resins prepared in a similar-manner and the reaction was very fast. If DSW 621275 STLCOPCB4095305 |Kr% | t 9 rf If Line in-small' amounts, 0.15 to 0.205* based on epoxy, was effeo- l | tlve in catalyzing the esterification in some cases. It improved | ' the reaction rate in Run 367970 in which "n" O.Oo. However, " ' lime was not so effective in other cases; for examole. Runs 367960, 367968 and 969. ' Only the esterification of the oxirane oxygen has been described to this point. Epoxy esters having up to PC-H of their functional groups esterified are the usual commercial products. The preparation of a number of the mo; esters is tabulated in Table VII ' omu _ > esterified 'ABLE vri ESTEHS 0? PATTY ACID w/LPCXY RESINS (0.5 to equlvale:. t) Patty N3P Derivation Ac Id 359C97 35*861 35*362 359077 359080 Epotl n 100M * tl n Aroclor n 126n 8 Soya 1* n it 36*202 36*203 367972 3739*5 EDon 100* Aroclor 1268 tl tl ft It Tall t . Soya ft Es 11 r* Leve 1 90% 6C% 50$ 8056 90* 905* 90* 805* 70% A:id At Max Value Terns Teino . Comment - 11 55 *.3 5.8 15-8 12.5 "3 5' a 'i * 50 5-7 - -' ^ ri-/> i~\ 2 c-C 2 OC w .. 280 n = C.35, 56 hydroxyl n = 0.35, 5* nydrcxyl 9.6 ~ c 220 15.3 I; .25 2 a 0 8.8 . 2CC n = 0.68 n = C.06, Ca(0H)jj pre n - 0 .2 We found that the 50-905* complete esterification of the Aroclor- epoxies takes place at temperatures 60C lower than the similar esterification of Epons! This7isfnoteworthy and could be of con siderable economic importance. The examples cited are based on 90{* dlhydrolyzed Arpclor with the .exception of 367972. This was il^yd^'o "; jne lowQsj Iftl vali $* DSW 621276 m *i STLCOPCB4095306 tfheaej t The ,Aroclor-epoxies all show good stability during, the esterifi cation reaction. Neither HC1 nor Cl2 was ever detected in the off gases from the reaction. However, the esters ido reaoh a minimum acid value and if held at high temperature for periods beyond this, the acid values increase. No explanation is offered to account for this phenomena. These esters are quite fluid by comparison with Epon esters. This does not seem to be attributable to molecular size as we at first supposed Number 3739^5 is a 0.7 equivalent soya fatty acid ester of an Aroclor 1268-epoxy. Its molecular size is Intermediate between Epons 1C01 ana 100^ and should be similar to an ester of Epon 1004. Its viscosity is only A~3 on the Oardner-Hoidt scale compared to a literature (Ref. 8) viscosity of I to K for the Epon ester. This fluidity would be advan tageous in spray or dip applications wpere a high solids loading with low viscosity is desirable. We attempted to Increase the viscos: :hese esters by re placing part of the soya fatty acid with dimer acid. The vis cosity in Run 767988 was Increased from A-3 to Q. Another approach to increasing viscosity by increasing the molecular sice was made in experiment 367967 by reacting 0.141 mol of digit-', idyl ether of "n" value of 0..06 with 0.023 mol of phthailc ar.nydrIde . This "enlarged" dlglycidyl ether was esterified with soya fatty acid Viscosity was not Improved, remaining at A-4. However, this reaction demonstrated that phthalic anhydride reac ts readily with digly idyl ethers, at least in limited amounts. This property may be useful in making mixed epoxy-alkya type resins. Skinning and gelation of the solutions of the esters of the Aroclor-epoxies takes place to a lesser degree than In Epon esters. No quantitative measure of th.13 was made, but the results were definite.. - The Arocior-epoxy esters are more soluble in hydrocarbon sol vents than conventional epoxies, as shown; > t -!* f 1 mm DSW 621277 STLCOPCB4095307 # '#MMr tm m % w;':* *.** > -: m& vi :- - ": .* *-?r ' f # gi Epoxy SOLUBILITY OF EPOXIES IN XYLENE AND ACETONE Mol. Wt,, Xylene M1 Acetone n wt. (gms) to Dissolve 1001 1004 Aroclor 1270 Aroclor 1268 2.3 5.5 0.76 3.2 990 1900 834 2160 1. 3 This tolerance for hydrocarbon solvents by the Aroclor-epoxy could be of considerable economic Importance "Varnolene" may be used Instead of xylene, with similar results We have shown in Experiment 3739^6 that the Aroclor 1268-epoxy will cure hlgh-acid-value alkyds. American Cyanamld 3 Rezyl 1102-5, with an acid value of **3, was ased. TABLE IX CURE OP ALKYDS Rezyl Xylene 1102-5 (ml) Curing Agent Res 1 stance to Solvents Xyiere A'_ etone ;035 NaOH 5i ah . 50 gms 35 none Good Palled Failed Good 50 gms 35 Pb, Co, Mn driers Good Pa 1 ? ed Pal led Good 50 gms 35 3.2 gms* O.63 equiv. Good Fa 1.ea Improve i Good 50 gms 35 6.4 gms* 1.3 equiv. Sol? tenea - Improved Good 50 gms 35 | -M 12.7 gms* 2.5 equiv. So1tened t IM. i - Best - Good No Attack ^Aroclor tl268-Epoxy I` I: :jffl I'1 y A i i i jt- DSW 621278 y y-yf-f r w. r n STLCOPCB4095308 i- & .. _ . . MW Ifh other lllqras and eurltfg is e^Wted niyfwe^se of laefcaaffi* or` psr' urea resins. Sueh an alkyd system was successfully cured'by the Aroblor epoxy to give coatings of high gloss, fair hardness, good detergent resistance, and good flexibility. Resistance to xylene and alkali was only fair. J ; Coatings The esters prepared from the diglycidyl ethers of 90 and 100# dihydrolyzed Aroclors were evaluated as surface coatings Cobalt naphthenate drier was added equal to 0.04$ cobalt based on the weight of resin. Advance Solvent and Chemical Corporation anti-skinning agent was used at a concentration of 0.255^ based on vehicle solids The vehicles prepared from the 100$ dihydrolyzed material were also tested as enamels pigmented with Tltanox RA 50. The finishes were coated on steel panels and air dried. They were evaluated as to dry time, physioal and protective properties, and exposure in the Weatherometer and on the outdoor test, ra: * . The following soya esters of the 90ft Aroclor-epoxy along with controls based on Epons we-? evaluated. All are oxlrane esters ___ except 354862 which Is a 50$ equivalent ester. "ABLE X DRY TIMES FOR VARNISHES MADE FROM EPOXY SOYA ESTERS Es ter No D' ree Derivation of Set Print Free Tack Free 35U873 154862 ^54881 Epon 1004 Epon 1004 Epon 828 Gx .: d ne '-:$ Ox 1 rane 354868-1 * Aroclor 1270 754868-2 * Aroclor 1270 354874 Aroc.or 1268 354872 Aroclor 1262 354875 tetrachloro- bisphenol A 354377 1 1 tetrachloro- dlphenol model compound M r It n M ft :0 min 0 ,t. 1 n 2.5 r.rs 10 min 20 min 1 5 hrs 2 hrs 2.5 hrs 2.75 hrs 1 hr 2 hrs overnight 3 hrs 3 hrs overnight overnight overnight. ' *I overnight 2 hrs overnight not after 4 days overnight overnight overnight overnight not after ^ | days not after days hi DSW 621279 STLCOPCB4095309  4, i f* * t fTABLE XI CHARACTERISTICS OP EPOXY ESTER COATINOS Ester No. Mar ResisDerivation tance Sward Alkali Hard- Resistance Water Adhesion ness 2% NaOH Resistanc 354873 354862 1004 Average V.Oood 15 Good 1004 Good Qood 18 Good 354881 828 Poor Fair 4 Poor 354868-1 354868-2 1270 1270 Average V.Oood Average V.Qood 17 Good 18 Good 354874 1268 Average Good 10 Poor 354872 1262 Average Good 13 Poor 354875 TCBPA Poor Good 10 Poor 354877 TCBP Poor Fair 6 Poor Excellent Excellent Fair Good Good Poor Fair Fair Poor In general. while the esters of Epon 1004 were best. they were closely followed by the derivatives of hydrolyzed Aroclor 1270, 1262 and 1268 in that order. The hydrolyzed Aroclor ester films were much better than those of tetrachlorobisphenol A, tetrachlorodiphenol and Epon 828. Mar resistance is a qualitative measure of the resistance of the films to fingernail scratching. The ratings reflect the surface toughness of the longer chain Epons as opposed to the shorter chain chlorinated epoxies. Adhesion was measured in two ways; first by scratching with a knife and attempting to broaden the scratch, and by bending a thin panel around a 1/4" mandrel. The chlorinated epoxies were equivalent to the Epons in this respect. The hardness of the Aroclor-epoxy esters was as good as that of the esters based on Epon 1004. Epon 828 and TCBP both gave soft films. The materials rating good in alkali resistance resisted soften ing for ten minutes in 2% aqueous NaOH. Those listed as poor were dissolved inffive minutes or less/ The esters based on hydrolyzed Aroclor 1270 were almost as good as those based on the Epon 1004 and much better than Epon 828. . ' I I i-.J DSW 621280 K STLCOPCB4095310 P4#-' ;| * _........................................... ; While walfer Ibfe- closely followedJoyy thefesters based orf tne l^i^roayxedjAroal All the films were more resistant to water than they had been to alkali. '- TABLE XII Ester No. 354873 354862 354881 ^54868-1 tsa868-2 35^874 154872 354875 354877 PERFORMANCE IN VEATHEROMETER Derivation Time to Failure or Completion of Test Epon 1004 Epon 1004 Epon 828 Aroclor 1270 Aroclor 1270 Aroclor 1268 Aroclor 1262 TCBPA Model Compound 640 640 640 160 120 12C 120 12C 120 hou11 rs 11 11 It II 11 M 11 Comment (Based on Visual Exam.) No failure n II Failure Failure was accompanied by a complete loss of gloss and the development of a dark brown color, starting at the center of the panel and spreading to the outer edges of the panel as exposure continued. The esters based on IOO56 dlhydrolyzed Aroclor 1268 and on Epon 100^ were similarly evaluated. Clears and pigmented esters at the oxlrane oxygen and 0 7 equivalent level were tested. A commercial epoxy ("Eponal"-Lewe Brothers Paint Co.) was Included for comparison. See Table XIII The plgmen*:ed enamels were male up according to the following formulation: Parts by Weight Tltanox RA 50 Resin Solution (5056 soln.) 142 284 Either xylene or methyl lsobutyl ketone (MIBK) was added to give a suitable viscosity for grinding. Grind 24 hours in a pebble mill, adding cobalt drier and antl-skinnlng agent as ;eded. DSW 621281 f f t iff 1 .y STLCOPCB4095311 % STLCOPCB4095312 Mm CHARACTERISTICS OP ENAMELS MADE FROM EPOXY ESTERS & White No. NBP 367974 367976 367978 367979 367982 367983 367981 Viscosity No. 4 Ford Cud lbs/sal 19 9.74 52 IO.38 14 11.08 13 11.00 77 10.43 59 10.48 13 10.85 C-17238 - Lowe Bros. ] Solvent (Other than Xylene) MIBK MIBK These varnishes and enamels were coated on steel panels. The dry time, and water resistance was as follows: " TABLE XV DRY TIKES AND WATER AND ALKALI RESISTANCE Coatlns Epoxy Dry Time (Hours) Water Alkali Dust Free Dry Hard Resistance Resistant C-l C-2 C-3 C-4 C-5 C-6 C-7 C-8 10C4 1CG4 1268 1268 1004 1CC4 1268 Commercial 0.5 0.5 >2<l8 >2<l8 >1<17 2 2 1 1.5 1.5 >2<l8 >2<l8 >1<17 2 4 Excellent II Good If Excellent ft ft ft Excenllent Good II Excen llent Poor Excellent ( if W-l V-2 V-3 V-6 - . 100^ 1C04 1268 1268 1004 1004 0.5 0.5 >6<22 > 6<22 2 2 l 1.5 1.5 >6<22 >6 <22 2 2 >4<20 s S it Excellent fl Good Excellent' If n. Excen llent Good *Poor Excellent V" toe ^ 7 "? 1 7^ DSW 621283 t STLCOPCB4095313 TKefpt tablet i films aw' given in the fgillowing | -- - TABLE XVI PROPERTIES OP COATINGS AS PRKPARwn Coating No. Derivation Degree of Ester Sward Flexl- Hardneas blllty C-l C-2 C-3 C-4 0-5 C-6 C-7 C-8 w-i W-2 v-3 W-4 w-5 w-6 w-7 w-8 Epon n 1004 Aroclor 1268 ft Epon n 1004 Aroclor 1268 Commerclal Oxlrtlane n ti 0.7 egulv. Epon 1004 ft Aroctllor 1268 Epon 1004 ft Aroclor 1268 Commercial Oxlrane 0.7 ecjuiv, 41 39 6 8 10 10 6 4 21 24 9 9 10 8 6 22 Good Pair Good Pair Good Pair Adhesion Panlled Passed ft Passed ft The marked difference between the hardness of the oxlrane esters ^understandable In terms of oil content or i-hf Ki-i* *?on e3ters are very low in oil content, while Sl0K"ePXleS ar medlum 011 length materials and should be softer because of the increased oil content. Note Table XIII. The Aroclor-epoxies are excellent with respect to flexibility ana adhesion, equal to or better than the Epon esters. 7 rTrPle030 Vheouls.0&ted n 3t6el ^ PlaCSd ln the Weatherometer f Ii I I * I I I ? ? -7 - n t r - ' Tf : El STLCOPCB4095314 I' >*' PROPERTIES OF COATINGS AFTER 400 HOTOS IN WEATHEROMETER i " 1 > Coating No. Derivation Sward Hardness 0 200 oo 400 C-l C-2 C-3 C-4 C-5 C-6 C-7 C-8 Epon 1004* ri # 41 50 60 50 37 50 68 40 Aroclor 1268* 6 8 6 " * 7 10 12 4 4 Epon 1004 10 38 50 40 10 48 58 50 Aroclor 1268 6 22 38 24 Commercial 4 44 62 48 W-l W-2 W-3 W-4 W-5 w-6 W-7 w-8 Epon 1004* "* 21 26 34 32 24 36 36 38 Aroclor 1268* g 10 12 1C n * Q 18 16 12 Epon 1004 10 26 34 32 8 36 38 30 Aroclor 1268 6 12 18 16 Commercial 22 34 36 28 Oxirane esters Plexibilitv Palled It n tt Passed II Failed II Palled II Passed tt It II II Palled Adhesion Failed 11 n n Passed n Palled II Palled If Passed II II 11 11 Failed The performance of the pigmented Aroclor epoxies with regards to flexibility and adhesion was most gratifying. They are equal to or better than the corresponding Epons or the control. ! H i ii hr *t " ? *f* DSW 621285 STLCOPCB4095315 - if i * ' f i > 4 <5 . PERFORMANCE OH OUT-OP-DOOR EXPOSURE (DAYTON, OHIO) Coating No. Derivation Sward Hardness Weeks Exposure 02 4 68 After 8 Weeks Flexibility Adhesion C-l C-2 C-3 C-4 C-5 C-6 C-7 C-8 W-l V-2 W-3 W-4 W-5 w-6 W-7 w-8 Epon 1004* "* 4l 58 50 56 50 41 58 52 46 42 Aroclor 1268* 6 14 14 8 1G " * 8 22 12 8 12 Epon 1004 n 10 22 24 26 30 9 24 26 22 26 Aroclor 1268 5 20 12 6 8 Commercial 4 16 18 :6 24 Epon 1GC4-* 20 26 24 12 12 " 22 34 22 24 Aroclor 1268* * 9 12 14 12 .4 9 12 14 1 eL 18 Epon 1CC4 It 10 14 18 i 6 24 8 16 18 18 28 Aroclor 1268 6 10 12 1c * vQ--` Commerelal 21 30 32 32 28 Poor II fl 11 Good n Poor Good P1o1 or Good It n fl fl 11 Good n Poor II Fair If Poor Fair Goft od It ft fl If If Fair Oxlrane oxygen This data parallels that for the Weatherometer very closely. We conclude that the Aroclor-epoxy-esters are as good or better than the corresponding Epon-esters, including the commercial sample, as far as physical durability is concerned. STLCOPCB4095316 ' r i' Theappearanbe bfthe esters is indicated table: :he following' TABLE XIX jating No. Derivation Initial Color G loss 400 Hours Weatherometer Color Gloss 8 Weeks Out-of- Doors Color Gloss C-l C-2 C-3 C-4 C-5 C-6 c-7 c-8 Epon 1004 Clear tl Aroclor 1268 rt II It Epon 11 1004 it n Aroclor 1268 Commercial it 95 100 100 100 99 QQ 08 ICO W-l W-2 W-3 W-4 . W-5 V-6 W-7 W-8 Epon ri 1004 White It Aroclor 1268 II Epon 1004 II . It Aroclor 1268 II Commercial Qrey 83 98 q6 95 93 95 94 77 Clear rt Brown It Clear It Brown Clear WhMite It 11 II It II Grey 81 Clear 85 It 9 Brown 6 It 83 Clear 85 It 21 Brown 88 Clear 44 55 WhM ite 13 27 SI.Yen llow 83 White 82 11 65 SI.Yellow 5 Grey 88 91 28 28 95 95 23 95 75 94 95 90 92 92 93 50 The clear Aroclor-epoxies developed a grainy type of alligatoring. The clear Aroclor-epoxy finishes could not be recommended for outdoor service, due to their loss of gloss and darkening. The pigmented finishes, however, stand up quite well out of doors. These results illustrate the problem of correlating Weatherometer data with outdoor performance. The soya ester of the oxirane oxygen of Aroclor 1268-epoxy has a chlorine content of 5?^- This should give a good degree of flameproofing to a paint prepared from it. A paint was made up based on the recommended "formula of The Baltimore Paint and Varnish Production Club Technical Committee on Fire-Retardant Paints (Ref. 9)., .A control was also made containing a conven- , tional non-chlorinated epoxy ester. The formulas are as follows: 1 III `I fir 'y i DSW 621287 fI&tfI STLCOPCB4095317 M Antimony Oxide TItanox RA-50 ' Magnesium Silicate Vehicle in Lnat (Parts by Weight) 13.5 9.0 13.5 64.0 (No. 3679573 oxlrane ester Epon 1004, 50* solids) no chlorine 13 ,5 v ' 9.0 13.5 64.0 (No. 367970, oxirane ester D.Q.E. hyd.. Aroclor 1268, 50% solids) 8.6^ chlorine Mill the ingredients in a pebble mill for 24 hours and add driers and solvents as required. Sticks of pine 3/4" square by 18" long were used in testing the paints. Each stick was given two coats of the test paint The sticks were dried in air for 1 week following the second coat Tests were run in the hood with an aluminum foil shield around three sides of the test piece. The stick was hung in a vertical position. The procedure was a much simplified version of the ASTM Method D 1361-55T. Ignition was by means of a gauze wick soaked with 4 ml of 95% alcohol. The height of the flame was taken at 10 second intervals for the first 100 seconds after ig nition. The weight loss of the sticks was also observed Photo graphs of the test pieces before and after the test show the results of the fire tests (Figures 8 and 9). The designation of the sticks are as follows: B is an unpainted stick, E the Epon ester, and C the Aroclor epoxy ester. The results are tabulated below: TABLE XX Height of Flame (lnohes) for Sample B Sample E Sample C sec . sec . ri lo; f! >ox 2 [orfl DSW 621288 STLCOPCB4095318 STLCOPCB4095319 w $ ............ Th ...Shofb_gfap. hs and the rest s^3.0frsoxe 'XX? show ( ^ Aroclor-epoxies are superior to conventional epoxies in fFame $ retardanoy as measured by flame height, degree of charring and * weight loss. &' Resin Curing A brief discussion of the mechanism of curing epoxy resins should afford the reader a better Insight to the considerations Involved In a curing study of ep-'xy resins. Amine cures Primary or secondary amines react with the epoxy group giving a secondary or tertiary amine and an a-hydroxy group. R-NH2 + CH2-CH-R' OH RNH-CHo-iH-c. R~>-NH + CHo-CH-R1 V/ C7 -- RoN-CHo-CH-R The nitrogen of a tertiary amine can fur: nor re a.: t wl: h an epoxy ring to form a quaternary ammonium compound followed by chain propagation. Polyfunctional primary and set. r.iary amines can react as above or as cross-linking agents tv toe foil owlng reaction: H2NRNH2 + 2CH2-CH-R' R1-CH-CH-CHc-N-R-N-CHo-CH-R T I l I d\ OH H H OH If R' groups in the foregoing reaction contain other epoxy groups, a cross-linked, thermosetting polymer obtained. From this brief discussion It can be appreo ed that a long chain will be forme! if the ratio of diamir. c epoxy Is stoichic metric (ratio of available amino hydrogens epoxide oxygen Is 1:1} and that less than the equivalent quan v of diamine would result In gaps in the chain. Furthermore, if the equivalent amount^ of diamine Is doubled,! each epoxy group could react with an amine group without any coupling involved. ; i pthpi? j'^acfIons which in|oildl| can as slsti jriqg* in accomplishing . at. J -1* cure Include J-j fer c HI ini I Ir'E; e r. DSW 621290 i-w ww- ~ t - -?1 STLCOPCB4095320 "S' Sir? rrpmjthe in< peifl * 4 f ` ' * ~ : " ? .... _ _ ^ Jp'-- : f 7 Curing generalizations - ... !?amino* grpup Ireacta with ar epoxide group at room temperature |* : ' ' - Polyfunctional amines by themselves, or those produced in the initial reaction, can couple epoxy resins. "1 At elevated temperatures and in the presence of certain amines, other coupling reactions such as those involving the hydroxyl groups can occur. Dibasic acids cure epoxy resins by esteri fication . Coupling can occur if two hydroxyls are si iated on different oolymer molecules. Acids may also -eact with an epoxy group b; icnating a hydrogen wi'-.h the formation of an ester linkage. The total cure with an acid anhydride probably involves bath of the above mechanisms. In general, acid curing requires higher temperatures than amine cures. A study of the variables in the curing of epoxy resins involves mar.;- unknowns. Anhydride-type equivalency (per epoxy equivalent) is about 85# of theory, due to epoxy hydroxyls which, as they form, serve as reactive centers and account for about 15# of the reaction. Amine equivalency appears near normal generally, tut equivalencies can deviate from 65 to 125#, depending pri marily on steric factors. BF* adducts, t-amines, polymeric curing agents, etc., all require detailed study. Ir. general, in regards to the variables mentioned, Aroclor-epoxy resins resemble commercial epoxies. Experience gained in curing studies of commercial epoxies can be applied to the Arcclor- etoxies within reasonable limits. | The presence of the chlorine in the Aroclor-epoxy resins apparently' does not greatly alter its general reactivity with a wide variety of common curing agents for epoxy resins. However,! in some instances the curing aments for Aroclor-epoxy resins ape too reactive, so that insolu bility occurs before cross-linkng| cant occur.^ f ; # t if f ' f faj r f I ftIf If i LAroclor-enoxies containing some monofunctlonal! species ? Large mumber ,of ci lied ising .a&varleti DSW 621291 STLCOPCB4095321 prf iL '$ # A ... T phyl ethyl ketcrnis) was* a<Jdea|`jp6 th# Solid Arb'clor- witE the curing agent, and then evaporated prior to curing. ; This always led to superior oured produots compared to that obtained from a dry blend of the,reactants. , v | . A low melting epoxy resin (92C) prepared from hydrolyzed Aroclor 1268 (90^ dlfunctlonal) was cured with 6-15jt by weight of various polyfunctional amines. Optimum cure tlme3 were de pendent on the size of the mass, the rate of development of exothermic heat, and the dissipation of this heat. Best results In terms of "high softening temperature products" were obtained when using 6% triethylene tetramlne, dlmethylaminopropylamine or dlethylamlnopropyl amine. As the concentration of the amine was Increased to 10#, the softening temperature of the cured product was significantly lowered. This was probably due to a high aralne/epoxy equivalence, leading to a lower degree of cure. With the development of a reliable oxlrane-oxygen analysis for Aroclor epoxies, curing studies could be performed on a quanti tative basis. The primary variable for study was then reduced to the effect of epoxy content on the properties of the cured -- Droduc t. Our early work had Indicated that with maleic anhydride curing systems, the best results were obtained with Aroclor-epoxies of fairly low epoxy content and high hydroxyl content. (Such materials contain considerable amounts of the Di-a-glycerolmonochlorohydrin derivative (GMH)). Prom the stoichiometry. It appeared some additional reaction must be occurring since a large excess of anhydride was required for curing. If stoichiometric equivalence of anhydride to epoxide oxygen was used, abnormally long curing cycles were required and lower heat resistance was exhibited by the cured product. It was concluded that In these low epoxy resins tne excess an hydride must be reacting with the hydroxyl present and that the products resulting from excess anhydride was not some phenomenon unique with the Aroclor-epoxy system. Duplication of earlier Aroclor/polyfunctional amine curing studies indicated several problems with Aroclor-epoxies. As resins of increasing epoxy content became available, the ex pected Improvement in properties of the cured product, did not (materialize. Aroclor-epoxies with epoxy'contents approaching* those of commercial resins behaved quite differently In the sense that extremely brittle products were always obtained. Jo y&rlfLtlons in coring agent-andicurln^jtime appeared yto *alt * re amp m v DSW 621292 . f y-f ! t W * r - STLCOPCB4095322 jj^^sIderations al-e- -rivative--s in... the --* .yCeCt Annrojuciluorr were: tditha poor properties of (the; epoxy by aotins' as 4,' torsiS I ' i ? , i, 4- i ' * '*&&! I ' * .* I ? < f' -1 ' ` - ` ;* Difunotional Aroolor-epoxle3 / t- ' '- *J:.'-.' i`` In order, to verify the pre dicted importance of complete dlfunctionallty in Aroclor-epoxy *5- |resins, an evaluation was made on the bisglyoidyl ether of 4,4*-bi-(2,6-dichlorophenol). This model compound (made via chlorination of dihydroxy biphenyl) should be free of the mono functional, chain-terminating epoxides known to be present in the early Aroclor-epoxles. A model-compound epoxy, containing 91 556 of the calculated epoxy content and very low (0.4156) hydrolyzable chlorine, exhibited excellent resin-forming properties with several curing agents (phthalic anhydride, p-phenylene diamine and HET anhydride). Best results, in terms of high heat resistance and strength, were obtained with epoxy compound cured with HET annvdrlde at 85* stoichiometric equivalence of anhydride. It was noted that as the epoxy content of the model ;cmpound increases the melting point rises, probably a consequence of the increased molecular symmetry; a melting point of :83C was found for 6.7* epoxy material. It should be recognized that this high melting point imposes some limitation as to the type of curing agents one may use. Many polyfunctional amines used in conventional curing formulations are volatile at tempera tures considerably lower than this (l85C). For this reason, the curing agents used in this study were restricted to higher melting solids - particularly anhydrides - which could be dry blended or formulated with the resin in the melt. Curing studies on an epoxy resin derived from completely dlhydric Aroclor 1268 containing 83. 556 of the calculated epoxy content and very low hydrolyzable chlorine (0.1756) were en couraging. This dlfunctional epoxy exhibited excellent resin forming properties with several curing agents, particularly iphthalic anhydride at 85 to 100* stoicnlometrlc equivalence of anhydride to epoxide oxygen. With the preparation of a larger size baton (3.5 lb.) of cora- pletely dlhydrio Aroclor 1268 (5-42* oxirane oxyg>een and 0,1856 on larSer 3lze oasjtfngg.4 suiittaabbllee forj fphyS&sl measurements, could be made . This Aroclor-epoxy ex- ' |hib^ted excellent'resin-forming properties withiphtnalic . v ' idrlde; Ibls.a&L-lfrie or. methylene . diamline* j. Compariao, illline afadibilH if DSW 621293 STLCOPCB4095323 Iigible"affipi6t firistanoes. Heatfa latgrtion temperat Aroolor- "1--; I?C whenf cure? ^biaanlllne^il..^ oa. Lthan the distorfcic epoxy cured' In' `'"liar fashion.^*" * srature forfth - -V :Pi Adhesive' , Lee ;I - fi &h Aroclor-epoxles containing monofunotional species Early experiments indicated that the Aroclor epoxy resins Sid not possess the inherent adhesive characteristics normally associated with conventional epoxy resins (Appendix C, Table 4). Our initial study on metal-to-metal epoxy-adhesives was directed largely toward obtaining good aluminum to aluminum bonds with the Aroclor epoxies. Efforts zo obtain the necessary shear strengths to meet military standards were unsuccessful. The highest shear strengths obtained with the Aroclor-epoxy were ca. 400 psl as compared with 2000-5000 psi for commercial epoxy _ resins (Qov1t minimum is 2500 psi). Several experiments were made, designed to "build-in" additional adhesiveness into Aroclor epoxy resins. A formulation containing Aroclor epoxy (ca. 40# by wt.), plus liquid phenolic resin and hexamethylene tetramlne did exhibit adhesion to wood, paper, and aluminum. However, the shear strengths obtained of aluminum/ aluminum bonds (200-400 psi) were far below target values. Fifty parts by weight of Aroclor epoxy resin (FD-4) was blended with commercial Epon 834. The two components are compatible in all proportions. We hoped that by using mixtures of Aroclor epoxies with commercial epoxies that synergistic effects of adhesion to metal, high heat resistance and flame-proofness might be realized with the cured product. It was found that the shear strength of an aluminum/aluminum bond of Shell Epoxy/Aroclor epoxy mixture decreases (almost linearly) with the amount of Aroclor-epoxy .. present. (Appei^lix C, Table 4)..f . i | i , | t| i ;< To an epoxy Aroclor (218392) was added 33# by weight of finely divided aluminum" powder. This was dorie by melting the epoxy . r and stirring 5in ^aluminum powder until 4a paste-like consistency* was obtained.! SjT'l'bp:^ttjhis mixture was |iaddd<e^defddji,mdiemthe3trhlaymla:mJ inopropyl amir^|afe|a!!fifr|ip;fagents iThe.i..m.....i..x...i.u*rl^f'al ithen used 'as faibj^ondi agent/for aluminum| Very poor bond strengths were obtained (Appendix. .C| |Tat|le,l4)_ Slight impijoyements-were, obtai J ,M [jth BD Q samids, 1 l; DSW 621294 STLCOPCB4095324 Lor-epoxy , yhe adhesive character ht g-model* (was investig_ated. This completely- ictZonal? epbaSMBEB^fyl^^roa '4,4* -bi - (2,6-dlc hlorophenot] fcaimed 91.5% ofl^i^alaulated epoxycontent and very low*( hydrolyzable ohlQ^uvs and exhibited good resin-forming propertiet with several different ouring agents. However, no improvement j in metal adhesioniwas obtained. ; f::# X / t :H - - :-Z. 7 !! $ Difunotional epoxy resin derived from Aroclor 1271 and Aroclor 1268, like the model compound, exhibited poor adhesion to aluminum. However, the epoxy resin derived from Aroclor 1268 (5.42jtf oxirane oxygen and 0.18# hydrolyzable chlorine) appeared to have fair adhesion to glass and porcelain. Equipment limita tions precluded actual measurement of these values. The shear strengths of the aluminum to aluminum bonds were ca. 110 psi" with phthalic anhydride cures. Some improvement was obtained with bisanlline A cures in which values of 550-600 psl were measured. These values are still far below the 2500 psl target strengths. These results cast a pessimistic outlook to the use of Aroclor epoxy resins in an unmodified form in any adhesive applications. Because of the limited time available for our study of completely difunctional Aroclor-epoxy resins, no study was made on formu lation variations that might be expected to lead to improved adhesion. However, on the basis of data obtained to date, the use of Aroclor epoxy resins as metal adhesives does not appear promising. Casting and Potting' Applications t. Aroclor epoxies containing monofunctional species Attempts were made to translate findings from curing studies development or casting and potting applications. to Miniature electronic-components were encapsulated using a variety r of Aroclor/ouri'ngfa'gent systems. In most instances the residual - stresses developed, during curing were sufficient to crack and rupture the castings} Variations in curing conditions did not . change the^inherentjbrittleness of the final products. Similar l effects Jwere no^;eaSS.n hi'n. films and glass laminates prepared ; - \ 4 from i theselArlcBLb^lt^oyv^e sins :l * the cured products exhlb|Lt&)| ft i ! very poor strei^tlj^ndf were *too brittle! for testing . ' '*? f * f tL* u DSW 621295 STLCOPCB4095325 IW&t: I: sTdbbei ____ tof&li^res^na < f^~ -- "' "~ Metal?InletsV in&o^ppratetf int| were firmlys anchored with no vlBdaElp^p|di| noted contacts. However, metal adhesion lcraddltlon tplatrej^thf Is a necessary prerequisite. The earlier results indioatad|poor adhesion; of/ Aroclor ,spoxy to metal mlgh^restriotitsusc fin this type application. This Is particularly true ;in electrical applications where Incomplete adhesion of the resin to the potted component results in minute "free spaces" in the casting which can lead to electrical break-downs or performance varia tions . This area of application would require considerable additional study before the potential of Aroclor-epoxy resins could be fully evaluated for potting use. : Hydrolyzed Aroclor as an epoxy curing agent Curing studies have shown that the hydrolyzed Aroclors themselves ef- fectlvely cure Epon 828 to give hard castings. These castings have Improved flame resistance similar to that obtained when HET-anhydride is used as a curing agent. Our data are listed in Table XXI. - `' TABLE XXI (NBP - 342792) *' -! ' Epoxy .* Curing Curing Agent Agent Soluble , 1 Comment - Epon 828 Hydrolyzed 20 Yes-700 slowly: aU-19O-20O0 - dark "Aroclor 1268 r' ' - red, lb'M.ttle''.^ * ? & ? 33 - rapid, . 19p-2pQ dark red ' : brittle^ !rvon^f|.ammable. ; 37 , Yes } ; cure3 slowlyt at^.^0rl60C . 'hd-ciire^ragfe^^^QrljS hrs(l | *f cure 3 sl'owlyl a^pQCpC * he rmopl a st ic.vj icxire * * ' DSW 621296 STLCOPCB4095326 fand ` the other? {PPlyfttiMtjOTaB?ii>h^bl8 IsmnQtaWartfS^We Attribute its higher sc iwp^ediy. lfere^^<iiHH$lro. .'The pH value a of #*. igfj'the |three|l|lpheno]S|tested are ;as;followsti ; Iff fl J ;;?''#! i pM . M&& ' I" ' ; Bisphenol A . *'.' 3.58 'r Tetraohlorobisphenol A 3.49 ? : ' Hydrolyzed Aroclor 1270 3.37 The values were determined In 1-1 dloxane-water solution, us.lng a glass electrode. Less than stoichiometric amounts of hydrolyzed Aroclor 1268 could be caused to react with and modify conventional Epons. I If Aroolor-Bpoxy Resin as a Stabilizer The epoxy resins have been used as HC1 acceptors and stabilizers for chlorine containing resins, particularly polyvinyl chloride. The dlglyc'dyl ethers of tetraohlorobisphenol A and hydrolyzed Aroclor 1268 were equal to Paraplex G 60 and Epon 828 In stabilizing polyvinyl chloride plasticized with trlcresyl phosphate against the effect of heat and light. TABLE XXII PVC - 70 parts Trlcresyl Phosphate - 30 parts Stabilizer - 2 parts Weather- Heat Stability Ometer Outdoor No. Stabilizer 5 Hrs at 150C 100 Hrs. 2 Weeks 1 Paraplex G-60. clear-reddish brown dear- very slight water white bleach Epon 828 - |? H equals 1 f' TetrSShlorob'iSphe hoi^ t ; Afdiglycidyl ether < equals 1 not as good as 1. i* h Ij equal DSW 621297 II STLCOPCB4095327 a* r j, -------- E* II' ft i ***** f * ' ;. % -' \ ^ ^,, , .,, HYDROLYSIS V; 1' The experimenaf ^eta"iis| on hydro] _ ^ previously^reported (Ref. l) and are repeated here fol of completeness. 1.5-Llter Batch (Run No. 14, Table I) Into a 1-liter Erlenraeyer flask was placed 144 g reagent grade NaOH pellets (3.6 moles, 80# excess) and 400 g methanol (C.P., Carbide 8c Carbon). The temperature rose on shaking, due to heat of solu tion. After a short boiling period on the steam bath, some makeup methanol was added and the mefnanollc NaOH was decanted Into a 1.5-liter, 316 stainless steel bottom-stirred Aminco autoclave. The residual NaOH was dissolved in about 35 g dis tilled water and this solution was charged to the autoclave. Monsanto production grade (Lot 86) Aroclor 1268 (224 g or 0.5 mole) was charged to the autoclave which was then sealed, flushes out thrice with Ng (to 300 psig) and heated with stirring as follows: Time 0845 0945 0951 1001 1016 1101 1308 Temp. C 35 173 181 190 192 193 25 Pressure Psig__ 0 345 420 510 550 605 40 Remarks - Both heaters on Fixed heat off, 97 volts on variac Slight cooling required Cooling water on if The bomb was vented slowly without agitation, opened and the dark solution poured out, and the bomb was rinsed out with water (ca. 200 ml), which dissolved the solid NaCl. Distilla tion to a head temperature of 98C yielded 423 g of metlianol plus water. The thick residue was diluted with distilled water to 450 ml and filtered on coarBe filter paper at about 50-60C, to remove (the small amount of solid. The solution was fyeated to boilings and acidlfied^atreflux, (Ref. 3) with5 190 .`mil cone . HC!E (fela^CTtfgrade) |thenfic| ml| exti'a sHCl was added find-?tiding M Mil continued 'at freflux If or about 30 min." ;The still*acid mixture ; was allowedvto stand overnight and: decanted. The lower! layer iwai, xsM<mMth i5l xjjOQ. ml>|poiiti^ris|.c^|h(^ti (50C Lstl3.1e/ DSW 621298 ref-i STLCOPCB4095328 IdlBtlllpt --^pot temperatt V^ll^ waa iain- MtfjO miiwftei Raaid5#edlfied?:193|8'Ma 9V *' fyield. ;S 2-Gallon Batoh" (Run No. 17. Table II) ? Into two 4-liter Erlenneya^. flaska were divided 100b g (25.2 moles or 80J< ex cess )NaOH and' 2800 g methanol. The mixtures were shaken, heated gently on steam baths and, after addition of some makeup methanol, combined. Most of the remaining NaOH was dissolved In about 250 ml distilled water and all the solu tions plus undissolved NaOH were charged to a 2-gallon, topstirred Autoclave Engineers autoclave. Aroclor 1268 (1568 g, 3.5 moles) was then charged and the autoclave sealed, flushed with 3 x 300 psig No and then heated with stirring as follows: Time Temp. Pressure C P3lg Remarks - 1130 1205 1212 1225 1227 35 116 144 166 1228 1235 1245 1248:30 178 181 187 190 1255 ' 1300 1338:30 5 193 ; 192 192 atmos 50 100 200 400 425 500 525 600 600 625 95 Volts on both heaters Cut back to 90 volts Fixed heat off, exothermic reaction to 186C Intermittent heat on Cracked cooling water inlet valve for a few seconds to control temp. Cooling water on The bomb was vented through a dry ice-acetone trap in which 30 g of a low boiling material, probably dimethyl ether, col lected.! The ^bomb was opened, the batch was siphoned out and the reactor was washed out with one" liter of distilled water. The methanoliwas distilled to a head temperature of 94C and 2565 g of|methanoland water (17# water by Karl Fisher titra tion)*, recovered.f The batch was diluted with water to a final If volume ipflabouta 5 liters and- filtered Ihot on poarse .filter : , paper!. ffiW^efjPsmali ^amount; fbf *s6flLTi&iterlLaly wefsfrelnov^ea; f i The flltratfetwaa fac ldlf led ;carefully at about 95-lOOpC with aboutfc iJSfiJal fcojlc,. |HC1| Iff then 25 mlfexcess^acid .was padded. i.2i ce L DSW 621299 STLCOPCB4095329 distXllef tire, halide( te| 100 ml benzerieli benzene distill* the residue was| yield was 1300 g of ! tned on|t. 'las6$|er wan J'the wafeffwas alfCc^fopjei jjl temperature ol _ . . IHm for about 30 minutes, * " * ' i-.i? fry iiri aonvf I* & GLYCIDYLAT ION l/a n t. Expt NBP >67 489 DihydroIyzed Aror'ii-r 1268 (0,5 mole, 206 g, N.E,, 204, made by J. R. LeB'.anr NEP 366970) and Shell epiohloro- hydrln (6.0 moles, 556 g) were placed in a 1-liter, 4-necked ( ' flask equipped with a mechanical stirrer, dropping funnel, eon*-: . denser and tnermometer. The mixture was heated to 90C, then; 0 5 mole of NaOH as a 50% aqueous solution was added over.a : ! 14 minute period A mildly exct nermi-t reaction took place and brought the mixture to ref: ;x In 3^ minutes, the solution be came neutral. Excess epic hi roi.ydrin and water were distilled at 20 mr pressure, and at a maximum pot temperature of 100C.. -- The hot residue was diluted wltn 230 g of benzene and one mole., of NaOH. as a 30% aqueous .eolation (a 50% molar excess) was added- The mixture was refluxed wit--, stirring :or 7 5 hours and water was removed azeotropical1y The solution was diluted with benzene and washed to neutrality wim . 8-Iiters of water, at which point, tne neutral water wash remains clear on acidification. The benzene solution was dried by distillation, then the solvent was removed Final traces of benzene were removed at 120-30C at 2-4 mm, or at 150C at 20 mm. for 20 minutes The product , - ,, (221 g) is a light amber resin, m p 46-7C. The yield is '* 84 5% based on dihydrolyzed Aro .'.or 1268 charged. ' Anal Founds % Epoxy t 00 % Hydro . y zac-1 e Cl 0 11: 012 % Total C: - 42 9*- 43.16 Expt NBP 367491 This is a tnree-fold scale up ex^perl\mke<nnL] - from tne previous-jWith tne dif fereu' e that a lower (ls6). hydrolyzed Arcelor 1268 to epic nlorohydrla molar ratio was used*I In this run, the Sodium ahloride fortred was, extracted with witter; before ther excMsjfejplsdilorohydrin used waa distilled off?. K separation^was p'aroffit'*'T^e residue >was dllpitedi with benzol treated .with the xcess Sodi um nyarox ide'as* in'previous expel ment then was; Hashed, to neutral ity wi th watgr j(emulsion): jjqs.feive j epoxy.and DSW 621300 STLCOPCB4095330 .... , ..... _ -as;Sl#^ti'iSaxfbn bf ^ the guls prooesf Ref|, 2^| Bihydr^5!|*aftlAi?oolor 1268 adefbyi J. R.iLtBMicfe|?65970) 1$ # *^eal556n||^fc^|feed at 90c. P j Tpithfswas addedf] ^of NaOH as a 50J f&queous solution over ^ a 17t minute period ( [idly exothermic reaction took place after 7 minutes when*'4.,. .il NaOH had been added, and the mixture refluxed. In 40 minutes^) the solution becamerneutral. Excess eplchlorohydrln and water were distilled offf at 20 ram pressure at a maximum of 100C. ; The hot solution was diluted with ben zene (4.4 moles, J40 g) refluxed, and 2 moles of NaOH was added as a 50% aqueous solution (a 50% molar excess). This was re fluxed for 6.5 hours as given in experiment 367489. The solution was diluted with 400 g of benzene, and the residue washed with more benzene. The filtration was slow and the washing was In complete since the filtration residue contained 57 g of water - Insoluble material (glycidyl ether). The filtrate was dried by distillation, then the solvent was removed. Final traces of benzene were removed at 1-2 mm at 100C. The product (442 g) is a light amber resin, m.p. 49-5GC. The yield is 84.556 based on dihydrolyzed Aroclor charged. The material 13 slightly alkaline. Anal.. Found: * Epoxy - 5.33; 5.47 % Hydrolyzable Cl - 0.07; 0.09 Alkalinity - 0.013 MUllequivaient NaOHgr, resin The identical experiment (NBP 367467) made at a 0.25 mole scale gave a 92.6% yield due to more complete washing of the filter cake; this cake contained 4.4 g of water insoluble material. The final product analyzed 5.44^ epoxy oxygen and 0.10% hydrolyz able Cl. ,% ESTERIFICATION Our esterifications follow the method recom mended In Reference 8. Expt. NBP 354873- ^ | i- -2 . li * v sEsterificatlon of oxirane oxygen of Epon 1004 Z: Jr . , . . ) The reaction was carried out In a four-necked flask, fitted with a stirrer, ga3 iplej Itube,ivent tube, and thermometer. Nitrogen at approximate lyO tc&2.bubbiesj/second was. passedj|through the reaction mixture! HI'ft * MH M ii 15p gms Ep^i|MQffi|(Shell) Chemical Cq.)| DSW 621301 STLCOPCB4095331 The reactants were hea^edEf^o &OC^I^^^tf fj^3vAtTtnI*g<SMI ture until a. low acid |raluS was njMed. |/This ' requiret^^ 5 minutes to reach an aold|vaiu^i^^i.l6.f This was '3 wgii: 5 minutes after reaching 200C. fipi^estaV'was then co<ap dissolved in an equal weight of xylene. Viscosity - 1005tcerit poises (Z5 - Gardner-Holdt), Color 7* Expt. NBP 354868 Esterification' of oxirane oxygen of dlglycidyl ether of hydrolyzed Aroclor 1270. 125 gms Dlglycidyl ether of hydrolyzed Aroclor 1270, 5.8556 Epoxy "n" value O.76 84 gms Soya Patty Acid (Armour Neo-Pat 127). Initial acid value =* 8l The resin and fatty acid were heated to 200C. After a total time of 1 hr, 55 min, and 50 min after reaching 200, the acid value was 1.47. After this the acid value started to rise. At a value of 2.25, the ester was divided, one-half was dis solved in an equal weight of xylene and the other half heated for a total of 5 hrs, 15 min at 200, and 8 hrs, 5 min at 260. The acid value finally rose to 20.0'. The off gases were scrubbed to remove HC1 during this experiment; no indication of chloride ion was found in the gas scrubber. The viscosity of the portion heated to 20CC was 55 centipoises, GardnerHoldt N. The color was 7. Expt. NBP 559094 Esterification of oxirane oxygen of dlglycidyl ether of hydrolyzed Aroclor 1268. 50 gms Dlglycidyl ether of hydrolyzed Aroclor 1268, 5.57# Epoxy oxygen, "n" value = O.85 29.5 gras Soya Fatty Acid (Armour Neo-Pat 127). Initial acid value -57.0 The previously described cycle was followed. After 1 hr, 10 min total, and only 15 min at 200C, the acid value was 2.15. During the next 50 minutes the acid value started to rise. The-ester was cooled and dissolved in an equal weight of xylene. . - < 4 * Jr! i t i. ` | t I 1 I | i ti -if it i !- * t S i Kli ii i ' i i I Expt. NBP 564202 Tall fatty-ester of Shell Epon 1004,* * I * ? ? . h I 0.9 equivalent. ki&ki: | fe-tefc'j,i t J.l 1 1 j| hi 1 , DSW 621302 1 STLCOPCB4095332 H ;?he reaction wasjoI ati 280?C'.* 280C), the ae id' value'1 was'* 12.5*. The realatIon was 'slopped^ a^^ % tfila point and the^eater dissolved in an*equalt^weight of xylene; ni % M* Expt. NBP 36420^ Esterification of diglyeidyl ether of hydrolyzed Aroclor 1268,with Tall Patty Acid 0.91 eater. 150 gins Diglyeidyl ether of hydrolyzed Aroclor 1268, 3.3J< Epoxy "n" 0.88 143.8 gms Tall Patty Acid (Arizona Chemical Ac Intel FA 2) Initial acid value - 98 The usual procedure was followed. The temperature was held, at 220C. The total time to an acid value of 9.6 was 10 hrs, 10 min, of which 7 hrs, 50 rain was at 220C. If the reaction had been stopped at an acid value of 13.8 comparable to the 12.5 of the preceding e3ter of Epon 1004, the time would have been 6 hrs, 25 min at 220C. The ester was dissolved in an equal weight of xylene. Expt. NBP 354899-4 Preparation of polyester. 4.2 gms Hydrolyzed Aroclor 1270 diglyeidyl ether 353864 1.3 gms Benthal 0.5 gms Fumarlc acid The ingredients were heated on a hot plate 15 minutes at 400F. The product was cooled and ground In a mortar. 2.1 gms of above resin was dissolved in 0.9 gms of Styrene. One drop of cobalt naphthenate and a trace of benzoyl peroxide was added. The material was heated in a hot water bath to yield a rubbery resin which became hard on standing overnight. This material would burn in a flame, but would not support combustion. Expt, NBP 359094 Preparation of polyester. f$ | 4.17 gms Diglyeidyl ether of hydrolyzed Aroclor 1270 1.29 gms Benthal ; 0.25 gms Maleic anhydride ; ,0 i 30j gms ,Ph|haldi4|^^lc|g * |. * f- f f - i f || 11 Heated on hot-plate 10 minutes and dissolved In 2.4 gms of j ' ` olufcloiii Wlthtcpha.lt naphthenat* aMohl cats 111 DSW 621303 STLCOPCB4095333 jv s-PS Eater of _ r 7ol * 1 * i Aroclor (with Soya Patty Acid and;D1 i f V % f:% did.* i .. . . : '* i?55 gs Diglyeidyl ether of hy . , red; Aroolor 1268--'* ' '40 gms Soya Patty Acid (Armour^Ndo-Fat 127) 5 gms Dimer Acid (Emery - M461-R) ' Materials were heated at 260C for one hour to obtain an acid value of 13.24. The acid value did not'fall any lower but began to rise. The final viscosity of this material was a Gardner-Holdt Q. CURING STUDIES Melt Mixing Curing studies on Aroclor-epoxy resins were per formed on a 5 g scale by several different techniques, dictated largely by the solid state of the resin. In the melt technique, five grains of resin were added to a test tube (which had been well lubricated with silicone), melted and the curing agent added. This method Imposed several limitations; one, if the melting point of the Aroclor was higher than that of the boiling point of the curing agent (which is often the case with liquid amines), the curing agent would be partially flashed, and the stoichiometry of the curing reaction would be unknown; two, at the temperature required for melting the Aroclor epoxy, addition of the curing agent often led to violently exothermic reaction. In addition, at the high mixing temperatures, the curing agents are often so reactive that insolubility occurred before sufficient reaction time ensued. For these reasons, melt mixing was re stricted to slower-reacting, solid curing agents such as acid anhydrides. Dry Blending with Liquid Curing Agents In instances where a low-boiling, liquid curing agent was to be used, it was dry- blended with the Aroclor-epoxy, at room temperature. Although this method leaves a lot to be desired in terms of dispersion of the curing agent, it did lead to fairly satisfactory results. . *S ' An alternate procedure was to dissolve the Aroclor-epoxy and * the (cuijijig *|tgent' in a mutual solvents, (evaporate* the ^solvent?, f * and .'recover!the mixture. One difficulty! encountered in this - .method lis/finding a sufficiently low-boiling solvent which can Itie^ayanointed under icondltlons) in? which no pre-cure (oftthe I l^sfeiMMiitiMke lid)* m I jjyii ill DSW 621304 V':'V STLCOPCB4095334 igent waa.^phyaioally thej^lWdeslrabltT foi,p. ... l$,, e.poxle#> f** itHa|jBiiring^gertare first ground;. ;through :60pieah' sc|*oj|hs .aud^hen|d^-hieuded via tumbling. , ' Solution miking wai`falsa Inrestiga^l^wjSihjaolid ouringfagehts^ however/ the difficulties mentionedfa^ve/In solution blending/ are similarly troublesome. '*\ Irf | f ' ` 5$fp ;J ' 4i$' A ' 'i > fI I ; The Use of Reactive Diluents The use of reactive diluents as both solvent and curing agent, such as allyl glycidyl ether, etc., was not extensively studied. Such reactive diluents are known to lead to lower heat distortion products. Since heat resistance was one of the expeoted "plus" values of Aroclor epoxies, use of such diluents could effectively mask this important property. r Curing Conditions Optimum curing temperatures and times must be determined empirically for each curing agent. In general, acid and anhydride cures require longer times and higher tempera tures than amine cures. The addition of catalytic amounts (<1.056) of amine will sometimes accelerate anhydride cures Evaluation of Cured Products Our criterion for cure, adapted to a micro-scale, was the behavior of the cured product on a melting point bar and its solubility characteristics in a series of selected solvents. In general, a product which exhibited no visual change in appearance at temperatures up to 300C, and re mained unaffected by the selected solvents, was considered fully cured. Most of the early Aroclor epoxies (found later to con- i tain monofunctional derivatives) led to cured products which were very brittle in nature and "soluble in" or "attacked by" ` a variety of solvents. Such behavior is not associated with high molecular weight, highly cross-linked epoxy resins. In the evaluation of the completely difunctional Aroclor-epoxy, larger size castings were made, from which test specimens for heat-distortion temperatures were ct. ^The measurements were | made In accordance with A.S.T.M.^procedure. :* f f- Adhesive Studies :Our method consisted of lap-bonding aluminum . strips (4"1 V X 1/ 16") which havelfir^t been degreased In ' I I trlchlprof lenefin!. then etched$ ' " " aoid/d' iohronia|;e^4 | f solution. [The htrip;* are lapped.??1-Irion ftojjglve a 1-Inch .square^ 1 ` test areal The shea strength of the, bonded aluminum is measured, r on th Ins foot | DSW 621305 STLCOPCB4095335 J, If-t Ms. Iwelghing;^ and^n(3apsulatibn*xp^rj gd&by&.embeddjng Miniature electronic :part|s wilj^" ^T-apw'jyj^pmulatiSus. i Silicone -coated aluminum fwerei'used for molds Li 0X: Qlaa8 Iaalhatlng Both wet and dry lay-up procedures were employed in;preparation of glass laminates. t ^ Sr <r- In one lnstanoe glass cloth was oo.vered with a dry blend of Aroolor-epoxy and curing agent, and a sandwich structure built up by adding alternate layers of glass cloth and epoxy mixture. The sandwich" was contained in a mold and cured. Another method consisted of dipping glass cloth into a solution of Aroclor-epoxy and curing agent, evaporating the solvent, and repeating the operation until a sufficient thickness of resin on glass was obtained. In both of these instances, the cured products were too brittle to be of any Interest. There was no opportunity to evaluate the completely difunctional Aroclor epoxies In this application.I fS; #fW t *t | i ' ! ` / f * ! L $ " " i * f 4t * ! ' I >; l m DSW 621306 . .-T^ I STLCOPCB4095336 "ff c $ -% * m ality. (D2472) tS?^i'S';%or ;rnr ia68- p"1?glve co"Plete phenolio difunction- 1 **- - - f- >-f - . ;n - : .. if - ' Patent disclosures have also been submitted for uses of hydrolyzed Aroolort the following 1. Tanhe epreoaxyctigornouop,f ehy.gd.roelyjzlecdj^Arorohcjldorrl^wi-hh(M869) ^ 2. Hydrolyzed Aroclor as Industrial preservatives. (D2078) 3. Glycidyl ethers derived from hydrolyzed Arocln r as (M32C Dj?2[)r PlyVlnyl ahlorwe and a plasticizers. I ofehydrro?yLtdLSolStty(D2335)3terS f dlSlyoldyl The preparation of casting resins by the reaction of equlmolecular quantities of ... cion 01 wlth^styrene(D23555 an<1 fw&ric acld and cosS-liSing The preparation of a thermosetting aoldd!g%13l6)ether f hydI'ly"!d Diglycidyl ethe^rrsd for epoxy resins; or\ ff* hydro1lyz__e__jd .Aroc_lo (D2477, D2506) . r as a curing agent (D2567) f catalyti0 counts of lime in epoxy reactions. cureUMghfaclfJa"fa?SdIS <= rH ft'!! I'M 1 M DSW 621307 m STLCOPCB4095337 E8 fit? RD-57-^in No/81061, by J. R. LeBlano, dated July;511, '1957. 2. j "H R. C. Cass, St. Louis Research Report No. P-762, "Suggested Procedure of the Preparation of Diglycidyl Ether of Tetra- ehlorobisphenol A." '` 3. Dazzi (1), "New Epoxy Compounds Derived from Aroclor", Memo to the Idea Review Committee dated 11/16/54. 4. Dazzi, "New Organic Chlorine (Aroclor) Derivatives", Memo to the Idea Review Committee dated 4/26/55. 5. "The Aroclors and Other Halogenated Aromatics as Chemical Intermediates," Memo to Dr. C. A. Hochwalt from J. Dazzi, dated I/I6/56; Part II - continuation, "Sulfur Containing Compounds", dated 12/14/56; Part III - continuation, "Nitrogen Containing Compounds", dated 11/7/57- ___. 6. Schecter, L. and Wynstra, J , Division of Paints, Plastics and Printing Ink Chemistry, Preprint Book, Vol. 15, No. 1, p. 280. 7. Letter - John F. Palmer to J K. Craver, March 28, 1957. 8. "Epon Resin Ester3 for Surface Coatings," Shell Chemical Corp. 9. "Studies in Fire Retardant Paints III," Official Digest, Vol. 28, p. 942. 10. "High Strength Epon Laminates," W.A.D.C. Report No. 52-5 Supplement 3. 11. Earhart, K. A., and Montague, L,, G., Division of Paint, - Plastics and Printing Ink Chemistry, Preprint Book, Vol. 16, No. 2, p. 200. 12. Kosmln, M., R&E Division, Research Dep; i "Hydrolyzed Aroclor 1268 Process." Iwf 5 J: f Wf!H TR-9, R&E 1957, Ink ii f- Jf DSW 621308 STLCOPCB4095338 i 4 - f^ f * f > ** * :, , f- * - .. t ,.i?i |. i *, * * I Epoxy Constants :? v I ' 1. \ * B. Calculations for Preparation of 300 Gallons of Epoxy . ,.; Ester (0.8 equivalent) % 4* ; 'I T'f ' * ^ 'if C. Key to Abbreviations Used in Tables Curing Studies - Tables 1, 2, and \ - , , Shear Strength of Epoxy Aluminum/Aluminun Bonds - Table 4 D. Biological Toxicant Evaluation E. Esterification Tables 6, 7 and 8 P. Notebook Pages G. List of Raw Materials H. List of Tables I. List of Figures J. Table of Contents ,* 3 - i h I i is ^ t ' ' lit I f -ti I t* i1 DSW 621309 STLCOPCB4095339 |||f^ tIE " , Epoxy Constants^ resins^have'r a number of values that I -serve to oharaoterizeftkKS "'resin. `" ` f" Percent epoxy - determined by analysis, reported as per cent oxlrane oxygen. Epoxy value - Epoxide oxygen content (# Epoxy) 15 Epoxy equivalent weight - Ep03,:y1Yalu7 - J Epfxy x 100 (grams of resin containing one gram equivalent of epoxide) Equivalent weight - grams of resin required to esterify completely, one gram mol. of a monobasic acid, e.g., 60 gms acetic acid. Calculated molecular weight - 2 x Epoxy equivalent. "n" number - Mols blsphenol-1, or Mols eplchlorohydrln-2, or Number of hydroxyl groups, or Calculated M.W. - Wt. outside bracket In formula Weight Inside bracket In formula Functionality - Mols eipchlorohydrln + 2, or _ J |Mols blphenolv+ 3, or r n + 4. ' Percent hydroxyl determined by analysis. ; < Percent hydrolyzable chlorine ** amount of chlorine, other than that attached to the aromatic nuclei, left in the resin* i . - ? ^ r ir?' h - : .Calculations for Epoxy Esters Ref, l) , oxlrane ester of the diglycldyl ether3, ! 1 follows: ;* ' ' 1 ' ^ Wt t ? M M f wtt For preparation of the the calculation Is as * Ivii - : f'16 ft s&Epoxy l! Epoxy Eq, W* 1 , w. . DSW 621310 -w \ ,,4 STLCOPCB4095340 Epoxy Eq. Vt. caloulatinglthefamlmnt of fat tv * A 4vi t-v, l . If the equivalent weight is known:' f ' T?** ^ | Required Patty Acid - Weight diglyeidvl ether ,, * Eq" Wt.-------------- --- * Wt. of Patty Acid mayILtr`olloied:lent W6lSht 13 n0t knoWn' the following .course Functionality - n + 4 by definition Calculated Molecular wt- n+4 ----------- = Eq. Wt. (calculated) Wt. cf diglyeidvl ethe" Eq Wt. (calculated)-1" x Eq wt- of Fatty Acid x Fractional amount of complete esterification APFENDIX B calculations * for PrGDap?*Jia-. n b equivalent)--------- ----------- ~---of ^0l- Gallons of Epoxy Ester '-aarge to kettle 483 lb Epon 1004 627 lb Soya Fatty Acid ^ Charge to kettle - ; 627 lb diglycidyl ether o30 lb Soya Patty Acid agitation. Brin^to^equi-ed temDe*ft-meltlnS iS ~omPlete> start acid value. requ,.ea temperature and hold for desired 260C (500F) Time for up heat -2.5 hours Acid value - 6-9 Time to jac|d|ja;l^'i|9 hours >I Cool - 1 hour|(to 350) spirits - ,*0.5 .{ f 5 I I I I I I I I I I I I STLCOPCB4095341 if * > m Keyfe "Abbreviations Vied in* Tables! f r t. f ill MA . Kalelo Anhydride TETA DMAP Triethylene Tetramlne . Dlmethylaminopropyl Amine ** DEAP Diethylaminopropyl Amine P Phthalic Anhydride PA Pyromelletlc Dianhydride DD Durene Diamine DET Diethylene Triamine PD Phenylene Diamine M Maleic Acid DADPS DA Diamlnodiphenylsulfone Dimethyl Aniline If M `\ t-f Ift4* H f -1t STLCOPCB4095342 o o H oc c O H> <L> C~\Q P U 4J = O rt CO CO to to WMWtowtoeonwcoM O<-i OmOinOmLOnmOLOnOmOmOmOm C\JrHr-tCVJr-|r-4I-H(\J^ H'H VVVVVV VVVVV ' I i M*tOC .-i- M'Pvw'JOr4 f O jz mo mo n --t o o' <~t cm oi --t *0= = = = = = = vViO i'. irT-*j 1> c" ff) r-'i r-l -i 1 - ffj 73 | o c "*1 o xl o moOmOK\OK\OmOK\OKNOm ^O OOOOOOOK", * 1 <--t i f--i <--i t-- --i f--i i--i i--i i--i *--i i--i i--t i--i cm c\j oj cm c\j cm cm O OOOO *n rr\ .a ,v\ .r -* .d- -=* -f -j < vu -'j c\j \i m i`j oi CURING STUDIES ON AROCLOR EPOXIES 4> . ,, <1/ o< <C <C < < CL, Cl. CL, CX. H E-* o f-. < C <C <! O >ij *r 1 U ;0 , J cl f~ C 0) ou r-l < 2C O in in in o in i_n t - inr-a> o a* o o in t j r in t/M ' `M t - O! a.' >- ij -i X c s: (c D o >M >* I! C nH Q o *r-1 0; M C 03 H >R x: X, Cu o sH r-i * r--1 x: <o <1--u c 03 d) u > X. X oo 3 ? IHI| iO Cl C\? .-i o, ( OJ A j ;. ml I iff I &A a-r DSW 621313 EUgT STLCOPCB4095343 STLCOPCB4095344 STLCOPCB4095345 STLCOPCB4095346 tiOE C i y i tam ni l CO" O' - - ^ pOjCQ bOCO CO CO Oft CO CO ft) v* CV>J aO ' * OJ CkVJvOg iCnVJ CiVnJ iCVfJ CVJ CVJ CinVJ CinVJ ioc|, VVv VVVVV/V . v ~f vo=f= = = =,= s! ooooo in in tn in in or JM cvj OJ CU CVJ CVJ bO C 4-> H C Jj OJ 3 iq O< 0<_4 < < < <ww as X X fr> E-< ooooooooocooooo inintnininminininminininin rtHHHHHHHHHHHHH CVJCVJCVjnjCVJCVJCVJCVJCvjfVJCVJCVJCMCVJCVj XXXXX XXXXa,a.(i,ais:s; <+ <+ <+ <4t.4c. XXXXX oooooooo in in m in m m m in HHHrlHHHH CVJ CVJ CJ CVJ CVJ CVJ CVJ CVJ ;<<<<< !XXXXX <--* CV -=f VOW4-OH O H Ol H OJ <5 in2 ?\ cC1 in 2 a, 2i ^ ^ ................................... _ * cvj r-i .*ni OOhoiOOhojQOOh CVJ in cu ao vo m mj inco c-- vo o, CM 0) > rv "J3L3 j^-> oX X0 Q CJ > &T o ' *f Nft. i K\ cvj in m f OJ in m -1 v' '*i 1! i?. #*. m| t--i co, <1) O rOO--J* OJ vO Ot--Ji r>-->< u biO Ho tio--H G* O ou O O 9 V < *< f? DSW 621317 r-y. y rm n<& f- STLCOPCB4095347 STLCOPCB4095348 STLCOPCB4095349 TABLEl m is if M: -r EFFECT OP CURING VARIATIONS' >N. ?* 3REPOXIES *t1* A t ~*m I Sample PPH of | Ciirlng Softening ,,Identl- Curing .. Conditions Temperature flcatlon Agent Curing Agentfr ^ * *Hrs~a c>C Off Aroclor 310448 12 1-Cyanoguanidine 12 1-Cyanoguanidine 0 .5 1 .0 1 .5 2 .0 2 .5 0 .5 1 .0 1 .5 2 .0 2 .5 113 113 115 115 115 150 150 15c 150 150 92 96 100 94 93 109 136 145 144 144 12 1-Cyanoguanldine 0 -5 1 .0 15 2 ,0 2 ,5 175 175 175 175 175 150 151 151 150 150 20 1-Cyanoguanldlne 0 .5 1 .0 1 .5 2 .0 2 .5 150 150 150 150 150 123 144 146 140 134 1-Cyanoguanidine 16 150 16 150 150 155 p-Phenylene Diamine 0 -5 .cf 1 .5' 2 .4 2 150 150 150 150 150 120 124 128 127 126 e124 n-Phenylene Diamine $f % ST>- 1, I DSW 621320 - rt STLCOPCB4095350 Sample f fPPH of Identi- * fCuring ficatlori iAgent CABLE' $ 4$* ? > i -i 4 If ' Curing Agent f Curing Conditions Hri C~ % * > % : Softer Tempera! oc Aroclor 12 310448 j; | 12 Aroclor 31CAMS (1268) 20 6 10 20 6 10 20 o-Phenylene Diamine 0.5 1.0 1.5 2.0 2.5 Olutaric Acid 0.5 1.0 1.5 2.0 2.5 Ethylene Maleic Anhydride 2.5 Triethylene Tetramlne 2.0 2.0 2.0 Diethylaminopropyl Amine 2.0 2.0 2.0 150 150 150 150 150 150 150 150 150 150 150 15c 150 150 150 150 150 104 ; 105 105 110 110 95 110 110 113 109 103 234 192 159 214 196 158 6 Dimethylaminopropyl 2.0 150 10 Amine 2.0 15c 20 2.0 150 220 203 150 Aroclor 316097 10 Maleic Anhydride 0.5 150 1.0 150 1.5 150 2.0 150 2.5 150 105 105 106 107 110 - : 10 i Ethylene Malelmide 0.5 150 id 1 .0 15c 103 1.5 150 105 -. if 3i M f !* ' 2.0 150 106 \ l4i^. ti 11 -1# * 2.5 fU 150 t1 -i rf 108 jh fe| i: -3^ oyt: ,-.l & ' DSW 621321 111 STLCOPCB4095351 STLCOPCB4095352 STLCOPCB4095353 STLCOPCB4095354 STLCOPCB4095355 STLCOPCB4095356 STLCOPCB4095357 g-llfW ft Gabbert - 370312^7 me f i &2W * " ~ t. ? H LeBlanc - 359388-9, 91-4,' 96-99, 356400; 365951, -2, -4, -5, * *-8, -9,.-63, -69; 365970, -1, -3, -4, -5, -6. *,: &<& ! - .i ffi--a . v- - ' $.* If ?5^ `7- ?'".e Schwendeman (-'354851 Ho 354881 inol., 354884 to 354900 lncl., 359051 to 359054lncl., 359056 to 359083 lncl., 359094 to 359100lncl., 364201 to 364207 lncl., 364245 to 364247lncl., 366951 to 366952 lncl., 36695575366972 to 366974 lncl., 366977-78, v 366991 to 366995 lncl., 373940, 373945. Grant - 367951 to 52 lncl., 367955 to 367994 lncl., 379803, 379080. Craver - 342751, 342752, 342792, 342797, 342798. Herblg - 322801t07, 10-13., 23-27, -32, -34, -43; . | 33.05017^-04,^-07, -08, -12, -15, 19-21, .? i |.,.336401-05-, -09-14, -18,; -48; ` *' 346552;1-78, -89; ' : % 37800^j 380715-16. v . 23-44, 46; t 4.* * *'I ? 1r ' ? |; Koch - 3425831; 92-93 - 1 | 5*-- "5 Dazzi - 218932;'; 310448; 316057, 098; .323108, 13I; 328956; i v I#ipP^i^|1g|04 346529; ^527 78; 353864, 8654 889, 896, ' f ? tff37?|#f 35|5^5 5f7,^5487 367467, 367775.1* .i i it * |;| DSW 621328 STLCOPCB4095358 >4 u i $l, ;LiS0':Ol if Arodi^^^^270/^^^68^and 1262 I jMolisalito' _ mm i&d Blsphendlff f W- *-' *s 5 Diglycidyl 'htiiep tetraohlorobisphenolr j Monsanto .. . A Experimental iSj-rdE * ^3i ? r Epon 1004 and 828 ' ' ' ' '; ' Sc""hevl'l*'' Eplchlorohydrln Soya Patty Acid Tall Patty Acid . / i: sheiks: I if Armour r s ' Arizona Chemical Methanol, '' . Dimer Acid M-461-R W#i'BeiTC * * * Allyl Glycldyl Ether Maleic Anhydride Triethylene tetramine ; .ShWll: , ' ' k Mbfisarito' Carbide Dimethylaminopropyl Amine Carbide Dlethylaminopropyl Amine Phthalic Anhydride Benthal , Fumaric Acid Styrene Pyromellitic Anhydride Durene diamine Diethylene triamine Phenylene Diamine Maleic Acid Diamino diphenyl sulfone Dimethyl Aniline f : Rezyl 1105; Carbide Monsanto Monsanto i . Monsanto Monsanto DuPont Experimental Carbide f Malllnckrodt Tenne 8 see-Eas tman Experimental , Eastmanf f ;#^^ | # , I '= ml American Cyanamld i ; rir vS ? I U I I *i DSW 621329 STLCOPCB4095359 ... Table V Table VI i Table VII $ f^TalJ'tviII i* |, Table IX, " * Table X Table XI Table XII Table XIII Table XIV Table XV Table XVI /.^Table XVII w , f- , \ Table XVIII KTable XIX "If f- i : ?atai XX L [ I , #able XXI , ; ? Table XXII l Mfu f/ ffi ' I 'Appendix | I $ i i s. , . i ' t-f-frable / |Ta^|y^ rable 1.5 Liter Autoclave' \ Hydrolysis|of sole 268, 2 Gallon Autoclave' ;Soya?P&ttylAd s Soya '^SttirTAe --J;'ssa-i-tt'ew|s_Sr3s-rTMJw S>f> Glyoldyl Ethers fi-Dlglycldyl Ethers5^^- of Hydrolysed Arodlor 1268 ' s Soya Patty AcldiEsters of Tetrachloro- ' bisphenolAji / . Esters of Dlglyoidyl Ethers (lOOjC dlfunotlonal) Esters of Patty Acid %/Epoxy Resins : 0.5 to 0 ^Equivalent ^ Solubility of `Epoxies on Xylene and Acetone % Cure of Alkyds.;-j| i 4'4 ' Dry Times for 'Varnishes"Made from Epoxy Soya Esters; Characteristics of Epoxy Ester Coatings Performance In Weatherometer Esters Used In Evaluation of Coatings Characteristics of Enamels Made from - Epoxy Esters Dry Times and Water and Alkali Resistance Properties of Coatings as Prepared Properties of Coatings After 400 Hours In Weatherometer Performance on Out-of-Door Exposure Appearance afteivWeatherometer and Out-of-Dooxf Exposure 1 Resuits ofiPlrel* 'es&oifertjl J ' t Hydrolyzed Arocljbr as an Epoxy Curing Agent Aroclor-Epoxyfast Vinyl Stabilizers 19 29 30 31 32 J if4 Curln ;Curin Studie s io * Aroclor 'Aroclor Epoxies Epoxy : ; i Effectf of Curing'Variations on Aroclor - \ fEpOXies^ i I i ; . : t Shear .Strength of Epoxy;, Aluminum/Aluminum I Luations "of |] |f | 75 ,78 | jfDiglyrldyl Ethers by DSW 621330 STLCOPCB4095360 ** Figure Figure Figure Figure Figure Lst^of figure! 4 /age 1:* ^ No-. Effect of Excess NaOH on Hydrolysis of Aroclor 1268 ,l Time-Temperature-Pressure-Data Hydrolysis of Aroclor 1268 | Cooking Schedule - Qlycidyl Ether of Several Biphenols plus Soya Fatty Acid Cooking Schedule - Qlycidyl Ethers of Hydrolyzed Aroclor 1268 + Soya Fatty Acid Unreacted Fatty Acid in Reacted Mix vs. Time at 200, Glycidyl Ethers of Tetrachlorobisphenol A Esterification as a Function of Time and "n" Value Cooking Schedule of Diglycldyi Ethers and Soya Fatty Acid Wood Sticks Before Fire Test Wood Sticks After Fire Test 1C 13 18 20 j2 2? 2^ 39 .... ...... ?H f ft IMS f *H*f I f v -i DSW 621331 STLCOPCB4095361 > rtt tj- | fable of Contents k if :Vw *> -i- . S&S ' ^ $ < i vf',4 Page No. i. PURPOSE ii. SUMMARY hi. CONCLUSIONS IV. RECOMMENDATIONS *i DATA AND DISCUSSION Hydrolysis Biological Evaluation Glycldylation Application Studies Esterification Coatings Resin Curing Amine cures Curing generalizations Aroclor-epoxles containing some monofunctional species Dlfunctlonal Aroclor epoxies Adhesive Studies * Aroclor-epoxles containing , monofunctional species Difunctional Aroclor-epoxy Casting and Potting Applications Aroclor-epoxles containing monofunctional species Dlfunctlonal Aroclor-epoxy Hydrolyzed Aroclor as an epoxy curing agent Aroolor-Epoxy Resin as a Stabilizer ' f v- i*i m ff it' t- i itM i|f f ! *M DSW 621332 STLCOPCB4095362 VI. EXPERIMENTAL DETAILS Hydrolysis }\ t 1.5 Liter' Batoh v 2.0 Gallon Batoh Glycidylation Expt. NBP 367489 Expt. NBP 367^91 Expt. NBP 367469 Esterification Expt. NBP 354873 Expt. NBP 354868 Expt. NBP 359094 Expt. NBP 364202 Expt. NBP 364203 Expt. NBP 354899-4 Expt. NBP 359094 Expt. NBP 367988 Curing Studies Melt Mixing Dry Blending with Liquid Curing Agents Dry Blending of Solid Curing Agents The Use of Reactive Diluent Curing Conditions Evaluation of Cured .Products Adhesive Studies Potting Experiment 1 Glass Laminating VII. PATENT STATUS VIII. REFERENCES ,49 i 49 * 49 T 50 31 51 51 52 52 52 53 53 53 54 54 54 55 55 55 55 56 56 56 56 56 57 57 STLCOPCB4095363 't ->* h HAPPEND1X3 tT , : i . Epoxj|,Constants ' | | - ' * Calculations for Epoxy Esters * Calculation for Preparation of 300 Oallons of Ester - Key to Abbreviations Curing Studies on Aroclor Epoxies Curing Studies on Aroclor Epoxy Effect of Curing Variations on Aroclor Epoxies Shear Strength of Epoxy-Aluminum/Aluralnum Bonds Biological Toxicant Evaluations of Hydrolyzed Aroclor Esterification of Dlglycidyl Ethers (oxirane oxygen only) ' Esterification of Dlglycidyl Ethers (oxirane oxygen only) Higher Esters of Dlglycidyl Ethers List of Notebook Pages List of Raw Materials List of Tables List of Figures Table of Contents No, 8: ?f f. : l p ** t- tkk i iL A > m Iff!|-iff DSW 621334 STLCOPCB4095364