17. Analysis: the method decides the finding

The determination of chromium(VI) is not an elemental determination but speciation analysis. What is measured is an oxidation state, and that state can change during sample preparation. What the instrument finally registers is not necessarily what was present in the sample.

The American procedure EPA SW-846 Method 3060A, on which alkaline digestion is based, names both directions of error explicitly. On the one hand the extraction conditions must “not induce reduction of native Cr(VI) to Cr(III)”, on the other the method must “not cause oxidation of native Cr(III) contained in the sample to Cr(VI)”. On the second direction it adds: “Cr(VI) results obtained using this method may be biased high due to method-induced oxidation.” [83]

Reduction during digestion

If the sample contains reducing constituents, the chromate is converted before it can be measured. In coatings these are above all metallic zinc and aluminium, together with iron(II) and organic constituents. The round robin test on the Dutch NEN 5617 shows the order of magnitude: for a zinc-free paint sample the participating laboratories achieved an average recovery rate of 58 per cent, and 39 per cent at 50,000 mg/kg of zinc. In a matrix with 16 per cent zinc the recovery rate under NEN 5617 was 0.2 per cent, under a further accredited method 6.5 per cent and under the method of SEEF B.V. 35.3 per cent. For this reason NEN 5617:2022 classifies itself as semi-quantitative. [84, 85]

The insulation material matrix

For thermal insulation materials the situation is sharper, because the matrix is itself the reaction partner: rich in calcium, basic and, once in service, permeated with soot, oil and corrosion products. No standardised method exists for this matrix; there is neither a NEN nor an ISO nor a NIOSH method tailored to insulation materials. ISO 15192:2021 is intended for waste and soil. According to the laboratory concerned with it, transferred procedures give results that are too low by 20 to 30 per cent in mild cases and values close to zero in strongly calcium-containing samples. One Dutch laboratory, SEEF B.V., has developed its own method for this matrix which stabilises chromium(VI) throughout the analytical procedure and, according to its own data, achieves recovery rates structurally above 80 per cent on spiked insulation material samples. [85]

Oil films and apolar phases

Besides reduction there is a second route by which a sample reads too low. Where a surface is treated with a penetrating oil before sampling, the chromate passes into an apolar phase and is no longer available to the aqueous measurement. Chemically nothing happens; the analyte is merely withdrawn from the determination.

SEEF B.V. has quantified this for WD-40. Solid chromium(VI) in direct contact with 2 ml of WD-40, then dissolved in water without digestion chemistry, gave a recovery rate of 5.8 per cent. At higher volumes no reliable concentration could be determined at all, because the chromate clumped together in the oil phase. With chromium(VI) already dissolved, and with full digestion, the effect did not occur (section 26). [91]

For practice it follows that sampling after an oil or grease treatment gives no usable result. This applies to penetrating oils as well as to assembly pastes and preservatives.

What follows for the risk assessment

An analytical result for chromium(VI) without a statement of the recovery rate in the same matrix says nothing about the content. The EPA method draws the consequence and requires spiked samples before digestion, soluble and insoluble, with an acceptance range for the recovery rate of 75 to 125 per cent; outside that range the entire batch must be repeated. [83]

Where this statement is missing, two false conclusions are possible. A result that is too low leads to an existing contamination being classified as insignificant and protective measures being omitted. A methodically elevated result leads to measures being applied in the wrong place. For occupational safety the first case weighs more heavily, because it can lead to a release that is not based on any actual freedom from chromate.

This is not an assessment of individual laboratories. The procedures named are suitable for the matrices for which they were developed and are carried out properly. The point is a different one: chromium(VI) is matrix-dependent, and a procedure developed for soil or waste does not give a comparable result on a used, calcium-rich insulation material.

From this follows the requirement for every test report: method, matrix and recovery rate in that matrix belong in the report, not only the numerical value. And from this follows at the same time why the German TRGS 561 does not rely on a measurement but on the presumption of presence: as long as the recovery rate in this matrix is not demonstrated, a negative result is no proof of absence.