Talanta

ELSEVIER Talanta 44 (1997) 2103 2108

The determination of wear metals in used lubricating oils by flame atomic absorption spectrometry using sulphanilic acid as ashing agent
E.J. Ekanem *, J . A . L o r i , S . A . T h o m a s
Department ~[~Chemisto,, Ahmadu Bello Unit,ersity, Zaria, Nigeria

Received 5 November 1996: received in revised form 24 February 1997: accepted 27 February 1997

Abstract

A simple and reliable ashing procedure is proposed for the preparation of used lubricating oil samples for the determination of calcium, magnesium, zinc, iron, chromium and nickel by flame atomic absorption spectrometry. Sulphanilic acid was added to oil samples and the mixture coked and the coke ashed at 550C. The solutions of the ash were analysed by flame AAS for the metals. The release of calcium, zinc, iron and chromium was improved by the addition of sulphanilic acid to samples. The relative standard deviations of metal concentration results in the initial oil samples were 1.5% for Ca (1500 mg 1 ~ level), 0.3% for Mg (100 mg 1 " l level), 3.1% for Zn (1500 mg 1 level), 0.7% for Fe (500 mg 1-1 level), 0.02% for Cr (50 mg I i level) and 0.002% for Ni (10 mg 1 i level). The optimum sample size for efficient metal release was 20 g while the optimum sulphanilic acid to oil ratio was 0.05 g per gram of oil for Zn and Cr and 0.10 g for Ca and Fe. Results obtained by this procedure were highly reproducible and comparable with those obtained for the same samples using standard procedures. 1997 Elsevier Science B.V.
Keywords: Aided-ashing; Lubricant; Metal determination

1. Introduction

The performance characteristics o f lubricating oils are usually improved by the additives incorporated to improve the colour, pour point, viscosity, antiwear, antifrictional, antifoaming, oxidation and corrosion inhibition properties. These additives contain various metals in varying amounts. In use, a lot more o f some o f these metals are added to the lubricating oils from * Corresponding author.

friction and wear [1,2]. A simple, rapid and accurate analytical procedure is needed for routine determination o f these elements in order to provide an indication o f the degree of wear o f the oil wetted parts within an engine. A specific application is in the analysis o f used lubricating oils to indicate potential failure in internal combustion engines as well as the metal pollution potentials o f such oils if they are disposed directly into surface water or on land. A l t h o u g h wear metal particles are not uniformly distributed in oils, they are generally deter-

0039-9140/97/$17.00 {2 1997 Elsevier Science B.V. All rights reserved. PII S0039-9140(97)00086-6

2104

E.J. Ekanem et al./ Talanta 44 (1997) 2103 2108 2.2. Working standard solutions

mined by using techniques traditionally applicable to the determination of metals in homogeneous solutions [3]. Various methods have been reported for the preparation of samples prior to instrumental analysis [3-7]. The possible loss of certain naturally occurring volatile constituents during unaided ashing of petroleum or its products has been reported [8-10]. Hence, the determination of trace metals in petroleum oils has relied on ashing procedures that aim at preventing metal losses by volatilization [11,12] during ashing. Benzenesulphonic acid has been applied as ashing agent for the determination of Ni and V in petroleum distillates [11,13] while p-xylenesulphonic acid has been used for ashing petroleum distillates [12] and crude petroleum [14] for the determination of trace metals. Unlike these reagents which are hygroscopic and difficult to handle and store [15], sulphanilic acid has a very long shelf life and was investigated as ashing reagent in this work. The purpose of the present work was to establish the optimum conditions for the ashing of used lubricating oils applying sulphanilic acid to contain any volatilization losses.

ZnO, (NH4)2SO 4. and chromium powder were dissolved appropriately and diluted with water to obtain 1000 mg 1-J stock solutions of Ca, Mg, Zn, Fe, Ni [16] and Cr [17]. Working standard solutions were obtained from the respective stock solutions by appropriately diluting them with water introducing 5% (m/v) lanthanum as the nitrate in the case of Ca and Mg to control the interferences of ions like phosphate, aluminate and silicate. CaCO3, MgC126H20,

FeSO4"6H20, Ni(NO3)2"6H20

2.3. Recommended procedure

2. Experimental
2.1. Reagents, apparatus and samples

All reagents used in this work were of analytical grade. Sulphanilic acid was obtained from BDH. Doubly distilled water was used in preparing all aqueous solutions. The muffle furnace used was a Carbolite (Sheffield, UK) capable of maintaining constant temperatures in the range 20-1200C. The Perkin Elmer 290 B atomic absorption spectrophotometer, equipped with a premix chamber and a single slot burner head operated on airacetylene flame, with automatic background correction and mono-element hollow cathode lamps was used for solution analysis. The parameters selected on the spectrophotometer for the determination of the test metals conform with those recommended by Perkin Elmer for this spectrometer. The oil samples analysed were collected from private vehicles and automechanic workshops in Zaria, Nigeria.

A mixture of 10.0 g used lubricating oil and 1.0 g sulphanilic acid in a 250 cm 3 pyrex beaker was heated in a heating mantle in the fume cupboard. The smoke issuing from the beaker was ignited and the set-up left until the flame was extinguished. The coke formed was muffled at 550C until the ash was free of carbon particles; this required 45 min. The ash was allowed to cool before it was dissolved in 20 cm 3 of 6M HC1 and digested in a boiling water bath to obtain a clear solution. This solution was made up to 50 cm 3 with distilled water in a volumetric flask, introducing 5% (m/v) La and analysed by FAAS for Ca, Mg, Zn, Fe, Cr and Ni against a blank solution prepared by digesting 20.0 cm 3 of the 6M HC1 in the boiling water bath in the same way as the sample digests and making up to 50 cm 3 with distilled water as done for the sample solution.

3. Results and discussion

3.1. Optimization of ashing conditions 3. I. 1. Temperature and time Optimization of parameters was based on 10 g replicates [18] of samples. Temperatures were observed in the range 400-650C for durations of 15-120 min. The progress and completeness of ashing was visually determined. The lowest muffling temperature that yielded carbon-free ash was 550C and this was achieved

E.J. Ekanem et al./ Talanta 44 (1997) 2103 2108

2105

Table 1 Effect of ignition Sample treatment Coking time (min) Metal concentration (mg 1 ~) Ca Without ignition With ignition 20 18 1470 1470 Mg 88 88 Zn 1350 1390 Fe 387 426 Cr 32.5 35.6 Ni 9.5 9.6

in the shortest time of 45 min. Though a higher muffling temperature could be selected for a more rapid ashing, 550C was regularly applied in this work to avoid the disfiguration of the pyrex beakers that may occur at higher temperatures [19]. The oxides of the metals determined are known to be stable even above 1000C [20], hence volatilization of the metals would be prevented by rendering them as oxides in the ash at 550C. This ashing temperature has been recommended by the Institute of Petroleum [21] in a procedure that does not specify coking and muffling times and includes an overnight delay and a filtration step. The present procedure has eliminated the need for ash solution filtration and the attendant delay by achieving carbon-free ash in a short time. The chances of error arising from contamination during the filtration are removed in this procedure.

3.1.2. Ignition of samples Coking time was the duration up to the cessation of smoke when the smoke issuing from the samples was not ignited. In the cases where the smoke was ignited as it issued from the sample, coking time was the time taken for the smoke to finish burning and the flame to be extinguished. Satisfactory coking was achieved in 20 min without ignition and in 18 min when the smoke was ignited. Ignition did not, however, alter muffling time because the coke derived with and without ignition both yielded good ash in 45 min of muffling. The ashes obtained were completely soluble in 20 cm 3 of hydrochloric acid and a clear solution was obtained after 2 min of digestion. Therefore, there was no need to filter the solution as recommended by the direct ignition method for the determination of additive metals in unused lubri-

cating oils [21]. The results presented in Table 1 for the determination of the test metals in sample number 1 were typical of those obtained for all samples in this work and show that there was increased metal retention in the ignited sample particularly for Zn, Fe and Cr. For these metals, the ignition allows metal complexes that would have escaped in the smoke to burn and return their metal constituents to the coke as metal oxides. The differences in the determinant levels of ignited and unignited samples give an indication of which determinants are more prone to volatilization losses in lubricating oils. Though there was no change in the levels of some of the test metals, the enhancement in the levels of even a few of the test metals advised the adoption of ignition of lubricating oils for determining trace metals. A similar advantage of sample ignition while coking has been reported for petroleum crudes [14] and petroleum distillates [22].

3.2, Sulphanilic acid as ashing reagent

Various levels of sulphanilic acid were added to 10 g replicates of oil analysed by the recommended procedure and yielded results as detailed in Table 2. The addition of sulphanilic acid to the samples did not alter the ashing efficiency of the samples: it did not alter either the ignition time or the muffling time and the ash produced in the presence of it was still completely soluble in 20.0 cm 3 of 6M HC1. The muffling of 1.0 g of sulphanilic acid alone at 550C under similar conditions to those applied for ashing oil samples left no residue at the end of 45 min. This implies that SA did not contribute to the ash. The digest of 2.0 g of sulphanilic acid in 6 M HC1 yielded blank levels of all the metals determined. The sulphanilic

2106 Table 2 Sulphanilic acid (SA) as ashing reagent Mass of SA added (g)

E.J. Ekanem et al. / Talanta 44 (1997) 2103 2108

Metal concentration (mg 1 ~) Ca Mg 88 88 87 88 88 88 89 89 89 89 89 Zn 1390 1390 1393 1395 1397 1400 1400 1400 1400 1400 1400 Fe 426 428 428 434 455 456 457 458 468 468 468 Cr 35.2 35.7 36.1 36.3 36.7 37.3 37.3 37.0 37.0 37.3 37.3 Ni 9.6 9.6 9.6 9.7 9.7 9.7 9.7 9.7 9.7 9.7 9.7

0 0.I 0.2 0.3 0.4 0.5 0.6 0.8 1.0 1.5 2.0

1470 1472 1475 1485 1498 1500 1500 1501 1502 1500 1501

acid was therefore shown to be free from the determinants. The results presented in Table 2 for 10.0 g o f o i l sample 1 show that there was an improvement in metal release especially for Ca, Zn, Fe and Cr as the sulphanilic acid to oil ratio increased until peak values were obtained at 0.05 g of SA for Zn and Cr and 0.1 g for Ca and Fe per gram of oil. Once obtained, the peak determinant levels persisted when higher ratios of sulphanilic acid were applied, up to 0.2 g of SA per gram of oil. The same trend was observed also for samples other than sample number 1. That peak metal release is achieved at different ratios of releasing agent for different determinants reflects the differences in stabilities of the metal complexes in which the determinant species are encaged in the oil samples [23]. These results suggest that the zinc and chromium metallocomplexes are less thermodynamically stable than the calcium and iron ones. Similar differences in the stabilities of porphyrins and metallocomplexes of metals occurring in crude oil have been reported [14]. The addition of 0.1 g of sulphanilic acid per gram of oil to samples was adopted in this work and is recommended for the ashing of used lubricating oils for the determination of trace metals by FAAS. Its stability and long shelf life make it preferable to other similar reagents like benzene-sulphonic and p-xylenesulphonic acids, which are hygroscopic and require special handling [14].

3.3. Evaluation o f the procedure

When the same oil was ashed in six replicates and the ash analysed for the test metals in the replicates, the results showed that the recommended procedure is highly reproducible with relative standard deviations (% rsd) of 1.5, 0.3, 3.1, 0.7, 0.02 and 0.002 for Ca, Mg, Zn, Fe, Cr and Ni, respectively at the respective concentration levels 1500, 100, 1500, 500, 50 and 10 mg 1 J. All five samples were analysed in triplicates by this procedure, the direct ignition procedure [22] and the Universal Oil Product (UOP) method 800-79 [24] and the average results are compared in Table 3. F r o m observation, the results indicate that the values obtained by the direct ignition procedure are generally lower than those obtained by either the recommended or U O P method 80079 procedures; the values obtained by the recommended procedure are generally closer to those obtained by the UOP procedure. A regression analysis of the results reveals a closer fit between the recommended procedure and the standard UOP procedure than between the direct ignition and the U O P procedures. The respective regression lines are y = 1.004x r - 1.884 and
y = 1.017x i + 5.356

r = 0.9993

r = 0.9982

E.J. Ekanem et al./Talanta 44 (1997)2103 2108

2107

Table 3 Comparison of sample results obtained by three ashing procedures Sample No. Determinant Measured levels (mg 1 ~) Recommended procedure 1 Ca Mg Zn Fe Cr Ni Ca Mg Zn Fe Cr Ni Ca Mg Zn Fe Cr Ni Ca Mg Zn Fe Cr Ni Ca Mg Zn Fe Cr Ni 1508.4 90.0 1411.8 472.2 36.0 9.7 1739.9 93.6 1475.5 440.5 28.0 9.3 1638.5 85.0 1441.2 412.7 22.0 9.3 878.4 32.1 1034.3 281.8 64.0 11.3 641.9 28.5 872.6 388.9 67.0 11.7 Direct ignition procedure 1503.4 87.7 1372.6 346.8 36.0 9.3 1706.1 92.6 1446.1 428.6 26.0 9.3 1621.6 82.0 1446.1 412.7 22.0 9.3 810.8 30.5 1029.4 264.8 63.0 10.3 608.1 29.8 853.0 388.9 66.0 11.3 UOP method 800-79 1513.4 90.7 1401.9 460.3 35.0 10.3 1756.8 94.9 1446.1 452.4 28.0 9.3 1706.1 88.7 1446.1 420.6 21.0 9.3 844.6 31.1 1012.8 281.8 64.0 11.3 641.9 29.5 857.9 388.9 69.0 11.7

where y represents U O P values, x r represents values f r o m the r e c o m m e n d e d p r o c e d u r e , x i represents values f r o m the direct ignition p r o c e d u r e a n d r is the c o r r e l a t i o n coefficient. T h e a p p l i c a t i o n o f S t u d e n t ' s t-test to the d a t a o f T a b l e 3 reveals t h a t the results o b t a i n e d b y the direct ignition p r o c e d u r e are significantly different f r o m those o b t a i n e d by the U O P m e t h o d 800-79 while there is no significant difference between the results o b t a i n e d by the r e c o m m e n d e d a n d the U O P procedures. These statistics s u p p o r t the a r g u m e n t that while the U O P m e t h o d 800-79 a n d the r e c o m m e n d e d p r o c e d u r e s realise m a x i m u m release o f metals, the direct ignition p r o c e d u r e fails to achieve maxi-

m u m m e t a l release, hence the generally low values o b t a i n e d by the latter p r o c e d u r e a n d the signific a n t difference between it a n d the f o r m e r two procedures. T h e a d v a n t a g e o f sulphanilic acid causing e n h a n c e d m e t a l release as i n d i c a t e d in T a b l e 2 is reflected in the results o f these statistical analyses. T h e r e c o m m e n d e d p r o c e d u r e has an a d d i t i o n a l a d v a n t a g e in being faster t h a n the U O P m e t h o d 800-79.

4. Conclusion
U s e d l u b r i c a t i n g oil m a y be c o n v e n i e n t l y c o k e d a n d ashed at 550C in 45 min for d e t e r m i n i n g

2108

E.J. Ekanem et a l . / Talanta 44 (1997) 2103-2108

wear and other metals by flame atomic absorption spectrometry. Sulphanilic acid applied as ashing agent improves the release of metals and the accuracy of the analytical results. The procedure reported is simple, rapid and precise.

References
[1] O.I. Milner, Analysis of Petroleum for Trace Elements, International Series of Monographs on Analytical Chemistry, vol. 14, Pergamon Press, Oxford, 1963. [2] C.R. Hodgkins, J. Hansen, Anal. Chem. 26 (1954) 1759. [3] J. Hernandez-Mendez, L. Polo-Diez, A. Bernal-Melchor, Anal. Chim. Acta 108 (1979) 39. [4] C.S. Saba, W.E. Rhine, K.J. Eisentraut, Anal. Chem. 53 (1981) 1099. [5] C.M. De la Gaurdia, C.A. Salvalor, N.V. Berenguer, Analusis 8 (1980) 448. [6] J.L. Fabec, M.L. Ruschak, Anal. Chem. 57 (1985) 1853. [7] C.S. Saba, K.J. Eisentraut, Anal. Chem. 49 (1977) 454. [8] J.H. Karchmer, E.L. Gunn, Anal. Chem. 24 (1952) 1733. [9] O.I. Milner, J.R. Glass Jr., J.P. Kirchner, A.N. Yurich, Anal. Chem. 24 (1952) 1728. [10] L.W. Gamble, W.H. Jones, Anal. Chem. 27 (1955) 1456. [l l] J.E. Shott Jr., T.J. Garland, R.O. Clark, Anal. Chem. 33 (1961) 506. [12] W.A. Row, K.P. Yates, Anal. Chem. 36 (1963) 368.

[13] Institute of Petroleum, Method IP 286/77(a), Standards for Petroleum and its Products, Part 1, Methods for Analysis and Testing, vol. 2, 1985, pp. 286. [14] A.P. Udoh, S.A. Thomas, E.J. Ekanem, Talanta 39 (1992) 1591. [15] A.P. Udoh, Ph.D. Thesis, Ahmadu Bello University, Zaria, 1989. [16] P.J. Whitehead, Pye Unicam Atomic Absorption Data Book, Pye Unicam, Cambridge, 1975. [17] A.I. Vogel, A Textbook of Quantitative Inorganic Analysis, 3rd ed., ELBS-Longmans, London, 1961. [18] ASTM method D-811-82, Annual Book of ASTM Standards, vol. 5.01, 1984, pp. 341. [19] A.P. Udoh, S.A. Thomas, E.J. Ekanem, Bull. Chem. Soc. Ethiop. 4 (1990) 13. [20] R.C. Weast (Ed.), Handbook of Chemistry and Physics, 55th Edition, CRC Press, Ohio, 1974, pp. B. 63. [21] Institute of Petroleum, Methods IP 110/82 (a), IP 111/82 (b) and IP 117/82 (c), Standards for Petroleum and its Products, Part 1, Methods for Analysis and Testing, 45th ed., vol. 1, 1985, pp. 110.1 (a), 111.1 (b) and 117.1 (c). [22] A.P. Udoh, S.A. Thomas, E.J. Ekanem, AJST (B) 5 (2) (1991) 8. [23] R.H. Filby, in: T.F. Yen (Ed.), The Role of Trace Metals in Petroleum, Ann Arbor Science Publishers, Ann Arbor, Michigan, 1975, pp. 31. [24] Universal Oil Products, UOP method 800-79, Laboratory Test Methods for Petroleum and its Products, vol. 2, UOP Process Division, 1979, pp. 1-12.