42 Analyst, January, 1973, Vol. 98, pp. 4248 The Direct Complexometric Determination of Metals in Unused Lubricating Oils and Additives by Automatic Non-aqueous Potentiometric Titration* BY J. Tu’. WILSON AND C. 2. MARCZEU’SKI (The British Petroleum Company Limited, BP Research Centre, Chertsey Road, Sunbury-on-Thames, Middlesex) Methods are described for the determination of calcium, barium, zinc and magnesium in any combination in additive concentrates and unused lubricating oils. The samples are dissolved in a mixed aqueous - organic solvent and titrated directly with 1,Z-bis (2-aminoethoxy)ethane-NNN’N’- tetraacetic acid [ethylene glycol bis(2-aminoethy1)tetraacetic acid ; EGTA] or 1 , 2-diaminocyclohexane-NNN’N’-tetraacetic acid (DCTA) , thc end-points being indicated by metal ion sensitive electrodes.By use of a masking agent and the appropriate buffer solutions the individual metals in any combination can be determined following a maximum of three titrations. The results are compared with those obtained by established procedures. The accuracy of the complexometric methods is comparable with that of established wet-chemical methods, but the former are up to ten times faster and are therefore suitable for blending control. THE blending of organometallic additives with automotive lubricants to give a closely specified product has presented a problem in process control analysis that has persisted for many years. The required control is best effected by the determination of the metals, usually calcium, barium and zinc, that are present, sometimes singly, but more often in combination.Magnesium may also be present, and an increase in the use of this metal seems likely to take place, All of these metals can readily be determined by spectroscopic methods. However, it may be uneconomical or impractical to install and maintain the necessary equipment at small or remotely situated blending centres with a limited output in both range and amount. Blending control at such centres frequently depends upon the determination of sulphated ash, which gives an indication of the total metal content. Because such a method is time consuming and can give misleading results, there remains the need for a procedure that is rapid, selective, simple and inexpensive. The use of EDTA and similar complexing agents for this purpose has recently received considerable attention, and several titrimetric procedures have been propo~ed.l-~ Of these, the non-aqueous procedure’ is non-selective, while those involving extraction with acid2s3 create problems of emulsion formation and incomplete recovery of the metals.More recently, very simple, direct methods have been used to determine calcium and ~ i n c . ~ - ~ However, these methods are less satisfactory for samples that contain barium, and are limited to the determination of total metals and zinc if more than two metals are present. This paper describes a direct potentiometric method that is capable of determining any combination of the metals calcium, barium, zinc and magnesium in either additive concen- trates or unused lubricating oils.A wide range of oils and additives has been analysed to determine calcium, barium and zinc and a limited number of determinations of magnesium have also been carried out. The results compare well with those obtained by use of estab- lished wet-chemical procedures or atomic-absorption spectroscopy. Duplicate analyses to determine calcium, barium and zinc can be made in 21, hours, and in proportionately less time for fewer metals. Only a little additional analysis time is required if magnesium is also to be determined. OUTLINE OF METHOD- The oil or additive is dissolved in a mixed aqueous - organic solvent, buffered to pH 7 to 8 and titrated potentiometrically with EGTA [ethylene glycol bis(2-aminoethy1)tetraacetic acid; 1,2-bis(2-aminoethoxy)ethane-NNN’N’-tetraacetic acid]. This titration enables the EXPERIMENTAL * Presented a t the Third SAC Conference, Durham, July 12th to 16th.1971. Q SAC and the authors.WILSON AND MARCZEWSKI 43 total metal content to be calculated. A second inflection is observed if magnesium is present. The titration is repeated after the addition of sulphate ions to mask barium, which can be determined by difference. Zinc is determined separately at pH 4 to 496 by titration with DCTA (1,2-diaminocyclohexane-NNN~W-tetraacetic acid). Calcium is determined either directly or from the difference between appropriate titrations, depending upon the other metals present. Dialkyldithiophosphate ions interfere in the determination and, if present, are removed from the analytical solutions by an anion-exchange procedure before titration.APPARATUS- A block diagram of the apparatus is shown in Fig. 1. The solution to be titrated is contained in a 250-ml tall-form beaker and stirred on a magnetic stirrer. The titrant is added from an automatic incremental dispenser at the rate of 0.05 ml per 10 s (Fison or Hook and Tucker dispensers are suitable). The calomel reference electrode has a modified salt bridge comprising a mixture of 2-methylpropan-2-01 and water (9 + 1) saturated with potassium chloride. Two indicating electrodes are required: when zinc is to be determined with DCTA an electrode of stout silver wire (99.9 per cent. silver) is used whereas a J-type mercury-pool electrode, as described by Reilley, Schmid and Lam~on,~ is used for all other determinations.The electrodes are connected to a millivoltmeter and recorder (100 mV full scale), fitted with a source of backing- off potential and a sensitivity adjustment. The glass column used in the anion-exchange process consists of a 1 cm diameter glass tube of about 25ml capacity, fitted with a straight burette tap with a PTFE key and a 1.6 mm diameter jet. A reservoir of 250 ml capacity is fitted to the top of the tube (Fig. 2). I Stirrer I Chart recorder H Millivoltmeter Fig. 1. Block diagram of apparatus for the determination of additive metals REAGENTS- Fig. 2. Anion- exchange column All reagents should be of analytical-reagent grade unless otherwise specified. Methanol. 2-Methylpropan-2-oI, laboratory-reagent grade. Chlorobenxene, laboratory-reagent grade.Nitric acid, 4.5 per cent. V/V. Acetic acid, 10 per cent. V/V. Sulphuric acid, 0.02 N. Amberlyst A-29 anion-exchange resin, Rohm and Haas, U.S. A .-This material is obtainable in the U.K. from BDH (Chemicals) Limited. Ethanolamine, 8.8 per cent. V/V-Dilute 8.8 ml of laboratory-reagent grade ethanolamine to 100 ml with a mixture of 2-methylpropan-2-01 and chlorobenzene (1 + 4). Prepare a fresh solution weekly.4.4 [Analyst, Vol. 98 Dithizone (diphenyzthiocarbaxone) solution, 0.02 per cent. m/V-Dissolve 0.02 g of dithizone in 100ml of chlorobenzene. Store in darkness. Standard calcium solution, 0-001 M-Dry calcium carbonate (CaCO,) in an oven for 2 hours at 120 "C and cool it in a desiccator. Dissolve 0-1001 g in the minimum amount of warm 2 N nitric acid and dilute the solution to 1 litre with water. Standard zinc solution, 0.001 M-Treat granular zinc with 2 N hydrochloric acid, decant off the acid and wash the zinc thoroughly with water, methanol and finally diethyl ether.Dry the granules under vacuum, dissolve 0.1308 g in the minimum amount of warm 2 N nitric acid and dilute the solution to 2 litres with water. EGTA solution, about 0.001 M-Dissolve 0.38 g of laboratory-reagent grade EGTA (free acid) in 20 ml of warm 0.2 N sodium hydroxide solution, add 80 ml of water and dilute to 1 litre with methanol. Determine the pH with a pH 6 to 8 indicator paper (the pH should be within the range 6.5 to 7.5). If necessary, adjust the pH by the dropwise addition of 0.2 N sodium hydroxide solution. Standardise the solution against the standard calcium solution by the procedure described under Standardisation of titrants.DCTA solution, about 0.001 M-Dissolve 0.35 g of laboratory-reagent grade DCTA (free acid), dilute to 1 litre and check the pH, exactly as described for the preparation of EGTA solution. Standardise the solution against the standard zinc solution by the procedure described under Standardisation of titrants. CONDITIONS OF ANALYSIS- Complexing agents-Because of the relatively low stability constant of the EGTA - magnesium complex, it is possible to obtain the second inflection that distinguishes this metal from calcium, barium and zinc (Fig. 3, B). EGTA cannot be used to determine zinc in the presence of the other metals; for this purpose it is necessary to use DCTA, which forms an unusually stable zinc complex, and therefore enables this metal to be titrated in the presence of the others under relatively acidic conditions.WILSON AND MARCZEWSKI : DIRECT COMPLEXOMETRIC DETERMINATION Fig. 3. Typical titration curves : A, total metals; and B, calcium and magnesium, showing second inflection Electrodes-The mercury-pool electrode is used extensively to indicate metal-ion con- centration and gives reproducible end-points for all of the titrations with EGTA. It is not, however, satisfactory for the titration of zinc with DCTA, for which the recommended silver electrode gives excellent results. The salt bridge of the calomel reference electrode is modified to control the amount of water entering the titration medium.January, 19731 OF METALS I N UNUSED LUBRICATING OILS AND ADDITIVES 45 Titration medium-A solvent mixture of 20ml of chlorobenzene, 40 ml of water and 130 ml of 2-methylpropan-2-01 was adopted after extensive experimentation.It was designed to give homogeneous solutions of up to 1 g of oil, yielding sharp, reproducible end-points. PREPARATION OF ANALYTICAL SOLUTIONS- Weigh accurately into a 100-ml calibrated flask sufficient oil or additive to provide a maximum of 0-2 mmol of total metals, and ideally 0.02 to 0.07 mmol of each individual metal. Add about 50ml of chlorobenzene, swirling the flask to dissolve the sample, and make up to the mark with chlorobenzene. In the absence of dialkyldithiophosphates-For each determination required (see Table I), transfer by pipette duplicate 5-ml aliquots of the chlorobenzene solution into 250-ml tall-form beakers.To each aliquot add 15 ml of chlorobenzene, 130 ml of 2-methylpropan-2-01 and 40ml of water, mix thoroughly, and complete the analysis by the appropriate procedure described under Method I or Method 11. In the presence of dialkyldithiophosfihate-Prepare a 20-ml bed of Amberlyst A-29 anion- exchange resin in the chloride form (as supplied) as follows. Place a small pad of glass-wool in the glass column (see under Apparatus), locating it immediately above the tap. Transfer an aqueous slurry containing 9 g of resin to the column with water, opening the tap sufficiently to allow a gentle flow so that the resin settles into an evenly distributed bed free from channels and air bubbles.During this and all subsequent operations do not allow the liquid level to fall below the top of the resin bed. Depending on the nature of the eluting agent, the maximum flow-rate through an exchange column to this specification will be up to 20 ml min-l. Wash the bed successively with 100 ml of methanol and 100 ml of chlorobenzene, applying each solvent in several portions and a t the maximum flow-rate. Close the tap as the liquid level reaches the top of the bed after the final addition of chlorobenzene. Transfer to the column, by pipette, a 5-ml aliquot of the sample solution. Open the tap so that the sample enters the column over a period of 20 to 30 s and collect the effluent in a 250-ml tall-form beaker containing 90 ml of 2-methylpropan-2-01, Wash the column consecutively with 5 ml of chlorobenzene and, in 10-ml portions, 40 ml of 2-methylpropan-2-01 and 50ml of water, continuing to collect the effluent as before.Elute the column at the maximum flow-rate, but close the tap before adding each different solvent as the liquid level reaches the top of the resin bed. Mix the total effluent thoroughly and complete the analysis by the appropriate procedure described under Method I or Method I1 (Note 1). Prepare the column for re-use by repeating the washing cycle with 100 ml of methanol and 100 ml of chlorobenzene. Process a duplicate 5-ml aliquot of the sample solution and prepare similar duplicate solutions for each determination required (see Table I). METHOD I- Determination of total metals and magnesium-To the prepared solution add, by use of a pipette, 1 ml of 4.5 per cent.nitric acid solution and 1 ml of 8.8 per cent. ethanolamine solution. Stir the mixture on the magnetic stirrer. The pH should be 7 to 8. Introduce the calomel and mercury electrodes into the solution and titrate with 0.001 M EGTA solution while stirring rapidly. Record the volume of titrant added. Carry out a blank determination by using 5 ml of chlorobenzene in place of the 5-ml aliquot of sample solution, including the ion-exchange procedure when appropriate. Determination of barium in the presence of zinc and calcium-Repeat the determination of total metals as described above, with the addition of 1.0 ml of 0.02 N sulphuric acid to the sample solution following the addition of nitric acid and ethanolamine (Note 2).Again, carry out a blank determination, including 1.0 ml of 0-02 N sulphuric acid in the reagents used. NOTES- solvent environments. to be constant at 10 ml. 2-methylpropan-2-01 and 40ml of water, the column retaining the additional 10ml of water. it is necessary to record only the first end-point. 1. Unlike gel-type resins, macroreticular resins do not swell or shrink appreciably in different The solvent retention volume of the spccified resin bed can be considered The analytical solution thus contains 20 ml of chlorobenzene, 140 ml of 2. If magnesium is present a second inflection may be observed. For the barium determination46 WILSON AND MARCZEWSKI : DIRECT COMPLEXOMETRIC DETERMINATION [ A n a & f , Vol. 98 METHOD II- Determination of zinc in the +resence of calcium and barium-To the prepared solution add, by use of a pipette, 1 ml of 10 per cent.acetic acid solution and 1 ml of the 0.02 per cent. dithizone solution. Stir the solution on the magnetic stirrer. The pH should be about 4.5. Introduce the silver and calomel electrodes into the solution and titrate with O-OOl M DCTA solution while stirring rapidly. Record the volume of titrant added. Carry out a blank determination by using 5 ml of chlorobenzene in place of the aliquot of sample, including the ion-exchange procedure when appropriate. STANDARDISATION OF TITRANTS- Standardise the EGTA solution against 5-ml aliquots of the standard calcium solution by using Method I, and the DCTA solution against 5-ml aliquots of the standard zinc solution by Method 11.Dilute each aliquot with 35 ml of water, 140 ml of 2-methylpropan-2-01 and 20 ml of chlorobenzene and add the appropriate buffer solution. Then determine the reagent blanks . CALCULATION OF RESULTS- of two or more of the additive metals. Table I shows the determinations that are necessary in order to analyse any combination TABLE I DETERMINATIONS REQUIRED TO ANALYSE VARIOUS METAL COMBINATIONS Metals present Determine Calcium ; barium Calcium; zinc Total metals: zinc Barium; zinc Total metals; zinc Calcium; barium ; zinc Total metals ; barium Total metals; barium; zinc The additional presence of magnesium does not affect the required determinations. The calculation of results will vary with the metals present, as follows. Calcium and barium present- Calcium, per cent.m/m Barium, per cent. m/m Calcium and zinc present- .. .. .. - (V3 - b2) x 80.16 x MI W - [(V, - bJ - (V3 - b,)] x 274.72 x M I W .. - - Calcium, per cent. m/m Zinc, per cent. m/m Barium and zinc present- Barium, per cent. m/m (V4 - b3) x 130.76 x M , W .. .. . . - IS [(Vl - b,)+-- $(V4 - b3)] x 274.72 x M , W - - Zinc, per cent. m/m, see equation (4). Calcium, barium and zinc present- Calcium, per cent. m/m = MI M2 [(V3 - b,) -- (V4 - b 3 ) ] x 80.16 x MI W .. . . .. .. .. .. Barium, per cent. mlm., see equation (2); zinc, per cent. m/m, see equation (4).January, 19731 OF METALS IN UNUSED LUBRICATING OILS AND ADDITIVES 47 Magnesium present-The magnesium content is indicated by the second end-point in the titration of total metals, irrespective of the other metals present.(V, - VJ x 48.64 x M i W Magnesium, per cent. m/m = and V , and V , are the volumes of EGTA taken to reach the first and second end-points, respectively, in the titration of total metals; V , is the volume of EGTA used in the barium titration; V , is the volume of DCTA used in the zinc titration; b,, b, and b, are respectively the total metals, barium and zinc contents blanks; M , and M , are the molarities of the EGTA and DCTA solutions, respectively; and W is the mass of the sample in grams. RESULTS ADDITIVE CONCENTRATES- The results of the analysis of several additive concentrates containing one or two metals are shown in Table 11. All the determinations were carried out by using Method I, zinc being determined in sample numbers 8 and 9 by masking the barium with sulphate. The alternative wet-chemical procedures were I.P.110 (for barium), I.P. 111 (for calcium), polarography with the Shandon Southern Instruments A 1660 cathode-ray polarograph (for zinc) and A.S.T.M. D 811 (for magnesium). TABLE I1 DETERMINATION OF METALS IN ADDITIVE CONCENTRATES Sample No. 1 2 3 4 5 6 7 8 9 10 11 Metal sought Calcium Calcium Barium Magnesium Zinc Calcium Barium Calcium Barium Barium Zinc Barium Zinc Barium Magnesium Calcium Magnesium Metal content, per cent. mlm, found by- established EGTA titration wet-chemical (Method I) methods * r A > 11.5 11.4, 11.7 11-5 11.5, 11.6 3-46 3.43, 3.64 7-60 7-60, 7-40 7.09 7-05, 7.76 5.54 6.64, 6.63 3-78 3.76, 3.76 0.48 0.48, 0.61 6.70 6.96, 7-00 1.40 1.33, 1-38 0.04 0.30, 0.32 7-60 7-89 7.60 7-30 12.7 23-0, 13.1 13.4 13.2, 13.2 11.6 11.6 11.5 12.0 UNUSED LUBRICATING OILS- Eight samples of unused lubricating oils, providing a comprehensive range of additive formulations, were analysed to determine calcium, barium and zinc.All contained zinc dialkyldithiophosphate and were subjected to the ion-exchange procedure. For comparison purposes they were analysed by the wet-chemical methods and also by atomic-absorption spectroscopy. The results are given in Table 111. A limited statistical programme, which took into account results from several blends, indicated that for all three metals duplicate determinations should not differ from one another by more than 5 per cent. of the mean value (95 per cent. probability). The reproducibility has not been determined.DISCUSSION The determination of metals in lubricating oils by the complexometric methods described is appreciably more rapid than by the classical procedures. Atomic-absorption spectroscopy is a faster technique when a series of similar blends are to be analysed, although not when48 WILSON AND MARCZEWSKI TABLE I11 DETERMINATION OF METALS I N UNUSED LUBRICATING OILS Metal, per cent. mlm Sample No. 1 2 3 4 5 6 7 8 7 Atomic- absorp- tion spectro- scopy 0.023 0-023 0.056 0.056 0.095 0.095 0.12 0.12 0.19 0.20 0.54 0.55 1.07 1.12 2-23 2-28 7 Calcium Estab- lished wet- chemical methods 0.023 0.023 0.055 0.056 0.094 0.091 0.12 0.12 0.19 0.19 0.54 0.54 1.18 1-18 2.23 2.23 A EGTA titration 0.022 0.022 0.052 0.053 0.096 0-093 0.092 0.10 0.18 0.18 0.51 0.50 1-16 1-18 2.32 2.36 7 Atomic- absorp- tion spectro- scopy 0.12 0.13 0.041 0-041 0.092 0.093 0.97 0.95 0.18 0.18 0.36 0.36 0.18 0.19 0.62 0.63 Barium Estab- lished wet- chemical methods 0.13 0.13 0.043 0.041 0.10 0.10 1.01 1.00 0.21 0.2 1 0.39 0-38 0.18 0.19 0.63 0.66 A 1 EGTA titration 0.13 0.14 0.040 0.036 0.088 0.088 1.11 1.11 0.18 0.18 0.39 0.38 0.19 0.18 0.63 0.61 r Atomic- absorp- tion spectro- scopy 0.030 0-030 0.13 0.13 0.070 0.072 0.063 0.062 0-15 0-14 0.22 0.22 0.091 0.093 0.42 0.38 1 Zinc Estab- lished wet- chcmical methods 0.028 0.028 0.12 0.12 0.068 0.069 A- 1 DCTA titration 0.028 0.027 0.12 0.12 0.069 0.069 0.059 0.059 0.060 0.058 0.14 0.13 0-13 0.13 0.22 0.21 0.22 0-21 0.092 0.091 0.090 0,087 0.43 0.44 0.42 0.43 the numbers of samples are small.The complexometric apparatus is simpler and easier to maintain than an atomic-absorption spectrophotometer and the precisions of the three methods are comparable. Two samples gave poor results. Sample number 9 of the additive concentrates (Table 11) in which the barium to zinc ratio was 20:1, gave a very low value for zinc by Method I. A similarly low result was obtained when the zinc determination was repeated by using Method 11. Hooks and Noar5 have noted a depression of the zinc results in the presence of much larger amounts of barium, although not to the same extent. In sample number 4 of the lubricating oils (Table 111) the barium to calcium ratio was about 10 : 1. In comparison with the values obtained by the other methods used, the value for barium was high while that for calcium was low. Co-precipitation of calcium with barium in the presence of sulphate ions could, in part, explain this phenomenon. Neither of these anomalous results seriously interferes with the suitability of the method for most control purposes. Permission to publish this paper has been given by the British Petroleum Company Limited. 1. 2. 3. 4. 5 . 6. 7. REFERENCES Gerhardt, P. B., and Hartmann, E. R., Analyt. Chem., 1957, 29, 1223. Fisher, W., J . Inst. Petrol., 1962, 48, 290. A.S.T.M. Standards, Part 17, Appendix IV, Proposed Method, American Society for Testing Materials, Philadelphia, Pa., 1964. Crump, G. B., Analyst, 1969, 94, 465. Hooks, R. W., and Noar, J. W., Ibid., 1969, 94, 473. Crump, G. B., and Noar, J. W., J. Inst. Petrol., 1970, 56, 180. Reilley, C. N., Schmid, R. W., and Lamson, D. W., Analyt. Chem., 1958, 30, 953. Received July 31st, 1972 Accepted August 24th, 1972