Anal. Chem. 1991, 63, i999-2001
1999
Determination of Ethylenethiourea in Crops Using Particle Beam Liquid Chromatography/Mass Spectrometry Daniel R. Doerge* and Carl J. Miles Department of Environmental Biochemistry, University of Hawaii, 1800 East West Road, Honolulu, Hawaii 96822
Several food crops were analyzed for residues of ethyienethbwea (ETU), a suspect thyrold and lvercarchogen present in EBDC fungicldos, using a commercial particle beam (PB) LC/MS method. The PB/LC/MS detection lhnits for ETU In crops (5 ppb, 1.25 ng) are comparable to those obtained by LC with electrochemical detection. Spectra obtained from crop samples containing as little as 5 ng of ETU were matched with the NBS library reference E1 spectrum. 190toplcaily labeled ETU was used as an internal standard for quantitatlon and determinatlon of recoverles. No enhancement of molecular kn m a l lntendty hwn unlabeled ETU was obeerved upon codutkn wtlh the WoplcaHy labeled variant. Thk MS method permits detection of ETU with Increased selectivity without compromising sensitlvity .
INTRODUCTION Ethylenebis(dithimbamate)(EBDC)fungicides are widely used on fruits, vegetables, cereals and other food crops (I). EBDC's are of regulatory concem because of ethylenethiourea (ETU,imidamlidine-2-thione), a formulation contaminant and environmental metabolite. ETU inhibits thyroid hormone production (2) is a rodent carcinogen (thyroid and liver), and has been classified as a probable human carcinogen (3). The analytical determination of ETU residues in foods by GC is hampered by its low volatility, which necessitates derivatization (4). LC with UV and electrochemical detection has proved useful in the quantitation of ETU residues in food crops (5,6).These LC methods, while sensitive, often do not give the selectivity and specificity required for method validation that is possible only with mass spectrometric detection. Particle beam (PB) mass spectrometry can provide compound identifiation and method validation for polar nonvolatile analytes that can be separated only by LC (7-9). This report describes the sensitive detection, quantitation, and confiiation of ETU residues in several important food crops using PB/LC/MS. EXPERIMENTAL SECTION Materials. ETU was obtained from Aldrich Chemical Co. (Milwaukee, WI) and recrystallized from water before use. lSCLabeled ETU was synthesized as previously described from 99% atom ex'%Sa (IO). LC solvents, acetonitrole (Optima,Fieher Scientific, Fairlawn, NJ) and Milli-Q water (Millipore Inc., Bedford, MA), were degassed by helium sparging. Ammonium acetate (HPLC grade) was obtained from Sigma Chemical Co. (St. Louis, MO). Liquid Chromatography. LC separation of ETU was performed by using a Perkin-Elmer Series 10 isocratic pump with a 4-mm4.d. X Wmm OmniPac PAX 500 column (Dionex Corp., Sunnyvale, CA) using 5 % acetonitrile in water at a flow rate of 0.25 mL/min. It was determined that m / z 102 intensity increased (and peak shape improved) with increasing acetonitrile concentration in the mobile phase. Since the presence of these higher amounts of acetonitrile eliminated retention of ETU on the column, acetonitrilewas added postcolumn at 0.2 mL/min with
* Author to whom correspondence should be addressed. 0003-2700/91/0363-1999$02.50/0
a Waters 6000 pump through a zero dead volume mixing tee to give a final mobile-phase composition of 47% acetonitrile/water into the nebulizer. However, sensitivity decreased when total flow rates >0.5 mL/min were introduced into the PB interface. Therefore, the mobile-phase composition and flow rate used represents a compromise between these two constraints. A UV detector (Model WIS 200, Linear Instruments, Reno, NV) was used for off-line detection of ET1 at 232 nm. Mass Spectrometry. Mass spectrometry was performed by using a VG Trio 2A (VG BioTech, Altrincham, U.K.) equipped with the LINC PB interface (VG BioTech) and a Hildebrand grid nebulizer (Leeman Labs). The effect of helium flow into the nebulizer on m / z 102 ion intensity was determined. It was found that sensitivity increased as He flow was increased up to 40 psi, after which no further effect was observed. The temperature of the desolvation chamber was maintained at 30 O C , and typical operating preesures were 8,0.9, and 3 X lod mbar at the fmbstage momentum separator pump, the second-stagemomentum separator pump, and the ion source housing, respectively. Positive ion spectra were obtained by using E1 (70eV) and full scan (m/z 70-300 in 1 s) conditions. The amount of ETU was determined from the area under the mass chromatogram of the M+ ion (m/z 102 or 103 for native or labeled ETU, respectively). The mass spectrometer was tuned and calibrated daily by using perfluorotributylamine,and the performance of the particle beam interface was monitored by periodic injection of 10 ng of l%XTU. The response to ETU was not significantly affected by source temperature over the range of 150-250 O C , and a source temperature of 200 O C was used throughout. Sample Preparation. Extraction and purification of ETU residues from crop samples was performed by using the AOAC method (4) as revised by Krause (5). Recovery data were calculated by linear regression by using the method of standard additions to crop control samples or by adding 10 ng of %-ETU as an internal standard prior to the extraction and purification procedures. All crops were obtained from commercial grocery stores. Briefly, a 50-g portion of vegetable was homogenized with aqueous methanol and filtered and an aliquot corresponding to 10 g processed further. The bulk of the solvent was removed in vacuo and the remnant applied onto Gas-Chrom S. The ETU residues were eluted through a column of alumina with methylene chloride, which was removed in vacuo prior to dissolution in water (4 mL) for LC/MS analysis (100-pL aliquots).
RESULTS AND DISCUSSION PB/LC/MS was employed to determine ETU in four economically significantcrops by using the AOAC method for extraction of residues. The background signal from solvent ions was minimal in the region m/z 102-103, which contributed to the low detection limits. Application of a small repeller potential (2-8 V) stabilized ion transmission into the mass spectrometer, resulting in lower pulsations and decreased noise. The crop fortification levels for ETU (10-50 ppb) were chosen to be relevant to the maximum residue limits (MRL) for many countries (12). The PB method is robust and does not suffer from interference by crop coextractives that can invalidate other detection methods (see Figure 1). UV detection was often sufficient to detect ETU in crop extracts but applesauce contained an interference (see Figure 1). However, this interfering peak did not affect detection by PB/MS (see Figure 1). Spectra obtained from as little as 5 ng added to crop extracts matched the NBS E1 library 0 I991 Amerlcan Chemical Society
2000
E
ANALYTICAL CHEMISTRY, VOL. 63, NO. 18, SEPTEMBER 15, 1991 A
UV 232
B
UV 232
D
MRlO2
-
60,
/I
I %FS
188
%FS E
I 0
-1. InterfaceshtheWdetectknofETUlnappleseuce. Panel A: ETU (5 ng on-column) with 232-nm UV detection. Panel B: Applesauce control extract with 232-nm UV detection. Panel C: Mass chromatogram(mlr 102) of applesauce control extract with FWMS detection. Panel D Applesauce control extract fortified with 5 ng of ETU (equivalent to 20 ppb).
2
4
6 [nul ( n d
10
8
Flgwe 3. Calibration curve for standard additions of ETU to a banana extract. Data shown represents the m n f standard deviations (n 2 3). A
M E 103
188
XFS
'YI
A
1881 I
I %BS
e l lee1
, , , , , . , , , , , , , , , , ,, I
n
1E8\ 7?
87
I
,)E4
145
158
188
%FS
B l0El
I
XFSI
A
8
Win
1.EE
'
2.88
n
n
3.88
4.88
i,
\ n n n , n ,I - ,
5.88
6.88
?.BE
Flgure 2. NBS E1 Ilbrary-searchable spectrum from 5 ng of ETU In a forwled lettuce extract. Panel A: Acquired spectrum (mlz 70-300) (top)andtheNBSEIlbmyspectrmofETU(bottom).Panels Maas chromatogram (mlz 102) of 5 ng of ETU (equivalent to 20 ppb).
Table I. Percent Recovery of ETU from Crop Samples crop lettuce applesauce banana pulp papaya pulp
ETU added, pg/kg
mean & SD, 5%
10 20 10 20 10 20 50 10 20 40 50
81 18 86 16 123 & 32 85 28 67 f 45 93 37 65 7
*
** 86+0 83 30 89 * 15 82 & 1
n 5 3 3 3 2 2 2 2 2 2 2
spectrum for ETU (see Figure 2). The methods used to extract residues from crops are those published by the AOAC (4) and revised by Krause (5). Table I shows the recoveries from crop samples fortified with known amounts of ETU prior to sample extraction. These recoveries (mean = 85 f 15%) are typical for those reported for extraction of ETU residues from a variety of crops ( 4 6 ) . The 10 ppb fortification corresponds to a total of 2.5 ng injected on-column. The detection limit (SIN = 3) for ETU in these crops was estimated to be 1.25 ng (5 ppb) on-column. These detection limits represent significant improvement over those determined for ETU by using two other commercial PB/ LC/MS systems (9). The mass spectrometer response was linear (13 = 0.984.99) and reproducible from 0-12.5 ng for standard additions to crop samples (see Figure 3). Linear standard curves were observed for quantities of ETU up to
-
.-
,
.
,
ME 102
Anal. Chem. 1991, 63,2001-2007
PB interfaces (12).Bellar et al. (13)observed enhancements of ca. 2-fold for coeluting isotopic variants of the same analyte. Their study also showed an enhancing effect on analyte response (including ETU) when ammonium acetate (1-50 mM) was present in the mobile phase. In the present study, added ammonium acetate (10-40 mM) had no effect on the PB response for 10 ng of ETU. The physical basis for this difference is presently unknown but could result from the different nebulizer used, the 40-70-fold lower mass of ETU analyzed in the present study, or interface hardware differences. This work demonstrates the utility of PB/LC/MS for use in determinii potential carcinogenic pesticide residues in food crops. With the recent advances made in production of bench-top PB/LC/MS instrumentation, this technique has high potential for routine application in pesticide residue laboratories.
ACKNOWLEDGMENT We thank Steve Bajic, VG BioTech, for helpful discussions and Linda Groves, Department of Environmental Biochemistry, for technical assistance. Registry No. ETU, 96-45-7.
200 1
LITERATURE CITED Natlonal Research Council. Reguktlng Pes/n Food; National Academy Press: Washington, D. C., 1987; pp 208-214. Doecge, Daniel R.; Takazawa, Richard S. CY”. Res. Toxlool. 1000, 3, 98-101. U.S. €PA. Fed. Regist. 1080, 54, 52158-52185. AOAC OfffcelMethods of Ana&sk. 15th ed; Association of Official Chemists Inc.: Arlington, VA, 1990 Vol. 978.18, pp 300-301. Krause. Rlchard, T. J . Assoc. Off. Anal. Chem. 1080, 72,975-979. Rev. Bottomly, Peter; Hoodless, Richard A,; Smart, Nlgel A. Res1085, 95, 45-64. Voyksner, Robert D.; Smlth, Cynthia S.; Knox, Patrick C. Bkmed. Enk l e Spectrom. ~ 1000, 79, 523-534. Klm, In Suk; Sadnos, Fassii I.; Stephens, Robert D.; Brown, Mark A. J . A@. Food Chem. 1000, 38, 1223-1255. Behymr, Thomas D.; Beliar, Thomas A.; Budde, William L. Anal. them. 1000.462, 1686-1690. Doerge, Daniel R.; Coocay. Niranjala M.; Yea, Austln 8. K.; Nlemczua, Walter P. J . Leb. Compounds Rad@hann. 1000, 28, 739-742. Krause, Richard T.; Wang, Yi. J. L i q . chrometug. 1088, 7 7 , 349-362. Ch. Lentza-Rizos, Rev. EnviLOn. Contam. T O W . 1000, 775, 1-37. Beliar, Thomas A.; Behymer, Thomas D.; Budde, William L. J . Am. SOC. Me= Spe~from.1000, 1. 92-96.
*.
RECEIVED for review March 22,1991. Accepted June 22,1991. This paper is submitted as Journal Series No. 3581 from the Hawaii Institute of Tropical Agriculture and Human Resources.
Isolated Dual Trapped Ion Cell Assembly for Fourier Transform Ion Cyclotron Resonance Mass Spectrometry ,
Steven A. Hofstadler and David A. Laude, Jr.* Department of Chemistry and Biochemistry, The University of Texas, Austin, Texas 78712
An electrically isolated dual trapped lon cell positioned in the strong magnetic field of a Fourier transform Ion cyclotron resonance (FTICR) mass spectrometer Is demonstrated to o v e r c m an Inefflclent electron ionization duty cycle that Is common to previous dual cell configurations. A repeller grid posltloned between dlfferentlally pumped trapped Ion cells aHows continuous Ion formation to occur In the source while excluding electrons from the analyzer cell during concurrent data acquldtkn. I n a qstematk study of dual cell performance krdudkrg trader Mkiency, mass selectMty, and signal enhancement, the high-field isolated dual cell deslgn was compared to conventlonal dual cdl and frlnglng field extemal cell arrangements. Several Ion trapping and manipulation schemes were Investigated. For the high-field Isolated cell, gated transfer and equlllbrlum pulse sequence transfer efflclencles routinely exceeded 80 % and 30 %, respectively. These values were slmHar to the performance achieved with the conventional dual cell deslgn. I n contrast, transfer efflckncies for the hinging lleld source were a factor of 5 poorer. However, when time-of-flight-based mas selection was deCdred, wperkr performance was derlved from the mghlg field source because of Increased path length. Finally, large lonlzatlon duty cycle Increases were achieved for continuous beam experiments In the Isolated dual cell with consequent S / N Improvements as high as a factor of 30 compared to the conventional dual cell In otherwise Identical high-resolution measurements. 0003-2700/91/0363-2001$02.50/0
INTRODUCTION The localization of ion formation, mass selection, and detection events is a unique feature of ion trap mass spectrometers including the Fourier transform ion cyclotron resonance (FTICR) instrument designed by Comisarow and Marshall (I,2). Although this physical consolidation of the experiment at first promised simplified and inexpensive operation as advantages, reduced performance for many mass spectrometry applications has since prompted the evolution to more complex and versatile FTICR instruments. These include external source spectrometers in which ions are formed outaide the magnetic field and then directed into the trapped ion cell (3-10)and multiple trapped ion cell configurations that exploit the strong magnetic field to trap ions in separate regions of the vacuum chamber (11-16).The simplest of these designs, developed by Nicolet Analytical Instruments, consists of adjacent cubic trapped ion cells that share a common trap plate which also serves as a conductance limit for differential pumping (11). At least three advantages are realized with this dual cell compared to the original single section cell. Of primary importance, differential pumping offers a solution to the pressure mismatch between high gas load sources or inlet systems and the conditions necessary for high-resolution FTICR detection. A second problem addressed by isolating reaction and detection cells is the high chemical interference background that arises due to uncontrolled ion-molecule reactions in a contaminated analyzer region. Finally, the limited dynamic range of the single cell has been addressed through @ 1991 American Chemical Society
