1120
Anal. Chem. 1981, 53, 1120-1 122
tions where a “Barker-Bolzan peak” and depressed limiting current in the plateau region of the wave occur. However, under the conditions used in Figures 1 and 3, the surface excess was measured at drop times of 2, 3, and 5 s and was found to decrease by less than 3% with increasing drop time. The present results have convinced us that the adsorption equilibrium of NAD+ is nearly truly achieved. It should be noted that in many cases Qd values large enough to measure accurately are obtained only when adsorption equilibrium is not reached; in such cases, NPP-C with a DME is not reliable. Obviously, NPP-C is valid when results are in agreement with those based on conventional techniques, Le., chronocoulometry or chronopotentiometry. When adsorption equilibrium is not achieved a t a DME, the NPP-C pulse train can be applied to a stationary electrode or a static mercury drop electrode (14);with the latter, the advantage of a renewable electrode surface is not lost.
LITERATURE CITED (1) Cummlngs, T. E.; Bresnahan, W. T.; Suh, S. Y.; Elving, P. J. J . Electroanal. Chem. 1080, 106, 71-83. (2) Chrlstle, J. H.; L a w , G.; Ostwoung, . - R. A,; Anson, F. C. Anal. Chem. 1983, 35, 1979. (3) Christie, J. H.; Lauer, G.; Osteryoung, R. A. J. Nectroanal. Chem. 1.-M-A.., 7. . , 60-72. - - . -. (4) Anson, F. C. Anal. Chem. 1084, 36, 932-935.
(5) Murray, R. W. in “Techniques of Chemlstry”; Weissberger, A., Rosslter. E., Eds.; Wllev-Intersclence: New York, 1971; Vol. I. Part IIA, Chapter 8. (6) Barker, 0. C.; Bolzan, J. A. Z . Anal. Chem. 1988, 216, 215-238. (7) Flanagan, J. E.; Takahashi, K.; Anson, F. C. J . €/ectroanal. Chem. 1977, 65, 257-286. (8) Temmerman, E.; Abel, R.; Osteryoung, R. A. J. Electroanal. Chem. 1974, 55, 173-186. (9) Cummlngs, T. E.; Jensen, M. A.; Elvlng, P. J. €lectroch/m. Acta 1978, 23, 1173-1184. (10) Schmakel, C. 0.; Santhanam, K. S. V.; Elvlng, J. P. J . Am. Chem. SOC. 1975, 97, 5083-5092. (11) Bresnahan, W. T.: Elvlng, P. J. J . Am. Chem. Soc.,In press. (12) Bresnahan, W. T.; Elvlng, P. J., unpubllshed results. (13) Wilson, A. M.; Epple, D. 0. Siochemistty 1968, 5 , 3170-3175. (14) Peterson, W. M. Am. Lab. (Fairflew, Conn.) 1070, 11 (Dec), 69-78.
‘
Present address: 19899.
Research Center, Hercules, Inc., Wllmlngton, DE
William T. Bresnahan’ Philip J. Elving* Department of Chemistry University of Michigan Ann Arbor, Michigan 48109
RECEIVED for review May 8, 1980. Resubmitted December 15, 1980. Accepted February 23, 1981. The support of the National Science Foundation is gratefully acknowledged.
Ribbon Storage Techniques for Liquid Chromatography-Mass Spectrometry Sir: Liquid chromatography-mass spectrometry (LC-MS) is a rapidly developing technique for the analysis of complex mixtures not amenable to gas chromatography-mass spectrometry (GC-MS) techniques ( I ) . Existing commercial LCMS interfaces use either (a) a direct introduction of a small flow of the LC effluent directly into a heated chemical ionization (CI) source ( 2 , 3 )or (b) deposition of the effluent on a moving ribbon with evaporation of the volatile mobile phase on progression through a series of vacuum locks and volatilization (or pyrolysis) of the material in (or adjacent to) a conventional CI (4) or electron impact (EI) (5) ion source. A number of other LC-MS interfaces using a variety of approaches have also been reported (6-10). The moving ribbon technique has the advantage that the LC flow rate is, in principle, not restricted and relatively large samples can be transported to the ion source due to nearly complete removal of the mobile phase. One difficulty with this approach, however, is that the source (or “flash heater”) temperature must be optimized for each compound of interest. Too high a temperature may cause the sample to pyrolyze or be volatilized outside the source while too low a temperature will result in only partial volatilization. We have developed a new LC-MS interface which allows semipermanent storage of the chromatographically separated material on a moving ribbon permitting multiple temperature analyses of a single LC separation. The new interface removes the major disadvantage of conventional moving ribbon devices by allowing analysis of a single LC separation at several different temperatures. EXPERIMENTAL SECTION Figure 1is a schematic illustration of the LC-MS interface. The system incorporates a number of departures from conventional practice including a new aerosol liquid deposition device (11), SIMS analysis as an alternate ionization mode (12),and a triple quadrupole mass spectrometer for improved selectivity (12). The 0003-2700/81/0353-1120$01.25/0
effluent from a Spectra Physics Model 8700 HPLC is sprayed on a slowly moving (5-60 cm/min) continuous ribbon (0.63 cm wide, 0.008 cm thick, 320 cm long). One unique feature of the interface is the inclusion of a 120 cm long region before the first vacuum lock. This increased ribbon length allows the semipermanent The aerosol deposition method requires only gas to effect evaporation of the liquid effluent in nearly all cases. Tests show that one can readily evaporate a variety of liquids (hexane, methylene chloride, 2-propanol, etc.) deposited at more than 2 cma/min and at a ribbon speed of 5 cm/min prior to the f i t vacuum chamber (11). The evaporation of the mobile phase is essentiallyinstantaneous under most flow conditions (11)allowing the use of very slow ribbon speeds (i.e., >10 cm/min). At the slowest speeds, some loss of chromatographic resolution appears unavoidable due to the size of the spray deposition area (-0.3 cm2) and the length of ribbon in the volatilization region (1cm). Thus, for a ribbon speed of 10 cm/min one would expect a maximum peak broadening of approximately 8 s, Most “sharp” LC peaks in our work are 10-30 s wide at half-height, and comparison of UV and reconstructed ion chromatograms shows that at ribbon speeds >20 cm/min such effects are usually negligible. Three regions of differential pumping are employed prior to the high-vacuum region. The first two regions are pumped at 10 L/s by “hot pumps”, maintained at >lo0 O C to limit the effects of condensable vapors during long-term pump operation, and the third by a 1500 L/s turbomolecular pump. Vacuum slits (-0.03 cm X 0.68cm) are machined from Teflon. The main drive wheel is also used to adjust ribbon tension and is motor driven through three universal joints; at no point does the sample surface of the ribbon contact another surface. Typical working pressures are approximately 40, 1, and torr in the three differentially pumped regions. The pressure in the high vacuum chamber is approximately lo-’ torr. The interface also incorporates a rhenium “cleanup” heater for removal of material after completion of the analysis. The flash heater is designed to rapidly heat a 1cm length of ribbon; heat shields and two copper wheels serve as thermal “drains” and prevent heating of the ribbon outside of the flash 0 I981 American Chemlcal Society
ANALYTICAL CHEMISTRY, VOL. 53, NO. 7, JUNE 1981
- COUNTING_.... BESSEL BOX
I
DEPOSIT1ON ATOMIZER
AUXILIARY AUXl L l ARY FLASH HEATER
ELECTRClN MUYIPLIER
I
I
--
. I
GUN
~~
\
LA Q U I D CHROMATOGRAPH INERT GAS INLET
Y
\ -
- SI LVER'VAPOR DEPOSITION SOURCE AND CLEAN-UP HEATER
"'I'
PUMP
I
STRIP HEATERS
1121
THERMOCOUPLE LOCATIONS
Schematic illustration of the LC-MS Interface. The system also includes a molecular-SIMS ionization capability and a triple quadrupole mass spectrometer not used in the present study. Figure 1.
I T
IIY
I
"11
518
BENZOlaiPYRENE
367
346 335 291 274
7 AMINO-4-HYDROXY QUINOLINE
252 246
IC4
)-HYDROXY QUINOLINE
578 )-HYDROXY I N D O L L I -
367 3M 335
124
144
164
184
204
244
224
264
-
291 274
285' C
a2
246 295
578
367
315
335
335
375
-
435
415
455
35OoC
An
4
38
_/I
) 4 a p
145
144
335
c
133
io8 485
505
525
545
545
585
605
621
645
SCAN NUMBER Figure 3. Reconstructed ion chromatograms for eight typical Ions In an analysis of a biomass product for a single Injection at three different ribbon passes through the "flash heater".
the flash heater at increasing temperatures, (b) decoupling of the moving ribbon speed for LC deposition from the desired rate through the flash heater (in this case MS analysis would commence one ribbon rotation after completing liquid deposition), and (c) a removal of the requirement to optimize the flash heater temperature for each compound of interest. Figures 2 and 3 illustrate the application of the ribbon storage technique with multiple flash heater temperatures to a standard mixture of eight compoundg and a raw synthetic fuel sample (a complex and as yet uncharacterized "biomass" wood liquefaction bottom product, i.e., the material remaining after removal of distillate fractions). The mass spectrometer is typically scanned over a mass range of 50-900 daltons. The HPLC separations used a 5-pm amino column, flow rates of 1mL/min, and a ternary gradient elution program (hexane,
1122
Anal. Chem. 1981, 53, 1122-1125
dichloromethane, and 2-propanol). Figure 2 shows the reconstructed ion chromatograms for eight ions indicative of the eight components in the standard trace mixture (there is also evidence of impurities or reaction products leading to additional peaks) for the same injection at two different temperatures. The ribbon speed for this analysis was approximately 10 cm/min and the mass spectrometer scans required 6 s. The quantity injected was approximately 1pg for each component, and the identity was confirmed by comparison with elution times for individual standards. The results show that these relatively volatile compounds are primarily desorbed during the f i t ribbon pass through the flash heater. The second pass at a significantly higher temperature shows evidence for only three of the more polar and less volatile components, clearly showing that the bulk of the material was desorbed during the fiit ribbon pass. In contrast, Figure 3 shows a wide range of volatilities for the wood liquefaction products during a similar gradient elution program. In this case, the behavior is illustrated by eight typical reconstructed ion chromatograms for a single injection and three ribbon passes through the "flash heater". For this analysis the ribbon speed was approximately 10 cm/min and the mass spectrometer scan speed was 10 s / m . The multiple analpis of the chromatographicallyseparated material clearly provides much additional data on the volatility of the components. The appearance of peaks at the same location at different temperatures may be due to the same compound or different compounds, depending upon their volatilities, and can be resolved by analysis of the mass spectra. As expected, one generally observes that the higher molecular weight components are desorbed at the highest temperature but, as illustrated in Figure 3, the actual situation can be much more complex. The data in Figure 3 were obtained by using the aerosol liquid deposition device; the improved performance may be observed by comparison with Figure 2 which shows the distorted peak shapes which resulted from the use of conventional deposition methods (11). The results in Figures 2 and 3 clearly illustrate the advantages of the ribbon storage technique. As demonstrated, compounds are selectively observed a t certain flash heater temperatures and discriminated against a t higher or lower temperatures. This technique can assist in interpreting complex chromatograms since it provides differential information related to compound volatility. For experiments where mass spectrometer scan requirements are more demanding (high resolution, simultaneous collision induced dissociation
in a triple quadrupole, or increased sampling time), the LC effluent can be deposited on the moving ribbon and the mass spectrometer analysis performed at a different ribbon speed after completion of the liquid deposition, totally decoupling mass spectrometer and HPLC operation. The technique should also be useful in the application of new "soft"ionization techniques which apparently produce ions directly from the solid surface (12). These applications are currently being investigated in our laboratory. ACKNO WLEDGMEN" We thank W. D. Felix for helpful discussions concerning the design, the assistance of A. W. Madsen and E. N. Sullivan in the design and construction, and K. A. Loss and J. E. Burger in the assembly of the LC-MS interface. LITERATURE CITED (1) McFadden, W. H. J . Chrometogr. Sci. 1979, 77, 2. (2) Mclafferty. F. W.; Knutti, R.; Venkataraghaven. R.; Arpbro, P. J.; Dawkins, B. G. Anal. Chem. 1975, 47, 1503-1505. (3) Melera, A. "Chemical and Analytical Applications of an LCMS Interface", Presented at Twenty-Seventh Annual Conference on Mass Spectrometry and Allied Topics, SeatUe, WA, June 3-8, 1979. (4) Yorke, D. A,; Bwns, P.; Millington, D. S. "a New HPLCMS Interface", Presented at Twenty-Seventh Annual Conference on Mass Spectrometry and Allied Topics, Seattle, WA, June 3-8, 1979. (5) McFadden, W. H.; Schwartz, H. L.; Evans, S. J . Chromatogr. 1976. 722, 389-390. (0) Henion. J. D. Anal. Chem. 1978, 50. 1687-1693. (7) Tsuge, S.; Hirata, Y.; Takeuchi, T. Anal. Chem. 1079, 51, 166-109. (8) Christensen, R. G.; Hertz, H. S.; Meiselman, S.; Whlte, E. "LC-MS UsIng Continuous Sample Preconcentration", Presented at Twenty-Seventh Annual Conference on Mass Spectrometry and Allled T o p b , Seattle, WA. June 3-8, 1979. (9) Blekely, C. R.; Adams, M. J.; Vestal, M. L. "LC-MS Interface Uslng Mdecular Beam Techniques", Presented at Twenty-Seventh Annual Conference on Mass Spectrometry and Allied Topics, Seattle,WA, June 3-8, 1979. (10) . . Blaketv. C. R.: Carmodv. J. J.: Vestal. M. L. Anal. Chem. 1960. 52, 1036-.1641. (11) Smith. R. D.; Johnson, A. L. Anal. Chem. 1081, 53, 731. (12) Smith. R. D.; Burger, J. E.; Johnson, A. L., submitted for publlcatbn in Anal. Chem.
Richard D. Smith* Allen L. Johnson Chemical Methods and Kinetics Section Physical Sciences Department Pacific Northwest Laboratory Richland, Washington 99352
RECEIVED for review December 22,1980. Accepted March 9, 1981. This paper is based on work performed under United Statm Department of Energy Contract DEACM-76RLO 1830.
Determination of Isocyanates in Air by Liquid Chromatography with Fluorescence Detection Sir: Isocyanates are extensively used in the manufacture of polyurethane foams, paints, elastomers, and fibers. With this widespread use, respiratory and allergic reactions have been reported by some workers exposed to these chemical substances. For diisocyanates, NOSH recommended a vapor concentration of 5 ppb as a time weighted average (TWA) concentration for a 10-h workshift, 40-h work week, and 20 ppb as a ceiling concentration for any 10-min sampling period (1). Most of the analytical methods used for determination of isocyanates in air are based on ultraviolet or visible absorption spectrometry (2-10) except for the tape monitors 0003-2700/81/0353-1122$01.25/0
marketed by MDA Scientific, Inc. (11). These techniques are often near their lower limits of detection when used for isocyanate concentrations at or below the 5-ppb range. Thus, utmost care in handling as well as expertise on the part of the analyst is generally required to obtain accurate results. In a contribution paving the way to markedly improved sensitivities, Levine and co-workers (12)introduced a method to determine aliphatic isocyanates in air by fluorescence detection. The isocyanates were converted to stable 1naphthalenemethylamine (NMA) urea derivatives and analyzed using isocratic reverse-phase liquid chromatography. 0 1981 Amerlcan Chemical Society