Document 4aZeDbV6ky2m7oayo6ez2gL1V

FILE NAME: Oil Industry and American Petroleum Institute (API) DATE: 1957 Apr 10 DOC#: API133 DOCUMENT DESCRIPTION: Preparation of the Interim Report for the API Research Project MC-1 PREPARATION OF THE INTERIM REPORT FOR THE A PI RESEARCH PR O JE C T MC-1 A P R IL 10,1957 P relim inary Comment T his status rep o rt is being p rep ared so as to include details reserv ed hitherto for the annual rep o rt presen ted in Septem ber afte r the d o s e of the fiscal y e a r. In o rd e r to assem b le the la rg e m a ss of experim ental data in a useful docum ent, the fo rm norm ally followed in scientific jo u rn als is being used. An introductory statem ent of the background and broad objectives of the investigation is follow ed by a descrip tio n of the pursuit of an experim ental p ro g ram tow ard the specific goal of the analytical tool req u ired fo r a rap id estim ation of the h aza rd of skin can cer asso ciated with the handling of any type of hydrocarbon oil. The introduction is concluded with a very b rie f sum m ary of the m ain contributions m ade thus fa r to the achievem ent of the objectives as set fo rth , and to the p o rtray a l of the general p icture of the rela tio n ship betw een com position and carcinogenic potency as it is understood at this tim e. The experim ental m ethods em ployed in the cu rren t investigation and the re su lts of th e ir application to a v arie ty of petroleum oils a re discussed in detail in a se rie s of appendices. The p resen tatio n of the proposed p ro g ram fo r 1957-58 is divided into a statem ent of the gen eral p ro g ram fo r the fu rth e r developm ent of co rrelatio n s betw een the com position and potency of th ese oils and an outline of the specific a re a s to be attacked and the ex p erim en tal techniques to be used. It is hoped that this m ethod of form ulating the re p o rt will serv e the dual needs of the A m erican P etroleum Institute fo r a reasonably concise statem ent of the p ro g re ss tow ards the objectives of the investigation, on the one hand, and a detailed technical d iscu ssio n of the experim ental m ethods and re s u lts, on the other. INDEX Introduction - O bjectives and G eneral Sum m ary P age No. ^ P roposed P ro g ram for 1957-58 5 A ppendix 1 - E x p erim en tal M ethods U sed to D eterm ine the R elative C ontributions M ade by the C arcinogens and the M edium to the P otency of a C om plex Oil Section A. - T otal C arcinogens (1) Section B. - M easu rem en t of the R elative Spreading (5) T endencies of V arious M aterials Section C. - Individual C la sse s of C arcinogens (6) S ectio n D. - C a ta ly tic In flu en ce of the N o n -c a rc in o g e n ic (8) M edium Section E. - M aterials (9) F ig u re 1 - The Rate of Induction of P apillom as in C3H (10) M ice by the R epeated A pplication of 3-M ethyl- cholanthrene and Benzo(a)pyrene in Non acceleratin g Solvents Appendix 2 - R esults of E xperim ents Appendix 3 - D iscussion Section A. - R efined L u b rican ts and White O ils (2) S ectio n B. - U n refin ed an d P a r tia lly R efin ed L u b ricatin g (2) Stocks and Slack Waxes Section C. - S traigh t Run R esidua (3) S e c tio n D. - P r o d u c ts of T h e r m a l an d C a ta ly tic C ra c k in g (3) and D elayed Coking S ectio n E . - P r o g r e s s in the D evelopm ent of a Q u an titativ e (5) A nalytical Tool 1 i Section F. - P ro g re s s T ow ard the A nalytical E stim ation (10) of the Content of C arcinogens in any Given P etro leu m Oil I API 06578 A PI RESEARCH PR O JE C T MC-1 Introduction R epented contact w ith c e rta in petroleum oila is known to have produced cancer of the akin of m an under certain occupational condition!. Inv estig ato r! in various countries have gathered evidence of the existence in these oils of c e rta in types of aro m atic hydrocarbons of com plex s tru c tu re, w hich a re capable of inducing can cer of the skin of m ice when applied thereon under suitable experim ental conditions. Some y ears ago, fu rth er experim ental evidence fro m a num ber of sources indicated, o r at le a st suggested, that c e rta in products of m odern refining p ro cesses (e.g . catalytic cracking) in the p etro leu m in d u stry w ere capable of inducing experim ental cancer of the skin of m ice w ith considerable speed and u n i fo rm ity . This experim ental fact, w hile not alarm ing in itse lf, and w hile lacking any confirm ation of its p ra c tic a l significance in the te rm s of experience in the p etro leu m in d u stry , cam e to be reg ard ed , am ong responr sible people in the industry, a s the ju stificatio n for a fa irly com prehensive investigation of a w ide range of products and p ro c e sse s, from the asp ect of the potential h azard of such ca n c e r inducing substances as m ight be found th e re in . Such an in v estig atio n w as in itiated in 1946 in The K e tte r ing L aboratory, in the D epartm ent of P rev en tiv e M edicine and Industrial H ealth of the U niversity of C incinnati, under the financial sponsorship of the A m erican P etroleum Institute. The objectives of this investigation, as they w ere established initially, w ere three in num ber: 1. The f ir s t of th ese w as th a t of determ ining, on the b asis of the experim ental response of the skin of m ice to applications of specific and w ell identified petroleum pro d u cts, those which p o sse ssed ca rc in o genic capacity. It w as hoped, of co u rse, that such capacity could be estim ated in quantitative or, at least, com parative te rm s. -2- 2. The second objective w as that of developing an analytical m ethod or tool, o r possibly a se rie s of such tools, w hereby the carcin o genic capacity of various oils could be detected, and w hereby ultim ately, if possible, carcinogenic substances could be classified, identified, and assesse d quantitatively. 3. A th ird and cru cial objective w as that of ascertaining the extent of the co rre la tio n which m ight ex ist betw een the experim ental induction of cancer in the skin of m ice and the capacity of petroleum o ils to induce hum an cancer of the skin under occupational o r other environm ental conditions w hich m ight be found to have a ris e n h is to r i cally or currently in industry or elsew here. The f ir s t two of th e se o b jectiv es w ere to be p u rsu ed by the application, adaptation and refinem ent of biological and technological m ethods, w hile the th ird req u ire d the gathering of the available in fo r m ation from the lite ra tu re and from contem porary in v estig ato rs, as w ell as fro m the p ast and c u rre n t ex p erien ce of the p etro leu m in d u stry . The approach to the experience of the ind u stry contem plated an epide m iological investigation of extensive scope w hich will be the subject of a later report. In the background of the specific o b jectiv es which determ ined the p a tte rn of the ex p erim en tal w ork, w ere two g en eral purposes, (1) that of pin-pointing the foci and sev erity of the hazard of skin can cer, if such h aza rd existed in fact, ra th e r than a s a potentiality, w ithin the industry, on the one hand; and (2) that of revealing g en eral o r specific te c h n iq u e s four c o n tro llin g s u c h h a z a r d a s m ig h t b e found, w h e th e r by the m ethods of technology o r by those of in d u strial hygiene and p rev en tive m edicine. As the w ork in the L aboratory p ro g ressed , these p u r poses underw ent som e fu rth e r definition. Thus, out of the biological side of the w ork cam e indubitable evidence of the existence and the nature of facto rs w hich influenced significantly the ability of specific -3 - petro leu m oil* to induce can c er in the akin of m ice. M oreover out of the technological w ork cam e co rrelatio n s betw een the carcinogenic potenciea of oila (for the induction of can cera of the akin of m ice) and c e rta in p h y si cal and chem ical c h a ra c te ria tic a of theae oila that led to the identification of certain of th e ir com ponente a a being actually carcinogenic or capable of prom oting or inhibiting carcinogenic activity. It ia evident that auch experim ental reau lta, w hich w ill be deacribed in auitable detail la te r, may provide atrategic m eana for the control or perhapa the elim ination of c e rta in occupational h azard a, to the extent that the la tte r a re c o m p ar able in type or deg ree to the conditiona im posed experim entally upon the m ouse. Thereby some b asis for the selection of p ro cess or product may be afforded to the petroleum technologist, in accordance w ith his needs, purposes and judgm ents or, altern ativ ely , for the application of safeguards by the in d u strial hygienist. It ia of som e significance to recognize that a considerable body of inform ation, g athered by experience and experim ent over the y ears, h as rev ealed strong co rre la tio n s betw een hum an experience w ith the cancer-inducing p ro p erties of coal ta r, shale oil, and crude paraffinic d istilla te s, and the ex p erim en tal re s u lts of the application of theae m a te ria ls upon the skin of m ic e . It is n e c e ss a ry to m a in tain a carefu lly guarded c ritiq u e ag ain st too lite ra l an accep tan ce of the view that the hum an lesio n s of th is type th at have o c c u rre d can be rep ro d u ced in the m ouse under p recisely p a ra lle l conditions, or vice v e rsa . N ever-thele ss , the use of th ese anim als in following the objectives of this and other com parable investigations has been ju stified fully, and it is fair to say that the co rre la tio n s betw een hum an experience and anim al e x p e ri m ents in this field have been strengthened by im provem ents in the design and technique of ex p erim en ts. Thus one tends to assu m e, w ith som e reserv atio n of judgm ent, that the m ore rapidly an oil w ill produce can c e r of the skin of m ice, the m o re activ e it w ill be upon the skin of m en. API 06581 -4- On the foregoing background, the statu s of the in v estig ativ e p ro g ram m ay be sketched b riefly in the following m anner: The ra te at w hich tu m o rs of the skin of the m ouse can be induced by a given oil depends b asically upon the concentration of c e r tain active carcinogens which a re polycyclic arom atic hydrocarbons. The basic rate m ay be in creased as m uch as five-fold by the influence of the other com ponents of the oil even though the la tte r a re com pletely non-carcinogenic when freed of the polycyclic aro m atic fraction. This catalytic activity of the m edium is contributed in p a rt by alkyl d eriv a tives of the low er a ro m a tic s (naphthalenes and benzenes) as w ell a s by satu rated com pounds, such as the n o rm al p araffin s. On the other hand, the effect of th ese " a c c e le ra to rs" m ay be counteracted alm o st com pletely by other satu rated hydrocarbons (the polycycloparaffins) which in som e m anner inhibit the cataly sis and lim it the potency of the oil to a level little higher than th at due to the carcin o g en s. A few y e a rs ago re s e a rc h in the a re a of satu rated hydro carbons of high m olecular w eight w as extrem ely difficult. However, the developm ent of high m a ss sp ec tro m e try during the p a st four y e a rs h as m ade it p ossible to obtain useful inform ation on the com po sition of satu rated fractio n s of refin e ry stre a m s., This technique of analysis has also played an im portant ro le, together w ith absorption spectrophotom etry, in the co n cu rren t investigation of the aro m atic c a r c i.n o g e n s . 1 1. T he r e s e a r c h d e p a rtm e n ts of a n u m b e r of th e m a jo r oil co m p an ies furnished the m ass sp ectra req u ire d in this p ro ject. PROPOSED PROGRAM FOR 1957-58 In approaching the goal of an analytical tool for prediction of the ra te a t which an oil w ill induce tu m o rs of the skin, two p a ra lle l and closely rela ted paths a re being followed. The firs t is the in v esti gation of the identity of the polycyclic carcinogens, which m u st be p resen t if the oil is to have any cancer-inducing activity w hatsoever. The second path is that leading to an understanding in quantitative te rm s of the relatio n sh ip s betw een the com position of the m edium in which the a ro m a tic carcinogens a re d isso lv ed and the extent to which it is capable of effecting an acceleratio n of the rate of the developm ent of tum ors. API 06583 -6- The three m ajor classes of carcinogens m ay be described as the 5 -rin g ed , the 4- (and 4 1/2) ringed, and the 3 1 /2 -rin g ed com pounds. The developm ent of an analytical m ethod fo r the 5-ringed benzopyrenes com pleted the w ork on the firs t class. C onsiderable p ro g re ss has been m ade during the p a st y ear tow ards the ch aracterizatio n of the 3 1 /2 -rin g ed class (those p o ly cy d ics b a s e d upon a n u c le u s h av in g 16 o r 17 c a rb o n a to m s ). Two a d d i tional cycles of biological te sts (about 9 m onths for each) are estim ated to be sufficient to com plete the identification of this c la ss. The attention that is being devoted to the 3 l/2 -rin g e d group of carcinogens is ju stified by the evidence that alm ost all petroleum oils that a re carcinogenic contain significant quantities of this group. A m ajo r p a rt of the 4 - and 4 1 /2 -rin g ed carcinogens m ay be ex tra cted fro m an oil with m aleic anhydride. Such e x tra c ts from a v ariety of oils a re being tested to determ ine what p a rt of the basic potency they contribute. D uring the com ing y e a r the experim ental w ork on the second im p o rtan t a sp ec t of th is p roblem (the cataly tic activity of the m edium ) w ill be concerned with the developm ent of quantitative data that define the influence of the polycycloparaffins w hich se e m s to in te rfe re with catalytic activity. A technical white oil th at w as shown to have this "inhibitory" p ro p erty has been fractionated to concentrate the su sp ected m a te ria ls. T hese w ill be te ste d in synthetic blends API 06584 -7with a lo n g -ch ain a c c e le ra to r and a stan d ard carcin o g en . M ass sp e c tro m e tric an aly sis w ill be applied in p a ra lle l to a v a rie ty of fractio n s alread y te ste d fo r th e ir cataly tic activ ity , in an effort to quantitate these relationships. API 06585 T APPENDIX 1 EXPERIM EN TA L METHODS USED TO DETERMINE THE RELATIVE CONTRIBUTIONS MADE BY THE CARCINOGENS AND THE MEDIUM TO THE POTENCY OF A COM PLEX OIL A. Total Carcinogens The te rm carcinogen is applied to specific types of poly cyclic aro m atic com ponents of petroleum oils known o r suspected to have cancer-inducing p ro p e rtie s. The following stru c tu re s indicate im portant types in cracked oils: 3 1 /2 -ringed 4 and 4 1/2-ringed 5 -ringed Ben (a)an th racen es known B cn zo (a)p y ren c known Such compounds are easily separated from m ore satu rated com ponents having le ss than 3 rings p er m olecule by chrom atography on alum ina (Alcoa F -2 0 ). The polycyclic fractio n is then blended into a n o n -acceleratin g m edium (benzene fo r fastest resu lts) at a selected level of concentration. The rate at which the blend w ill induce pap illo m as of the skin of C3H m ice is determ ined in the usual m anner. API 06587 A PPENDIX 1 Pag (2) Since the ra te a t which a polycyclic carcinogen w ill induce tu m o rs is d irectly re la te d to the dosage p e r unit a re a of the skin (see F ig u re 1), the d iffe re n c e ip th e a r e a c o v e re d by eq u al q u an tities of th e blend in bensene and the original oil m u st be taken into account. (The m ethod u se d to m e a su re the re la tiv e ten d en cies of d ifferen t m a te ria ls to sp read upon the skin of C3H m ice is d escrib ed in section B of this ap p en d ix .) The spreading tendencies of the oils which have been fractio n ated in the m anner described above have not been m easured as yet, but reasonable estim ates of the relative tendencies m ay be m ade fo r im m ediate purposes. The ra te a t which the o rig in al oil would have induced papillom as if the m edium had had no catalytic influence is estim ated by correcting the ra te at which the blend of the po ly cy clics in benzene acted upon the skin. The estim ated ra te (sym bolized by e^) is slow er than the ra te fo r the blend in benzene because the blend usually u sed has a higher concentration of the carcinogens than the original oil. M oreover, the blend in benzene yields a la rg e r dose of th e carcinogen p er unit a re a of the skin because it sp read s to a le s s e r extent than the original oil. To obtain e re fe re n c e is m ade to the d o se-resp o n se curves for the stan d ard carcinogens, benzopyrene and m ethylcholanthrene, in relativ ely non-volatile, n o n -acceleratin g solvents, such as white m in e ra l oil (340 S. S. U. a t 100), decahydronaphthalene, o r s e c . a m y lb e n z e n e (F ig u r e 1). It is a s s u m e d f o r th e p u rp o s e of th is e s tim a te that the corresponding curves fo r the carcinogens of the oils have slopes interm ediate betw een those of m ethylcholanthrene and benzopyrene. API 06588 A PPEN D IX 1 P ag e (3) The curve fo r such a hypothetical re fe re n c e carcin o g en is shown on T ab le 1. The following exam ple should illu s tra te the use of the relationships discu ssed above in arriv in g at the value of e : The polycyclic aro m atic fractio n of oil A PI-89 is obtained in 19 p e r c e n t y ie ld by a c h ro m a to g ra p h ic fr a c tio n a tio n . A b le n d c o n ta in ing 40 p erce n t of th is fractio n and 60 p e rc e n t of benzene (by weight) is applied rep eated ly upon the skin of m ice (th ree 20 m g. ap p licatio n s each week). The av erag e duration of tu m o r-fre e life p ro v es to be 27. 1 w eeks. F rom the relationships in F igure 1 it is noted that this rate of resp o n se co rresp o n d s to that obtained with approxim ately 0. 54 m g. of the re fe re n c e carcin o g en (applied p e r unit a r e a of the skin each w eek , th e u n it a r e a b ein g d efin e d a s th a t c o v e re d by 100 m g . of th e referen ce solutions having a rela tiv e sp read equal to one). The o rig in al oil h as only on e-h alf as m uch carcinogen as the blend. F u rth e r, it is found that its tendency to sp re a d over the skin is approxim ately 1.6 tim es that of the blend. T h ere fo re the d o se , --do , of c a rc in o g e n p e r u n it a r e a w hen th e o rig in a l o il is a p p lie d 0. 54 is e q u iv a le n t to 2~x"T ~ 6 ' r ^ m 8* * t*le r e *e r e n c e c a rc in o g e n s p e r w eek. T his ra te of ap p licatio n of th is dosage would re s u lt in the in d u c tio n of p a p illo m a s in a p p r o x im a te ly 43 w eek s (fro m F ig u r e 1), w hich i s th e v a lu e of th e e s tim a te d e ffe c t, --ec , of o il A P I-8 9 if it h a s no catalytic influence on the activity of its polycyclic carcinogens. API 06589 A PPEN D IX 1 Page (4) The total content of carcinogens Crc (ex p ressed as the equivalent percentage of the re feren c e carcinogen) also m ay be derived fro m the value of 6^ , using the following relationship: d^^C rc ,1) lo w here f is the num ber of applications p e r week (three) and s -- --o the rela tiv e sp re a d of the oil. In the case of oil A P I-89, s --o is e s tim a te d to b e 1. 3; th e r e f o r e , th e c o n te n t of c a rc in o g e n s , C rc 0. 16 3 1. 3 , is equivalent to 0.07 p e rc e n t of the re fe re n c e carcinogen. A lte rn a tiv e ly C r c an d e^. m a y be c a lc u la te d directly from the equation for the dose-resp o n se curve of the reference carcinogen. d 10 e - 13 As above, w here d^ is the rate of dosage of carcinogen obtained when the blend of the polycyclics in benzene was applied upon the skin (0. 54 m g. p e r unit a re a p e r week in the case of oil A PI-89) resulting in an average d u ra tio n of tu m o r - f r e e lif e , --e",O; -s,O a n d --s a r e th e estim ated relative sp read of the blend and the original o il, r e s p e c tiv e ly ; a n d c^ a n d c^j a r e th e c o n c e n tra tio n of the polycyclic fractio n in the blend and the original oil, respectively. C om bining eq u atio n s (1) an d (3), I Cr c r *b ' *b b API 06590 A PPEN D IX 1 Page () an d su b stitu tin g th e e x p re s s io n fo r d of equation (2) fo r d^i the resulting equation fo r the calculation of Crc becom es: , (4) Since e is the average duration of tu m o r-fre e life resulting from a rate of dosage, d , these factors a re also rela ted by equation (2). T h erefo re, an expression fo r m ay be derived by the com bination of eq u atio n s (1) an d (2), a s follow s: e (5) the value of Crc being obtained from the calculation of equation (4). In the case of the o il, A P I-89, 0. 07 and e B. M easu rem en t of the R elativ e S preading T en dencies of V arious M aterials A d irec t m easu rem en t of this facto r is being m ade in selected cases by the following p ro ced u re: A m e a s u re d quantity of the m a te ria l (2-10/Ul. ) is applied upon the skin of each of 20 m ice (during the r e s t ing p hase of the h a ir cycle), w hich have been a n esth e tiz ed with carbon dioxide to im m obilize them . A fter in terv als of 30 an d 60 m in u te s , th e a n im a ls a r e ag a in a n e s th e tiz e d . E ach is then placed under a g la ss slide and the apparent a re a of distribution of the polycyclic hydrocarbon is tra c e d on the Blide by its flu o re sc e n c e u n d e r an u ltra v io le t lam p. Several tracings of each spot are m ade, m easu red with a p lan im eter, and the av erag e a re a calcu lated . F rom the average for the group of m ice u sed in a given te st and the quantity of solution ap p lied to each, the av era g e a re a of skin covered per^u l. of solution is calculated. API 06591 A PPEN D IX 1 P ag e (6) The absolute values obtained vary from one opera to r to an o th er, depending in p a rt upon the experience in observing the flu o rescen ce of the a ro m a tic hydrocarbon. H ence, an in stru m e n ta l pick-up of th is rad iatio n would be n e c e ssa ry to obtain significant absolute values. H ow ever, it is the relativ e spreading of the different m a te ria ls that is the im portant consideration in th is application. By use of standard referen c e solutions, the relativ e values obtained by two ex p erien ced o p e ra to rs a re in good ag reem en t. C. Individual C lasses of C arcinogens It has proved to be useful to co n sid er the b asic (non-accelerated) carcinogenic potency of an oil as a sum m ation of the concentrations of th re e individual c la s s e s of carcin o g en s (those in d icated in sectio n A of this appendix) when each concentration is ex p ressed in equivalent w eights of a hypothetical referen ce carcinogen having a d o se-resp o n se curve interm ediate between that of m ethylcholanthrene and benzopyrene. The content of the fiv e-rin g ed benzopyrenes is determ ined by an analytical procedure developed in an e a rlie r period of this in v esti gation^. The equivalent value of the referen ce carcinogen is obtained from the d o se-resp o n se cu rv es in F ig u re 1 (the value is essen tially equal to the content of b en zo p y ren es u n le ss th e la tte r ex ceed s 0.10%; for higher v alues, the rate of dosage p er unit a re a m ust be taken into account). The available evidence indicates that m ost of the 4- (and 4 1/2-) ringed carcinogens a re derivatives of benz(a)anthracene and m ay be ex tra cted from an oil by form ation of an adduct with m aleic anhydride. The carcinogens m ay be concentrated fu rth er by chrom atographic fra c tionation, which elim in ates m o st of the 3 -ringed non-carcinogens (determ ined by absorption and m a ss sp ectra). The yield of the API 06592 APPENDIX 1 P ag e (7) 4 - r i n g com pounds is noted and the fractio n is then blended with benzene at a definite level of concentration. The ra te a t which the blend w ill induce papillom as in the skin of C3H m ice is then determ ined and ex p ressed in equivalents of the referen ce carcinogen. In o rd e r to a rriv e a t the contribution of th is cla ss of carcinogens to the potency of the o rig in al oil, th is value m u st be c o rre c te d in the m an n er d escrib e d prev io u sly with re fe re n c e to the total carcinogens (Section A, page 2). In attacking the classificatio n of the 4- and 5 -rin g ed c a rc in o gens it has been p ossible to p roceed with a considerable background of published knowledge on the carcinogenic potencies of pure com pounds. In c o n tra st, only a very few of the m eth y l- o r dim ethyl- d eriv ativ es of the 3 1 /2 -rin g ed cla ss have been te ste d fo r th e ir potency, and none of these w ere rep o rte d to be active. Since it is apparent that a sig n i ficant p a rt, and in som e cases a m a jo r p a rt, of the potency of p etro leu m oils is due to som e com pounds in th is g en eral c la ss it is e ssen tial that th ese carcinogens be identified to p e rm it the develop m ent of adequate analytical m ethods. T herefore a system atic approach to the problem is being m ade. By fractio n al d istillatio n , ch ro m ato graphy, and ex tractio n with cold, co n cen trated sulfuric acid , the various 3 1/2-ringed classes, fluoranthenes, cyclopentanophenanthrenes, pyrenes, and benzofluorenes, have been sep arated and a re being tested to determ in e th e ir rela tiv e contributions to the potency of the fractio n distilling fro m 70-760 F . of the cataly tically crack ed oil, A PI-8. API 06593 A PPEN D IX 1 P ag e (8) D. The C atalytic Influence of the N on-carcinogenic M edium By the "m edium " is m eant the com bination of all of the non-carcinogenic com ponents of an oil. The influence of the m edium is d e te rm in e d by c o m p a rin g c> th e r a te of tu m o r in d u c tio n ex p e c te d on the b a sis of the carcinogens p re se n t if the m edium had no catalytic activity (as determ ined by the m ethod discu ssed in section A of this appendix), with the observed ra te of induction of tu m o rs by the oil. Thus the re la tiv e catalytic activ ity , A, is e x p re sse d sim ply as the ratio betw een the expected and observed duration of tu m o r-fre e life: .lobs. (6) T hus, in the exam ple at the end of section A of this appendix the expected duration of tu m o r-fre e life fo r oil A PI-89 was 43 w eeks if th ere was no cataly tic contribution to its activ ity . A ctually the o rig in al oil induced tu m o rs quite rap id ly (e . * 12. 3 w eeks). The re la tiv e ' obSi catalytic activity of the m edium of this oil is obtained from the calculation: A = L = 3. 5 . 12. 3 Solutions of 0. 05 p e rc e n t m eth y lch o lan th ren e in the non carcinogenic fraction sep arated chrom atographically from the polycyclic aro m atics a re being applied to m ice. F rom these te sts is determ ined the extent to which this fra c tio n , in the ab sen ce of c e rta in m o re strongly absorbed non-carcinogens, is capable of accelerating the rate of carcinogenesis. API 06594 J y I A PPEN D IX 1 P ag e (9) E. M aterial The m ethods d iscu ssed in the previous sections have been applied to a sp ectru m of the oils which w ere te sted fo r th e ir effective carcinogenic potency in an e a rlie r p erio d of this investigation. Sam ples have been included from Straight Run D istillation, T herm al C racking, C atalytic C racking, and D elayed Coking. An effort has been m ade to cover a sufficient variety of oils to be su re of the general applicability of analytical m ethods for estim ating carcinogenic potency developed from this investigation. API 06595 APPENDIX 1 Pag* (10) API 06596 FIGUBK 1 THE RATE OF INDUCTION OF PAPILLOMAS IN C3H MICE BY THE REPEATED APPLICATION OF 3 -METHYICHOLANTORKNE AND BENZO(a)PYRENE IN NON-ACCKLJ3RATING SOLVENTS d , RATE OF DOSAGE OF POLYCYCLIC CARCINOGEN ( a . p e r " u n i t a r e a " p e r w e e k ) APPENDIX 2 RESULTS A com parison of oils from various refining operations according to the relativ e ra te s a t which they induced tu m o rs of the skin of C3H and CFW m ice is shown in Table 1 (the m e asu re m en ts involved rep eated applications without w ashing). The ra te s are ex p ressed as th e ir re cip ro cals, i. e. , the fractio n s of the average life sp an of th e m ic e ('vSS w eeks) r e q u ir e d to induce p a p illo m a s. TABLE 1 NUMBER OF OILS 11 2 2 REFINERY OPERATION S tr a i g h t Run D i s t i l l a t i o n H If It ! It tt PRODUCT FROM FRACTIONATOR AVERAGE DURATION OF TUMOR-FREE LIFE, e --o b s . (weeks) Lube D i s t i ll a t e 15-81 FRACTION OF LIFE SPAN OF MICE TO INDUCE TUMORS1 < o b s/85> 0.2 - 1.0 G asoline (>85) > 1 .0 Residuum 70-(>85) 0.9 - 1.0 7 Solvent E x tractio n E xtract 11-78 0.13- 0.9 2 L ubricant M anufacture F in ish ed Lub 31-(>85) 0.4 - 1.0 ricatin g o il 3 W hite O il M anufacture W hite M ineral (>85) O il > 1 .0 6 Thermal C racking of V irg in Gas O il Residuum 15-53 0.2 - 0.6 6 Thermal Reforming Residuum 22-27 0 .2 6 -0 .3 2 12 Thermal C racking of C a ta ly tic Gas O il Residuum 9-34 0.1 - 0.4 10 C o k in g 11 C a ta ly tic C racking 27 C ycle Gas O il High B o llin g S id e stre a m B o ttom s 10-30 10-43 5-36 0.1 - 0.35 0.1 - 0.5 0 .06- 0.4 A PPEN D IX 2 P age (2) The polycyclic arom atic fractions w ere separated from a num ber of th ese o ils. In Table 2 a re shown the estim ated ra te s , e , at --c which the carcinogens in th e se fra c tio n s would induce tu m o rs if applied upon the skin of m ice in a non -acceleratin g solvent (at a concentration chosen so that the dosage applied p er unit are a p er week was the sam e as that obtained when the o rig in al oil w as applied). The catalytic a c tiv ity of th e to ta l m e d iu m , A, is c a lc u la te d a s th e r a t i o _____ .fobs. (a value of A - 2 in d ic a te s th a t the ra te w as doubled). In ad d itio n , the catalytic activ ity , a, of the fractio n of the oil sep arated from the polycyclics was determ ined. TABU! 2 API SAMPLE NUMBER DESCRIPTION OF ORIGINAL OIL 89 FCC Heavy Gas O il OBSERVED DURATION OF TUMOR-FREE LIFE FOR TOTAL O IL ,1 fo b s. (weeks) ESTIMATED RELATIVE DURATION CATALYTIC DUS TO ACTIVITY CARCINOGENS OF alone/ TOTAL ec MEDIUM, (w eeks) A RELATIVE CATALYTIC ACTIVITY OF PART OF THE MEDIUM,2 a 1 2 .3 1 1 .3 43 3.5 2 .3 33 C oker Gas O il 1 0 .2 1 1 .0 36 3.5 2.4 120 Lube S to ck (n a p h th e n ic ) 115 It 3 3 .2 1 3 .4 S56X ^443 /V I. 3 < 1 .4 ^/1 .5 105 Lube S to c k 2 3 .2 1 4 .4 2 .0 (paraffin ic) 65 T h e ra a l T ar ^ 2 IX 26 ^ 1 .2 ^ 2 .0 1. F or t e s t s In v o lv in g th re e 50 o r 100 a g . a p p lic a tio n s each week. 2 . That p art containing le s s than 3 rin g s per ao lec u le (te ste d as the s o lv e n t in a s o lu tio n of 0.05% a e th y lc h o la n th r e n e ) . 3 . The e x p e ria e n ts fro a which th e v a lu e s preceded by r j are d eriv e d a re not ceapleted. The values derived fo r e^, A, and a in Table 2 should be reg ard ed as having relativ e, but not absolute significance since the correction A PPEN D IX 2 Pag (3) facto r* which m u st be ap p lied fo r the ip rea d in g c h a ra c te ristic * of the different m aterial* w ere estim ates. M easurem ents of these spreading c h a ra c te ristic s a re being m ade in the c u rre n t p rogram . Inform ation th at h as been obtained (by m a ss sp e c tro m e tric an aly sis) on the com position of c e rta in n o n -carcin o g en ic chro m ato g rap hic fra c tio n s of the o ils is shown in T ab le 3. D ata a r e a ls o in clu d ed on the com position of the tech n ica l w hite oil A P I-80 (95 SSU a t 100), which p rev io u sly h as been shown to contain acc e le ra tin g fra c tio n s, on the one hand, and com ponents capable of counteracting acc elera tio n , on the other. TABLE 3 PERCENT OF CLASS IN NON-CARCINOGENIC FRACTIONS OF CLASS OF HYDROCARBONS P a r a f f i n s ( n - fc I s o - ) API -89 1 + 2 + 3 1 1 (e stl1 2 m ated) 53 5 38 API -120 A P I- 65 API-80 1 + 2 1 + 2 + 3 + 3 1 1 (e stl- 2 1 1 (e stl- 2 1 2 mated) 1 2 m ated) 40 3 42 21 14 12 A lkylcyclopentanes 22 ( ) 3 18 34 ( ) 3 24 ( )3 28 A Ik y 1c y c 1o h ex a n es 8 29 8 15 0 10 Jb 11 VJ s 23 Condensed C yclop a ra ffin s A lkylbenzenes 15 ( ) 3 14 2 42 5 43 (8) 35 4 84 25 20 (21) 20 33 3 32 15 3 A lkylnaphthalenes - 23 12 - 8 3 1 15 25 1 3-rlnged A rom atics -- 5 - - - - - 11 - Percent of fractio n In o rig in a l o il 61 8 81 45 22 67 2 .5 27 44 (100) 1. F ra c tio n 1, most e a s ily e lu te d ; fra c tio n 2, next most e a s ily e lu te d . 2 . T e ste d as a s o lv e n t f o r 0.05% n e th y lc h o la n th r e n e . 3. V alues fo r cyclopentanes and condensed c y c lo p a ra ffin s are u n c e rta in In the fra c tio n s In d icated due to In te rfe re n c e fro a the arom atics. As indicated by its an aly sis the th e rm al ta r A PI-65 h as a com position that is som ew hat unique. In ch rom atography v ery little of the oil could API 06599 A PPEN D IX 2 Page (4) TABLE 4 - THE CATALYTIC ACTIVITY 0 ? SOLUTIONS OF REFERENCE CARCINOGENS API EXPERI- KENT NUMBER SOLVENT SCHEDULE . OF CONCEN TRATION OF APPLICATION OF SOLUTION CARCIN A p p li > 8 . RELATIVE OGEN c a t i o n s P e r SPREAD % MC Per A ppli OF lo b a . le OR BP1 Week c a t i o n SOLUTION (w eeks) [weeks) A 204a Benzene 0 .1 7 MC 3 100 0 .5 1 5 .7 1 1 .6 18 1 .1 222 sec-A y 1benzene 0 .1 8 MC 3 100 1.0 2 6 .7 2 .7 25 0 .9 249a 249b t-D odecyl benzene2 tl 0 .1 7 MC 3 0 .1 7 MC 3 100 1.4 20 1.4 8 .8 1 0 .9 30 3 .4 9 .9 1 0 .9 30 3 .0 296 W hite O il I I 0 .2 0 BP 3 (340S S U /100) 299 tr 0 .0 8 BP 3 602 n -O ctan e 0 .2 1 BP 3 278 n-Decane 0 .2 1 BP 3 295 n-Dodecane 666 tt 665 M 0 .2 0 BP 3 0 .0 8 BP 2 0 .0 4 BP 2 50 1 .0 2 8 .0 1 3 .4 29 1 .0 5 1.0 3 7 .3 1 2 .0 38 1 .0 100 0 .8 e s t . 2 7 .1 1 1 .8 27 1 .0 50 1 .4 e s t . 1 6 .6 1 1 .2 31 1 .9 50 1 .6 1 4 .6 1 1 .1 33 2 .3 50 1 .6 2 0 .4 1 2 .2 50 2 .5 50 1 .6 2 3 .7 1 1 .8 60 2 .5 661 20% Dodecane 0 .0 8 BP 3 100 0 .8 e s t . 2 4 .0 1 1 .5 35 1 .5 80% B enzene 659 50% Dodecane 0 .0 8 BP 2 50% D e c a lin 50 1 .3 e s t . 2 0 .0 1 2 .6 48 2 .4 660 50% D odecane 0 .0 8 BP 2 50% b - a e t h y l - naphthalene 50 1 .3 e s t . 19.113.1 48 2 .5 297 50% D odecane 0.20 IP 3 50% W h ite O il I I 50 1 .3 e s t . 2 0 .6 1 2 .2 31 1 .5 614 75% D odecane 0 .2 0 BP 3 25% W hite O il I I 50 1 .4 e s t . 2 0 .4 1 1 .3 31.5 1 .5 258 API-8 0 , W hite 0 .2 3 BP 1 O i l I (95SSU/100 B) 50 1 .3 35.1+ 1.9 42 1 . 2 1 . M e th y lc h o la n th re n e (MC) o r b en z o p y ren e (B P). 2 . Product of alk y la tio n of benzene w ith the te tra a e r of propylene. API 06600 APPENDIX 2 Page (5) be rem o v ed fro m alu m in a b efo re com pounds containing one o r two arom atic rings appeared. , In T able 4 is shown the cataly tic activ ity of v ario u s solutions of referen c e carcinogens. This data w ill facilitate the in terp reta tio n of the cataly tic activ ity of the com plex oils in Tahle 2 and 3 in te rm s of their com position. In the following p ages, the re su lts of the investigation of individual classes of polycyclic carcinogens from various oils are p resen ted . In Table 5 is shown the proportion of the total carcinogens rep resen ted by benzopyrenes, and the 4 - ringed benzanthracenes. API SAMPLE NUMBER DESCRIPTION 89 FCC Heavy Gas O il TABLE 5 TOTAL CONTENT12 OF CARCINOGENS, <C%r)c 0.07 CONTENT1 OF 4 -RINGED CARCINOGENS3 (%) 0.007 CONTENT OF BENZO PYRENES <%> 0.001 33 C oker Gas O il 0.11 0.07 0.008 65 T herm al T a r 0.20 , 0.04 0.05 120 Lube S to ck (n a p h th e n ic ) 0.05 0.004 =0.003 8 FCC C l a r i f i e d O il 44 FCC C l a r i f i e d O il 45 T ar f r o a th e rm a l c ra c k in g o f API-44 0.6 e s t. _3 _3 0.10 Incom plete incom plete 0.42 0.15 0 .3 8 1. Expressed in q u iv alen ts of the referen ce carcinogen of F igure 1. 2 . E x tra c te d from th e o i l by fo rm atio n of adduct w ith a a le ic anhydride. 3. Not determ ined to d a te . API 06601 A PPEN D IX 2 Page (6) The investigation of the 3 1 /2 -rin g ed class of carcinogens (by a p p ro p ria te fra c tio n a tio n of the FCC c la rifie d o il, A PI-8) has n arro w ed the field to th ree groups of hy d ro carb o n s, fluoranthenes, cyclopentanophenanthrenes, and pyrenes, the la st being the le a st likely to be significantly carcinogenic. The benzofluorenes, which have been se p arated from the above by chrom atography, apparently contribute only a sm all fractio n of the potency due to th is c la ss as a w hole. Pyrene B en zo (a)flu o ren e The pyrenes have been ex tracted on a sm all scale from the fraction containing the f i r s t th re e g ro u p s by cold c o n c e n tra te d (90%) su lfu ric acid. Some lo ss due to sulfonation o ccu rs; th e re fo re , the use of liquid hydrogen fluoride for this purpose will be investigated. M ass and absorption sp ectra have shown that the raffinate from this acid extraction contains about equal am ounts of flu o r anthenes and cyclopentanophenanthrenes to g eth er with a sm all am ount of a group having m olecular w eights of 220 and 234. Initially it was supposed th at the m e m b e rs of th is la s t group w ere d e riv a tiv e s of p h en a n th re n e (3 o r 4 alk y l c a rb o n s ), but th e ir c h ro m a to g ra p h ic APPENDIX 2 Page (7) behaviour m akes this assum ption questionable. F o r a ll groups in the raffinate (from the m o st carcinogenic fractio n of A PI-8 in the d istillatio n range 727 to 759F. ) the 18-carbon atom d eriv a tiv e s predom inate (in the case of fluoranthene these would be dim ethyl or ethyl). The separation of the fluoranthenes from the cyclopentanophenanthrenes by use of chrom atography and picric acid has proved to be feasib le. T hese groups w ill be te ste d se p a ra te ly on m ice to determ ine th e ir respective potencies. API 06603 APPENDIX 3 DISCUSSION The evidence developed in the course of this investigation indicates that an analytical estim atio n of the rate a t which a given oil m ay be expected to induce tu m o rs in the skin of m ice m u st take into account two d ifferen t types of in flu en ces. The b asic and e sse n tia l one . is that of the polycyclic arom atic carcinogens. As th eir concentration in c re a se s, so is the rate of induction of tum ors expected to in crease. An augm entation of th is effect m ay , and freq u en tly does, occur if the rem aining n o n -carcin o g en ic portio n of the oil, i. e. , the "m edium " in which the carcin o g en s a re dissolved, is of such com position a s to act as a cataly st in the m echanism of carc in o g en esis. Although the nature of this m echanism is not understood, considerable p ro g ress has been m ade in describing the types of com ponents which ex e rt this accelerating effect. It has been found to be u seful, fo r the purpose of orienting the collection of ex p erim en tal in fo rm atio n , to ex p re ss the contributions of the carcin o g en s and of the m edium as a sim ple equation which e x p re s s e s ra te . T he ra te a t w hich a given oil m ay be ex p ected to induce tu m o rs, r , , --o b s. r = Ar , (7) ~obs. ~c w here A is the facto r due to the catalytic influence of the m edium , and the rate , r , is that at which the oil would produce papillom as if its --c com position w ere such that A equals one. The ra te s at which c e rta in c la sse s of p etro leu m oils induce can cer of the skin of m ice is dealt w ith sem iquantitatively in the API 06604 A P P E N D IX 3 P a g e (2) following sectio n s in te rm s of the g en era l re la tio n sh ip s e x p re sse d by equation (7). In a subsequent section the p ro g re ss tow ard a quantitative analytical tool will be d iscussed in detail. A. R efined L u b rican ts and W hite O ils A few oils of the types indicated by this heading have been tested and found to be com pletely non-carcinogenic. Solvent extraction of lu b ricatin g stocks ap p ea rs to be p a rtic u la rly effective in elim inating the polycyclic aro m atic carcinogens . B. U nrefined and P a rtia lly R efined L ubricating Stocks and Slack W axes When m a te ria ls of th is c la ss a re applied rep eated ly upon the skin of m ice, the rate of induction of tu m o rs v arie s over an ex tre m ely wide ran g e, depending upon the com position of the sam ple. Among the unrefined stocks (having a given d istillatio n range) th e re is as m uch a s a th re e -fo ld v a ria tio n in the. content of polycyclic carcin o g en s. The ra tio of naphthenes to p a ra ffin s a lso a p p e a rs to be of m a jo r im portance. It seem s probable that the naphthenes (poly cycloparaffins) of high m o lecu lar w eight co u n teract m o st of the catalytic effect of any long-chain a c c e le ra to rs p re se n t in the oil (see A ppendix 2). The two p araffin ic d is tilla te s which have been te ste d p ro duced tu m o rs fa irly rap id ly (in 4 to 6 m onths on the av erag e). Although the content of polycyclic a ro m a tic s in these oils is relativ ely API 06605 T A P P E N D IX 3 P a g e (3) low, the ratio of long-chain a c c e le ra to rs to polycycloparaffins is probably high. The p re s s oil from , dewaxing of one of th e se d istilla te s showed the sam e level of activ ity a s the o rig in al oil. The available evidence indicates that solvent extraction of such stocks produces an oil having little o r no carcinogenic activ ity . It is u n n ece ssary to c a rry the refining to the stage of a white oil to elim inate this p ro p erty effectively. C. Straight Run Residua In all likelihood the stra ig h t run naphthenic resid u a which we have tested contained the sam e types of carcinogens, in somewhat low er concentration, as w ere p resen t in the raw lubricating stocks. - H owever, the potencies of these resid u a in g en eral have been negligible. The possible presence of inhibitors has been considered in an e a rlie r In te rim R eport (S eptem ber 24, 1952). The re c e n t findings indicate that the ap p aren t "inhibition" m ay be due to the dilution of a c c e le ra to rs by polycycloparaffins. D. P ro d u cts of T h erm al and C ataly tic C racking and D elayed Coking Wide v aria tio n s have been found in the potencies of the p ro d u cts of the crack in g of heav y o ils by th e se p r o c e s s e s (0. 06 to 0. 6 of a life span re q u ire d to induce p a p illo m a s in the skin of m ice). With one exception, the ta rs produced by the th e rm al cracking p ro c e ss have failed to induce tu m o rs in le s s than 20 w eeks, on the a v e ra g e (so long as the stock charged to the unit was f r e e of cataly tically crack ed __ com ponents). In com paring the effect of th e rm a l cracking with that API 06606 A PPENDIX 3 Page (4) of catalytic cracking it is im portant to note the rela tiv e fate of the paraffin s and the polycycloparaffins in the ch arg e. Although both c la sse s a re decom posed m uch m o re rapidly in the catalytic than in the th erm al p ro c e ss, the in c re a se in rate fo r the naphthenes is m uch g re a te r than the in c re a se in ra te fo r the p araffin s (2000 tim es v e rsu s 50 tim e s , re sp e c tiv e ly ). * T h u s, the c a ta ly tic o p eratio n ten d s to in c re a se the ratio of a c c e le ra to rs to those com pounds su sp ected of in terferin g with acceleratio n , w h ereas the th e rm al p ro c e ss does not. C atalytic cracking also builds up the concentration of polycyclic carcinogens som ew hat m o re rapidly than does the th erm al p ro cess. Hence the pro d u cts of catalytic cracking generally will have higher potency than ta rs from therm al cracking. The one exceptionally active therm al ta r m entioned above was produced from a charge stock which w as relativ ely fre e of naphthenes (P ennsylvania crude). C atalytic cracking usually provides an effective build-up of ca rc in o g e n s to g e th e r w ith an in c re a s e in-'the c a ta ly tic a c tiv ity of the m edium to the point w hereby the p roduct, in the extrem e c a se s, m a y in d u ce tu m o r s of th e sk in of m ic e in l e s s th a n 10 w e e k s. M o re severe catalytic cracking m ay re su lt eventually in com plete destruction of all long-chain com pounds, reducing the catalytic activ ity of the m edium to a low lev el. U sually the concentration of highly active c a rc in o g e n s in su ch o ils h a s r e a c h e d v e ry high le v e ls ( 7 O. 5 p e rc e n t). Hence such m a te ria ls will still induce tu m o rs alm ost as rapidly as those produced by m ild er cracking conditions. 1. B .S . G re e n s fe ld e r, et a l, In d u s tria l an d E n g in eerin g C h e m is try , Voi. 37, p a g e s 514, 983, 1083, a n d 1168, 1945. API 06607 APPENDIX 3 P ag e (5) The effect of th erm al cracking on the potency of a cataly t ically crack ed charge stock w ill vary, depending upon the relative ra te s of two com peting phenom ena. The carcinogen content (as m easu red by the benzopyrenes) norm ally in c re ases during this operation. However, the catalytic activity, if any, usually d ecreases. The com position and effective potency of recy cle gas oils from delayed coking un its ap p ear to be s im ila r to those of the heavy catalytically crack ed sid estrea m s. A pparently, the coke building up in the drum h as a cataly tic influence on the directio n of the decom position of the v ario u s c la sse s of com pounds in the feed . com parable to that of the silica-alu m in a cataly sts. The one sam ple of a coker gas oil which h as been exam ined in detail h as shown one unusual c h a ra c te ristic , in that tw o -th ird s of its basic carcinogenic activity can be a sc rib e d to com pounds e x tra c te d by m a leic anhydride. E. P ro g re s s in the D evelopm ent of a Q uantitative A nalytical Tool The expression of the relativ e influences of the carcinogens and the m e d iu m , sy m b o lized by equ atio n (7) (th is A ppendix), m u s t be e x p re sse d in p a ra m e te rs which can be m e a su re d ex p erim en tally . In the typicsLl te s t, an o il of a given co m p o sitio n is ap p lie d re p e a te d ly upon the skin of m ice until papillom as ap p ea r on the back of each m ouse. The average tim e, calculated by standard statistical m ethods, is called the average duration (or expectancy) of tu m o r-fre e life, e ^ B Obviously the m o re rap id the rate of induction of tu m o rs, the sm aller th e c a lc u la te d v alu e ^.5iobs * ^ o th e r w o rd s , th e r a te i s in v e rs e ly API 06608 A PPENDIX 3 Page (6) p ro p o rtio n a l to e ^ g . Thus equation (7) m ay be re s ta te d in te rm s of the ex p erim en tal m e asu re m en ts in the following m anner: __ 1 _ ^obs. _1 Axe e e q u iv a le n t to (6) or, obs. A F o r the sake of c la rity , the two im p o rta n t p a ra m e te rs , A and e --c will be d iscu ssed sep arately in the following p arag rap h s. An understanding of the relatio n sh ip between the com position of the m ed iu m and re s u ltin g value of A h a s been ap p ro ach ed by two p a ra lle l paths. To furnish a background of referen ce, various solutions of these synthetic carcinogens, m ethylcholanthrene and benzopyrene, have been te sted to determ in e the relativ e catalytic a c tiv itie s, A, of the solvents. The solvents em ployed have included pure com pounds and blends of these com pounds to sim u late the v ariety of com positions of the n o n -carcin o g en ic p o rtio n s of p etro leu m oils. The re su lts of th ese te s ts , shown in T able 4 of A ppendix 2, have d em o n strated that the cataly tic activ ity of a solution ris e s a s the m o le c u la r stru c tu re of its chief com ponent approaches that of the typical 4 - ringed carcinogen in general size. The sim ilarity in m olecular size between carcinogens and a c c e le ra to rs will be noted from the exam ples on the following page. A v a rie ty of com pounds of s im ila r m o lecu la r size have been found to be capable of in c re asin g the rate at which the polycyclic carcin o g en s m ay induce tu m o rs. L a rg e r com pounds, e .g . , d i-sec. -nonylnaphthalene, do not seem to s h a re th is p ro p e rty . API 06609 CARCINOGENS A PPEN D IX 3 Page (7) ACCELERATORS C. --------------------------------\ c I C^ o - IC C -C ^ I c t-D odecylbenzene A cr C C ^ C etan e C f ,c c ( A lm ost all refin ery stream s will include som e quantities of lo n g -ch ain a c c e le ra to rs of this type. H ow ever, the a c c e le ra tin g , or cataly tic, activ ity of the total m edium w ill vary over a wide range depending p a rtly upon the co n cen tratio n of such a c c e le ra to rs , which in som e ca se s m ay be quite high, and also to a v ery im p o rtan t extent upon the relativ e concentration of various c la sse s of those com ponents which in th em selv es have n e ith e r acc e le ra tin g n o r carcinogenic activity. The re su lts of ex p erim en ts in which the long-chain a c c e le ra to rs , t-dodecylbenzene and n -dodecane, w ere diluted with various solvents having little o r no catalytic activ ity , led to a reco g n i tion of the p o ten tial im p o rtan ce of such diluents. In T able 4 it w ill be noted that, when the diluent has a m olecular stru ctu re appreciably s m a lle r than th a t of ty p ical c a rc in o g e n s, no sig n ifican t re d u c tio n of A o c c u rs u n til the c o n c e n tra tio n of the d ilu en t ex ceed s 50 p e rc e n t. A PPENDIX 3 Page (8) In c o n tra st with the effect of diluents p o sse sse d of only one o r two rin g s in th e ir m o lecu la r s tru c tu re , is the v ery m ark ed in te rfe re n c e with the a c c elera tin g activ ity of dodecane (and J-dodecylbenzene) on dilution with sa tu ra te d oils containing app reciab le am ounts of condensed cycloparaffins com parable in m o lecu lar size to the typical carcinogens (experim ents 297 and 614). It should be em phasized that this inhibitory property counteracts the catalytic influence of a c c e le ra to rs , w ithout reducing the b asic activ ity of an oil due to the carcinogens th em selv es. D irect proof that it is the content of condensed polycyclo p araffin s of these w hite oils w hich is resp o n sib le fo r th e ir ab ility to counteract the influence of long-chain a c c e le ra to rs has not as yet been obtained (fractio n atio n of the w hite oil A P I-80 haB been in itiated , but the biological te sts of the fractio n s a re still to be c a rrie d out). H ow ever, the tentative application of this assum ption to the p ro b lem of the rela tio n sh ip betw een the corr^position of the non-carcinogenic fractions of com plex oils and their com parative catalytic activ ity h a s proved to be fru itfu l. Thus, as h as been indicated, the testin g of sy n th etic solutions of p u re com pounds h as been p a ra lle le d by te sts on non-carcinogenic fractio n s of a v ariety of refin e ry s tre a m s. The re s u lts of in itia l work on the com position of th ese fra c tio n s (by m a n sp e c tro m e tric an aly sis) h as c o rre la te d w ell with the explanation of the inhibitory p ro p e rtie s of the white oils in te rm s of th e ir content of condensed cycloparaffins. API 06611 I ._ A PPEN D IX 3 Page (9) In T ables 2 and 3 a re shown the data which have been collected on the com position and relativ e catalytic a c tiv itie s of the fractio n of a se rie s of o ils containing le ss than th ree rings p e r m olecule (sep arated from the polycyclic aro m a tic com ponents by chrom atography). The com po sition of the fractio n from the naphthenic lube stock, API-120, is com parable to that of the white oil, A P I-80, each containing about 50 p e rc e n t of m o n o -c y c lo p a ra ffin s (som e of w hich m ay a c t a s a c c e le r a to r s ) a n d ab o u t 35 p e r c e n t of co n d e n se d c y c lo p a ra ffin s . The catalytic activ ity A of th is fractio n of API-120 was found to be e s s e n tia lly the s a m e a s th a t of A P I-8 0 , a p p ro x im a te ly 1. 3. It is expected that the com position of the no n -carcin o g en ic fractio n of the o th e r naphthenic lube sto ck , API-115, w ill have a s im ila r p a tte rn . In c o n tra st, that of the waxy lube stock, A PI-105, w ill be expected to show a high ratio of long-chain p araffin s and m onocycloparaffins to condensed cycloparaffins. In line with the tentative hypothesis, the catalytic activity of the fractio n of API-105, containing le ss than 3 rings p e r m olecule, was alm ost as high as that of pure g-dodecane. As indicated in section D of this Appendix, catalytic cracking tends to in c re a se the ra tio of long-chain com pounds to condensed cy clo p araffin s. T his is shown by the com position of the n o n -carcin o g en ic fra c tio n of A P I-8 9 in T able 3. The vigorous catalytic activity p red icted fo r such a com position is confirm ed by the re s u lt of the bio lo g ical te s t shown in Table 2. S im ilarly it is to be expected that the intense cataly tic activ ity of the non-carcin o g en ic API 06612 I T | I --- APPENDIX 3 Page (10) f r a c tio n of A P I - 33 is due to a c o m p o sitio n co n tain in g a r e la tiv e ly sm all quantity of condensed cycloparaffins. Difficulty has been experiences in obtaining satisfacto ry biological m easu rem en ts of the potency of th erm al ta rs because of th e ir toxicity fo r the m ice. H ow ever, it ap p ea rs that the influence ex erted by cataly sts in th e ir total carcinogenic effect is v ery m uch le ss than it is in oils produced by catalytic cracking o r coking. The com position of the n o n -carcin o g en ic fra c tio n of one fa irly typical exam ple, A PI-65, shows an unusually high ratio of alkylbenzenes and naphthalenes to sa tu ra te d h y d ro carb o n s. Although such a fractio n m ight be expected to show cataly tic activ ity , it is ap p aren t that its influence, in the total com position of A PI-65 (of which it re p re se n ts le s s th an 50 p e rc e n t), is re la tiv e ly s m a ll. The one th e rm a l ta r , A PI-72, which has been found to be significantly m o re potent than the av erag e m a te ria l fro m th is type of p ro c e s s , is expected to be re la tiv e ly fre e of condensed c y c lo p a ra ffin s, on the b a s is of its cru d e so u rce. A d ire c t an aly sis w ill be m ade by m a ss sp ectro sco p y to te st this assum ption. F. P ro g re s s Tow ard the A nalytical E stim atio n of the Content of C arcin o g en s in any G iven P e tro le u m Oil The application of biological m ethods to estim ate the ra te (i. e. , i t s r e c i p r o c a l, e^_) a t w hich a g iv e n o il m a y be e x p e c te d to p ro d u c e tu m o r s if it s c a ta ly tic a c tiv ity A w e re eq u al to 1, h a s b een d e s c r ib e d in A p p en d ix 1. T he p r o c e d u r e in v o lv e d a n e s tim a te of the API 06613 A PPENDIX 3 P ag e (11) in term ed iate p a ra m e te r, d^, the dosage of carcinogen in m illig ram s p er "unit a re a " p e r week supplied by the oil in question. The approach to chem ical m ethods to estim ate e^ is designed sim ila rly to evaluate as an in term ed iate step. The p a ra m e te rs which m u st be m e asu re d by an aly tical m ethods in o rd e r to a rriv e a t the value of a re the to tal content of ca rc in o g e n s C rc> and the re la tiv e sp read in g tendency of the oil, The investigation has been designed to estim ate C rc by adding up the contributions of th ree sep arate c la sse s of polycyclic arom atic carcinogens (expressed in equivalents of the referen ce carcinogen): Crc " C5 / C4 / C3iy2. (8) The m ethod of estim ating the value of Cg fro m the m easu re d co n ten t of b e n z o p y re n e s h a s b e e n d e s c r ib e d in s e c tio n C of A p pendix 1. The d e v elo p m en t of m e th o d s f o r C 4 an d C 3 j/2 *B ^ ein g p u r s u e d vigorously. The p relim in a ry re s u lts indicate the m a jo r im portance of th e 3 1 /2 - r in g e d c a rc in o g e n s in a lm o s t a l l typ<ss of p e tro le u m o ils (the re la tiv e contribution of the 4 - ringed b en zan th racen es and the 5 -ringed benzopyrenes in c re a se s with in creasin g sev erity of catalytic cracking). The p ro g re ss m ade in concentrating the 3 1 /2 -ringed carcinogens indicates that th e ir identification will be com pleted s a tis fa c to rily in the c o u rse of two additional cy c le s of b iological te s ts (about 9 m onths each). Since the investigation of the chem ical and physical p ro p e rtie s of th ese fractio n s w ill be c a rrie d on co n cu rren tly , the API 06614 A PPEN D IX 3 Page (12) step from the identification to the an aly tical determ ination of this j c la ss (and the value of ^ in equation (3)) should be a sh o rt one. The contribution of the 4^-rin g ed com pounds ex tracted by m aleic anhydride v arie s from alm o st nil in the case of raw lube stocks, such a s A PI-120, to as m uch a s 2 /3 of the b asic potency in the c a se of the coker gas oil, A PI-33 (T able 5). The lack of significant quantities of the 4- and 5 -rin g ed carcinogens in A PI-89 m ay be explained by its low fin al boiling point (7 5 0 F . ). Only the c a rc in o g e n s of lo w er m o lecu lar weight (the 3 1 /2 -rin g e d class) a re p re se n t. C u rren tly additional r e s ts a re being c a rrie d out to d eterm in e w hether the content of 4 -rin g ed carcinogens in c re a se s during the therm al cracking of a catalytically cracked charge stock. The content of 5 -rin g ed benzopyrenes m o re than doubled. In addition, the te s ts , which a re being c a rrie d out at a s e rie s of co n cen tratio n s, should provide in fo rm atio n on the slope of the d o se -re sp o n se curve fo r the 4-ringed class of carcinogens. - From the Kettering Laboratory in the Department of Preventive Medicine and Industrial Health, College of Medicine, University of Cincinnati, Cincinnati, Ohio Experimental Work and Report by* Bernard H. Braun, Ph. D. Frank P. Cleveland, M. D. Ralph T. Denham, B. S. Mary Jane Graf, B. S. A. Wesley Horton, Ph. D. Fred Shaffer, M. S. Klaus L. Steamer, M. D. Russell Tye, M. S. Patricia Clapsaddle Lenore Hull Irvin Rapien Effie Wst Date; April 10. 1957 Director