ANALYST, JANUARY 1991, VOL. 116 21 Application of Tryptamine as a Derivatizing Agent for Airborne Isocyanate Determination Part 4." Evaluation of Major High-performance Liquid Chromatographic Methods Regarding Airborne Isocyanate Determination With Specific Investigation of the Competitive Rate of Derivatization Weh S. Wu, Robert E. Stoyanofft and Virindar S. Gaind Occupational Health Laboratory, Ontario Ministry of Labour, I0 1 Resources Road, Weston, Ontario M9P 3T1, Canada Several amines were investigated to determine their competitive rates in the derivatization of isocyanates. The a mines studied included N-(p-nitro benzy1)-N-propylam i ne, 1 -(2-pyridyl)piperazi ne, 1 -( 2-met hoxy- phenyl)piperazine, N-a-methyl tryptamine and tryptamine. Phenyl isocyanate, which might find application in future studies on sampling isocyanate vapour on solid absorbents, because it possesses a much higher vapour pressure than any of the industrially used isocyanates, was employed as the reference isocyanate.This approach was adopted because only the relative, rather than actual, rates of derivatization were of interest. By comparing the significant features of the methods for the determination of isocyanates using high-performance liquid chromatographic techniques, it was concluded that the proposed method, which uses tryptamine (and possibly N-co-tryptamine), was the most promising for practical application. The theoretical background of the proposed method was based on the isolation of a selected n-system in a derivative for specific detection. This approach should be applicable to other areas involving analysis with chromatographic techniques.The competitive rate study has also provided a better assessment in selecting a particular amine for further research in personal sampling of isocyanates on amine-coated solid absorbents. Keywords: Isocyanate derivatization; competitive derivatization rate; isolation of selected n-system High-performance liquid chromatography (HPLC) tech- niques for determining airborne isocyanates have been routinely practiced in occupational health laboratories. The reason for this is that high-performance liquid chromato- graphy is capable of handling larger molecules or compounds that are unstable to heat such as the industrially used isocyanates or their common derivatives.This has become even more important in recent years because the monitoring of airborne polymeric isocyanates in workplaces has been in demand. In spite of the many published isocyanate monitoring methods using HPLC techniques, the representative methods that have involved major developments are few and are discussed below in order to draw some distinct comparisons to the attention of the reader. Dunlap er al.1 developed the earliest significant method using N-(p-nitrobenzy1)-N-propylamine (NNNP) to derivatize isocyanates. Quantification of the derivatives is carried out by reversed-phase high-performance liquid chromatography with ultraviolet (UV) detection. This method has been applied almost exclusively to the determination of monomeric isocy- anates.The NNNP reagent, however, can only be purchased in the form of a salt owing to the instability of the amine itself. An additional disadvantage of this method is that it lacks the sensitivity of detection often required for the determination of extremely low levels of airborne isocyanates. Despite the drawbacks of this method, viewed by present day standards, many laboratories are still accustomed to applying it as the ultimate regulatory method for monitoring isocyanates in the workplace. Subsequently, several HPLC methods have been published, with the emphasis on improving the sensitivity of detection, * For Part 3 of this series see reference 6. t Present address: Concord Scientific Corporation, 2 Tippett Road, Toronto, Ontario M3H 2V2, Canada. but not for determining polymeric isocyanates. The most important methods are those of Levin er a1.,2 which uses the fluorescent reagent 1-naphthalenemethylamine to derivatize isocyanates, and Goldberg et al.,3 which uses 1-(2-pyridyl)- piperazine (PP) as the derivatizing agent to enhance the UV chromophore for detection. Interestingly, the latter workers have also used PP to derivatize a number of polymeric isocyanates.4 However, procedures for quantifying the deri- vatized polymeric components have not been addressed. In order to satisfy the increasing demand for monitoring both monomeric and polymeric isocyanates, Bagon et al.5 have devised a dual-detection HPLC method using 1-(2- methoxypheny1)piperazine (MPP) as the derivatizing agent. Polymeric components of a particular isocyanate are ident- ified by the ratio of the responses from both the amperometric oxidation and UV detectors. Each of these components is quantified to the corresponding amount of -NCO by calibra- tion with the parent monomer.However, this method has some major drawbacks; hence the reliability of the analysis is often open to question. Establishing a reliable response ratio from amperometric oxidation and UV detection is, in general, not an easy task. The reason for this is that the former is regarded as a detector with 'low stability' and the latter lacks both sensitivity and selectivity. This method, regardless of its drawbacks, has, nevertheless, provided an unprecedented approach to the determination of various polymeric iso- cyanates together with their monomers.In fact, in the UK, this method has been designated as the official regulatory method for monitoring total isocyanates exposure in work- places apparently because no other methods are available. Recently, Wu et a1.6 developed a method for determining total isocyanates with considerable reliability. This method uses tryptamine to react with all the isocyanates and the derivatives are quantified by a dual-detection system consist- ing of a fluorescence and an amperometric oxidation detector. As both detectors are very sensitive and selective, very low22 ANALYST, JANUARY 1991, VOL. 116 detection levels can be attained with minimum detection interferences. The uniqueness of this method is that all the tryptamine-derivatized isocyanates can be quantified by cali- bration against a single, pure standard such as tryptamine- derivatized toluene diisocyanate.The amount of isocyanato groups (-NCO) for individual components of the sample can be quantified without necessarily identifying the appropriate type of isocyanate. The theoretical basis of this method is that the n-orbitals of the indolyl moiety of tryptamine, responsible for the fluorescence and amperometric oxidation activities, are unperturbed before and after derivatization.7.8 In the present study this approach has been generalized to analysis with chromatographic techniques as the isolation of a selected n-system in a derivative for specific detection. The evaluation of analytical methods for determining isocyanates would be more conclusive if the derivatization reaction rates could be compared.When collecting airborne isocyanates in solutions of various derivatizing agents, the derivatization reaction rates would affect the over-all ef- ficiency of the methods. This is particularly important for sampling airborne isocyanates using a solid absorbent coated with a derivatizing agent, which would become indispensable for personal sampling in the workplace. A relatively slow reaction involving a reagent coated on a solid absorbent would be even slower than the same reaction occurring in solution. Unfortunately, information regarding competitive derivatiza- tion rates has not appeared in any of the published methods, which clearly indicates that some research in this area is necessary. Derivatization of an isocyanate (RNCO) with an amine (Am) is a second-order chemical reaction and the reaction rate can be written as follows: -d[RNCO]ldt = k[KNCO] [Am] (1) where k is the rate constant.Assuming the initial concentra- tions of amine and isocyanate are a and i, respectively, and the concentration of the derivative formed after a given time, t , is x, the reaction rate can be expressed as duldt = k(a - ~ ) ( i - X) ( 2 ) However, if a second amine is also involved in the derivatiza- tion to compete with the first amine, the rate kinetics for both reactions are given by equations ( 3 ) and (4), respectively duldt = k(a - X) (i - x - y ) dyldt = k’(b - y ) (i - x - y ) (3) (4) where b is the initial concentration of the second amine, y the concentration of the corresponding amine-derivatized iso- cyanate and k’ the corresponding rate constant. As both reactions proceed concurrently at different rates, the combined rate kinetics can be expressed by equation (5) with the time variable, t , being considered as a constant (5) By appropriately integrating equation ( 5 ) as shown in equa- tion (6), the relative rate (klk’) for the competitive derivatiza- tion reaction9910 is finally obtained [equation (7)] klk’ = log[(a - x ) / ~ ] / l o g [ ( b - y)lb] (7) Experimentally, it is unnecessary to conduct the rate study on industrially used diisocyanates because multiple derivatives would be produced due to random attack on the isocyanato groups by the amines thus complicating the investigation.As only the relative derivatization rate with the isocyanato functional group was of interest, a relatively pure and stable mono-isocyanate, i.e . , phenyl isocyanate, was preferred for this work. Experimental and Results Chemicals and Apparatus Tryptamine and N-a-methyl tryptamine (NMTP) were pur- chased from Sigma (St. Louis, MO, USA). Phenyl isocyanate, 1-(2-pyridyl)piperazine (PP) and 1-(2-methoxyphenyl)piper- azine (MPP) were from Aldrich (Milwaukee, WI, USA). N-4-Nitrobenzyl-N-propylamine (NNNP) was a Regis prod- uct, obtained through Caledon Laboratories (Georgetown, Ontario, Canada) as the hydrochloride salt. Amine solutions of NNNP were freshly prepared immediately before use. Final dilutions of these solutions were made in acetonitrile. All solvents were of glass-distilled quality and were obtained from Caledon Laboratories.Water was doubly distilled after treatment with KMn04. The HPLC system consisted of a Beckman 112 solvent delivery module, a Scientific System LP-21 pulse damper and a Shoeffel 970 fluorescence detector. The excitation wavelength was set at 275 nm and the emission wavelength was filtered at 320 nm. A 5 pm Hypersil-ODS column (25 cm x 4.6 mm i.d.) from Chromatography Science was used. The mobile phase was acetonitrile-water (50 + 50). The flow-rate was set at 0.8 ml min-1. The infrared (IR) spectrometer was a Beckman Model 4240 instrument and the spectrum of tryptamine-derivatized phenyl isocyanate (KBr disc) was obtained at a scan rate of 600 cm-1 min-1. Preparation of the Tryptamine Derivative of Phenyl Isocyanate A solution of phenyl isocyanate (1 g) in 10 ml of acetonitrile was added dropwise, with stirring, to 100 ml of an acetonitrile solution containing 0.5 g of tryptamine.The solution was allowed to stand for 1 h and the derivative was recrystallized from acetonitrile. The urea derivative [m.p. 196 “C (decomp.)] was also identified by the IR band at about 1650 cm-1. Study of Competitive Reaction Rates by Derivatizing Phenyl Isocyanate With Amines There were several options for conducting the experiments for this study. The competitive reaction rate is a relative rate between two competing amine reagents reacting concurrently with the isocyanate. Therefore, only two derivatizing agents were used for each specific rate study. To a set of 50 ml calibrated flasks, each containing a mixture of amines, 5 ml aliquots of acetonitrile solutions of phenyl isocyanate were added individually. For the sake of simplicity, the amount of the individual amines in each flask was kept constant while the amount of phenyl isocyanate added was varied.The contents of the flasks were allowed to react for 1 h before diluting to volume with acetonitrile. All solutions were diluted with acetonitrile to the appropriate concentrations for HPLC evaluation. A small amount of acetic anhydride (in acetonitrile), sufficient to remove the excess of amine, was also added to each flask before HPLC analysis. The data for the competitive derivatization rates for tryptamine and MPP, tryptamine and PP, tryptamine and NNNP, and NMTP and MPP are listed in Tables 1 4 . A typical HPLC trace of a solution of phenyl isocyanate containing tryptamine and MPP is shown in Fig.1. Study of Competitive Reaction Rates by Derivatizing Phenyl Isocyanate With Tryptamine and Water As water is known to react with isocyanates and exists in the atmosphere, it is important to have some knowledge about its relative reaction rate in comparison with those of the amines. It is also known that isocyanates react much more slowly with water than with amines. Therefore, the amount of water used for the experiments has to be fairly large in order toANALYST, JANUARY 1991, VOL. 116 Table 1 Relative rate for derivatizing phenyl isocyanate with tryptamine and MPP 23 Phenyl iso- Tryptamine MPP (b)l cyanate (i)/ PI-TP* (x)l PI-MPPt (y)l (a)lpmol p o l pmol pmol p o l Log[(a - x)/a] Log[@ - y)/b] 1.563 1.115 1.050 0.599 0.451 - 0.2 1006 - 0.22485 -0.17512 1.563 1.115 0.840 0.470 0.370 -0.15534 1.563 1.115 0.630 0.349 0.281 -0.10974 -0.12611 - 0.08240 1.563 1.115 0.420 0.227 0.193 -0.06825 1.563 1.115 0.210 0.105 0.105 -0.03029 - 0.04282 klk' = 0.979$ * Tryptamine derivative of phenyl isocyanate.t MPP derivative of phenyl isocyanate, obtained with the assumption that i = x + y . $ Obtained from the slope of the log[(a - x)/a] versus log[(b - y)/b] plot. Table 2 Relative rate for derivatizing phenyl isocyanate with tryptamine and PP Phenyl iso- Tryptamine PP (b)/ cyanate PI-TP* ( x ) l PI-PPt ( y ) l (a)/ pmol pmol (i)/pmol pmol pmol Log[(a - x)/a] Log[(b - y)lb] 1.563 1.154 1.050 0.724 0.326 -0.27020 -0.14418 1.563 1.154 0.840 0.569 0.271 -0.19657 -0.11625 1.563 1.154 0.630 0.407 0.223 -0.13100 -0.09326 1.563 1.154 0.420 0.280 0.140 -0.08573 - 0.056 17 1.563 1.154 0.210 0.128 0.082 -0.03711 - 0.03201 klk' = 2.011 * Tryptamine derivative of phenyl isocyanate.t PP derivative of phenyl isocyanate, obtained with the assumption that i = x + y . Table 3 Relative rate €or derivatizing phenyl isocyanate with tryptamine and NNNP Phenyl iso- Tryptamine NNNP (b)/ cyanate PI-TP* (x)/ PI-NNNP? (y)/ (a)lpmol pmol (i)/pmol pmol pmol Log[@ - x)/a] Log[@ - y ) h ] -0.00976 1.563 15.03 1.050 0.716 0.334 - 0.26608 1.563 15.03 0.840 0.560 0.280 -0.19266 - 0.0081 7 1.563 15.03 0.630 0.449 0.181 -0.14707 - 0 .00526 -0.09949 -0.00290 1.563 15.03 0.420 0.320 0. 100 1.563 15.03 0.210 0.163 0.047 -0.04783 -0.00136 klk' = 23.63 * Tryptamine derivative of phenyl isocyanate.t NNNP derivative of phenyl isocyanate, obtained with the assumption that i = x + y . Table 4 Relative rate for derivatizing phenyl isocyanate with NMTP and MPP Phenyl iso- NMTP (a)/ MPP (b)l cyanate PI-NMTP* PI-MPPt pmol pmol (i)/pmol bJ)/Pmol (x)lpmol Log[(a - x>la] Log[(b - y ) h ] - - 1.437 O$ 1.050$ 1.050 0 1.437 2.800 1.050 0.640 0.410 -0.25600 -0.06876 1.437 2.800 0.840 0.512 0.328 -0.19132 - 0.0541 1 - 0.03790 1.437 2.800 0.630 0.396 0.234 -0.14001 1.437 2.800 0.420 0.259 0.161 -0.08631 -0.02572 -0.01 195 1.437 2.800 0.210 0.134 0.076 - 0.0425 1 klk = 3.744 * NMTP derivative of phenyl isocyanate. t MPP derivative of phenyl isocyanate. $ Derivatization without addition of MPP to the reaction mixture, used for calibrating the amount of PI-NMTP produced in the set.differentiate the rate from that of the amine. The competitive reaction rate for tryptamine and water is shown in Table 5. Summary of Competitive Derivatization Rates and Over-all Comparison of the Methods The competitive derivatization rates obtained above would be more meaningful if relative rates could be assigned to all the amines investigated. Table 6 shows the relative rate constants listed in descending order; the rate for MPP was arbitrarily assigned a value of 100. An over-all comparison of the various HPLC methods is presented in Table 7. Experimental data reflecting the relative rate constant of klk' are plotted in Fig. 2. Discussion Although many HPLC methods for determining airborne isocyanates have been published, no study of the competitive derivatization rate of amines has been reported.This work has shown that the relative rates for the derivatization of phenyl isocyanate by MPP and tryptamine are almost identical and24 ANALYST, JANUARY 1991, VOL. 116 -TP ,TP tlmin Fig. 1 Chromatogram of competitive derivatization of phenyliso- cyanate with tryptamine and MPP. (a) Solution containing 1.05, 1.56 and 1.12 pmol of phenyl isocyanate, tryptamine and MPP, respec- tively, before dilution; and (b) solution containing 0.63, 1.56 and 1.12 , pmol of phenyl isocyanate, tryptamine and MPP, respectively, before dilution Table 5 Relative rate for derivatizing phenyl isocyanate with tryptamine and water Phenyl iso- Tryptamine H20 (6)l cyanate (i)l PI-TP* (x)/ PI-H20t (a)lpmol pmol pmol pmol (y)/pmol 1.563 1.39 x 104 1.050 AS B§ 1.563 1.39 x 104 0.840 A B 1.563 1.39 x 104 0.630 A B 1.563 1 .3 9 ~ 104 0.420 A B 1.563 1.39 x 104 0.210 A B Log[(a - x)/a] > - 0.47 log[(b - y)/6] 2 - 3 x 10-67; klk' > 1 x 10s * Tryptamine derivative of phenyl isocyanate. t Derivative of phenyl isocyanate with water. S Approximately 100% yield. 0 Approximately zero. 7 Assuming the HPLC technique fails to differentiate up to a 5% yield of PI-H20. Table 6 Relative reaction rates for derivatizing phenyl isocyanate with various amines Derivatizing agent Relative rate constant, k NMTP MPP Tryptamine PP NNNP Water 374 100 98 49 4 a x 10-5 are two and 25 times faster than those for PP and NNNP, respectively. Experiments conducted earliela by using a Test Atmosphere Generation System indicated that the recoveries of airborne toluene diisocyanate in the respective impinger solutions of MPP and tryptamine were indeed very competi- tive.As can be seen from Table 6, NMTP is the most reactive of the amines investigated, being about four times more reactive than MPP. However, a detailed study of the application of NMTP to the determination of isocyanates has not been conducted because of the high cost of this amine. Further, the amines used for derivatization are always present in a large excess, which is unlikely to affect the over-all yield of the derivatives caused by reactions with slightly slower rates. On the other hand, a very much slower derivatization rate was obtained with NNNP. The main concern is that the NNNP method is still being used widely for quantifying airborne isocyanates in workplaces and may not be able to reflect the true exposure.The results indicate that a much smaller amount of tryptamine or MPP is required than of NNNP to efficiently derivatize an equivalent amount of isocyanate. This would also be of benefit in the HPLC system, because less material would need to be loaded on to the column if tryptamine were to be used for derivatization. It has been observed in previous experimental work6 that the derivatization of isocyanates with NNNP is less efficient than with tryptamine or MPP, as reflected by the consistently lower results. It was suspected that the solvated water might have been partly responsible as NNNP has to be extracted from water before use.However, the relative derivatization rate study indicates that this is not the situation as it would have affected the rate of the other competing amines 300 I L I I 0 100 200 [-Log(a - x)/al x 100 Fig. 2 Plot of second-order reaction kinetics for competitive rates for derivatizing phenyl isocyanate with amines. A, Tryptamine and NNNP; B, NMTP and MPP; C, tryptamine and PP; and D, tryptamine and MPP Table 7 Comparison of representative HPLC methods for determining isocyanate Derivatizing agent Availability for Type of No. of specific determining of unidentified detection detection modes pol yisocyanate isocyanate Availability for quantifying NCO* NNNP U1 traviolet NAMAS Fluorescence PP Ultraviolet MPP Ultraviolet and amperometric TryptamineP Fluorescence and amperometric ot 1 0 1 2 No No No Yes Yes No No No No Yes * Reactive isocyanato functional group.t UV absorption common for organic chemicals, regarded as a non-specific detection mode. $ 1-Naphthalenemethylamine. P Employed in the proposed method.ANALYST, JANUARY 1991, VOL. 116 25 simultaneously if the derivatization rate for water were competitive. In fact, the relative rate for water to compete with amine for the derivatization of isocyanate is negligible. For instance, our work shows that the derivatization of phenyl isocyanate by tryptamine is at least 1 x 105 times faster than by water. A slower derivatization rate for NNNP could also be caused by the instability of the amine form of the reagent. However, as the NNNP solutions for derivatization were always pre- pared freshly before use, the instability of the amine form of the reagent does not appear to be the reason for the slower rate.Therefore, it is more probable that the nature of the reaction kinetics is responsible, but the exact cause is still unknown. It should be noted that in all the experiments conducted on competitive derivatization rates the actual yield of tryptamine- derivatized phenyl isocyanate was used as the basis for all the necessary calculations. This was because of the highly fluorescent nature of this derivative. The analytical inter- ferences on high-performance liquid chromatography would be minimized by fluorescence detection. One of the most important reasons for performing the relative rate study was to provide the best possible assessment in developing personal sampling techniques for isocyanate exposure in the workplace.The preferred air collection media for personal sampling at work sites are solid absorbents instead of impinger solutions. For a reagent-coated solid absorbent, this implies that the derivatization would occur in the unfavourable solid phase of the reagent. Amines with faster derivatization rates would therefore be more suitable for coating the solid absorbents. Considering all the factors such as the sensitivity and selectivity of detection of the methods discussed for isocyanates, the use of tryptamine is recommended for future studies on sampling airborne iso- cyanates using solid absorbents. By comparing all the features of the HPLC methods shown in Table 7, the proposed method is the most suitable for practical applications. Moreover, the proposed concept of the isolation of a selected x-system in a derivative for specific detection has promoted a new area for exploration in analysis with chromatographic techniques. References 1 Dunlap, D . A . , Sandridge, R. L., and Keller, J . , Anal. Chem., 1976,48, 497. 2 Levine, S. P., Hoggatt, J . H . , Chladek, E., Jungclaus, G . , and Gerlock, J . L., Anal., Chem., 1979, 51, 1106. 3 Goldberg, P. A . , Walker, R. F., Ellwood, P. A . , and Hardy, H. L., J. Chromatogr., 1981,212,93. 4 Walker, R. F., Ellwood, P. A., Hardy, H. L., and Goldberg, P. A . , J. Chromatogr., 1984,301,485. 5 Bagon, D. A . , Warwick, C. J . , and Brown, R. H., Am. Znd. Hyg. Assoc. J . , 1984, 45, 39. 6 Wu, W. S., Stoyanoff, R. E., Szklar, R. S . , Gaind, V. S . , and Rakanovic, M., Analyst, 1990, 115, 801. 7 Wu, W. S . , Nazar, M. A . , Gaind, V. S . , and Calovini, L., Analyst, 1987, 112, 863. 8 Wu, W. S . , Szklar, R. S., and Gaind, V. S., Analyst, 1988,113, 1209. 9 Wibaut, M. J. P., Rec. Trav. Chim., 1915, 34, 241. 10 Ingold, C. K., and Smith, M. S . , J. Chem. SOC., 1938,905. NOTE-References 6, 7 and 8 are to Parts 3, 1 and 2 of this series, respectively. Paper 01031 62 C Received July 13th, 1990 Accepted September 7th, 1990