ISOCYANAT ES, TOTA L (MAP): METHO D 5525, Issue 1, dated 15 March 2003 - Page 9 of 17 shortened in the strong mobile phase relative to the weak m obile phase. A model monofunctional isocyanate (octyl isocyanate) eluted 14 times faster and a model diisocyanate (MDI) eluted several hundred times faster using the strong mobile phase. Re-equilibration time is only a few minutes when switching from the strong m obile phas e ba ck to the weak m obile phas e. Detector baselines are very stable during the gradient. The fluorescence baseline is unchanged and the UV253 baseline is disturbed less than 1 mV . The UV response of MAP derivatives has been found to be independent of the pH gradient. However, the fluorescence response is highly dependent on the pH of the m obile phase. This is corrected by post-column addition of a cid to the m obile phas e prior to detection . Co m pou nd-to -Co m pou nd U V R esp ons e Va riability: Evaluation of UV response for several MAP-derivatized isocyanates, including monofunctional isocyanates and diisocyanates [5,7] showed a compound-to-compound RSD of 3.0-3.5% at the maxim um 254-256 nm and an RSD of 1.0-2.2% at the less intense but more selective wavelengths of 352 nm , 368 nm , and 388 nm . In other studies, the UV response of MDI triisocyanate oligomer varied from 6% higher than the mean of several m onom ers [7] to ab out 10% lower than the M DI m ono m er [8]. Th e o verwhe lm ing m ajority of the da ta sug ges ts very little variability in UV response among MAP-derivatized isocyanate compounds, supporting the validity of quantifying non-monom eric isocyanate species by using the UV area of the monom er standard. Reco very of MAP-D erivatized Mon om ers: Quartz-fiber and glass fiber filters were spiked with HDI-MAP, 2,4-TDI-MAP, and MDI-MAP at 630 ng NCO group per compon ent per filter along with 500 :g MAP. Recovery ranged from 97-98% for the quartz-fiber filters and 97-9 9% for the glass -fiber filters [8]. In another study, MAP-coated quartz-fiber filters were spiked with a mixture of HDI, 2,4-TDI, and MDI at nominal levels of 21, 63, 210, 630, and 2100 ng NCO group per component. Re co veries averaged 93% for HDI, 91% for 2,4-TDI, and 92% for MDI with no correlation between recovery and spiking level over the range studied. As part of an International Organization for Standardization (ISO) intercomparison stu dy, tw o com m ercial pro ducts conta ining only m onom eric isocyanates were studied [8]. Levels spiked on filters were 2100, 1050, 525, and 105 ng N CO group per filter, corresponding to 2X , 1X, 0.5X, and 0.1X the U.K STEL level. For Desm odur H (HDI monom er), the mean recoveries were 104% at 2X, 103% at 1X, 102% at 0.5X, and 99% at 0.1X. For D esm odur T -80 (a m ixture of 2,6-TDI and 2,4-TDI), the mean recoveries were 101% at 2X, 104% at 1X, 103% at 0.5X, and 108% at 0.1X. Be llo et al. found HDI monom er recovery averaging 108% in the range of 0.45 to 42 ng NCO per filter and, in a se para te study, a m ean reco very of 100 % in the ra nge of 8 to 250 ng N CO per filter [7]. Recovery of Total Isocyanate Group: Losses of isocyanates can occur at several stages in the sampling and analytical pro cess [1,2 ]. Be llo et al. evaluated total reactive isocyanate group (TRIG) recoveries from spiked filters and impingers as well as recoveries from HPLC analysis [7]. Desmodur N3300 (HDI isocyanurate) was derivatized with MAP and spiked onto quartz-fiber filters. Ex tractio n and analysis fo und re covery ranging from 91-9 3% com pare d to directly analyzed product for spiking levels ranging from 90 ng to 2.34 :g NCO per filter. Sim ilarly, a bulk product containing HDI isocyanurate and IPDI oligomers was derivatized with MAP, spiked into MAPcontaining impingers, processed by solid-phase extraction, and analyzed by HPLC . The recove ry of the TR IG varied from 92-106% for the HDI isocyanurate in the range 18-2 500 ng NCO/mL and 76-89% for the IPDI oligomer in the range of 16-1900 ng NCO/m L. Recoveries from HPLC analysis were determined by comparison with isocyanate conte nt obta ined by bulk p roduct titration . The pro ducts investiga ted and their respective recoveries were Bayer N100 (HDI biuret) 75%, Bayer N3300 (HDI isocyanurate) 94%, Bayer N3400 (HDI isocyanurate + uretidinedione) 95%, Bayer Z4470 (IPDI oligomers) 100%, and B26-2 (HDI isocyanurate + IPDI oligomers) 96%. These HPLC recoveries were conducted at high analyte levels. It is recognized that quantitative recoveries of complex products are not possible at low levels because many components fall below the lim it of detection [7]. As part of an International Organization for Standardization (ISO) intercomparison study, two comm ercial products containing oligomeric isocyanates we re stu died [8]. Le vels spiked on filters were 2100, 1050, 525, and 105 ng NCO group per filter, corresponding to 2X, 1X, 0.5X, and 0.1X the U.K. STEL level. For Desmodur N3300 (HDI isocyanurate), the mean recoveries were 52% at 2X, 43% at 1X, 35% at 0.5X, and 28% at 0.1X. These results are very different than those obtained by Bello et al. [7] and require re-exam ination. For Sup rasec 50 30 (polymeric M DI), the mean recoveries were 65% at 2X, 62% a t 1 X, 5 9% at 0.5 X, and 54% at 0.1X . Stu dies are underway to account for m issing isocyanate in oligomeric products. NIOSH Manual of Analytical Methods (NMAM), Fourth Edition
