Document 3Q1JBL6K3EBajVy8gkGkwj5za

6 ,* i UC UV-7 INTIRNAL C RVtSSPONOENCK BAKELITE COMPANY RIVER ROAD, BOUND BROOK, NEW JERSEY To i'Nam*) Compjny location Copy to Dr. T. W. Nale Union Carbide Corporation 300 Madison Avenue New York, N. Y. Cat* OtiginQiing 0pL Aftiw*rt*9 lottor dof* Mr. Mr. Dr. Mr. Mr. Mr. ...Mr. Mr. Mr. Dr. Mr. G. R. R. J, K. R. H. G. N. H. R. Sub/acf P. Bigelow - 9E. 41 A. Calsibet - 9E. 41 F. Clash, Jr. - BB M. Holmes - 607 V. McCullough - BB L. Norum - Clifton F. Robertson - 406 T. Scriba - 1901 R. Smith - BB F. Smythe, Jr. - Mellon D. Swain - Phi la. December 5, Development Dear Dr. Nale: You will recall that Mr. H. F. Robertson initiated a testing program with U. S. Testing Co., Inc. to determine the weight loss in a VMCH coating when tested under the Continental Can Company extraction procedure. This was instituted as a result of a letter, Lehman to Nale dated June 11, 1957. Two copies of the report from U. S. Testing, Report No. 33740 dated November 20, 1957 are enclosed which describe the resultBof their investigation. Weight losses were found to be negligible and residues recovered from the extracting solutions were free of resin as determined by inflra-red spectra. Very truly yours, WBA /bhm Enc. P &. W. B. Ackart ucc 030893 -*t ' - ' . f- BAKE LITE COMPANY DIVN. OF UNION CARBIDE CORP. 30 EAST 42nd STREET NEW YORK, NEW YORK .1, <4 >, ^ * . ;*rrr . C REPCRT #33740 - NOVEMBER 20, 1957 t' - * av,`v^ j * V'- ";.V , , " * </-: f` - *- -, -' i"=V?,i4 *. " * /-ift- x, vA>> t . - : i1 ... - ,, . ?` ''? '-Tv,* * , * .* ^ ' ' ,vT. r ;,s ,V V* r*ri3.1 ci : --* UNITED STATES TESTIN,\G'w' (^X), Inc , ; ,nyfm-(i V.*1 , K vt* v* (ESTABLISHED 1HO r* ,, " - _-v- i HOBOKEN, NEW JERSEY : `''. i: V-'-tf- NEW YORK CHICAGO BOSTON . ------ -------- ; / DALLAS PHILADELPHIA ;; MEMPHIS DENVER ' ` " PROVIDENCE a-A' ,r,, LOS ANGELES ~ ..-v T ' >1 -. ; - fj - - - 4' UVv l .-"s .*", 030894 I ti 1 %> iti * United NEW YORK BOSTON DENVER DALLAS MEMPHIS States Testing Company, ESTABLISHED 1*80 HOBOKEN, N. J. TELEPHONE SWarthmore 2-2AOO ---------------+--------------- Inc. PHILADELPHIA PROVIDENCE LOS ANGELES SAN ANGELO BROWNSVILLE REPORT NUMBER 33740 IQYEMBSR 2Qf 1957. ______________ DIVISION OF UNION CAMBIDE CORP. 30 EAST 42nd STREET MEW YORK, NEW TORE MATERIAL ONE (1) SAMPLE "VMCH RESIN". PROJECT DETERMINATION OF EXTRACT ABIES IN A CAN LINING RESIN. i. f- ERVISED BY C.A/ Gerard! General Chemistry United States Testing Company, Inc. y T*/Ryan ETTERS AND REPORTS ARE FOR TNC EXCLUSIVE USE OP THE CLIENT TO WHOM THEY ARE ADDRESSED. AND THEIR COMMUNICATION TO ANT OTHERS, E USE OP THE NAME OP UNITED STATES TESTING COMPANY. INC.. MUST RECEIVE OUR PRIOR WRITTEN APPROVAL. OUR LETTERS AND REPORTS APPLY TO THE SAMPLE TESTED AND ARE NOT NECESSARILY INDICATIVE OP THE QUALITIES OP APPARENTLY IDENTICAL OR SIMILAR PRODUCTS. SAMPLES NOT DYED IN TESTING ARE RETAINED A MAXIMUM OP THIRTY (SOI DAYS. ' \ 4.*7 s ucc 03089? UNITED STATES TESTING COMPANY, INC. 1415 PARK AVENUE HOBOKEN, N. J. i fkE'Lh. or COHTSMTS Introduction SuBaary and Concluaiona Data Diacuaaion Appendix - Teat Procedure Pace #1 Page #3 Pag* #5 Page 012 'C ' ,'TT i 1^ 1" 1[ 1 ^7 I I JrV 41 B 030896 United States Testing Company, Inc. Page *1 Introduction In the evaluation of a resin or plastic notarial as a satisfactory lining for food containers, it is obvious that a prime consideration must be tne absence of toxicity as sociated with the use of the material under evaluation. It is essential that if any material which may be toxic la present In the can lining, it moat not be transferred to the food product. Obviously if it can bo shown that moms, or a negligible amoiat, of the material forming the liner of the can is soluble in the food for which the ean is used, then there is no risk of toxicity. This method of evaluating the auitability of a ean liner has been recognised end is referred to by Or. A.J. Lehman, Chief, Division of Pharmacology, Food and Drug Administra tion, Bureau of Biological and Physical Sciences, Department of Health, Education and Welfare, Washington, D.C., in hia article entitled "Food Packaging* (Association of Food ad Drug Officials of the United States fol. XX, bo. A, October 1956) In which the principle of weight loss as a measure of insolubility Is discussed as a technique for the evaluation of resins for use in food packaging. The weight loss method is the basis of the study which is reported here. Infrared examination has been used as a supplementary technique, even where the weight loss was negligible, further to confirm that such residues as wera ucc 03039 United States Testing Company, Lnc. p** #2 Introducti n (Ccntd) recovered tram toe test solutioms Aid mot rosult from the solution of the resin under test sor any of its components. Vo ore indebted to the Research Division of the Con tinental Can Conpany who were kind enough to give their perm! ssion to use s method developed in their laboratories. This procedure, as modified in minor ways found to be necessary in connection with the particular type of rosin being tested, is show as an Appendix to this report. United States Testing Company, Inc. PMe n So--*ry ind Conclusions The material wider test ms supplied by the elleat --d ms Identified as VMCH fcesia", la the form nf s 20* solutioa of the resin In a solvent mixture consisting ef equal parts ef toluene and methyl iso butyl ketone. Following the technique developed by the Continental Can Company, with such modifications ss mere found necessary, considering the particular type ef coating being examined, a study ef *THG8" resin ms made te determine quantita tively the amount of can enamel extractables resulting from the contact of this manel, after baking on pure tin foil, with solutions simulating a variety of food products. Details on the preparation of the film on tin foil, the conditions of extraction, the ncthod for determining weight loss, and supplementary examination are found in the Appendix, "Test Procedure." All extractions and weight loss data were taken in quad ruplicate. The results are presented first in the form of data obtained for each type of simulated food product, followed by discussion end interpretation of this data. In all eases, the weight loss ms found to be negligible. Confirmatory examination of such snail amounts of residues that were obtained fron the recovery of enamel extractables from the extracting solutions showed these to be negative, in that the infrared spectra of these residues indicated the absence of the resin used in the enamel and any of the other components. ucc 030899 United States Testing Company, Inc. Pw #4 3u--try and Conclusions (Contd) It is concluded that the TMCi resin", when applied Ter tin foil in the fen of a 20% solution of the renin la a solvent mixture consisting of equal parts of toluene and methyl iso butyl ketone, and subJacted to the conditions prescribed herein, followed by the "extraction" of the coated tin foil otrlps with the simulated food produets used in this study, yielded a weight loss after extraction which may be classified as negligible* |1 C o p Y 030900 United States Testing Company, Inc. *5 TABUS I Extraction Being Redistilled Water A. Enamelled strips air dried 1$ ninutee; baked 275*7. for 2 minutes a first aid* and 6 slant** on a*ccad aid*; oven dried at 221*F. to coastant weight. B. Weight loss after extraction: OrLiggiinaaali Ww< eight (grans) Weight final Weight Change --foraST" (nllllgran*) .1. 22.2116 2 23.1160 3. 22.9141 4. 22.556 22.6117 23.1160 22.9146 22.5563 .1 0.0 .5 - -5 .1 Averago gain .3 Average deviation from mean value. C. filter Residue - #200 nesh stainless steel screens: Original Weight (grans) Weight final Weight Change (grans) (nXTIXgraas) .1. 1.1550 2 1.1535 3. 1.1620 4. 1.1181 1.1550 1.1536 1.1622 1.1160 0.0 .1 .2 - .1 .05 Average gain Examination of filter residue*: Evidence of e slight a- nount of ataospherlc dust; no tin or resin particles present. C P Y D. Recovery of extractahles by evaporation: Gain Blank (600 ml. redistilled water) Net Gain 1. Sample destroyed 2. 5.9 g. 3. 6.2 4. 5.2 3.1 kg. 3.1 3.1 3.1 Average 2.6 ng 5.1 2.1 3.3 ng. ucc 030901 United States Testing Company, Inc. fw -ft.....-... TABLE II Extraction Being li Salt Solution A. Saamellad atrip# air dried 25 minutee; bafced at 275F. for 2 minutes on first aids and 8 minutes on aacond aids; oven drlad at 180F. to constant weight. ! B. ; ! Weight loss after extraction: Original Weight (grams) Final Weight (grams) Velg* Change (milligramo) 1. 2. 3. I 4* 22.2939 21.7751 22.1650 22.6419 22.2940 21.7751 22.1651 22.6421 .1 0.0 .1 .2 .1 average gain .05 Average devia tion from mean value. C. r i: 1. 2. s 3. i 4' 1iI; Filter Residua - #200 mesh stainless steel screens: Original Weight (grama) Weight Final Wei? ht Change (grama) (milligrams) 1.1550 1.1536 1.1619 1.1179 1.1551 1.1557 1.1627 1.1185 .1 2.1 .8 .6 .9 Average gaim Examination of filter residues: Evidence of alight a- nount of atmospheric dust; no tin or resin particles present. D, Recovery of Enamel Lxtractables by Liquid-Liquid Extraction: No conclusive weight measurements could be obtained on | this recovery (see discussion", page 14 i. However, Infrared ucc 03090 i * .1 ot *r .S .X .c .4 .0 .1 . *.eVX1 orr -- i 1 United States Testing Company, Inc. j. 1, Table II (Contd) ^.... ..........-.... analysis (see fi*. 5 ) shows conclusively that no resin Is present In the resides, and that the portion of the resides which was soluble in the toleene-MIBi solvent Mixture* is the sane Material that Is pre*t in the blank run on th extraction solution In the absence of the enas*llsd foil. j' Ji j, *>:IBK " methyl iso butyl ketone: Mixture la equal parts* !l ! 0 P Y i; 1 e i' 3~d2 1 ucc 030904 United States Testing Company, Inc. T TABLL HI Extraction using Salt, IQi Lard Solution \k. Eaauelled strips air drisd 25 sdnutes; Waked at 275F. for 2 minutes on first side and 6 niautes on soeond side; i Ten dried at 180P. to constant weight e B. Weight loss aftar extraction: Weight 1 Original Weight (grans) Final Weight Change (grans) (nilllgrans) j 1* 24.0110 2. 23.2107 !l 43.. 22.7280 22.7292 24.0103 23.2105 22.7283 22.7292 -0.7 - .2 .3 0.0 JG.T5 Average loss 3 Average deviation froa man value j c. Filter Residue aot applieablo. D. Recovery of anawel extractable* not applicable. C P y iJCC 030905 United States Testing Company. Inc. P*e.........^10 ........ TABLE it Extractive Being 10^ Bueroee, Citrate Solution A. Enamelled atrip* air dried 25 minutes; baked at 275#J* for 2 minutes <m firat aid* and 6 adnntas os sacond aida; cron driad at 1B0*F., to constant wight. B. Weight lose after extraction: Original height {grams) 1. 21.$957 2. 23.467$ 3. 23.0120 4. 23.7190 final height l grama) 21.$920 23.4670 23.0120 23.7163 height <fes-.i -3.7 - . 0.0 -.7 -1.1 Average lees 1.2 Average devla tion from moai value Corrected for the tin blank (aae "discussion*, page 17 ), ! of 0.3 ngr., the above average net loss becone* Q,$ mgr. C. niter Residua: Determination of this value by weight was found invalid i1 i: as evidenced by stains on the neshea. The residue weights I |i were inconsistent with the apparent amount of residue, e.g., r : Ca 7 mg. weight gain was obtained for a mesh which contained only a slight amount of dust and fibars. Examination of realduee: Evidence of some dust and fiber; no tin or resin particles present. ucc 030906 United Sates Testing Company, Inc. fable IT (Cent!) Page #u D. Recovery of extraetables by liquid-liquid extraction: Vo accurate and conclusive weight ueasuranests could be i obtained by evaporation of the ehlerefor* phase (see I "discussion", page 14). Vevever, lafeared analysis (see ! fig. 7 ) proved eaaclusivaly that no resin was present in t ; the residue. I tJCC 03090? United States Testing Company, Inc p*e ^12 ..... ftlscuaslon j: The method developed by the Continental Can Company was followed in detail eo far as possible; however, since the enamel studied by Continental Can yielded fairly large a| Mounts of extractables, whereas in this study, the aasel under test yielded negligible amounts, certain daviations from the Continental Company netted were deeaed noeeasary i in order to obtain eoacluaive rasults. In other cases, the applicability of certain procsdures was doubtful; banes, these procedures were Modified or li j ositted as deeaed aeceaeary, and as explained in this section. i A. Discoloration of Coating j In the early stages of this study, the tin foil strips were baked at 300F. for 2 minutes on one side and 6 minutes on the other side to give a total baking tine of 10 minutes and 8 minutes respectively* The strips were than dried at 221F. for 15 minute periods* Two adverse effects resulted from this procedure: The first was s marked discoloration, ranging from s blue to black which was evident on all strips so treated, which could not be accounted for. I The second effect was that constant weight could not be obtained on the strips. In successive 15 minute drying i periods at 221*F., the discoloration of the coatings in1 creased, and a fairly constant weight loss of about 1 mg. ` or more was obtained after each period, which continued United States Testing Company, Inc. 413Pe....... -........ About six drying periods. It Mt eescluded that the dis coloration and weight loss resulted from the same muss, probably not fros loss of Moisture or solvt; hence conditions were sought to oorroct both offsets. After experi seat al work, it was found that e baking temperature of 275F., for 2 minute end 6 niaute internals sufficed for 1 baking, and that the drying tsaporat in-e could he redwood I to 180F. This proved to be the key factor, for although sone discoloration persisted at the new beking tenperstur , it was found that the discoloration alnost eenpletely dis appeared at tne reduced drying tanpsrsturs. The strips were dried at 160?., for 30 ninutes end weighed, then dried at 1$0F. for 15 ninutes and reweighed prior to extraction. Constant weight was obtained by this procedure in alnost all eases; where necessary additional drying was carried out wtil constant weight was obtained. In addition, it was found that insufficient air drying subsequent to coating and prior to baking resulted In bubbles on the coating, further tests established 15*25 ninutes, dependent upon weather conditions, as the optlnun air drying tine to give a bubble-free filn with niniaun exposure to the air. B. filter Residue The filter residue was obtained in all cases except for the salt-lard nixture, as per the Continental Can procsdur ; however, in most cases it was found that while the data was not useful for weight determinations or corrections, the ucc 030909 n "" -r United Sates Testing Company, Inc. Fi*e.--.---- -^14 procedure was a convenient Method by which to laevre that no tin or resin particles were present. In the distilled water extraction, only atmospheric test one evident in the filter residue and an avaraga chance la weight of only 0.05 ag. was obtained^ which did not affect the results of the extraction. With the auger solution, and to sees extant with the salt solution, unduly large filter residue weights were ob tained which were unsubstantiated by the apparent amount of residue present. lo evidmice of enamel er tin particles was found. I C. Recovery of hnanel Extractable* (1) Recovery bv Evaporation The redistilled water extraction solutions were e- vaporated in accordance with the Continental Can procedure and apparatus; however, the results were inconclusive. The evaporation process was unduly lengthy and hence left room for contamination by duBt and by the stream of air, me matter how effectively filtered, hence the residues wars high. About 3 mg. material was unaccounted for, and is believed to be contamination contracted during the process. Later studies on liquid-liquid extraction residues confirm ed that the precision of this type of procedure is not better than 2-3 milligrams. (2) Recovery by Liquid-Liquid Extraction This procedure appears to be applicable mainly for samples yielding appreciable amounts of extractable material I li- B ucc 030910 / United States Testing Company, Inc. Page #15 In this study, where the material sought, if uj, amounted to a few tenths of a milligram, the procedure an presented i was of academic value only. In the esse of the salt solution, involving 800 ml. of sample, shout 8 grans of solid eaterial was present. In the ease of the sugar solu tion, over 50 grans of solid material was present. The nethod Involves the extraction of "resin axtractablsa" for 48 hours, followed by evaporation and weighing of the aqueous layer and the ehloroforn layer. It should not bo expeeted that a few tenths of s nilligran of oxtraetablos could bo accurately accounted for, oven by blanks, which thensslvss I j are accurate only to a few tenths of a nilligran. Then too, it was found that a slight anount of solid notarial | was carried over fron the solution being extracted, into the ehloroforn phase and was present In the residue. Thoro- j fore, the aqueous layer was discarded and only tho chloro form layer was evaporated, experimental work was done however, to ascertain the nature of the ehloroforn residues. j The water soluble portion was extracted and the chloride content determined by Mohr titration, and the petroleum ether soluble portion was also extracted. After this treatment about 2 milligrams of material remained unac counted for. Consequently, it was held that the results of this recovery could not be deemed conclusive evidence of the absence of resin in the residue and that infrared analysis would be carried out. i An infrared spectrum was obtained for the pure resin j ucc 03091 United States Testing Company, Inc. .......#16...... (fig. I), for the chlerofar-m residue Alone (fig. 2} and for the tclueae-MIBg* solvent residue (fig. }). Vhon the apectre of the blanks (fig. 4-7) and the eaaples wore coopered, it was evident that no resin was present in the residues, and that tne spectra of the samples were identical to the eurv a of the blanks. Coopering these spectra with those of the chloroform and solvent, it would readily ho seen that the residue did. In fact, cooe froo the chloroform and solvent. It should be noted that prior to rwanlng infrared spectra, the residues were noshed with distilled water and petroleta ether to reoove salt, sugar and other foreign natter, since the resin Itself was insoluble in both. The remainder of the residue was dissolved In toluene-MIBC solvent, evaporat ed, and the speetra obtained. D. Blanks (1) Drying blanks These were fotad to be aboot 0.10 milligram; since the gross loss In weight was nil in all eases, it was not necessary to apply this blank. (2) Solvent blank Again, the loss in weight was found to be nil for the salt-lard extraction, and it was not necessary to apply a blank. (3) Recovery of Bxtractables Blank Blanks were run on all extraction solutions by carrying BOO milliliters through the entire process in the *&IBI methyl iso butyl ketone. ucc United States Testing Company. Inc. p-e....#17..-..... absence of the snamsllsd fall. It was fotad that the resi gns rari ad to such an extent that the blank could not alidljr be applied, e.g., the sucrose blank waa 15.2 mg. vhareas an actual sample had a 10.5 *g. residue of "enamel extractables." (4) Chiorofora Blank The chloroform blank mas fbuad to ba .6 mg./XOO ml. CHCl^. However, it was noted tnat this blank is mnnacassary and In fact cancel# out In the calculations when recovery blanks are obtained, according to the following equation: U-b) - (B-b) - X-B where X the residue of a sample B the blank for that extraction solution b - CHClj blank (5) The tin blank was obtained in duplicate for oaeh of the four citrate solutions: Tin content of extraction solutions. Tin: mg./liter Aver, X 0.800 1. 3.70, 2. 1.75, 3- 1-35, 4. 1.90, 3.65 1.75 1.35 1.65 3.68 1.75 1.J5 1.78 2.94 1.40 1.08 1.42 Avsrags 1.71 mg. Results <of tin blank (run on separate strips) Original Weight (grams) Final .eight (grans) Gross Veight Chan#* (milligrams; Tin Blank Bet Loss (milligrams) (milligrams 1. 21.3722 2. 22.9330 3. 22.8295 4. 23.8508 21.3676 22.9819 22.8290 23.8491 Average -4.6 -1.1 - .5 -1.7 -2.0 2.9 -1.7 1.4 0.3 1.1 0.6 1.4 -0.3 -0.3 on trrr ; !!v _L. --d j ^t M |1 , i a4 4J N j; ii 'Ll j4 j! Ii tT i -j; iil iiill . . 1a - ii U i-41t *4i fdi 1--i ii i 1 r 1 iialii liU 1, d d 4u , U34rli4. tIlt-+juiiu _*. 1 j ELIE 11Mk L__ i n1 a, -1 B i 1f 4 J>4 ** ,; 4i i l| | l _Qpi ' r 1 1 '1 i 1 Uli **L 4=i p- L--r ii. 1-1: ji i 'ii ' ,; ii i T -v *, j rr4 T* : !' i1 "~5 T -- [' tr* 1 1 T I ,i ; ' i i. 1 4 i T' j i 1 r -- T~ L. -- -- -- -- ] , ii .i i , ' 1. -A- -- -*- "4 1 1 .1" -ii' 4, , , --- - 4 L._ i -J -- -- --i i__ .| -- -- . i -1 -- --- -- j i.. _j - *i -T*-. ' | iLi pj4* i-i aL 44pt jiij - f'l?1.-. - -- -- --3p- 14 1 ~{t | l| -- ,i "1 .j 1 T. | HMt > i ': 'j 1 iii: 1 -- \ ii j. *ttr ~\ rl! wre rp/rtA'? r` |im rr r , | ii ir --_ -- -- J .... ,\ .di. V -- <umt rmm n f f* r-- J l .. MKNtUWMVI*"0 i. J z' 4i Ljs-J 44 tr* -4-4 ._L Fig. 1 - PURE VMCH RESIN 030914 ` ,o I (3 'c<-Ooo o o Fig. 2 - REDISTILLED CHLOROFORM RESIDUE rs 030917 f Fig. 5 - CHLOROFORM RESIDUE FROM EXTRACTABLES RECOVERY SAMPIE #2, 1^ SALT SOLUTION CO ucc 03091 Fi?. 6 - RECOVERY CF EXTRASTABLES BLANK ON 1056 SUCROSE - CITRATE SOLUTION vO Sample #1, 10% SUCRCEE 1$ CITRATE SOLUTIOR United States Testing Company, Inc P*e.......... ^1............. APPKMJU Tsat Procedure The test method used la this evaluation follow closely the method developed in 1953 by th Eeseareta Division of the Continental Can Company, hoadod *A Method for the Quantita tive Determination of Can Enamel Extractties. For the particular type of coating being ovalaated la ear tdyt aone minor variations of procedure were found necessary mad are described at the proper places in the report. This method is based upon the use of weight less to dotermine the moot of material that may be extracted from a can enamel by a food product. In the aetual teat pro- j cedure various media simulating a variety of food products | are used. Where any weight loss, however small, was obI served supplementary examination was made to determine j whether the loss resulted from actual extraction of the I coating or was due to small amounts of other impurities. jj The major steps in the procedure are (1) the prepara tion of the enamel film for extraction, (2) the extraction of this enamel film using four types of simulated food products and (3) the measurement of the net weight loss from the enamel film. Preparation of the enamel film. Pure tin foil 0.002 inches thick and 3 inches wide was | used as tne material upon which to apply the enamel. This Form OJ>2 B : ir'r 0309 United States Testing Company, Inc. pe #2 Appendix (Contd) foil vts first cleaned to remove sny possible lubricating material, dust or other contaminants by passing it through an electrolytic cleaning bath (see photograph, Plate X), in which the tin foil was sade the eathode and passed between carbon rod anodes submerged in a 1% solution of nodion carbonate, at a apeed of approximately 1 foot per ninnte. The electrolysing current was furnished by the power supply unit which was regulated to deliver a currant ef approxi mately 5 amperes, thus giving a currant density of about 0.15 amperes per square inch. The cathodically cleaned foil was out into 36 inch lengths and then rinsed thoroughly, first in tap watsr and than in distilled water, and air dried. These strips were stored In a cabinet to protect them from atmospheric dust and were handled at all times with clean cotton gloves to prevent finger narks. The cleaned foil strips were then eoated with the enamel under test to a miform dry film thickness of batwoen 0.5 and 1.0 mils (0.0005-0.0010"), on each side, the average thickness being 0.00075". After coating on one side of the strip the enamel was baked in an oven for two minutes at 275P. to "set" the enamel on the first side. The atrip was then coated on the other side and a final baking time of six minutes at 275P> was given to completely set and bake the film on both sides. To prevent tne formation of bubbles In the filn, the strips were air-dried after the enamel applica- uec 03092 \A Form Gh62 B United States Testing Company, Inc. Page ......^3" Appendix (Contd) tion and before baking for a period of approximately 13-25 minutes. The baking tine ar.d temperature are tboae rcom mend *d by the manufacturer for the type of onanel being tested. Tboae onanelled strips were then examined for inperfeetions and discarded if ao found* The remaining atrips wore then trimed to a final length of 30 inches, thus giving an onanel area of 2 I 3" x 30" or ISO square inches, which ia equivalent to the interior area of three Mo. 2 cana. four solutions simulating various typos of food products" wore used for extraction. These wore: a. Redistilled water. b. Ijt salt solution. Ten grans of raagant grade of salt (not free flowing or iodised)were dissolved in one liter of distilled water. c. l salt, 10* lard solution. This mixture was prspared immediately before use. 1% salt solution was hasted to near boiling, 720 nl. were transferred to a baring Blendor Jar, BO grans of nelted lard were added, blended for thirty sec onds and transferred to the extraction vessel. d. lit citrate. 10 sucrose solution of pH 3.5. One hundred grans of sucrose and 11.3 grams of citric acid wer dissolved in one liter of distilled water. In another liter of distilled water 100 grams of sucrose and 15.6 grams of sodium citrate were dissolved. While stirring slowly, the sodium citrate containing solution were added to the citric acid solution until the pH had been adjusted to 3*5, as measured by the glass 1 ctrode. C o p Y ucc 030924 United States Testing Company, Inc. Page ......... -#4--------- Appendix (Contd) gTtMrtion vessel. (see photograph, Plat* II). Th* extraction vessel consists of a standard Pyrex pip* sap, 6 inches high, A inches inside disaster, 1/4 inch rails, | having a capacity of approximately 1,000 nl. A 1/2 inch I thick cover plate of Pyrex glass fitted in a Teflon "snap- #on gasket pieced over the vessel spelling it held in place with a fiber gasket insulated 1/4 inch steel plate which is fastened tightly to a cast iron retaining flange to horaeti- eally coal the vessel. j The raenelled strips were prepared fer extraction by rinsing in four aoparate portions of petroleum ether, after which th* strips were colled and oven dried at l60p. to I constant weight, Eight hundred nl, of th* extracting solution was placed in the extraction vessel and deaerated by passing nitrogra through the solution for 10 ninutos, Just prior to in sertion in the extract ion vassal, the previously weighed and coiled rarael strips were again dried and reweighed, j the enamelled strip was then imsersad in the air-free solu tion and the extraction vessel assembled aa shown In photo graph, Plata II to fern a sealed extraction systra. Four extractions for each of the media shorn above were nade by placing the four extraction vessels in an air oven at a temperature of 160F. for a period of two hours, except for the sucrose-citrate solution where the extraction tine was 3d ninutes. The extraction vessels were then removed from the oven end cooled in a running water bath for ona hour. Form C-42 B 1JCC 030926 United States Testing Company. Inc. Appendix (Contd) P*e -... & ...... After extrectlen the stripe mtc rmo Ted fron the extree* tion tom el and rinsed (see photograph, Plate III) accord ing to the folioeing schedule: a. Redistilled water extraction. The extracted strips were rinsed in redistilled water. b. li salt solution extraction. The extracted strips were rinsed in fire to six separate portions of redistilled water, the rinsing baths being changed after every fear I strips. c. Ik citrate. 10j sucrose solution extraction. The extracted strips were rinsed in fire to six portions of ro ll distilled water, the rinsing baths being changed after every four strips. ! d. li salt. lOjC lard aqueous Mixture extraction. Host j of tae solid lard was renoTed by gently squeegeeing the strip lengthwise between the fingers. ach strip was then rinsed I according t# the following schedule: rinsed in four portions I of petroleun ether, followed by two portions of water, by two i portions of petroleim other, by two portions of watsr, by ono portion of potroleun ether, then dried and weighed. Rinsed again in four portions of petroleun ether, and again dried and weighed. The first four petroleun ether rinsing batha were replaced, after every two strips, and the remaining solvent and water rinses were replaced after every four strips. Excessive exposure of the enamel filn to the solvent, as vould be obtained with several hours of lnnersion, was avoided. ucc 030928 United States Testing Company, Inc Appendix (Coatd) Page *6 ...... Measurement of net weight lota. After rinsing, thu strips were dried to constant weight at 160F., cooled to room temperature in a desiccator, and weighed. Blanks. a. Drying blank - Affect of drying open the film After pre-nmatlng the amelltd strips at 1S0P. for the Initial weighing before extraction, two strips were re served to serve ss blanks. These were redried in the sane wanner as the extracted saaples and reweighed. The change j in weight was taken as the blank to be deducted. jl |! b. Dissolved tin blank. i The tin content of the citrate-sucrose extracting solution was determined polarigraphically using the method of Godar and Alexander, 2nd. ng. Chew., Anal. Ad., 16 661 ! (1946). This was done on the extracting solution after the iextraction step in which four coated strips (other than those used for the weight loss extraction) were used as I blanks. The solutions were first filtered and measured to the nearest five sis. and the tin content determined In terms of milligrams of tin per liter of solution. c. Kilter residue. After removal of the enamelled strips from the ex traction vessels, the solutions were examined for the pre sence of any visible particles of enamel or tin foil. If these were apparent the determination was discarded. If o particles were visible the solutions were then filtered B I.1CC 030929 United States Testing Company, Inc. PW....ft ............ Appendix (Contd) throngh * weighed 200 mesh atainleea steal Kirs doth ad any residue en the cloth washed several tines with redistil led water, after which the screens were dried et 105*C. for one hour and weighed* This determination is not a blank in tha true sense but was used es a precautionary meaetre to detect any possible fmilt in the wanel application or mishandling of the foil during extraction, d. Solvent blank* This la the effect of the solvent on the vnextraeted enanel filn and is applicable only to the lard-salt extract ing solution. For this blank the strips are carried through the identical rinsing schedule mead In reaoving the lard fron the extracted strip specimens. Any weight lose was considered as a blank to be deducted from the gross weight loss of the lard-salt extracted foil ntrlps. According to the above, the total amount of enanel ex tra eta bias was considered to be equal to the gross weight loss ninus the epplicsble blanks. needurea for Recovery In cases where an appreciable weight loss may be en countered ns the result of extraction, supplementary pro cedures can be used to confirm this weight loss and further to Identify the nature of the extracted material, particu larly to determine whether the material lost by extraction Is the sane in composition as the enamel or one of its ucc 030930 -PLATE IV RECOVERY OF EXTRACTABLE3, BY LIO UID-I.IQ UID EXTRACTION United States Testing Company, Inc. nPage ............ Appendix (Camtd) components. Ia the eat* of the redistilled water extractlag eolation, the amount of eolable extractable material way be determined either by staple evaporation ef the entire solution or by liquid-liquid extraction with chloro form. Za the ease of the ether extracting solutions* liquid-liquid extraction eust be used because of the large amount of solids present in the original extracting aedia. However, liquid-liquid extraction eaanot be used for the salt-lard solution since the lard itself is chloroform soluble. The liquid-liquid extraction apparatus is illustratod in photograph, Plata ZV. Zn this procedure, the entire 600 ml. of extracting solution is transfsrred to ths ex tractor and extracted for approximately 46 hours with re distilled chloroform by introducing about 200 ml. of chloroform in the collecting flask and a sufficient quan tity of chloroform in the extractor to prevent water from coming over into the collecting flask. After 46 hours extraction the chloroform phase is eaporated to dryness and the residue weighed. In eases where an appreciable weight loss has been encountered during extraction it is then possible, by infrared ex amination of the residue, to establish the nature of this residue by comparison of the infrared spectra with those of the enamel components. Form C-62 B ucc 030932 PLATE V - SAMPLES UNDERGOING EXTRACTION IN OVEN'AT lR O F . United States Testing Company, Inc. Appendix (Ccetd) fw.....-............... fr eaeh extraction Media, a blank determination van made in which 600 ml. of the extracting solution vaa sub jected to the entire extraction and liquid-liquid ex traction stages nithout the presence of the enamelled strip. The liquid-11quid extraction procedure vaa used where applicable and the details surrounding the use of this procedure are given in the hedy ef the report. Although the recovery of extraetables by liquid-liquid extraction offers a method of verification end identification, it la obvious that where the weight loos encountered daring ex traction was negligible, the nee of the liquid-liquid ex traction step ia not necessary. ucc 030934 United States Tasting Company, lac. Pt*e ucc 030935