DOCUMENT 7 A declassified letter report dated December 15 1981 from Philip S Thayer PhD Consultant at Arthur D Little Inc to Mr Steve Singer American Broadcasting Company attention “Rain of Terror” ABC News Closeup Steve Singer Producer in CBW Box 5 Reports Articles Papers and Memoranda 1982-84 1986-87 and 1989 National Security Archive Poem “A Moment Still Waiting for You” on page 99 of book YELLOW RAIN Page 1 March 17 1982 I Mr Richar d Bur Direct or of Politic o-Mil itary Affair s United States Depart ment of State Washin gton D C 20520 j j 1 Dear Mr Burt Thanks for your note and your comnen ts about ''Rain of Te or I have enclos ed a copy of Arthur D Little Inc 's report to us on the chemic al analys es done on the ABC News - Whitne y sample The result s of Dr Rosen 's analys is -are part of the Arthur D Little report Dr Rosen has submit ted his report for public ation in scien tific journa ls but as yet no public ation has occu ed He has asked us to hold his full report until after it has been printe d We are doing that but feel that circul ation of the Arthu r D Little report is not incon sisten t with his wishes I see by the New York TiJDes that you may be transferred to anothe r post in State Maybe our profession al paths will cross again if you move Or maybe even if you don't -- - Sincer ely • f s I _I 'vl c s 1_ - '_I ' I _ 21 Page 2 Arthur D Little lnc -c - ··· • · v ' 3 · •• c2• o s 5 n o •E EX 92 1436 December 15 1981 American Broadcasting Company 7 W 66th Stree t ASCAP 2 New York NY 10023 Attention Rain of Terror ABC News Closeup Steve Sing er Producer Dear Mr Singer AOL Ref 87134 This is submitted as the final repo rt on our activ ities on your beha lf in connection with chemical analy of a sample of 111aterial prov through you There were several ses ided tasks outli ned in our proposal dated October 21 1981 and authorized by you on Octo ber 26 Each will be discussed sepa ratel y 1 Arrange for the anaL yses desir ed to be perfo rmed at an appr opria te laborotol ' or Zabo ratoz -ies There were three categ ories compounds for which analyses were origi nally desir ed seve ral oftrich othecenes of the T-2 toxin type macrocyclic trich othec enes and othe Analysis of the carry ing compounds r non-mycotoxin chemical agen ts or matr ix of the sample was also desir ed if possible afte r the 'init ial analy ses It soon became that no singl e labo rator y was expe r rienced and qual ified to do all clea analy ses and fina lly when the actua the 1 size of the samp le was known it was decided that the third category of analyses would be dropped In the search for appr opria te labo rator ies seve ral crite ria were applied 1 experience with the ral spec ific trich othe cene s of inter est and avai labil ity of referseve ence samp les 2 willi ngne ss to work as rapid ly as poss ible and 3 repu tatio n in the field Or Ches ter Mirocha of the Univ ersit y of Minn esota would have fit all of these eminently but since we wanted to obta prein obje ctive conf irma of his findings with othe r samp les he was not approached tion or not Before the final selec tion of Or Bruc e Jarv is of the Univ ersit y of Maryland and Dr Joseph Rosen of Cook fifte en labo rator ies were cons idere d Coll ege Rutgers Univ er it y or approached Or Edward Smalley Univ ersit y of isco nsin was capable of analy zing only for d xynfvale ol _ Vom1toxin for T-2 toxin by rad10-111111unoassay and was just developing a techn ique RIA He al so would requ ire a large r sample than we had because of the lower sens itivi ty of his method 22 Page 3 rtb r D L d re Oettt11ber 15 1981 - 2Mr Steve Singer American Broadcasting Company or Wi l l iam Roush Massachusetts Insti tute of Technology had no chemical st ndards and was not routinely analyzing crude materials being a syntheti c organic chemist He did suggest other workers including Or Jarvis Two others were a government employee and a chemist at a pharmaceutical company neither of which we considered appropriate sites for the task Or William Busby of MIT has published one of the definitive reviews in the field and thus has an excellent overview of the scientific c01T1Tiunity of interest He did not have anal yt ical capabilities but suggested a number of leading workers we might approach They are listed as the next eight entries Or F S Chu University of Wisconsin was contacted but could only analyze T-2 toxin by radio-inrnunoassay Ors Ronald Vesonder and Harland Burmeister of the Northern Regional Research Laboratories of the U S Department of Agriculture Peoria IL were not approached because of potential complications due to their affil i ation with the U S Government Or Robert Epply of the U S Food and Drug Administration was not approached for the same reason Dr M A Hayes and Dr H B Schliefer University of Saskatchewan were called but found to work only with T-2 bioassay by animal skin application which lacks both the sensitivity and specificity we needed Or James Bamburg of Colorado State University the original isolator of T-2 toxin lacked experience in the most recent analytical methodology Ors P M Scott and J Harwig of the Canadian Food Directorate equivalent to the U S F O A were not approached for reasons similar to those applying to U S Goverrvnent personnel · Or Bruce Jarvis was ·contacted in regard to the T-2 type analyses for which he was not equipped We later went back to him primarily upon Dr Mirocha's rec ndation for the macrocyclfc analyses and engaged him for that task It 1s generally agreed by others we talked to that he is the pre-eminent worker in this specialized area Dr Joseph Rosen was orig1n1lly suggested by Busby Initial contacts indica tea a capability of lllllyz 2 VOllitoxin deoxynivalenol and DAS di1c1tox cf _ 1 e with these analyses had been gained with 1111lyz primary of milk and com He was not uPe alenol 23 Page 4 Oecember 15 1981 -3- Mr Steve Singer Ar-erican Broadcast ing Company Wisconsi n Conversations were held with Raltech Scientific Services of Madison the nonprimarily with James Kinsinger They would have been appropriate for trichothec ene analyses but were not further considered when that aspect was dropped A Dr Romer of Ral ston Purina was suggested by Rosen as one of three laborawas not aptories capable of doing the multiple trichothecene analyses aHeCOIT'ITlercia l proached because of the slight chance of the laboratori es of product company being willing to participa te The other two laborator ies indicated by Rosen were his own and Mirocha's of the We establishe d contact through our Tokyo office with Dr Yoshio Ueno of the Tocyo University of Science He expressed willingness to performofmost analyses we requested but we did not carry this through becauseed the l ogistics and communication difficulti es likely to be encounter analyses The laborator ies of ICI in England were suggested at least for DAS Attempts were made to contact Ors Grove and Aldridge but were abandoned because of finding the U S laborator ies and because of the co11111ercial implications as with Ralston Purina Thus the selection of the two laborator ies was arrived at after an international search in which all the experts in the field were evaluated Dr Jarvis was a clear choice for the macrocyclic analyses Dr Rosen washadequally clear for the T-2-type analyses especially after other laborator ies per-been eliminated for the various reasons and criteria indicated Both have data formed in the most professio nal manner and we are satisfied that their are the result of carefully planned and executed analyses using the most advanced methods and instrumentation available 2 Op 11 and sub-divid th• sampl• for si bmissio11 to the analytica l Zaborato n The sample carried by Hr Charles Whitney was received by Mark Goldman ofwith our laborator ies on October 23 19B1 The sa111 le was in a glass bottleportions a rubber stopper The contents were removed in a Bio-hazard hood and transferre d to five tared pre-weighed ampules and sealed with teflon-lin ed rubber stoppers and capped with al11111inum An Arthur D Little Inc andspecial safety plan for handling toatc chtlllicals was followed The receipt f11aed by your crew transfer 1111re • tnll 24 Page 5 L- -h • S $ _ nt - ghing of the pules gave the foll owi ng net contents Weight mg 21 3 3 4 s 20 6 15 7 17 7 12 1 Thus the total weight of the ori9inal sample as portioned out was 87 4 mg O 3 ounce On Noveillber 4 the vials were packed boxed and sealed and given to Ms Diane Rhodes for delivery to Ors Jarvis and Rosen Vials 13 and 4 to Jarvis #1 2 and S to Rosen Receipts were received by us indicating elivery to Or Rosen on November 4 and Dr Jarvis on November 5 Or Bruce Jarvis of the University of Maryland looked for macrocyclic trithothecenes in the sample anc concluded that there were none His letter report attached describes his findings Or Mirocha present using GCiMS found no evidence of any of the macrocyclic colfl ounds Jarvis' techniques described in a recent·publication Science 214 460-461 1981 were not designed to separate the simple trichothecenesTT-2 toxin etc so that analysis of his preparations would not necessarily be to confirm analyses done specifica lly for only these substances expected Rosen see below on examining Jarvis' Science paper has suggested Or that the simple trichothecenes were most likely to be still on the chromatog raphic column when Jarvis stopped collecting fractions because no UV-absorbing materials were coming off A higher concentration 20% of methanol han Jarvis used 2-6S would be required to elute T-2 and related materials Dr Rosen has reported tht presenct of three trichothec enes a non-trich othectnoid inycotoxin and a Polytthyleneglycol-derived substance in the sample The toxins art discussed In his letter of November 23 attached Thest Obs1rv1tlons ere pr1stntld 1n some technical detail have visited Dr Rosen Incl reviwed his rtsults in even more detail but We only present here llltwlt Ml btltt _ 11 t tl to an understanding of the analytica l ntcet1es 1nvol pulations is to prepare a solution hly enriched as possible with the susl1ct1d Ion Monitoring technique on the 25 Page 6 A nlll l L ctlc ire Ooe1mb1r IS 1981 -5- Mr St1v1 Singer Amerle4n Broadcasting Company mass spectrometer would reveal their presence The various steps involved -extraction with methanol removal of fatty mate r ials with hexane reve ephase chromatography on a Setpack col umn and thin-layer chromatography TLC for some of the suspected com onents -- narrow the likelihood that extraneous materials would be present • The gas chromatograph mass spectrometer GC MS system as employed by Rosen involves the use of a GC column for which the elution time is another diagnostic for the presence of trichothecenes and the use of an internal standard deuterated silyl derivatives of authent ic sa les of the suspected components of the unknown This is a potent method for the identification of fragments produced by the electron impact EI method of analysis in the mass spectrometer which sorts out molecular fragments according to their weight The deuterated reference materials come out of the GC column very close to the silylated materials from the sample and thus provide another cross-check on the Identity of the fragments observed For each of the trichothecenes there were two or three fragment peaks looked for on the mass spectrometer and in each case the results in comparison with the deuterated standard were convincing Thus we accept Or Rosen's conclusion that T-2 toxin vomitoxin deoxynivalenol and DAS diacetoxyscirpenol are present in the sample Zearalenone is a separate situation since Dr Rosen did not have a deuterated standard available for comparison However its level of occurrence was high enough to pennit a full-scale mass spectrum which provided conclusive evidence of its presence and the quantity present calculated by reference to authentic material Or Rosen attempted to detect nivalenol in the TLC cut containing the other materials using a reference sample supplied by Dr Mirocha but did not find positive indications Because he has not developed methods for nivalenol this is an uncertain finding The quantftation of trichothecenes identified in the sample is dependirnt on comparison of peak heights for the sample with those of the internal standard which is present in a known quantity Confinnation is obtained by deten iining whether the ratios of the heights of the various peaks are the same for sample and standard These criteria were met in all cases The calculated amount of each in ppm was zearalenone 228 T-2 toxin 48 OAS 42 and deoxyniva lenol 58 After the results on the aycotoxins were reported orally it became of Interest to detel'Wltne whit else was present in the sample as received since the toxins accowited for less than o os of the mass involved Discussions between A t 1 Inc Dr Rosen and ABC News su9gested a search 26 Page 7 l ret l1n lYCol•rel1ttd attrials This was indicated because PEG nly used as surface-active agents detergents a its eri at vts are c - lsifiers etc ind thus ight be •n e petted component of a mixture esi ed for spreading and persistence of active material as is sometimes d ne in pestici e applications i r The terial anal 2td as the s -ne solution fron the thin-layer chromato· Sand zearalenone The conclusions from gN which contained T•2 sen' s data are based on agreement with tion two properties of PEG after silyla- • There is successive loss of a repeating monomeric unit of 44 a ic- ss•units ar u equivalent to the ethol 'late group · ·OI -0- • Silyl ation of PEG or derivatives yields a fragment of 161 amu Ol3 3-Si-0-CH2-0l 2-0-CH2-0l 2-J Furthenrore if there is another substituent on the terminal nonsilylated end of the polyetho late fragments will be produced which decrease successively by 44 amu 60th the reference standard of PEG 400 average amu and the unknown show peas at 161 205 and 49 amu rt Presenting the silylated fragment plus one and two ethol 'late units see Figure In the spectrum of the unknown sample there are peaks showing the successive loss of 44 amu for example 279 235 and 191 in one GC peak see Figure and 478 435 391 346 302 and 258 in another From these data it cannot be determined what the original substituent on the polyethol 'late chain was but it is clear that the chain is losing fra Jllents of the size of ethoxylate and that something other thin a silyl group is attached Quantttation of PEG or its derivatives by mass spectrometry is difficult because any such p t contains 1 • ixture of polymers of various chainlengths which Ny tlutt fro• 1 coliai at different times and also fragment dtfftrtntly tn tht • ass spectro• ettr Dr Rosen however by analogy with tht results wtth PEG 400 1stt• 1t1s that the unknown sample contained a • int- of 1250 111111 of tht PEG•ltke • 1teri1 I rest11ct of PEG-based materials in the sample tly introduced during the handling of the It CIIIS tn110lved in preparation for the GC MS Ink sa- ilt and found no indications of 27 Page 8 · DL r ' re Arthur December 15 1981 -7- Mr Steve Singer Ameri can Broadcasti ng Company PEG that i s no peak at 161 amu nor any peaks showing increments or decrenents of 44 amu The preparation of this blank sample involved all se steps from ext raction of methanol with hexane through Setpack reverse-pha chromatography thin-layer chromatography silylation and gas chromatography Prior to the chemical processing all physical handling of the sample both in our laboratory and his was with the dry powder · The vials in which the sample was transmitted to him were new and not washed and had Teflon-line d stoppers We conclude that the PEG-like material was not introduced by any of the procedures used 4 Provide you r rith our assessment of the vaZi dity and significanc e of the resuZts insofar as possibZe • All four of the compounds found by Rosen 3 trichothece nes and zearalenone are known to be produced by Fusarium roseum Vomitoxin according to the review by Busby and Wogan 1979 has not been reported from F tricinctum It is thus possible if not ·necessaril y likely that this material could derive from a single culture source • The absence of the macrocyclic compounds is not surpris i ng since they have not been reported to be produced by the same micro-organisms as the T-2 type trichothecenes are This may be taken to indicate that•the sample is not a gross sample of soil or a contaminated natural material • The fact that the trichothece nes are found in nearly equivalent amounts is believed to be unexpected from a natural source They could be artificiall y produced tn these amounts either by manipulation of growth conditions in a single culture or by pooling cultures grown under different conditions • The presence of a polyethoxylate material PEG-like is completely unexpected in a natural sample It is a man-made chemical polyme r and the only one known to us which would give repeated fragment losses of 44 mass units as reported by Rosen If the PEG-like material is tntrtnsic to the sample the implication s of how and why it is there a st be considered Sincerely yours 7 faf fc 4 S Thayer Ph D Phili Consultant 28