Document gbDrgYVqZDJpgdQ2b86Oe5E7e

ENYIROmENTAL FATE OF COMBUSTION-6E&RATED POLYCHLORINATED OIOXINS AND FURANS by Jean M. Czuczwa and Ronald A. Hites* School of Public and Environmental A ffa irs and Department o f Chemistry Indiana U niversity Bloomington, IN 47405 iV X i . ABSTRACT P p ly c h lo rin a te d d io x in s and fu ra n s were found in sediments from the Saginaw River and Bay ahd from Lake Huron. The congener d is trib u tio n s o f the dioxins and furans indicate thatACombustion may be the major source o f these compounds, the depth vs. c o n c e n tra tio n p r o f ile s in dated sediment cores showed th a t emission o f dioxins and furans has increased g re a tly since 1940. This h is t o r ic a l increase is s im ila r to tre n d s fo r the p ro d u c tio n , use, and d is p o s a l o f c h lo rin a te d o rg a n ic compounds and suggests th a t c h lo r in a te d precursors o f dioxins and furans, present 1n in c in e ra to r combustion fu e ls , may be the main source o f the dioxins and furans found in these sediments. 1 *2 INTRODUCTION P olychlorinated dibenzodioxins (PCOD) and d ibe nzo fu ran s (PCDF) are th e subject o f a recent, often- heated debate because some o f these compounds.are i very to x ic . For example, 2,3,7,8 tetrachlorodibenzodioxin (2,3,7,8-TCDD) has been found to be a cn e g e n ic to humans ( 2 ) , t e r a to g e n ic to m ice (2) carcinogenic to ra ts (3), and acutely to x ic to guinea pigs (4). Other isomers o f PCDD (75 t o t a l ) show d if f e r in g degrees o f t o x i c i t y ; isomers o f PCOF (135 to ta l) are g e n e ra lly as to x ic as the corresponding PCDD. I n i t i a l l y , PCDD and PCDF were'discovered as trace im p uritie s in various chlorinate d aromatic compounds. We w i l l c a ll these "in d u s tria lly -g e n e ra te d " d io x in s and fu ra n s . PCDD and PCDF were found in c h i orophenol s ( 5 - 7 ) , herbicides (8-10), and PCB's (11). In d u s tria lly -g e n e ra te d PCDD and PCOF have entered the environment through accidental release during 2,4,5-T manufacture (12). a e ria l a p p lic a tio n o f phenoxy-herb 1cides (13), and improper disposal o f wastes (14). These events tend to be sporadic and lo c a liz e d . More r e c e n tly , PCDD and PCDF have been id e n t if ie d in e f f lu e n ts from combustion processes. In p a rtic u la r, dioxins and furans have been found in the f l y ash and flu e gas o f municipal in cin e ra to rs ( 15-18). The PCDD and PCDF may be associated with small p a rtic u la te s , which have long residence times in the atmosphere, and 1n this"manner, com bustion-generated d io x in s and fu ra n s could become d is tr ib u te d over la rg e areas. Thus, combustion may have made PCDD and PCDF ubiquitous in the envIronment. The i n i t i a l reports o f PCDD and PCDF in m unicipal in c in e ra to r f l y ash led to an in v e s tig a tio n o f a v a rie ty o f combustion processes each o f which was a p o s s ib le source o f PCDD and PCDF ( 19). PCDD and PCDF were measured in p a rtic u la te s from the combustion o f municipal and chemical wastes and fo s s il f u e ls , and in some unusual samples such as c ig a r e tte smoke and c h a rc o a l- 2/ bro ile d steak. The researchers concluded th a t PCDD and PCOF'are ubiquitous products o f th e combustion o f org a n ic m a te ria ls . In an in te rv ie w (20), an author o f t h is paper s ta te d , HWe now th in k th a t d io x in s have been w ith us since the advent o f fird ." These reports sparked a debate centered on whether 2,3,7,8-TCOD is formed 1n c o a l- f ir e d combustion sources. No TCDD was found in the f l y ash. frorrua c o a l-fire d * pi ant burning T ow -sulfur coal (21) or in a p la n t burning a coa l/refu se derived fu e l m ixture (22). These fin d in g s suggested th a t th e re may be fewer sources o f combustion-generated dioxins than i n i t i a l l y thought. A fundamental point in th is debate centers on the mechanism o f form ation o f d io x in s and fu ra n s in combustion sources. PCDD and PCDF may be formed by the c y c liz a tio n o f c h lo rin a te d precursors present in the fu e l o r by th e re a c tio n o f o rg a n ic compounds w ith in o rg a n ic c h lo r in e both present in the f u e l. I t is l i k e l y th a t the f i r s t mechanism is o p e ra tiv e ; model p y r o ly s is experim ents have shown th a t PCDD a n d /o r PCDF are formed by p y ro ly .z in g chlorinated precursors such as'chlorobenzenes (2 3 ), ch i orophenol s (2 4 ), and PCB's (25). There is l i t t l e experimental evidence fo r the second mechanism, but i t cannot y e t be excl uded. In any case, the f i r s t mechanism is l i k e l y to form PCDD and PCDF in higher y ie ld s than-the second mechanism.. Because o f the t o x ic ity o f these compounds, i t is Important to know th e ir environmental fa te . We propose the fo llo w in g paradigm: Once emitted from a combustion source, the p a r tic u la te s (c a rry in g t h e ir load o f d io x in s and furans) can t r a v e l some d is ta n c e , which is a stro n g fu n c tio n o f the s iz e o f the p a rtic le . Larger p a rtic le s w i l l s e ttle close to the source w h ile small p a r t ic le s may have s u f f ic ie n t reside nce tim es in the atmosphere to be transported to remote lo catio ns. Thus, PCDD and PCDF may be carried by d ir e c t a irb o rn e tra n s p o rt to u ltim a te e n viro n m e n ta l sin ks such as the oceans or lakes. A fte r deposition in these aquatic systems, the dioxins and furans w i l l 3 s e t t le to the bottom sediments. As e a r li e r sediments become b u rie d by Si m aterials deposited in subsequent years, an h is to ric a l record o f d io xin and furan inputs to the environment w i l l be preserved. i: Some caveats should be sta te d re g a rd in g t h is general model f o r . t h e environmental fa te o f PCDD and PCOF. F ir s t, the sources are h ig h ly v a ria b le . The q u a n tity and isomeric d is trib u tio n o f PCDD and PCDF emitted w i l l depend on combustor design, operating conditions, composition o f the fu e l, and degree o f emission c o n tr o l. Second, environm ental a lte r a tio n s in. the a ir or water column due to photodecom position, b io d e g ra d a tio n , v o l a t i l i z a t i o n , or bioaccunulation may occur. Third, we have assumed th a t these compounds are not s u b je c t to d e g ra d a tio n once they are in the sediments. T his is p ro b a b ly a good assum ption; a re c e n t summary o f th e en vironm ental che m istry o f PCDD' j suggests th a t m ic ro b ia l de gra da tion 1s n e g lig ib le (26). F o u rth , sediment mixing processes introduce some, usual ly minor, uncertainty in in te rp re tin g the h is to r ic a l trends o f compounds deposited in sediments. The goal o f our study is to approach two o f the present q u e s tio n s concerning PCDD and PCDF: (a) Are the PCDD and PCDF which are present in the environment the r e s u lt o f In d u s tria l production or combustion? (b) Is there evidence regarding the h is to r ic a l in p u t o f these m a te ria ls in to the environment which could more c le a rly define the mechanism o f th e ir formation? To address these q u e s tio n s , we measured PCDD and PCDF in samples from combustion sources and from la c u s trin e sediments. F ly ash from a municipal in c in e ra to r and from a c o a l-fire d power p la n t were analyzed to study the dioxin and furan congener d is trib u tio n s ty p ic a l o f combustion samples. These d is t r ib u t io n s can be used to determ ine i f combustion is a source o f PCDD and PCDF in environm ental samples. The analysis o f sediments, however, was the main focus o f th is study. 4 S ix sediment samples from the Saginaw R iv e r and Bay and from southern Lake Huron were analyzed3and several types o f information*were obtained. . One was the PCDD and PCDF congener d is t r ib u t io n , which we used to id e n t if y sources. Another was pCDD and PCDF concentrations as a function" o f distance from anthropogenic a c t i v i t y . The most v a lu a b le in fo rm a tio n came from the analysis o f sediment cores. Sections o f cores were analyzed fo r PCDD and PCDF and dated by ra d io -is o to p ic techniques. Thus, the h is to ric a l input o f dioxins and fura ns was obtained. We used these data to d is tin g u is h between anthropogenic and natural inputs o f PCDD and PCDF. EXPERIMENTAL F ly Ash. Four f l y .ash samples were o b tain ed from B. J. Kim ble (Laboratory o f Energy-Re la ted Health Research, Davis, CA). Samples 1 and 2 were c o lle c te d from the cyclo n e stage (c y c lo n e ash) and e le c t r o s t a t ic p r e c ip it a t o r ' (ESP hopper ash), r e s p e c tiv e ly , o f a m id-w estern m u n ic ip a l in c in e ra to r. Samples 3 and 4 were from two c o a l-fire d power p la n ts burning western (low s u lfu r , low ch lo rin e ) coal. Approximately 10 grams o f f l y ash were spiked w ith 100 ng o f 37c1o. q(;qo (KOR Isotopes, Cambridge, MA), allowed to dry, and soxhlet extracted w ith 200 'ml " d i s t i l l e d in g la s s " grade benzene (MCB Reagents, C in c in n a ti, OH) f o r 24 hours. The re s u ltin g e xtra ct was concentrated by ro ta ry evaporation to less than 1 ml and the solvent exchanged to hexane fo r alumina fra c tio n a tio n . Pre-extracted neutral alumina (Brockman A c t i v i t y I , F ish e r S c ie n t if ic ) was activa te d at 250oc fo r 2 hours. I t was then deactivated w ith 1% by weight d i s t i l l e d water and a llo w e d to e q u ilib r a t e fo r 16 hours. A 0.5 x 6.5 cm micro-column was packed and washed with hexane. The sample was Introduced and e lu te d w ith 8 ml each o f hexane, 2% m ethylene c h lo r id e in hexane, and 40% methylene c h lo rid e in hexane. Dioxins and furans eluted in the 40% fra c tio n , 5 which was concentrated to 100 ul by slo w l.y passing a stream o f p u r if ie d Ng over the sample, the sample was then ready fo r analysis by methane negative y chemical io nizatio n GC/MS (NCI-GC/MS). I* ' Sediments. The sam pling s ite s are shown in F igure 1. A sediment grab sample, nominal depth o f 8 cm, was c o lle c te d in 1981 from the Saginaw River (S ta tio n 161) by C. P. Rice (G reat Lakes Research D iv is io n , U n iv e r s ity o f Michigan, Ann Arbor, MI). The lo ca tio n was close to the mouth o f the Saginaw R iv e r (43o 39'N, 83o 5VW). A sediment core from Saginaw Bay (S ta tio n 30A) (43o 52'N, 83o 40'W) and two companion sediment cores from southern Lake Huron (SLH-75-46L, and H, 43o 30'N, 81 o 55'W, f u r t h e r re fe rre d to as cores 1 and 2, re s p e c tiv e ly ) were c o lle c te d by J. A. Robbins (National Ocean and Atmospheric A dm inistration, Great Lakes Environmental Research Laboratory,'Ann Arbor, MI). We used o n ly the top 1 cm s e c tio n o f th e Saginaw Bay core. Two cores from southern Lake Huron were co l le c te d by S. J. E is e n re lc h (Department o f C i v i l and Mineral Engineering, U n iv e rs ity o f Minnesota, Minneapolis, MN) and. J. A. Robbins. These samples .were c o lle c te d in 1981 using a box c o re r at co o rd in a te s 43o59'N, 8lo5'9'W (core 3) and 43o59'N, 82olOW* (core 4). A l l cores were sectioned in to in te rv a ls o f 1 an to a depth o f 10 cm, and then in to 2-5 cm segnents, depending on depth. Approximately 50 g (wet weight) o f sediment were placed in glass soxhlet thim bles and spiked w ith between 2 ng and 200 pg o f th 37c 18"^CDD standard, depending on the expected d io xin and furan le v e ls . The samples were extracted fo r the f i r s t 24 hours with 200 ml o f isopropanol to remove water, follow ed by e x tra c tio n with 200 ml o f methylene c h lo rid e fo r an ad ditional 24 hours. The methylene chloride and isoprppanol extracts were combined and then reduced to 2 ml by r o ta ry e v a p o ra tio n and subjected to a th re e -s te p chrom atographic clean-up. 6  N atural- sediments sometimes c o n ta in s ig n if ic a n t amounts o f e lem e ntal s u lf u r which can In te r fe r e in the a n a ly s is . Thus, s u lf u r was removed by an activa ted copper column. F ifty grams o f copper (p u rifie d e le c tro ly t-ic dust, Fisher S c ie n tific ) was .'activated w ith concentrated HC1. A glass column (1 x. 25 cm) was f i l l e d w ith the copper s lu r r y . The sediment e x tra c t was passed through the column and eluted w ith 150 ml o f methylene ch lo rid e . The eluent was concentrated to 2 m l, and the s o lv e n t was exchanged to hexane fo r fra c tio n a tio n on s ilic a . Pre-extracted s ilic a gel (Davidson Chemical, Baltim ore, MO) was a ctiv a te d a t 160oC f o r 16 hours, d e a c tiv a te d w ith IX w a te r, and loaded in to a 1.5 x 25 cm column w ith hexan. The sample was Introduced and eluted with 75 ml each o f hexane, 15X methylene ch lo rid e in hexane, and methylene chloride;- Dioxins and furans' el uted in the second fra c tio n .... The so lve n t was again reduced and exchanged to hexane. The fin a l step was a lw iin a chromatography as described * above. NCI-GC/MS A nalysis. Negative chemical io n iza tio n (NCI) mass spectrometry 1s p a r tic u la r ly se n s itiv e to molecules th a t have, a high electron capture cross s e c tio n due to el e c tro p h i 11 ic atoms such as c h lo r in e (27). Thus, NCI is an Id e a l technique fo r the a n a ly s is o f PCDD and PCDF. The enhancement o f s e n s itiv ity is e s p e c ia lly pronounced fo r the more h ig h ly chlorinate d dioxins and fu ra n s . In our la b o ra to ry , OCDD showed a th o u s a n d -fo ld in cre a se in s e n s itiv ity over th a t o f electron impact (E l). A ll the analyses were obtained on a Hewlett-Packard 5985B GC/MS system. Chromatographic separation was achieved on a 30 m X 0.25 mn D8-5 fused s il ic a column (J & W S c ie n t if ic , Rancho Cordova, CA) w ith he lium c a r r ie r gas ( s p l. it le s s in je c t io n at 30oc, is o th e rm a l fo r 4 m in utes, 4oc/min to 28Qoc, iso th e rm a l fo r 20 min). The ion source tem perature was 25QQC, and the pressure o f methane was ty p ic a l l y m aintained at 0.7 t o r r in the ion source. 7 t To fu rth e r Increase s e n s itiv ity , se le cte d .io n monitoring was used.- Ion's were monitored fo r te tra c h lo ro - through octachlorodioxins and furans, in clu d in g a c o n firm in g ion [M - o r ((M -C l)-, depending on the isom er]. .The ions used fo r , q u a n tita tio n were: te tra c h lo ro d io x in s (TCDO), m/e = 322; pentachlorodioxins (PnCDD), m/e- 356; h e x a*c h lo ro d io x in s (HxCDD), m/e = 355; h e p ta c h lo ro d io x in s (HpCDD), m/e * 389; o c ta c h lo ro d io x in (OCDO), m/e 423; te tr a c h lo ro fu r a n s (TCDF), m/e * 308; pe nta chl o ro fu ra n s (PnCDF), m/e = 340; hexachl o ro fu ra n s (HxCDF), m/e * 374; heptachl orofurans (HpCOF), m/e * 408; and >octachlorofuran (OCDF), m/e * 444. In a d d itio n , m/e 435 was monitored, representing the M-Cl ion o f the in te rn a l standard. Dioxins and furans were quantitated by ra tio in g the appropriate peak area to th a t o f the in te rn a l standard and co rre ctin g fo r r e la tiv e response factors which were obtained from a standard m ixture o f PCDD and PCDF (one Isomer per congener cla ss). D ating o f Sediment Cores. Sedim entation ra te s fo r the cores were A determ ined by J. A. Robbins and K. A. Johansen (NOAA, Ann A rbor, MI) using th e Cs-137 and Pb-210 tech niq ue s o f Robbins and Edgington (28). S edim entation rates varied from 0.15 - 0.41 cm/yr and w i l l be discussed below. Q u a lity Assurance. The a n a ly tic a l work follow ed the guidelines suggested by the ACS Committee on Environmental Improvement (29). Experiments included re p lic a te s , p ro ce d u ra l b la n k s , and re c o v e ry measurements. The re c o v e ry averaged 75% even fo r the lowest le v e l samples.. The average re p ro d u c ib ility was b e tte r than +30%, the le a s t re p ro d u c ib le being the t e t r a c h lo r o - and pentachloro- PCDD and PCDF. The l i m i t o f detection was 20 pg fo r 1,2,3,4-TCDD and 0.2 pg f o r 0CDD. S u bse que ntly, i t was found th a t lo w e rin g the GC/MS ion source temperature to 150oC re su lte d in increased s e n s itiv ity , lowering th lim its o f detection t 0.01 pg fo r 1,2,3,4-TCDD and 0.03 pg fo r 0CD0. Method v a lid a tio n included an in te rla b o ra to ry c a lib ra tio n experiment. In 8 Table I, dioxin concentrations measured by the above procedure in a sample o f St. Louis a ir p a rtic u la te s (National Bureau of Standards, Standard Reference M a te ria l #1648) are compared w ith those rep orted by Bumb ^et a_K ( 19). A lthough the e x tra c .tip h , cle a n -u p , and mass s p e c tro m e tric techniques are d iffe r e n t, the re s u lts agree w ith in the measurement e rro r o f the procedures. The te tra c h lo ro d io xin s may be an exception; in th is case, the El analysis used by the comparison la b o ra to ry was more s e n s itiv e . However,, i t w i l l be shown la t e r th a t th e le s s c h lo rin a te d d io x in s are th e le a s t abundant compounds encountered in our work. RESULTS AND DISCUSSION Combustion Sources. Figure 2 shows the re s u lts o f the f l y ash analyses; note th a t 4 d iffe r e n t concentration scales are used in th is fig u re . Sample 1, c o l l ected a t th e c y c l one stage o f a m u n ic ip a l in c in e r a to r con tain ed lo w e r c o n c e n tra tio n s o f PCDD and PCDF than the e le c t r o s t a t ic p r e c ip it a t o r (ESP) hopper ash (sample 2) from the same p la n t. This can be e xp la in e d by th e fo llo w in g mechanism:. A t the c y c lo n e s ta g e , m a te ria ls are c lo s e to the combustion chamber and are col 1ected a t a higher temperature than at the ESP. In the cyclone, the dioxins and furans could be p rim a rily in the vapor phase and, th e re fo re , not associated w ith the p a rtic u la te s . As the e fflu e n ts reach the ESP, the temperature has decreased, and PCOO and PCOF are now condensed on the p a rtic u la te phase and, thus, are c o lle c te d . I t is u s e fu l to note some tre n d s in the PCDD and PCDF congener d is trib u tio n s in the f l y ash samples (see Figure 2, samples 2 to 4). F ir s t, a la rg e number o f isomers were d e te cte d f o r each PCDD and PCDF congener. Second, OCDD is th e most abundant d io x in , and OCDF is present in much low er c o n c e n tra tio n s than OCDD. And t h i r d , the hexa- or heptachl orodibenzofurans are th e most abundant fu ra n s . S im ila r d io x in congener p r o f ile s were a ls o 9 measured in f l y ash samples from a powerhouse, a r o ta r y k i l n , and waste in cin e ra tin g f a c il it ie s in a recent study (19) and in municipal in cin e ra to r and f l y ash ,(15-18). .These general trends w i l l be compared to "th e p r o file s found in environmental samples to d is tin g u is h among possible sources. I t is important,-however, to recognize th a t these samples represent on ly a few sam pling occasions and, th e re fo re , may not be re p re s e n ta tiv e o f combustion processes in g e n e ra l. F urtherm ore, an e x tr a p o la tio n o f these trends to those found 1n environmental samples does not take in to account the p ro b a b ility th a t fly .a s h c o lle c te d in an e le c tro s ta tic p re c ip ita to r may not accurately r e f le c t the PCDD and PCDF th a t are emitted. These data w i l l sim ply be used to compare combustion processes with environmental samples. The coal f l y ash samples (Figure 2, samples 3 and 4) d if f e r s ig n ific a n tly from the municipal in c in e ra to r ash samples. Although some RCDD and PCDF were detected, no te tra ch lo ro ^ or p e n ta c h lo ro - d io x in s or fu ra n s were d e te c te d , w ith li m i t s o f d e te c tio n o f 100 ppt ( t e t r a ) and 10 ppt (p e n ta ). D io xin s and furans, when present, were 1n much lower concentrations than in the m unicipal in c in e ra to r ash. For example, the le v e ls o f 0CD0 in the coal f l y ash samples (2.2 and 3.8 ppb 1n samples 3 and 4, r e s p e c tiv e ly ) were a t le a s t 100 tim es lower than those found in the municipal In cin e ra to r ash (440 ppb). A lthough coal f l y ash c le a r ly c o n ta in s le s s 0CDD then m u n ic ip a l in c in e r a to r ash, i t s t i l l could be a s ig n if ic a n t source o f 0C0D to the environment, since coal combustion is so prevalent. In 1974, the amount o f p a r tic u la te s em itted from coal combustion was estim ated to be 2.4 x 109 kg (30). The amount o f p a r tic u la te s em itte d from s o lid waste in c in e r a tio n in 1971 was estim ated at 7 x IO** kg (3JJ. Thus, a p p ro x im a te ly 3 tim es more p a rtic u la te s are em itted from coal combustion than from s o lid waste combustion. I f the OCDO c o n c e n tra tio n on coal p a r t ic u la t e s is 100 tim es 10 lo w e r, then t o t a l em ission o f OCDD from coal combustion would s t i l l be approximately 30 times lower than th a t from municipal waste in c in e ra tio n . F in a lly , we should, address the ongoing debate regarding _2,3,7,8-TCDO in ii t coal f l y ash. No isomer o f TCDD was detected in these samples with a l i m i t o f d e te c tio n o f about 100 p p t. This re a ffirm s s im ila r fin d in g s [ 21,221 and suggests th a t' coal combustion is not a s ig n ific a n t source o f 2,3,7,8-TCDO to the environment. Surface Sediments. Figure 3 shows the dioxin and furan congener p r o file s obtained from s u r f i c i a l sediments from the Saginaw R ive r and Bay and from southern Lake Huron. PCDO and PCDF are u b iq u ito u s in the samples s tu d ie d , in clu din g the most remote lo ca tio n s. The p r o file s are s im ila r to those shown 1n F igu re 2. The c o n c e n tra tio n s o f PCDO and PCDF are h ig h e s t in those sediments c o lle c te d c lo s e s t to urban areas (161 and 30A) and lo w e s t in th e open lake cores. This indicates th a t the PCDD and PCDF found, in these samples are anthropogenic 1n o rig in . Figure 4 shows the mass .chromatograms fo r the dioxins and furans' in the s u r f ic i a l segment o f core 1. These data 111 u s tra te the t y p ic a l com bustion " fin g e rp rirT t" discussed e a r li e r . A number o f isomers f o r each d io x in and fu ra n congener c la s s are d e te cte d . One fin d s a predominance o f OCDD and* HpCDF, and g re a te r le v e ls o f TCOF and PnCDF tha n-the corresponding d io x in s . In g e n e ra l, the PCDD and PCDF isomer d is t r ib u t io n s , even in the most remote sam ples, are s im ila r to each o th e r and are in d ic a t iv e o f com bustion. We, the refore , conclude th a t combustion is probably the major source o f PCDD and PCDF to these lo ca tio n s. These data emphasize the importance o f determining the e n tir e PCDD/PCDF p r o f i l e ra th e r than ju s t th e 2,3,7,8-TCDD c o n te n t. Future work w i l l include isom er-specific q u a n tita tio n o f a l l te tra - through o c ta c h lo ro -d io x in s and fu ra n s , thus in c re a s in g our a b i l i t y to d is tin g u is h among sources. 11 . ' Sediment Cores. As o u tlin e d above, we have obtained data oh sediment cores to determ ine the h is t o r ic a l in p u t o f PCDD and PCDF, O b v io u s ly , th e -amount and the com position o f fu e ls have changed w ith tim e. The e f f e c t o f these changes on the input o f dioxins and furans to the environments should be re fle c te d in the sedimentary record. S im ila r work by Hites ;et _al_. (3-21 showed tha t sedimentary p o ly c y c lic aromatic hydrocarbons re fle cte d the changing use of fossi 1 fuels. The most abundant PCDD and PCDF in cores 1 to 4 were HpCDO, HpCDF, and OCDD. The depth vs. c o n c e n tra tio n p r o f ile s f o r these species are shown in Figure 5. The sedimentation rate fo r cores 1 and 2 (companion cores from th same lo ca tio n ) was ca lcu la te d from the rad io * is o to p ic dataand was found to be 0.15 an/yr. This was used to estimatethe year o f deposition corresponding to each d e p th , and the se da ta are a ls o p lo t t e d in F igure 5 (to p ). The m ix in g depth was 2.8 cm and is an in d ic a to r o f th e degree o f movement o f m a te ria l s a f t e r d e p o s itio n . T his in d ic a te s th a t th e re may be an averaging o f in p u ts over as much as a 15 year in te r v a l. In both cores 1 and 2, there is an abrupt increase in PCDD and PCDF concentrations around 1940. Lake Huron core 3 had a sedimentation rate o f 0.21 an/yr and a mixing depth o f 6.4 cm. This core also showed th a t PCDD and PCDF^ in p u ts increased around 1940 (see F ig u re 5, bottom , l e f t ) . Core 4 had a s u b s t a n t ia lly h ig h e r sedim entation r a te (0.41 cm /yr, m ixin g depth o f 5.9 cm), and thus o ffe re d g re a te r tim e r e s o lu tio n f o r t h is tre n d . From t h is core , i t is apparent th a t PCDD and PCDF in p u ts increased s lo w ly during th e 1940's and e a rly 1950's to thepresent le v e ls (see Figure 5, bottom, rig h t) . In general, the concentrationso f PCDDand PCDF in core sections corresponding to deposition before 1940 are low, representing a much lower input o f these m aterials before th is time. These changes cannot be accounted fo r by _in s itu degradation o f PCDD and 12 * PCOF in the sediment,. The congener p r o file s in a l l cores were s im ila r to each o th e r and were c o n s is te n t along the depth o f the core,. T his is best illu s tr a te d in Figure 6 which shows the congener d is trib u tio n s fo r Core 4 as a f u n c t io n o f d e p th , -Note th e s i m i l a r i t y o f th e p a tte r n w ith d e p th . Furthermore, isomer ra tio s were c a lcu la te d at each depth in core 4, and were found to be co n sta n t. For example, th e r a t io o f 1 ,2 ,3 ,4 ,6 ,7 ,8-HpCDF to 1 ,2,3,4,6,8,9-HpCDF was 0.71 +; 0.27, the r a t io o f 1 ,2,3,4,6,7,9-HpCDD to 1 ,2,3,4,6,7,8-HpCOD was 0.72+^0.06, and the r a t io o f 0CD0 to 1 ,2 ,3 ,4 ,6 ,7 ,8 HpCOD was 2.8 +^ 0.36. There were no tre n d s in the r a t io s w ith In c re a s in g depth fo r anyo f the hepta- to o c ta - d io x in s and fu ra n s . Thus, th e re is no evidence o f degradation o f PCDD or PCOF in these Lake Huron cores. C le a rly , the dioxin and furan inputs have changed considerably over time; much more were deposited since 1940. I t seems l i k e l y t h a t a m ajor source began in th e 1940's.and increased u n t i l the present tim e. What is such a source? The burning o f coal has been a m ajor combustion process sin ce the l ast century, and.coal f l y ash, as shown e a r lie r , contains seme PCDD and PCOF. The tren d f o r U.S. coal consumption f o r the la s t c e n tu ry is shown in F ig u re 7, top (33). Coal use in the Great Lakes area para! 1el s th a t o f the nation. For example, in 1930, I l l i n o i s , Indiana, Ohio, Michigan, and Wisconsin accounted fo r 402 o f the nation's coal consumption; in 1957, 30X; and in 1970, 30% (33). Therefore, we may s a fe ly compare U.S. coal consumption to sedimentary dioxins i and fu ra n s in th e Great Lakes. F ig u re 7 compares coal consumption w ith the t o t a l HpCDD, HpCDF, 0CDD and 0CDF found in the fo u r Lake Huron cores (see F ig u re 7, bottom ). I t is obvious th a t coal use cannot account fo r the increase in d io xin and furan concentration since 1940. Indeed, coal use has been r e la t iv e ly constant since 1910. The U.S. T a r r if f Commission Reports o f Production and Sales o f Synthetic 13 1 Organic Chemicals (34), published since 1918, re v e a l th a t the chemical In d u s try grew g r e a tly beginning in 1940. S ta rtin g a t t h is tim e , the p ro d u c tio n o f c h lo rin a te d organic compounds such as chlorobenzenes and i, chlorophenols increased s u b s ta n tia lly (see Figure 7, middle). These compounds are used in a v a rie ty o f products, in clu d in g b u ild in g supplies, he rbicide s, and packaging. Much o f these m a te ria l s e ve n tu a l l y become In c o rp o ra te d in s o lid wastes. The trend f o r the p ro d u ctio n o f c h io ro -o rg a n ic compounds is v e ry s im ila r to the sedim entary PCDO and PCDF p r o f ile s (compare F ig u re 7 m id d le and bottom ). The agreement between these two tre n d s is c o n v in c in g d e s p ite the u n c e rta in tie s Introd uce d by sediment m ixing and th e e rro rs inherent in the dating and q u a n tita tio n techniques. From these data, we conclude th a t the input o f dioxins and fu r ans to the sedim entary environm ent is p ro b a b ly due to the combustion o f c h lo rin a te d org a n ic products present 1n v a rio u s wastes. These wastes may be m u n ic ip a l waste from Saginaw, Bay C ity , or other urban areas; or they may be in d u s tria l wastes from chemical m an u fa ctu rin g ta k in g p la c e in c e n tra l M ichigan. The d ir e c t dumping o f chem ical wastes ( f o r example, from p e n ta ch lo ro p h e n o l production) 1s an a lte rn a te , but u n lik e ly , in te rp re ta tio n o f our re s u lts . The agreement o f the congener and Isomer p r o f ile s o f PCDD and PCDF in the sediments w ith those in combustion e f f lu e n ts and 1n a i r p a r t ic u la t e s (see Table I) and the coincidence o f the production and concentration p r o file s (see F ig u re 7) are persuasive pieces o f evidence th a t combustion is th e m ajor source. D ire c t dumping and coal or n a tu ra l combustion may be re a l sources, but we b e lie v e them to be m inor. In any case, i t is c le a r th a t the high le v e ls o f dioxins and furans found in presently accumulating sediments are not due to the " advent o f f ir e . " 14 3 ACKNOWLEDGEMENTS We are g r a te fu l to B. J. Kimble fo r the f l y ash samples, to S. J. E is e n re ic h , P. A. Meyers, C. P. Rice, and J. A, Robbins fo r the v a rio u s sediment samples, to B. D. McVeety fo r in s tru m e n ta l a s s is ta n c e , and to S, L. Sikes fo r c le r ic a l support. 15 Table I. In te r! aboratory Comparison o f Dioxin Concentrations in a St. Loui s Urban A ir P a rtic u la te Sonple (NBS, SRM #*1648). i: 1i PCDD Concentration (ppb) ~~ This work Ref. 19 TCDD . <2 * 0.2 PnCDD . <0.2 HxCDD 8 2 HpCDD 42 34 OCDD 250 210 16 REFERENCES 1. Kimmig, J.; Schulz, K. H. Dermatoloqica 1957, 115; 540-546. 2. Schwetz, B. A.; N d r r is , J. M.; Sparschu, G. L.; Rowe, V. K .fG e h rin g , P, J . ; Emerson, J. L. ; Serbig, C. G. Adv. Chem. Ser. 1973, 120, 55-69. 3. Van Hi 11 e r , J. P.; La i ic h , J. 0.; A.1 1en, *J. R. Chemosphere 1977, _> 537-. 544. 4. Gupta, B. N.; Vos, J. G.; Moore, J. A.; Z in k l , J.. G.; B u llo c k , B. C. Environ. Health Persp. 1973, J5, 125-140. 5. B laser,-W . W.; Bredeweg, R. A.; S h a d o ff, L. A.; S te h l, R. H. A n a l. Chem. 1976, 48, 984-986. 6. Buser, H. R. J. Chromatoqr. 1975, 107, 295-310. 7. Buser, H. R.; Bosshardt, H. P. J. Assoc. O ff. A n a l, Chem. 1976, ^ 9 , 562- 569. 8. Rappe, C .; Buser, H. R.; Bosshardt, H. P. Chemosphere 1978, 7_, 431-438. 9. Buser, H. R.; Bosshardt, H. P. J. Chrocnatogr. 1974, 90, 71-77. 10. Yam agishi, T,; M iy a z a k i, T.; Akiyama, K.; M o rita , M.; Nakagawa, J .; H o rii, S .; Kaneko, S. Chemosphere 1981, J , 1137-1144. 11. Vos, J. G.; Koeman, J. H.; Van der Maas, H. L.; ten Noever de Brauw; de Vos, R. H. Fd. Cosmet. T o x ic o l, 1970,., 625-633. 12. C a rre ri, V. In "D ioxin: T oxicological and Chemical Aspects"; C attabeni, F., C a y a lla ro , A.; Gal 11, G., eds.; SP M edical and S c ie n t if ic Books. Spectrum P u b lica tio n s, In c .: New York, 1978; pp. 1-4. 13. Baughman, R. W. ; Meselson, M. Environ. Health Perspect. 1973, 5_, 27-35. 14. C a rte r, C. D.; Kimbough, R. 0.; L id d le , J. A.; C lin e , R. E.; Zack, M. M.; B a r th e l, W. F.; Koehl e r, R. E.; P h i l l i p s , P. E. Science 1975, 188, 738740. 15. 01ie , K. ; Vermeulen, P. L . ; Hutzinger, 0. Chemosphere 1977, 6, 455-459. 17 16. Buser, H. R.; Bosshardt, H. P.; Rappe, C. Chemosphere 1978, ]_, 165-172. 17. Eiceman, G. A.; Clem ent, R. E.; Karasek, F. W. A n al, Chem, 1979, 5J_, 2343-2350. ` ti ` 18. C a v e lla ro , A.; Bandi, G.; I n v e r n iz z i, G.; L u c ia n i, L.; M o n g in i, E.; Gorni, A. Chemosphere, 1980, 9^ 611-621. 19. Burnt), R. R.; Crummett, W. B.; C u tle , S. S.; G le d h il, J. R.; Hummel, R. H,; K a g e l, R. 0 .; L a m p a rs k i, L. L .; Luoma, E. V .; M i l l e r , 0. L .; N e s tric k , J. J.; S h a d o ff, L. A.; S t e h l, R . H.; Woods, J. S. S cie nce, 1980, 2J0, 385-390. 20. "Dow Finds S upport, Doubt f o r D ioxin Ideas," Chem. Eng. News Feb. 12, 1979, 23-29. 21. Kimble, B. J . ; Gross, M. L. Science 1979 , 207 , 59-61. 22. Junk, G. A .; Richard, J. J. Chemosphere 1981, _1, 1237-1241. 23. Buser, H. R. Chemosphere 1979, , 415-424. *24^ Buser, H. R. J . Chromatogr. 1975, 114, 95-108. 25. Buser, H. R.; B osshardt, H. P.; Rappe, C.; L in d a h l, R. Chemosphere 1978, 419-429. 26.. Kearney, P. C.; Presented a t 2nd In te r n a tio n a l Workshop on.Chi o rin a te d Dioxins and Related -Compounds, A rlin g to n , V irg in ia , October 25-29, 1981. 27. Dougherty, R. C. Biomedl Mass Spectrom. 1981, 283-292. 28. Robbins, J. A.; E dglngton, D. N. Geochim. Cosmochim. A cta 1975, 39_. 285- 304. 29. ACS Committee on E nvironm ental Improvement; "G u id e lin e s f o r Data A cquisition and Data Q ua lity Evaluation in Environmental Chemistry" Anal. Chem. 1980, 52, 2242-2249.' 30. Chrisp, C. E .; F ishe r, G. L . ; Lanmert, J. E. Science 1977, 199, 73-75. 18 31. " C o m p ila tio n of. A ir P o llu tio n Emission F a c to rs , 2nd Ed.1' U.S. EPA: Washington, D.C., 1973. 32. H ite s , R. A.; Laflamme, R. E.'; F a rrin g to n , 0. W. Science 1977, 198, 829- 831.- ;; ' '~ 33. Miner, Yearb,. U. S. Bureau o f Mines, 1870-1980. 34. P roduction and Sales o f S y n th e tic Organic Chem icals, U. !. T a r i f f Commission, 1919-1980. CREDIT This work was supported by the U.S. Department o f Energy (Grant No. 80EV10449). 19 ' FIGURE CAPTIONS' Fig. 1 Map showing the Lake Huron sample s ite s . Fig. 2 C oncentration p r o f ile s o f PCDD and PCDF in combust ion. p a rti cul ate samples from: 1, a midwestern municipal in c in e ra to r cyclone; 2, the e le c tro s ta tic p re c ip ita to r from the same in c in e ra to r; 3 and 4, two c o a l-fire d power plan ts burning western coal. Fig. . PCDD and PCDF congener p r o f ile s in s ix s u r f ic i a l sediments (see Figure 1 for s ite locations). Fig. 4 Mass chromatograms showing PCDD. and PCDF in the s u r f ic ia l segment from core 1. These data have not been c o rre c te d fo r response factors. Fig. 5 Hp.CDF, HpCOD and OCDD c o n c e n tra tio n s in fo u r Lake Huron cores vs. depth ( l e f t a x is ) and d e p o s ltlo n a l age ( r ig h t a x is ). Symbols: __ _________ .____________ H pC D F;_______________________ HPCDD; _____ . _____ . _____. OCDD. F ig. 6 PCDD and* PCDF congener p ro file s in -co re 4 as a fu n ctio n o f depth. Fig. 7 U.S. consumption o f coal and p ro d u c tio n o f s y n th e tic c h lo r in a te d o rg a n ics ( in c l udes. c h io ro - and dichlorobenzenes; 2 ,4 -d ich l oro- and 2 , 4 , 5 - t r ic h la r d p h e n o x y a c e t ic a c id , e s t e r s , and s a l t s ; and pentachlorpphenol) compared to the to ta l PCDD-and PCDF_in the four Lake Huron cores as a fun ction o f time ( a l l are p lo tte d on a decade b a s is ). 20 200 TCDF PnCDF HxCDF HpCOF OCOF TCDD PnCDD HxCOO HpCDD OCOO 600 400"' z warn a ______1 J iB m -- mm H TCDF PnCDF HxCDF HpCDF OCDF TCDD PnCDD HxCDD HpCDD OCDD CONC. (ppb) TCDF PnCOF HxCDF HpCDF OCDF TCDD PnCDD HxCDD HpCDD OCDD QONC. ( w * ) . TCOF PnCOF HxCDF HpCOF OCOF TCDO PnCOO HxCDO HpCOO OCDO TCOF PnCO F HxCDF HpCO F OCOF TCOO PnCOO HxCDO HpCOO OCOD TCOF PnCOF HxCDF HpCOF OCOF TCOD PnCOO HxCOD HpCDO OCOD CORE 1 TCOF PnC O F HxC DF HpCOF OCOF TCOO PnCOO HxCOO HpCOO OCOO FURANS DIOXINS 123468 123689 123478 1234679 -- 1234678 12346789 . o croo -rro- Depth in core (cm) io\ i *e-n* ro oo ^ a D e p o sitio n a l Aqa ro o CD Depth in core (cm) r o r o o * e--n r--o oo O) pB O3 *ToJ rt- JOOO 00 a a trooo IUUO)1 to . atn Depositional Aqe (CUOO1 1380 o Depth in coro (cm) o uoi iu\tj ow . u*-t o*-- at 200 400 600 Cone. (ppt) 800 1975 0 ,--- ----- 0 200 400 600 800 Cone. (ppt) Depositional Aqe Depth in core (cm) uai Mat wa at o at JT Y ro L COOD CUUDII CD a tauoi to o Depositional Aqe -- - :--- :-- ,1375 4 1600 0-1 CM 1200" 800" 400" TCOF PnCDF HxCDF HpCDF OCDF TCDD PnCOD HxCOD HpCDD OCOD 1200 2 -3 CM 600" 400" TCOF PnCOF HxCOF HpCOF OCDF TCOO PnCOD HxCOD HpCDD OCOD 1200 4 -5 CM eoo- 400" CONC. (ppt) 1200 800" TCOF PnCOF HxCDF HpCDF OCDF TCOD PnCDO HxCOD HpCDD OCOD 6 -7 CM 400" TCDF PnCOF HxCDF HpCOF OCOF TCOD PnCOD HxCOD HpCDD OCDD 800 8 -9 CM 400" 800 400" TCOF PnCDF HxCDF HpCDF OCDF TCDD PnCDD HxCDO HpCDD OCDD 1 0 -1 2 CM TCDF PnCDF HxCOF HpCDF OCDF TCDD PnCDD HxCDO HpCDD OCDD * i YEAR CONC. (PPT) UtLUONS OF J.. t; jf'K: M. L i a r K TO: DTF & c c 1s U niversity of Illinois at Chicago 2121 West Tnvlor Street (3l2j996-66ku 0820 SCHOOL OF PUBLIC HEALTH July 7, 1983 Mailing Address: Post Office Box 6998 Chicago, Illinois 60680 Ronald A. Hites Jean M. Czuczwa School of Public and Environmental Affairs and Department of Chemistry Indiana University Bloomington, Indiana 47405 Dear Dr. Hites/Ms. Czuczwa: I read with great interest a draft of your paper "Environmental Fate of Combustion-Generated Polychlorinated Dioxins and Furans." Let me say that your type of research and findings are critically important in the understanding of the true scources of PCDDs and PCDFs in our environment. I would like to offer the following comments for your consideration if your paper is to be published: 1. The PCDDs and PCDFs isomer distributions, which decreases as the distance from Saginaw increases would also be interpreted as indi cating a discrete water pollution source-- discharging pentachlorophenol or other similar chemicals to the Saginaw Tittabawassee River System. The isomer distribution of PCDDs and PCDFs which you found in the sedi ments match as closely to those found in pentachlorophenol (technical grade) as to the pattern you reported in your combustion samples. Dow Chemical of Midland has been a major producer of pentachlorophenol and data from Dow, US EPA, and the State of Michigan have confirmed the presence of pentachlorophenol and other PCDD/PCDF containing chemicals in the effluent of Dow Chemical and in sediment of the Tittabawassee River. You may wish to analyze the sediment cores for PCDD/PCDF chemicals and determine if the ratio of the PCDD/PCDF chemicals to PCDD/PCDFs in sediments closely approximates the ratio of PCDDs and PCDFs found in the commercial products which Dow manufactures in Midland. 2. It is not possible from the data collected to conclude that the PCDDs and PCDFs found in sediments are from combustion sources (follows argument presented above). US EPA (see enclosure) has recently found PCDDs and PCDFs in Dow's wastewater and in caged fish placed in Dow's effluent. Process water from chemical manufacture and scrubber water from Dow's chemical incinerators are also discharged to the Tittabawassee River. Therefore, at this time there is not enough data to conclude that the source of some PCDDs and PCDFs in sediments are solely from combustion processes. t Dr. Hites/Ms. Czuczwa #" Page two July 7, 1983 I hope these suggestions are helpful. If you have any questions, please do not hesitate to contact me (312/996-0820). Congratulations for initiating the breakdown of the "Trace Chemistries of Fire" theory. Sincerely, y J. Milton Clark, Ph.D. Environmental and Occupational Health Sciences JMC:vld Enc-