Document jg1RYNVneNEamGR8v41x2VRLN

Monsanto Company Organic Chemicals Division St. Louis Research Department St. Louis Research Report No. 3067 PINAL REPORT ON ANALYTICAL CHEMISTRY AND SPECTROSCOPY INVESTIGATIONS - 1962 Part II - Special Studies Job No. 2-0?-760.01RPd^-II January, 1964 Work done and written by: R. S. Keller 0. W. Ashworth B. Katlafsky M. W. Dietrich D. B. Hines C H. Brackblll 0. Klnast 0. Hicks 0692835 HARTOLDMONOOQ5342 DISTRIBUTION OP REPORT NO. 3067 1. Pile 2. R. E. Keller 5. H. L. Hubbard - Pile 4. Duplicate Pile ..i.frfr. Central Technical Pllen 6. R. 5 Uelsman - J. P. QueenyPlant 7. W. Gresham - W. 0. Krummrlch Plant 6. MCL - England - via D. Danna 9. MCL - England - via D. Darina 10. Extra 11. Extra 12. Extra 15. Extra 14. Extra This report contains confidential Information which Is the property of the Monsanto Company and which shall be disclosed only to authorized persons. The recipient la held accountable for the filing and safe custody of the report which must be returned on demand. 0692836 HARTOLDMON0005343 TABLE OE CONTENTS Page Wo. INTRODUCTION 1 GENERAL SUWARY 1 ANALYTICAL STUPIE3 Section A - Aroclor 1242 Product Variation - Infrared and Oas Chromatography Study 2 Section B - Penzlmidar.olines - VfR Study of Substl tuted 10 Section C - Blolite Analytical Investigations 22 Suction D Plbutyl Phenyl Phosphate - Analysis of Reaction Products 34 Section E - Competitive Hydrocarbon Ester Fluids Identification of Inhibitors 35 Section F - Evaluation of Infotronlcs CPS 1 Inte grator for NHR Signal Integration 42 Section 0 - Hethyl Lactate - TO Determination in First Step Reaction Mixture 56 Section H Fentachlorophenol X-Ray Diffraction Studies 57 Section I - para-There*, tdlne I lent 1 fl cat inn of Contaminants 65 Section .1 Plasticizers snl Other High Bolling Esters Identification 67 Section K - Resorcinol in Sodium Sulfite Solution Ln'rared Stuly of Reaction Rate 66 Section L - Solium Pherate and Sodl :rt, t"tr?eeesulfonate - Attenuated Total Reflectance Infrared Study 99 Section K - Resorcinol Thin Layer "hromatography Investigations 111 Section N - Trlcosane Oxidation Study - Infrared Analysts 114 Section 0 - vanadium Oxide Components ir. 'atalyst Fixtures 116 0692837 J VTROD'JCT.'CN 1. Special Investigations were carries out. by the Analytical Chemistry and Spectroscopy Croup during 1962 to support Organic Research Depart ment and Manufacturing Department prefects. The studies also included certain projects listed In the September 29, 1961 FA ("nelassifled Anal. Chem. - 5a). SUMMARY Fifteen analytical studies covering Organic. Division products were carried out to provide Information reeled for Research Department and Manufacturing Department projects. Chemical and Instrumental techniques were . ed toi . show Aroclor 12A? product variations . characterize the structure of substituted bentImidazolines . evaluate Rtollte unien use situations . show the composition of dlbutyl phenyl phosphate reaction products " ........... ' . identify Inhibitors in competitive funct 1 onsl__flulds measure methyll__laotVe In process reaction mixtures . show the crystal form and translator of pentachlorophenol . Identify cont smlnanas In jj- pher.et ldlre . measure reaction n^es of resorcin-l In sodium sulfite solution . show the oxidation products form-1 f-om trl_cosne . show the components t ri vans 11 >m or.: 1e catalyst mixtures . evaluate an Jrifot.ror ! cs fRS-1 j ri* eg; ator for NMR . set up an anal/Mcai sphere for identification and measurement of jilastl elzers and oner high belling esters . demonstrate the potential ctilitv c' irfrered Attentuated Total Reflectance ss so Jnstresm monitoring device lor s o<jjurn pKerate end soil urn benzenes-1`q-ste "~ demorst.rate coin '.aye- r1 rir'/p.i1.ciV techniques. 0692838 HARTOLDMONOOQ5345 SECTION AROCLOR 1242 PRODUCT VARIATION - INFRARED AND OAS CHROMATOORAPHY STUDY 2. Introduction Infrared absorption and OLPC (gas liquid partition chromatography) methods were used to study compositional variations between five lots of Aroclor 1242 representing typical production material shipped to Oeneral Electric from the Anniston plant over a six month period. The study was Initiated to determine If abnormal variations were present and. If so, the cause of the difference. IR and OLPC are complementary techniques for such a study, since IR can detect gross Isomeric distribution differences and OLPC can Identify and measure the Individual Isomers. Summary An Infrared and OLPC (gas-llquld partition chromatography) study of five normal production samples of Aroclor 124? from Anniston showed that most of the variations In Isomer content were small and random. However, one pronounced periodic variation was observed In the Infrared spectra In the region of 14 5-14.4 microns which was shown by a combination of Infrared and OLPC analysis to be the result of variations In the lower chlorination level Isomers, namely the 2-chloro, 2,2'-dlchloro and the 2,41-dlchloroblphenyls. Analytical Studies Reserve samples of lots D-122, P-i45 and D-185 were obtained from Anniston to represent normal production material. Since no sample of lot D-155,the subject of a recent General Electric complaint, was available.sampleB of lots D-lb2 and D-1^4 were substituted. Infrared spectra were run under conditions such that the gross Infrared spectra could be compared rather than under special conditions to study a particular Isomer. The OLPC analysis was carried out using a 150' x 0.020" stainless steel capillary column coated with silicone gum rubber and a f'ame Ionization detector. A critical comparison of the IR spectra showed primarily the minor random variations expected for small differences In Isomer distribution. However, in the region of 14.3 to 14.4 microns, two weak absorption bands were observed whose intensities varied periodically with lot number. The attached IR spectrum shows the superlmposltlon of the spectra of the five lots of Aroclor 124? studied run under conditions to amplify the difference. Examination of the spectra in the region 0692839 HARTOLDMONOOQ5346 3 of 14.3-1*. 4 micron* shows that the Intensity of the absorption bands at these wavelengths Increases sharply from lot D-l?2 to D-14b- The absorption reaches a maximum In lot D-lb?, then tapei* off in lots D-1J>4 and D-185. The maximum variance observed betwian lots P-1?? and D-lbS was }Ojt In terms of optical absorption units. The OLPC data for the five samples shows the random variations of Isomers also detected by 1R. No variations In a single component was observed that coulJ explain the difference found in the 1R spectra. The OLPC data Is tabulated In Table I. Where possible, the OLPC fraction has been Identified as a particular Isomer on the basis of retention time. Despite the fact that the positive Identification or the com ponents Involved In the observed periodic variation In the 1R spectra could not be made, structural characteristics of the materials contributing to this variation could be assigned, based on Infrared interpretation of aromatic absorption patterns. The chlorine distribution patterns In Aroclor 1242 fit the following general classes: Type I Type II Type III Type IV Cl x Of these four general classes of chlorine distribution patterns, only Types I and II (the monosubstltuted aromatics arid the ortho dlsubstltuted aromatics) produce absorption bands In the 14.3-14.4 micron region. Although It Is possible to detect and differentiate the Types I and II materials In other regions of the IH spectra. Interferences and overlapping of bands prevents this l'n the bulx sample. 0692840 HARTOLDMON0005347 4. Ir. order to resolve this problem and Identify the components causing the observed variance In the IR spectra, lot D-152 was fractionally distilled by Nr. Qeorge Ashworth under reduced pressure and four top fractions .representing a total of S.AJf of the charge, were Isolated for subsequent analysis. The OLPC analytical data which was confirmed by IR analysis shows that the mono and dlchloroblphenyls having the type I and II structures postulated by IR Inter pretation of the bulk sample were the major contributors to the variation observed In the IR spectra of the bulk sample. The QLPC data Is tabulated In Table II. A combination of IR and QLPC analysis has demonstrated that the periodic variation observed In Aroclor 1242 lots D-122, D-l*5, D-152, D-1&4 and D-185 are the cumulative effects of variations primarily of the 2-chloro, 2,2'-dlchloro and the 2,4'-dlchlorobiphenyls. 2/62 - R. E. Keller, B. Katlafsky, E. H. Emery 06928*1 HARTOLDMONOOQ5348 5 Table Z AROCLOR 1242 Practlon/Component 1 Biphenyl 2 Ortho 3) Meta 4)Para 5 2,2* 6 2.5 7 (2,4 or 2,5 or 2,3*) 8 2,4* 9 (3,5 or 5,3*) 10 (5,4 or 2,5,2*) 11 4-4* 12 15 2,3,2* ' 14) 15) 16) 2,5,4* 17) 18 5.4,2* 19 2,5,4 20 21 22 25 24 25 , 26 5,4,4* 27 28 29 30 5312) 33 Lot D-'22 0.04 0.8 0.2 4.5 0.8 1.4 9-1 0.7 10.2 5-3 0.2 4.8 1.5 15-3 6.7 3-0 0.7 0.2 3.6 2.7 1.9 3-5 3-3 3.4 0.6 0.1 0.1 5.1 3-9 lot D-145 0.05 1.0 0.3 4.9 0.8 1.8 9-6 0.8 10.0 5-7 0.5 5-8 1.8 Lot . D-152 0.3 1.9 0.6 5-2 1.0 Lot D-154 0.2 1.8 0.5 4.8 1.0 1.9 10.5 0.6 9.4 5.2 0.3 5.1 1.5 1.8 * 9.6 0.7 9-5 5-3 0.4 5.2 1.4 14.9 6.6 3.0 0.7 tj3_.4, 2.4 1.5 3-3 3-2 3-2 0.7 0.1 0.05 4 .6 3.5 13.7 6.0 2.7 0.6 0.2 3-4 2.4 1.7 3-3 3-1 3-1 0.6 0.1 0.05 4.9 3.7 14.1 6.1 2.7 0.7 [3.9 2 .6 1.8 3.4 3-0 3-3 0.6 0.1 0.05 5-1 4.0 Lot D-185 0.05 0.9 0.2 4.6 0.8 1.6 9-5 0.7 10.1 5.4 0.5 5.7 1.6 0 6.6 3.0 0.7 0.2 3.5 2.6 1.8 3-3 3-2 3-4 0.8 C.l 0.1 4.9 3.8 0692842 HARTOLDMONOOQ5349 6 Fraotlon/Component 34 35 36 37 38 39 40 41 42 43 44 48 52 Lot D-12C 0.2 2.7 0.3 0.5 0.2 0.03 0.2 0.4 0.2 0.8 0.1 0.6 0.3 Table I (cont'd) Lot D-145 Lot D-152 Lot D-154 0.1 2.4 0.3 0.5 0.2 0.05 0.2 0.3 0.2 0.8 0.2 0.4 0.4 0.2 2.6 0.3 0.7 0.5 0.1 0.3 0.5 0.3 1.1 0.1 0.3 0.3- 0.2 2.6 0.3 0.1 0.3 0.05 0.2 0.4 0.2 0.9 0.2 0.5 0-3 Lot D-185 0.2 2.6 0.3 0.5 0.2 0.02 0.2 0.3 0.1 0.1 0.1 0.4 0.2 0692843 HARTOLDMON0005350  t: Table II ARCOCLOR 1242 LOT D-152 TOPPIKQ EX PERCENT Practlon/Component 1 Biphenyl 2 ortho 3) meta 4) para 5 2,2l 6 2,5 7(2,4 or 2,3 or 2,3`) 8 2,4i 9 (3,5 or 3.31) 10(3,4 or 2,5,2i) 11 4,4i 12 13 2,3,2i 141 15) 1$) 2,5,4i 16 3,4,2l 19 2,3,4 20 21} 22 5 23 24 25 26 3.M1 2d 29 30 31) 32 33 Starting F#1 F#2 F#3 F#4 Mat'l 1 -956 1.7* 2.3* 2.5* 0.3* 1.9 0.6 5.2 1.0 8.6* 29-7 6.5 36.8 2.8 0.8* 26.0 7-3 43.0 3.4 traae traoe 7.1* 1.3* 5-9 4.1 41.3 36.0 5*6 6.6 1-9 10.3 0.6 9.4 5.2 0.3 5.1 1.5 2.9 4.2 6.9 11.5 14.1 27-4 0.3 !j 0.8 1 0 0.3 ;1 i n [0 0 i.i 3-5 0.6 0.1 trace trace 0.3 0 9-2 35-0 *O 4 .7 1.2 0.1 C. 4 0 Residue 91.6* 0 0 trace 0.2 0.2 0.9 5-7 0.5 9-9 6.6 0.5 6.1 2.1 13.7 6.0 2.7 0.6 3.6 2.4 1.7 3-3 3-1 3.1 0.6 0.1 0.1 4.9 3.7 trace trace 17.6 7.7 2.1 0.7 ^5 3-1 2.0 4.1 3-8 3.8 0.7 U10 ,1 6.1 4.5 0692644 HARTOLDMONOOQ5351 -i*'\KlA\viDr\Kt-C\O* \ Ci*Tw\0 Table II (cont'd) 8. Starting F#1 F#2 F#3 F#4 Residue Fraction/Component Material 0.2 0.2 2.6 3.1 0.3 0.2 0.7 0.6 0.5 0.3 0.1 trace 0.3 0.3 0.5 0.4 0.3 0.2 1.1 1.0 0.1 0.1 0.3 0.3 0.5 * instrument shut down before this peak was eluted from the column. eH<\JKVfrCO ^ ^ -4- iCAM 0692845 HARTOLDMON0005352 9. l\-l *r 4i 2 100 80 0 40 20 SPECTRUM NO SAMPLE AKocloR izt/z ORIGIN PURITY PRoPUC 7" tOKi snPLfS PHASE UQUlD THICKNESS O.OS'nm. lUTStttz Mav-ISH aioTP-zvr s: 34.T t>./S2 DATE l/lf/bZ OPERATOR 0K remarks _Z8/? Trans, S_CA$ RF(!ivc6 prism N/a.C.1 RESOLUTION *?27 RESPONSE gain____ UL SPEED 31 SUPPRESSION V SCALE 1 0 cm. f/l- fr ------------ PIlC Rons CHT| tc. Rurr.io r * PR-1100 069284b HARTOLDMONOOQ5353 10. SECTION B NMR STUDY OF SUBSTITUTED BENZIMIDAZOLINONES Introduction Research to define the scope and limitation of S.^.A'-trlchloro- carbanlllde (TCC) as a soap bacterlostat has established that TCC loses Its bacteriostatic activity when used In conjunction with hypochlorite bleach. A chemical study of the action of hypochlorite on TCC Indicated that this loss of bacteriostatic activity was due to the conversion of TCC to a trlchlorlnated l-phenyl-2-benzlmldazollnone which Is biologically Inactive. This NHR study was Insti gated to provide additional evidence for the proposed ring system, and to determine the location of the 3 chlorine atoms In this system. Summary Proton NMR spectra were obtained from the product Isolated from the reaction of TCC with bleach, from a related monochloro compound obtained from the reaction of carbanlllde with bleach, the mono acetate derivatives of these compounds, and appropriate reference compounds. Analysis of these spectra Indicated the structure of the compound formed In the reaction of carbanlllde with hypochlorite ion to be - and the compound formed In the reaction of TCC with hypocnlorlte Ion to be - 0692847 HARTOLDMONOOQ5354 IX. "These compounds are shown as o-ecei-ate, but may be N-acetate. The benzimidazole derlvatlvesof II and IV were prepared by the following reactions (II) (IV) B. NMH Measurements All measurements were made on 0.50 K solutions in dimethyl acetamide unless otherwise state. Chemical shifts are reported relative to (CHs)Sl (WS) defined as *10*(Htks-H) i.e. 0. Peak area measurements, not reported, we?e 0692848 12. : made on all compound and reference peaks with a Varlan V-J521 Integrator, and are consistent with the proposed structural assignments to +< In all cases. C. Reference Compounds The NMR parameters of the reference materials examined are giver In Table I. As Is shown the peaks due to the two rings In the 1-phenyl-2-benzlmldazollnone are well separated. Unless complicated by spin-spin Interactions, this should also be true for the chlorinated compounds In question. The acetate derivative displays a well separated peak for proton C, that nearest the Introduced unsaturation. TABLE 1 NMR Data from Reference Compounds Compounds Chemical Shift (ppm) a________b________ c J (cps) NHa 7-05 6.57 ' Cl-^ ^ ^JCI rrro JAB - 8.7' Jab " 9*o 6.8l 6.58 7.11 JAB ' 8,5< JAC * 0; 'C-O 6.86 Indicates the average peak displacement; In all cases but peak (A) of the last reference - a single narrow peak Is observed. 0692649 HARTOLDMON0005356 1 13. Table 1 (cont'd.) NKR Data from Reference Compounds' 0692850 HARTOLDMONOOQ5357 D. Proof of Structure of (II) Cl 0 14. Other physical and chemical measurements Indicate the structure shown above, with 1 ohlorlne atom moleoule, but the location of the chlorine as to ring and position on ring Is uncertain. Figure All. Aromatic proton spectrum of (II) Three peaks are seen in this spectrum. Based on their absorption positions the two high field peaks are assigned to the benzlmidazollnone ring and the lower field peak Is assigned to the phenyl ring. Area measurements of these peaks show 3.0 protons to 5-0 protons. This shows positively that the single chlorine atom is on the benzlmazollnone ring. However, because of uncertain spin-spin coupling, 'no definite assignment as to the ring position can be made. Figure B) Aromatic proton spectrum of (V) From measurements of peak area and chemical shift, peaks 4,5, and 7 are assigned to protons in the benzimidazole ring; the remaining peak Is assigned to phenyl ring protons. The areas of peaks 4,5, and 7 are identical and correspond to 1 proton/peak. The analysis of- this first order spectrum for the benzimidazole ring protons leads directly to the indicated assignment. Peak 4 is assigned as Indicated because of the large shift to lower field observed when unsaturation Is Introduced one atom removed from this position. This then allows an unambiguous assignment of structure (II) as l-phenyl-6-chloro-2-benzlmldazollnnne. H 'C-0 0692851 HARTOLDMONOOQ5358 15. E. Proof of Structure of (IV) :0 A* in Structure (ll) the problem le to assign the position of the > chlorine atoms In the two-rir.g system. The nature of the reaction Is such that there Bhould be 2 chlorine atoms In 1 ring, and 1 chlorine atom In the other ring. Figure C) Aromatic proton spectrum of (IV) Four peaks of varying area are seen in this spectra. Based on their absorption positions the high field peak Is assigned to the benzlmldazollnone ring; the remaining peaks are assigned to the phenyl ring. Area measurements of these peaks Indicate one chlorine atom/benzlmldazollnone ring and two chlorine atoms/phenyl ring. No detailed analysis as to substitution positions was attempted. . Figure D) Aromatic proton spectrum of (VI) Seven distinct absorption peaks are seen. No Interpretation Is attempted but comparison to Figures B and C suggests a monochlorlnated benzimidazole ring and a dlchlorlnated phenyl ring. Area measurements are In agreement with this Interpretation. F. Solvent Effects Before trying to build a convincing argument for the chlorine ring positioning necessary to explain spectra C and D, an attempt was made to utilize the slight variation In chemical shift of o, m, and p protons to be expected In various solvents. Since a strong Interaction between solvent and the N atoms Is likely as well as between the aromatic IT electrons and solvent, noticable effects can be expected. These effects may r*move the overlap of peaks thought to be present In spectra C and D. Spectra of IV and VI in several solvents were obtained. Solvent effects on the observed spectra were observed. The spectrum of VI In dloxane gave the best resolved spectrum; this spectrum Is shown as Figure E. 0692852 HARTOLDMONnnn.'V^Q 16. Figure E) Aromatic proton apectrum of (VI) In dloxane (''.7 M). Comparison to spectrum B Indicates that peaks 4,5, and 7 are due to the aromatic protons In the benzimidazole ring while the remaining peaks are due to the phenyl ring protons. Determination of protons/peak by Integration and observation of repetitive peak separations, indicating spin-spin coupling, shows that each ring exhibits a first order spectrum. The Interpretation of this spectrum Is Indicated. Calculations using the observed chemical shifts and coupling constants Indicates the peak Intensities In the phenyl ring spectrum are as predicted. This allows an unambiguous assignment of Structure IV as l-(5,4-dlchlorophenyl'-6-chloro-2-benzlmldazolinone. 6/62 - R. E. Keller, M. W. Dietrich 0692853 | ft i 8 HARTOLDMON0005360 ///!*- fO$\Tw>/ (ffM ro ) 0692854 I HARTOLDMONOOQ5361 fACfbui T*, >'/****' pl*K f0SiT/^d (ft* TC ) 0692855 HARTOLDMONOOQ5362 0692856 HARTOLDMONOOQ5363 0692857 *i| HARTOLDMONOOQ5364 0692858 HARTOLDMONOOQ5365 22 SECTION C BIOLITE ANALYTICAL INVESTIOATIONS Introduction Basic data were needed to define the scope and limitations of Blollte as a bacterlostat. This study covered the following problems! 1. Development of i.:ethod for ppm pentachlorophenol and/or Blollte on cloth. 2. Stability studies - a. Blollte In detergents. b. Coated Blollte In detergents. c. Experimental funglstat In detergent. 3. Retention studies - a. Blollte on cloth from Terg-O-Toneter washes. b. Blollte on cloth from full scale washing machine washes. 4. Miscellaneous studies - a. Effect of sterilization of cloth on Blollte analyses. b. Stability of Blolite-detergent-water at l4o*F. c. Determination of Blollte by reduction with sodium biphenyl and measurement of the liberated chloride. Summary 1. A method has been developed for the determination of pcntachlorophenol and/or Blollte (2-chloroethyl pentachlorophenyl car bonate used as a bacterlostat-funglstat) on cloth. The method Is based on the hydrolysis of Blollte to pentachlorophenol followed by extraction with chloroform and measurement by ultraviolet absorp tion In alkaline aqueous solution. The method sensitivity Is a few parts per million Blollte and the accuracy Is believed to be within * 1056 of the amount present. 2. Stability tests run at 80*F. In the humidity cabinet on ljt Blollte (99% assay and <100 mesh) In All showed that approximately 50S6 of the Blollte remained after two months. Approximately 70% remained after two months with 90% assay, 20-100 mesh Blollte In All, Fab and Tide. 0692859 HARTOLDMONOOQ5366 23. Accelerated stability tests (one week at 60*C. which the soap manufacturers claim equivalent to a one-year shelf life) were run on 1% Blollte In Tide. The Blollte was coated with various materials which might prevent decomposition. A 10% coating of polyvinyl alcohol Increases the stability of the Blollte in the presence of the detergent from 4ojf to 80j. Accelerated stability tests shewed approximately the same decomposition for an experimental phenyl pentachlorophenyl carbonate as for Blollte. 3. The Terg-O-Tometer which Is used to determine the effective ness of various detergents In laundering was used to evaluate Blollte In the detergents. Cloth laundered In the Terg-O-Tometer and air dried contained approximately 17 ppm Blollte while the samples washed In an automatic washer and dried in a commercial drier contained 100 to 140 ppm Blollte. Cloth washed In the Terg-O-Tometer where the wash as well as all the rinse solutions are poured (filtered) through the cloth to simulate the full scale washer gave 60 to 160 ppm Blollte on the cloth. These findings Indicate that the Terg-O-Tometer Is not equivalent to standard washing machines for studies of this type. The analysis of cloth from ten successive washes showed no build-up of Blollte. Three full scale washer experiments resulted In the following concl"slons: (1) Less than B ppm (calculated as Blollte) Is present on cloth as pentachlorophenol after the wash, the first and second rinse and the air drying steps. This represents about 1% of the total Blollte present. (2) The pentachlorophenol content of the heat dried cloth Increased to 20 ppm or about 18# of the total Blollte present. (3) Upon sterilization, the pentachlorophenol Increases Ir. the washed and dried samples to &0 ppm. About 10% of the Blollte Is hydrolyzed. This shows that the biological testing was carried out with cloth containing 10% pentachlorophenol and ~}0% Blollte. (4) The explanation for twice the Blollte content In heat dried samples In comparison to the air dried samples Is not known. Analytical method error and/or difficulties in sampling due to the non-uniform distribution of the Blollte may account for this. 0692860 4. Analyses show that sterilization does not change the recovery of known amounts of Blollte added to cloth. Approximately 5# of the Blollte In detergent In wash solution Is decomposed In 10 minutes at l4o*F. The state of subdivision of the Blollte affects this decomposition. 16% decomposition occurs In 10 minutes If the Blollte Is added as a solution In acetone. The reduction of Blollte with sodium biphenyl and titration of the resulting chlorides with standard silver nitrate has been used as an Independent method to check the validity of the extractionultraviolet method. Both methods give essentially the same results for the analysis of cloth from washing tests. Analytical Studies 1. Development of Method for ppm Pentachlorophenol and/or Blollte on Cloth '' Preliminary results were obtained using a method based on extraction of the clotn in a Soxhlet extractor with ethanol. The optical brlghtener offers serious interference with ultraviolet measurements. Attempts to effect purification by extraction pro cedures proved unsuccessful due to emulsions. Extraction of the pentachlorophenol from the cloth with chloro form In the presence of water eliminates most of the above difficulties. Quantitative extraction is slow because of the entrainment of the chloroform by the cloth and because equilibrium In the extraction Is established slowly. This requires p minutes of vigorous shaking. To accomplish equilibrium on the Wrist Action Shaker, the samples must be shaken at least 1 hour lb minutes and preferably 2 hours. All the pentachlorophenol is found In the chloroform layer so that If a known amount of chloroform has been added, quantitative analysis can be made on a measured aliquot. Blollte does not Interfere if the solution Is acid. The chloroform solution Is washed with 10 ml. 0.1 N sodium hydroxide and the ultraviolet absorption of the aqueous alkaline phase Is measured at 520 nyj. Por the determination of Blollte and pentachlorophenol, the cloth Is moistened with 0.1 N methanol-HjO-sodlum hydroxide. After standing a few minutes, the hydrolyzed sample Is acidified and the pentachlorophenol extracted In the usual manner. There Is very little background Interference using this procedure. The reproducibility of the method is believed to be within + 10jC. The method Is as follows: 0692861 HARTOI DMDNnnn^RR METHOD NO. 62-4 Determination of Blollte and/or Pentachlorophenol on Cloth Method # 3 SCOPE The method la designed to determine the Blollte and/or pentachloro phenol present In cloth at the level of 5 ppm or higher. It la a modified and Improved version of Method # 2 developed earlier. PRINCIPLE The pentachlorophenol Is extracted with chloroform directly from the cloth suspended In water, the chloroform washed with water and finally with a small volume of dilute sodium hydroxide solution. The ultraviolet absorption of this alkaline solution Is measured at J20 nju where sodium pentachiorophenate has maximum absorption. Blollte does not Interfere. The cloth Is moistened with alcoholic sodium hydroxide and the Blollte Is decomposed Instantly to pentachlorophenol. The total pentachlorophenol Is determined. The Blollte Is the difference between the total pentachlorophenol and the pentachlorophenol found directly. REAOENTS Chloroform - A.R. Sodium Hydroxide - A.R. 50* Sol'n. Hydrochloric Acid - A.R. Methyl Orange Indicator Methyl Alcohol - A.R. APPARATUS Burell Wrist Action Shaker 1 - 250 ml. Separatory Funnel 5 - 125 ml. Separatory Funnels ... Cary Recording Spectrophotometer with 2.0 cm. Fused Silica Cells 8 oz. Narrow Mouth Bottles with Polyethylene Closures Usual Laboratory Olassware STANDARD SOLUTIONS 0.1 N Sodljm Hydroxide - No. 1 Add 2.6 ml. 50* sodium hydroxide to 500 ml. distilled water and mix thoroughly. No. 2 Add 2.6 ml. 50* sodium hydroxide to 250 ml. distilled water and add 250 ml. methyl alcohol. 0692862 HARTOLDMONOOQ5369 METHOD NO. 62-4 Ipg. 2J PROCEDURE A. Pentachlorophenol 1. Take a 5.0 g. sample of cloth and cut Into suitable slzi pieces or strips and place in a clean dry 8 oz. bottle. 2. Add 150 ml. of distilled water and 6 drops of 1-1 hydrochloric acid. 3. Add, by pipette, 2? ml. of chloroform and shake vigorously for 30 seconds or until cloth Is dispersed throughout water and chloroform. 4. Clamp in a horizontal position on the Wrist Action Shaker and shake with maximum shaking action for 1.5 hours (8 samples may be shaken at one time). 5- Removethe cloth from the bottle with the aid of a wire hook. Squeeze the cloth against the neck of the bottle during the withdrawal to minimize the amount of chloroform discarded with the cloth. 6. Transfer the contents of the bottle to a clean dry 250 ml. separatory funnel. 7. Drain the clear chloroform layer Into a clean dry 25 ml. .'-aduate and record the volume to the nearest one tenth of a ml. 8. Place 25-30 ml. of distilled water Into each of two 125 ml. separatory funnels. 9- Pipette 10 ml. of 0.1 N aqueous sodium hydroxide (sol'n. No. 1) Into a third 125 ml. clean dry separatory funnel. 10. Transfer the chloroform from the graduate quantitatively, with the aid of a little cnloroforn, to the 1st wash water funnel, stopper and shake for 15 seconds. 11. Drain the chloroform Into the second wash water funnel. 12. Add approximately 25 ml. of fresh chloroform to the first wash water funnel. 13. Shake both the first and second wash funnel for 15 seconds (these can both be shaken at the same time;. 0692863 HARTO L D MON6005370 METHOD NO. 62-** [pg. 5) 14. Drain the chloroform In the second wash funnel Into the third funnel containing the sodium hydroxide (# 9 above). 15- Drain the chloroform from the first wash funnel Into the second wash funnel. 16. Shake the second wash funnel and the third funnel containing the sodium hydroxide for 15 seconds (these can be shaken at the same time). 17. Discard the chloroform layer from the alkaline funnel. 18. Transfer the chloroform from the second wash funnel to the third funnel containing the alkali. 19. Shake for 15 seconds, and discard the chloroform. 20. Apply vacuum to the top of the separatory funnel and open the bottom stopcock to allow air to bubble up through the solu tion for several minutes or until all the chloroform has been removed (odor). 21. Pour the solution Into a 2 cm. cell and measure the absorp tion In the range of 400 to 233 iyj using distilled water In the reference. 22. Draw In the background which Is a straight line extension of the curve from 400 to 350 njp. The absorbance measured at the maximum of 320 Is corrected for the background from the curve at this wavelength and the difference used In calcu lating the results. Blollte and Pentachlorophenol - Total 1. Taxe approximately a 5 gram sample of cloth, weighed to the nearest 0.1 gram, cut Into suitable size pieces or strips and place In a clean dry 8 oz. bottle. 2. Add 10 ml. of 0.1 N sodium hydroxide In 50^ methanol-water mixture (sol'n. No. 2). Stir the cloth around with a stainless steel spatula to make certain that the cloth Is thoroughly wet with the alcoholic sodium hydroxide solution. 3. Rln3e off the spatula with water using a total volume of 150 ml. Add 1 drop methyl orange Indicator solution. 4. Add approximately 6-10 drops 1-1 hydrochloric acid solution and shake. The indicator snouli be red, showing the solution Is acid. 5. Proceed with the analysis starting with Procedure A, Penta chlorophenol, # 3 0692864 HARTOLDMONOOQ5371 METHOD NO. 62-4 Ipg*) CALCULATIONS ppm pentaehlorophenol - Vol. of Chloroform (0. D. at 320 - Sol'n. Extracted Vol.of 0.1 N NaOH O.D. at 320 of background) x 23 x Taken In ml.______ x ID4 (Absorption coefficient4 of ~ Cell length ~TTT pentaehlorophenol at 320 nyj) In cm. x sample In g. Corr. O.D. at 320 x 10 x 10* x fraction chloroform extracted 200 x 2.0 x 50 Corr. O.D. at 320 x 30 x chloroform fraction * Absorption coefficient - (g./leffml?) [fiellHeSgCK.-cnirT PRECISION AND ACCURACY No data were obtained for statistical calculation of accuracy or pre cision. The five extracts give better than 90)6 recovery with known pentaehlorophenol from cloth at the 30 ppm level. The reproduci bility Is within 1 ppm at the 30 ppm level. The accuracy and repro ducibility are believed better than with Method # 1 and equivalent to Method # 2. DISCUSSION Pentaehlorophenol can be quantitatively extracted from water solution by chloroform In one 13 second extraction. However, In the actual extraction of the pentaehlorophenol from the cloth, only 6056 Is ex tracted In the one minute time. Quantitative extraction Is obtained using 30 minute shaking. No Intermediate shaking times have been run but a 3 minute shaking would probably give quantitative transfer from the cloth to the chloroform solution. In shaking the cloth, water and chloroform, with 23 ml. chloroform, about 21 ml. of chloroform la the mo3t that can be drained forward. This means that the practical recovery 13 only 83# of the amount present. Thus, two additional extractions are needed for quanti tative transfer of the pentaehlorophenol. An alternative procedure could be used where an aliquot of the chloroform Is washed and the pentaehlorophenol transferred to the sodium hydroxide solution for UV measurement. This Is the basis of the present method. 0692865 HARTOLDMONOOQ5372 METHOD NO. 6?-4 Ipg. $J When extractions are shaken by hand, 5 min. Is sufficient for complete extraction whereas 1.5 hrs. 13 required using the Burell Wrist Action Shaker. The shaking action must disperse the cloth throughout the solutions,otherwise the extraction will be Incom plete in 1.5 hrs. The purpose of the 1 minute vigorous shaking Initially Is to thoroughly disperse the sample before the mechanical shaking Is started. Monsanto Chemical Company Organic Research Department St. Louis, Missouri 12/62 - 0. W. Ashworth, R. E. Keller 0692866 HARTOLDMON0005373 25- 2. Stability Studies a. Blollte In Detergents The stability of 1)6 Blollte In standard All and dry mix All was determined. The samples were prepared In Individual 8 oz. bottles by accurately weighing in 0.2 g. Blollte and adding 20.0 g. of the detergent. At the end of the test period, the entire sample was used for analysis. The dry mix All contained no brlghtener or perfumes. The control sample was kept tightly capped while the sample (cap removed),was placed In the humidity cabinet. The results of the first series are as follows. The Blollte used was 9956 assay and less than 100 mesh. Percent Blollte In Sample Time of Test Standard All Control Humidity Dry Mix All Control Humidity Initial 1 day 2 days 3 days 1 week 2 weeks 4 weeks 8 weeks 0.97 0.91 0.88 0.87 0.91 0.81 0.81 0.83 0.90 0.86 0.82 0.77 0.67 0.63 0.46 0.98 0.89 0.82 0.84 0.74 0.58 0.38 0.31 0.87 0.85 0.84 0.78 0.65 0.61 0.33 The results Indicate moisture may be a contributing factor In the decomposition. A second series of tests were run using 90% Blollte [90% pentachloro - the balance being the tetrachloro analogue of Blollte) with mesh size of 20- 100 . Several detergents were used In this series. The :results are given In the following table Analyses of Blollte Heavy-Duty Detergents Up Through Two Months (Containing 1$ Blollte after 3 days, 1 week, 2 weeks, and 1 month at 80*F. In the humidity cabinet.) Percent Blollte 3 days 1 wk. 2 wks. 1 mo. 2 mo. All - 9996 Blollte - 20-100 mesh All - 90% Blollte - 20-100 mesh Fab - 90% Blollte - 20-100 mesh Tide- 90% Blollte - 20-100 mesh Cllmalene - 90% Blollte - 20-100 mesh Beads O'Bleach (1) - 90% Bio- llte - 20-100 mesh Snowy Bleach (2J - 90% Blollte - 20-100 mesh 0.97% 0.92 0.90 0.90 0.92 0.90 0.91 0.93X 0.85 0.91 0.90 0.92% 0.84 0.85 0.88 0.92 0.90 0.95 0.95 1.01 1.00 (1) Assay corrected by 0.20% as blank correction. 2; Assay corrected by 0.29% as blank correction. 0.83S 0.77 0o..e749 0.6956 0.65 0.70 0.70 0.85 0.76 0.90 0.78 0.79 0.69 0692867 HARTOLDMONOOQ5374 26. The above results do not show as nuch decomposition as when the fine mesh (< 100) Blollte was used. The results at the two-month period show that the decomposition Is approximately 70$ for all the detergents. The 4(4 obtained on the first series Is pro bably due to the fine mesh size of the Blollte. b. Coated Blollte In Detergents . Dr. J. Baker had 90% Blollte coated with various coating agents at the Wisconsin Alumni Research Foundation. This Blollte was tested by an accelerated test used by the soap manufacturers. The sample Is heated at 140*P. (6o*C.) for one week. It Is claimed that this test Is equivalent to a one-year shelf life. The analytical data for this test using the method based on measurement of CO* are given In the following table. Heat Stability Studies % Blollte In Detergent ___ Sample Originally At 60 C. 1 wk. 9036 Blollte - 20-100 mesh 1.00 0755" 99% Blollte - 20-100 mesh 99% Blollte - <100 mesh 1.00 1.00 0.59 0.24 Control - Blollte used In sub- sequent tests 1.00 0.56 9% Carbowax 6000 0.95 0.56 10$ Carbowax 6000 0.90 0.56 2.5$ Methocel - Ethocel 9% Methocel - Ethocel 10$ Methocel - Ethocel 0.975 0.95 0.90 0.59 0.58 0.51 2.5$ Carbowax 6000 Methocel- Ethocel 0.975 5$ Carbowax 6000 + Methocel- 0.55 Ethocel 0.95 0.56 7.5$ Carbowax 6000 + Methocel- Ethocel 0.925 0.55 10$ Carbowax 6000 + Methocel- Ethocel 0.90 0.54 2.5$ Polyvinyl Alcohol 5$ Polyvinyl Alcohol 7.5$ Polyvinyl Alcohol 10$ Polyvinyl Alcohol 0.975 0.95 0.925 0.90 0.54 0.62 0.71 0.76 5$ Sugar 10$ Sugar 15$ Sugar 20$ Sugar 0.95 O. 90 0.85 0.80 0.48 0.49 0.51 0.48 25$ Sugar 0.75 0.59 % of Original Remaining 46 59 24 56 40 51 40 51 46 46 48 49 49 45 65 0.92) 77 0.95 84 0.91) 51 54 60 60 52 5$ Dextrin 600 0.95 0.52 55 Remarks $ Blollte In original formulation Coated Blollt' Stlcky-neceosary to grind before formulatlng. 0692868 HARTOLDMON0005375 27. One percent formulations in Tide of the 5. 7-5 and 10# polyvinyl alcohol coated Blollte samples were analyzed by the regu lar procedure. The values obtained - 0.92 In place of 0.95; 0.93 vs. 0.925 and 0-91 In place of 0.90 - are In close agreement with the theoretical values. The other samples were not analyzed. The percentage of the coating was assumed to be correct and this value was used In determining the percentage of the original remaining. As a check on the uniformity of the coating, three Indi vidual samples (0.2 g. ) of the 10i6 coated mateilal were weighed out, dissolved In approximately 25 ml. of alcohol and 10 ml. of 0.1 N sodium hydroxide were added. This hydrolyzed the Blollte to pentachlorophenol which was determined by measurement of Its ultraviolet absorption with an appropriate dilution. The assay figures obtained by this method were 87-0, 86.0 and 86.3# Blollte. These are In close agreement as well as with the 90# nominal figure and the 91# obtained by the method Involving the measurement of carbon dioxide. Of the coated materials tested, polyvinyl alcohol appears to be the most promising. There Is a trend showing less decomposition as the amount of the polyvinyl alcohol Is Increased up to the 10# level. c. Experimental Funglstat In Detergent The accelerated test - one week at l^o'P. - was run on 1# phenyl pentachlorophenyl carbonate In Tide. 33# of the funglstat remained after one week. This is approximately the same as obtained for Blollte. 3. Retention Studies a. Blollte on Cloth from Terg-O-Tometer Washes Experimental washings were run In the Terg-0-Tometer to te3t the effect of part'cle size of the Blollte. An average value of 17 + 2 ppm Blollte was found on the cloth while cloth laundered In an automatic washer contains 50-100 ppm. Thus, the Terg-0-Tometer cannot be used to duplicate the results of a full scale washer. To check the deposition of Blollte on cloth by filtration, eleven circular swatches of cloth were washed In the Terg-0-Tometer but the samples were arranged on a Buchner funnel after each cycle In the same order and the solution poured through the swatches after the wash as well as after tl.e first and second rinse. Blollte < 100 mesh was used on one test and 20-100 mesh material was used In a second experiment. The results are given In the following table. 0692869 1 % I I HARTOLDMON0005376 28. Cloth Sample No. Prom Top to Bottom ppm Total Blollte <100 Mesh 20-106 Mesh 1 8* 166 2 66 131 10 50 61 11 31 60 The range of 60*166 ppm Is higher than the 17 ppm obtained when the Terg-O-Tometer la operated In the usual manner. The data Indicate that Blollte la retained for the cloth by a filtering action during standard washing machine tests. This work shows that the Terg-O-Tometer testB must be changed In order to simulate the retention of Blollte on cloth laundered In a standard automatic washer. b. Blollte on Cloth from Full Scale Washing Machine Tests The analysis of cloth from ten successive washes showed no build-up of Blollte with each successive wash. The average amount found was 6 ppm which Is so low that there Is a large experimental error In the determinations. Only one-eighth the normal amount of Blollte ws3 used In this series of experiments. (0.35^ Blollte In detergent and 0.25# used In the wash.) The results of three full scale washer experiments carried out to show the retention of Blollte on cloth are summarized In the following table. 0692870 1 1 Cloth Sample wash 1st rinse Pnd rinse Washer #1 Penta- chloro- Total phenol as as Blollte Blollte , (PPm.) 8 109 8 5 Washer $2 Total aa Blollte Std. Method Tv. (ppm.) (Ppm.) 54,l4?,6l 5?,75,88 75,75.01 86 + J8 7?+~13 06+10 Std. Method (prm.) Washer #3 Na bi Re- phenyl Av. analysis Method 1PCt-1 (ppm.) (ppm.) 2nd rinse sterilized Air dried 7 >0,49 70,97 70,68.66 84 . 1J 68 1 58,58,56(2) 37 1 61.57 86 A*r dried sterilized 82,8?,79(1) 81 t 1 Air dried sterilized 8r no 8H.66,70'11 75 i 7 112,64,75 85 19 Air dried sterilized 63 hrs. at 76* Meat dried ?0 Heat dried f sterilized 76 64,69 116 156,99,107 67 i 5 1U + lb 159,157,148l5) 1*11 * *1 1?9 no 1?5,106,100 11? + 7 156,166,160 16? 5 1?0 (1) Two sets of samples analyzed. (?) Re-analysis of the samples gave 13,9,7 ppm. total. (3) Re-analysle of the samples gave 9 ppm. Blollte - essentially all Blollte removed. ro vo 0692671 HARTOLDMONOOQ5378 >0. Cloth from the first washing experiment was analyzed for both pentachlorophenol and Blollte. Samples taken up through the air dried step contained no more than 8 ppm pentachlorophenol or 7$ of the total Blollte. The air dried sterilized sample contained 2 ppm pentachlorophenol or about 70$ of the Blollte. The heat jrled sample contained 20 ppm Blollte or 18$ of the total Blollte. The funglstat activity of all samples were thus determined on cloth con taining 70$ of the Blollte present as pentachlorophenol as all samples were sterilized before running the biological tests. In Washer Test # 1 only one sample of each was analyzed. Tne erratic values make It difficult to draw conclusions. The air dried sample contains about one-half the Blollte found In the steri lized sample and the heat dried and sterilized sample. Washer Test i 2, involving triplicate samples, was made to determine the reproducibility of the sampling. The large difference between air dried and heat dried samples is again noted and unexplained. There Is good agreement between the sterilized and unsterlllzed air dried samples. The reproducibility of the wash sample before rinsing Is poor. This value does not decrease rapidly with two rinses which shows the tenacity with which the Blollte Is held by the cloth. This cannot be called adsorption because, If it were, the amount on the cloth would be the same when using the Terg-O-Tometer under similar concentrations. The results of the third washer experiment show a maximum variation of four-fold between air dried arid heat dried sterilized samples. Two additional samples of the air dried were run at a later date and the values of b7 and fcl agree nicely but do not agree with the average of V + 1 obtained originally. A check of the analyti cal method disclosed no serious errors in the procedure. A very non-uniform deposition of the Blollte on the cloth might contribute materially to a non-representatlve sampling error. The analysis of an additional sample of heat dried cloth agreed better with the value obtained previously. The reduction of Blollte with sodium biphenyl and titra tion of the liberated chloride with standard silver nitrate wa3 developed and used 83 an Independent method of analysis. Results showing 66 ppm and 120 ppm for air dried and heat dried samples from Run #3 confirm that the difference In Blollte content of air dried and heat dried clcth Is real and that the data obtained with the extraction-ultraviolet method are of the correct order of magnitude. Samples of the air diled and heat dried clotn were re analyzed and only a small amount of Blollte was found. This shows that all extractable Blollte was removed from the cloth by the analytical method. 0692672 HARTOLDMONnnn.^ya 51. 4. Miscellaneous Studies a. Effect of Sterilization of Cloth on Blollte Analyses To determine the effect of sterilization on the analysis, a series of experiments were run where a known amount of Blollte was added In acetone solution to a swatch of cloth and this, In turn, was analyzed. The swatches used were unwashed, washed with water only and washed with Tide. Residual amounts of detergent or other Impuri ties left on the c'oth were considered as possible sources of trouble with the method. The results given In the following table show that reco veries are generally better than within _+ 10% of the amount added with outside variations of + 20%. Effect of Sterilization of Cloth on Blollte Analyses Sample and Treatment Pla11 in " " CloII th it it - Unwashed Unwashed Unwashed Unwashed - Sterilized Acetone Acetone - M II M " " it ti it ti n it Sterilized - Unwashed . Unwashed - Sterilized Unwashed - Unwashed - Sterilized - Washed with Water - Washed with Water - M n N Acetone - Washed with Water - Washed with Water - 11 N n Sterilized Washed with Water - Washed with Water - it Sterilized - Washed with Tide It it Bleach - Washed with Tide II n Bleach + Acetone - Washed with Tide + H ii Bleach - Washed with Tide Bleach - Sterilized " Washed with Tide W it Bleach - Washed with Tide + Bleach - Sterilized Blollte Added as ppm 0 0 0 0 50 50 100 100 0 0 50 50 100 100 0 0 50 50 100 100 Total Pentachlorophenol and Blollte Reported as ppm ______ Blollte 4 5 4 5 49 57 94 105 0 0 66 56 122 107 0 0 67 56 116 75 0692873 ___________ j HARTOLDMON0005380 32. b. Stability of Blollte-Detergent-Water at 14q*F. A study to determine the extr.t of degradation of Blollte In All during the washing step at l4o*P. was carried out. In one experiment, the Blollte was added to the detergent-water solution (ljS) at l4o*P. as a solid. In the second experiment, the Blollte was added In an acetone solution. Vhls wash solution turned turbid Immediately and stayed turbid more than 10 minutes. This probably Is caused by excess Blollte precipitated In colloidal form which hydrolyzed faster than Blollte added as a solid. The results In the following table show that less than b% of the Blollte would be decomposed In the normal 10-mlnute wash cycle. Percent Blollte Decomposed to Fentachlorophenol Time In Minutes Added as Solid Added In Acetone Solution 52 10 6 20 10 -30 16 A0 22 60 29 60 32 90 46 120 67 11 16 24 31 38 44 48 66 74 c. Determination of Blollte by Reduction with Sodium Biphenyl and Measurement of the Liberated Chloride The sodium biphenyl method for the analysis of Blollte on treated cloth consists of the following operations. A 5.0 g. cloth sample (diced; Is covered with 30 ml. chlorine-free xylene and stirred for several minutes. A 10 ml. aliquot of sodium biphenyl reagent Is added and the sample mixed 6 minutes. The excess sodium Is reacted with 1 ml. ethyl alcohol. Five ml. of fresh hydrogen peroxide is added and the sample Is boiled 10 minutes. After cooling, 100 ml. acetone and 1 ml. strong sulfuric acid (l to 1 concentrated sulfuric acid to water by volume) are added. The sample Is stirred thoroughly and the solution Is decanted off. The cloth is extracted four more tlme3 using for each extract 6 ml. of distilled water, 100 ml. acetone and 1 ml. 1 to 1 sulfuric acid. The separate extracts are collected and titrated with 0.0025 N silver nitrate to potentlometrlc endpoint using glass-silver electrode system and an L & N expanded scale potentiometer. Analytical data obtained using this procedure are reported In the following table. 0692874 HARTOLDMONOOQ5381 33. Sample 1. Xylene Solution of Pentachlorophenol 2. Xylene Solution of Pentachlorophenol Plus 6-0 Orams Control Cloth 3- Same as Sample 2 Ml. Sodium Biphenyl 5-0 5.0 10.0 Equivalent Biol tc- Added Found vPPml 100 98 100 98 100 77 Remarks CloM th AbsMent 6.0 Oram Cloth Control Present 100 93 6.0 Gram Cloth Control Present A. Treated Cloth Air Dried, 3rd Washer 10.0 -- 86 36,38,57,61 ppm Blollte by UV Method 5. Treated Cloth Heat Dried, 3rd Washer 10.0 - 120 129,137,14b ppm Blollte by UV Method This work was carried out to provide data for studies under Job No. 4162 - "Soap and Detergent Bacterlostats". Messrs. K. L. Godfrey, D. P. Roman, J. W. Baker, J. P. Speyrer and J. R Wiedemann carried out the Terg-O-Tometer and full scale washing tests, the special sample preparation. Including the sterilization studies, and the preparation of coated Blollte which are given In this report. 12/62 - R. E. Keller, D. B. Hines, 0. W. Ashworth 0692875 HARTOLDMONOOQ5382 SECTION D P31 NMR Analysis of Process Reaction Producta Introduction In the development of the new dlbutyl phenyl phosphate process, analytical methods were needed to evaluate the two major steps of the process: (1) P0C1* + phenol, and (2) (0O)POC1 + butanol. In both cases several by-products are formed along with the major product. A method was needed to characterize the reaction mixtures and to determine the yield of the desired product. Summary P31 NMR methods were developed to analyze mixtures from both steps in the process. Several by-products were identified. Major and minor components were determined quantitatively. The following components were identified in the reaction mixtures examined: P0C1*, (0OjPOCla, (0O)aPOCl, (0O)9PO, (Bu0)9P0, (BuO;,(0O)PO, (BuO;(0O)2PO and two pyro compounds. The developed methods will measure 0-100# of a given component, and are considered accurate to about _+ 2# absolute [based on total phosphorus containing compounds). However, it was not found possible to analyze accurately for both P0C19 and (0O)POC1* since the P31 NMR peaks of these materials overlap. A combined analysis of these materials can be made. The developed methods provide a fast, comparatively accurate, non destructive method for quantitative analysis of sample mixtures of phosphorus compounds. The various components are identified by inspec tion of the spectra. The method Is unaffected by components which do not contain phosphorus. Analytical Studies The developed methods are based on the difference In chemical shift of the various phosphorus containing components. In a mix ture these components were Identified using chemical shift data ob tained from the pure materials. Quantitative analysis was performed by measurement of the area of each peak, the area being proportional to the concentration of the component. Samples from the laboratory and pilot plant were examined, covering both steps of the process. The data obtained were used to optimize yields and to determine the nature and quantity of Impurities present. Data and additional details can be found on Analytical Data Sheets filed in the Analytical and Spectroscopy Oroup. 0692876 HARTOLDMON0005383 METHOD NO. 62-33 NMR Determination of Reaction Products from the Dlbutyl Phenyl Phosphate Process A. Step I - Reaction of POCls with Phenol SCOPE This method describes a phosphorus NMR procedure for the determina tion of the amounts of the various phosphorus components formed In the reaction of POCls with phenol. The phosphorus components can be determined to about 2 mole % (of the total phosphorus compo nents) In the range 0-1005<. PRINCIPLE The P*1 NMR spectrum of a typical reaction mixture (see Plgure 1) contains peaks for each of the phosphorus containing components pre sent. These peaks were Identified by reference to spectra of the pure materials. The concentration of each component Is determined by measuring the area of the peak associated with each component and comparing to the area of all peak3 in the spectrum. APPARATUS AND REAOENTS High-Resolution NMR Spectrometer capable of measuring PS1 (a Varlan HR-60 was used). Equipment for measuring area such as a planlmeter or electronic Integrator. PROCEDURE 1. Instrument Settings Choose optimum settings to give signal display most convenient for the Integration method to be used. 2. Calibration No calibration is necessary. 0692877 HARTOLDMONOOQ5384 METHOD NO. 62-32 Ipg. 2} ?. Sample Analysis Pill mm. thin walled NKR cells with sntple (add CC1 to reduce viscosity If necessary). Adjust spectrometer for maximum homo geneity and minimum drift. Introduce sample and record spectrum (and Integral). Add HPO capillary to sample, rerun and obtain spectrum with scale calibration. Calibrate spectra and Identify components by chemical shift values. Determine area of each component by means chosen. Sum up areas of all peaks In spectrum. Determine the mole % of a given component using the equation: Mole 0 X Area X 1______ lf.n nx ' Total Area ` luu where n number of phosphorus atoms per molecule Total Area - ^ Area 1 nl where 1 - the phosphorus compounds present If both POCls and (0O)POCla are present, the amounts of these components are reported as a sum of the two concentrations, since the peaks of these species are Incompletely resolved. Area measurements were made by planlmeter in the examples reported here. B. Step II - Reaction of (0OjPOCl with Butanol SCOPE This method describes a phosphorus NMR procedure for the determina tion of the amounts of the various phosphorus components formed In the reaction of (0O;POC1 with butanol. The phosphorus components can be determined to about + 2 mole i absolute (of the total phos phorus components) In the range 0-1000. PRINCIPLE See Section A. A NMR spectrum obtained from a typical sample Is shown In Figure II. APPARATUS AND REAGENTS See Section A. 0692878 HARTOLDMON0005385 PROCEDURE 1. Instrument SettInga See Section A. 2. Calibration No calibration is necessary. >. Sample Analysis See Section A. METHOD NO. 68-32 Ipg" 3) Monsanto Chemical Company Organic Chemicals Division St. Louis Research Department 12/62 - M. W. Dietrich, R. E. Keller 0692879 HARTOLDMONOOQ5386 0692860 HARTOLDMONOOQ5387 f<*E . I a T<v H| 1M M *.( r, ft,-'''. t> J 'C 0692881 SECTION E 3S Idenf 1 flcat'on of Tr.hl tit ora In Competitive Hydrocarton Ester Fluids jw rod uctlon This study was undertaken to characterize the oxidation Inhi bitors present In competitive turhofar fluids. The Information was needed to better select an Inhibitor for a Monsanto Co alkane product. . The techniques applied In the 1 lentlf 1 cat.Ion were column chroma tographv, ultraviolet absorption spctroscopy, Infrared spectroscopy and nuclear magnetic resonance spectroscopy Summary Esso, Texaco and Socony turboban fluids were found to con'aln one or more components with an aromatic secondary amine structure. The findings suggest components of the type - P i (A-?6?i (A-19b) Analvtlcal Studies Following Is a chart of the work carried out. ai30 attached are the ultraviolet spectra of the three turbofan fluids examined and the typical ultraviolet spectra of the suggested components. 0692882 E*per i me?.!*) r'<rpeU 1 ve _Turtof an PluMs Structure Data for Inhibitor L--7** Bssq______ "li--7'^7 "Ttxsco . L-793 Socony ^ Column Chromatography Ultraviolet. Malarial "Par 3e" Two aromatic inhibitors Spectra characteristic Spectra characteristic Spaotra characteristic of H ____ i) p-w-^ of " H ------ - H :) R N_<^ K-R of a mixture of - V?-/ 1J J Aromatic system (allph^tl'' sromatto tertiary amine) (hlsubstltuted aroma tic secondary amine "or^ho" position phenyl e*>e diamine type) It) Arom*tic system Infrared Chromatographic Fractions P <y\* A-J96 NMR Chromatographic Fractions and Material "par 8a1 J.T) Aromatic s) Multi-'system with per* and trl-sub^ stltutl^n. J b) MjHI system - all 3 trl--substitution ,2 c) Heterocyclic nitro gen system, ho -prl. or sec. amine carbonyl or ether oxygen. T*o aromatic proton One aromatic proton One aromatic proton signal, signals, 1 and II. signal, 1. but different from I and 11. II, ratio aromatic Signal strength Indicates hydrogen to aliphatic 1/3 amount present In hydrogen, at least 3/l L 795 and L-?*2 Conclusion: The fluids contain one or more components as Inhibitors with an aromatlc_secondary amine structure. The findings suggest components of the type ................. ' ' A J n It; 0692683 I HARTOLDMON0005390 0692884 HARTOLDMONOOQ5391 t'-'i 0692885 i| HARTOLDMONOOQ5392 1I 0692886 I HARTOLDMONOOQ5393 o 0692687 HART0 L D M O N0005394 0692888 HARTOLDMONOOQ5395 SECTION P Evaluation of Infotronlcs CRS-1 Integrator for NMR Signals MS. Introduction The Infotronlcs CRS-1 Digital Chromatograph Readout System Is designed for Integration of signals obtained from a gas chromatograph. It has been found to be very satisfactory In this use. Since It possesses potential advantages of speed and accuracy over other methods presently available for Integration of NMR signals, It has been evaluated for this use. Summary The Infotronlcs Corporation Model CRS-1 Digital Chromato graph Readout System has been evaluated as a method of Integrating NMR signals. This Instrument was evaluated with and without signal preamplification using both the signal taken directly from the Varlan HR-60 NMR spectro meter and the base line stabilised signal obtained from the Varlan Model V-?5?l integrator. Two major limitations In usefulness of this Instrument, were found. Difficulty In zeroing the integrator before Integration was observed when using the signal obtained directly from the NMR spec trometer as well as a tendency of the zero position to drift. These difficulties were minimized by using the base line stabilized signal from th< Varlan Integrator. The second limitation observed was a tendency to discriminate against peaks with a low signal to noise ratio. In a sample containing both a strong and a weak peak, the weaker peak may either be omitted, or an improperly low integral value recor ded. With sufficiently concentrated samples of high signal to noise ratio, this instrument was found to perform well, being both accurate and rapid. 0692889 HARTOLDMONOOQ5396 Analytic*! Stadias A. Description of the Infotronlcs Integrator The Infotronlcs Model CRS-1 Integrator Is designed specifically as an adjunct to laboratory gas chromatographs. The Instrument Is intended to produce a printed record of the areas of the chromatographic perks, with the time of emergence of each peak shown opposite the recorder area. Input to the Integrator Is an electrical voltage of 30 milli volts nominal maximum (100 millivolts Is permissible); the time integral of the Input voltage Is produced, the recorded value being 100 times the value of the integral In millivolt-seconds. Positive Input voltages only are integrated. Structurally, the Integrator Includes two components, the Integrator proper and the printer which produces a printed tape listing the integrals of successive peaks. Functionally, the Integrator compon.nt In turn contains two principal elements, an integrating unit and a peak-sensing unit. The integrating unit comprises a voltage-tofrequency converter and a counter; the peak detector Is essentially a differentiating element which detects changes of slope Indicative of the start, top and end of a peak. The sequence of operations of the Integrator and printer Is as follows: Input Operation Start of peak (positive slope exceeding selected level). Start counting, Top of peak (zero slope). Transfer time reading to printer. End of peak (negative slope less than selected level). Transfer accumulated count to printer, print count and time. Controls for operating the Integrator manually are also provided, and the printer Includes a control for manual printing of total counts for a series of peaks. B. Connection of the Integrator to the Kra Spectrometer The radio frequency unit of the NMR spectrometer provides outputs for both oscilloscope and recorder. Tne arrangement of these (when using phase detection of the radio frequency signal).Is shown In Figure 1. 0692690 ( V HARTOLDMON0005397 F M. i The total realctar.ee ir. series with the spectror.etcr output Is less at the oscilloscope terminal than at the recorder terminal, so the CRS-1 Integrator was first connected directly to the oscilloscope ternlnal. Initial trials, with the voltage-frequency converter and slope detector inputs of the integrator connected together, showed an apparent large aero offset. This was found to be caused by an alternating voltage present m the slope detector input. (The slope detector includes a magnetic amplifier energized at the line frequency.) Connection of a capacitor across the slope detector input served to bypass the alternating voltage present here, but to avoid excessive damping of the oscilloscope it was necessary to add a resistance in series, a3 shown in Figure 2(a). Because the input resistance of the voltage-frequer.cy converter Is ICO kllohms and thef input resistance of the slope detector is 2 to po kllohns, depending upon the sensitivity setting, the result of using this circuit was a decrease of Integrator sensitivity combined with an apparent decrease of recorder and oscilloscope sensitivity. A similar result was obtained by connecting the integrator directly to the recorder terminal of the spectrometer; the shunt capacitance at that terminal (see Figure 1/ provided the required a.e. bypass. A possible means cf avoiding a.c. feedback from the slope detector and simultaneously reducing loading cf the spectrometer output circuit was to place an Isolating amplifier between tr.e voltagefrequer.cy converter and the slope detector. A transistor emitter follower was tried, as shown ir. Figure 2^,; it was effective for isolating the slope detector from, the integrator input circuit, but the transistor base current was sufficient to produce an objectionable offset of the zero of the voltage-frequer.cy converter. A more elaborate isolating amplifier -as tried in the form of a Leeds ar.d Northrup No. 93ji-2 electrcric D-C null detector. It was connected as shown in Figure 2(c,. Tr.e 1 m.cgohm. resistor In series with the input to the null cetector served to reduce the sensitivity of this unit to tr.e required level. It also slowed the response of the null detector; a voltage divider might have teen preferable for reducing the input voltage, ir. general, how ever, performance of the apparatus assembled as in Figure 2(c) was satisfactory. Because the null detector r.as beer. bar. owed from tr.e manufacturer ar.d had to be returned after a short time, some alternative amplifier was sougr.t. Tr.e alternative chosen was tr.e use of a Leeds ar.d Nor t hr -p No. 7^*01 pH meter as a preamplifier. Tr.e connections were as shown ir. Figure J. The good linearity and low output resistance of tr.e pH meter permitted It to be used to drive- both tr.e voltage-frequer.cy converter and the slope detector. As set up, tr.e range cf tr.e pH meter was - 50 millivolts. A range perhaps twice tr.is wide would have teen Fetter to prevent tr.e meter being driver, off-scale on large peaks. In other respects, performance of the pH meter as a preamplifier was satisfactory. :j ; 0692891 HARTOLDMONOOQ5398 *5. With all of the circuits tried, It was difficult to obtain a suitable compromise between sensitivity to small peaks and sensitivity to r.olse. The effect of large amounts of damping was not tried In the present work, but presumably the discrimination between noise and spectral peaks could be Improved by use of greater damping with correspondingly slower sweep rates. A filtered output is available from the spectrometer at the recorder terminal, as indicated Ir. Figure 1. To make the test use of this output, it would be necessary to apply a preamplifier because of the high internal resistance of the spectrometer and the lew resistance of the integrator. C. Evaluation 1. General Comments All measurements were made with the zero of Ir.fotronlcs integrator adjusted so as to give a counting rate of <10 counts/sec. (manu facturer's recommendation). The effect of higher zero settings is discussed later. By recording the CH.- quartet in CK3CH2OH at various sweep rates, the resolution of ti.. Integrator was determined to be better than one peak every 2.0 seconds. Tr.e manufacturer reports or.e peak every 1.5 seconds and that resolution car. be Improved considerably If data Is recorded on magnetic tape. Accurate phase settings appear unnecessary, see Figure A. 2. Operation without Prear.pllflcatlor. With the Infotronlcs integrator connected to the recorder output of the HR-60 spectrometer a sample was examined which contained two peaks in t!ie area ratio 6.0/1.0. Signal Intensity was varied by adjusting n?. The results obtained are given In Table I; increasing total count Indicates Increasing signal. Table ~ Effect of Signal Intensity or. Aocura Total Counts or. Integrator 1p,000 + cOO 35.000 + 1,5:0 53.000 " 2,000 162,000 + 10,000 Observed Area Ratio* Infotror.tcs integrator 7lar.lr.eter o-o t.tl + 0.^5 5-90 6.08 p.83 + 0.20 5.55 + 0.30 ----------- * Actual area ratio 6.00 When the base line stabilized signal obtained from the Varlan V-3521 Integrator (spectrum position) was used as In Input signal for the ir.fotronlcs integrator, the same biasing against weak peaks was observed. 0692892 I I I? % ''hARTOL DM ON0005399 46 5* Operation with Prear.pltflcatlon . pH Meter At low counting rates (weak signals) t!-.e sarr.e bias against weak peaks was observed. At high counting rates a bias against stro:.0 peaks wa6 observed apparently due to over loading of the pH meter. For these reasons, no detailed study was made. b. P. C. Null Meter A synthetic mixture was prepared containing components of 63.1, JO.O and 6.9 mole percent phosphorus. Signal Intensity was varied by adjusting nt, signal to r.olse ratio was varied by adjusting the frequency response. Representative spectra are found in Figures p-S; the data obtained are summarized In Table II (attached;. The results obtained are considerably better than those obtained with no preampllflcatlon, but the bias against weak signals remains. To determine the effect of background level on accuracy, the relative areas of several of t.ve multiple! components In the P31 spectrum were obtained at various background readings. ~s the background was Increased, the area ratio decreased approaching the correct area ratio (with and without back ' ground subtraction,. However, at the high background level needed to obtain the correct area ratio, the background was a major part of the total count. Since peaks with different line widths would have different backgrounds, no simple method Is available to accurately subtract the contribution of the background from the observed count. D. Surrested Operating Procedure a. Connect recorder output of Co to input, of a preamplifier. b. Connect output of preac.plifier to input of Infotrcnics integrator. c. After warm-up of equipment, adjust zero setting of Infotronies integrator to give a counting rate c-f less than ten counts/second (adjusting the red probe paooles if necessary;. d. Switch function control switch of Infotror.lcs integrator to automatic (integration;. e. Record spectrum and integral. f. Press total button on calculator. (If spurious peaks are recorded, adjust the sensitivity control of the Infotror.ios integrator to . give decreasing sensitivity and repeat steps c-f., 0692893 I ' t <? ji. i i l 1,1 ' i HARTOLDMON0005400 47. E. Feasibility and Recommendations Under the operating conditions Investigated, the Xnfotronlcs CRS-1 Integrator Is not satisfactory as an lnstruaent for general use In Integrating EUR signals. In certain cases of samples with strong signals. Its operating speed nay prove useful In obtaining area measurements fro* a large set of similar sasg>les (where Its accuracy on one of the samples can be Independently determined). 12/62 - N. W. Dietrich, W. M. Trump, R. E. Keller 0692894 TabU II Data Obtained Using 0. C. Null M tor Instrument 6b in.,, m - 70 6b in.,, HR - 4 5p 6b in.,, na - l 70 6b in.,, HR - 1 --------- A^'t^Ti---------- ---------------ItTn?Srrflif eKntor------------------ - T5~HiimroT57~ A 61.l B _C_ ;o.o 8.9 A 62.2 . 1.3 63.4 t 0.6 62.8 t 1.2 TT 29-T 0-7 30.2 o.a >0.4 o.a TT 8.1 + 0.8 6.4 + 0.5 6.8 + 0.6 A 64.9 60.7 60.0 B 25.1 30.8 29-7 C 10.0 8.4 10.3 67.1 0.4 29.7 . 0.4 3.3 > 0.2 63.7 26.6 9-7 ri nl \ !lf 0692895 l HARTOLDMON0005402 Pig. 1. Output Circuit of V*>311 Rartlcpreiucncjf Unit (HR Phase Detector). h I' if? A.' 0692896 HARTOLDMON0005403 O- VVNA*++OK <b o- i \ 0 /> /W 0pf*S f* *.*. //vm r^ o **,W. o- d+itr*e A>r //^*r <S 0" .....^V1- ' " ^ oT 0- Fig. 2. Circuits for avoiding feedback of s.c. from slope detector into Integrator. 0692897 HARTQLDMON0005404 Pig. 3- Connection of pH noter as prcaaplirie'r. 0692898 HARTOL D M O N0005405 0692899 HARTOLDMON0005406 (fitioxSo 1 r 0692900 HARTOLDMON0005407 0692901 >/ i. HARTOLDMON0005408 HARTOLDMON0005409 56 SECTION 0 NMR Determination of Methyl Lactate In First Step Reaction Mixtures Introduction In the development of a synthetic lactic acid process. It mss found that an analytical method Mas needed to determine the methyl lactate (KL) In first step reaction mixtures. NMR spectra of the pure materials, as well as typical reaction mixtures, were obtained and analyzed. It was found that an NMR peak specific for the methyl of the ester In the presence of methyl alcohol and residual lactonltrile could be used for analytical purposes. Summary A method was developed based on the measurement of the area of the methyl peak of the ester In ML. The weight percent ML Is determined from a calibration curve obtained from area measurements of standard mixtures of ML and methyl alcohol. With this method it Is possible to determine the weight ML covering the range 0-100jl with an accuracy of about + 1J< absolute. Other methyl esters are the only components expected to Interfere with this method. The method developed is preferred to a gas chromatography pro cedure when precise data are required. The NMR method Is less rapid than the gas chromatography procedure and It will probably not be the method of choice where many samples and the need for less precise results are Involved. Analytical Studies t'o analytical studies were carried out. The VPC method developed was chosen for routine analysis for the reasons given above. The NMR method will be used to provide more accurate data when needed, such as confirmation of doubtful VPC results. 0692903 f HARTOLDMONOOQ5410 METHOD WO. 62-2? NMR Determination of Methyl Lactate In Flrat Step Reaction Mixtures SCOPE This method describes a proton NMR procedure for the determination of the concentration of methyl lactate In reaction mixtures con* talnlng methyl lactate, methanol, and Impurities at relatively low concentration levels. The methyl lactate can be determined to about + 1% absolute In the range 0-100%. PRINCIPLE The proton NMR spectrum of methyl lactate (see Figure I) contains peaks for the OH, H and two types of CH* present In methyl lactate. The Bame spectrum (OH peak shifted) Is observed for reaction mix tures containing methanol and Impurities (see Figure II). This analysis Is based on the peak area of the methyl group of the ester. The other peaks In the methyl lactate spectrum cannot be used since Impurities expected to be present would Interfere In these measurements (other methyl esters would Interfere with the measurement described). APPARATUS AND REAOENTS Hlgh-Resolutlon NMR Spectrometer (a Varlan HR-60 was used). Methyl Lactate - Reference of at least 98 purity. Methanol - reagent grade. Synthetic mixtures covering the concentration range of Interest. The mixtures used were: Weight, % Methyl Lactate Methanol ?5.0 40.2 45.9 47.9 51.7 64.3 0692904 HARTOLDMONOOQ5411 METHOD WO. 62-22 lPg.2) PROCEDURE A. Instrument Settings The aettlnga given ere for e Verlan HR-60 Spectrometer equipped with V-3521 Integrator and a Moaely Autograph X-Y Recorder. V-4311 RP Unit: 80 db In Recorder Level: Maximum Receiver Oaln: 3 Frequency Reaponae: WL0 Detector V-3521 Integrator: Input Level: 650 Output: Coarae X.l Pine X.l Frequency Reaponae: 20 cpa Autograph XY Recorder: X: 100 aec. full acale Y: 500 mVo - full acale Slow Sweep Unit: Increase 1 x 100 Sample: 5 mm., spinning B. Calibration Pill precision NMR cells (Varlan or equivalent) with the prepared synthetic mixtures. Adjust the spectrometer carefully for maxi mum homogeneity and minimum drift. Using the reference mixture with highest concentration methyl lactate, record the Integral curve. Make any adjustments necessary so that this Integral Is near recorder full scale. Also, make any adjustments neces sary to optimize the appearance of the Integral curve. In sequence, run the set of standard mixtures making spectro meter adjustments when necessary. Obtain as many Integral traces for each sample as needed to give the desired measure of precision. On each Integral trace, draw parallel lines through the base line before and after integration of the ester methyl peak In methyl lactate. Measure the distance between these baselines for each trace with an accurate ruler. Calculate the standard deviation (If desired) for each standard mixture. On coordinate paper, plot the areas of each sample against the percent, by weight, methyl lactate. Draw the best smooth curve through these points. The working calibration curve set up for the method Is shown In Figure III. 0692905 HARTOLDMONOOQ5412 METHOD MO. 68-22 CptTJT C. Sample Analysis Prepare samples, obtain and analyze spectra by the procedure described for the synthetic mixtures. (If the samples and standards are run on different days, one or more of the standard mixtures should be re-run and the results compared to those obtained previously. If necessary, appropriate corrections In the areas obtained for the samples should be made.) Compute the peak area for each sample and equate this to the weight percent methyl lactate using the calibration curve. Monsanto Chemical Co. Otganlc Research Department St. Louis, Missouri 8/62 - M. Dietrich, R. E. Keller 0692906 HARTOLDMONOOQ5413 '* ?(!' HARTOLDMONOOQ5414 0692908 HARTOLDMONOOQ5415 0692909 HARTOLDMONOOQ5416 57. SECTION H X-Ray Diffraction Studies of Pentachlorophenol Introduction Caking and sublimation problems with plant pentachlorophenol were thought to be caused by crystalline fora transitions. X-ray diffraction studies with a Norelco X-ray Diffractometer were carried out to provide basic data about the crystal forms and their transformation. Special sample preparation techniques were developed to overcome particle size and preferred orientation problems. A heated sample holder operating In the range of 25-100*C. was constructed to permit direct measurement of crystal transitions too rapid to be followed by Intermittent sampling. Summary Results show that different diffraction patterns are given by pentachlorophenol when held above or below the transition tempera ture of approximately 60*C. The transition Is reversible. The more complex diffraction pattern of technical pentachlorophenol Indicates the presence of one or more crystalline Impurities. Quantitative measurements In the 29 range of 52.0* to 54.5' were used to evaluate effects on the transition rate by chemical Impurity, storage form, storage temperature, and crystal seeding. These studies Indicate that the crystal change in Itself Is not the cause of caking of technical "Penta" In storage. Details of this work not reported here appear In Research Progress Report "*504, No. 1, entitled, "Pentachlorophenol" by M. E. Olbbs, R. W. Bucknell and D. B. Hines. 0692910 f>y: ! Sr HARTOLDMONOOQ5417 Analytical Studies A . Development, of Techniques 1. X-rav diffraction studies by pressed disk techniques Problem of sample preparation were encountered In Initial X-ray powder diffraction experiments with pentaehlorophenol. The crystals are soft and It Is difficult to grind them to the fine powder (200 mesh or smaller) required for powder diffraction work. The crystals are also highly susceptible to align ment along crystal planes (preferred orientation). Accurate determination of relative intensity of the various diffraction peaks necessary for quantitative analysis depends on uniformly small particles with completely random orientation. It was found that l?0 mesh "Peers" powder could be pressed into smooth, dense disks at pressures as high as 16,000 pounds per square Inch, the maximum pressure developed by the laboratory Carver Press. Three grams of powder pressed in a 1 3/8" diameter, non-evacuable die gave a disk approximately 2 mm. thick. This disk can be substituted for the specimen holder In the Horelco Diffractometer and Its pattern readily measured. This technique, however, produced diffraction patterns exhibiting extreme preferred orientation with all peaks highly suppressed except for the one at 12.2* 20. Putting this technique on a quantitative basis appeared rather difficult, and alternative solutions were explored. However, the pressed disks look promising for future applications as a means of Increasing sensitivity for a minor component or perhaps In the suppression of the diffraction pattern of an lnterferrlng component. 2. X-ray diffraction studies with nu.lol mulls of pentachlorophenol The problems of particle site reduction and preferred orientation in "Penta" diffraction samples were largely o 0692911 HARTOLDMON0005418 59- overcome by preparation of a mull of Penta and NuJol. The mull la prepared by weighing 0.2 grams Hujol Into a 2 ml. stainless steel WIG-L-BUG grinding vial. Pants is added equivalent to exactly 3.5 times the weight of NuJol. Prilled Penta may be used directly In pre paring the mull but flaked Penta was pre-ground on a Wiley intermediate mill to 40 mesh. After the grinding vial is charged, two 1/4" stainless steel jails are added. The vial Is capped and the contents ground for 3 minutes In a model 3110B WIG-L-BUO. The contents of the vial after grinding are transferred to a Norelco standard aluminum powder diffraction specimen holder and a plane sample surface Is prepared by pressing with a microscope slide or other smooth flat-surfaced object. The diffraction pattern Is recorded using the following Instrument settings on the Norelco Diffractometer and Electronic Panel: 33 kilovolts, 23 mllllamperes, medium milllsnpere range control, 2 ratemeter scale factor, 0.8 multiplier, 4 time constant, proportional counter, 1675 counter voltage. If the total diffraction pattern Is of Interest the sample Is scanned over a 20 angle range of 10-50* at a rate of 1* per minute. For measurement of the ratio of the two crystalline forms present, the 20 regions of 32.0 to 34.5 are recorded at a scan rate of 1/4* per minute. The peak height ratio of the intensity peak at 32.7* 20 to the peak at 33.0* 20 was used as an empirical representation of the amount of the higher temperature crystal form present. The work necessary to put the method on an absolute basis was considered Impractical because of the difficulty of preparation of standards with known ratios of crystalline forms. 3 X-rav diffraction studies at variable temperature conditions In the exploration of means to accelerate the crystal form transition, a study was made of the transition rate at temperatures Just below the transition temperature. A heated specimen holder. Fig. 1, was constructed to permit measurement of the change In diffraction pattern with time as the temperature was varied. The special 0692912 HARTOLDMONOOQ5419 holder was milled from a solid 5/8" x 1 5/8" type 2024-T4 rectangular aluminum alloy bar. The holder la heated with a Watlow QIE2* Cartridge Heater, 100 watt, 115 volt, available from the Watlow Electrle Kfg. Co., St. Louis, Mo. The raised "Island" In the sample cavity prevents flssurlng of molten samples upon solidification and also houses a No. >0 tefloninsulated, lron-constantlne thermocouple probe. The temperature Is registered on an Leeds and Northrup Speed-0-Max Model H recorder equipped with a 0-200*C. lron-constantlne scale. Electrical connection of heater and thermocouple were made through a hole drilled In the diffractometer x-ray scatter shield. Pig. 2. Experimental Findings 1. Confirmation of crystalline form transition Preliminary experiments carried out with pure penta chlorophenol and with Monsanto and Relchhold technical "Penta" confirmed that exposure of Penta to a temperature of 75*C. for several hours or melting it gives a product with a changed diffraction pattern. Examination of samples undergoing transition shows several Intensity peaks characteristic of both crystal forms and their ratio changes with time. The two regions In the pattern where this change is most evident is the 20 range of 11.5 to 1J.0* and 32.0 to 34.5*. The 11.5* to 13-0* region had limited use In quantitative measure ment because Intensities here seemed more affected by sample surface reproducibility. However, this region was favorable for study of samples showing marked preferred orientation such as pressed disks or solidified melts because of the greater sensitivity. 2. Effects of purity on transition rate Study of transitions at room temperature of solidified melts of pure pentachlorophenol and technical "Penta" from Relchhold and Monsanto exhibited sharp rate differences. Pure pentachlorophenol underwent essentially complete transition In about 4 hours where as Monsanto and Relchhold Penta still showed sub stantial amounts of the high temperature fora after 14 and 20 days respectively. 0692913 HARTOLDMON0005420 61. >. Effects of storage form on transition rate Monsanto Psnts stored as flakes at room temperature showed essenttniy no transition from the high temperature form In a five day period following flaking. But a portion of the same sample ground to a 40 mesh powder Immediately after flaking underwent nearly complete transition at room temperature In the same five day period. . Monsanto flaked Penta stored for '9 months contained small amounts of the high temperature form. Hujol mulls on pure pentachlorophenol In the high temperature form showed nearly complete transition within the forty five minute period required to prepare the mull and obtain Its diffraction pattern. Transition of a solidified melt of this same material took four hours. 4. Exploration of practical methods to accelerate crystal fora transition The study of the transition rate of pure pentachlorophenol In the high temperature fora showed essentially no change In 3 hours time while maintained at 55*C. which Is Just below the transition temperature. This same sample underwent nearly complete transition at room temperature In about four hours. 5. Transition of Monsanto Prilled "Penta" Freshly prilled Monsanto "Penta" showed only slight transition from the high temperature fora over a 30 hour test period signifying that this physical fora doss not appreciably accelerate the crystal form transition. C. Future Applications of X-ray Diffraction Techniques This Investigation Illustrates several X-ray diffraction techniques that can be used to investigate crystal structure problems of the type described In this report. 0692914 HARTOLDMONOOQ5421 62 1. The powder diffraction pattern la characteristic of a pure crystalline material and la useful In a) making Identification, b) determination of purity, o) examina tion for crystalline changes brought about by decompo sition or reaction to form different crystalline pro ducts, d) following appearance or disappearance of a crystalline phase, e) measuring transformation of a given substance to a different crystalline form, f) formation of solid solutions or amorphous phases, etc. 2. Measurement of selected diffraction pattern Intensities ' often permits Quantitative measurement of one or more crystalline components, the extent of a crystal form transition, rates of changes of crystalline phases and transition temperatures. >. The special techniques of sample preparation devised should be useful In future studies of organic crystalline matter. 4. The production of diffraction patterns under variable temperature conditions should have future utility. 0. Acknowledgements Acknowledgement Is gratefully given to the following persons for helpful suggestions and support In this problemi 1. VI. M. Trump for recommendation of heating and temperatuie measurement components. 2. To A. J. Blndbeutel for construction of the heated sample holder. J. V. W. Saeger for suggestions pertaining to preparation of pressed disks of pentachlorophenol. 4. E. D. Plerron and R H. Munch for suggestions on sample preparation by the Nujol Mull Technique. 12/62 - D. B. Hines, R. E. Keller 0692915 HARTOLDMON00054?? HHTISk *1 6>. x-Rav oiPPRAcrtou hsatbO sample, uql&bk V-- SAKAPue v-jeu. (A) . IMMJ t* THE**AOCOUPl.fc VJRLL (ft) ' EN& (IUSO /ettAAi r MEATfcfc (LACTPlDeR WELL <.) .QPRON,r, C*/r* k----------------------------------------------------------- 5j SCALE: 1 INCH HARTOLDMONOOQ5423 64. l 0692917 HARTOLDMONOOQ5424 i SKCT10N 1 ldentll lcatlon of Contaminants In pata-Phcmtldlnc Intrqducl Ion In Un continuous hydrogenation of jr-phenotldlne, a precipitate war. for mod that caused a quality problem In the W.O.K. -phenctldlnc. Analytical at.Umptu to Identify the organic structures of the contaminant.-, in the j>-plu net Id Inc were made by application of Infra red spic t r or copy These Identifications were attempted using a sample of polymer extracted from -phenetidlne that had been returned from Nltro. Summary Solid state Infrared spectra (as KBr discs) showed the following structural characteristics of contaminants in i>-phcnolldl n- 1. Secondary amine 2. Aromatic nucleus Aliphatic hydrog.n 4. Ethoxy group L>. j>nra-:.ut:itltu < d f r om.tt.lc and the possible additional structures1. nltro group 2. 1,2,4-trlsubstltuted aromatic ring. Other techniques - NMR, thin layer chromatography - were unsuc cessful. Analytical Studies para-PhenetIdlne Contaminant Sample - Polymer from pat a - Phene t Id lne, b/24/62, R. Schubert (from material returned from Nltro;. Solubl llty - Estimated less than soluble In water, b<n..tr,i. , chloroform, acetone, acetone i HC1, dloxane, benzene, DMF and totinhydrofuran. Molecular Weight - Average molecular weight range Is VJ? to 600 using Menzles apparatus and boiling toluene. 0692918 HARTOLDMONOOQ5425 o6. Number of Components - At least 3 or 4 as determined by solubl1lty and thin layer chromatography. Structures pound - Determined by Infrared using KBr pressed discs of the samples. Ro satisfactory solvent was found for a nuclear magnetic resonance study. ' il) -N-, secondary amine , aromatic nucleus C-H, aliphatic hydrogen -O-CcH,, ethoxy group tre-substituted aromatic J03slblc Additional Structures 6. -NOp, nltro group r f ' H , !,?, trl subr.tl t.ut *-d aromatic rlr.g P^o? - R. S. Keller, P. Katlafsky 0692919 HARTOLDMONOOQ5426 67SECTION J Identification of Plasticizers and Other High Bolling E3ters Introduction A meeting of Research Supporting Technical Services personnel was held to discuss our current ability to analyze mixtures of plasticizers. The conclusions drawn also apply to other high boiling esters such as functional fluid base stocks. The three techniques having wide applicability are Infrared absorption spectroscopy, nuclear magnetic resonance spectroscopy, and gas chromatography. The purpose of this discussion was to make certain that the Organic Research Department personnel were well oriented with respect to the advantages and limitations of each technique. Because It was felt that the Information presented at the meeting would be of Interest to others, this report was drafted for dis tribution to staff members, group leaders and Individual chemists In the plasticizer, resin materials and functional fluids applica tive. and process groups. Summary Infrared absorption spectroscopy, gas-liquid partition chroma tography (OLPC) and nuclear magnetic resonance iNMR; are powerful tools for detecting. Identifying, characterizing, and measuring plasticizers. Each method has Its advantages and limitations. This report Is offered as a guide to those working the plasticizer field In selecting the best physical analytical method of analysis to fit a particular problem. Analytical Studies The plasticizers In common use at present can be divided Into the following classes: 1. Phthalate Esters 2. Isophthalate Esters 3. Terephthalate Esters 4. Phthalyl Olycolate Esters 6. Benzoate Esters 6. Maleate Esters 7e.. Fumarate E3ters Adipate Esters 9- Azelate Esters 10. Sebacate Esters 11. Citrate Esters 12. Acetyl Citrate Esters 13. Phosphate Esters 14. Sulfonamides 16. Chlorinated Hydrocarbons (Ar> 16. Hydrocarbons 17. Polyesters 18. Epoxy Compounds Table I lists the common members of each class showing those of Monsanto origin, those of competitive origin, and the number of competitive producers of each (data taken from the Encyclopedia of Modern Plastics, Vol. 39, No. i-a, 1962;. 0692920 HARTOLDMONOOQ5427 68. Type Phthlates TABUS I COMMON PLASTICIZERS Monsanto Origin Competitive Origin dl-methyl dl-ethyl dl-butyl dl-2-ethylhexyl (OOP) dl-lsooctyl (DIOP) dl-lsodecyl (DIDP) dl-phenyl dl-cyclohexyl dl-trldecyl butylbenzyl (S-160) butyloctyl (S-165) lsohexylbenzyl (S-260) lsooctylbenzyl (S-261) laodecylbenzyl (S-262) No. of Competitors 10 11 28 25 21 22 none 6 7 4 2 ... ... ... 0692921 f HARTOLDMONOOQ5428 69 Type Table I (cont'd) Monsanto Origin Competitive Origin (methylcarbltyl) benzyl{S-263) . dlcapryl No. or Competltora 6 butylcyclohexyl 2 butylphenoxyethy1 ethyleneglycol bls-(butyl) Iaophthalates none dl-2-ethylhexyl 1 Terephthalates none dl-2-ethylhexyl Phththaallyyl lycoiiates E-15lethyl-phthalyl ethyl glycoiate) B-l6(butylphthalyl butyl glycolate) M-17(methylphthalyl ethyl glycolate) Benzoates none dletnyleneglycol dlbenzoate 1 dlpropyleneglycol dlbenzoate 3 ethyleneglycol dlbenzoate 1 0692922 HARTOLDMONOOQ5429 70. Type Adipates Maleatea Fumaratea Azelates Sebacates j 1 i Table X (cont'd) Monsanto Origin Competlve Origin dl-2-ethylhexy) (DOA) dl-lsodecyl (DIDA) dl-n-octyl-n-decyl dl-lsobutyl dlbutyl dloctyl dlbutyl dloctyl none none dl-2-ethylhexyl dl-lsooctyl dl-lsobutyl dlbutyl dloctyl dl-laooctyl No. of Coapetltora 23 17 lit 8 5 1 3 3 10 7 2 14 14 0692923 HARTOLDMON0005430 71. Table I (cont'd) J&BS. Cltratea Monsanto Origin Competitive Origin No. of Competitora none triethyl 2 trl-n-butyl 3 Acetyl Cltratea none acetyl trlethyl 1 acetyl trl-n-butyl 1 Phosphates acetyl tri-(2-ethylhexyl) 1 triphenyl trleresyl S-l40(cresyldlphenyl) S-l41(2-ethylhexyldlphenyl) 6 12 7 S-144(lsoctyl diphenyl) Sulfonamides S-l-H(N-cyclohexyl-ptoluene sulfonamide) none S-3(N-ethyl-p-toluene sulfonamide) none S-8(N-ethyl-o,p-toluene sulfonamide) none S-9(o,p-toluene sulfonamide) none N-butyl benzenesulfonamide 1 069292** HARTOLDMON0005431 SMWHKSr r 72 Tabla I (coat'd) .Type Monsanto Origin Chlorinated Hydrocarbons Aroclor 1221 Aroclor 1232 Aroclor 1242 Aroclor 1248 Aroclor 1254 Aroclor 1260 Aroclor 1262 Aroclor 1268 Aroclor 4469 Aroclor 5442 Aroclor 5460 Hydrocarbons Aroclor 2565 HB-20 HB-40 dodecylbenzene Competitive Origin misc. No. of Competitors none none none none none none none none none none none none none 1 0692925 HARTOLDMON0005432 Type Table X (eont'd) Monsanto Origin ' Competitive Origin No. of Competitors Polyesters S-*K>5)l*3-butane<Jlol S-409)adlplc acid S-**ll)polyesters mlsc. Epoxy Compounds epoxldlzed soy bean oil epoxy tallates* epoxy tetrahydrophthalates 20 4 2 epoxy stearates mlsc 3 11 0692926 HARTOLDMON0005433 7<*. A. Tnf rired Absorption Spectroscopy In general. Infrared absorption spectra offer a rapid method of determining the class type of an unknown plasticizer (parent acid of the esters type, sulfonamides, phosphates, etc.). The absorption patterns are very consistent for a given class and serve as proof for Identification purposes. Identifica tion of a particular compound within a class requires a critical comparison of the Bpectrum of an unknown material with that of a reference spectrum. Often the positive Identification of a given compound In a class Is not possible because unless the IR spectrum Is obtained under special conditions, the small differ ences that etlst are not easily discernible. Por example, the spectra of dllsooctyl and dllsodecyl phthalste are practically Identical . Por mixtures, each case must be considered almost Indi vidually. However, the following general rules will apply: 0692927 HARTOLDMONOOQ5434 75. 1. For mixtures of two or more compounds In a class (l.e., butylbenzyl, dlbutyl and dlbenzyl phthalates) Infrared will not detect or , differentiate the components, only Indicate that a phthalate Is ' present. 2. In mixtures of plasticizers or different classes (l.e., fumarates and maleates) Infrared spectra will often, but not In every case, be useful In detecting the presence of each of the several classes of plasticizers present. 3. Hydrocarbons In the presence of the other classes of plasticizers cannot be detected directly. Quantitative analysis of such a mixture will give a differential measurement of the hydrocarbon. A qualitative Infrared absorption scheme Is attached which can be used to identify the majority of the plasticizer classes on the basis of the presence or absence of key absorption bands In the spectrum. A bar graph Is attached showing the characteristic absorption patterns for each plasticizer class. The wave length In microns for these characteristic bands are: Phthalates 5.8 S'J-8.9 9-3-9-4 9-6-9-7 13.4-13.5 14.2-14.3 2. Isophthalates 5.8-5.9 J.7-7-9 (usually a doublet) 8.6 8.8 9-1 9-3 ,, 13.7-13.8 14.1 3. Terephthalates 5.8-5.9 7.9-8.0 (doublet) 8.g-9.1 (doublet) 13'.7 0692928 l1 HARTOLDMONOOQ5435 76. 4. Phthalyl Olycolates 5. Benzoates 5.8-5-9 7.6 .8-7-9 9-0 9-3 9-7 14.0-14.1 6. Maleates 5-75-5.8p 6.1 Broad band with four distinct peaks at: V . U -U.X 8.25-8.55 8.6 -8.7 7. Fumarates 8.75-5.85 6.1 Broad band with two peaks at: 7.7-7 .8 7.9-6.0 8.2 Broad band with two peaks at: 8.5-8.6 8.7 10.2 8. Adipates 5.75-5.85 8.0 -8.1 8.| -8.6 (usually most Intense; 0692929 HARTOLDMONOOQ5436 9- Azelatcs |.75-5-85 8*5 -8.6 8.8 9-1 10. Sebacates 77- 11. Citrates 2.9 -5-0 5.75-5.85 8.4 -8.5 8.9 -9-0 12. Acetyl Citrates 2.9 -3.0 (weak) 5.75-5-85 lj. Phosphates No band In 5-7-6.0 region. 7.7- 7.8 10.3-10.6 14. Sulfonamides 3.0 No band In 5-7-6.0 region. 15. & 16. Hydrocarbons and Chlorinated Hydrocarbons No absorption In 5.7-6.0 region. No strong absorption In 7.5-9-0 region. 0692930 HARTOLDMONOOQ5437 17. Polyesters Spectra of polyesters resemble those of the parent acid derivatives (l.e., adipic acid polyesters spectra will resemble those of the simpler adipates). 18. Epoxy Compounds Insufficient data available to characterize this class. 0692931 HARTOLDMONOOQ5438 Present Band at $.77--55'J CARBOXYLIC ACID tarns Present Band *t 9,6-10.5 Absent Band at 716-7.6 r-eaent phosphates Absent nLio sulphokAmides Wvl.n*th In nloron. - band. r usually th atron*.t In th apcotrua. i-.4uwuuN)n'n>e 0692932 HARTOLDMON0005439 0692933 HARTOLDMON0005440 CHARACTERISTIC INFRARED ABSORTION 81. B. Nuclear Magnetic Resonance Spectroscopy Nuclear magnetic resonance Is a useful analytical tool for determining the structure of unknown compounds and for determining concentrations of components present in a mixture. A particular NMR measurement Is sensitive to only one nuclear species; at present, the Organic Research Department Is equipped to perform analysis of proton, phosphorus, and fluorine containing compounds. In struc tural determination, use Is made of the position (chemical shift; and the multiplicity of an absorption. The absorption position Is characteristic of the molecular environment of particular nuclei; the peak multiplicity gives Information concerning the presence or absence of nearby magnetic nuclei. NMR Is useful as a quantitative tool since the area observed for an NMR absorption Is directly proportional to the number of nuclei giving rise to the absorption. In using NMR for structure elucidation of plasticizers, a given plasticizer class cannot be assigned a characteristic absorp tion pattern which Is unique to that class, as Is possible with Infrared spectroscopy. Instead, the chemical shifts of the peaks observed In the NMR spectrum are used to Identify the structural groups present. Using this Information, together with peak multi plicity and relative peak areas, a definite structural assignment can often be made. . The approximate absorption intervals of the structural groups commonly found In commercial plasticizers are Indicated In t;-. attached figure. If more than one proton type Is present, the Indicated absorption Is due to the underlined proton type. The second attached figure Is a graphical representation of the NMR spectra of > series of phthalate esters showing their characteristic chemical shills and peak multiplicity; no attempt was made to repre sent relative p.ak ares . quantitatively In this figure. The com pounds listed ar , dimethyl phthalate; DIP, diethyl phthalate, DBP, dlbutyl phthalate; DIOP, dllsooctyl phthalate; r>?r, diphenyl phthalate; and S-160, butyl benzyl phthalate. There are several limitations to the Information obtainable by NMR. In mixtures, or very complex molecules, because of the finite line width of the observed NMR signals. It Is not always possible to distinguish between structural groups having similar chemical shifts. Also, in general. It Is not possible to dis tinguish between a pure compound and a mixture of compounds contalninlng the same structural groups. yr i "i | 4 i) fl 0692934 HARTOLDMONOOQ5441 PRCTOH MW SPECTRA OP SOKE FHTHALATK I4 Al I -< I g-UJj 3-160 A .1. A J, I. A *5'. i i w tt7 400 JOO 35 JOO Chemical Shift to (CH)4S1 at 60mc I *<* I DIOP **4-A +*> III DEP c<l Mr 47 <00 0692936 HARTOLDMONOOQ5443 8<t. C. Cas Chromatography The following table summarlres our present capability In gas chromatography with rerpect to plastl^lzers. This represents the scope of our present experience ar.d rot necessarily the limits of capability of ps chromatography, further experimenta tion and dvelopment of *hls technique may likely broalen its applications to plasMclzers. 3/63 B. Kat.lefskv, * Met rich, E. Eme'v, W. E. Koerner, h. E. teilt" 069293? HARTOLDMONOOQ5444 Plasticiser Systems Dialkyl Phthalates Dlalkyl Pumaratea and Maleates Alkyl Phosphates Diw?kyl Sueclu'tes, Adipates, Sebacstes, etc. Qrs Chromatoftra,hy Capability in Plasticisers Advantages 1. Resolution of mixed esters from the corresponding sym metrical esters (e.g., isooctylbenzyl and butyl-2ethylhexyl). 2. General resolution of a mixture of DOP, DIOP, DIDP, S-100, TCP, dibutyl-, dibensyl- and ditrldecylphthalates. Limitations Estimated Time Required per Sample Elapsed on " 1 "1 Instrument (min.) Man-mln. 30 30 1. Detn, of unreacted esters in bisfumarates. 2. Detn. of unreacted alcohol and other impurities. 3. Resolution of mixed esters from the corresponding symmetrical esters. Resolution of most mixed phosphates, such ac the methylphenyl and butyl-phenyl series. Do not resolve the higher maleates from the corresponding fumarates. life can resolve the dlbutyl and diethyl, but not the dl-2-ethylhexyl and dlisohexyl). Cannot as yet reoolve the 2-ethylhexyl-phenyl series (S-1^1y. JO :o }0 jo The same capability should apply in general as with the dlalkyl phthalates. 0692938 HARTOLDMONOOQ5445 SECTION K 66. Infrared Study of the Reaction Rate of Resorcinol In Sodium Sulfite Solution Introduction The literature reports that resorcinol In the presence of sodium bisulfite Is sulfonated In several Intermediate steps to yield sodium l,3-dlhydroxy-l,3,5-cyclohexanetrlsulfonate and m-phenol sodium sulfonate as stable end products. Confirmation of this side reaction and reaction rate data were needed to optimize process conditions and minimize resorcinol losses In the acidification and quenched liquor storage steps of the resorcinol process. Summary An extraction procedure and Infrared absorption method were developed for the measurement of resorcinol In saturated sodium sulfite solutions which provided analytical data to calculate rate constants. In the temperature range of 50 to 103*C. and the pH range of 6 to 12 the data Indicated that at constant pH the reac tion rate Increased with temperature, and at constant temperature, the rate Increased with acidity. Under the tentative operating conditions of pH f.0 and 50*C., resorcinol Is sulfonated In a saturated sodium sulfite solution at a rate of l.o per hour. NKK analysis of samples of resorcinol In a saturated sodium sulfite solution refluxed at pH 7-0 showed that sodium l,3-dlhydroxy-l,3,5cyoI'>hexanetrlsulfonate and sodium nr-phenol sulfonate were the sulfona;*on products formed. Analytical Studies A. Infrared Spectroscopy Infrared absorption methods were used to study the disappearance of resorcinol In aqueous solutions saturated with sodium sulfite. This study was requested to provide reaction rate data as a function of pH and temperature to assist In optimizing process conditions and to minimize losses of resorcinol In the acidification and quenched liquor storage stages of the resorcinol process. The literature N. Uflmtseb, Zhur. Prlktad. Khlm., 20, 1199 (19i*7J7 reports that resorcinol,in the presence of sodium"Elsulflte, is sulfonated. The sulfonatlon proceeds through several steps and 1.5- dlhydroxy-5-(1,3-cyclohexadlene sodium sulfonate) and 2,5-dlhydroxy- 2.5- (cyclohexene dlsodlum sulfonate; have been Identified as Inter mediate products. The stable end products of the reaction are 1,3- dihydroxy-l,3,5-cyclohexanetrlsulfonate and m-phenol sodium sulfonate. 0692939 HARTOL DM ON0005446 87 Infrared methods were Investigated to measure the resorcinol directly In aqueous solution. Attenuated Total Reflectance techniques proved unsatisfactory due to the poor long term stability of the optical bench settings which resulted In poor reproducibility. Direct transmission measurements of the solutions using thin cells with Irtran-2 windows were ruled out because measurements of either the 6.7 micron C-C ring stretching band or the 7.8 micron 0-11 deformation band showed Interference from intermediate products. An extraction method using isopropyl ether was used to isolate the resorcinol from the reaction mixture. The resulting ether solution was analyzed directly using conventional infrared transmission techniques to determine the resorcinol concentration. In the procedure used to Isolate the resorcinol from the reaction mixture, a 20.0 0.02 gram aliquot of the reaction mixture was cooled to room temperature in a water bath and the pH adjusted to 6.0 by bubbling In sulfur dioxide (S0) and using a pH meter. (In this study the resctlon mixtures contained approximately 4# resorcinol and were saturated with sodium sulfite.) The solution was transferred to a 76 ml. separatory funnel, the beaker was rinsed with 10 ml. of Isopropyl ether snd the ether was transferred to the funnel. The system was shaken for 50 seconds, the phases permitted to separate and then shaken for an additional 50 second period. The ether layer was filtered through a cotton plug containing a layer of anhydrous sodium sulfate Into a 2`> ml. volumetric flask. The water layer was extracted In turn with a second 10 ml. portion of Isopropyl ether and finally with, a 6 ml. portion of solvent. The funnel was rinsed with a small portion ol lsopropvl ether ar.d the rinsings were added to the filter. The Isopropyl ether volume in the flask was aoj 'sted by adding additional ether dropwise to the filter. Anhydrous sodium sulfate w.o added to the ftasa end the mixture Bhaken. A portion ol th0 Isopropyl ether solution was transferred toa 2 dram vial containing a fresh portion of anhydrous sodium sulfate and shaken. This solution was decanted and equilibrated with finely ground sodium chloride In another 2 dram vial. Resorcinol recoveries of 9?-955( were obtained by this extraction and drying procedure. The dried Isopropyl ether solution was transferred to a 0.082 mm. rock salt Infrared cell and the Infra; ud spectrum was obtained uslrg a 0.C76 mm. solvent compensating cell through the region of 1C. 1-10.7 microns. Measurements were made at 10.58 microns and a background correction measured at 10.20 microns to determine- the resorcinol concentration. The instrument ccr.lltions used were: 0692940 ! HARTOLDMONOOQ5447 68. Instruments Optics: Silt Program: Attenuator Speed! Scanning Speed: Osin: Suppression: Chart Scale: Sample Cell: Reference Cell: Solvent: Perkln-Elmer Model 221 Double Beam Infrared Spectrophotometer NaCl prism 990 A seconds full scale deflection A mlnutes/mlcron 4 0 10 cm./micron 0.082 mm. 0.076 mm. Isopropyl Ether The resorclnol-sodlum sulfite reaction was run under the following conditions: H Temperature 4 6 7 8 9 10 12.5 70* 50 50, 70 and 90 5, 70 % 90, 105.5 (Reflux) 105.5 105-5 The large excess of sodium sulfite present In the reaction mixture does not buffer the solution. As the reaction of resorcinol proceeds, the pH rises approximately 0.5 to 0.5 pH units when about 25$ of the resorcinol has reacted. A plot ',f the analytical data. In the form of the logarithm of the rt'or^inol concentration as a function of time for pH ranges of 6-12.3 and temperatures of 50-105.5*0., shows that the reaction Is first order with respect to resorcinol. These plots are shown In Figures 1-4. Plots of the resorcinol loss as a function of time under the conditions Indicated above are shown In Figures 3-8. The rate constants calculated for each pH and temperature condition studied by: pH Temperature Rate ConstantA Rate PercentB 6 7 8 7 8 7 9 9 10 12.5 50*C. 0.0180/hour 50*C. 0.0101 ft 50*C. 0.0044 ft 70*C. 0.0504 " 70*C. 0.0114 ft 90*C. 0.118 tl 90*0. 0.0445 105.5 C.(Reflux)0.141 It 105-5 C. ft 0.0900 ft 105.5*C. ft 0.0017 ft 1.80/hour 1.01 ft 0.44 tl 5.04 tt 1.14 " 11.8 tt 4.45 II 14.1 M 9.00 tt 0.17 M 0692941 HARTOLDMONOOQ5448 89. (A) Rate Constant k 2.303 . t-t,` 105 where Ci - concentration at time ti and Ci concentration at time t* (B) Rate Percent - 100 k - percent resorcinol reacting per hour. A graphical plot of the rate constant vs. pH le shown In Figure 9- The data obtained at 103.5*C. are curvilinear and suggest the data for the lower temperatures are also actually curvilinear with an extrapolation to a common zero point. The data obtained under acid conditions of pH Aat 70*C. are somewhat confusing. After one hour under these conditions, the resorcinol concentration Is reduced about 1 It. During the next three hours the resorcinol Is regenerated so that at the end of four hours the resorcinol concentration Is the same as the zero time sample. If bisulfite is the active reactant, species In this reaction, then at pH A,where the equilibrium should be shifted entirely to the sulfite Ion form, some other complexlng mechanism must be operating. Since resorcinol Is regenerated under these conditions, the resorcinol complex must be loosely coupled. B. NMR Spectroscopy Analysis of NMR spectra obtained from samples of resorcinol In saturated aqueous NaaSO* solution Indicate the following trans- formation- Resorcinol (A) (B) (C) 71.J rates of these reactions Increase with Increased temperature. Three samples taken from a solution under reflux conditions at pH 7 Indicate the following approximate compositions: 3 hr. reflux ?A hr. reflux 72 hf. reflux A, tOS* P ^90* P. 10# *> ^70# P, 30* C - Analysis of"the NHR spectra end related reference materials Indicate that (B) lasodium j , 3-d 1 hydroxy 1 , 1. cj-c vcl ohexanet rl sul - fonate or some closely related material, and (?) Is sodium m-phenol sulfonate. Intermediate compounds between resorcinol and (B) and between (B) and (C) may exist. > 12/62 - P. Katlafsky, M. Dietrich, R. E. Keller ; ' E |> ( ! 1 > ! > | : ; \ ; , , ' t ; j j | 06929A2 f HARTOLDMONOOQ5449 HARTOLDMON0005450 HARTOLDMONOOQ5452 j HARTO L D MO N0005453 069294 7 HARTOLDMONOOQ5454 Fitufi C HARTOLDMONOOQ5455 Pi6uRe ~r HARTOLDMONOOQ5456 0692950 HARTOLDMONOOQ5457 0692951 t HARTOLDMONOOQ5458 99. SECTION L Attenuated Total Reflectance Infrared Study of Sodium Phenate and Sodium Benzenesulfonate Introduction Conventional Infrared analytical methods for laboratory and plant proceas aqueous streams have been limited by the lntenae eater absorptlon which nakes It necessary to use extremely thin cells. Problems In plugging or deposits of foreign material In cells of this type In process streaa applications result In serious analytical errors. A new Infrared technique, known as Attenuated Total Reflectance (ATR), Is based on Infrared spectra obtained by surface absorption and Is Independent of sample thickness. The latter feature permits design of sampling cells of sufficient thickness to overcome the sampling problems associated with the normal transmittance techniques. Sodlua benzenesulfonate (NaBS) and sodlua phenate solutions were Investi gated to demonstrate the analytical feasibility of this technique, since better control methods for these siaterlals In phenol process streams could result In Improved phenol yields. Summary ATR infrared studies of aqueous solutions of sodium phenate and sodium benzenesulfonate (NaBS) have demonstrated this tschnlque would bs a specific and practical method for monitoring certain phenol process streams. Sodium phenate and NaBS In the concentration range of 10-250 can be measured to within +0.50 absolute. In NaBS streams, the components NaS0, Na*S0s, Ns5h, Ns* CO* and sodium phenate would Introduce only negligible errors In measuring NaBS. In s'.dlum phenate streams, small corrections could be made for the ttrlbutlon of NaOH and NaaSOs to the sodium phenate measurement. The Connecticut Instrument Corporation ATR equipment used In this work should ba replaced by a structurally more stable unit for use as a plant streaa analyser. Analytical Studies Attenuated Total Reflectance (ATR) Infrared methods were Investi gated to determine the feasibility of this technique for monitoring aqueous NaBS (sodium benzenesulfonate) and sodium phenate process streams in the phsnol process. Excellent results obtained with laboratory scale equipment warrants a recommendation to Dr. Fowler's Instrumentation Oroup to consider construction of an Instrument using the ATR principle as a process stresm analyzer. 0692952 HARTOLDMONOOQ5459 100. ATR spectra were obtained using a Connecticut Instrument Corpora tion Model ATR-1 attachment on a Perkln-Klmex Model 221 Infrared Spectrophotometer. Silver chloride, silicon and Trtran-2 (Eastman Kodak) prisms were Investigated. A silver chloride prism was used first., since this was available as a component of the basic ATR-1 kit. Na3S gave no difficulty, but sodium phenate reacted with the silver chloride, leaving a deposit on the back face of the prism. A heml-cyllnder made from zone refined silicon In which the diametrical face acts as the reflecting surface was obtained by Dr. Fowler from the Inorganic Division. Surface oxidation on the exposed faces of the silicon resulted In intense absorption bands In the regions of Interest for NaBS and sodium phenate making this optical material unusable for this application. Eastman Kodak's Jrtrsn-? optical material is Infrared transparent In the regions of Interest and Is relatively unaffected by caustic materials. Two lrtran-2 prisma were obtained from the Connecticut Instrument Corporation and have been used with out any adverse effects In this study. The angle of Incidence relative to the reflecting face of the prism was found to have a large effect on the Intensity of the absorption bands observed. The bands cecame more Intense as the angle of Incidence became nn',,1er and an angle of HO*, thi minimum angle that can be set up with the ATR-1 unit, was used. The phenol process streams of Interes* have the following typical compositions: Process Stream W~Nafig----- 6"f~Nag5 Pherate Slurry Temperature 95l 100* 120* Component Water NaBS Sodium phenate Sodium sulfite Sodium sulfate Sodium chloride Sodium hydroxide Sodium carbonate Diphenyl sulfone Na dlbenzenesulfonate Na dlhydroxybenzene Na o phenylpher.ol Tars l 49 76 44.68 0.06 0. 41 "51 0.06 C. 16 0.06 0.25 0 06 2 44 6? 61.60 -------- C.56 P.hR 0.07 0.23 o.o8 0. >4 0.06 ........... -- ' 26.30 0.23 20.90 34.82 2.76 1.54 3.60 0.41 0.06 0.03 0.16 0692953 HARTOLDMON0005460 101. Since no provision la made for thermoatatlng the ATR-1 unit or the liquid ATR cells, the laboratory studies Mere made at room tem perature. Solubility limitations at room temperature loitered the concentration ranges In which NaBS and sodium sulfite could be studied. Water solutions In the following concentration ranges were prepared for the principle components In these streams: NaBS Sodium phenate Sodium sulfite Sodium sulfate Sodium hydroxide Sodium carbonate 1005* 10-35* 10-20* 1-5* 1-8* 0.2-1* - Two ATR units were set up to permit running the spectra of the solutions against water In the reference unit to compensate the water absorption. The ATR spectrum of water Is shown In Figure 1 of Progress Report 2-02-760.01-4285, No. 16. The following ATR and Infrared spectrophotometer parameters were used In this study: ATR prism Angle of Incidence Solvent Compensator Slit program Scanning speed Chart scale IR Instrument Irtran-2 40* Water Water 980 1 mlnute/mlcron 10 cm./micron Pcrkln-Elmer Model 221 Spectral Results . Sodium Phenate ATR absorption bands are found at 6.77 microns for the phenyl ring and at 7.90 microns for the Ar-0 stretching absorption. Either band Is suitable for analytical purposes. However, serious errors due to the presence of sodium sulfite would be Introduced to measure ments made at 6.77 microns while at 7-90 these effects would be small. Figure 1 shows the superimposed spectra of a series of aqueous solu tions of sodium phenate. Figure 2 Is a plot of the optical density at 7.90 microns for these solutions versus the sodium phenate con centration (all measurements are made relative to the water blank at this wave). A linear function Is obtained showing that Beer's Law Is obeyed at this wavelength. 0692954 HARTOLDMONOOQ5461 102. 2. NaBS ATR absorption bands are found at 8.5? microns for the sul fonate group and at 8.90 microns for what In probably due to a vibration of the aromatic ring. Either band could be used to set up an analytical method, but m this system the presence of sodium sulfate would produce serious errors In measurements made at 8.9 microns. The Interference due to the sulfate would be small for the 8.52 micron band. Figure 3 shows the superimposed spectra of a series of aqueous solutions of NaBS. Figure 4 is a plot of the optical density of the 8.52 micron band versus the concentration of NaBS (all measurements are relative to the water blank at this wavelength). A linear function Is obtained showing that Beer's Law Is obeyed at this wavelength. 3. Sodium Sulfate and Sodium Sulfite Figures 5 and 6 sho.* the superimposed spectra of a series of aqueous solutions containing l-5m sodium sulfate and 10-20# sodium sulfite. At the analytical wavelengths (7.90 and 8.52 microns). It can be seen that interference errors would be small. 4. Sodium Hydroxide and Sodium Carbonate Figure 7 shows the spectra of 7.8* sodium hydroxide and i# sodium carbonate in the spectral region of interest. Sodium hydroxide produces no finite absorption band but merely depresses the background. Its presence would have about equal effect at 7.90 or 8.52 microns. Sodium carbonate produces an absorption band peaking at 7-?5 microns which tapers off to a point where its presence would Introduce only a slight error In measurements at the two analytical wavelengths. From the ATR spectra of sodium phenate and NaBS obtained under laboratory conditions. It can be seen that both materials could be easily measured to + 0.5^ absolute. Calculations were made'to deter mine the approximate error that would be Introduced by the presence of BOdlum sulfate, sodium sulfite, sodium hydroxide and sodium carbonate at the concentration levels present In a .typical stream in the measurement, of sodium phenate and NaBS. The results are: NaBS Measured at 8.5? /> 44# NaBS Stream 0.4# sodium sulfite 0.?# sodium hydroxide 0.06# sodium carbonate 0.06# noMum phenate Is equivalent to 0.006# Is equivalent to 0.06# is equivalent to 0.02# Is equivalent to 0.01# NaBS NaBS NsB3 NaBS HARTOLDMONOOCI5462 540 NaBS Strati 0.60 sodium sulfite Is equivalent to 0.010 NaBS 0.20 sodium hydroxide Is equivalent to 0.06% NaBS 0.09% sodium carbonate Is equivalent to 0.030 NaBS 2.7% sodium sulfate - negligible contribution Sodium Phenate Measured at 7.90 Microns Phenste Slurry 34.80 sodium sulfite Is equivalent to 0.50 phenate 3.80 sodium hydroxide Is equivalent to 1.00 phenate 0.40 sodium carbonate Is equivalent to 0.10 phenate 2.8% sodium sulfate negligible contribution 0.2% NaBS - negligible contribution These calculations, even If the approximations were In error by a factor of 2, show that a NaBS stream analysis by an ATR method would not be affected by the presence of the other compo nents In the system. Corrections would be necessary In a phenate slurry stream to correct for errors Introduced by the presence of sodium hydroxide and sodium sulfite. The results obtained in this Investigation Indicate that ATR infrared methods are feasible for the analysis of NaBS and sodium phenate streams In the phenol process. Presently available commercial equipment, however, would not be practical for long term use unless frequent calibration checks were made. The optical bench and the optical mounts of the ATR-1 unit from Connecticut Instrument Corporation are not sturdy enough for long term stability. The ''.eslgn and construction of a more rugged ATR unit should be Inves tigated for application as a plant stream analyzer. 8/62 - B. Katlafsky, R. E. Keller 0692956 s: f i HARTOLDMONOOQ5463 HARTOLDMONOOQ5464 fc- 0692958 . { HARTOLDMONOOQ5465 WAVELENGTH (MICRONS) 0692959 HARTOLDMONOOQ5466 i V 0692960 v t %' % l I HARTO L D M O N0005467 loti. 11.1.1111III I i 1 I I I 1 I I I I 1 I I I I I I WAVELENGTH (MICRONS) 0692961 HARTOLDMON00054fi8 109 4) \ WAVELENGTH (MIRONS) 0692962 HARTOLDMONOOQ5469 110 WAVELENGTH (MICRONS) ------------- -rr''> " w. ' 0692963 HARTOLDMON0005470 I SECTION M Thin Layer Chromatography - Evaluation 111. Introduction and Summary Thin layer chromatography (TLC) Involving the use of an adsor bent spread In a thin layer on a glass plate has been given a preliminary evaluation. Results obtained Mlth silicic acid as the adsorbent show that mixtures of resorcinol, pyrocatechol, phenol, hydroqulnone and phlorogluclnol can be separated Into components and Identified within *5 minutes using 105f n-butanol In benzene as the eluant. Seml-quantltatlve estimations can be made by visual Inspection. TTie prime advantages over paper chromatography appear to be speed and Increased resolution. Water con*ent of the adsor bent Is a significant variable that must be controlled. As with paper chromatography, accurate quantitative measurements may prove to be difficult. Analytical Studies The thin films on glass plates used for TLC may be prepared with a number of supports including silica gel 0, aluminum oxide, Kleselguhr 0 and cellulose powder. The films for this work were prepared with the "Desaga" apparatus designed by Stahl. The samples are applied to the film with a Hamilton mlcrollter syringe with sample volume ranging from 1 to SO mlcrollters. The chromatographic separation Is dependent upon the amount of moisture In the film and upon the mobile solvent used. In the work with resorcinol, dry plates produced poor separation and tailing. The freshly prepared wet plates were put in a desiccator containing a dish of water to control the moisture. This technique will keep the film sufficiently wet for a few days. As In paper chromatography, the selection of a suitable solvent Is based on knowledge of solu bilities, partition coefficients and finally by trial and error. In the case of resorcinol, n-butanol in benzene (vol./vol.)was us-d for our paper chromatography work. The amount of n-butanol was varied and It. was found that \0% n-butanol In benzene gave the best separations using the silica gel TLC scheme. The separated components were detected by color development with a coupling agent (^-nitrobenzene diazonlum fluoroborate) followed by alcoholic potassium hydroxide. The TLC chromatograms are difficult, to keep and are somewhat fragile. A solution to this problem Is the transferral of the chromatogram to paper with a Xerox 91<* copier. Figure I shows a Xerox copy of a chromatogram of Individual components and an equal mixture of the resorcinol and related components. 0692964 I 112. The TLC procedure requires about 45 minutes total elapsed time per chromatogram as compared with 3-4 hours per paper chromatogram. The sensitivity appears greater than that obtainable using paper. Preliminary work on quantitative measurements shows that the problems common to paper chromatograms also exist with TLC schemes. Removal of the spot can be accomplished easily with a razor blade. Elution of the component can be made with a suitable solvent for subsequent ultraviolet analysis or similar measurement. However, the microgram quantities normally separated make It necessary to use very sensitive methods for measurement. Work at higher concentration levels showed that bands Instead of spots could be separated with the purpose of producing larger concentrations of separated compo nents. Area measurements for quantitative determinations apparently can be carried out with about the same accuracy obtainable with paper chromatography. No specific area measurements were carried out as a part of this study. Work Is underway on determining the feasibility of Infrared attenu ated total reflectance as a means of Identifying separated components directly on the silica gel. l?/62 - 0. Hicks, R. E. Keller 0692965 t; i ! HARTO L D M O N0005472 113. Figure 1 TLC Chromatogram for Rrrorclnol and Related Components (copied with Zerox 914 copier) SOLVENT - 10 Butyl Alcohol in Benzene CARRIER - Silica Qel G Preparation - 1 part silica el/2 parts water, film prepared ana held 1 hour at room temperature conditions. DEVELOPMENT Til - 30 Minutes SPOT DEVELOPMENT - p-Nltrobenzene diazonium fluoroborate followed by alcoholic KOH. jr A\ u PHENOL I F/Rccmwi. I , j RSOC.INOt- | UiDRcWiMUZ PHLOkOG-LUClNOL f ccmpcrznts 1. Resorcinol, 500V in 10 microliters acetone. 2. Phloroglucinol, 500V in 10 microliters acetone. s. Pyrocatechol, " " " " " Hydroqulnone, " "" " " 5. Phenol, " "" " " .4. Equal mixture of 1, 2, 3, 4 and 5, " total 0692966 * HARTOLDMONOOQ5473 SECTION N Evaluation of Trlcosane Oxidation Study by Infrared Analysis Introduction Basic structure data were needed to assist In the selection of an oxidation Inhibitor for Trlcosane. Arrangements were made to examine a Co alkane that had been analyzed In the laboratory using the Dornte apparatus by a differential Infrared absorption procedure. Summary Infrared data were collected from five Dornte oxidation runs. The samples were analyzed as capillary films in 0.2 mm. cells using the starting material for compensation. Samples were taken periodically during the laboratory run until the Infrared absorp tion went off-scale. Optical densities were plotted graphically for five wavelength positions representative of the functional groups formed during the oxidation. The family of curves show that the various functions form at somewhat Independent rates. Saturated ketones Increase at & fairly steady rate. Alcohols rise rapidly Initially. Type III olefins rapidly reach a maximum and eventually attenuate to a small fraction of the peak value. Analytical Studies Ketone, alcohol and the olefin, RRCCH, form Immediately and become the dominant structures early In the oxidation. Near the peak of olefin production a wide variety of oxidized products appear and esters become significant. There Is no evidence for conjugated ketones or aldehydes. Infrared does not provide good evidence for peroxides. None are evident. Eventually a wide variety of carbonyls appear. Unsaturation occurs almost entirely In the Type III form, plus a minor amount of trans olefin. One gas sample from the laboratory oxidation run was analyzed. The carbon monoxide content of the gas collected from the apparatus was six times as great as the carbon dioxide content. Methane, ethylene and propylene each amounted to about one tenth the carbon dioxide content. An unknown alcohol of more than three carbons appeared In trace quantity. 0692967 HARTOLDMONOOQ5474 U5, Punctlon -0-H -C-0 -C-0 -C-0 -C-H Table of Wavelengths Studies Wavelength Structure 2.95 ml crone Alcohol hydroxyl 5.82 Saturate ketone and ester carbonyl 8.62 Ester C-0 9-50 Alcohol and ether C-0 11.50 Type III olefin 12/62 - 0. Klnast, R. E. Keller 0692968 HARTOLDMONOOQ5475 SECTION 0 116. Vanadium Oxide Components In Catalyst Mixtures Introduction Vanadium can form In addition to Va0*, V0 and V>0 other lesser known oxides with empirical formulas that are intermediate between Va0a and Va0* and Va0 and VaOs. X-ray diffraction data indicate the possibility of six oxides between Va0 and Va0 and one between Va04 and Va0*. Since these intermediate oxide compounds could be present in conventional vanadium oxide catalyst mixtures, methods for detecting and identifying these species would be helpful. In addition, characterization and correlation of composition with activity would be useful in the design of better catalysts. An Infrared study was initiated to elucldAte the chemical bondltv, in a series of intermediate oxides. Summary Infrared solid state spectra were obtained for Va0a, V*0 and Va0 dispersed in KBr in the 2-35 micron region. Subsequent studies show that radical spectral differences exist for prepared intermolecular complexes including VaOs-s VaOa.ss, V*0S., V0*., Va0a.a and Va04.se* This work corroborated X-ray diffraction data and indicates new chemical compounds are produced in the systems VaOa-Va04 and VaO-VaOs after heating. Analytical Studies Vanadium is known to form,in addition to the common oxides Va0a, Va0 and VaOs, a series of intermediate or suboxldes on heating the systems: n Va0a + m Va0 or n VaC4 + m Va0$ where n and m are lntergers. Since these suboxldes have catalytical application potentials, an Infrared study was initiated to detect and characterizes these oxides. The oxide systems studied were prepared by Dr. R. H. Munch and Included: V,,09 Va0.s (V0 + Va0* - V0a> Va0*.,, (2Va0 VaOs - Ve0XSJ Va0., (7V0 + 3Va0s - Vao04S) va0.a AVaO va0s - V10Oa, Va04 i (9V*0 + Va0* VaoOi i va0 0692969 HARTOLDMONOOQ5476 117. VaO,. iV,0, + 9V,0 - V,o0,,) V0,. (v,0, + 4V,0 - VloO.) V0,.T (3V,Oj + 7V0 - V,o0,7) VaO,.*r (V,0, + 2V,0 VOn) VaO,., (2V,0, 3V,0 - Vlo01B) V*0,.. V,0, + V,04 - V07) v0,. (3V0 + 2V0 - V10Oit) V,0,.,, (2V,0, + V,0 - V*0,o) VaO,., (7vOs + ?V,0 - VjoOjj) VaO,., (4V,0, + v0 - Vlo0lB) VaO,.! (9VaO, + V,0 - VtoO,j) V,0, The Infrared spectra were obtained In the 2-35 micron region for the oxides dispersed In KBr (potassium bromide) and pressed Into thin discs. The most significant data were obtained In the rock salt region (2-15 microns). The Infrared absorption data observed In this region for the vanadium oxide systems are repre sented in bar graph form in Figures I and II. A. VaOt-VaOs System (Figure I) Vanadium pentoxlde (V,0,) has a very Intense, clearly defined absorption spectrum with absorption bands at 9-80 and 12.30 microns. The tetroxlde (v,0) shows a very weak, broad band at about 9.6 microns and a slightly stronger band at about 15 microns. The Infrared spectra of the oxides obtained having the empirical formu lae V0.s, V,0.,, V*0., and V,04.i Indicate that these materials are mixtures of V0 and VgOs. The spectrum obtained for the material having the empirical formula V,0.,, shows radical changes compared to those of V,0 and V,09. Two weak bands of approximately the same Intensity appear at b.65 and 8.90 microns and a stronger band at 11.24 witii a shoulder at 11.70 microns are observed. No absorption Is found at 9-80 or 12.30 microns. Indicating that no V,0s Is present. This spectral evidence supports X-ray diffraction data that the vanadium oxide having the empirical formula V,0.s, Is a new chemical entity. B. VgOs-VgO* System (Figure IIJ Vanadium trloxlde (V,0,j Is practl ally opaque throughout the spectral region studied and shows only an extremely weak absorp tion band at about 1C microns. The tetroxlde (V*0),as noted before, has a weak band at about 15 microns and a very weak, broad band peaking at about 9-6 microns. The molecular complexes corresponding to the empirical formula0 of V,0,.,, V,0,.7, Va0,.*7> V0,.s, V,0,.a and Vg0,.i produce Infrared spectra that are devoid of character. Indicating that these are merely mixtures of the V,0, and V,0 enti tles. The species VaO,.*, VaO,.* and V,0S. have a medium intensity 0692970 HARTOLDMON0005477 110. band at 10.05 microns and VaOs.s has a strong band at 14.0 microns. Absorption at these wavelengths Is absent In both VtOs and V*0 Indicating that these four entitles are new chemical compounds. This data supports the conclusion drawn from X-ray defraction data for these materials. The species corresponding to the empirical formula Vc0s.ss has Infrared absorption bands at both 10.05 and 14.0 microns. Although the spectral data can be interpreted either as indicating a new chemical entity or to Indicate a mixture of the VaOs.s and Va0s. species. X-ray diffraction data clearly indicate that VaOs.ss is a new chemical compound. The appearance of an absorption band in the 10 micron region In the Va0s.a, VaOs.s, V80*.4 and VaOa.ss species is Interpreted as Indicating that these suboxides have some V0 bonding. This conclusion Is based on data compiled In the reference text: 0. Herzberg, "Spectra of Diatomic Molecules", 2nd edition, pp. 501-581 (Van Nostrand, New York, 1950) In this compilation, spectral data observed for the electronic states of diatomic molecules are used to calculate the vibrational constants for the system. These constants can be used In turn to calculate the vibrational frequency (infrared absorption) of the system. This frequency Is conventionally reported In terms of reciprocal centimeters, cm."1 (cm."1 - 104 x the reciprocal of the wavelength In microns). The vibrational frequency calculated for the M0 entity of vanadium and Its neighbors In the first row of transition elements In the periodic table are: M (-0) Frequency In cm."1 Se 96" T1 100" V 1003 Cr 086 Mn 631 Fe 870 Co No data N1 No data observed Infrared data for Va03 and the 's-ss, V0S. , V0 s. and VaOs-s are: Oxide Preqjcr.cv In cm."1 V0, V*0 . V0,.3, VaOs. Va03. s VaOs.s 1021 1123, 1156 995 995 995 995 0692971 HARTOLDMONOOQ5478 119 The close agreement between the observed Inrrared absorption frequency and the calculated frequency for the V-0 vibration is Interpreted to indicate that the suboxides V*0.ss, V0*.*, V0S., VjO*.# and VaOs. contain some V-0 bonds. This infrared study has confirmed X-ray diffraction data that a minimum of five suboxides can be obtained by heating stoichio metric quantities of V*Os and V0 and one between V0* and v*0. The observed Infrared data have been interpreted to Indicate the presence of V-0 bonding in five of these suboxides. Further work would be required with known oxide systems to fully characterize the oxygen bridge bond In the vanadium oxides. 12/62 - B. Katlafsky, R. E. Keller 0692972 HARTOLDMONOOQ5479 W O T -tilG WAVELENGTH IN MICRONS 0692973 HARTOLDMONOnn.Man 121 WAVELENGTH IN MICRONS 0692974 HARTOLDMONOOQ5481 K-H