The first step in the analysis of a solution containing derivatized isocyanates is the separation. The separation technique isolates individual compounds from a complex mixture to enable correct identification and accurate quantification of each component. A total isocyanate method will be biased low if the separation technique does not deliver all the derivatized isocyanate species to the detector(s) as identifiable and quantifiable peaks. Reversed-phase HPLC has been the dominant separation technique in isocyanate analysis.1,3,4 A reversed-phase HPLC analysis can be isocratic, i.e., having a constant mobile phase strength, or gradient, i.e., having the mobile phase strength increasing during the course of the analysis. Isocyanate samples frequently contain compounds of greatly varied molecular weight, which translates to greatly varied retention in an isocratic analysis. Such an analysis is necessarily long, with late eluting peaks that are broad and difficult to detect and quantify accurately. Gradient elution is frequently used when compounds of widely differing retention need to be determined in the same analysis. Weakly retained compounds are eluted early in the chromatogram with a relatively weak mobile phase. Then the mobile phase is strengthened to accelerate the elution of more highly retained compounds. Not only do these compounds elute faster than in an isocratic analysis, resulting in shorter analysis times, but also the peaks are taller and narrower, improving the LOD and facilitating peak integration. Several total isocyanate methods have been evaluated for recovery of derivatized oligomeric isocyanate species. An evaluation of MDHS 25 found an average recovery of 105% for three isocyanate prepolymers.83 In contrast, two separate investigations of prepolymers using NIOSH 5521 (which is very similar to MDHS 25) found 58% and 60% recoveries, respectively.84,85 An evaluation of NIOSH 5522 found average recoveries for five prepolymer products of 62%.85 MDHS 25 and NIOSH 5521 are isocratic methods, whereas NIOSH 5522 increases the strength of the mobile phase slightly during the analysis. These evaluations did not distinguish between actual physical losses of isocyanate and apparent losses resulting from prepolymers having lower detector response factors than the monomers used for calibration. However, the similarities in recoveries, comparing NIOSH 5521 and NIOSH 5522, suggest that physical losses are likely. If this is the case, a method using a stronger gradient may improve recoveries. This contrasts with the experience of the developers of MDHS 25/2 who have found no additional isocyanate-derived peaks even after using a gradient ending and holding at 100% acetonitrile.86 Gradient elution most frequently involves increasing the percent of organic modifier in the mobile phase. There are several disadvantages to this approach relative to isocratic analyses. It requires preparation of at least two mobile phases. Although the analytes elute faster, there may be considerable time required between analyses to allow for the system to reequilibrate to the initial conditions.87 The baseline of the chromatogram is likely to change during the gradient, making it more difficult to integrate small peaks. Finally, artifact peaks can elute as the mobile phase is strengthened.88 There are also some disadvantages of gradient elution of particular importance in isocyanate analysis. Several methods for total isocyanate (MDHS 25,78 NIOSH 5522,79 and the Ontario Ministry of Labour tryptamine method82) use electrochemical (EC) detectors. Unfortunately, EC detectors are somewhat incompatible with gradient elution because they are especially sensitive to changes in mobile phase. Also, HPLC-based total isocyanate methods quantify isocyanate species for which analytical standards are not available by assuming the detector response per isocyanate group is the same as that of a derivatized monomer standard. Even if the responses are the same in the same mobile phase, the response of a late-eluting isocyanate species may be quite different than that of the derivatized monomer if the two are eluting in substantially different mobile phases. Therefore, there is a potential for quantification errors.
1/15/98
126
NIOSH Manual of Analytical Methods
