Document N2oq5GGjvgmYpgd6JLoz4bQjQ

INDUSTRIAL AND POWER CAPACITOR DEPARTMENT - GENERAL ELECTRIC COMPANY Technical information for use by Monsanto Chemical Limited of United Kingdom In License Agreement on Chlorinated Diphenyl with Enoxlde Additive - dated February 10. 1969. All Information contained In this book, as well as subsequent additions and oral Information submitted In technology discussions, la to be kept confidential. The copy of all written Information submitted under this agreement will be retained In duplicate record by Ccneral Electric Company (att, M.E, Scovlllc, Consulting Engineer). M,. ocovllle February 10, 1969 HONS 034780 IO T E X HONS 034781 Industrial and Power Capacitor Department - General Electric Company Technical Information Index Engineering General 1, Column Chromatographic Refining of Fyranol II Part I - Report HF-68-808-JE Fagcl 2, Column Chromatographic Refining of Pyranol II - Part II - Report HF-68-822 -JE Fagel 3. Technique for Measuring Dissipation of 1699 Pyranol and of Pyranol II Report 11F-68-872-JE Fagel 4, IR Spectrum of Pyranol II in 3-6 Micron Region-Report HF-66-962-JE Fagel 5. Epoxide 206 - Pyranol II - Fuller's Earth: Calorimetric and Adsorption Study - Report IIF-66-835 - JE Fagcl Material Specifications 1.. Pyranol 41699 (41262 Arochlor) . 2. Pyranol II (41262 + 46206 Epoxide) 3. Epoxide 46206 (Vlnylcyclohcxenc Dioxide) 6. Fuller's Earth (Attapulgus Clay) A13B1-S7 A50PC8A-E3 D5D71-S3 D50FC12-S1 Manufacturing Process Instructions 1. Flow Chart 2. Characteristics of Pyranol with and without Stabilizer 3. Tcst6 on Pyranol A13B1C(1699) 6. Refining Pyranol A13B1C (1699) 5. Acceptance Tests of Pyranols 6. Refining of Pyranol II 7. In Process Testing of Pyranol II 8. Safety Rules for Handling Pyranol II G/.S6367 GAS6-77-66 P3CPC23 P26DPC6 F3BrC? P76DPC5 P3CPC60 P2APC10 Quality Control Instructions 1. Determination of Epoxide 6206 in Pyranol II 2. Aliphatic Content in Pyranol II 3. Unstable Chlorine Compounds In Aekarels 6. Contamination of Capacitor Liquids S. Resistivity of Electrical Insulating Liquids E50PC72 E50PC25-S1 E6C68-S2 E6C66-S6 E6D2-S2 Laboratory Test Methods 1. Free Chloride Teste 7. Contamination of 1699 Pyranol by Paper 3. Pyranol-Thermal-Chemlcal Stability Test 4. Volume Resistivity of Dielectric Liquids 5. Epoxide #6206 Test Methods > CLTM-l CLTM-5 CLTM- 70 CLTM-37 T1B-80G206-1 M.B.Scovllle 7/10/69 MONS 034782 I t'J Ci t t*J l') W W PI > yj r / HONS 034783 MEMO REPORT HF-68-80S April 8, 1968 COLUMN CHROMATOGRAPHIC REFINING OF PYRANOL II PART I J. E. Fogcl Material Physics and Chemistry Unit ABSTRACT Epoxide ?06 and "aliphatic** concentration data arc reported ot Pyrnnol II chromatographic effluents from columns filled with fuller's earth (as received and thermally activated), silica gel, alumina, and magnesium silicate. Probable explanations for typical Manufacturing experiences in the Pyrnnol II refining operation arc derived from the duta* ** * * Capacitor Department Engineering Section Advance Development Laboratory Hudson Falls, New York Copy No. HONS 034784 1 ymonucnoN The refining of Pyranol II to yield a dielectric of acceptable resistivity and power factor Is complicated by the Department's commitment to tank treating* Economics dictate recycling the lerge excesses of used Pyranol II following each treating operation, but a difficulty exists: refining methods successful In removing power dissipating impurities from the liquid also remove epoxide 206* Presaturatlon of the refining system with epoxide 206 is therefore necessary to permit Pyranol II purification without un acceptable, concomitant loss of epoxide* Fortunately, it appears possible to do this (remove power absorbers from PIT. without gross loss of epoxide); but, unfortunately, not without coinpLication(s) resulting In part, at least, from the mode of reaction between epoxide 206 and the refining adsorbent (fuller's enrth)--- cf. Memo Report -835 Purification of liquids by column chromatography is now commonplace, efficient, and economical In many industrial, nnd even home environments* The Capacitor Deportment, in essence, employe column chromatography in its Pyranol refining operations, hut in a grossly inefficient manner. Tt is highly probable that modern column chromatographic techniques and equipment (Hey, Culllgan Man!) would substantially improve the Department's Pyranol refining >y yielding higher quality oil with less fuss and bother, and without the ,,li.phatic8M contamination that often characterises the current Sparkler prenn operation* (No particular claim to originality is made for this state ment; It scl'ios to he well known Ami appreciated throughout the Department*) The aim of A*D*L* work, the first part of which is here reported, Is to point the way to a column procedure that can be Introduced to the Department's Pyranol refining operation in plocc of the present system* Further work will detail the results of column chromatographic purification of Pyranol and Pyranol II with respect to electrical characteristics, and ghe attempt to find correlation between such characteristics and chemical composition* Emphasis, however, will be on defining a practicable method for efficiently producing high quality Pyranol II both from virgin materials and from recycled oil* . EXPERIMENTAL The chromatographic work was done with commercial glass columns 25 nan* in diameter* The columns were fitted with a simple constant-head device (water cooler principle) constructed from a conical funnel, and a separatory funnel as a reservoir* Desired How rates were achieved by adjusting the height of the collection vessel (flexible, Teflon tubing connected to exit valve at column base). In each case, column beds contained 50 grams of ad sorbent, 60-80 mesh; bed heights averaged about 20 cm* Flow rates were nominally about one ml per minute (extremes of 0*3 and 2 ml per minute)* The constant head waa approximately one meter. M0NS 034785 -2 - The chromatographic adsorbents used were: 1* fuller's earth; os received In Capacitor Department 2* fuller's earth; thermally activated at 450eC, stored in air at 170*C 3. alllea gel} 60-80 mesh, MC6B #9311 4. alumina; 60-80 mesh, Alcon Type F-l, MC&E #9288 5. alumina; 60-80 mesh, Alcon Type 11*151* MC&B #9293 6. magnesium silicate (Florisil); 60-100 mesh, MC&B #9109 The two aluminas arc essentially alike except H-151 has greater adsorptive capacity than F-l (F-l is normally used for gases; 11-151 for liquids). Epoxide and aliphatic concentrations were determined by conventional IR pectrogrnphy (Test Report NF-64-962). Fyranol 11 (about 0.3 wt/wt percent) was synthesized by mixing one gallon quantities of refined storage Pyranol and epoxide 206 (vlnylcyclohexene dioxide). All column chromatography was performed at room temperature (19eC to 28*C average close to 25*C) Column flow was interrupted overnight for all runs:partly because continuous fraction collection equipment was not available, and partly because It was desirable to learn the effects, If any, of such IntcM'iRittaul operation. In normal ue, a colui.a'i would be. operated more nearly continuously (there is normally no Interest In collecting effluent fractions). Single pass operation was exclusive. Data compilations, concentration calculations, integrated column retention value calculations, and plotting of column elution curves were all simultane ously performed via time-sharing computer. RESULTS The data arc presented graphically in the accompanying elution curves. Effluent concentrations, expressed as fractions of starting (Influent) con centrations, are plotted against effluent volume (25 ml fractions). The printed nutubera represent grams of material retained by the column (Integrated for each act of five effluent fractions). The BASIC computer program XYPAfiS, appended, is a marriage of the library program XYPL0T and necessary additions to perform IR calibration conversions, Integrated column retention calculations, and printout. HONS 034786 -3 - CONCLUSIONS AND DISCUSSION A number of conclusions and projections are possible from this work, and the earllor work cited above (Memo Report HF-64-835). From the earlier work It is evident that the principal fate of epoxide 206; on retention by fuller's earth, is polymerization. The exact mechanism of the polymerization reaction Is not known, but presumably is closely related to documented heterogeneous polymerizations catalyzed on the surfaces of many metal oxides and similar solids) (e.g., Encyclopedia of Polymer Science and Technology, Interscience, Vol. 6, p.llOff., 1967). Runaway, violent polymerization does not occur except when critical quantities of fuller's earth and undiluted epoxide 206 ovc mixed; in particular, violence erupts when epoxide 206 Is added to a slurry of epoxide 206 and fuller's earth - for reasons that arc not immediately clear, but in attention - demanding fashion. In any event, epoxide 206 seems to be irreversibly adsorbed by fuller's earth even in non-violent reactions. Historically, there has been considerable Department Interest in the presence of aliphatic hydrocarbons In Pyranol. The total aromatic character of 1699 Pyranol mokes the infrared spcctrophotomctrlc detection of Impurity aliphatic* (of which vacuum pump oil is one example) a simple task, and the determination of relatively small amounts of such Impurities is routinely and sensitively performed. Epoxide 206 contains both active epoxide rings and aliphatic hydrocarbon functions In a single molecule (In addition, the roughly five percent Impurities in commercial epoxide 206 bear further aliphatic functions). Therefore, "active epoxide" and "aliphatics" are synonymous in fresh Pyranol II (neglecting the five percent Impurities in commercial epoxide 206). What, then, Is the cause of increased aliphatics so often reported for Pyranol II? Pump oil is a probable "sometimes" answer, but obviously ridiculous as a general explanation. Polymerized epoxide 206 Is tho likely culprit. Recently there has been interest In the relative merits of as received and of thermally activated fuller's earth. Reportedly, the use of as received material eliminates (or greatly lessens) the normal problem of high aliphatic content in refined Pyranol II. Why? The answer seems clear: activated earth would be expected to exert a greater catalytic effect on epoxide 206 polymeri sation than would unactlvated earth, and hence would yield more polymer to contaminate the Pyranol II (or possibly polymer of readier solubility in Fyronol II). The hypothesis is this) polymerized epoxide 206 Is generally tetained by fuller's earth, but does exhibit some tendency to be stripped off by fresh Pyranol IX. That is, polymerized epoxide 206, formed by reaction of Pyranol II with fuller's earth, Is in small measure picked up by passage of subsequent Pyranol II over the earth on (in) which the polymer is largely re tained. The polymerized epoxide 206 has lost none of the aliphatic character of the original monomer, but has lost all of its active epoxide character; hence the concentration of active epoxide is changed In the Pyranol II, but the aliphatic hydrocarbon concentration is|increased, ? $ t O' C/M r. w{ M0NS 034707 - 3A Examination of the appended elution curves (X through VI) reveals the following! 1. The breakthrough for as received fuller's earth comes considerably sooner than for thermally activated earth. (Activation Increases the adsorptive capacity of the earth.) 2. The breakthrough for the as received earth is sharp and clean typical of physical adsorption without significant chemical reaction. 3. The breakthrough for the thermally activated earth is notably unsharp - characteristic of significant chemical change as well as physical adsorption. 4. After breakthrough, the active epoxide and aliphatic concen trations are essentially the same from as received fuller's earth columns (curves I & II, V & VI). 5. After breakthrough, active epoxide concentrations are notably less than the aliphatic concentrations from thermally activated fuller's earth columns (curves III & IV). 6. After long term elution (curves V & VI), there is a slight increase in aliphatic concentrations over active epoxide con centrations even for the as received fuller's earth. The difference Is not at all comparable to that observed for the thermally activated earth; it is significant, however. In fact, the level of aliphatic is seen to exceed the epoxide (hence aliphatic) level in the Influent - fact that could result only from stripping the column of previously retained siaterlal. HONS 034788 -4 - The conclusion is that fuller's earth (particularly thermally activated fuller's earth), causes polymerisation of epoxide 206 In Pyranol II; that the polymerised material Is largely retained by the earth, but portions arc subsequently stripped Into fresh Pyranol II thereby increasing the aliphatic/ epoxide ratio ("delta") of the lattir. Magnesium silicate Is a compound similar to fuller1s earth - the OBrth being a natural clay (a complex silicate of aluminum, magnesium, and potassium). Blution curves VII and VIII reveal the expected: Pyranol II responds to magnesium silicate In substantially identical fashion at to fuller* earth. (The magnesium silicate was used as received). Epoxides arc polymerised by acids and by various oxides. Silica gel is both acidic and an oxide; one would therefore predict a response of Pyranol II to silica gel much like that observed for activated fuller's earth. Elution curves IX and X, for as received silica gel, tend to confirm the prediction. Quantitative comparison Is difficult at best for this, or in fact any other solid adsorbent system, because of uncertainties regarding the actual state of activation of the adsorbent. ThermAlly activated adsorbent is more active than non-activated material depending on time and temperature of the activa tion and on the length and quality of storage after activation. The significant and often unsuspected degradation of an activated solid during even brief storage under presumably good conditions was mentioned specifically in HP-64-835, and of course lo generally well known. Undoubtedly many seemingly strange behaviors are readily accounted for by not-so-cctivc "activated" adsorbents. Activated silica gel, as received from one commercial source, In any event, behaved much like Capacitor Department activated and stored fuller's earth. Alcoa Type 11-151 activated alumina, as received, exhibits an unusual and most Interesting behavior in the column chromatography of Pyranol II (elution curves XX and XII). Breakthrough (alleging for Interrupted column flow) Is sharp Indicating little if any chemical reaction - as one might predict for a nominally basic compound. Following the elution curves after breakthrough, one notes an unusual feature: the active epoxide effluent concentration tends to become greater than the aliphatic hydrocarbon concentration. It is easy to see how active epoxide can be lost through catalysed polymerization, but how can the aliphatic function disappear without concomitant loss of active cpoxlde7 The onswer is that it can't; but the aliphatic included in the five percent of Impurities in the commercial epoxide 206 (part of which are known to be acidic) can be adsorbed independently. Since the initial aliphatic concen tration in Pyranol II synthesized from refined Pyranol (aliphatic-free) and commercial epoxide 206 Includes both the aliphatic functions of the epoxide and indistinguishable (from the IK measurement viewpoint, at least) aliphatic contributions from the five percent impurities in the epoxide, the selective adsorption of the Latter by an epoxide-"saturated" column would yield the unique result of an aliphatic/epoxide ratio less than one. MONS 034789 5 The basic alumina la expected to remove acidic compounds, whereas acidic adsorbents like silica gel, fuller's earth, etc*, are not* The dote of curves XI and XII are apparently the first non-electrical mepsure of the specific removal of Impurities (more than likely power absorbing Impurities, at least in part) from Pyranol II. Work In progress will hopefully confirm the electrical quality of euch purified Pyronol II. A precedent exists supporting In a small way the above claims. ES Dunham has recently shored with the writer essential details of work roported by Monsanto concerning their application of aluminA-filled columns for chromatographic purification of 1499 Pyranol. Monsanto's experience (not In the least unusual or surprising In view of common chromatographic knowledge) has been that the alumina column purification has exhibited long term, high volume capacity and yielded dielectric of very high quality. -- The cited elution curves for Pyranol II strongly suggest chromatographic purification with alumina should yield results comparable to Monsanto's for 1499* Aside from the general advantages of single-pass column operation over the recirculating Sparkler press procedure, the possibility of enjoying added gain by using alumina instead of fuller's earth Is of increasing interest. Many engineering problems undoubtedly lie ahead, not to mention the Inevitable economic factors that usually confuse otherwise good technical situations, but there may well be a small pot of gold worth seeking at the end of this rainbow. The elution curves for all the adsorbents clearly reveal the effect of inter* rupted column flow. Weekend Interruptions arc discernible from overnight interruptions, even. The only immediate explanation for the sharp falloff of epoxide concentration coincident with each flow interruption is that given sufficient time (hours instead of minutes) of contact, the solids catalyse polymerisation of significant amounts of epoxide. The fact that this behavior Continues apparently ad infinitum indicates that too fast column flow is not the answer - In the indicated flow range, chromatographic equilibrium seems certain. In practice, then, Pyranol II could be refined by passage through an epoxide 206 saturated column without loss of epoxide as long as flow was con tinuous. Prolonged static contact of Pyranol II and any of the adsorbents should obviously be avoided. The planned second part of this report will include electrical measurement data on column chromatographlcally purified Pyranol and Pyranol II, and will discuss specifics of column chromatography applicable to the Capacitor Depart ment's refining operation* ACKNOWLEDGMENT The routine of column operations was handled mainly by E. Hawk and V* Shedlglan, both of whom also performed the many IR measurements* Helpful discussions with many Department personnel (particularly including X*L. Guiles, G*A* Smith, E.S. Dunham and W.O* Solberg) are gratefully acknow ledged. MONS 034790 X Y3A0S 9156 TUESDAY 04-08 r> hek ymm data# ljnk isoi links 160-399; correct line iso 10 LET S*0 30 LM A*0 . .10 LET T*0 /|0 'IKAD ft# CO 50 (foro iso * 60 READ H 70 IF H> 999 THEN 790 HO LET A*AM . ,, 90 LET S5*H 100 LET Y* H/tft*CO> 110 IF A<>5 THEN 110 130 LET R* 35*1*39*C51 .00314-S/27400) 130 LET l*TR 300 DATA 999 310 nU'I'!-:.V , 330 PRINT 330 PR I Ml "v-HAl OR E Y.. I W, YE,AY , .YE 1 N XftAX * DELX" l 300 |NI1 . *0 01# 03 * 03# 04 350 |,M 05 <ni-00>/60 ` 360 LFf 06*0 370 FO* X * 03 TO 03 STK 04 . 3ho Gur.on on 390 IF 06 * 0 THEN 650 400 IF 06 = 30 THEN 430 ' 410 *>RINT M "1 430 601U 450 430 POINT..............I 440 LEI 06 e 10 4JVO IK Y > 01 THEN 630 ' 460 IK i < 00 THEN 630 ' 470 LM 07 * 00 ' 4HO LET Tn7<0.S*>5 490 IF T<Y THEN 590 . -- hOO LET (36 06 H 510 IF Z-Y>*P-I0S THEN 570 '. 530 IF /-Y>*-*5 l HEN 550 530 PRINT " "/ 540 f-UiO 710 550 Ptl.vl " 560 t-.OTO 710 570 ' INI ""J 5*o i;CiTm 710 590 L.M 07 * 07 3*05 600 *M VT ; ` 610 6010 *HO 630 O'lNT "OK* SCALKi (X * Y) * "1XJ"# "I Y 630 LE! 06 oft * I 640 M7io 650 51 NT 660 i'i INI Eu'> V| TO*9 "fOP| OOTlOFi ";Q3J" INCREMENT "J'N/I *70 *" | NT "FOE YI LF El "JOOJ" p ! 6RT * "J01J" INCREMENT * *'105 640 V* INI Ann S' | NT " ........................ 1.................................................................................................... 1............... t 700 H>10 4 no 710 IF AS THEN 70 . 730 *0IN1 T in 60T0 770 7nO *>Mn1 K 750 LM A*0 7v |.n :;o 770 next y 760 `'HINT >n p INI "GRAND TOTAL HF TAlNED*"ITJ"GRAMS" no vi* ` , | I ! * ' ' ' HONS 034792 FOR X I FOR Y l TOP = 0 LEFT O -HOTTOK 78 RIGHT 1 *8 INCREMENT INCREMENT c/c ~ \> * -v Atotl-At-T. SO** MOMS 034793 FOR X : FOR Y : TOP 0 LEFT a 0 BOTTOM * 72 RIGHT 1 *2 INCREMENT INCREMENT AutPM. fltoA/-- flcT- act. r.e. o c/c. MONS 034794 I act, re. auf*. c/e ifLi HONS 034795 i I :S. - 4t zec" / HONS 034796 /If ! /s nf oS HONS 034797 /re . - as *** I FO`t Xt T<> I FCM Y* LF.FT O ptfTIOM - 340 INCHFf.rNT 1 !! IUHT ! tNCUMr.NT *08 3U V W I\/,, c-c -A \/ 0 (-/-JSL-/1Z-) HONS 0 3 4 8 0 0 ' AuPMAr/C A. Ala. ' So^ HONS 0 3 4 8 0 1 is a ra n ts HONS 034802 7?9 VzMS MONS 034803 % i *i N HONS 034805 MEW REPORT HF-68-822 July 16, 1968 COLUMN CHROMATOGRAPHIC REFINING OK PYKANOL II. TAUT II. J. E. Fogel Mnicrlnl Physics And Chemistry ABSTRACT Single pans, voom temperature column chromatographic purification of Fyrnnol II with both fuller** earth And alumina yields full strength Pll with no ha1iphaticsM, and 100*0-100 ])z dissipation of 0.6Z. Minimum unrful capacity Is 4 liters per fifty gramr. of each adsorbent; breakthrough occurs at about 1 liter, and full strength effluent is obtained after 3 liters. ***** This J.k Class IV technical report. The Information la proprietary And in I'ov General Electric, employees only. Distribution within the Company is restricted* No dtr.clo&ure should be made tounnuthorls'.ed prv'iont. ' Capacitor Department Engineering Section Advance Development Laboratory Hudson Kalis, New York lN*pr No. % HONS 034806 m-RomicnoN: The results of Initial column chromatographic efforts to purify Pyranol 11 were reported In Memo Report HF-68-808. Data In that report concerned column behavior, effluent concentrations of active epoxide and "aliphatic*", holdup and breakthrough characteristics, but nothing nbout the electrical quality of the effluents. This report, and its companion describing the technique by which dissipation data were finally obtained, essentially com pletes the story started earlier. The conclusion, although not unexpected, is nonetheless happy. KXPm^NTAj,: Two columns, identical to those described in HF-68-808, wore set up for the present work: Thu first filled with 30 g of 170*0 nir-dried fuller's earth, the second filled with 50 g of activated alumina 11-151. Flow rates through each column (approximately 20 cm deep, each) were 0.3 ml/win under a constant head of roughly one meter. Percent dissipation at 100C-100 Ur (and 1000 Hr.), and epoxide and aliphatic concentration (IR) were determined for effluent samples from each column over s period of three weeks. Column flow was con tinuous throughout the entire period for each adsorbent (contrasted with the interrupted flow reported in HF-68-808). H&PXCS* The data arc tabulated in Tables I and II for each column. DISCUSSION: < The % epoxide (column 3) values listed in the two tables were derived from IR absorption measurements of effluent samples taken from the Balsbaugh cell following dissipation measurement. Comparison samples measured before dissi pation determinations showed full strength effluents after about 3000 ml. In other words, a roughly ton percent: loss of activc epoxide (no "aliphatic0 change) ns a comicqunnro of the dissipation measurement is clearly revealed. The cause and/or mode of the loan ore unknown; the Ions, however. Is as real as it Is suvprJ nine,, and raison n number of interesting questions regarding the actual in-service fate of epoxide 20(> a* a capacitor dielectric component. The Immediate-:!nteiost significance of the loss seems minor, however; the loan certainly does uot. contradict earlier column chromntogrnphic data. The data further reveal effective purifying capacities of at least four-plus liters of Pyrnnol. 11 per 50 grams of each adsorbent. The runs were terminated arbitrarily after the passage of roughly seven liters of PII; one might guess the actual capacities to he greater than the minimum value demonstrated, however. M0NS 034807 2 The important, If not surprising, result of this demonstration is that a single, room temperature pass through modest-depth beds of either fuller's earth or alumina yields full strength Fyranol II of excellent 100*0-100 Hz dissipation. Q.E.D.! ' What about the prognosis of the technique as a useful Department tool? The writer confidently expects the near-future Implementation of the basic (columnar) approach by Manufacturing in some non-eritlcal area, and gradually the complete replacement of existing refining facilities by a neat, efficient easily manageable, automated column operation. ~ Sealing columns to useful dimensions will of course uncover nuisance problems of the sort always encountered In the transition from laboratory through pilot plant to production. There Is certainly no reason why anything greater than nuisance will he encountered, however, and at least a good trial Installation should be eonstructablc at modest cost and effort. The following specifics arc suggested for openers! 1. A column roughly 1?" in diameter and about 6* tall. 2. Fuller's earth activated at about 1S0*C (the usual higher tempera.turr activation results only In unnecessary destruction of epoxide and the consequent generation of "allphatics"). 3. A simple pumping system for efficient liquid handling and pressurizing. 4. A simple heat exchanger to warm Fyranol to lower viscosity and hence speed column throughput (100*C la unnecessarily hot and possibly troublesome-Increased undcslrcd reaction ratcs--50*C to 75*C maximum Is suggested). 5. Provision (pumping and storage) for recirculating liquid through column to facilitate Initial "saturation" of adsorbent with epoxide. 6. At least 10x P1I (that Is, about 37. epoxide) for column "saturation." Probably considerably greater epoxide concentrations (10%-20%-----?) could profitably and safely be used, but continuous recognition must be given to certain consequences (all bad?) that result when "too much" epoxide is mixed with active adsorbent (Memo Report HF-64-835). i 7. In line with No. 6, consideration should be given to the real possibility of making a concentrated "PII" that profitably could be added to refined 1499 to yield the desired final PII, It might he more economical to run the operation this way than conventionally (smaller volume of liquid through "Pll" column, hence smaller column, etc, etc.). The conccntrotc concept warrants serious Investigation, and reportedly (Monsanto personnel.) Is already employed in Europe. HONS 034808 3 As a first bnbystep, the writer has constructed an all glass column about 2M In diameter and 3,5* tall to produce usable quantities (laboratory scale, to be sure) of purified oil* An adsorbent charge of 500 grams of fuller's earth ()60*C air dried) yields a bed about 17 deep, and a room temperature throughput of 300-plua ml per hour under a hcod of roughly four feet. This represents about a 15-fold output increase over the small (1" column, 50 g charge) setups originally used and described (Memo Report HF-68-808; first part of this report). The initial output from this column has a 100*0100 llz dissipotion of 0.4%. . The data In Tables 1 and IT offer further suggestion that alumina may be better suited to the PIZ purifying Job than fuller's earth, and should be actively considered for future application. For the moment, however, there is no apparent reason not to proceed simply and directly using material on hand and well proven by long, direct experience. In other words, to summarize, the work reported and the suggestions offered say little more than this: A modest, reasonable change (to full column technique) of the present, basically proven procedure of pyranol refining will yield n more efficient, reliable, And easily manageable system than presently available while permitting greater flexibility to cope with new problems the future will undoubtedly bring. i MQNS 034809 ML Throur.lt Column 380 520 905 1300 1450 2445 2965 3400 3675 4325 3475 6305 6755 7205 100*0 - 100 Hz DISSimilON fuller's EARTH COLUMN X Epoxide* X Alxpha .8 0 0 .7 0 0 .4 0 0 .9 47 73 .7 63 82 .8 76 66 .5 84 ) 95 .7 87 ) 100 .5 87 ) 95 1.0 89 > 100 .6 92 ) 100 100 .6 89 ) 100 .6 89 ) 100 .8 92 ) 100 Influent %D *= 2.5 * Relative to Influent <? about 0.35% EP 206. (3 See discussion. J. K. Fnfjcl 7-16-68 HONS 034810 1. Tltvoofth Column___ 325 A 05 815 1275 2355 2875 3325 3785 $735 5315 6215 0005 7115 -5TABLE XX 100*C - 100 Ha DISSIPATION ALUMINA COLUMN IJlsL .8 .7 .5 .5 .5 6 1 -5 .6 *7 5 .6 .5 .8 7. Epoxide* 0 0 58 Vt 87 89 89 ) 92 ) 89 ) 92 ) 100 @ 89 ) 92 ) 89 ) % Aliphatic* 0 0 6$ 77 91 93 100 98 100 100 98 100 100 Influent ZD 2* 5 * Kelfttlvo to Influent (? about 0.35Z F.P 206. @ See discussion. J. I'. Knt'.cl 7-10-08 MON5 034811 MEMO RETORT HF-68-821 July 16, 1968 TECHNIQUE FOR MEASURING DISSIPATION OF 1499 PYRANOL AND OF FYRANOL II J. E. Fagel Material Physics And Chemistry ABSTRACT Cell denning inadequacies are the principal source of contamination leftdin.; to the frequent experience of low precision dissipation measurement a. A technique honed on tills premise has been derived and found to yield sell able measurements routinely* * ** * This Jn n Clnss IV technical report. The information is proprietary and in for Crnrrnt Electric employees only. Distribution within the Company if restricted. No disclosure should be made to unauthorized persons. Capacitor Department Engineering Sretion Advance Development Laboratory Hudson Fnlls, New York Copy No. HONS 034812 I INTRODUCTION; The author In recent months has had need for reliable dissipation measurements, of relatively small volumes of Pyranol I and II, in support of column chroma tographic purification studies. At the start of these studies, the naive assump tion was made that the electrical measurements need would easily and routinely be met; In order to continue from the first part of the chromatographic work (Memo Report 1IP-60-800) It: was necessary to develop a reliable dissipation measure ment technique. Chemical and related measures of additive and of Impurities concentrations in Pyranol* are Interesting and even pertinent, but the real proof of the pudding is of course electrical dissipation. The brief discussion offered in this memo relates a successful technique derived from the hypothesis that assenttally the only reasonable cause for low quality dissipation measure ments is contamination resulting from--or as residual of--ccll cleaning inade quacies. APPARATUS: The Balnbaugh 03HV35 (Monel) dielectric liquid test cell was chosen because of minimal size (roughly 3ft mil of liquid required), proven performance (Capacitor Department; ASTM), and immediate availability (two such cells have been located In the Department). Instrumentation and associated equipment (completely shielded cell cage, for example) chosen was the GR Type 1605 Impedance Comparator setup in the QC Lab.-uxed with the generous approval and assistance of QC Lab. personnel. TECHNIQUE; The writer is of the opinion that reliable dissipation measurements must be possible given a clean test cell and (of course) functional, quality instrumenta tion. Mkewlfic, measurement* that fall short must suffer from contamination that in turn in directly attributable to inadequate celt cleaning. As previously indicated, the technique iter*! described la based on the thcals of adequate cell cleaning, and hanlcally conslsta of two parts: (l) Initial cleaning of cell, and (?) between measurementa cleaning, of cell. 1. Ini.t.ial. ^Ctonnln^ General experience, coupled with the dictates of the chemistry of the mctal(s) of which Lyplcnl cells arc constructed, suggests an alkaline cleaning medium as most suitable for initial cleaning of the cell. (The Balsbaugh 3HV35 is Monel; 316 stainless steel would be preferred by the writer, and reportedly otherr, Monsanto personnel for example). Accordingly, the following is done: The disaasembled cell Is first cleaned with acetone to ensure removal of organic residues (principally Pyranol). The cell parts ore then placed in a gently boiling aqueous solution (in a stainless steel beaker) containing roughly 107. (wt/vol) sodium HONS 034813 -2- hydroxide and 1% ethylenedisminetctraacetlc acid--the latter to stabilize cations and ensure thnt they stay in solution. Following boiling for approxi mately one-half hour, the parts are removed from the caustic solution, rinsed in water thoroughly, then rinsed in dilute, aqueous hydrochloric acid (1:10), and finally rinsed thoroughly again in distilled water, and oven dried at about 100*C. 2 Between Measurements Cleaning. , Before the first measurement, and between each and evory subsequent measurement, the entire (assembled) cell is cleaned with generous quantities of acetone and then hexane. The acetone is dispensed from a wash bottle in a fine stream to gain the benefit of motion in addition to solublllty--this Is an essential feature of the technique: Pyranol is not readily removed from the cell, It turns out. Cleaning is judged adequate when, (a) Pyranol is no longer precipitated by water from the acetone wash liquid, and (b) no rraidue la dinccrnihlc on tho cell surfaces (viewed obliquely from several different angles) following evaporation of the acetone. Repeated acetone washing la required to meet condition (a) and even further washing is essential . to satisfy (b). Hot air (laboratory heat gun) is blown over the cell to prevent condensation of moisture from the air as a result of cooling caused by the evaporation of acetone. , Generous rinsirgoi all parts of the cell with hexane, and hot air drying, completes the cleaning procedure. The cell is then rinsed with two small portions (10 ml) of the sample to be measured , and finally filled with the sample. The filled cell Is placed in an oil bath, at about 105*C for 30 minutes. A thermal Jacket (constructed with a glass beaker inside a larger, steel beaker--the annular space filled with glass wool) Is kept In an oven at 100*C. The cell, on removal front the oil bath, Is placed in the thermal jacket, and immediately transported to the comparator setup. Tho completely closed cell cage precludes the convenient use of a controlled bath during measurement, but the Initial, slight overheating of the cell, and the heated thermal jacket makes 100*C measurement quite simple and repeatable in the following manner: The dissipation at LOO Hz and at 1000 Hz is measured at the time in the cooling period of the cell when the observed capacitance equals the value Initially measured for the cell at 100*C. This initial capaci tance measurement was made at tho time (about *> minutes) after the removal of the cell from the oil bath when an enclosed thermocouple Indicated n sample tempera ture of 100*0. The coll used for the. measurements reported exhibited a capacitance of IB7.? pf, at 100C, filled with either 1499 Pyranol or with Tyranol 11. A duplicate cell has essentially the same capacitance. It Is true that the actual temperature during measurement may not be exactly 100*0, but it Is equally true that, (1) tW temperature is close to 100*0 (withIn a degree or two at most), and (?) it is repeatable whatever the exact value. Furthermore, it is repeatedly observed that tho pleasured dissipation changes but little over the entire tempera ture range during the measurement; the absolute accuracy of the determined 7J> fully acceptable, in other words. HONS 034814 -3- RESULTS: Typical, replicated data obtained over a period of several days for a ran domly selected sample of 1499 Pyranol (obviously of low quality) are tabulated In Table I along with calculated averages and standard deviations. Although the data are pooled, they actually come from two separate (and statistically distinct) populations: The first 13 values represent results obtained without washing the cell between measurements; the last 0 values were obtained following I ho previously described technique for cell cleaning. In other words, the 1071 relative standard deviation is a worst-case expression of precision. IHSCUSSION: One fundamental fault was discovered in the course of the work with the Balsbaugh cell: The large insulator (a filled polymer) separating the outer and guard electrodes of the cell was found to retain Impurities to such an extent that its resistance dropped from 10** ohios to 10* ohms! No washing procedure was found that would successfully clean the insulator; but the high resistance could be restored by beating for at least one hour at 125*C in a vacuum oven. To circumvent the problem altogether, even though the vacuum-heat treatment was completely successful, new insulators were machined from Teflon--the same polymer being the Insulation originally used between the inner and guard electrodes of the cull. After prolonged use (three weeks; five measurements/day) and despite the fact nothing but AC ever was applied to the cell, an unmistakable yellow-brown film had formed on the walls of the celt, along with a few darker colored spots on tne bottom of the cell. Both blemish conditions were easily removed by mild abrasion with sink cleaner (Bab-o; special institutional formula, to be exact). Initial cleaning procedure (1) was repeated following the abrasive cleanup-the need for which was signalled by significantly high dissipation values for samples known to be of relatively low 100C, 100 llz dissipation. The nature and cause of the contamination buildup arc not known, but presumably relate to oxidation of ttic cell. Nonetheless the consequences arc unmistakable. One wonders how often this sort of difficulty might have led in the past to erroneous results. Undoubtedly, other approaches could successfully be used to gain the same result here reported; the point is that It is possible, routinely, to achieve cleanliness ami that reliable dissipation measurements follow as a natural consequence. At least Ibis is the conclusion from one practical, inductive process. Vandon Shodiglan made nil the dissipation measurements. MQNS 03<815 -3 - RESULTS: Typical* replicated data obtained over a period of several days for a ran4only selected sample of 1499 Pyranol (obviously of low quality) are tabulated in Tabic 1 Along with calculated averages and standard deviation*. Although the data arc pooled* they actually come from two separate (and statistically distinct) populatlona: The first 13 value* represent results obtained without washing the cell between measurements; the last 6 values were'Obtained following the previously described tecbnlquo for cell cleaning. In other words* the 102 relative standard deviation la a worst-case expression of precision. EISCUSSION: One fundamental fault was discovered In the course of the work with the Balsbaugh cell! The large Insulator (a filled polymer) separating the outer and guard electrodes of tho cell was found to retain impurities to such an extent that its resistance dropped from 10** ohms to 10^ ohms! Mo washing procedure was found that would successfully clean the Insulator; but the high resistance could he restored by heating for at leost ono hour at 125*C In a vacuum oven* To circumvent the problem altogether* even though the vacuum-heat treatment was completely successful* new Insulators were maehined from Teflon--the same polymer being tho insulation originally used between the Inner end guard electrodes of the cell. After prolonged use (three weeks; five measureraents/day) and despite the fact nothing but AC ever was applied to the cell* an unmistakable yellow-brown film hpd formed on the walls of the cell* along with a few darker colored spots on the bottom of the cell* Both blemish conditions were easily removed by mild abrasion with sink cleaner (Bab-o; special institutional formula* to be exact). Initial cleaning procedure (1) was repeated following the abrasive cleanup-- the need for which was signalled by significantly high dissipation values for samples known to be of relatively low 100*C* 100 Hs dissipation. The nature end cause of the contamination buildup arc not known* but presumably relate to oxidation of the cell. Nonetheless the consequences are unmistakable. One wonders how often this sort of difficulty might have led In the post to erroneous results. Undoubtedly* other approaches could successfully he used to gain the seme result here reported; the point is that it Is possible, routinely* to achieve cleanliness and that reliable dissipation measurements follow as a natural consequence. At leeet this Is the conclusion from ono practical* inductive process. ACIOIOWUnCRMENT: , ' Vandos Shedigtan made all the dissipation measurements. HONS 034816 -4 tabu; i 100 Hu *10 2.45 2.75 2-15 2.10 2.75 2.45 2.70 2.42 2.04 2.40 2.60 . . 2.2. . 2.35 2.05 2.21 2.30 2.00 Mi PKRCF.NT DISSIPATION - 100*C 1000 117. 0.22 .25 .29 .23 .22 .29 .25 .28 .25 .21 .25 .27 .25 -.25' .22 .23 .24 .21 .22 Avr. r 2.33 Std. Dev. 0.24 % S.D. 10.3 0.24 .025 10.4 J.R. F(!rc1 7-io-f.n MGNS 034817 i Advance IHvelorrr.cint laboratory______ SUB - SECT I ON NO . m m t r x f r mgc*vr*c* *o? asstcm* ""Vyjorss 8tSTi?,T?* rr.T- outer To evaluate significance of observed spectral difference between Fort EdvarA PX1 and Tilot Shop PII. CWCLtftfWS - BCSVITS - DISCUSSION COKCl/USIOHS: Tito two peak* obfer/ed for PII in typical differential IR spsetra in th 3 to 4 micron rccion are duo to C-H vibrations. The 3o~ ''"length peak (3*42 microns) in attributable to aliphatic C-H. The she. .r * vv . ,nglh peak (3*34 microns) ha?- been found to correlate with titratable epoxies. jg, Howk hrts for nor:? time reported a notable difference in the IR abcorption spectra of Pyranol II from Fort Edward era from the Hudson Falls Pilot Shop. Twp peaks arc observed in the 3 to 4 micron region for the Fort Edward material, whereas the rilot Shop PII usually chows only one peak with a small poorly resolved p:.ok or shoulder on its low wavelength side. The longer wavelength peak (3.42 r-J.cron") h5 regularly been uced by the Quality Control Laboratory n3 a measure of oliphalic(c) ^ In PIT. The undersigned was asked to review the observation of Howk, and to attempt to elucidate the difference in the IR spectra of Fort Edward and Pilot Shop Pyranol II samplea. D. Harms, IST-G Department, M fc P Laboratory, and P. launer, Silicone Products Department, R & D Laboratory, were consulted in ad lition to personnel of th;'j Laboratory. Following detailed discussions with those interested in the problem, the following hypotheses emerged as probable explanations of the observed facts: 1* The longer wavelength IR peak is due to aliphatic C-H vibration, and is a measure of added epoxide plus any other aliphatic C-H containing molecules. 2, The shorter wavelength peak is due to titratable or active epoxide C-H. 3. The appearance of the shorter wavelength peak as (l) a shoulder on the longer wavelength peak or (2) as a distinctly resolved peak depsnla on tbs relative concentrations of titratable epoxide and aliphatic?, ox.d, of course, on the effective dispersion and resolution of the IR instrument. To test these hypotheses, the following were done: A. Synthetic (Epoxide 206 and 1499 Pyranol) solutions were prepared by E. Howk and by R. Guiles. IR spectra were recorded, and epoxide content determined HfMi by: J.K. Fagcl AfrsovAi: HONS 034818 HF-6I1-962 Page 2 by H ))r titration for each prepared psmplr;# B. The IR spectrum of cyclohexane in 1499 Pyranol was recorded in the 3 to *4 micron region. C. Data for approximately fifty ee.mplen of Fort Edward and Pilot Shop Pyranol II were obtained fro'n the Q.C. Laboratory files (liovk). The measured height of the charter wavelength IR punk or shoulder was compared with the H Br titration value for percent epoxide for each sample. ' RB8U1TS: 1. The IR spectrum of cyclohexane in 149? Pyra&o.l reveals a single peak (in the region of concern) at 3*42 microns. The alignment of the 3.42 micron __ peak to aliphatic C-H iff corroborated, accordingly. _ In. In the course of running the new calibration standards for percent nllphnllcs, it was discovered (Howk) the original calibration curve had shifted. Howk han traced the cauae of the shift to increased instru mental sensitivity resulting from the addition of a new Nerast glower on Kirch 30, 1964 (Corrections to reported percent allphatics for the period April 1 to April 1.5 havs been male by liovk). 2. The measured peak heights of the shorter (3*34 micron) wavelength peak correlate linearly with titratablo epoxide as Illustrated in the accompanying graph. Standards made by Howk (triangles) and synthetics by Guidon (square) fit will on the straight line about which cluster the routine PII samples (circles) from Fort Edvard and the Pilot Shop. (All circled values above 0.30# Epoxide represent Fort Edward PII; all below originated in the Pilot Shop). 4. Spectra recorded on the Model 521 spectrometer exhibit distinctly resolved peaks for all PII samples (lowest epoxide content 0.11#). (The Model 521 instrument gives significantly higher dispersion than the modified Model 21 that is regularly used in the Q.C. laboratory for routine FII Analysis). DISCUSSION: Without attempting to elucidate fundamentals of IR absorption, it teemo highly probable on empirical grounds that the cited hypotheses are correct. The correlation between peak height at 3*34 microns and percent titratable epoxide is clearly excellent for the relatively large number of samples and stendarde tested. Likewise, there can be little, if any, doubt of the basic correctness of tho Q.C. Laboratory use of the 3.42 micron IR peak as a measure of aliphatic C-H. HONS 034819 i etrsznr. risie ^cmcTAar. . 4, g,ax HONS 034820 MEMO REPORT HF-64-035 November ly, 1964 EPOXIDE 206 - PYRANOL II - FULLER'D EARTH: CALORIMETRIC AND ADSORPTION STUDY J. B. Fagel Materials Application Unit ABSTRACT Calorimetric and adr.orptlon data for Epoxide 206, Pyranol II and Fuller's Earth arc reported and discussed with particular reference to the preparation of Epoxide 206 - eaturated fuller's earth and the troutment of Pyranol II. * *** This is n Class IV technical report. The information is proprietary end i for Gcncrr.l Electric employees only. Distribution vi.thin the Company is restricted. No disclosure should he made to unauthorized persons. Capacitor Department Engineering Section Advance Development Laboratory Hudson Falls, N.Y. Copy Wo. *7 HONS 034321 -1- C0NCI.USI0N8 " The following baoic data were obtained as a result of the study aimed at lcnrniG details of the mixing of Epoxide 206 and Fyranol II with fuller's earth: 1, The specific heat (C ) of Kpoxlde 206 = O.36 cal/gC. 2, c 1*199 tyranol c 0.21 cal/g-*c. 3, The hent of wetting of fuller's earth by Kpoxlde 206 is 200 calories/ gram for slurries containing a maximum of about 1*4$ (w/w) fuller's earth. Polymerisation of the epoxide is initiated in slurries con taining greater than 1*$ fuller's earth. *1. Thermal runaway (uncontrolled polymerization) does not occur with slurrico containing as much as 20 weight percent fuller's earth. Thermal runaway cannot be initiated by preheating the Epoxide 206 before treating with fuller's earth. Thermal runaway cannot be initiated by heating a slurry to the normal boiling point of Epoxide 206 (227*C). Thermal runaway can be initiated by adding fresh Kpoxlde 206 to u hot slurry of reacted 206 and fuller's earth. The reaction of fuller's earth with PII solutions containing 0 to 70$ (w/w) Epoxide 206 follows the Langmuir adsorption isotherm. 6. Activated fuller's earth saturated with Epoxide 206 and dried by suction filtration and pressing weighs 101$ more than its initial weight. Overnight vacuum drying at room temperature reoulto In a weight loos that reduces the 101$vcalculated gain to 71$. 7. by )IHr titration, activated fuller's earth removes up to about 15$ of its own maos of Epoxide 206. 0. Fuller's earth removes Epoxide 206 from PII solutions to the maximum extent of 0.23 groms/gram. Tho following conclusions arc based on the results of this study: 1. The storage of thermally activated fuller's earth Is troublesome: unless storage conditions are carefully controlled, the activity of the earth con be drastically changed. 2. Fuller's earth canr.ot by itself be used to purify Unox 206 to any gross extent, (excluding the reported lowering of the acid number). 3 Fuller's earth takeB up about 70$ of its own mass of Unox 206 upon slurrying with the as received liquid. k* Rapid, thorough mixing of fuller's earth and Unox 206 is essential to minimise localised reaction and consequent major heat evolution. The preferred laboratory scale order of mixing is to add liquid (one quick pour) to powder with continuous, effective stirring to achieve intimate total-contact in a minimum time. 5* The slurrying of fuller's earth with undiluted Unox 206, like many chemical operations, is potentially dangcrouo; especially in view of our limited current understanding of the reaction(s) involved and how such reactions are initiated. Accordingly, normal safety precautions should be guaranteed by competent p/erconnel attempting such mixing. Adoqu 1to stirring, heat dissipation, ventilation, and perhaps protective clothing arc considered minimum needs. HONS 0348ZZ CONCLUSIONS (oont'd.) 6. The runnwny polymerization of Epoxide 206 appears to demand an acidic environment. Under no circumstance should free acid be added to Unox 206 nor should frech Unox 206 be added to a reacted slurry of 206 and fuller's earth. (The latter probably results from improper mixing of earth and 206, and is the cause of runaway observed on some past, occur,ionn.) imTOucrTM The treatment of Epoxide 206 with activated fuller's earth has resulted on occasion in thermal runaway. The work described herein was undertaken to learn some of the fundamental physical chemical facts that characterize the oyntems fuller's cai-th - Epoxide 206 and fuller's earth - Pyranol II. (Throughout this report the term Pyranol II is used to identify any con centration of Epoxide 206 in-11199 lyrur.ol.) -- .. The basic information sought included meensures of specific heats, heats of miction or adsorption, tlw nature of the reactions (physical and/or chemical) and the relative quantities of the reactants. Reasonable answers have been Obtuined for the questions osked, end hopefully will be useful in further laboratory and manufacturing applications of the materials. mymiMr.j apparati in;__and wispHonr. To cnouxe reliable on^^prcser.tative results, care was taken to obtain Itnd employ in this t$"dy only virgin materials typical of those used in normal Department c/erations Epoxide 206 (Vlnylcyc lohexc^T^lfepoxide ) 1*199 ryranol Fuller's earth - An unopened 1*5 pound can of Unox 206 obtained from A. Abbott (can marked "3"j on Fittsfield tag 005 S/n 7912). - A carefully cleaned milk can was filled with refined storage liquid. - Approximately 25 pounds of Attapulgus 60/90 mesh were obtained from a new bag supplied by C. Fuller. A small sample of 200/325 mesh material was supplied by V. Shedigian. The activation, in small quantity lots as needed, of the fuller's earth was econgdlshed by heating in air at i25-,i50*C for four hours. Storage cf the activated earth was in air at 150C. Immediately before use, the earth was removed from the 150*0 oven and cooled to room temperature in a vacuum desiccator. (Storage of the activated fuller's earth in air at 150C is probably less desirable than storage in vacuum. Failure of the only vacuum oven Available for exclusive use in this study forced the air oven storage to ensure uniformity of product throughout the prolonged experimental period.) Active epoxide was determined in the Unox 206 end in all Pyranol II solutions by the standard HBr titration method regularly employed in the Department(). HONS 034023 -3- COHCIAJOIOMS (cont'd.) Calorimetric measurements were made with ti simple Dewar flask calorimeter eonntnicted for this study. The essential components of the apparatus included a stirring assembly (both magnetic and propeller rod stirring were successfully used), thermometer, and a glass enclosed, helically wound plunar heating element (fabricated to order in the Research laboratory Clans Chop). The power input to the heating element was variable transformer controlled and measured with a OK type P-3 single phase wattmeter. KXMffllMKNTAL TMCHNKiinSS * The adsorption isotherm data were obtaijfefc^by stirring 10 gram quantities of activated fuller's earth with POO g) ir./ <r each PII solution. Equili bration of the slurries was achieved o.'V nmUng within two hours following an Initial one-half hour vigorous stir in';. Final titrations were performed one to two weeks following the mixing/'f/-he reactants. The titrations were performed with about/ 5 gram filtered portions of oupernntunt liquid to determine the amountin'.' active Epoxide removed by the fuller's earth. Ml samples were equilibrated and titrated at room temperotuiv. The data were treated conventionally: the amount of Epoxide 206 removed per gram of fuller's earth was plotted against the equilibrium concentration of 206 in the P31 following the reaction (Figure 1). Calorimetric measurements were made toy observing the maximum temperature reached an a result of oach reaction, and then heating the cooled reaction mixture to the measured maximum temperature with the calorimeter heating elew'nt. The rate of heating was chosen to duplicate the observed rate of the studied reaction; the measured power (wattmeter) multiplied by the required time to heat the mixture to the endpoint temperature_is directly equal to the energy of the reaction. Initial calibration of the Dewar flask calorimeter (necessary to enable calculation of the specific heats of Kpoxidc 206 and 1^99 Pyranol) was accomplished by measuring temperature Increases upon dilution of aqueous sulfuric acid solutions. The difference between observed and calculated temperature increases is a measure of heat loss to the calorimeter. msuitc 1. Specific heat determinations at atmospheric pressure. A* Epoxide 206 199 grams, 25*5*C to 8^.t*C, 6.4 keal.; loss to calorimeter - 2.2 kcal p * ` 0-36 CRl-/e-acs c. B. 1*199 ryranol 199 grmtm, 25-3*C to 09.1'C, 4.9 kcnl. C,, (4.9-2.2)______________ (199)(09.1-25.3) " 0,21 col./g-deg C. MONS 034824 WSSUI-Tfi (cont'd.) 2. Percent active or titratnble epoxide in Unox 206. l'he average of several determinations made over a period of many weeks is 95.3i t>y weight. No attempt was made to isolate or to identify the non-titratoblc constituent or constituents. 3. Adsorption Isotherm. The basic data characterising the reaction of activated fuller's earth and NI solutions are tabulated in Table 3 and shown graphically in Figure l, Kach datum point was obtained with an equilibrated slurry of roughly 10 grume of fuller's earth and 200 grams of PII solution. The straight line plot indicates the data conform to the Langmuir equation that describes the adsorption of a monolayer on a crystalline solid. The calculated constants for the Langmuir equation, obtained by measuring the slope and intercept of the straight line, arc as follows: a . 7.0 x.1.0"'' b * 2W The J/mgiwIr equal.Ion is expressed in linear form as: S--/-X * iC + i (x/m) b a c the equilibrium concentration of the adsorbed species x r mass of adsorbed species taken up by the adsorbent m I-, maos of the adsorbent a,b r constants characteristic of a particular adsorption system (b is proportional to the number of gram moles of adsorbed spccico for square centimeter of adsorbent). Saturation of fuller's earth with ftpoxidc 206. A* Kiss Increase of fuller's earth. 38.1 grams of activated fuller'b earth were slurried with 200 ml. of Kpoxldc 206 in an icc bath. The solution was stirred continuously for about 20 minutes and then allowed to stand undisturbed about one hour. The solids were then collected by suction filtration, and pressed between sheets of dry filter paper to remove excess liquids. The product finally obtained was apparently dry, and had a mass of 7^.3 grams. After standing overnight at room temperature in an evacuated vacuum desiccator, the mass of the saturated earth was -- approximately 6$ grams. The moss increase of the fuller'a. earth following saturation by Upoxldc 206 equals: 7!>-3 30-1 . o, 6* or, after vacuum drying; 6.i> 30 . 71# 30 MQNS 034825 WSUI.TS (cont'd.) Subsequent vacuum drying (an additional 24 hours) did not further reduce the mace of the saturated earth. Repeated determinations of the maos increase of fuller's earth upon saturation with Epoxide 206 gave results Of 301.6# and 102.6#; thus yielding an average maos increase of 100.6#. - - B. Loss of Titratnblc Epoxide Following Saturation of fuller's earth. (a) 1$>8 grams of 206 were slurried at room temperature with 50.2 grams of fuller's earth. After equilibration, the filtered supernatant liquid van found by imr titration 92.1# active epoxide. The active epoxide lost wao accordingly; (95*3#-92*lft)(.V;f0)-''j.3 gram;; or 100 r 12.6# of mass of fuller's earth. *50^2 (b) 201 grams of 206 were similarly slurried with 6l.O grams of fuller's earth. The epoxide lost was: (9!>-3-90.6#(20l) * 9-5 grams or (9.5) 100 * 25.5# of mass of fuller's earth. 63 .0 (c) 199.9 grams of 206 were similarly slurried with 41.0 grams of fuller's earth. The epoxide loGt was: (95 3#-9J* -0#) (199 9) 1 -0 grams or (1.0) 100 * 2.4# of mass of fuller's earth. Ml .0 5. 1499 Fyrnnol - Kpoxlde 206 Saturated fuller's earth. The possible desorption of active 206 from fuller's earth by 1499 Pyranol was examined by slurrying 10 grams of the saturated earth with 100 ml of J.499. After standing overnight, portions of the liquid were titrated by the standard HBr method. Results are tabulated below: Blank Titration 1499 Pyranol - 0.00# Epoxide Fuller's earth 1499 i Epoxide in Equ: 10 g 10 g 10 g* 10 g* 100 ml 100 ml + 500 pjpa V 100 ml ' 100 ml 4 500 ppm v 0.75 0.72 0.42 0.44 Stored about 48 hours in vacuum desiccator. 6. Calorimetric ItoaBuremcnts. The bulk of the data obtained with slurries of Epoxide 206 and fuller'g earth arc summarized in Table II and plotted in Figure 2. HONS 034826 -6- ktow.ts (cont'd.) Delta obtained for PII solutions mixed vith fuller's earth fit in well with the results for the undiluted 206 and fuller's earth as noted. The effect of particle slide of the fuller's earth was observed in one experiment in which a sample of 200/325 mesh earth was used Instead of the usual 60/9O menh material, (insufficient material was obtainable for any further work.) The heat evolved was significantly greater than for compar able experiments with 60/90 mesh fuller's earth, but not propprtional to the calculated surfaces ratio: 60/90 mesh * average particle diameter about 200 microns 200/325 mesh - average particle diameter about 50 microns assuming spherical particles and constant density, the ratio of surface areas for equal masses of the two materials is (200)3 (50)^ , tsojr tsmF * u The ratio of heats evolved was 3*7/2.6 * 1J1. Several attempts were made to Initiate the runaway polymerization of the epoxide; all failed except two. Preheating the Epoxide 206 prior to mixing vith fuller's earth resulted in no greater heat evolution than obnerved for comparable quantities of room temperature mixed reactants, preheating was tried from 50*C to the normal boiling of the epoxide (227eC) without effect. < Heutinc of the reaction slurry to temperatures as high as 227*C-likewise had no measurable effect on the apparent nature or extent of the reaction. The order of mixing the reactants also hod no apparent effect (contrary to Shedigian's experience - cf. 2). Comparable heat evolution was measured for slurries made by adding liquid to solid and by adding solid to liquid. Thcimial runaway, resulting in the violent boil off of liquid and the pro duction of a nearly intractable ma6s, occurred only twice in the course of this work. Both Instances (one unplanned, the other deliberate) had a common background: fresh Epoxide 206 was added to a hot slurry of reacted 206 and fuller's earth. In the planned experiment, 10 grams of activated fuller's earth were slurried with 200 grams of Epoxide 206 preheated to 125*C. After one hour, the temperature of the slurry had dropped (in steady fashion) to 87*0. At this stage it appeared nothing was likely to happon, so heat was applied to raise the temperature to 150*C, then to 200*C, and subsequently to the boiling point with a holding period at each temperature of several minutes. At no time during the experiment was evidence obtained that indicated any reaction beyond the initial adsorption illustrated in the tabulated data; that in, as noon as heat van no longer applied from the calorimeter heater, the temperature of the slurry started to decline. HONS 034827 -7- 1? rUM.'r* (cont'd.) After approximately 20 minutes, following unsuccessful thermal attempts to Initiate polymerisation .of the epoxide, 25 grams of fresh room temper ature Epoxide 206 were added to the 6lurry at a temperature of 220C. Immediately the slurry temperature dropped to 200*C, and then gradually rose over the next hour to a maximum of 273C. On subsequent inspection of the contents of the calorimeter, only a colid, dark brown maos wao ob served. The polymerised epoxide fused the fuller's earth particles into the nearly intractable material mentioned earlier. (Equipment cleanup was achieved only by cooking in hot caustic and by mechanically chipping away the "comont.") The reaction terminated short of complete violence only because the excess liquid reactant boiled off before it could react. Continued boiling of Epoxide 206 in air also fails to result in uncontrolled polymerisation. Treatment of 206 with aqueous potassium hydroxide leads to the formation of yellowish, gelatinous material after many hours, but nothing more. Tho addition of aqueous sulfuric acid (concentrated) to excess Epoxide 206 yieldo within a few minutes red colored, gelatinous material. Shortly after the formation of the gel, a violent exothermic reaction suddenly occurs leaving only a black (largely carbon) residue. The addition of dilute aqueous sulfuric acid to excess Epoxide results in on instantaneous explosive reaction and the formation of a resinous material. DISCUSSION The data of Table I are at least prima facie evidence that the reaction between fuller's earth and Fyranol 11 solutions is strictly physical - as one might predict. Saturation occurs at equilibrium concentrations greater than about 1*<$ Epoxide 206 and corresponds to 0.23 grem6 of epoxide per gram of fuller's earth. From the shape of the curve, it is evident only the Epoxide 206 is adsorbed from the solutions, and there 16 no desorption of active epoxide by 1^99 Fyranol over the concentration range 0 to 70$ Epoxide 206. The saturation of fuller's earth by Epoxide 206 is difficult to evaluate uniquely as evidenced by the variation in results obtained by different techniques: 100$ to 70$ approximate tnkcup by mass increase measurement, 2,ii$ to 15*5$ by titration, and 23$ from the adsorption isotherm. It seems reasonable to assume the unusually high value obtained by mass Increase determination might be a consequence of the presence of excess Epoxide 206 lit the saturated earth despite best attempts to dry. It is also possible that the temperature control (ice bath) used during the mixing of the 206 and fuller's earth for the mass gain determination may have influenced the adsorption sufficiently to account for the observed result. From the standpoint of reconciliation of results, the 100$ and 70$ values might be dismissed; they have practical utility, however, from the point of view of the preparation of saturated earth for manufacturing application. (6co added discussion in addendum.) MONS 034628 -0- DISCUSSION (cont'd.) The difference between the isotherm value (23#) and the direct titration values for raw Epoxide 206 and fuller's earth is not readily explained. Aside from the evidence presented by the isotherm (that I*i99 1b not ad sorbed from III solutions), one would predict the effect, if any, of 1**99 would be to Interfere with end lessen the adsorption of Epoxide 206 by fuller's earth. Actually, the opposite appears to be the case; but the results nuty be confounded by the temperature difference during slurrying os noted curlier. (See addendum.) The results of the experiment aimed at elucidating the response of Epoxide ;o6 saturated fuller's earth to 1**99 *Vranol with and without added water arc also partly indeterminant. Prom the only data currently available, one cannot tell whether 1**99; with or without added water, strips little epoxide from the saturated earth or whether posBibly stripped material in no longer HBr tjtratable. It i6 possible the adsorption of Epoxide 206 by fuller's earth is non-reversible, but it may be the present dntri should be interpreted simply as indicating Epoxide 206 is tenaciously held by fuller's earth and cannot be stripped by 1**99 Pyranol and a small amount of water. These came data also sited additional light on the inter pretation of the takeup of Epoxide 206 during saturation of fuller's earth. Assuming the previously discussed 70# takeup figure is erroneously high because of occluded free Epoxide VO6, and the correct measure of adsorbed epoxide is approximately 20# of the mass of fuller's earth saturated, one calculates that 10 grains of vacuum dried saturated earth contains about 2 groins of free Epoxide 206. Such free Epoxide 206 should be readily soluble in 1**99 ryranol, and the concentration of epoxide in the equili brated 1**99 should be at least 1.**# instead of the measured 0.**#. The assumption of a significant amount of mechanically held Epoxide 206 in saturated fuller's earth is accordingly untenable, and one should perhaps not be hasty in dismissing the 70# takeup value. The calorimetric data arc essentially straightforward and agree veil with the usually noted values for heats of wetting or adsorption up to the l*i# fuller's earth concentration. The differences in observed heat evolutions and the extrapolated heat of adsorption above the 1**# point are of the correct order for heats of polymerization (roughly 10 kcal/mole). That the large amount of heat released, especially in runaway situations, is attributable to polymerization of the epoxide is only conjectural. The evidences, coupled with possible reactions, however, suggest there is little likelihood this conjecture is incorrect. The real question is what causes the runaway condition that is occasionally encountered? The basic facts Beem to dictate acidic catallzatlon as necessary for poly merization of Epoxide 206. To weave this in with the observation that runaway occurs when fresh Epoxide is added to a hot slurry of reacted fuller's earth and 206 is not immediately straightforward. Among other possibilities, suspicion is directed toward the roughly 5# of non-titrat-able constituent^) in os received Unox 206, toward conceivable reaction product(s) MONS 034829 9' WSCUSSION (cont' d.) renulting from admixture of 206 and fuller's earth, end toward some inter action between these constituents and products. Reacted slurries are in deed mildly acidic as determined by aqueous extraction o.nd conventional pH determinations. If acidity is the key, why then is it necessary to add fresh Kpoxido 206 to a slurry already containing & gross excess of 206 before runaway polymerization occurs? A pooaible explanation lies in the reported lowering of the acid number of Unox 206 following fuller's earth treatment* The reacted slurry may have lost acidic contaminants present in the original Unox 206 to such an extent that polymerization io not possible; the addition of fresh 206 could, however, provide sufficient acid to CBtfiblliih the final necessary condition for runaway polymerization. Obviously, little is contributed by continued conjecture concerning a limited store of facto. Further work is clearly dictated to elucidate the details of the runaway reaction. In the interim it appears safe to con clude that any practical application of the subject materials should be cautiously steered around the condition of fresh Epoxide 206 being added to n hot, reacted slurry of 206 and fuller's earth. Without regard to the mcchnnicnl details involved (which of course can be limiting), the preferred method of slurrying fuller's earth and Epoxide 206 seems to be to odd Urn liquid quickly to the solid rather than the solid to the liquid (with apologies again to Shedigian, 2). Rapid mixing of the bulk of the two materials in the essence of the technique required to avoid localized effects that often lend to partial fuming and charring if not full blown runaway. The transfer of the liquid rather than the powder tends to minimize localized effcctfi better than the opposite approach - at least on the* lnborutory scale here reported. External cooling to limit the temperature rise of the slurry is of course immediately suggested as a nccensnry adjunct to the safe handling of quantity lots of the materials. The calorimetric data should permit reliable estimation of the required cooling capacity for a desired operation. It i6 a moot point, however, whether or not external cooling can fully arrest polymerization once initiated. Kindly, some coniocnt on the effective activity of any batch of fuller's earth is in order. The bulk of the work reported here was done during the period of July through August, and many small batches of fuller's earth were activated and stored as earlier described. On several occasions results were obtained that clearly indicated the activity of the earth is particularly sensitive to time and conditions of storage following 'nitial high temjierature activation. Storage in a good vacuum oven at perhaps 100*C would be the writer's choice to ensure optimum activity of the earth. Social care would bo required, however, in handling such materiel to void certain, probably rapid deterioration on exposure to air (particularly during high humidity periods). As noted, air storage at 1^0eC was em ployed for the activated fuller's earth used in this study; it is con sidered that all the data here reported may likely have been significantly (but uniformly) influenced by this storage. That is, greater adsorption and increased heats of wetting may well be observed for comparable reactions involving activated fuller's earth stored in vacuum above room temperature, and exposed to air for a minimum time prior to reaction. HONS 034830 -10CONCLUSIORS Much remains that can be done to further elucidate the nature of the studied reactions. The question is whether or not the information coined would be worth the required efforts. For tho moment, ot least, the writer in of the opinion that enough time has already been spent and any further effort should be dictated by a specific, practical need involving n particular feature of the basic reactions. Further titration data and a one shot infrared absorption analysis hope fully can be added in addendum to aid in answering the question posed regarding tho roaction between Epoxide 206 - saturated fuller's earth and lll99 pyranol. Aside from this eleventh hour burst, the study is probably best considered closed. HONS 034831 -11- addendum The following data were obtained after the writing of the main text waB completed. fintmvitiarLof Fuller's Earth by Epoxide 206 ])O,0 grams of fuller's earth were.* o)urrlocl with 200.grume of /.po/iido 206 Jn on ice bath. After standing overnight, the percent active rpoxidc in the supernatant liquid (qb determined by llflr titration) was only 0.nr,j less than .'n the original Unox 206 - corresponding to lcco than 2# takeup by the fuller's earth, (it io highly doubtful the observed 0.*# difference io significant. Accordingly there ie no real evidence of any preferential takeup of active epoxide in thin experiment - or in the earlier noted experiment in which n calculated 2.V/> takeup was recorded.) ' In another preparation of Epoxide 206 - saturated fuller's earth, an excess of 206 vac elurried with an unknown mass of fuller's earth. After overnight standing, the percent active epoxide in the supernatant liquid was found by titration 5/> less than in the aturting Epoxide 206. The four separate titration determinations of loss of active epoxide following saturation of fuller's earth will be noted to constitute two pairs: (1) about 1V/> loss end (2) about 0# to 2yJ loss. It will further bo observed that the 0j to 2>* loss pull* were slurries that experienced relatively small temperature rises (one because of ice bath cooling, the other because the proportions of fuller's earth and Epoxide ?06 yielded little more than the heat of wetting), whereas the lh<i average lo69 pair originated from slurries that reached temperatures at least 100*0 (because the percentage fuller's earth in each slurry was well above the level at which polymerisation of the epoxide is initiated - Figure 2). One con conclude from these observations that fuller's earth does not selectively adsorb active epoxide from Unox 206, and that the titration Ions measured for slurries that have been relatively hot is due not to preferential adsorption of active epoxide but to loss of active epoxide through partial polymerization. The final experiment involved the mixing of 10 grams of Epoxide 206 saturated fuller's earth with 100 grams of 1^99 Pyranol. After overnight standing, aliquots of liquid were taken for HBr titration and for infrared absorption analysis. The results arc tabulated below: Epoxide by titration: Epoxide by 1R: Aliphatic by IR: 1.265!; 1.275! (v/w) 1.295;; 1.305; 1.335i; 1 3'.* The difference between active epoxide and percent oliphatics is most likely not significant, and accordingly suggests there is no stripping of inactive epoxide from the fuller's earth by 1499 pyranol. The relatively high values (i'.$;) undoubtedly were a result of the use of saturated earth that had been suction filtered and pressed but not vacuum dried; that is, excess free MONS 034832 epoxide van obviously present in the earth when slurried with the Pyranol. If one in fact assumes the 1.3 crams of titratablc epoxide represent Mechanically held 206, the calculated amount of Epoxide 206 lost on slurrying with fuller's earth is 71*# of the mass of the earth (maos increase of...earlh after slurrying 100',therefore 10 crams of saturated earth contains !> gramr. of cuith plus 5 grains of contained Epoxide 206 - but 1.3 Grams of the 5 Grams of 206 are soluble, and titratablc in 1**99 Pyranol; 3*7 grams of 200 are therefore held by the fuller's earth). The agreement between this 7'*Vj and the maso increase value for vacuum dried saturated earth is at least an interesting coincidence! ' The final conclusions based on all the reported results is that fuller's earth non-selectively sorbs and tenaciously retains about 70 its own weight of Unox 206. The non-selective sorption leads to no change in composition of the 206, and accordingly the apparent discrepancy between mass increase and titration determinations of 206 uptake by fuller's earth is explained. The titration change noted in some cases is the result of lots of active epoxide through polymerization. The sorption of Kpcrldc 206 from 1*199 Pyranol solutions is selective os evidenced by the adsorption isotherm data (Table I - Figure l). Why the sorption ir. selective from solutions and non-selcctive from undiluted Unox P0(y in not obvious. It oeomr. obvious, however, that on* cannot expect to purify Unox 206 simply by equilibration with fuller's earth. ACKH0UT>;W1^KT All of the experimental work here reported was performed by A.J. Munichiello whose assistance is gratefully acknowledged. A. Kfttchman provided details of related unpublished work, and other helpful information and advice. X. Guiles performed the 1H analyses. S2!ES3iS2 1. Toot Report HF-63-67, P. Scherer, Stock Qualification Data for Epoxide 2<XS. 2. Test Report HF-6*-1007, V. Shedlgian, Purification of Epoxide 206 with Activated Puller's Earth. 3. Memo Report HK-63-931# A. Katchman, Epoxide Stabilization of Capacitors. 4. Durbetaki, A.J., Anal. Chem., 28, 2000 (1956). HONS 034833 C(6/100 b) 0.0 0.1. 0.5 0.5 -3 J( .0 i|.3 5.7 6.5 0.0 9- ao.it 13.1 33-3 13.3 13.6 15.1 i?8.6 22.7 1.6.0 I|0. u 69.8 TABLE X ADSORPTION ISOTHERM x/n (e/e) 0.0 0.17 0.10 0.09 0.14 0.10 0.19 0.22 0.19 0.21 0.17 0.25 0.23 0.23 0.25 0.23 0.22 (0.38)7 0.24 (0.43)? 0.23 0.21 . x7m (b/s) _ 2.4 5.0 5.6 16.4 22.2 22.6 25.9 3*1.2 41.0 5J.o 41.6 57.0 57.9 53-2 59-2 68.7 (70.6)7 94.8 (107. )7 202. 332. TABLE II CA1PRIMKTMC MEASUREMENTS (Epoxide 20<S~+ fuller's earth) 5 Puller'0 earth kcal Evolved Time to Reach Temperature Max. 4.60 4.06 5-15 6.90 7.03* 10.3 12.0* 13.5 13.0 17.0 20.2 22.0 23.3 24.6 27.6 9.2- 1.32 1.24 1.76 1.44 1.50 3.16 3.60 5.30 4.05 5-67 9.14 11.72 10.85 IO.85 15.1 ^ 3.74 1800 sees. 900 9900 420 13100 2640 2580 >2800 1100 660 540 420 400 360 *P11 solutions, about 71# by weight Epoxide 206 (# f.e. relative to Epoxide ?o6, not to Fll). /Calculated fron Cp data a:.d observed tem;n;>.ture iquxlu'um of reaction. ^800/325 mesh fuller's earth. MONS 034S34 S C 8 *e o SNOW FN-i:i-3 UM i scN9t nccmc cc*****. seaciccnw. n. r, sja. i. C. :c..c; o 7/i'.~ : .: i--- C ' " , I{ tx > I l MQNS 034837 GEKERAL v, ; ELECTRIC A1 331 - S'/ SPECM'K.ATION' CHLORINATED DIPHENYL LIQUID INSULATION ___________________________ _____________________________ ____ _ Superset-.''3 r.:i2]::c O-v Material A13H1 Identifies chlorinated dlprmyl liquid Insulation, at follows A13B1 - 12 C pov.r point A13B1C - Minus 1< C pour point AlSttlE - 25 C pour point WioPrUWirr.s Specific eravlly at 65/15.5 C ----------------------Iwfraetlvo Index at ?f* C .......... ...................... ......... TOUT, acid !!<.. tnR KOH/r.* "X --------------------------- A13B1 1.495*1.505 3.(370-1.6390 .010 A13P1C 1.58)-].5jJ (1) 3.6240-1.6260 .010 A13MS 1.S55-1.5-56 (3) 1.6455-1.47 .014 Krm> chlorl.toe, ppm, max .............-........................... \T)A (3) flulfalert --...................................------------------------- Mor.e .05 None HDA (5) - Vain*. pp:, max --- ....................-........................ 35 35 ss VtncoMty, Saybolt Uiilvonjal, sccondr.? At 37. C ...................................... -................................ 82-92 . At 66.9 ft .........................................-............................. 44-<a Mr.ttliailon ran,'.c, d.f. C: ' 72-76 tCru di: tilled -.............................................. fcf.Ii' dtHllliad ................................................................ tCrf dl.' tilled ................................................................ JeC-JTS 371-J8J 3* ?-394 325 r.tn - 350 max set-555 (<) 390-404 (4) sod-(,) pour point, d*6 C, max -- --------- -- ---- FLiriipelr,:. deft 0 ................. -..............--.................. - -14 170-200 25-34 _ Mire point, deft C, win .............-- -- ------ - Color, r.,..r ............ --*----------- ------------------------------ <0 Nona 40 . 350 150 ? Melees rsc atrenftth, 20 C, KV, min ----------------- 35 WiI ret Me constant at iV' c ---......... ............ 4.15-4.35 35 4.70-4.50 50 (5) !.; . * . *at *rc> ft, ehm-ern x 10^, nin ---- 500 DreV.lerlr.jt *oi>, ppn, max ----------------- ------------------- 500 1.0 500 . Cen'cMc; stability - After a ar.mple cont.ilnir. ; ene srom of aluminum fell is heat ed for si/, ho:: at ?K* C, II'.'* alur.lnar. foil shall be free fr v eorroalon a ; determined vinutiliy or by onr Ir. well*,hi and the sample shall conform to the f >1 lowing reoui rements: TolJl or Id Ko., m.; KOH/ft, max -------------- - Mme chloride?!, t*P?* max --............................ - Colo*, n.ix ----------------------------*----------- -------------- . . - .014 NOA (3) 150 Con.il Men - --- ------ - - Clear ' {)) At MVlJi.h C. (?) At AsVifi.ft C. (3) So >*< ,'ctntle omenit. {*.) / d!: ! 11,M 0- mined or. a wel-'ht bar,in ('.) At &3 ft. ka;:. P.iWM * M <:] arlra V/ 1 .ilph*-.y; W lmjM .-K : - MMtrlol shall bo clone ar.i free i.oi,jV*.r f.-:rt*.:lej or ow-:* crntr.!;u:! App:\\ y - After a r.itn-ial hat, been approved or rif.'nir.ii .7.rt;.oft;t bo r.l; uithej: r t*;?, dll lnj iri.cun lr.puritiee such nr. - n-i-'r thin specification, no charge ir or cpproval sf the purchaser. Rm.wi: l Irr (Al'ilC) (Continue. HONS 034033 GENERA!./'; t IfCYRIC r.:?r.r.*:Yi. iiom i?:volavic *>;s:-r: Pec*.- 2 Vj:fr7-s'. f: ("?.:inurd) ' ....................................................................... ................................................................-................ pour point .............. ............................................ -- -- *---------------------------------- Mt.rh ft Fire pclr.t ------------------------------------------------------------------------------------------------ --------- -------------- fMcr ........................ -............................................................................................................................. .................. Corral- olcbimy .......... -............................................................................................................................ Dielectric Mrcnr.lh ....................-................................................................................................. *................ pleleetrie constant ---------------------------------------------------------------*------------------------------------------------------------ Ncfilati `.'1 ty ----------- ----------- ----------------------------- ------------------------------------------------------------------------------- 1* < hiorinet Ion ............ ......................... .................... ................................................-..............-................ AS7K D2C-56 (?) ASiV. r?7-S7 (8) ASTM D92-S7 ASTM D901-56 O-E S4C26 AS7K D877-49 ASTK t>?24-49 O-A L'4D2 {i-Z KtC'.S (?) p.y.eept that percent distilled is determined on a weight basic with the following .T.odlflcitlcme In proceduret Correct rebuilt. Tor rricrc.frt thermometer stem and barorretrlc pressure. For MSME or.ly ute a hooted condenser eor.ci rt J.rg of a center glass tube 5/8 Inch 1"; by 5/.; Inch OD ty *8 1/4 Inches lore, cut cfr at 45 cicr.t on lower end. V.'rap for 18 Inches with 48 lurnr spaced 3/8 Inch on center with flat Nichrome wire .CC5 y. 1/3 Inch. Frlnj*. out ende of wire through holes In outer Jacket. Outer Bless Jacket jr. 1 3/6 inches OD and Is reeled to center tube ut upper end. Lower end it drawn down to fit rnugly against center tube but. \t r it ccr.lcd to It. 7hc coridcrucr heat la controlled by p Vnriae to an Internal t'M:.rerature of 125 t 10 C which *> rtanted when the heat to the flask ir. turned or.. Adjust the distillation rite to 50*70 drops per minute and collect the distillate Ir. a tared flask set on a bear, talanev. (0) hxcept ut'Jr.f, only a single bath continually cooled to pour point. jamuj.*, AVVHCV.M s (A3 3l`l and A13MC only) TlK' supplier shall forward to the laboratory of the purclu .*?r at the point or delivery a two quart terple tsVer. from each r.snufacturir.6 bstch of this materia: scheduled for shipment to the Gtreral Electric fr.pahy cr others designated by the Crr.crel Electric company. This sample shall be de* Jesw.ted ir. "Advance rumple, Lot Ko.___ 0*E K.'ttcrlcl w Apjr.*,T.* ef thin iidvnrcr rnrpte by the General Tile Ctrl? Ce'vnny Coon r.c`. constitute approval cf the Shipment. Ccr.ern) Fleetrlc Company approval will be based upon the natcri/O ea received i.t Itf. dert lurit lor mwl will tr booed not only or conformity to the properties specified herein but on conformity to nil performance characterlntlca deemed essrr.tli.l by the purchaser, CKAfTfMATF OK TliiTl . ?Ik vendor Mifill submit promptly to the laboratory of the purchiser at the point of delivery a cert 1 Mr/'Ve of tent in duplicate so that It will bo received on or before receipt of shipment. Tne cirUMCiit** of test ohnlJ Include the followlnc Infonr.atlon. ..},\.lf:c gravity C:)or CM.Utiun /.?!(! number 1,;;.t! 1 l.'.tlon rr.r.fe f<.-fv.' Cvl ve Index !.<e chlrt'lden Cvrvotilon atweility (A1JME only) Voter content Dielectric strength (Al.tDl /; A131C or.ly) Dielectric cor.rtu.nt (Al.Vsl k AluhlC or.ly) Resistivity (Al:-M -i AicHIC only) Pour point Viscosity (AliMS only) Flash point (.U/MC only) Dechlorination (/.l.f;':l ft Ali?lC only) ThJt cev.IflCMe shrill also contain the C-K cierlenwiioti, purchor.*; order number, tank c.*-r number, 3*.*. and d.iti shipped, so the. the certificate may te identified wltn tut bh'.pncr.t. ,".i' eri; 1 iihaii be ihipped In t: nk ccr;* or ?rural), i s specifled or. the purchase order. Ali cI Mr., ctuil) n;i!'we to IOC 0"*; And shall be of a type having, nr- delete vir..w effect .*!. the r.fterl:; l. Kx'.rc.r care r.hsll :t c <.. re! r.*3 hy the aupjlier In thv ce 1 or. ar.ti u:;r- of cohta livers or c:.:.',. !r.ru jrov`.i-;i by the Kujpiifr. 3ucl. cor:t.' '-h-.-a 'r.-Ml ; c:.:*c- !' ir.; : jt*:-.! by the uui pl'.er to or.v.urt: th..: they are free of Injurious T-n i^r. ir;.-n.er. .n't. L-f the r.jteriiii, tn spcclfiei r.w. In, upor. t.rrt .ml of t;-,o r.hi;.:.c:.t i.i pd.iv.s vf dell .tr C*s: .n-.-i ton i.nd the lot number. HONS 034839 GEFJERAL0 ELECTRIC CAPACITOR DEPARTMENT SPECIFICATION FOR FYRAHOL II 4 WtltHdl idMUfln epoxide modified chlorinated diplienyl liquid insulation AJOPOflA - Pyrenol 11 - Epoxide 0.35* A50rC8N - fyrsnol II - Epoxide 0.25* A5OrC0C - Pyrenol II Epoxide 1.00* follows: >4Awt50M PUCiiii8w-iSXt3' IM IXI( CmHi NmMm February 11* 196) Supersedes AfrOPCS-S? fiW'Bffiai AX)pc6A peeifie Orevlly..........-.....................-........................--...............----- iTyjY -1.3S Refraction Index ............................................. .................................1.6?1*0 - i.( Potal Aetd Ho, ,* K0*l/g, {box).................--................................------ Wster Oontsnl, MM, Max ...................................... .05 20 Sulfate...............................................................................*.................................... None them*I Stability Chlorldoa (2 hr*, st 940 C) WN, stax ....... MM Active Epoxide Content - * by weight ......................... 0.3)1.03 Mcleetrlc strength, 95C, KV, siln 35 Mclrelrle Constant st 100 C --------------------------.........-------------- I*.70 - 4.90 Resistivity st 100 C, elw-tn X 10$, min ......................................... 1100 Poor Point deg. C, nox..................................................................................................14 Flashpoint deg. C ........................................-------.............-- 370 - 200 rire faint ........................................ Aliphatic Oontent, mi * -Actual Asttvs Epoxide Content * plus Hone .01* MAIMrACTVHEi A5%Q1PfCcrB6-rbB1.39? 1.6240 l.fefo .05 90 Hons ' NDA 0.231.03 35 4.70 - 4.90 3500 -14 170-200 Hone .05* A5o>rrcc8Occ l-. ,Wl - 1. 39* 1.6240 - 1.6260 .05 90 None MM 1.01.10 35 4.70 - 4.90 1)00 -14 170-200 None .15* Hsterlol - Shall be blond of ehlorlnotsd diphenyl (A13B1C) end epoxide (l>5071A) Impurities Miterlsl shell be clear snd free from all Injurious Impurities such ss Mtslllc or non-Moullle psrtleles or other contamination. myixtMi Hirrwouii tempi 114...............................-- ...................................................................................... Specific Orovity ........................ --.......----------- ..................... Refmotive index ........................ ...........-----........---------------------------Aeld number------- -----------------------------....------------------------ ....---------------Sulfate................................................................ ........... ......... .............................. Free Chlorides.............. ................................ ......------------- .....----------... Water ...................................................................................................................... Pour point -.......... --------------......... ..--....----------------------------- -------flash end fire point ............................ ....... .............. .......------------- -.... Mcleetrlc strength ........----------------------------------------------- .......... Melretrle constant ........................ ............. .....--.................... Neelstlrlly.............. -................................................................... -..................... Active Kpoxlds content ...................... --.......................... ----------------------Alt|datlc content ...................... ......... ..........--------- ---------------------- --. (]) Except use 0.0) N aqueous KQtl (9) Except using only single bath continually cooled to pour point Asm D90I-56 Asm 1901-56 Asm 1901-56 Asm 1974-55 <i) Asm D67M9 0-B Efccfcic ,0-B B4C45 . Asm 097-57 (2) Asm 199-51 Asm te77*49 Asm 1994-49 0-G G4ce 0-E G4C50 0-B E509C95 Wllill HONS 034040 o t' I G HONS 0 3 4 8 4 1 GENERAL^ ELECTRIC SPECIFICATION' VINYLCYCLOHEXENE DIOXIDE 0S071 Super.odes S.D: DO NOT SEND OUTSIDE GENERAL ELECTRIC COMPANY 0-E Material DSD71 Identifies vlnyleyctohexene dioxide, as follows: DSDT1A - General purpose rtlOl'CHl'IKJi Specific gravity, 20/20 C ............................................... ......... Viscosity ut 2*o C, eentlutoke, max ----........----- Total epoxide (os oxlranc oxygen), %, by weignt, min Acid number, %, max -.................................... ..............................- Xsdlnc number, max --------------- .....------------- --------.-------------- Color (platinum - cobalt), max --................. --.......... Suspended matter .......... ................ --................. .. 1-C0-1 -10 21.7 ................ 100 Substantially free MANdPACTUKKs .. Material - Shall bo a vinyleycJohvxene dioxide. Impurities - Material shall be clear and free from ell Injurious Imparities as metallic ar n;r>- motaUlo particles or other contamination. Approval - After a material has been approved unner this specification, no change In mmufa.ir:-.g or refining methods shall be made without prior approval of the purchaser. nepsmse mcthod3: ' ,- .. .. Sampling .................................................. -..............................-...................... ...................................... ......... ................ ASTMoa;i Specific gravity -.............----............. .........................................-.................. ................................................ ASTI*. DiaiO Viscosity -............... -.....................-............-........................................................................................-........... ..-- /S':\v Du;, ' Color ---.......................... ..................--.............................--'.................................................................................... AS?y !>d: ;> Acid number: Introduce 60 gm of the sample weighed to the nearest 0.1 gm Into a 260-al Krlemcver flask containing 50 ml of methanol which has been previously neutralized to a faint pink erJ point using 1.0 percent phenolphthaleln indicator In methanol. Swirl to effuoc solution. Add a few additional drops of the indicator and titrate with standard 0.1 K alcoholic potassium hydroxide to the first pink end point permanent for at least is seconds. Calculation: - f l i acidity, % by weight as acetic acid pa samplo A ml Of N Norr-al XOH retired Tocal epoxide: . CAUTION: This procedure Invelvos the use of perchloric acid, pfi MOT use this reajent unless ~ aware of Its potential hazards. 1, Hydrogen bromido solution, O.S N in acetic acid: By means of a graduate carefully add 67 ml of cp bromine to two liters of glacial acetle acid. Add reagent grade phenol in 10-gm Increments until the solutl. becomes light straw in color, (approximately 100 gm) and add 10 gr: In excess. Mix the ac-ltulon aftor each addition of the phenol and allow the final solution to stand overnight before ueing. 0.2 N sodium acetate in acetic acid: Dissolve 16 g,n of anhydrous sodium acetate in sufficient glcclul acetic acid to make ens liter of solution. Standardise this solution against standard 0.2 N perchloric acid using 1.0 pureent crystal violet Indicator In arctic acid. Procedure: * Prepare a sufficient number of clean dry 250-ml Iodine flasks to perforu all sample r.nu blar* determinations In duplicate. Carefully plpet 2S ml of the hydrogin trcmlde solution into each of tho flasks, using the same ptpec Tor eerh transfer. Reserve two or the flasks for the blank dete-minatlon. Into each of the other flasks Introduce O.i to 0.4 gn of the sample wel ;hed to the rcarewt 0.1 mg by means Of a suitable welgnlny, plpet a."d swlcl to effect solution. Stopper the flasks usln< 5 ml of glacial acetic cld as the llr.uld seal and allow the sxr.pl'j*.* to stand with the blar.Ra at room temperature for lb mlnuVia. A-/, i ir p. f. S. CC- GENERAL EUCYRIC VlfA'i^VCli.'riV.-.KN.-: iMuXJbK DblW 1 i Pag* <1 RKpCKEK Nl-rntODS: (Continued) Total epoxide: (Continued) Procedure: (Continued) Carefully remove the stopper from each flask and wash down the stopper and inside walls of tte flasks with 25 ml of glacial acetic acid. To each flask add 5 or 6 drops of crystal violet indicator and titrate with standard 0.2 N odium acetate in acetic acid to the first blileh-greon end point. Calculation: (B - /ON x 1.6 gm sample total epoxide, % by weight, as oxlrnne oxygen A ml of N noranl sodium acetate acquired for the sample B avenge ml of K noc.nsl sodium ocitote required for the biar.k Iodine number: VIJs oolutlon: Dissolve 13 pj of resubllmi.-d iodine In 1000 ml of glacial acetic add. Gentle hoot say be necessary to effect solution. Allow to cool and remove ny.proxlmately 200 ml oT th*j tolf.lor.. Pans dry chlorine gas Into the rerunof the Lodlr.e solution until the original titration la not quite doubled. Perform the titration as directed be law. A characteristic color ` change taken place in the Wljs solution whrn the d vjlrod .mount of chlorine has t Thin may be used as sn aid in Judging the end point. Add . small excoss of ehlori.-.e and bring Vowk to the doolreJ titration by addition of sc.ro of the original Inline solution. Into respective 500-ml Eriermoyer flanks plp-Jt 2S ml cf the solution before aid: Men of the chlorine and ?f> mi of the solution after /u! lit lor. of chlorine. Add ?0 ml of 15 percent potassium Iodide solution in dint! 11 -t water and 100 ml cf distilled water to each f.aa-'. Titrate the contents of each flask with standard 0.1 N thiosulfate until the yellow color almost disappear:!. Add two tO of 1.0 percent starch indicator solution and continued tit rating until the blue color disappears. Procedure: Heal all samples wi.lch are eull.l or 1 id u.- 'co..tr,:turbidity under ar. lnfr.trcl hnt : vip . or In A hot water bath until completely hoisaj^neous, and mix well. Prepare a sufficient number of clean dry SOO-irJ glass-stoppered Srlerjr.eysr flacks to perform 21 blank And sample detomlnncJons in duplicate. 1 To each flask add 20 ml of rengent grade carbon tetrachloride by means of a graduate. Roeorve two of the flasks for the blank determination. Into each of tho other flasks Introduce li to 16 gm of the sample weighed to the nearest C.01 0a and swirl to effect complete solution. ' . Into each flask plpet 25 ml of the Wljs solution using pressure rather than suction to fill the plpet and again swirl to effect solution. Allow the flasks to ctnnd in the dork fur 30 mlnut.es at room temperature. To each flask add 20 ml of the IS percent potassium iodide solution, and ICO ml of ClJl'.];*: water. Titrate Invwdlately with standard 0.1 H sodium thiosulfate until the yellow color a!.*i>s: dis appears. Add two ml of the starch indicator* solution and continue titrating to :l.e i'sn; poarance of the blue color. If the sample Mtratton Is less than SO percent of the eljn*. repoat the dctenr.Jnatlon using a omallor slr-s sample. Calculation: L_o-s-AblaiLm_p*_leAafia. . total unsa'uratlon s:* iodine number A ml of N nomal Ha^SjO^r-rquired for the Simple B average r.l of N norr..l N.igSgO^requlred for the blank 1 Suspended mat ter: Invert a bottle of the sample and examine by transmitted light. HONS 0 3 4 8 4 2 SAMPLE AP1R0VA1,: The eupplJor shall forward to tho laboratory of tho purchase; St the point of delivery a one pint sample taken from each manufacturing batch of `.his material scheduled for shipment to the* General Electric Company or others deiitgna-od f-y the G*mer.ni Electric Company. This ao.m;l<: hall be designated as "Advance Sample, Lot No. , G-S Material . Approval of this advar.ae sample by the General S'. *ctrlc Company does r.oc constitute approval Of the shipment. General Electric Company appr-v tl will be saved upon the material as received it Its destination ai*.d will be based not only on -MMorr.ity to the propurtlcs specified bero'.a bit on eor.fajenity to all performance ehsraotu.'isti-'a decnou essential by the pur:h*wp. Tu':< 'ur.v ,*,0. Si.-c 3) SEliEiUl ELECTRIC Si U.WiwAVIU.V vjnYLcyc:..^.'.oroxis:-: D5D71-, p*c CrirPIFlCATK OF TEST: vendor thnll submit promptly to the laboratory of the purchaser a'- the point of delivery a eereifleate of teat In duplicate 50 that it will be received on or before receipt of shipment. The certificate of test shall include the foilowirg information: Specific Gravity 20/30 C Acid number, %, max Color, max . Epoxide content, win, # Iodine number, max Viscosity at 25 C, centlatokc, max Suspended mat ter This oerilflcalc shall also contain the fl-K designation, purchase order number, tank car number, lot number, and date shipped, ao that the certificate may be identified with tfw shipment. PACKING AM* MARKING: Materia! ah,ill bo shipped lr. tank cert or drums, a* specified on the purchase orler. All eontblhors shall conform to ICC ro,;'*)at lo.ia and shall b*> of a typo having no ddieitrlous effect on the material. Extreme care shall ha ex*.roIf.ed by the supplier In tho rejection and WOO of cw.nnr. carrier containers or container: provided by the supplier. Such r.onu.ln*.rj an ill be carefully inspected by lho supplier to aJasrc that they are free of Injurious foreign matter. Adequate precautions si all be taker, by the s.ip; ller to assure retention of the prvp'iHIcs ard ChM'AtilcrlAtJ cs Of t.'H material, ar. specified herein, upon arrival of the ah: pw.-'ftt at point of dt 1 Ivory. Each conmlncr ahull be legibly marked with th. purchase orier number, the mahufauturer'a r.a:::e, the 0-F. dnrlf,nation and the lot number. SONS 034843 T GENERALELECTRIC CAPACITOR DEPARTMENT SPECIFICATION TOR FULLER'S EARTH t; q.( tutorial 030102 idenliflea fuller'* forth ae foilovos D50PC12-S1 ---**-*<>. 186i D50PCL2A - On/300 mah 09010211 60/90 m*h FNPWgtTl3 i,MXtirnn.......... IiniUon Loae, mx1mw> lgOFC12A lO.o 1C12R 3" 10.0 erenn Aiwiyole, 1> MtoM on U.8- No. 50 noabeeretn, oaxlMa-- Retained on U-9. No. Co moIi ocreen, wRlnu.' Retained on V.N. Virouth U.8. No. No. 100 00 Mh ocreen, MKlmw- mH umn, naxlm*----- Retained U>l. No. 329 Mih icrtM, MniMM- -3.0 -20.0 <0.0 J.O JO.O NMflffACWM IXtorUl ihiU to l\kllr'i brth of Uw attapulcuo eli| type, fro* frw dirt ond other eontanlnation. mma wnww NliNl............... IbUture eontcnt--Igaltlon Ion.......... , Renee# enel/ele (1 > ......................... -A3TK 0392 A...................... dRN IflW, Method A ---------------- 4I1H 090 .........................-A93M **, Method A (1) 90 P* COOpl#. boohed. CTNnrXCATl Of 8T * "luhaa requested, the aenufioturer ahnU eubnlt prooptty to the purchaser ot tho point of delivery 0 certificate 't J of toot In triplicate thawing tint the aoterUl oonfom* to thU apaaiflcation. eortiflonte NmU be iMniul to the section, unit miA person apooiflod on tho purchaaa order, and aha11 contain the 0-* and tho purehato order nunbar 00 that tho oortlfioato can ho identified with tho shipment. rActony wm. approval , approval of notarial to tble opeolfleatton vlll ho booed on factory trial. Once approval, no change obeli ho node in the oonpooltlen or proportfoo of-tbo approved notorial without tho prior knowledge *od cooooot Of tho purchaser. . PACKIND AND HARKIN8 ` tutorial aholl ho ahlppod in,9-ply paper bag# to noot XCC and earrlera* regulations. * tech hoc aball ho legibly mrkod with tho purehaao order winter, the nanufboturor'a bom, tho lot or tat:b nmtor, the weight, and tho 0-C doolcnatlon. WN20K HONS 034044 V HONS 034845 PROCSSS INSTRUCTIONS HONS 034846 3CAW.C a jmSSK-SKT 3!i.i|.!X:; W .LS, K YO: A?:n.'. 21, 1900 COPliS: A Xatchran Go tonroo BUUhCY: Character!sties of Pyrar.ol With and Without KP-20S BtahiiistJ _ Mr. K.S. Pun-hmi, Mir. AC,v. Process development osmif At your request, soar tests vere r.ade to determine the ehan-ea is test characicrisiico of ll>99 Pyranol after eddies SP-806 stabiliser. Tour interest in this subject eortas froa year proposed method of fiilirc uardfold treated units. It is possible, vith available types of r.rcerin;; c?uipasnt, to add the epoxide (accurately measured to the desired con centration) to the Pyranol i:::-1:r'r.tc-ly before uea. this vovld eiirirr.tc the rofJrsinj step, provided that t'..; characteristics of the lrprc;;aant vara sati ofactory. '" Po evaluate this proposal, I vorhod with Gerry Karroo of the >3nterii.3 ar.d Processes laboratory. 3?or the experiment, several tests were lire 0:1 .lil99 Pyrc-ci vnirj; the test cells he norrr.ljy uses on Pyr.ar.ol n, When the resistivity showed that the cells were conditioned, control tests were rate for power factor ar.d resistivity. Weighed amounts of frp-206 taken fror. the factory storase t'-r.l were then added ar.d the measurc-enis were repeated. Here arc the results of two tests: ea:p:p; Control - ho hP-205 P? ICO eye .2? Resist, xli a 000 ' Kepeat - 0.3>p 2P-806 .10 >3* 2U0 Celle Wire then rinsed three tir.es vith good 1499 Pyr&nol and then filled vith Pyranol. Control - Ko KP-206 .077 .73 3 tao hepeat - 0.35p r?-806 .105 >3* 200 the above results were shown to Art Katchran. he reuerfoerei scr.e recent tests that they had race in the sere tanner ar.i had s>on Shea show no their results: HONS 034847 1 - Control - Ko 3P-206 loo eye i.'f i sy, :'-'S Resist. :<:o9 JJ cOO " 8 - Repeat - 0.2# S?-206 >&,l ' Si 6 * - Mils RP-206 was from an old botch of Lab. distilled epoxide. 3 - Contiiil 2.4 6 000 4 - Repeat - 0.35p S?-20-t >6 64 * - As receive: '.:?-2C6 Above or.rpleb 2 and 4 were heated in an oven for (4) hours at 100C. to see if nr.y charge occurred. 2 - Vith distilled >:?-20t s.y 4o0 4 - V:th ac received !i?-20o >6 123 Addition!'.! heating over a weekend at 100c . produced no essentia! changes frer. above results. In sor.-.e previous work for Art Xatohmen, a 'batch of Pyrar.ol II war. prepared in the Pilot Shop refining system Sor.e interesting results or. this cww.riner.t were rsparted in ay Report K? 65-101 dated October 12, 1?S?. At that ti:.o, ve refined jeep Pyrar.ol to a resistivity of 4600 X lo" with a 100 cyele pawar factor of 0.4Yi>. Freshly distilled epoxide supplied by Ka.tahr.an was then added directly to the batch without going through the filter bed. 'he resistivity dropped to 2360 X 10? and the power factor was 0.46^. A couple hours of refining through the filter press actually dropped the resistivity to I960 X lo" ar.d the power factor rose to 0.54'!. Ac you can see, there are sons snide variables in the data. In the case of the pilot Shop material, there was a negligible drop in resistivity but tho other two ewperiu.ir.ts show decreases of r.ore than a decade in resistivity. You will probably vent to explore this further with the la>. people before deciding to proceed on a watering type of fill. O.A. Sr.ith . Adv. M'g. Process Develoor.-.er.t 10 dg. 10-2 bje MONS 03484b CAPACI TOR DEPARTBENT .JPr?.'.??. miiioiiii trtJrtT TOT ON MINERAL OIL, P7RANOL ADD CASKS OIL tMlf IMTHUCTIM It l Aril* ....WAtS) mm. w.....?3q:Pc(3.. .^ PARE mimu sv . ISSUED SV .. ~ c. After Treats Hudson rails - The treat operator will secure In bottles supplied by the freat Control Laboratory one pint samples of Pyranol from each after treat liner and label as to treat tank, liner treat number, and P.I. of the treat process* These samples vlll be held until the subject treat or liner unit samples bavs been fully tested and passed. >rt Edvard - The Laboratory technician will secure one pint sample dally from ith the automatic treat and the vapor treat for testing that day. Thermal chemical chlorides, ppm, max. (Note #3) Hudson Palls 7?T Fort Edvard 75 #1. For any tank with a resistivity helov thle control Unit, the following parties must be notified before release. .. ' ' 1 - Product Engineering 2 - Quality Control #2. Run at Fort Edward only. Higher values indicative of treat difficulties. Material will be rejected If higher than .1$. #3. In the event that a thermal chemical chloride result la over specifica tion an additional sample will ha run and the final answer will be an average of the sample* tested. - 2. A13B2A - 1*136 Pyranol a. Mixer Refractive Index 25C B-chlor, $, Methfed Hudson Falls (Check once a week or nore frequently on request) ASTM 0901- 1.6l85-1.6ei5 56 .9 - 1.1 Fort Edvard (check each nlxer) 1.6185-1.6215 .9 - 1.1 b. Refined storage - check eeh tank fori Teat Resistivity, oha-ca, nln. 100C Mater content, ppa, max. Refractive index 25C B-chlor, % Allphetlct, $, max. (by Central Lab) Dovtbern (#2) Method CLIN 37 G.B. B4CA5 ASTM 0901-56 CLTM 1A Hudson Falls 600x10 20 I.6I85-I.6215 .9 - 1.1 .02 Fort Edward 600x10 20 1.6185-1.6215 .9 - 1.1 .02 NDA GENERAL ELECTRICcar sis k it/si T-P, 8 cats <r issue ..AwwrX.R3i.A9KI....... utEMutt itavt or . .August. .23,. I960 HONS 034849 ` SWJffr : CAPACITOR DEPARTRENT Instruction V?'ST OS MINERAL OIL, PYHAJJOL AHI) CA3TC-R COIL INSTR. NO. FAOC WRITTEN RV ...fcj ISSUED DY .. P3C-PC23 THI INSTRUCTION IS VOID AFTER Process P3C-PC23 covers all in-process tests made by the Quality Control Materials lYocess Laboratory Unit on A13AS (5314 mineral oil), A13B1C (1499) and A13B2A (1436) Pyronols and A50PC7 castor oil, used In Building #1, Hudson Falls Plant and in the Fort Edvard plant. GENERAL Sstinpies ore taken and tests run by the Q.C. M&P factory control Laboratory representative unless otherwise indicated. Test methods are referenced in the material.specification except these listed below: Refined treating materials must be tested and passed as indicated before they may be used for impregnating capacitors. Out of control tanks, trends, or deviations shall be brought to the attention of the Treat Unit Manager and the responsible Quality Control personnel at once. iwicmuRg X. A13B1C - 1499 Pyranol a. Refined storage - check each tank for: Test Method 0*55-5- M*#U* \*FT01 Co***1 tb* b.lo--'1p0\*<* o***i1to' ms*** Control Limit Hudson Falls Ft. Edvard Resistivity, ohm-cm, min. 100C Water content, ppm, max. Rofraotive index 25C Fluorcocence B-ohlor Thermal chemical chlorides, ppm, max. (Note 13) CLTM 37 O.E. E4c45 ASTM 0901-56 CLTM SO 4oooxlo9(#l) 15 1.6240-1.6260 none NDA .75 2400xl09(#l) SO 1.6240-1.6260 none NDA .75 b. Return storage - check each day that material is taken from at Hudson Falls and four times weekly at Fort Edvard for: Test Method Hudson Falls Fort Edvard Water content, ppm, max. Refractive index 25C B-chlor Fluorescence Allphatlcs, %, max.(by Central Lab) Dovtherm, max. Thermal chomicol chlorides, ppm, max. (note #3) G.E. E4C45 ASTM D-901-56 CLTM 14 CLTM SO 60 1.6240-1.6260 NDA none none .75 60 1.6240-1.6260 NDA none none -*5 m .75 HONS 034850 GENERAL ELECTRICMR KEY I E/SI T-P,S MTE OF ISSUE ,JMUSry .?9.,. .J9?3..v........ SUPERSEDES ISSUE OT . .August. 2j, 19&)...... I* I' . CAPACITOR DCPARTfEKT I f?91 r of 11 on o: .-.j ...,v i. o.a, I'WtfOh ; :* casv ' ?:).. *ti* iwnwi'iD^ is vomj Ar?r.* __ Ir.lCrM INST* w.ec WHITTEN CV . ISSU'D BY .. y- P3P-fP`3... . ,JL . , I.Sohhyoa'"' c After Treat Refractive index C ?5C METHOD HUDSON FALLS ASTM D90X-56 I.6165-I.62I5 FORT EDWARD i"."6185-1.6215 #4. Manufacturing must obtain sample from each liner for test at Hudson Falls. Manufacturing must obtain sample from each available liner otherwise raw storage is checked. 3* AA3A2 - 3314 Mineral 031 * . Keflricd storage (//5) - check each tank for: Resintivity, ohm-cm, min. e 100C Water content, ppm, max. Refractive index 2|>C CLTM 37 O.E. E4D2 G.E. E4C45 ASTM D901-56 SOxlO12 20 1.4890-1.4922 50xio12 20 1.4890-1.4922 96 hour apino at 100C Resistivity, ohm-cm, min. Acidity, ms.K0H/(jm, max. Color, ASTM, max. CLTM 37 CLTM 36 ASTM D974 ASTM D901 7X1012 .014 1-5 7X1012 .014 1.5 jfa. Hudson Falls - Tanks containing all new material are tested for resistivity, water content and HI before release; 96 hour aging is run, but the tank not held for the result. If the tank contains reprocessed material, it is held for all tests. Fort Edward - Tanks are tested for resistivity, water content and RI before release; 96 hour aging 1b run, but the tank is not held for the result. 4. PSOrCY - Cantor Oil a* Since this material will not be refined at Hudson Falls, all approved material will be used directly from the shipping drums. For this reason only after treat samples are required. b. After Treat (#6) - check each liner for: Acid number, mg.KOH/gw, max. Resistivity, obm-cm, 100C, min. Water content, ppm, max. AOCS C*5*40 G.E. E4D2 G.E. E4c45 not available not applicable #6. . Hudson Falls - The treat operator will secure one gallon sample from each after treat liner and mark it with the treat number and the date. The sample will then be given to the Treat Laboratory operator for transmittal to the Central Q.C. M&P Laboratory, for testing and grading. The after treat castor oil will be pumped back into the original shipping drums and tagged with the treat number and the date* HONS 034851 CAP M REV l!/l T-P,S GENERAL ELECTRIC date or ISSUE SUPERSEDES ISSUE C? . , ?.?>. ..................... 2$,. 19V>P___ *9 CAPACITOR DEPARTMENT PROCESS P2GD-PC4 REFINING A13B1C (1499) PYRANOL AT FORT EDWARD PLANT Page 1 of \ ___ _______________ . .___________________________________ November 21, 1960 ' , Procenn P26D-PC4 covers the procedure for refining A13B1C (1499) Pyranol at the Fort Edward plant* REFERENCES* P3C-PC23 Refined Pyranol lnh'checks P26D-PC1 Refining of Pyranols 1436, 1476 and 1499 / * GENERAL; This Process Instruction covers only refining A13B1C (1499) Pyranol at the Fort Edward plant. * ' Refer to P26D-PC1 for refining other Pyranole at Fort Edward, and all Fyranola ' at Hudson Falls. ' Only A13B1C (1499) Pyranol will be refined In this refining.system. Material from the filter blow out lines may be reused. Material from other sources ouch as drippings from sealing benches and pumps must be scrapped. EQUIPMENT* Mixing tank with agitator Raw storage tank v' Refined storago tanks (3-10,000 gal. & 2-2,000 gal.) -.* Sparkler filter* Model 33-D-17 VPVy* Sparkler filter, Model VR-12-6 the Information eontj:n'.(* h'rcvn In propriety 10 (too Intcretl ! the Ccmir.'.l ElctUic Cempeny- M* MATERIAL: 4itclotor to unauthorised pcrsoi.ribou'tlbeuado. A13B1C (1499) Pyranol Fullor o Earth, 60/90 mosh dired 4 hours at 400 C or undrled. PROCEDURE: Charging the Sparkler Presses Prior to charging, the presses must be cleaned according to instructions. Fill tho mix tank Approximately 2/3 full and start agitation Add 150 pounds of 60/90 mesh Fuiller*s Earth. Use undrled earth if the primary objcctivo is to remove chlorides* Use dried earth if the primary objective Is to remove water. Agitate for approximately 1/2 hour, or until all the earth is well suspended. Circulate the mixture through tho Sparkler filter to be charged until all of tho earth is removed from the mix tank. When the charging is complete, transfer the clear Pyranol to a raw or used storage tank. Refining A13B1C (1499) Pyranol Throughout tho storage, refining and aging cycles, A13R1C (1499) will be held at a minimum of 50C. When filling, recirculating or aging a refined storage tank, a vacuum of 25M of mercury or higher will be maintained. ' Refined storage tanks will be filled from a raw or used storage tank. During this filling cycle, the Pyranol must pass first through the large Sparkler press (Model 33-D-17), and then through the small Sparkler press (Model VR-12-6). To refine a tank of Pyranol, it must be continuously circulated through a Sparkler press (model 33-D-17) for 12 hours. At tho end of thin tima, tho Proccnn Control Laboratory will take a sample for testing. If ihu Process Control Laboratory has approved tho nample, tho Pyranol may ho used to Impregnate capacitors after it has completed a 6 hour vacuum aging cycle. A log sheet Indicating date, time filled, time the aging cycle will be complete and ______ _ Procenti Control Laboratory approval will be obtained for each tank. *T-S CEWGI3AI.?i'.)nLOeVl3ie su.r.c.i.is.cors isr.uE of HONS 034852 CAPACITOR DEPARTMENT PROCESS r.ir.-rc? ACCEPTANCE TESTING OF PYRANOLS - A13D1 (147G) & A13IUC (1499) Pago 1 of 2 August 30,1* l r' Process P3B-PC2 covers the procedure for acceptance testing of tank car shipments of Pyranols A13B1 (1476) and A13BIC (1499) before handling. OTHER PROCESSES REFERENCED: P26D-PC1 - Refining of Pyranols P2A-FC6 - Handling tank car shipments of Pyranol GENERAL:' ^CIAS3_IV, NOTICE Tho Infomuttion cc;l:.!v-3 h?:r*n h pfsp.-lstsry |e (ha cl Kw C-r..?.T.l '?.*..<!:; Cmpary. JJ SUolatwre I 1r.ujfr.1nMi fstsrni^Ssul 1 L-c maio . Receiving: A sample for the acceptance tests vill bo taken from each incoming tank car by a properly Instructed technician of the M & PT Unit BEFORE UNLOADING. When the tank car is received. It Is spotted In the proper unloading areas at Hudson Falls and Fort Edward, equipped with steam, pumping and storage facilities. As stated in Process P2A-PC6, Pyranols must be heated to proper temperatures for unloading. . Handling: For sampling purposes, it is only necessary to heat Pyranols sufficiently for the M & PT representative to obtain a sample for test purposes. Testa must be performed and Pyranols approved by the M 6 PT Unit BEFORE THE MATERIAL IS UNLOADED. Procedure: The M & PT Unit ordinarily performs the tests listed in GE Specification A13B1 for GE Materials A13B1 and A13B1C except the tests for viscosity, fire point and sulfate. Those exceptions are performed only when requested by either the Product Engineer Manager or the Laboratory Manager. Tho following tests are ordinarllv performed on all incoming tank cars of A13B1C _^y.T Pyranol (1499) : Condition Water Refractive Index Dielectric Constant * .. Acid Number Resistivity Free Chlorides Dechlorination Corrosion Stability Infrared Acid No. . Free Chlorides Color Condition This test is performed but not Hated In the material specification. Tho following tests are ordinarily performed on Incoming tank cars of A13B1C Pyranol according to the multi-level continuous sampling plan H106: - Specific Gravity Pour Point HONS 034853 Color Flash Point Distillation Range Dielectric Strength The tests conducted on Incoming A13B1 (1476) tank cars are the same performed on A13B1C (1499) excepting flash point. Also, a 507. distillation determination is made. In the event that either material (A13B1 or A13B1C) fails to meet the specification Halts of the required tests, a second sample must be taken from tho tank car and all tosts which the first sample failed must be repeated. In addition, the advance sample kJ taken by the vendor when the tank car is loaded must be evaluated. In tho event that the only acceptance test failure Is the resistivity test, the material may still So accepted providing that the resistivity is more than 507. of specification limit. Experience has shown that low resistivity Pyranols can be raised to our process limit of resistivity by our Fuller's Earth filtering process (P26D-TC1). /4V' Af V* i'c gi:r]nnai,ninenisc supersedes issue of June 9,1953 I t t,jI.I.c2 . CAPACITOR DEPARTMENT PROCESS r*r.e 2 of 2 ACCEPTANCE TESTINC OF PYRAN01.S - A13B1 (1476) 6 AL3L1C (1499) Augu'st *0, 1959^ __ ______________________________ ______________________________________ In the event the material fails to meet the specifications, the results must be reported immediately by the. M FT Unit to the responsible Product Engineering , Manap.cr and the Laboratory Manager, or their respective Supervisors whom they have designated responsible. Product Engineering Managers will secure advice from the Laboratory Manager or his designated alternate, and will advise the Supervisor of the M & PT Unit to reject oi accept the tank car in question. This Engineering decision will be recorded on the test report Issued by the M & PT Unit. In addition, the M & PT Unit will issue n IR showing that the material falls to meet the required specification. The Engineering disposition will be recorded on the IR, In the event that a tank car meets all requirements, but in the opinion of the Supervisor of tho M & TT Unit exhibits unusual characteristics in any respect, the lot ' mist he held lip and referred to tho responsible Engineering Manager for disposition by the procedure described In this process for cases of failure to meet specification rcqulrementn Tank cars which meet all specification requirements and which exhibit no unusual characteristics, will be released by the M & TT Supervisor. The results of these tests will be published quarterly advising the responsible Product Engineering Manager, purchasing Agent, Manngcr of Quality Control and other Interested persons by means of a copy of the report. In the event that anything is found unusual or the material fails to meet any of the specifications, the pertinent information will be published in an individual test report issued by the M & PT Unit. D HONS 034854 J I-C GENERA f. ELGGVniC SUPERSETS ISSUE OF June 9 1958 ... _ CAPACITOn DEPARTKCHT JTOCKSS.. instruction runjcci .. RKF1HIT1Q OF PYRANOL II AT FORT EDWARD PLAMT iNttn. no , reGo-rcj......... r*oi . A.. VBITTCH or ,C*A:'Smith tUJO BY . E;S Dunham THIS INSTRUCTION 1$ VOID AFTER December 1967 h inlvimoHun mlaiiuJ. licteln is p.p!etn/ lu the InteieU el the Cornual L'ltcltic Company. V. proccso rsGn-PC5, 1" on outline of the procedure,liYir'lll*c,fT5'M1'iy11)'(^APC\0.A',rVy,Ate2",J''.. which hen been clablllr.cd with D50P132-SI matcrlnl. RKFURKUCKS; P30-l'CS3 * P3C-l'C|iO P2>A-rci F2A-PC10 Refined pyranol Lab Checks In-Process Testing Pyranol II Operation of Pyranol Resistivity Bridges Handling of Epoxide OKNKRAt.l Thin Process Instruction describes the refining of stabilised 1U99 Pyranol (Pyranol !!) Only Pyranol II will be refined In this system. HQUIPMKHT; . . .' v, Mixing tank and agitator Row and refined storage tanks Sparkler Filter, Model 330-16 Sparkler Filter, Model VR-12-3 Epoxide Storage Tank Transfer Pump Proportioning Pump MATOHIA1,: ' , . ' .' AX3B1C (1^99) D50P132-SI Unox 206 Epoxide (see Material Approval) Fuller's Forth, 60/90 mesh dried 4 hours at 400C. . ' PROCEDURE: Charging the Sparkler Presses Prior to charging, the preaseo must be cleaned of the spent fuller *8 earth fren previouM runs. The mix tank in filled with 70 - 6oC Pyranol, and the agitator 1* started. A max5mitn charge of 75 pounds of 60/90 mesh fuller's earth Is added to the mix tank. The slurry is agitated for one-half hour and then circulated through the . . large Sparkler filter. Pyranol Is transferred through the filter to the return storage tank. Saturating the Fuller's Earth with Epoxide MOMS 034855 . After the press has been charged with fresh fuller's earth, it must be saturated with epoxide before it eon be used for refining. This in done by circulating Pyranol H through the earth. ' The epoxide and aliphatic content of the Pyranol IT must be known before attempting to saturate nev* fuller's earth. thr a?.* mv iivii GfiHefiAL (j) ELECTRIC "T``SUfCROFDES 15*1/1 t7 October IV 19u3" L . CAPACITOR DEPARTMENT ' ' PROCESS... instfuction SURJCCT refining of rvnANOL n at fort edward piant THIS INSTRUCT ION IS VOID AF HR .P?.<!!'*er .*!!. INS1R. MO Pf?61>-PC5 . RAGE ____7._____ Or . . ? WRITTEN GY . .P- A. . Sfni th............ ISSUED BY .. E.S., . __Pmtfrvi1 Saturating the Fuller's Earth with Epoxide (continued) Epoxide in the amount of 10 - 12$ by veicht of the fuller's earth must be added to saturate the earth. For a 75 pound batch of earth, the quantity is 7.5 to 9*0 pounds. Addition is accomplished by using the system set up for this purpose. Circulate the return tank of known epoxide and aliphatic content through the net ^ fuller's earth and sample every hour, until three (3) successive measurements indicate no change in alophatic end epoxide content. ,, Operating the System All rcturnn for any civen day must be back in return storage before testing. The return storage tank is conntantly circulated through the circulation pump while returns arc being made. After all returns are in the storage tank, tokc a sample and test for epoxide concentration. Add necessary make-up 1**99 Pyranol from raw storage. Add amount of epoxide specified by Process Control. Circulate new blend for a minimum of two hours. Sample and test for epoxide and aliphatic concentrations specified in P3C-PC*tO, Continue circulation while tests arc being made. When teats aie okay, transfer Pyranol IX through filter press and cone to refined storage tank. After trans- . forcing, sample will be taken from refined tank. Continue circulation through filter and cono until Pyranol II la approved by Process Control by written autogram. Placo refined tank on aging for a minimum of U hours before use, SAFETY! ` Special care should be taken when handling Unox 206 stabiliser, potential hnrardo from this material are described in process Instruction P2A-PC10. Material Approval The Epoxide used in the preparation of Pyranol II must be certified in writing by the laboratory prior to use. . ) HONS 034856 CAF i/tV ?/! GRwnnAL'^ ELncmin f*ATl C* t>SUE IWVnl>i:r 3j iupcnsiucs issue or October 1, !So7 1 ' CAPACITOR DEPARTUENT $*/0,,ix%-i 7 $/, ",sT"' no' p3c'rc`, PROCESS., iniltuctlon IWJICT W-IWWKB TEST BY MATERIALS TESTING Ami ANALYSIS baci 1 or 2,, by . C. A. Smith % UNIT ON PYUANOL II - HUDSON FALLS A FORT FDNARD usuen by ...... E. S. XKnbfs/^ THIS IN5YOUCT IOM IS VOIOAfrCB ...........TUOC. .1,. .1?69. Process I'3C-rcl|0 covers nil ln-proceoo. tecta mode by the Mntcrlnlo Test Inc A Annlyals Unit on A'yOl'C'OA end AJOl'COn uoed In the Fort F.dwnrd Flnnt and thhlr.cn Falls Plant. OENBRAL Samples are taken and tosts run by the Materials Testing k Analysis Unit represen tative unless otherwise Indicated. Test Diethodfl ore referenced In the material specification. ' Refined treating materials must be tested and passed as Indicated before they may be used for Impregnating capacitors. Out of control tanks, trends, or deviations will be published In the monthly process control report. r_ ay.r,: tv raoaspuRKs ir Ti'cflned Storage - check each tank for: ,, n, tuoiM "****'F. V, InOlWl 1*1 O'* O. . ' ........A50PC8A CONTROL LIMIT TEST________________________________________ ________ Aliphatic Content, Infra-red Analysis, Resistivity, ohm-cra, Min. Max. See Note 1 1100 * 10^ Epoxide Content, 1> 0.35 + 0.05 Fluorescence NOV Power Factor, $, 100C, 100 cycles (Max) -- A5OPC0I) CONTROL LIMIT FOHRUTDSEODNWAFRAD1J.S See Note 1 . 2000 x 109 0.35 + .05 Nm P-75 2. Return Storage - check each day material is added to or taken from tanks for: Fluorescence Epoxide Content, S Aliphatic Content, Infra-red Analysis NBA 0.35 + 0.05 See Note 2 NBA 0.35 + 0.05 Sec Note 2 3A. After Treat - Round A Rectangular Tanks: The laboratory will supply clean pint sample bottles to Manufacturing for after treat samples. Manufacturing will secure one pint samples of Pyranol from each after treat and label as to treat tanks, run number and type of treat, and turn them over to the laboratory. After treat samples of Pyranol II will be tested on a reduced sampling plan or ea required by problems encountered for the following: Aliphatic Content, Infra-red Analysla Epoxide Content, Min. Fluorescence . See Note 2 0.25 NBA See Note 2 0.25 NBA 3B. After Treat-Automatic Treat: The Laboratory will supply cleon pint sample bottles to Manufacturing for after treat samples. Manufacturing will rncurn one pint nnmplcn of Pyranol from return storage tank at cnoplrtton of .vaeh ran of the automatic treat and label aa to treat, run number, uud typo of front and turn them over to the Laboratory. After treat camples of A5OFC0A will be tested an described In 3A. MONS 034857 CAB BIB BtV ll/ll Q-8-P GENERAL*' DATE OF IbSUE JUHC 1, 19-^ ELECT filC SUPERSEDES ISSUE Of Au^UCt >> ^9-"' . CAPACITOR OEPARTHEHT .PROCESS Instruction ' sunjrciD'-t'RC'CRSS TEST I'Y MATERIALS TBSTIMS WTO ANALYSIS UNIT OH PYHAIIOL II - HUDSON FALLS & FORT KWAHD iNsrn. no. . facc or . 2. written sy___ 9-. A.. Smith issuro by .... E*. S- lAtr.h'.-i THIS INSTRUCT I ON IS VOID AFTER . . JW At 19^9 PROCRITURKS (Contlmied) HOTf: .1: The Aliphatic content nhnll be hold tn less tlinn .0if> Above the Active ' epoxide concentrate for both Fort Edward nnd Hudson Foils, 22rB_2! ,7le aliphatic content of the after treat sample shall not exceed that of the pre-treat (Refined Storage) cample. Thcne tentn are conducted to determine the effect that treat has had on the A50PCOA (I'yrnnol II), nnd to establish the quality level for after treat material being returned to storage. HONS 034858 GEfJGnALQ ELECTflSCri rev u/r OATC Of ISSUE . i?' -9^0 ...... SUfE5E3S ISSUE Of . . . Au^UC* 5/ /- '/ CAP/ICI TOR DEPARTMENT InsTM. 110. PPA-i'ElO........... . J'roccsfi.. Instruction Mot ...1........ or a. fl"\ = SAFETY RUT.KS FOR liAIIDIJIKJ AiiD USE OF I'YRAROL II nniTTtii nv G.A. S.-iitu STAiiUJ.IZKR (EPOXIDE 20J). r ,S.S. .Dunh-oi nils mstnucrioo is voio rnn l/JJl/63 General Process Pf'A-l'ClO covers the procedure for the licindl) 112 and storage of Uno:< Epo:< Ide S06' Reference . , ''to...... Union Carbide Corporation Bulletins c"Wo;.... w,.'i V^0|| RafbtjLVilM3_; Epoxido rccinn ore primary akin irritants. Prolonged end repeated contact r y cause delayed and serious injury to the skin. In commercial practice, precautions which permit only intermittent skin contacts with these materials are effective in preventing thu occurrence of dermatitis among normal workers tree o. protracted period of tine. However, under these circumstances, a small number of Hypersensitive workers hv/o beer, observed to develop a sufficient decree of sensitivity, to bo affected adversely by there i:?J Unfortunately, such individuals cannot bo identified in any croup of workers until re current dermatitis is displayed, because of thin combination of circumstances, it is necessary for all workers to use protective measures to prevent nil contact of the rkir with the epoxides. . If thcce materials come into contact with the skin, they should be washed off imedi* atoly with soap and water or removed with a waterless skin cleaner. Clothing which is soiled or vet with these materials must be removed and not worn again until it has been thoroughly washed. Avoid inhalation of vapors or mists even though the hazard by inhalation is slight. The eyes arc moderately harmed by the undiluted fluid. Eye protection is required when handling the chemical. In ease of eye contact, the eye 6hould be given a 15 minute emergency washing in the dispensary, end o physician should see those cases in which discomfort persists. Pemono who are working directly with epoxides must wash hands before rating, drinking or omoklng. ' Eating, drinking or smoking in working area where epoxides arc used or prepared, Is not permitted. Ao shipped, Unox 20<$ is completely stable at room temperature lor periods f.xcceV.r.'j one year. Contact with acids, alkalies, water or water vapor may cause partial c^vVr.cation. Other stang catalytic agents such as Lewis acids or other acid or basic r.iltr may cause rapid polymerization and the rapid evolution of heat. Therefore, Epoxide 20 S should not be stored in any area where it may contact any of tho above materials. Any partially used drums or cans of this material should bn purged with dry nitrogen storage. i,r. yMlt 11 V;l I*I |iMir !. Im pi-ovlili-.l '.* f.r"t I t mj? ntrn t.i tit \ n I 1 : i < <! ilMl'.tO !., riA I(. wli.-n tl, >. >.:y poiv. nln>-; of mI'aMI i z. y L . JJ ____ _____________________ __ _____ _______ caV'mjT stv" fa/ii Mn * - *> GEr3ERALQoi.ECrniG OATf cr ISSUS . . J/.:.. swtnsmjfs issts cr fort ____ Vo-V;' HONS 034059 ISSTSi/CTIO.'IS MONS 034860 draft specification DIRECT TITRATION METHOD FOR THE DETKRKIUATIOH OF EPOXIDB SCAVENGER IH Ihy) PVnAf.Of, Apparatus; Karl Fiocher type burette equipped with drying tubes filled with Drierite or other suitable dessicant. Titration must be performed In a closed system to avoid lost of H2r. A rubber cap similar to Flochcr #ll-127-5 or a vented, one-hole rubber stopper can be used to attach the burette tip to the titration flask. 125 ml crlcnmcycr fli..:ko Magnetic stirrer Teflon-coated magnet bnr Reagents > Glacial acetic odd, reagent grade Chlorobenzene, Eaotran or Fischer Crystal violet indicator solution, O.lfi In glacial acetic acid Hydrogen Bromide, anhydrous, Mathccon Co;,-pony Prepare 0.1 N KBr in glacial acetic acid by bubbling HBr slowly through glacial acetic acid for a weight galnof 0.1 g of KBr/llter of glacial acetic acid* Sodiun Carbonate, anhydrous reagent grade. Dry at 120*C for at least three hours before using - store in dcoslcator. Si.oc;d.u <!!;.;> liuu: Standardise nBr solution against 0.1 g of sodium carbonato dissolved in 10 nl of glacial acetic acid. Use five drops of Indicator solution and titrate to a bluish green end point while stirring rapidly with magnetic otlrrer. Titration should be performed in a closed system. Procedure: Weigh approximately 35 CP* of Pyranol Into 125 ml flask. Add five drops of crystal violet Indicator solution, 10 ml Chlorobenzene, and stirring bar. Titrate with rapid stirring in a closed system to blue green end point. Calculation: * V.ltf** of aetlv. cpxoidc . (1000) (Wt. of Pyranol Sample) <* ***>) Equivalent wt. of Epoxide 206 70 Equivalent vt. of .Epoxide 269 ** 02 DDr solution require frequent standardisation dally for most accurate results. Exclusion of light ray retard loss of. strength. Drying tuboo filled with Indicating DirerIts or other suitable absorbent should be used on reservoir oad burette to exclude atmospheric moisture. . This method lo also applicable for the dteralnntlcn of epoxide purity of Incoming Epoxide 206. Ur.o approximately 0.1 of epoxide weighed accurately to the fourth declral plnco - dissolve In 10 ml of Chlorobenzene and follow method as out lined. * HONS 034661 117 '... i.*'*: i ' V'.'T.VlI IJ ^ >)C:irf2flt3 Gtf IJERAL ELECTRIC II hi* . CAPACITOR DEPARTMENT 46K TfT Mirmoo 4A,1J,4!t#20A,20J,20K,30A,30J.-?)i' 60 FFO0RR , _. L ALIPHATIC CONTENT IN PYRANOL II (A50PC8) E50!'C?;}-rU J. fCQfK: Thl# procedure determine* the quenttty, In $ by weight, of "Aliphatic** In Pyranol XI. The "Aliphatic*" are the expreeaton of the Infrared Abaorptlen In the 3-* micron <39?$ ea*1) region of the apeetra and are related to the CH(>.h stretching frequency. it. The following method* end procedure* have been cotebltahcd with the Ferktn-Clmcr Model 91 lnfrored peetrophotomotnr which ha* been modified by the addition of Optleol Scale Exfonelon. The Infrared Spectrophotometer require* Optleol Scale expansion, f. Sample eelle to notch the optlee. The cull* are to be a notched pair of 0.1C0 m spaclrg. 3. Cyrlrcea and mce**ry equtpncnt to load and clean told eells. 111. CAMPHATIOtt: feantttatlv* Infrared Spectra Analyela requires thet preclae standard* be established to cover the rang* of analyst*. . Tor the raw of thl* enelyele etendard* of 0.1, 0.9, 0.3, O.V and 0.5? by weight of * received Cpovide 90$ In 1*9? pyranol should be prepared. The weight* of Epoxide 20$ and 1*99 Pyranol are to be determined to the Detract o.l ng for the calculation of percentage*. IV. TEST SAKPUSt Th* tempi** ahould be token, handled and atored In compliance with standard practice* for purity preservation. V. pr&BIVWBi (Sptalfle for th* modified Model 91 Parkln-Elmar Infrared Spectrophotometer). All analyst* h*)l be of th* differential type, that 1*, versus a standard lot of lk99 Pyranol. A* Instrument Condition* J. p<tr?* - 3.0 k.O Mnwne t. Source Intensity * Maximum *)0.} amp*. 3. Optlesl Seal* Expansion - JX a. Silt Control* a)Auto blPIbtor out * e)Rc*elut!on - 10.00 ). Amplifier Control* a)Te*t elgnal* - off b)0*ln k.> - 9.0 nominal iRaaponaa h $. Scanning Control )Sp*ed 15 minute* biAutcoatle Drum Control e)Autenatle Suppression - 0 . , I. Sample Cell fetrems ear* must b* taken to Insure that th* sample and reference cell* be free fro* gll contamination. The cleaning of the reference eell, 1^99 Pyranol, ean be accomplished by a thorough washing with petroleum other but the sample call, F)r*nol IX, must first b* rinsed with 10 ml of benaan* than voahsd with th* potrolewm ether. n. ynrCTfnETAnoft or spectra; # * y drawing a b**o line aeroea the top of the absorption epnetra the depth of the absorption cum I* measured ferpendlculnr from the maximum absorption point to tho baso line. Expressing thl* in percent transmittance a direct comparison can bo made to the calibrated standard*. iranr HONS 034862 GENERAL^ ELECTRIC TEST METHOD r4C4o-sa unstable: chlorine compounds in askarels _,,________________ _______ Sv.pcrjQJcs HC-lfl-Sl C-E Test Method R4C48 governs the determination of the presence of unstable chlorine compounds In chlorinated biphenyls (aokorcls). * PIUNCIILK OK MirtlOD: This method Js based upon the hydrolysis of unstable chlorine compounds In osk&rcls by mcthunollc odium hydroxide. The resulting chloride Ion la determined potentiometrlcally by titration with liver nitrate solution In an essentially nonaqueous medium. The messured chloride Jon, reported as parts per million In the askarcl sample, 1s Indicative of the relative stability of the askarcl In dielectric system. APPARATUS: Beaker - 200 ml tall form (Berzelius type) Magnetic stirrer - Fisher Scientific Co, Catalog No. 14-513-1, or equivalent with ring stand base nd built-in rheostat. Set rheostat at full speed and operate through a varlac to adjust Its peed to prevent heating or the stirrer during the stirring operation. Magnetic stirring bar. Teflon - One piece molded construction, cylindrical In shape, one Inch lon, NRher Scientific Co, Catalog No. 9-311 -9 or equivalent. Microburet - Graduated in 0.01 ml divisions. Suitable buret may be obtained from Scientific Glass Apparatus Co, Catalog No. JM-&70. Silver electrode - Beckman Oliver billet electrode. Catalog No. 33261, preferred. Olsaa electrode - Standard glass electrode such as Beckman electrode, Catalog No. 40496. pH meter suitable for use with glass electrode - Model 03 Beckman pH meter preferred because Instru ment has expanded scale and provides greatest sensitivity to Incremental cmf changes. Somewhat less sensitive meter such as Beckman "Zeromatlc" or Leeds Northrup line operated pH meter may bo used. Water bath - Use an Individual glass water bath ISO mm in diameter, 75 mri high and containing BOO ml of water heated to 40 C t 1 C. Corning Olass Co, Catalog No. 31401 or equivalent. Flpct, 2S ml. * Luivl or pipit, graduated to deliver O.b ml. Waoh bottles foi pure acetone, methanol, and water. rkaokhtsi Methanol (chloride free) - Reflux 5 liters of methanol with 0.5 gram analytical reagent grade silver nitrate for 1/2 hour. Distill the methanol from the silver nitrate, discarding the first 100 ml to flush the apparatus. Distill 90% of the charge from the flask and discard the contents remain ing In the flask. Chloride Ion concentration should be less than 0.01 ml of O.OOSN silver nitrate por 300 ml of methanol. Sodium hydroxide reagent, o.l N, methanollc - Dissolve 4.0 grams of analytical reagent grade sodium hydroxide In one liter of chloride free methanol. Sulfuric acid, 50-50 by volume - Dilute analytical reagent grade concentrated sulfuric acid with ehlorldo-free (deionised or distilled) water. Precaution - Always pour acid into the water with eonetent stirring to prevent any dangerous build up of heat. Standard silver nitrate solution, 0.005N and 0.0025N - Prepare by diluting ampoule of concentrated aqueous silver nitrate available from Anachcmica Chemical Ltd, Champlain, Hew York. O.OOSN silver nitrate may also be prepared by dissolving 0.6495 gram of analytical reagent grade silver nitrate orystals In one liter of chlorldc-Tree water. Standardize against a pure chloride standard. A sodium chloride crystal such as used In Infrared spectrometer cell3 Is a good source of pure odium chloride. Check silver nitrate solutions at least monthly to assure a consistent reagent. Acetone (chloride free) - Distill from silver nitrate os described for methanol. Also check by potcntiomctrlc titration to assure optimum purity. Normally a chloride content of less than 0.01 ml of 0.0025N silver nitrate per 100 ml is derived by this method. Benzene - Use an analytical reagent grade which Is normally chloride free. Check by potentlometrlc titration. PKJXAUTKHJSt Exercise usual analytical precautions to prevent cross contamination from other sources of halogen In the laboratory. All glassware, apparatus, and the area In whleh thlo test is run should be analytically clean. (Continued on page 2) MONS 034863 E4C46-S2 Page 2 GENERAL ELECTRIC TEST METHOD UNSTABLE CHLORINE COMfOUHPJ Iff ASKAItELJ vflOCEDUnp.: Weigh 25 grama of askarel into a tarcd 200 ml beaker to the nearest 0.01 gram on a suitable ta:jnee. Add tho magnetic stirring bsr to the beaker containing the sample without the bar touching thr hinds. Notes When testing the more viscous ackarels, add S ml of benzene Immediately after the stirring ' bar. Hat the sample until dissolved, cool to room temperature .and then proceed as pre scribed. Add the benzene to the reagent blank al30. Add 25 ml of O.IN methanollc sodium hydroxide by means of a 25 ml pipet and cover the beaker with a watch glass. ' Place tho sample In the to t i C water bath to a depth of 1 1/4 Inches on a magnetic stirrer ard clamp securely to a firm support. Stir the sample at as fast a speed as possible, without pro nounced splashing, for one hour. (Alternatively, heat the sample to 40 C in a water bath, renove the temple and stir as above on a magnetic atlrrer. After the one hour stir, remove the sample beaker from the bath and add 0.5 ml of dilute suirurlc acid by means of a suitable pipet or buret. Add 125 ml of chloride-free acetone from a graduated cylinder. Titrate the aample with O.QOSH silver nitrate solution using the silver-glass electrode system. Normal samples of askarel require extremely small amounts of silver nitrate. Therefore, run the titration using 0.01 ml additions and allow aufricle.nt time for equilibrium to be established be fore recording the emf change. If a change of less than 1 mv per 0.01 ml addition Is observed for three or four 0.01 ml Increments, use larger additions of silver nitrate such aa 0.05 ml until such a change is observed. Then reduce the additions to 0.01 ml again to complete the titration. The endpoint normally Is defined by two 50 mv changes. A normal titration would yield the follow ing tyomai data: MV w66 0 560 6 552 6 541 11 sn ZZ 271 50 221 SO 201 20 165 16 ML dML dHV/dKL .06 0 .07 .01 0 6 .06 .01 . 6 .09 .01 11 .10 .01 20 .11 .01 50 .12 .01 SO .13 .01 20 .14 .01 16 (1) Using the Beckman OS pH meter the change Is measured In 0.2 mv units and hence the meter changes observed would be five times this value (l.e. 2S units for 5 mv). To calculate the change per 0.01 ml observed divide the mv change by the volume of silver nitrate. By plotting dmv/dml vs ml, determine the endpoint to the nearest 0.001 ml. (This gives a sensi tivity of t 0.00? ppm. To define the endpoint to i 0.01 ml, no plotting is necessary and a sensi tivity of t 0.07 ppm Is assumed.) Nun a reagent blank exactly as above omitting the askarel sample. CALCULATIONS: Subtract the reagent blank from the total volume of silver nitrate and for the sample, then: Net vol AgNOg x normality of AgUOj x 55.46 x 1000 Reactive chlorine (ppm) *. Weight" PRECISION: A rapid titration can be made to the nearest 0.1 ml using the normal potential at the equivalence point or use can be'made of an automatic tltrator for routine control procedures. The sensitivity in either ease should be within * 0.2 ppm of the value obtained by more refined techniques. REPORT: Report shall include the 0-E designation of the material tested, the manufacturer's name, the pur chase order number and the parts per million of reactive chlorine dechlorlnatcd. Hay 29, I960 MQNS <m86<. GENERAL^ELECTMC E4C44-S4 TEST METHOD CONTAMINATION OF CAPACITOR LIQUIDS '_____________________________________________________________________________________________ Supers 0-E Test Method E4C44 governs the determination of the contaminating effect of va 'lous materials on capacitor dielectric liquids. The following Test Methods sre referenced In this method; 0-E E4C41C - Chloride Content of Solutions AS714 D974 - Neutralization value by Color-Indicator Titration 0-E E4D2 - Resistivity of Electrical Insulating Liquids PRINCIPLE OP METHOD: A samplo of definite surface area or size is aged in the liquid for 9G hours at 100 0 and the re sistivity or specific conductance determined as a measure of the contaminating effect of the sample on.tho liquid. Othor testa such as acidity and free chlorides may also be *un as warran ted. Tho material under test Is visually examined for physical change. PRECAUTIONS: Containers must be tightly covered to prevent cross contamination. Utmost care must be taken, to prevont accidental contamination of the liquid, the sample under teat or the equlpnent. APPARATUS: Circulating air oven capable of maintaining a temperature or 100 * 2 c 600 ml beakers Aluminum foil RKAOKNTSl Capaeltor dielectric liquid specified for the test. SAKPLIN0: Samples shall be selected dopendlnc upon the material to be tested and unless other-rise specified shall bo the following size: Solids Material Material soluble in the test liquid Powdered or granular material Insoluble In tho test liquid Sample size *1.0 sq Inch of total surface rr-a 0.5# by weight l.OJf by weight Zn handling the specimens take great care that hands do not touch the area to be tested. PROCEDURE; Dry sample at ISO 2 C for a minimum of 16 hours. Wrap an 600 ml beaker with aluminum foil. Add 500 ml of the specified dielectric liquid and place the sample Into the boaker. (Note: The liquid must not have a water content exceeding 20 ppm and should have good electrical characteristics.) Cover the top of tho beaker with aluminum foil and age the sample for 9G hours at 100 i 2 C. Age a blank consisting of a similar beaker or the name liquid but without a sample. At the end of the aging period, examine the test liquid and the sample for any changes In physical condition and measure the resistivity In accordance with 0-E Test Method E4D2, if required, then determine the acidity of the liquid In accordance with ASTM 0974 and the chloride content in accordance with o-S E4C41C. CALCULATIONS: Calculate percent change In resistivity as follows: P. .rcen.t c.hange----1--0--0-- 1(-R----e--s---i-s- t--i-v--i-ty ofRbclaln.Uk v- ltRyeosifsbtilvaintyk---o--f---l-i-q- 3u--i--d----&----s--a--m- ple)- Calculate the specific conductance due to contamination as follows; Os " o Where Og Conductance due to contamination. On Measured conductance (reciprocal of resistance) of liquid which contained sample, & 0o - Measured conductance (reciprocal of resistance) of blank Spoeifie conductance due to contamination <1$ x K Where K * cell constant REPORT: The report shall Include the 0-E designation of the material tested, the manufactcrai's none or tradonomo, tho purohnse ordor number, tho pereent change in resistivity or the specific conductanoo and if determined, the acidity and chloride content of tho liquid. 6D60 (1) February 4. I960 HONS 034865 GENERAL ELECTRIC TEST METHOD E4D2-S2 . RESISTIVITY OF ELECTRICAL INSULATING LIQUIDS -- - Supersedes S402-SI O-B Test Method 402 governs the procedure tor measuring the resistivity of eleetrlc.il insu lating liquids, PRINCIPLE OP METHODt The volume resistance of the liquid lo determined between two concentric cylindrical electrodes and the resistivity calculated. PRECAUTIONSt Beaker, glass plate end electrodes must be thoroughly cleaned to assure correct readings. Os not use abrasive eleanera. APPARATUS! Megohm meter capable or measuring at least 100,000 megohms, such as Electronic Instruments Ltd 0,000,000 megohm meter or Oenertl Radio 544B. Capacitance bridge copable of measuring capacities in the order of 50-100 uuf. Sleetrode assembly aa shown In figure 1, as follows: Concentric cylindrical electrodes made of pollshod nickel-plated brasa, 3 1/2 inches high and vith three feet 1/4 Inch high equally spaced around the cylinder. The Inner electrode shall have an OD of approximately 2.610 Inch and the outer tleetrodo shall have an ZD of approxi mately 3,015 Inch so that there will be a gap of 0.10 lnoh between them. , Pyrex glass plate 3 1/2 inch in diameter with two concentric grooves 1/32 inch deep to position the cylindrical electrodes. Figure 1 UAOCNTSt .. Acetone solvent (0-E Material DSB23B) Aleenox, trlaodlum phosphate, or other suitable detergent. HONS 034866 DETERMINATION OP CELL CONSTANT! Clean beaker, glaee plete end electrodes se specified under Procedure. Place the fleas plate and elaatrodea in the beaker, positioning the electrodes carefully in the plate grooves so that there le e gap of 0.10 inch between them. Mini sure the air cop'*,lty of th cell using M** center electr**Ie t'aIculele tlv .'nil .u.fti/ti.i , n f..||tvni the high terminal. C.11 Ode <K) - The eell constant should be determined at least once a week, and oftancr for more precise measure_ mem a. Typical values for .Hr capacity sr 67 to 72 inf. 5060 (Continued on ?) 4D.-S5. ,rii 2 TEST Mt niOD RESISTIVITY OP ELECTRICAL IKSULAVINO LIQUIDS OCEDUnSi dim eft* beaker, glass plat* and electrodes In a trichloroethylene vapor degreaser and then vash vlth acetone solvent. Wash vitft * solution or Aieonox (approximately or other s-jiteb'.e detergent In vater until vater will wet the entire surface uniformly. Wash thoroughly with distilled water and dry at 105-125 C. Care should be taken to clean the top and bottom edges of the electrodes so well aa the lorger surfaces. Vrsp the eldes end bottom of the beaker in aluminum foil to shield It from light. Place the glass piste and electrodes In the beaker and rinse twice with fresh port tens or the liquid to be tested St 100 C. Position the electrodes In the grooves of the glass plate so that there Is a gap of 0.10 Inch between them. Pill the beaker vlth the eample and maintain at a temperature of 100 t 0.5 C m a constant tempera- tore bath or hava the sample slightly above 100 C and make resistivity measurement when the sample eoola to 100 C. , Measure the resistance or the sample at 500 volts d-c after one minute electrification using the center elee trod* as the high terminal. If the realsttvlty of the liquid la above specifications, the cell may be stored in the liquid until the next measurement without cleaning. . CALCULATION! Resistivity In ohm-em resistance in ohms x cell constant (k). t REPORT! The report should Include the 0-E designation of the material tested, the manufacturer's name, the purchase order number, the resistivity and the temperature at which It wee measured. June *e, 1955 HONS 034867 V HONS 034868 CLTM-l Date Issued: Oct. 1, 1059 ` Revised: ' Tentative O.B. Free Chloride Test Method (1) . for Boreal pyrsnol IOOg of Pyranol Is dissolved la 100 ml of acetone. The Pyronol Is weighed accurately to *01 ga In a 250 ml beaker* Tbs acetone used here is ACS grade acetone* It normally contains about 0.01 ppm or lest chloride. If the solvent blank Is high, It eay be purified by refluxing with AgHO* and distilling off the acetone. Other solvents of similar purity may be used, If sample solubility nec'essitates (l.e. Diegans, Dtaothylformaide, Olaciol Aeetlc Acid etc.). ` (2) 2 ml of 1ft HNO3 is added to the scetone-pyranol mixture and it is stirred 5 minutes to Insure proper mixing. The nitric acid used Is cade from concentrated nitric acid, diluted to give a solution of 1# strength with chloride free water (distilled or deionized). (3) The resulting mixture (from 2) Is titrated with 0.0025N AgXO^ using a sliver ' glass electrode system and a Model "0" or "03" Beckman ph meter. The glass electrode used Is a Beckman glass electrode Cat. {p*Cb$S. The silver electrode Is a Bookman nliver billet electrode Cat. 39261* Normal samples of Pyr&nol require small amounts of the AgfcO^ reagent. For this reason 0.0025?? AgKOs is used with a 1 ml automatic buret graduated to 0.01 ml. (Such a buret may be obtained from Scientific Glass Apparatus Co., Inc.) Additions of 0.01 ml of the Agltt>3 reagent are made allowing 1 minute between additions for pro per mixing before making the emf readings* If a change of less than 1 ev Is observed for an addition, larger additions are mode - for Instance 0.05 ml until a change of 1 mv/per .01 ml of solution Is observed. When a change of this site Is observed, the Increment Is Immediately decreased to 0.01 ml of AgKto again. The endpoint for this determination will ordinarily give a change of Uo-50 mv/.Ol ml addition. This change Is calculated continuously during the titration from the volume of ASWO3 added and emf changes observed. To calculate the change per 0.01 ml observed, the millivolt change observed Is divided by the volume of AgNOj added. For Instance, If a charge of 10 *sv it observed for 0.02 ml addition, the change of eaf/.Ol ml would be 5 sv/.Ol cl and the next increment should be adjusted downward to 0.01 ml to avoid missing the endpoint. The information would be recorded as follows: MV J MV ML i ML 1 mv/iSl 1000 "990 s. 80 790 760 700 745 , -_ 10 10 30 50 50 30 10 5 0.00 - - .02 .02 $ .01 .01 .05 *01 .0(3 .01 0? .01 .03 .01 0$ * . .01 0 5 10 30 50 50 30 10 5 'By plottirg d!-.V/dnl vs nl the endpoint eat. no found fid the nearest" 0.001 :*! easily. ' A little experience with this :.:-Jth<xl of chloride determination on Known solutions will prow valuable r.nd in a short time precision or .CC1 rl will be routine. HONS 034869 -a. thylcno vopor doffreocer and then v&oh vith acctono solvent. Hash with a golutlon of Alcoaox (approximately 2) or other suitable detergent In voter until voter vlll vet t*eo entlro surface uniformly. Rlnso thoroughly vith distilled voter end dry at 105*125*C. Mote 1 I Coro should bo taken to thoroughly clean all surfaces vhlch vlll coeie In contact vith the foil or Fyronol. Mote 2 : Tbo electrodes should bo conditioned as specified In E>tD2. Stack sufficient layoro of foil to make a weight of approximately 50 grams plus on outer vrsp. Holding thlo pad with forceps, discard the top aad bottom lsyors and tic a cord around tho stack to make a bundle. Heigh this bundle, cutting off ploees until a 50 * 0.1 gram weight Is obtained. . Mote 3: Extreme core should bo token to protect tho cample floa say sourco of contamination, including handling. Dry tbs specimen at 150 i 2'C for a minimum period of 16 hours. Vrop a beaker with aluminum foil to shield It from light. Add 500 ml. of A13B1C and hoot to 100 * 2*C. ` Kota b 1 Experience has shown that room temperature Fyranol will reach 100*0 In approximately an hour when placed In an oven. Mote 5 s The Fyranol used must have a minimum resistivity of 10C0 x 10^ . ohm-cm and a wator content not exceeding 20 PPM before heating. Place tho boakor containing the hot Fyranol In the oven containing the dried sample. Orasplng tho saoplo by the string, hold it over tho Pyranol sad cut the String so that only tho foil sample falls Into tho hot Fyranol. Discard the string. ' Replace the aluminum foil cover and ago the semple for $5 hours at 100 - 2*C. Ago a blank consisting of a similar bsaksr of the stuco Fyranol vlthout a sample. At the end of tho aging period, pour the Pyranol Into a hooker containing the conditional olectrodo assembly. Measure tho reslotlvlty In accordance vita EtD2. Koxt, datarmlno tbs acidity of tho Pyranol In accordance vith ASCI 97! and the chloride content In accordance with ASTM D878. CALCULATIONS i ' ' Calculate tho speolflc conductance as follows: 0B. - nO . 0. whore 0D_ measured'conductance (reciprocal resistance) of inauiatinPyranol vhlch contained tho sample. ` aeasured conductance (reciprocal resistance) of control scn:;_e. then Ox *K conductance due to contoalnntloa* specific conductance duo to cont*r.l nation. K - cell conotonta * Tho report chall Include the complete Identification of the material MONS 034670 -3- type of coll used, the specific conductance, the acidity and chlorldo conten of the Insulating Pyranol after test. 'T?/^ B.A. Larrlek 9/23/59 MONS 034871 o CLTH-5 . Date laDued: October lc59 Revised: ' Thlo tent method governs the determination of tho eontnalnatlon of GS Materiel A13B1C (1*99) Pyranol hy capacitor papor. - -- PKKiCIPLE OF METHOD: A sample Is aged In Pyranol for 96 hours at 100'C end the change In reals- tlvlty determined as a measure of the contaminating effect of the paper sample on the Pyranol- " APPARATUS: . '- 0 1 Circulating air oven capable of maintaining temperature of 150 end ICO 2*C 600 ml- beakers Megohm meter capable of measuring at leant 100,000 cegohmu, such es. Electronic Instrument, Ltd. 20,000,000 megohm meter. - Analytical hslence .' ' Forceps ' Scissors Aluminum foil Electrode assembly consisting of concentric cylindrical electrodes rr.de of pollohed nickel-plnted brass, 3-1/2 Inches high and with three feet l/i inch high cruelly space! arouud the cylinder. Tile lnnor electrode shall have cn . O.D. of approximately 2.8l inch end the outer electrode shall have am 1.3. of approximately 3-015 Inch so that thero will he a gap of 0.10 Inch between then. Fyrex glesa plate 3-l/2 Inch In dlamter with two concentric grooves 1/33 Inch deop to position the cylindrical electrodes, twino REA0E1.T3: '' ' .. ' -....-- 1*99 Pyranol, CB Material A13B1C -- Acetone solvent, GE Material D5B32B . Alconox, trl sodium phoophate or other suitable detergent Distilled water .. . ' ' - PRECAUTIONS: ' ' '' . ' Bookers, glass plate, electrodes, forceps, and scissors must be thoroughly cleared | to assure correct readlngo. Containers must be tightly covered to prevent cross ccn- tsmlnatlon. Utmost care oust bo taken to prevent accidental conteminatlcn of the Py- renal, the paper sample or the equipment. TR3T BPECIMEB: .. ' o Tho test specimen shall consist of stocked layers of paper approximate':- "! to 3 inches in width welching 25 groan. In handling the speelmeno, great care should bn taken that hands do not touch the paper to be tested. ' -' HONS 034872 CLTW-5 Page $2 PPOCSr/JPE: Clean the beaker, glaco plate, electrodes, forcepo end eclcooro In a trt- chlorethylcne vepor dccrcacer and then wash vlth acetone solvent. Wash vlth a solution of Aleonox (approximately 2$) or other suitable detergent In vater until voter vill vet the entire surface uniformly. Rinse thoroughly vlth dlotlilcd vater ar.d dry at 105-125*0. Koto I: Care should bo taken to thoroughly clean all surfaces which will come In contact vlth the poper or Pyranol. Koto 2; Tho cloctrodos should he conditioned ao specified in E4D2. Stack suff'elcnt sheets to make a weight of approximately 25 gra.-23 plus an outer wrap* Hold this pad with forceps, discard the top and bottora sheets and tie a cord, around the stack to make a bundle. Weigh thlo bundlo, cutting off piecesmtii a 25 i 0.1 gram weight Is obtained. Eoto 3? Extreme care should be taken to protect the cample from any source ~ of contamination Including handling. . Dry the specimen at 150 i 2*0 for a ninlraum period of 16 hours. Wrap a beaker with aluminum foil to shield it from light. Add 500 ml. of A1331C Pyranol and heat to 100 2*C. . Hoto hi Experience hao shown that room temperature Pyranol vill reach 1C0*C In approximately an hour when placed In an oven. Hote 5: The Pyranol used must have a minimum, resistivity of 1000 x 10^ olm-cn and vater content not exceeding 20 PPM before heating. -- Place the beaker containing the hot Pyranol In the oven containing the dried annuls. Grasping the sample by the string, hold It over the Pyranol and cut the string so that only : paper sample falls Into the hot Pyranol. Discard the string. Replace the aluminum foil cover and age the sample for 96 hours at 100 t 2*C. Age a blank consisting of a similar beaker of the some Pyranol without a sample. At the end of the aging period, place the resistivity electrodes in the beaker end measure the resistivity In accordance with EUD2. * CALCULATIONS: Calculate the percent change In resistivity as fellows: t ____ -100 (Resistivity of Blank-Reatstivity of Pyrc.nol ?*".* 1*' Pcrcu.t Choose - ---------5--------------- ofM5.k &aUtvt.y------------------------------------ HONS 03`*873 ci/m-5 Pogo 3 R5?0STi end The report shall lncludo tho the chongo In resistivity. conplcto identification of tho naterlal teoted R.A. Lnrrlek 10/25/59 HONS 034874 IHg^W'-CHBtgCAL.STAntT.IlY TEST FOR CLTX-20 Issued: 4-26-61 Revision: //I Mm?.0L X. SCOPS . This method doscribos a procedure for measuring the Thermal Stability of Fyrar.ols, by exposing tho Pyranol to boat In a scaled ampoule. The resulting chloride is doterdned potontioraetrlcally and its magnitude is a measure of the Thonral Stability of the Pyranol. II, APPARATUS A!CO BQUIPMSMT A. 32 x 200 rm fyrex tost tubo 8. Olassworkors torch, glassos and various equipment 0. Ovoft operational to 250C or ovor and controilablo to i 29C D. Balanco accurate to 0.01 gram. Torsion Balance Stylo 1L9 adoqu;.to. 2* 50 ml* glass syringe, 6" 14 gauge stainless stool noodlo and stopcock. F. Support for ampoules Wnilo in ovon. 0. Beaker 200 ml. tall form. (Borzolluo) H. Magnotic stirrer and stirring bars (1" Toflon covGrod) 7. Microburotto graduated In 0.01 ml. divisions J. Sllvor olcctrodo (billot type). Beckman //1261 adcqu.ito 1C. Olass olcctrodo. Bookman //4919V4 shielded adequato. L. pH tfeter sultablo for use with glass eloctrodo. ~ Kodol GS Bocter.in preferred because instrument has expanded scale for greatest sensitivity to Incremental EMF changes. Somewhat less sensitive meter, such as, Beckman "Zoromatlo*1 or Leeds Uorthrup lino operated pH meter may be used. H, Water bath - Use an Individual glass water bath 150 ma in diameter 75 rr\ high (crystaliaing dish) and containing 600 ml of water heated to 40C 1*C. . H. Pipette 25 ml. Fisher 08-741 adequate. O. Burette or pipette, graduated to deliver 0.5 ml. P. Wash bottles for pure Acetone, Kothanol, and wator (Polyethelene) Reattents A. Kathanol (chloride free) cay be deionised through an ion exchange column or may be distilled from AgiCO- crystals. Chlorido ion concentration should be less than 0.01 ml of 0.005 M AgKO* per 100 ml. of Methanol. B. Sodium Hydroxide, 0.1 M prepared by dissolving 4.0 grarj of analytical reagent gredo NaOH pellots in ono litor of chlorido free Methanol. C. Sulfuric Acid, 50-50 by volume - Dilute analytical reagent grade cor.- eentrated Sulfuric Add with chloride free (distilled or deionised) ' water, CAUTIOHt Always pour acid into the water with constant etlrring to prevent any dangerous bnlld-up of heat. MONS 034875 -2- Ill, 0. Standard Silver Nitrato solution. 0.005U. Prepare by diluting ampoule of concentrated aqueous silver nitrato. Standardize against a puro chloride standard. (0.1N Silvor Nitrato ar.pouloo may bo pur . chased from Anachomia, Montroal, Canada.) E. Acotono, chlorido froo. Chlorido ion concentration should be 0.01 ml of 0.005 K Silver Nitrato or loss. Acetone r.ny bo distilled from Silver Nitrate crystals if necessary. Solvents (Acotone it Methanol) should bo checked for initial free Cl" concentration to dotomd.no whothor refining la necessary. RSTSSE.'CFS .' Spec. E!C46**S1 - Unstable Chlorine Compounds in Askarels. IV. P.nDCSDC?3 - A. Ampoule preparation 1. Thoroughly clean the 32 x 200 mm toot tubo with soap or detergent and water. Rinse thoroughly with doionlzod or distilled water to make sure all traces Of cleaning agent aro removed. . 2. Plaoe tho oloaned test tube In an oven for at least; threo (3) hours at about 100C. . 3* After oven drying the test tubes are allowed to cool to comfortable handling temperature, 25-30C. 4* By standard glass blowing procedures a glass rod Is attached to the lip ef the test tube and a neck formed at approximately 11.4 cm from the base. The neck should not reduce the diameter by more than half. Seo diagram below. 5. The test tubas nay be prepared in advance, covered with aluminum foil and - stored. ' P| n *}</>**, fi&rjt firtr p[ ' /At/tt* Zsr I) (------- p :(&__________ ____ ,_M 8. Teat Proceduro 1. Into a tarred-necked teat tube weigh 50 grams 0.1 gram of Pyr&r.ol ujir.g a 50 si, ayrirgo with an QH - 14 gauge needlo for delivery. Care rust be taken not to wet tho neck of the test tube, because sealing temperatures will cause decomposition of the Pyranol. ' 2.Tho lower portion of the ampoule is covered with a wet papor towel to preset tho Pyranol from tho heat of soaling. ' 3. Soal the neck of the ampoule off with the torch by drawing. leakage. ' MQNS Chock for 034076 -3- 4* o extend the heating period beyond two (2)~minutes> "a*vi do not apply the flame into the open test tube. Extreme temperaturos trd intenso light can cause ryranol breakdown, and high chloride results. 5. The hermetically sealod ampoule is placed in an oven operating at 2/0C i 2C for two (2) hours. 4. When the two (2) hours have expired the ampoule ie removed from the oven and allowed to come to room temperature. 7. After the ampoule has reached room temperature the neck is scored with a filo and then the neck le struck sharply with a spatula to break off tho end. 6. Empty the Pyranol into a clean 200 el- Barsallua beaker. 9* Rinse the inside of the ampoule with 10 ml. of chloride free acetone, and empty this into tho beaker. 10. With a 25 ml. pipette and automatic plpetter transfer 25 ml. of rethar.oli 0.1N sodium hydroxide to tho beaker. LI. Add an additional 25 ml. of chloride free aceiono 12. Place a clean Teflon covered stirring bar into the beaker. Do not let the hands touch the stirring bar. Use forceps or other transfer agent. 13* The beaker Is now placed In a A0C 1C water bath. Sufficient water should be In the bath to cover the sample-occupied portion of the bath, 1.e., about 4 cm. in a 7 x 15 cm. crystallsing dish. 14. The bath and sample aro placed on a magnetic stirrer. 15. Stir the sample for five (5) minutes as fast as possible without causing pronounced splashing. 16. After five (5) minutes of stirring add 0.5 ml. of (50-50 by vol.) dilute sulfuric add. 17. With a graduate add ah additional 45 ail. of chloride free acetono. 18. Titrate for chlorides potentlonetrlcally using the silver-glass electrode pair as per General Electric Test Method EM2A6-S1 "Unstable Chlorino Compounds in Askarels". 19* Dotorwino tho rongont blank by performing tho titration stops above, but omitting the fyranol. i.e.. Add 35 ml. of acotor.o and 25 ml. of methanol 0.1N NaOH to a 200 ml. Dorcelius beaker, stir in a 40C bath for 5 min., acidify with 0.5 ml. of (50-50) dilute sulfuric 3cid, add an additional c of Aootono, and Titrate for CL*. . Determine the Pyrar.ol blank by weighing 50 rraas of tho original P ran.:! into a 200 ml. 3>rcelius beaker, adding 35 ml. of chloride free Aautcr.b plus 25 ml, of r.ithonalic 0.1M sodium hydroxide and retreating sterra /-12 throurh '1G. This represents in nr?*. of Cl" tho chemical cor.tr .'juticr. to t.t. w.il; .....- cj.ce a HONS 034877 20. If the Pyranol blank falls within a 2 sigma (<r") <Jovi&tton from th*> average (5:) valuo uoo tho average a tho Pyranol blank. 21. If the value falls outsldo the 2 sigma (o") limits use the actual value as the Pyranol blank. Calculate one . Gross PPM Chlorides not ml. AgKO-j x normality of x 35^60 Weight of Sanplo in grar.s \\ Weight of Sarn.plto in Grams ) " 22. Record the not^thormal chloride valuo and tho Pyranol blank in pps CV Not thermal Cl* Gross Thermal - Chomieal Cl" - (Pyranol blank + Reagent Slar.'r 23. If tho net Thermal Chloride value doviatos from tho 2 stem (e~) limit a second determination should bo mado, If tho second value falls outside the control limits (2<f) tho Pyranol will not bo reloased.without tho supervisors approval. - 24. New (2 tf*) two sigma and (x) averaga values for tho net Thermal Chlorides and Pyranol blank (chemical Cl" contribution) will bo calculated ever;thirty days or ovory fifty-two (52) samples, whichever is most frequent. 25. Ihe now control limits and any obvious trends in results obtained will be brought to tho attention of the unit supervisor. MONS 034878 CLTM-37 Date Issued; Revised: ^ VOLUME RESrSTIVm OP DIELECTRIC LIQUIDS Principle: The volume resistivity of o dielectric liquid lo an expression of Its ohmic resistance times a constant vhlch Is representative of tke volume occupied. The D.C. resistivity Is a rough cause, easily measured, of material purity In order to obtain a continuity between readings like conditions must be met with each consecutive set of readings, l.e., temperature, cell condition, exposure to light, etc., must be similar In each Instance. The values ob tained aro determined on the basis of on arbitrary set of conditions previously agreed upon. Altering any of these conditions lessons the continuity be tween readings. . Reference; ' 0.8. E4D8-32 "Physics''' E. Hausraan end E. P. Slack S. Van Nostrand Co., Inc. New York, New York (1946) Equipment; .. (l) 0. E. cylindrical rebletlylty eells, Catalog (* 1,559,63 nickel plated brass. (8) Laboratory thermometer operable to 0 - 100*C or 0 - 200*C 1*C and maoting ASTM Spec. El. . . (3) "Pyrex" glees spacer plate for above resistivity cells. (4) 600 ml "Pyrex" beaker. (3) Heating mantle with auto transformer for voltage adjustment or hot plats, me heating mantle Is preferable and should closely conform to BOO ml size. . (6) Resistance meter: .Electronic Instruments, Ltd. 20 million segohnster General Radio 544b. Procedure: 3efore testing may begin the resistivity cells, glass spacer, beaker ard thirr.ar.cter should be thoroughly cleaned. The following procedure should be HONS 034879 - CI,Ti-;-37 Page Two used In order to obtain a continuity of results between lnbo. Tf otter cleaning nothods are used, there may not be a continuity of readings between labs or with past results. 1. All parts such as the beaker, spacer, cells, etc., should be vapor degreased prior to washing. 2. The glass spacer, beaker and thermometer may be washed manually with "Aleonox" or nay bo washed by soaklnc In Ooklte bottle wash solution. 3. They should be rinsed with tap water and distilled water. It. Each Item should be checked for wettability with distilled water. Complete wetting should take place with no soli or slick spots remaining. 5. The resistivity cells should be washed by hand with "Aleonox" or other non-abraslvo cleaner. A soft cloth or paper toweling may be used to facilitate cleaning. Do not scratch or mar the bright nickel surface. Eadly oxidized or scratched cells with plating worn thin should be replaced. Hand washing is the only accepted way to clean these cells. Do not put In bottle soak or other strong caustic. 6. After washing,the various parts are put Into 100*C oven to dry for. oao to two hours. They should not bo left In tho oven overnight as this will fora an oxide and require rowoshlng. 7* After washing, the cells are assembled In the beaker and the air capacitance determined a3 per CLTK-35. 8. The cell constant (K) Is obtained by multiplying the capacitance In uuf x 11.3. K - C (uuf) 11.3 9. Prior to obtaining readings on samples the cells should be conditioned in the material they are destined to bo used in. Normally three or four rinsings and electrifications at 100*C will bring the cells to a steady state. If the resistivity meter Is calibrated and steady state readings cannot be obtained, the cells may have to be recleaned. A set of cells should be re served for one dielectric liquid and are not Interchangeable between dielectric liquids. ` ` 10. Prior to placing liquid to be tested into the hot cells (100*C), the liquid should be heated to 100*C In an oven or heating mantle. This vLi': bring cells and dielectric to the same temperature range and eliminate fluctuations In resistivity brought about by temperature cycling. 11. Pill the cell with the heated liquid to be measured. 12. 3rlto temperature specified - usually 100"C t 1*C. HONS 034880 CLTO-37 pngo flirce. 13. Attach leads (a) shielded lead to Inner electrode (b) guard wire to external guard, i.c*, foil or Insulated metal beaker (c) third lead ;o outer clectrodo. . . lU. Make sure resistivity meter le calibrated or zeroed prior to use* 15* Charge sample for 30" (a) on Electronic Instrument depress "A" button (b) on G.R. bridge switch selector knob to charge. l6. Bring meter Into the circuit (&) 00 Electronic Instrument depress "B" button while leaving "A" depressed (b) on G.R. bridge switch to operate. 17* After a total of one minute electrification obtain reading end calculate resistivity. __ , P (resistivity) u 11.3 x C x R P a XR where - K b cell constant R * resistance in megohms 11.3 b conversion factor from electrostatic units. 18. Report results as follows: Q a. ^Calculated resistivity usually reported at x ICr ohm cm for ] ronolGj x 101 oka cm for oils. b. Record temperaturo at which measurement was taken. Most di electrics are measured at 100*C. c. Record voltage. d. Identify material under teat, t.e., 1**99, A1331C or AR . 19* Cells may be stored in the dielectric liquid until the next use. ffiey should be covered and shielded from exposure to ultra-violet light. Deviation of Cell Constants; In arriving at a constant for resistivity cells the foilowing should be taken Into consideration: 1 Cylindrical Cells: The resistance between two coaxial cylinders of length 1. end radii and la: jr b resistivity 1 natural log 2F l1nn _ area R b resistance (ohms) K b cell constant MONS 03488X CI.TM-37 Pago Four . 2. or K 8.303 loC[10 sWTT or the resistivity S nay to written 03 the product of a cell constant K and the resistance.n. . 27fl 3. yf'> Kn where K . 2.303 lot^o rx cn. The cepacttonco between theso sene tvo cylinders is: 1 4.15 Ios^q r2 uuf If ve divide C eq 4 by K eq 2, we obtain 6.30 V K - 2.303 C or 3.677c - 11.3 x c K cell constant '1 C " coll capacitance In uuf. ' The determination of cell constant for Parallel Plate capacitors: . EA C ------4 TT's . C Capacitance esw E m Permltlvity 1 A - Area S " Thickness of dielectric'or apace between plates To convert C In esu to m.f., the divisor is multiplied by 9 x 10^. Then In order to convert the product to a familiar form commonly used uuf, the whole Is multiplied by 10 . C - EA x 10* 4Jf S x 9 X 10? or C EA x 10 EA x 1 4*78 x 9 S 3.67T HONS 034882 Mono l iKT-HAnn) EXA!!x;iAmcc:; of DIFXKCTBIC LlCUiro An-wratua: Bpcctrorseter, Fcrkln-Sltxr, l'odol 21 or othor doublc-bcan Instrument, profomhly vlth Ordlnato Scolo Srpenoloa FollcxrlnQ JIcCl cello: 1 patched pair of scaled cello (0.1m) 1 0.09? na coaled 1 Citountoblo cell (capillrry flL-i) Chartd, iM) ecalo (5 ca/u) cod Koyoort card (lc=/u) goot Snoci-xni Tho teat npccluon oball conolot of an aliquot (usually 1 <jt.) representative of tho Botorlcl. Procedure: A.djuot cpoctronoter to tlio folloulcj coadltlono: Itocolutloa - pS7 Pcoponoc - 3 ooc. full ocalo Coin _ ?-l/2(noalnal0 Speed - 1 dn/islcron (u) Suppression 8 Sourco current - 0.3 nap. Plaeo undiluted oa-tpleo In sealed 0.1 na, 0.09? cm end eeplllary fils cello. Uolnj lull ocalo chart end 0.1 cn cell record 2-1? ulcrca re*lcn. Ice O.CC? aa end capillary fllma to recolvo totally otsorblac bnr.do obeerved. A JinCl crystal of suitable tlilchassa lo usad ca reference for tlieco sod. HONS 034883 Cn the Ena clinrt rceox-d a differential epeetnu vith 0.1 :u notched cells, uolG tlio appropriate* otondard In tho refcronco bean. repeat t'.o differential enclyois at 1 cii/siicron on o Keytort card. Rmalne tho difforcntlpl oneetrua and noto my cni'hed departurco frea prcrlouo differential curdyoio, An aliphatic hand at 3****3.5 tilcrono and diphenyl bar-do . at 13.5 end 111.5 rxlcrons oitcvld ho ebaoat for nil Pyrnnol ocr-plso. In n-HltLer., ll| Pywraol in 1>i7o Pyrnnol can bo detected at -10 nicrono, and 1)17'. Pyrcnol In lfc59 Pj-mnol by a bind at 7.^5 - 7.5s nlcronn. Borort; Tho report chill inclulo tho cannlcto idcntifieatlcn of tho notorial tooted, t1 infrored chart nrjnbsr tod any cbnori'lltios ail/or chromes evident in the xxitcrlal. HONS 034884 Jle.JO ,,* Treated Toll Teat for Ctuvllty of Dlclccttlc Paper Purr-or.c i The purpose of this procedure is to obtain data on vound and treated ualto prior to accepting paper cs an indication of paper quality. Hilo technique can also bo uoed in evaluating mr.r rocliaes and proeceeos, and row types and rruten c: par The teoto uo.tunlly perfon^ft on the units nro dry absorption, treated povor fee to at tiro tci'tpore.tureo and ft C. dielectric ntreiiQtii. Procedure; thirteen units nro vound oeoordlne to tho ucrkcA thlclncsn and Table I below. Tho Information bolov la submitted vlth the vound unite. Eanplo nurtor Vendor tot mciter - Specifications Hiiekncaa . Poll Vendor Foil lot nur.bor Turns root Winding Oi'.te Winding operator Winding nacblne 3011 number The vlndor oaoirr.o ample niE&ers consecutively vlth on "if prefix, fills in the card and dallvuro the wound units, unused paper and record eerdn to the rsaecSlsr Tho noocublcr eocuhlco tho rolls to covora (U35.'13ClAfS5l) end cans (lCdElcri). Three of cut croup In the hex with tie eocene rn>rr, after .crrilng the ":f n.c-hc on tho cover are retained. Chic ton renaming units ore seamed, tagged, eni rout treat. MONS 034885 Aftor tho treat the tea unite o to Q.C. Fl\o ore hand lnbolod, removed, csd pieced In n 15*C oven " overnight. Hie next, nomine t!ia povor factor end capacity nro read nt 300 volts-por-ciil. The unite ere cooled overnight and rend cgaln tho next toy at 25*C. limn tho (Into la nvotlnblo certain calculations ore tvulo. 71(p poucr fr.ctoro for the tiro ter.^err.turoo end tho brcahdovn renalto o.ro everoced. Proa tho ir,*?.'' tvo quantities vi.ll bo oubtrocted. The first, called o polorcocponcnt, la dotcrnlnod frees the type of pulp nnd density according,to tho following: HO pulp (A50?Clh) Po .230 x density A graph of this one equation In nr.do ovalinblo go that calculations nro not necessary. The cocond quantity called tho Ionic ccnpocent, lo a function of the absorption reading as follows: X .064 X Absorption Tho 125*W Is obtained by subtracting a foil loss vhicli la a function of foil length end capacltsaco (tho winder records foil length froci an unwound unit) according to the following: t a 25*c D LC/4446 t lencth la foot 65 100 115 1B5 D - tc/3753 D > tC^310 D - tc/3151 D - lc/3052 C b Capacity In uf DgjSlI foil loss Proa thlo difference the above two coapononto are deducted nod the residual Is tensed "cstccsa paver fhetor". Tho ealculntlon of the polar component Is adjusted co that the oxceao lo very closo to soro for n nergal lot. The natural variation la power factor and absorption oasetlnes enueo tho enlculctcd euccon to bo negative. A hand tolly lo tnlntnlned of the nverojed and enlculotcd results erf ench lot. The paper XtCb calntalno chnrto by density of four of tho moults: 125* PF, C3.J rr excess povor factor end breakdown. Periodically distributions are torn up end 95,' (JO) lhiltn nro determined far control limits. MONS 034886 cri.lbrr.Mon: Tho oven ixed to test for power lector should have a rcecider-ccaRoller. 1'our units ere rood. A standard ofoorso typo to needed for tlso absorption eyotca that vUl detect Inaccuracies when tbo oyotea la operated, rather then aoparato choc to on each of the ceopoacnto. For power factor lead lccsco, tHrtcen (13) uslta voro selected vlth cpproslnately tho sar power factor levels hut vlth a full na^e of cc.paeltraco. If tbo dlffercnca between tfca actual povar factor reeding aai o a tarda value corxolatcs with capacitance, a lead loon lo Indicated. HONS 0348S7 gACTOar TBIAt WlilMIO -t;sfillrr.Ti-n-n Ksmirol - .375" Foil. -.fa) .a5H * 1.5" PP"P - tt) 1.75" x thickm tP* ' - 931X913 - P4 Sum*- - As Followst iqulrsd flilcknoas. ,20* .25 ' .30 .35 .40 .45 .50 .55 .to .& .75 .80 .83 1.00 Foil Turaa* 450 390 344 308 278 254 234 216 202 188 .1*7 158 145 130. Tso Turns** 288 250 220 193 178 1*3 150 138 130 121 107 101 93 83 Fast--* 113. 98. 66*3 77.3 70.0 3.8 >8.8 54.3 50.7 48.8 . 42.0 39.a 3*. 5 32.7 Foil tons asdir '.V*5y **t*p turns foll^turns ***F..t . 79.4 f .2tp'. W - Can iffilth - 1.17" t| Foil thlclcnssa . ,3513" tjr- FApor tM.ckn.sa > >i_ l*bls MONS 034888 u.vi'on c.gat:s cn;:ncAr.s company UBOituTOliV ClAXVAI. TtlfTAnViv SPECIFICATION METHODS ':is-8on;os-: Paco 1 of -. November 25. 1059 . EPOXIDE 200 VlyVLCVCLOI'EXEXB DIOXIDE 1 .SPECIFIC GnAVITY .. ") Dotcrm 1 no' nt 20'C by moans of a hydroaotor calinracoo tu . give the apparent spsclftc gravity at 20/20'C and capable of bqlng read to the ricarost 0.0005 unit. ' . b) Maintain the constant temporature bath at 20 0.05'C. c) Roteroach 9-B2-4 - -.. VISCOSITY . Vy a) Dolermlnc the kinematic vls^o**ll\\yy atlNyySfio spcci'flod tenpor- at urn by no.-ins of a calibrated capillary-type viscometer bavins the required contlKtlPito.^vnj:e. . b) Une cither an Vbbulohdc or Nut Outu.Md-Fcnrko .'viscometer. o) Hr port the vlnconi t y Iin peennttn: itokcu. ,6s TOTAI, EPOXIDE CAUTION: This proeedur<a^nvolv?s tho uso of perchloric acid. Do not uno thlVr^cont unless aware of its'potential hazards. . .......... a) Kvdrecnn bronldysolutaon, 0.5 K In acetic acid By naans of a 5rad-.mt0yi.nfeiul ly add 07 sl ot c.p. bromine to two liters of glacinfSacelle acid. Add rongont grade phenol in 10-go lncVcrtnnts until the solution boeonos light , -straw In r-Cj.ni.'(approximately 100 g) and add 10 gm in oxcewsl >(Vx tmr solution after each addition-of the phr nnl.yind nV}n the final solution to stand ovornigh iVs*.. f hi--* lining. .. . , ,b) 0,2 N -oC-'-.-iu--r--.-A- v--c--e--t-a---t-e- ---I-n----a--c--e---t-i-c---a-cid -D--l-c--s--o---lv--o- ---1--0---g--a-oof i ' itMlvy3ro&\/.u{Iu acotate in Nufriclenl glacial acetic. M.'l to nako one liter of'solution. Stanr-d,-a"rdJi,s"e- this doUi'len against standard 0.2 K pcrcblorlc acid using 1.0 rpeetY"-cceent crystal violet Iindicaattoorr..in acetic acid, e) P'-needy.1.-c Prepare a sufficient .-uir.brr of .rloan dry ' `jKH-mi iodine flasks to porfona all or.nple and blank.. >l*t originations In duplicate. * '' - d),Carefully ptpet 29 nl of the hydrogen broaldo solution ' into each of the flasks, using the sane pipot for each . transfer. . ) Reserve two.of the-flasks for the blank determination. f)f Into each oj\ths other flasks Introduce 0.3 to. 0.4 gn of tho sanpln.weighed to the .nearest 0-1 by Beans, iff-a. suitable weighing pipot and swirl to effoet . 1* iv:< .-. .soilluuttiipen. . ' .... '. r . rr. v 'fT U. <- <.' C: :AU:: mons oahaao Tin-300206-1 u2 of I LABOilATOllY MANUAL g) Stopper the flasks. using 5 ml of glnclnl acetic acid a3 . the liquid real,and allow tho Bisiplos to stand with tho blanks at room temperature for 15 minutes. ... - h) Carefully remove the stopper from each flask and wash down . the stopper and Inside walls of tho flasks with 25 nl of ' glacial acetic acid. i) To each fl.iwk add 5 or 6 drops of crystal violet Indicator and titrate with standard 0.2 N sodium acctato In acotic ' acid to the first bluish-green end point. J) Calculation ` (B - A).N x 1,6 . gin sample total epoxide, % by weight, as oxlrane oxygen A - ml of N normal sodium acetate required for tho sanple , B - average ml of K normal sodium acetate required for tho blank " 4 ACIDITY . a) Introduce SO gm of the sample weighed to the nearest 6.1 gn Into a 230-mi Srlenmeycr flnsk containing 5C . 1 .of nothar.ol which hnn been previously neutralized to a faint pink end point using 1.0 per cent phenolphihaloin lndlcr.tor In 1 ` .... nothmio'.. Swirl to offect solution. ... *t>) Add a low additional drops of the indicator nnd titrate with standard 0.1 H alcoholic potassium hydroxide to tho first pink end point pormanent for at least 15 seconds. ' c). Calculation ' gm~saupYo ' * by Acotlc acid A - ml et S normal KOI1 required 6' TOT.11. US3ATvT..lT10;i , . ' a) tfljs <i.ion. Dissolve 13 gm of rosubllmcd Iodine In ; "Tol'd in."f'"ctacini acetic acid. Centle beat may he nccos&ary i.r offoct solution. Allow to cool nnd rea'ovo apprbvlrr.te'.y 200 ml -it tho solution. Pass dry chlnrina . gas into tho remainder of the Iodine jolutlon until, the original titration,Is not qulto doubled. Perform tho . titration as directed bolow. . A characteristic color ' change tth-'i place In tho ffljs solution when tho desired .amount o1'.' chlorine has been added. This any be liscd as. an aid. to .lollglnt the end point. Add a small oxcesn .of chlorine nod hrlng baek.to tho desired titration hy . addition of -Hir.-o of the original lodlno solution. Into respective 0')0-mt ErJenncycr flasks pipit 25 ml of tho solution before addition of'the chlorine ar.d 25 ml of HONS 034890 LADW.T^iY M.'.N'.iiMi P3 of i. T13-fiQG200-l '.hr tin\uil<v sifter add l lion of rhlor lnc' Add ?.0 rl o i . iS jvr <*cnl pot Iodide /solution io dJntlHod water .ird 100 ml of fllstlllrd water to each flarrit. Titrate tho contents of cnch flnyk with ntnudarct 0.1 It tliio- ' ' ftulfctc until the yollo-f color almost disappears, Add ' Va til of 1.0 por cent s tarch indicator solution aatl continue titrating until the blue color disappears, b) ?rogotV.rc llcat all oamplcs which are solid or seal- sufl-i or contain turbldii y under an Infrarod heat lamp or In a hot wator hath ur. til conplotnly homogonooun, and mix sill. . ' ". ,c) Prepare a sufficient number of clean dry SOO-nl glass ; ' stoppered Erlonir.eyor flasks to perform nil blank and sar-ple determinations in duplicator : d) To om.'i flask add 20 ml of rosger.tt*:rr.v.!o carbon totra- ; chloride by neano of a graduate. J ) Rcenrve two of the flasks for thoVelanuMieternlnatlon. f) Into each of the other flasks ^intnvlucn 14 to 10 ga of ' tho sample weighed to the non to ? . 01 gm nnd st/irl tc effect coapletc solution. ' - C> Into each flask ptpot 25 rfl. ofVjc Vijs solution using ' prow *ure rathor than ce^atjOn to*fill the pf.pet and ti'jln s*.vlrl to offoct foluldon. . h) Aliev the flasks to stand In Pho dark for 30 nlnutes t room tenperaturo. . ' ' I) To each flask add 20 \rftio 15 per ecr.t potassium . t Iodide solution ,/ind ltd ml of dlstlllod water. - . J) ''Titrate lnmodlay:Ay with standard 0.1 N sodiun thlc ' until the yellow color almost disappears. Add two the starch inlV.cator*solution and continue titrating to tho dlsappe.^ratibw; of tho blue color. If the sample ' titration fSTleribtitan 50,por coat of the blank repeat. , . the de!c.rritna'Vpn using a sraallor.slr.c samplo. 10 O.Iculaj^oi^ 1 i*^-r'"*~jN:(p^- " total unsaturation as Iodine number x . - l>f N nncormal Kh^SgOg required for tha sample average ml of It normal Ma-,S-0*| required for tho bleak 301*0*1 , , a) Ti-nisfcr 100 ml of the samplo to one of two matched tall .*' form Messier*tubea. ' * . b) Fill tho second tubo to.tho mark with tho platinum-cobalt * : otnadard that appears to match the color of tho samplo. C) Compare tho colors of tho samplo nnd tha standard by ' viewing vertically down through tho tubes against a vhlto background. ' HONS 034891 d) tloternli'.o tiio exact color of tho snnplo by sucrcxulvo i-cpiae iror.t of tho standard .tn the second tuba until an exact mate ' ' In obtained, o) reference 5-Ji-l iUSPUKDEB l!ATi"::t , 'a)/lnvert a bet tic oi the ranplo and examine by transnittoo llpht. DSVZLOPMSHT DSrARTKSNT .. . tXCifh'ICAL CZXTER South Charleston, West Vlrolnta HONS 039892