Document 5LgqLa5g7x2q0qQ7qa5rGKzEz

Sth November, 1938. T.R.A. Bevati, Esq., Sthyl Export Corpn., Thames House. imtiknock Fluid - Belgium Dear Hr. Devan, In confirmation of our telephone conversation, I am- attach ing hereto a complete report of the investigation which has been made into the "antiknock fluid" received from Belgium. You will note from the summary this this material, when added to a gasoline of 65 octane number in the proportion of 5 ccs. per U.S. gallon produced a rise of 5 octane numbers; the antiknock effect appears to be due to a volatile lead compound, probably tetraethyl lead, of which about \Z% is present. The fluid appears to be a mixture of wood tar with tetra ethyl lead, ethylene dibrorrdde and a very small proportion of ether, and is itself of dark brown colour, containing suspended matter. On mixing with gasoline it rives a cloudy liquid. The material as received also contained a small proportion of iron rust, due apparently to corrosion of the container by the fluid. You will note that the analysis of this new sample differs from that of the former sample which was tested, in that the present sample does possess an arfiknock value, but that this is due to tetraethyl lead not present in the former sample; furthermore the composition of the material appears to have been changed in that no alcohols are present. It would be somewhat difficult to establish whether the lead present in the fluid has been added as T.k. L. or ethyl fluid, but from a review of the analytical figures we feel that it is more probable that it has been added in the form of tetraethyl lead, since there is no indi cation, for example, of the presence of ethylene dichloride which should be present if ethyl fluid were used to obtain the antiknock effect. Yours very truly, ^ 00134^0 N17451 Sample of alleged "Anti-knock fluid" ex ______ Sthyl Gasoline Corporation______ The fluid has been examined as to its antiknock properties, and, in con centration of 5 ml. par !T.S. gallon, has been found to raise the octane number of standard 65 octane spirit to 70.5 The following elements were shown to be present - carbon, hydrogen, (oxy gen) , nitrogen, chlorine, lead, and iron. The substance was reddish-brown in colour, having an odour of pyridine bases, and containing a quantity of crystalline material in suspension. On fil tration, followed by washing of the residue with industrial methylated spirits, colourless needles were obtained which appeared to be the hydrochloride of a secon dary or tertiary organic base. 1% of the sample was soluble in brine; the extract, however, contained no ethyl alcohol. 12% was soluble in hydrochloric acid representing the bases present and 4% was soluble in caustic soda solution representing the proportion of acids present. The following figures represent the results of the I.P. distillation test (fuming began at 16CPC and a crystalline non-volatile substance was separated) ;- Temperature Percentage I.B.Pt. 75?C 80 *0 90 100 110 120 150 140 150 160 170 180 190 200 210 220 230 240 250 260 1.0% 1.5 2.0 5.0 4.5 7.5 18.0 27.0 34.0 40.0 44.0 48.0 49.0 51.0 54.0 58.0 65.0 74.0 82.5 F.B.JPt. 264*0 Recovery 88% The filtered sample was submitted to fractionisation in vacuo. The re sults of tests upon the various fractions are summarised in the next Table. 2- Antiknock prop. Lead Content in cone. 3 rals. Gravity ___________ per U.S. gallon ' Remarkte. 35-40=0 120-12 5C nil nil 60-800 (Grams.) 80-1000 (6 mms.) 100-1200 (6 mms.) 120-135*0 (6ams.) Residue nil nil nil nil present nil ethereal odour nil 1.970 odour resembling chloroform nil nil 0.9805 nil 1.139 nil 1.184 The residue was insoluble in spirit; on extraction ith bot water the solution gave reaction for lead and chloride. It seemed, therefore, that the lead compound present in the original flu id underwent decomposition under the conditions of distillation with the separation of lead chloride. The lead content of the fluid was found to be 8.43$ which would corres pond to a tetraethyl lead content of 13.11%. In order to minimise decomposition a rapid distillation of the sample voa carried out using the I.P. apparatus. A 30% fraction distilling below 154C was taken off and both that and the residue were examined as to their 'antiknock proper ties. The former, in concentration of 3 ml./U.S. gallon raised the octane number of 65 octane spirit to 67; the latter to 70. Both fractions contained lead. In an endeavour to prevent decomposition the sample (100 ccs.) was fil tered, washed, with hydrochloric acid, 'water, caustic soda, and filtered through dry filter paper. The dry filtrate was submitted to fractionisation in vacuo. No sep aration of lead chloride or fuming occurred, the reactive halogen compound having been removed. "esults are summarised in the following table Distillation range Volume Lead Content Gravity Hemarks 35-135 25 ml. positive 2.043 35-90 (6 mm.) 19 cc. positive .1.604 90-110 (6 mm.) 25 cc. positive 1.228 90-110 (2 mm.) 13 cc. negative 1.263 'esidue slight fluid The lead content of the 35-90 (6 mm.) fraction was found to be 20.4% corresponding to a tetraethyl lead content of 31.9%. In concentration of 3 ml./U.S. gallon, the octane number of 65 octane pirit was raised to 75.0. 0013492 <s V >- - 3- . - AARY '. The fluid is a dark brown liquid of 1.496 specific gravity, in which sus pended material is present, with an odour similar to that of wood tar, and on mix ing with gasoline gives a cloudy liquid. The fluid contains a volatile compound of lead in admixture with wood tar, ethyl ether, and ethylene dibromide. There is present 2% ethyl ether and approxi mately 33% ethylene dibromide, while over 50% of wood tar containing about 25% or ganic bases and 10% organic acids is also present. It was proved that a volatile lead compound was present, and as much as 32% was concentrated in one particular fraction by careful fractionation. The original fluid produced a rise in octane number of 5 when 5 ml. was added to one U.S. gallon of a gasoline of 65 octane number; this indicates that most, if not all, of the antiknock effect of the fluid is due to tetraethyl lead, and it was estimated that 13% of lead tetraethyl was present in the original fluid. He 0013493