Annals of Occupational Hygiene Advance Access originally published online on September 27, 2007
Annals of Occupational Hygiene 2007 51(7):645-646; doi:10.1093/annhyg/mem040
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
LETTER TO THE EDITOR |
Isocyanate Sampling and Analysis
I read with considerable interest Creely et al. (2006) on assessing isocyanate exposure. The reason I was so interested is that the title and synopsis led me to believe that there might be some answers for the confusing status of isocyanates monitoring in the workplace and the difficulties of interpretation of the approved air monitoring method MDHS 25/3 (HSE, 1999). Isocyanates feature quite a lot in occupational hygiene practice, and in particular, in my work as an expert witness involved in the investigation of alleged occupational asthma cases, the contents of such technical papers could prove to be very important in litigation. Unfortunately, I find that the paper only serves to increase my confusion and does not assist me in forming expert opinions, which a court may need to consider liabilities in litigated occupational asthma cases.The reasons I find the paper confusing may, I trust, be summarized in the following notes.
1. MDHS 25 aims to determine all isocyanates as the –NCO group by the determination of the concentration of a derivative of that group with a piperazine complex. The derivative has a particular absorption on UV which enables detection and quantitation (quantification). As I understand the method, ‘suspect’ peaks on UV detection are simultaneously analysed by EC detector, and if the ratio of detection falls within a certain defined range, then that ‘peak’ is considered to be an isocyanate derivative rather than an interfering substance. According to White (2006) the detector has to be calibrated, and I understand this is done on the parent isocyanate. For example, in the case of car refinish paints this would be HDI monomer (hexamethylene di-isocyanate) and in the case of flexible foams this might be methylene bisphenyl diisocyanate or toluene diisocyanate (various isomers), or possibly the original bulk material.
We all know that the amount of free monomer in commercial systems is negligible, usually quoted by the chemical suppliers as <1% of the free monomer. In effect, what we are assuming is that the peaks on analysis by the method (which probably do not have the same spectral qualities as the monomer derivative, certainly on UV detection) can be not only classed as active isocyanates but also quantified in the same way as the monomer. It is difficult to see how this can be the case for a complex pre-polymer system where even the manufacturers do not know what the molecular distribution is—and in any event this is changing all the time in a reacting system. My problem is that I cannot see how any of this method, or the current paper, could be used in a properly argued case (where the results of air sampling are a critical issue in determining negligence and liability of an employer) without a clever and knowledgeable defence lawyer making mincemeat out of any technical allegations of negligence.
2. At p. 611, Creely et al. (2006) state ‘The detection limit for this method was established at 0.02 micrograms total –NCO in each filter sample.... The limit of quantitation (LOQ) for routine hygiene samples of this type was 0.001 mg/m3'. I would value their comments on the following argument.
If we take a ‘routine hygiene sample’ as being a filter only (for the good reasons stated in the paper that personal samplers using toluene in a bubbler is not safe) run at say 2 l min–1 for 4 h, giving a total of 480 l of sampled air (say 500 l or half a cubic metre to make the arithmetic easier), then according to the quoted laboratory filter detection limit this would mean a lower reportable concentration in air of 0.02/0.5 or 0.04 µg m–3 (0.00004 mg m–3), some 40 times lower than the limit quoted (and upon which the statistical analysis in the table is based, see following sections). If, however, short-period sampling was conducted (15 min at 1 l min–1), then the LOQ would have been around the limit quoted; there does not seem to be any indication from the Health and Safety Executive field samples used in this paper as to the air volumes taken.
3. Creely et al. (2006) state ‘Samples below the LOQ were assigned a value of 
LOQ (0.0005 mg/m3)'. A table showing some statistical analysis using this figure is shown (Table 2, top of p. 615). I can see no justification for assuming such a figure; in fact, using this in statistical analysis (where the vast majority of the samples were ‘non-detected ie half the LOQ’) can lead to completely erroneous conclusions. The only thing you can say with any scientific justification about non-detected results is exactly that: there was nothing detected, certainly not that an arbitrary half-figure is taken for subsequent statistical interpretation.
You could go on to say this does not mean the substance was not there, merely that if it was there, it could not be shown and quantified by the sampling method. It then becomes a matter of ‘faith’ if you come to the view that isocyanates were there at all. There is an exactly analogous argument which has run in the courts ad nauseam regarding asbestos fibres in air (in asbestos disease cases), where if an advocate asks the question as to whether there was any asbestos fibre present in air at all (and where there is no sampling evidence to show that there was), the honest, scientific and correct answer is to say ‘I don't know'. You may express a view (an opinion based on faith) that there were such fibres, but equally, you could not ever demonstrate by testing that such fibres existed. It is an untestable hypothesis which is always a dangerous ground for someone working in the scientific discipline.
4. MDHS 25/3 and learned papers discussing this method recommend where a mixture of aerosol isocyanates (i.e. pre-polymers and oligomers) and vapours (i.e. invariably monomers such as HDI and TDI, but probably not MDI as it has no significant vapour pressure) are sampled, a sample train consisting of a reagent bubbler and conditioned filter be used. The former is said to sample the particulate and the latter the vapours (see p. 611 of Creely et al.). In this particular exercise, some or all of the personal samplers were filter-only ‘this sampling train was not always used in our study owing to the risk of spillage of solvent onto the wearer'. It is not clear from the tables of results which were, or were not, using such sampling trains. On this basis there is no consistency between results and again no justification for the statistical analysis presented. It is possible that these inconsistencies in sampling may go some way to explaining the odd correlations between biological monitoring and the air sampling, perhaps because of dermal exposure to particulate matter not detected by the air sampling.
On a general point about sampling trains (i.e. bubbler, filter and pump), several questions arise. Has there been any attempt to calibrate the method on standard atmospheres? How are standards for monomers such as MDI (which has a vanishingly low vapour pressure at room temperature) produced and themselves quantified? Most puzzling of all, how are standard atmospheres of polymeric isocyanates produced? Can reacting species (e.g. in a car paint spray system), where the active isocyanate concentration reduces rapidly with time, ever be quantified as a standard atmosphere? I hope the answers to this will not lead me to the view that the revised method is based on faith rather than good science.
Occupational Hygiene Advisory Service Ltd, Kenilworth, Warwickshire, UK
E-mail: mike{at}ohas.demon.co.uk
Received March 2, 2007; in final form August 3, 2007
REFERENCES
Creely KS, Hughson GW, Cocker J, et al. Assessing isocyanate exposures in polyurethane industry sectors using biological and air monitoring methods. Ann Occup Hyg (2006) 50:609–21.
HSE. MDHS 25/3. Organic isocyanates in air (1999) Sudbury, UK: Health and Safety Executive Books.
White J. MDHS 25 revisited: development of MDHS 25/3, the determination of organic isocyanates in air. Ann Occup Hyg (2006) 50:15–27.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||