Document EddjQxx5LLrnX1LbRRwGnOO7b

I ETHYL cr.::r-'n,vr!PN I RESEARCH LASOSATUaiEa I OSTH O.- T. M t SH t n A N I LOT FOR r`U3UCATIOfl CALIFORNIA RESEARCH CORPORATION RICHMOND, CALIFORNIA LEAD-IN-AIR ANALYSIS FILS 121.102 COMPARISON ,OF TML AND : S _ ________________ JANUARY 27. 1959 Introduction There Is considerable interest in studying the concentrations of 'alkyl lead compounds in gasoline vapors under exposure conditions typical of service. Current Interest in TML* which is more /clatile than TEL, stems from its superior antiknock performance in srtain applications. Since TEL has been shown to be nontoxic under the conditions of exposure associated v;lth handling gasoline, a direct comparison of the two alkyl lead compounds under a variety of conditions is pertinent in assessing any increased health hazard. Analytical Method The analyses were performed by the method in general use for the field analysis of lead in air.* In principle the analysis consists of drawing an air sample which contains lead compounds through an Iodine solution, thus converting the lead alkyls quantitatively to lead iodide. The lead iodide Is then cornplexed and extracted into chloroform with dithizone reagent. The resulting colored solution 13 compared with standards in a color comparator and the lead concentration of the air sample is read directly. It was found necessary to heat the iodine solutions after collecting the air samples to assure complete conversion of the trapped alkyl lead compounds to lead iodide. It was also necessary to filter the solutions to remove solid colored material which presumably resulted from reaction of the iodine with hydrocarbon components of the sample. These solids were soluble in chloroform and interfered with the color comparisons. With these modifications, highly reproducible results were obtained under a variety of conditions. Webelieve this method is suitable for making comparisons between TEL and TML under the conditions of a field survey. Discussion The molar ratio of TML to TEL above gasoline under equilibrium conditions, calculated from vapor pressure data, is 87.*1 at 7QF (see Table X)* On tl>e other hand, whenever *L. J. Snyder, W. R. Barnes, and J. V. Tokos, Anal. Chem. 2 0 , 772 (1948). -- gasoline is soil d u : poured from one container to another, d certain part of' the whole gasoline, Including the lead compound, evaporates. This situation is entirely different iron', an equilibrium evaporation. To the extent that whole gaso',.ine evaporates, tvc lead alkyls of different vapor pressure will be vapor:zed in equal amounts. .An effect intermediate between these two situations is to be expected ..henever a draft of ai: removes gasoline vapors at such a rate that equilibrium evaporation cannot be established. In such a process the rapid lost of gasoline by evaporation causes entrai:.ment of heavier gasoline components, including tne lead alkyls. The higher boiling components show up in the vapor in treater concentration t.uan would be predicted from equilibrium sons d e r a t i o n s . We have termed this phenomenon the ''spewing effect. ' Table I The Effect of Temperature ____ ot. Vapor Pressure Temperature p 50 6o 70 OJ QO 100 Pressure (mm K) TML TEL -15 0.11 C .16 0.23 52 0.45 45 G .67 58 1.0 41e Ra r.io TML/TEL 118 1G0 86 74 64 5b ExperTme n tal In experiments designed to check the equilibrium calculations, solutions of lead alkyls in hydrocarbons were shaken in a separatory funnel with air until equilibrium was established. These vapors 'were then displaced by gasoline and passed through a scrubber containing the iodine solution which 'was then analyzed for lead. The results are shown In Table II. K 0002995 Table II Lead-in-Air Measurements Eau...i..l. .i..b...r..i...u..t...n.....S- It u ,d- i! es..--I : Isooc tane, Calc* Found Toluene, 0 ale ,)und Gasoline . laic found Estimated Experimen cal Error Lead Cone entration (Microgra ms/Cu Ft) -------T--M---L--------- | TEL 4200 5000 1 48 !j .50 5500 4400 ` 58 ; 50 Mole Ratio TML/TEL 87 100 87 88 5500 5500 1! ! t-200 i ! 40 ;* 1_0- 0** i -20 i. 87 51 Gasolines contained j?.lo grams lead per gallon. Equilibrium not attained. In solvents such as isooctane and toluene, the lead conoca i itions predicted for equilibrium conditions were closely =.ppreached for both TML and TEL. On tne other hand, the amount of *1Jj xn gasoline vapor is much larger than calculated. VJe attribute this to the fact that equilibration of the gasoline samples in the test apparatus was incomplete because of the high vapor pressure of the gasoline. This large, positive deviation is associated with the 'spewing effect." The effect is not noted for TML, probably because of the relatively small effect it would have on the total compared to the size of the experimental error associated with the measurements. From these studies with gasoline vapors, it is concluded that the extreme ratio of 87 moles of TML to 1 mole of TEL will not be achieved under ordinary circumstances, but ratios of less than 50:1 will occur in situations which approach the equilibrium case. Measurements of the lead in gasoline vapors inside automobile tanks and barrels are given in Table III. TML to TEL ratios cf 25' to_4Q_were found under normal conditions. Two extreme examples' of the "spewing effect" were measured. -3- ! i* Gasoline from a barrel which hah been standing in the sun wa3 pumped into an automobile gas tank which was considerably colder. The mole ratio of TML to TEL-in the vapors expelled from the gas tank -was about 11. Table III From Auto Fill Pine During Filling* . b6.'JF During Filling, oOF Filling Tarn Gasoline Into Cool Gasoline i From Half-Full Barrels At ''Equilibrium, ' 50 u? At "Equilibrium," 3oF Emptied of Vapors bef0 re Samp1ing Estimated Error Lead Concentration (Mierograms/Cu Ft) Mole Ratio TML TEL TML/TEL 2000, 1400 950 50 40 4o yyx c ; 90 ii 1400 l6 00 200* * ` +200 40 60 90 +20 55 07 : 2.2 1. t . *Gasolines contained 3 .1S g Pb per gallon Estimated error ^40 In a second experiment, gasoline vapors sampled immediately after removing the existing (equilibrium) vapors showed a molar ratio of about 2, as might be expected from the rapid nonequilibrium evaporation of gasoline. A marked effect of temperature on vapor pressure of lead alkyls is shown in Table I. The mole ratio of TML to TEL at 10CF is half that at 50F. Although this is undoubtedly an important effect, it was not detected in the "field11 studies reported in Table III. The "spewing effect4 and experimental error combine to wash out any observable temperature effect over the 20-p0F temperature range of the study. . Lead in the vapors from spilled gasoline and from " gasoline allowed to evaporate in open vessels probably const!- tutes the most serious health hazard associated with the . marketing cf alkyl lead compounds. Lead analyses of vapors above gasolines evaporating in open beakers are reported in Table IV. These low ratios of TML to TEL which result from the 0002997 -4- 'spewing effect' confirm preliminary da ta reported by anot he r laboratory. It appears that in cases representative of the most flagrant types of gasoline misuse, the difference .between TML and TEL is minimized to the extreme. Tao. Leao-ln-Air Above Gasoline in Coen Vessels ' mens ce 10 iS Lead Concentration {Mic rcgram3/Cu F t ) TML TEL 6o 20 15i ! "1 -J 2G 10 y 3 n y Mole Ratio TML/TEL 3-0 2.7 2.0 l.C 1 .0 burn~ A comparison was made of the amount of lead in vapors above gasolines containing TML and TEL. The calculated ratio of 67 moles of TML to 1 mole of TEL was not achieved over gasoline but. was approached with the le 3s volatile fuels, isooctane and toluene. Vapors collected from automobile gas tanks and barrels contained 25 to 40 times more lead as TML than as TEL. Under conditions 'which involve a rapid evolution of gasoline vapors, such as pouring a warm gasoline into cold gasoline, or drawing air across gasoline to cause rapid evaporation, the TML to TEL ratio is markedly reduced and approaches unity. The "spewing effect11 is here defined as the rapid evolution of gasoline vapors which takes place whenever the vapors are not in equilibrium with the liquid. This effect accounts for the much smaller- ratios of TML to TEL found by experiment than are calculated from vapor pressure considera tions. In cases of flagrant misuse of leaded gasolines involving spills or rapid evaporation from open containers, TEL closely approaches the same concentration as TML in the vaoors. K E 0002998 -b~ i These stad :.us one:-- one ratio oi TML to TEL above rasolines C0`!v;i.00 ridan : one v.iuae ..oad .level is consideiabljc Lower- tha n anticicated . Field sarr.oiins data and information cr. one re iative :cxic:r u3 or TEL and TML will be required to reach a 1seisin:-, on One relative exposure hazards associated with re pi tcing TEL wit: TML in gasoline. w. L. RICHARDSON R. E. NOBLE G . J. KAUTSKY 'J 0002999 -o -