by Prof. Dr. W. Maurer, Analytical Chemistry II, University of Siegen (Germany). Authors: B.A. Schubert, E. Hohaus, H.S. Dengel, W. Riepe, W. Maurer.

Cooling lubricants are indispensable in the metal-working industry and many other industrial processes. The typical cooling lubricant contains inorganic and organic substances from over thirty classes of compounds, but for proprietary reasons the exact makeup is rarely specified.¹ Occupational safety issues, stability maintenance, and quality control for the manufacturing and use of cooling lubricants demand a quick and efficient analytical monitoring technique.

A research team in Germany recently developed different methods for the analytical evaluation of cooling lubricants, including cooling-lubricant concentrates, mineral-oil-free, and mineral-oil-containing agents.^2-4 Their methodology employs capillary zone electrophoresis (CZE), i.e., for the simultaneous determination of 11 inorganic anions and seven carboxylic acids (Figure 1).

Tracking the Formation of Nitrosamines
Among cooling agents, the water-miscible cooling lubricants have been gaining increasing preference over the customary cutting oils. Their high water content offers more efficient heat removal for the higher cutting speeds of late-model machines. One important aspect of monitoring cooling lubricants is the formation of nitrosamines from secondary amines and nitrosating agents. Some literature suggests limits of 20 mg/L for nitrite, and 50 mg/L for a reliable quantitation of nitrate in aqueous cooling-lubricant emulsions.⁵

Short Analysis Times, Little Waste
Selective separation with CZE employs a special buffer system using PMA, NaOH, TEA, DMOH at pH 7.90. Virtually no sample preparation is needed, only appropriate dilution. The analysis times are short, and only very small quantities of wastes are produced to require disposal. The process offers manufacturers of cooling lubricants a relatively simple and reproducible quality-assurance tool for their products. In operational use, it permits easy monitoring of possible changes in coolant-circulating systems.

Substance	Theoretical Plate Number N	Resolution R
Thiosulphate	140000
Bromide	133600	4.36
Chloride	84000	2.93
Sulphate	151000	4.22
Nitrite	147000	2.67
Nitrate	152000	3.70
Oxalate	202000	2.78
Perchlorate	161000	8.94
Chlorate	134000	7.59
Citrate	114000	4.29
Malonate	128000	1.68
Phosphonate	83000	10.97
Tartrate	117000	1.86
Fluoride	47000	1.32
Formate	93000	2.50
Succinate	72000	0.83
Phosphate	48000	8.85
Adipinate	50000	9.95

Table 1. Theoretical plate numbers and resolutions of the capillary zone electrophoresis separation of the anions from Figure 1.

The monitoring capabilities are important for optimizing the service lives of cooling-lubricant systems. They possibly allow moving from empirical clues to scientifically predictable behavior patterns and properties of the lubricants and provide accurate inputs relating to industrial hygiene challenges.

Working Range: 1-20 mg/L
As Figure 1 shows, the migration times were less than 6 minutes. The theoretical plate numbers N in the electropherogram (Table 1) vary in the order of magnitude between about 47,000 and 202,000 -- customary for capillary zone electrophoresis. Under the electrophoretic conditions chosen, the working range stretched between 1 and 20 mg/L.

Cooling Lubricant No.	Substances Found	Concentration (mg/kg)
1	Chloride Sulphate Phosphate	121.8 45.1 38.1
2	Phosphate	341.8
3	Chloride Sulphate Phosphate	18.7 107.2 626.5
4	Chloride	114.7
5	Chloride Sulphate	55.3 213.9
6	Chloride Sulphate	663.2 441.6
7	Sulphate	632.2
8	Chloride Sulphate	387.3 14.2
9	Citrate Phosphonate Phosphate	8656.2 27.5 170.4
10	Chloride Sulphate	58.4 359.8
11	Chloride Sulphate	191.5 62.7
12	Thiosulphate	542.1
13	Thiosulphate	4078.3
14	Phosphate	460.1
15	Chloride Sulphate Nitrite Nitrate Oxalate Phosphate	164.9 127.8 11.9 70.7 5.0 19.8
16-24	Not detectable within the working range

Table 2. Experimental results for 23 cooling lubricants (1-14 and 16-24) and one cooling lubricant emulsion (15).

The results for the minimum detectable levels and the determination limits are within the lower mg/kg range for undiluted cooling lubricants. Variation coefficients for standard solutions in the working range from 1 to 20 mg/L are normally below 10% (statistical certainty P = 95%).

Experimental Conditions
CE Apparatus:	HP 3DCapillary Electrophoresis system with diode array detector
Software:	HP CE ChemStation Rev. A, Version 03.03
Capillary Tube:	Fused-silica capillary tube, 50 µm. i.d., 48.5 cm length, 40 cm to the detector; bubble factor, 3
Conditioning:	10 min with 1M NaOH, 10 min with 0.1 M NaOH, 5 min with buffer
Detection:	Indirect UV, wavelength 245 nm
Pressure Injection:	150 mbar/sec
Temperature:	25°C using capillary thermostat (air cooling)
Rinsing:	Capillary is rinsed 3 min with 0.1 NaOH, then 3 min with buffer before every analysis
Samples:	Standards and cooling lubricant samples dissolved in double- distilled water
Water:	Double-distillation unit Bi 18 T (Heraeus Instruments)
Buffer:	2.5 mM pyromellitic acid, 1.6 mM triethanolamine, 6.5 mM NaOH, 0.75 mM decamethonium hydroxide (DMOH), pH = 7.8-8.0
pH Meter:	pH 91 (WTW)
DMOH:	From decamenthonium bromide (DMBr) via 3 g anion exchanger
Preparation:	Dowex 1-4 (Conditioning: 10 mL 1M NaOH, then 10 mL H2O)

Recoveries Versus Dilution
The recoveries were determined for several cooling lubricant concentrates in different dilutions. Each concentrate was diluted appropriately in independent double batches and the measurements repeated three times per batch. The recoveries depend on the particular sample matrix and the dilution selected (Figure 2). With dilutions of about 1:10, they average between 50 and 60%; with 1:20, between 70 and 90%. Recoveries of 90 to 110% are achieved only with dilutions of 1:50, and with some anions at 1:100. Measurements for the nitrite determination must be performed within a few hours after sample preparation.

What 24 Lubricants Contained
A total of 24 samples of cooling lubricant concentrates and one cooling lubricant emulsion from the circuit of an induction welding plant were analyzed quantitatively for the compounds listed in Table 2. Some were present in 14 cooling lubricant concentrates, mostly at very low chloride, sulphate, and phosphate concentrations. Concentrates 12 and 13 contained thiosulphate. The manufacturer confirmed the addition of polysulphide to these products.

Phosphonic acid in sample 9 may have been added for binding the calcium in conditioning water to prevent the formation of unsoluble calcium soaps; the manufacturer confirmed the addition of less than 1% phosphonic acid. The sample also contained citrate, which is used as a complexing agent, added by the maker at about 1%. None of the cooling lubricants contained nitrite.

Supplements to Feature Article

• Performing the Measurements
  
• The Importance of Cooling Lubricant

References

1. Cooling Lubricants, Maximum Allowable Concentration Values (MAK Values); Toxicological Industrial Bases. Vol. 5, 20th Edition, VCH, Weinheim (1994).

2. Schubert, B.A., Dengel, H.S., Hohaus, E., Riepe, W., Maurer, W. GIT Fachz. Lab. 41, 742-747 (1997).

3. Schubert, B.A., Hohaus, E., Dengel, H.S., Riepe, W., Maurer, W. Gefahrstoffe-Reinhalt. Luft 56, 393-399 (1996).

4. Schubert, B.A., Hohaus, E., Dengel, H.S., Riepe. W., Maurer, W. Gefahrstoffe-Reinhalt. Luft 57, 511-515 (1997)

5. Technical Regulations for Hazardous Materials 611 (TRGS 611). Restrictions on Applications for Water- Miscible and Water-Mixed Coolants in the Use of Which n-Nitrosamines Can Arise. Carl Heymanns Verlag, Cologne (Germany), March 1993.

6. EN 26777 German Standard Methods for Water, Effluent and Sludge Investigation. Vol. II, D10, 29th Edition, VCH, Weinheim (1993).

7. Francois, C., Morin, Ph., Dreux, M. J. High Resolut. Chromatogr. 19, 5-16 (1996).

8. Jones, W. R., Jandik, P. J. Chromatogr. 546, 445-458 (1991).

9. Harrold, M. P., Wojtusik, M. J., Riviello, J., Henson, P. J. Chromatogr. 640, 463-471 (1993).

10. Cunat-Walter, M. A., Shoikhet, K., Engelhardt, H. GIT Fachz. Lab. 39, 914-921 (1995).

This article has been abstracted from extensive analytical work.² The authors appreciate, as do we, the support of the Hauptverband der gewerblichen Berufsgenossenschaften and the Süddeutsche Metallberufsgenossenschaft (Germany), as well as cooling-lubricant manufacturers and metal- processing companies.

Please direct questions relating to HP instrumentation to: Andrea Kohn, Hewlett-Packard GmbH, Hewlett-Packard-Strasse 8, 76337 Waldbronn, Germany; Fax: (49) 7243-602-666.