Table 111.
Toronto Harm-ell Lamont 1955 Lamont 1956 Toronto HarTyell Lamont
lntercalibration
(U. S. Geological Survey standard) 204 206 207 208 1.44(0) 23.69 22,54 52.33 1.45(5) 23.64 22.61 52.30 1.44(4) 23.50 22,55 52.52 1.44(5) 23.56 22.50 52.50 Pb(CH3)a Gas Average of 5 runs PbClz; PbIz Solid Average of 7 runs Pb(CH3)d Gas 1955 Average of 30 runs 1956 Average of 10 runs
Stieff of the United States Geological Survey for provision of the lead standard for interlaboratory calibration. Elizabeth Hodges, William Knox, Vanfred Gwinner, Barbara Wolf, and Kalter R. Eckelmann deserve credit for assistance in various phases of the work. Suggestions given by Paul Gast have been very helpful. LITERATURE CITED
(1) Bate, G. L., Kulp, J. L., “Variations
as measured by Instruments I and 11. It is also evident that identical results can be obtained for lead ion or trimethyllead ion. The trimethyllead ion spectrum is preferred, because the reproducibility is slightly better. More important, if radiogenic leads are being analyzed, the hydride correction must be assumed from the preceding and following common lead analyses. Since the corrections on the trimethyllead ion spectrum are much less than for the lead ion spectrum, i t is much less sensitive to any changes in source conditions. The results in Table I1 shoJV that the analyses of Nier (6, 8) have a syetematically higher lead-204 content (average of 1.6%) and consequently lower lead-208 (average of 0.6%). Although it is difficult to determine which assays are closer to the absolute values, the results presented above represent different sources and tubes, and suggest
that the lead iodide vapor technique used by Kier was subject to systematic discrimination. An experiment is in progress at this laboratory to measure the discrimination directly in the Lamont tetramethyllead analyses by accurately mixing lead-208-free lead with monazite lead which has a high lead-208-lead-206 ratio. The results of Table I11 on a n identical sample show that different laboratories using the tetramethyllead and solid techniques agree to within about tn-ice the errors obtained b y Lamont using tetramethyllead with different techniques. This is considered excellent agreement for the present, but i t appears possible to reduce the error by factors of 2 to 3 with current equipment, ACKNOWLEDGMENT
The authors are grateful to the late
J. P. Marble, who furnished the original lead salts used by Xier, and to L. R.
in the Isotopic Composition of Common Lead and the History of the Crust of the Earth,” Columbia University doctoral thesis, 1955. (2) Collins, C. B., Farquhar, R. AI., Russell, R. D., Bull. Geol. SOC. Amer. 65, 1-22 (1954). (3) Dibeler, V. H., Mohler, F. L., J . Research Natl. Bur. Standards 47, p. 337-42 (1951). (4) Ducheylard, G., Lazard, B., Roth, E., J . chim. phys. 50, No. 10, 497 119531. ( 5 ) Fiquhar, R. RI., Palmer, G. H., Sitken, K. L., h‘ature 172,860 (1953). O., J . A m . Chem. SOC.60,1571 (6) Sier, -4. (1938). ( 7 ) Sier, A. O., Rev. Sei. Znstr. 18, 398 11947). (8) Sier, A. O., Thompson, R . IT-,Rlurphey, B. F., Phys. Rev. 60, 112 (1941). (9) Palmer, G. H., Aitken, K. L., J . Scz. Instr. 30, 314 (1953). RECEIVED for review July 25, 1956. Accepted September 22, 1956. Lamont Geological Observatory Contribution 213. Research initiated under tenure of S a tional Science Foundation predoctoral fellowship. Kork supported by the Research Division, U. S. Atomic Energy Commission, under Contract AT(30-1)1114.
Determination of Copper in Fuel Oil and Other Petroleum Products DAVID M. ZALL, RUTH E. McMICHAEL, and D.
W. FISHER
Naval Engineering Experiment Station, Annapolis, Md.
b An accurate and simple spectrophotometric method for the determination of copper in petroleum products uses neocuproine (2,9-dimethyl-l,10phenanthroline) for the development of the copper-neocuproine complex, which has a maximum absorbance a t 450 mp. The method is rapid and direct and requires no ashing of the sample. No pH adjustment or extraction is necessary. The results compare favorably with those of other more tedious and time-consuming methods.
D
stored in copper-base alloy tanks gradually dissolves small amounts of copper. This proc88
IESEL FUEL
0
ANALYTICAL CHEMISTRY
ess is accelerated when water finds its way into the same storage tank. Extremely small amounts of copper are harmless and present no corrosion problem (8),but when the copper content becomes considerable, it causes heavy deposits in the precombustion chamber of one type of Diesel engine. The determination of copper in the fuel, therefore, becomes important. Although a number of methods for the determination of copper are available, they either lack specificity or are subject to interferences of one kind or another (4, 7 , 10, 14, 16). The neocuproine method proposed by Gahler (6), however, eliminates most inter-
ferences and simplifies the determination of copper. The use of neocuproine also has other advantages that are apparent n-hen applied to the determination of copper in fuel oil and other petroleum products. I n its application to the analysis of Diesel fuel or other petroleum products, the determination resolves itself into two parts: treatment and solution of the sample, and the spectrophotometric evaluation of the copper content. I n the treatment and solution of samples of organic materials, the three accepted procedures usually followed are wet ashing, dry ashing, or acid extraction, whether the final determination is made
The solution was cooled. diluted nitli distilled water, and transferred to a 100ml. volumetric flask and diluted to the mark. 9 suitable aliquot was transferred to a separatory funnel. The copper was then determined according to the procedure outlined by Gahlei ( 5 ) . The results ale tabulated in Table I iiishing RIethod). A second set of samples of the same fuel oil nas treated by the method outlined here. The results ale listed in Table I under Diiect Method. Comparison of the values obtaincd by the direct method with those obtained using the ashing procedure indicates that the former method is accurate nithin the limits of its repioFigure 1. Absorption spectrum of copper-neocuproine ducibilit y. comp Iex Copper naphthenate x i s dissolx-et1 in isopropyl alcohol to give a solution of about 30 p.p.ni., and the copper T\ as gram of 2,9-dimethyl-l,lO-phenanthro- determined electrolytically. The calspectrogi aphically, polarographically, or line (neocuproine) and 1 gram of hydrospectrophotometrically (1-3, 6, 11-12, ibration curve was prepared from this quinone in 1 liter of isopropyl alcohol. 13, 15). standard copper naphthenate solution. This solution is stable for Iveeks. In the spectrographic determination Samples of this solution nere ashed of copper in turbine oils Barney ( 2 ) and copper content R as determined by PROCEDURE ashed the sample in the preliminary the method of Gahler (6). Samples of treatment for analysis. Martens and A suitable sized sample of fuel oil the same solution neie treated directly Githens (10) employed m-et ashing as a or turbine oil is pipetted into a 25-ml. using the neiv approach. Table I volumetric flask. If copper content is preliminary treatment in the deters h o m a ieasonablc agieement between high the sample is pipetted into a 100mination of copper in dyes and rubber the tn-o methods. ml. flask and diluted to the mark with chemicals. Hackett ( 6 ) extracted the chloroform. A suitable aliquot of this copper from niineral oil v i t h alcoholic is taken and pipetted into the 25-m1. hydrochloric acid and determined it flask, to which is then added 10 ml. colorimetrically 11-ith 2,2 - diquinolyl. of neocuproine solution. The sample Table I. Comparison between Ashing Khile these methods of treatment are is diluted to the mark n-ith chloroform, Method and Direct Method inevitable in most cases, they are timethoroughly mixed, and allorred to stand Copper, P.P.RI. consuming and require considerable for about hour. Another sample of -%lahing Direct the same size, the blank, is pipetted manipulation to avoid loss of material. Snniple method method into another 25-m1. flask and diluted Therefore, attention has been directed to the mark with chloroform. The to the development of a rapid proce1 105.0 105 0 transmittance is read on the spectro2 28.0 28 5 dure that would eliminate ashing or photometer a t 454 nip (Figure 1) .? 4_ .94 9 acid extraction. against the sample blank in the ref4 0 87 0 84 B study was made of the compatibility erence cell. The copper concentration 5 1.0 1.02 of fuel oil with the solution of neo6 36.0 34 8 is determined by referring to a standard 7 42 0 43 2 cuproine in isopropyl alcohol. Chlorocurve prepared by taking different 8 40.0 40.2 form was used for dissolving or diluting volumes of a standard copper naph9 4 0 4.3 thenate solution through the given the fuel, which is not I-ery soluble in 10 5.0 5.6 procedure. isopropyl alcohol. The chloroform thus became a mutual solvent for the fuel Copper Xaphthenate Standard, 50 P.P.X. RESULTS and the neocuproine-copper complex. 49 50 When neocuproine was added to the 49 52 The results obtained on samples of 51 52 copper - containing fuel dissolved in diesel fuel and copper naphthenate 49 50 chloroform, the reaction was almost are shown in Table I. For the ashing instantaneous. Also, when copper method 1 to 5 grams of sample was dry naphthenate or lubricating oils were ashed-Le., ignited in a platinum dish treated with neocuproine and diluted and alloxved to burn to completion; INTERFERENCES with chloroform or a niixtuie of chlorothe remainder of oiganic matter was form and isopropyl alcohol, a color burned a t a low heat. The ash in the Of the 56 metals tested by Luke and development occurred within 5 minutes. platinum dish n as dissolved in hot Campbell (Q), none interfered in the Advantage was taken of this property of hydrochloric acid (1 to 1) and the concopper-neocuproine extraction. Howneocuproine to develop a method for tents were transferred to a 250-ml. ever, small amounts of sulfide, as rethe direct determination of soluble beaker. The solution in the beaker ported by Gahler (j),caused serious copper in fuel oil. n-as oxidized n-ith nitric acid, then 5 interference. Some organic inhibitors ml. of concentrated sulfuric acid wis used in the fuel oil, such as i\'-phenyl-1APPARATUS A N D REAGENTS naphthylamine, S,N'-tetramethyl-diaadded and contents allowed to evaporate to the appearance of white fumes of minodiphenylmethane and other amino All the absorbance measurements sulfur trioxide. This fuming n as necescompounds (below O.lyc),caused no tvere made with a Beckman Model DU sary in order to eliminate all traces of interference. Some mercaptans interspectrophotometer. All reagents connitric acid, n hich, if allowed to remain fered, hon-ever, causing low results. formed to ACS specifications. Xeocuproine Solution. Dissolve 1 in the sample, leads to low results. Mercaptobenzothiazole and its sodium VOL. 29, NO. 1, JANUARY 1957
89
salt (in concentrations below 0.1%) caused low results. This interference was overcome b y a large excess of neocuproine. I n the case of butyl zymate the color faded rapidly (within 2 to 5 minutes) and no amount of excess neocuproine was of any assistance in overcoming the interference. Barium sulfonate and sodium sulfonate interfered in concentrations above 0.1%. This interference was more of a retarding nature, for after standing for about 4 hours, the color was completely developed. ADVANTAGES
The simplicity of the method described is not only due t o the use of neocuproine, but also to the novel approach used in the treatment of the sample. It is thus possible to determine the copper content directly without resorting to any prior treatment of the sample. The advantages may be listed as follows:
No ashing or acid extraction
KO acid dissolution of ash KObuffering or chelating
KOp H adjustments No extraction It is essential that the proper sized sample be used. As a time saving measure and to determine the proper size t o take, a simple spot test was employed. A drop of a standard solution of copper naphthenate was placed on a spot plate and the color developed by adding 1 drop of neocuproine and 1drop of chloroform. A drop of the unknown !vas treated in a similar manner. Comparison indicated the size of the sample to be taken for analysis. Using this simplified procedure i t is possible to determine copper in fuel oil in a matter of a few minutes per sample. The copper in a single sample can be determined in less than 1 hour. The rapidity and simplicity of this method stand out when compared with the old ashing or acid extraction procedure. LITERATURE CITED
( 1 ) Andrus. S..dnalvst 80. 514 (1955). (2) Barney; J.' E., IT, A4xk~. CHEW'26, 567 (1954). (3) Breckenridge, J. G., Lewis, R. IT.>
Quick, L. A., Can. J . Reseurch B17, 258 (1939). (4) Elwell, W. T., Analyst 80,508 (1955). (.5 ,) Gahler. A. R.. ANAL. CHEM. 26. 577-9 (1954).' (6) Hackett. C. E. S.. Anal. Chim. Acta 12, 3k3-62 (1955). (7) Hague, J. L., Brovm, E. D., Bright, H. ii., J . Research h a t l . Bur. Standards 47, 380 (1951). (8) LaQue. F. L.. Corrosion 10. 391 (i954). (9) Luke, C. L., Campbell, bI. E., ANAL.CHEM.25, 1588 (1953). (10) Martens, R. I., Githens, R. E., Sr., Zbid.,24,991-3 (1952). (11) Melvin, E. H., Hawley, J. E., J . Am. Oil Chemists' SOC.2 8 , 346 (1951). (12) Michel, G. X, Maron, N., Anal. Chim. Acta 4, 542-50 (1950). (13) Pohl, Heinz, Zbid., 12, 54 (1955). (14) Rees, W. T., Analyst 75, 160-6 (1950). (15) Sinyakova, S. I., Borovaya, 11. S., Gavrikova. X. A . Zhur. Anal. Khim. 5, 330-8 (1950). (16) Rilkins, D. H., Smith, G. F., Anal. Chim. Acta 9, 538 (1953). RECEIVEDfor review May 17, 1956. Accepted September 6, 1956. Pittsburgh
Conference on Analytical Chemistry and Applied Spectroscopy, Pittsburgh, Pa., February 28, 1956. The opinions expressed in this paper are those of the authors and do not reflect the views of the Navy Department or of the Naval Service a t large.
Colorimetric Determination of Titanium in Polyethylene RICHARD A. ANDUZE Research and Engineering Division, Monsanfo Chemical Co., Dayton, Ohio
,Titanium can be determined colorimetrically in polyethylene by a method which involves ignition in an oxygen bomb, separation from interfering elements by coprecipitation with zirconium arsenate, and reaction of titanium with hydrogen peroxide in sulfuric acid solutions. The color produced is measured at 41 0 mp. Amounts as low as 3 p.p.m. can be determined.
V
colorimetric determinations of small amounts of titanium have been described in the literature (3, 6-7). Of the many color reagents proposed, hydrogen peroxide is the most readily available and needs no further purification. I n suIfuric acid-titanium solutions, hydrogen peroxide causes a yellow color to develop. This color is due to the formation of the complex peroxidic ion furnished by the peroxodisulfatotitanic acid freed b y the reaction between titanium, peroxide, and sulfuric acid. This colored ion has been considered an anion, but recent work indicates t h a t i t is a cation (2). The 90
ARIOUS
ANALYTICAL CHEMISTRY
color is stable for oyer 2 years ( 1 ) . However, the necessary separation of interfering cations is often tedious and may lead to losses of titanium (5). For the determination of titanium in polyethylene or other organic materials, a method based on the formation of the complex titanium-peroxide ion has been developed. It differs from methods heretofore published in the method of decomposing the sample, and in that the sensitivity of the method has been extended. This method will detect 3 p.p.m. of titanium in a 1.5-gram sample. Vanadium, nickel, chromium, manganese, molybdenum, copper, silicon, and iron do not interfere, because titanium is separated from them by coprecipitation with zirconium arsenate (8). The mixed titanium and zirconium arsenate are dissolved in hydrochloric acid, and the color of the complex peroxidic cation is developed by hydrogen peroxide in sulfuric acid and measured spectrophotometrically. APPARATUS AND REAGENTS
Parr oxygen bomb, self-sealing type,
with 5-ml. Vitreosil crucibles, wide-form. International clinical centrifuge, Model CL . Beckman spectrophotometer, Model B, with blue-sensitive phototube, 23-mm. vertical and 50-mm. horizontal cylindrical cells. Zirconium sulfate solution, 0.1M. Dissolve 2.83 grams of zirconium sulfate in 100 ml. of 2.4N hydrochloric acid. Warm if necessary to effect solution. Filter before use if not perfectly clear. Arsenic acid solution, 1 . 5 X . Dissolve 21.3 grams of arsenic acid in 100 ml. of water, heating to effect solution. Filter before use if not perfectly clear. Rash solution. Dilute 1 ml. of arsenic acid solution to 100 ml. Kith 2.4N hydrochloric acid. Hydrogen peroxide, 3% analytical reagent grade. Sulfuric acid, 36N and 3.6,!*. Hydrochloric acid, 2 . 4 5 and 1.2.\-. RECOMMENDED PROCEDURE
Sample Preparation. Weigh 1.5 grams of sample in a 5-ml., form Vitreosil crucible. R u n a concurrently, using 2 ml. of
about wideblank ethyl
