Journal of Analytical Toxicology, Vol. 21, January/February 1997
Reduction in Extraction Efficiency of Charged Particles
from the Ion Source as the Cause of Matrix Effects in the
GC-MS Analysisof Drugs*
Donna Ostheimer 1,z, Michael Cremese 2, Alan H.B. Wu 2, and Dennis W. Hill1, ~
IMicrochemistry Laboratory, University of Connecticut, Storrs, Connecticut 06269 and 2Department of Pathology and
Laboratory Medicine, Hartford Hospital, Hartford, Connecticut 06102
[Abstract
[
Previousstudies haveshown that the ion responseof a compound
can be supressedby the presence of a large amount of a coe|uting
substancein a gas chromatographic-massspectrometric (GC-MS)
system. In the presentstudy, the change in the ion current of a
constant amount of diazepam-ds in the presenceof a 100-fold
amount of diazepam was used to monitor this condition in the
Hewlett Packard massselective detector (MSD). It was observed
that a reduced recovery of ions occurred when the potentials of
the MSD source elemenls were establishedby the autotune
algorithm. Increasingthe ion focus or the entrance lens potentials
or both increasedthe recovery of the ion current of diazepam-ds
in the presence of large amounts of diazepam. The data suggested
that the decreased recovery of ion current observed when the
autotune source parameters were used was due to insufficient
energy on the focusinglensesto extract a constant fraction of the
ions from the sourcewhen a high concentration of molecules was
present.
Introduction
Wu et al. (1--4) recently reported that the coelution of high
concentrations of fluconazole with the trimethylsilyl (TMS)
derivative of benzoylecgonine interfered with the gas chromatographic-mass spectrometric (GC-MS) detection of benzoylecgonine in urine extracts. Preliminary studies using a
Hewlett Packard (Palo Alto, CA) 5890 GC and 5970B mass
selective detector (MSD) indicated that the interference most
likely occurred in the ion source (4). This was substantiated by
observing the interference of high concentrations of fluconazole with the detection of benzoylecgoninein a direct probe MS
(HP 5988B) analysis (3). Depletion of the derivatization reagent
was also ruled out as the cause of false negative results by
observing the negative interference after derivatizing benzoylecgonine and fluconazole separately and combining them
* Prese,nledin part at the 471~ Annual Meelin~ o( the AmericanAssociation(or Clinical Chemistry,
Anaheim, CA, July 1995.
t Author 1owhom correspondenceshould be addressed.
just before GC-MS analysis (3). In other studies, it was observed that the coelution of high concentrations of a
methadone metabolite caused a decrease in the ion current response of the TMS derivativeof benzoylecgonine (1). Two other
coeluting drug pairs, propoxyphene and propoxyphene-d7and
methaqualone and methaqualone-d4, were also studied and
showed a similar concentration-dependent ion current decrease in the MSD (4). Propoxyphene and methaqualone did
not require derivatization, which demonstrated that reagent
depletion did not play a role in the mechanism of interference.
Matrix effects in the MS have also been reported by investigators in the analysis of environmental compounds (5-8). Previous studies by Tondeur et at. (5) and Marbury et al. (6) identified the ion source as the cause of matrix effects in these
samples. The GC-MS instruments studied by these investigators were an HP 5700 GC with a VG Micromass (Altrinchem,
Cheshire, U.K.) ZAB-ZFMSand a Varian (Walnut, Creek, CA)
model 3400 GC with an Extrel 400-1 MS, respectively. In the
analysis of 2,3,7,8-tetrachlorodibenzodioxin (TCDD), a significant reduction (over 50% relative to pure standard) in the
response for 100 pg of TCDD mixed with various amounts of
methylstearate (400 ng to 1 rag) was observed (5). Marbury et
al. (6) observed unusual nonlinearity effects (enhancement in
response) at high concentrations in the analysis of TCDD.
Although this phenomenon is likely to occur in all mass
spectrometers, the magnitude of the effect would likely vary
depending on the ion source design of the instrument. However,
characterization of the mechanism of this interference should
provide a basis for the detection of this type of interference by
routine drug testing laboratories independent of the type of
instrumentation used. An investigation of the mechanism of
this interference, using the Hewlett-Packard mass selective
detector {HP 5970B) as a model, is presented in this study.
Experimental
Reagents and standards
Ethyl acetate and methanol were obtained as GC-MSgrade solvents from Burdick and Jackson (Muskegon, MI).
Reproduction (photocopying) of editorial content of this journal is prohibited without publisher's permission.
17
Journalof AnalyticalToxicology,Vol,21,January/February1997
Perfluorotributylamine (PFTBA)was obtained from HewlettPackard (Kennett Square, PA).
Stock diazepam (1.0 g/L in methanol) and diazepam-d5
(0.10 g/L in methanol) standard solutions were purchased
from Sigma (St. Louis, MO). Methanolic solutions containing
12.5 mg/L diazepam-d5with 0.00, 1.25, 2.50, 5.00, 12.5, 25, 50,
125, 250, 500, 750, 1000, or 1250 mg/L diazepam were prepared from appropriate dilutions of these stock standards. For
identification, each standard will be referred to as D5x, where
x represents the amount (ng) of diazepam injected on the
GC along with 12.5 ng of diazepam-ds (i.e., D512.5refers to
the analysis of 12.5 ng of diazepam-d5 along with 12.5 ng
diazeparn). Diazepam without diazeparn-ds present was prepared in methanol at 1.25, 12.5, and 1250 mg/L.
Instrumentation
Studies were performed on a Hewlett-Packard (Palo Alto, CA)
5970B quadrupole MSD controlled by HP-IJX series ChemStation software. The MSD was interfaced to a model 5890 GC that
was fitted with a model 7673Aautomatic injector. An HP Ultra1 capillary column (12 m x 0.2 ram; 0.33-mm film thickness)
was used with a carrier gas of helium at a flow rate of 0.7-1.0
mL/min. The injector was operated in the splitless mode at
275~ One microliter of sample was injected onto the column
at 150~ The column temperature was held at 150~ for
1 min, then increased to 280~ at 30~
and held at this
temperature for 1 rain. Electron impact ionization (EI) was
performed at 70 eV. Data were acquired in the selected-ion
monitoring (SIM) mode with a dwell time of 50 ms with low
mass resolution. The most intense ions of diazepam (m/z 284
M§ 256, and 221) and the corresponding deuterated ions in
diazepam-d5 (rn/z 289 M§ 261, and 226) were monitored.
Detector response was measured as peak area by baseline-tobaseline integrations of the ion chromatograms.
Methods
The study examined the effect of increasing concentrations
of diazepam on the ion current response of a fixed concentration of the coeluting deuterated analogue, diazepam-ds, after
ion source potentials were established by the autotune algorithm. The MSD was calibrated using the autotune algorithm.
Standard methanolic solutions (1 pL) containing 12.5 mg/L
diazepam-ds and increasing concentrations of diazepam (0,
1.25, 2.50, 5.00, 12.5, 25, 125, 250, 500, 750, 1000, and 1250
rag/L) were analyzed in duplicate on the GC-MSD. One microliter of 12.5, 125, and 1250 rng/L diazeparn was also analyzed
in duplicate. The relative peak-area ratios of the rn/z 226/221
ions, the m/z 261/256 ions, and the rn/z 289/284 ions were calculated and averaged from the respective ion chromatogams of
diazepam. At each concentration of diazepam coanalyzed with
diazepam-ds, the peak area contribution of diazepam to the rn/z
226, 261, and 289 ions was calculated and subtracted from
the total peak area of the respective ion chromatogram. The
average corrected peak areas of the m/z 226, 261, and 289 ion
chromatograms of the 12.5 ng of diazepam-ds coanalyzed with
each amount of diazepam were divided by the respective
average peak areas of the 12.5 ng of diazepam-d5 analyzed
without diazeparn. These ratios were multiplied by 100 to give
the D5x/Ds0 relative peak area (RPAx/0).The response factor
for each concentration of diazepam (m/z 284) and the 12.5 ng
of diazepam-d5 (m/z 289) was calculated for each analysis by
dividing the respective ion chromatogram peak area by the
mass (ng) of the injected compound.
The effect of varying the potentials of the ion source elements (Figure 1) on the ion response recovery of 12.5 ng
diazepam-d5 in the presence of 1250 ng diazepam was studied.
Methanolic solutions (1 pL) containing 12.5 mg/L of diazepamd5 and either 12.5 or 1250 mg/L were analyzed in duplicate on
the GC-MSD system following an MSD autotune calibration of
the mass spectrometer. The standard solutions were then
reanalyzed in duplicate on the same system with each source
element, in turn, set at selected potentials while leaving all
other elements at their autotune settings. Ion current recoveries were evaluated at repeller potentials of 0.0, 2.0, 4.0, 6.0,
8.0, and 10.2 V; ion focus potentials of 0.0, -50, -100, -150, and
-200 V; and entrance lens potentials of-25, --48, -75, -100,
Automated tuning of the MSD
The MSD was calibrated on the m/z 69, 219, and 502 ions of
PFTBAusing the Hewlett-Packard MSD autotune algorithm (9)
to adjust the potentials on the source elements. The accepted
abundances of the rn/z 219 and 502 ions relative to the m/z 69
ion in the mass spectrum of PFTBAwere reported to be 50.7
and 2.6%, respectively (10). The MSD autotune algorithm
adjusts the ion source and electron multiplier potentials to maximize the intensity
Collimating
Magnet
of the m/z 502 ion while maintaining the
abundances relative to the m/z 69 ion at
LensStack
Filament.~
>35% and >1% for the rn/z 219 and 502
Quadrupoles
ions, respectively. The autotune calibration
[
Entrance Drawoutl
relative abundances obtained for the rn/z
Sample
nIelt ~
elIr
Slit [ L e n s ~
219 and 502 ions on the MSD used in the
present study were frequently between 40
and 65% and 10 and 16%, respectively.The
ExitS,it
repeller potential was consistently established by the autotune algorithm at the
GroundedSource
EntranceLens
Chamber
r ~ TargetElectrode
maximum value of 10.2 V. For each experIon FocusLens
Contact Lens
iment, the reference potentials on the
source elements were determined by the
Figure 1. Diagramof the MSDion source.
MSD autotune program.
/I
18
I,
P
\
EMDetector
II
I I\
X-RayLens
Journal of Analytical Toxicology, Vol, 21, January/February 1997
than that of D51250at a -150 and -250 V. The D51250response
increased with increasing entrance lens potential from -25 to
-100 V and then decreased slightly at -125 and -250 V.
-125, and -255 V.The average peak area (m/z 289) of the 12.5
ng of diazepam-ds that was coanalyzed with 1250 ng of
diazepam was dividedby the average peak area (m/z 289) of the
12.5 ng of diazepam-ds that was coanalyzed with 12.5 ng of
diazepam multiplied by 100 to give the D512so/D512.srelative
peak area (RPA12so/lz.s).
Discussion
Previous work with benzoylecgonine, propoxyphene, and
methaqualone showed a decrease in mass spectrometric ion
current response of moderate amounts of these compounds in
the presence of high amounts of coeluting substances (1-5). In
the present study, diazepam and diazepam-dswere evaluated as
probes to monitor this effect. These compounds were selected
because they do not require derivatization and have low
adsorptive properties in the injection port, column, and mass
spectrometer. When using ion source potentials set by the
MSD autotune algorithm, these compounds exhibited the same
decrease in ion response (Figure 2) that was previously
observed. It was also observed that as the total amount of
Results
Figure 2 shows the RPA~/0of the m/z 226, 261, and 289 ions
of 12.5 ng of diazepam-ds in the presence of increasing
amounts of coetuting diazepam (0 to 1250 ng) followingtuning
of the mass spectrometer to PFTBAusing the MSD autotune
algorithm. A decrease in the RPAx/0of diazepam-d5 response
was observed for each of the ions when greater than 25 ng of
diazepam was present. Less than 10% of the initial diazepamds m/z 256 and rn/z 289 ion response was observed when
1250 ng of diazepam was present. The m/z 284 ion response
factor of diazepam (0 to 1250 ng) and
the m/z 289 ion response of diazepam-ds
(12.5 ng) are plotted against the total
140
amount of injected solutes in Figure 3. The
response factor for both compounds
decreased when greater than 37.5 ng of
total solutes were analyzed.
~ 70The effectof the repeller potential on the
RPA12s0n2.s of diazepam-ds is shown
in Figure 4. The RPA12so/12.sincreased from
36% at the autotune repeller potential
(10.2 V) to 106% at a repeller potential of
0
200
400
600
800
1000
1200
1400
4.0 V. The RPA1250n2.swas less than that of
Diazepam coinjected with 12.5 ng diazepam-ds (ng)
the autotune value at repeller potentials of
6, 8, and 10 V. Figure 5 shows the ion chroFigure2. RPAx/oof the m/z 221,261, and 289 ions of 12.5 ng of diazepam-ds in presenceof increasing
matogram peak area response of D512.5and
amounts of diazepam. Detector settingsestablished by autotune algorithm: electron multiplier, -1800
D512so observed at increasing repeller
V; repeller, 10.2 V; ion focus, 0.0 V; entrance lens, -58 V.
potentials. The greatest response for D51z5
and D512s0was observed at repeller potentials of 6.0 and 8.0 V.
A linear increase in RPA12sonz.5was
0.25
---II---Dlazepam m e- -Diazepam-d 5
observed with an increase in ion focus
potential (Figure 6). The maximum peak
~ 0.20
area response of D512.sand D51zsooccurred
~
/I X
at an ion focus potential of-100 V (Figure
~ 0.15 I
7) which corresponded to an RPA12so/12.5of
~
/~
%
approximately 60% (Figure 6).
~ O.10
As the entrance lens potential was increased, RPAlzs0/12.5initially decreased and
n. 0.05
then increased to a value greater than 100%
(Figure 8). Figure 9 shows that the ion
O.OO - 200
400
600
800
1000
1200
chromatogram peak areas of D512.s and
O
1400
D51250did not respond in the same manner
Diazepam coinjected with 12.5 ng diazepam-d5 (ng)
to the changes in entrance lens potentials.
Figure 3. Ion current response factor (peak area/ng) of 12.5 ng of diazepam-ds (m/z 289 ion) and of
The D512.sresponse initially increased dradiazepam (m/z 289 ion) with increasing amounts of diazepam injected on the column. Sourceelement
matically from -25 to -50 V potentials, and
settingsestablished by autotune algorithm: electron multiplier, -1800 V; repeller, 10.2 V; ion focus, 0.0
then slowly decreased with increasing
V; entrance lens, -58 V.
entrance lens potentials to a response lower
m
~p.
m/z221
-
m/z261
-
- 9
-
-
m/z289
19
Journal of Analytical Toxicology, Vol. 21, January/February 1997
solute in the source increased, the response per mass of
injected compound decreased for both the compound that was
increasing in concentration (diazepam) and the compound
that was maintained at a constant concentration (diazepam-ds).
This indicated that all of the compounds present in the source
were affected by this condition, not just compounds at low
concentration.
Decreased sensitivity attributed to the ion source may be
caused by one of the following factors: a reduction in the
extraction efficiency of charged particles from the ion source
because of space-charge effects or poor ion focusing; a low
fraction of total emission current traversing the ionization
chamber; a build-up of ions on the source elements which
reduces the electrical field strength; or a high source pressure which decreases ionization because of a reduction in the
mean free path (MFP) of the ionizing electrons (5,11-15). The
detector may be a source of low sensitivity if the electron
multiplier is contaminated, has a low gain, or becomes saturated (15,16).
Based on previous reports (5,11-15), it was postulated that
the recovery of ions from the ion source (Figure 1) could be
enhanced by increasing the repeller potential, by increasing the
size of electron entrance apertures and ion exit apertures, by
changing the volume of the source, or by adjusting the potentials of the focusing lenses (ion focus and entrance lenses). On
the MSD, the autotune calibration function routinely sets the
repeller potential to the maximum setting of 10.2 V. Higher
potentials could not be evaluated. Changing the size of the
apertures or source volume on the HP MSD was not feasible
and was not evaluated. The focusing lens potentials could be
adjusted on the MSD and potentials different from those set at
autotune calibration were studied.
Although the RPA125o/12.5was approximately 100% at a
repeller potential of 4 V, the peak area response of D512.swas 8
to 11 times less than that observed at repeller potentials of 6
and 8 V. Even though the D512.5and D51250responses were similar at the 4 V potential, the sensitivity of the D512.5response
was compromised at this potential. Interestingly, the D512.5
response at the autotune setting of 10.2 V was 3.5 to 4.8 times
less than that at the 6 and 8 V repeller potentials, although the
RPAs25on2.5at these three potentials were similar. It would
appear that the optimum repeller potential for these ions
(m/z 284 and m/z 289) was 6 to 8 V.
The autotune algorithm routinely sets the ion focus potential to 0.0 V. In evaluation of the effect of repeller potential on
RPA]250n~.5, the ion focus was set at this potential. It was
observed that, with the repeller potential set at 10.2 V,
RPA125o/]2.5increased with increasing ion focus potential
(Figure 6). At an ion focus potential of-200 V, RPA]250n2.swas
almost 100%. As the ion focus potential was increased from
120.0 ]
100.0
90.0 ]
75.0
60.0 "I
50.0
,
EE
m 3O.O 4
0.0 ]"
0.0
0.o
-
2.0
4.0
6.0
8.0
Repeller potential (V)
10.0
12.0
Figure 4. RPA1250/12.5 (m/z 289 ion) of 12.5 ng of diazepam-ds with
increasing repeller potential. Other detector elements held constant at:
electron multiplier, -2400 V; ion focus, 0.0 V; entrance lens, -56 V.
30.0 "1
x
25.0
. -
0
-50
-200
Figure 6. RPA12~0/12.5(m/z 289 ion) of 12.5 ng of diazepam-ds with increasing ion focus potential. Other detector elements held constant as follows: electron multiplier, -2000V; repeller, 10.2 V; entrance lens, -48 V.
10.0 1
-
-100
-150
Ion focus potential (V)
--_ D5(12.5)- e- -D5(1250)
22.5]
15.o 4
~_
7.5
0.0 ~
0.0
~_ 2.51
0.o
2.0
l.O
6.0
8.0
10.0
12.0
Repeller potential (V)
Figure 5. Ion current response (m/z 289 ion) of 12.5 ng of diazepam-d5
in the presenceof 12.5 ng and 1250 ng diazepam with increasing repeller
potential. Electron multiplier, -2400 V; ion focus, 0.0 V; entrance lens,
-56 V.
20
...f
I
,
0
-50
,
,
-100
-150
Ion focus potential (V)
-200
7. Ion current response (m/z 289 ion) of 12.5 ng of diazepam-ds
in the presence of 12.5 ng and 1250 ng diazepam with increasing ion
focus potential. Electron multiplier, -2000 V; repeller, 10.2 V; entrance
lens, -48 V.
Figure
Journal of Analytical Toxicology, Vol. 21, January/February1997
0 V to -100 V, there was a concurrent increase in the peak area
of D512.sand D5125o(Figure 7) that suggested that the additional energy applied down field from the source was effective
in extracting a greater fraction of the ions from the source. In
addition to aiding in the extraction of ions from the source, the
ion focus lens acts to change the ion beam dispersion. The ideal
condition would be a potential that collimates the ion beam to
a point that allows all of the ions leaving the ion source to pass
into the mass analyzer. In the case of the D512.5sample, the
molecular dispersion would be low and require less energy to
focus the greatest fraction of molecules into the mass analyzer. The molecular dispersion of the molecules in the D512s0
sample would be much greater and would require greater
energy to focus the maximum fraction of molecules into the
mass analyzer. The data suggested that at -100 V, the maximum fraction of molecules from each sample was focused
into the mass analyzer; however, the fraction of total solute
focused was not the same for the D512.5and D51250samples. As
the ion focus potential increased to -200 V, the fraction of
total solutes in sample D51250 focused into the mass analyzer
remained the same while it appeared that the fraction of total
solutes focused into the mass analyzer from sample D5~2.s
decreased (Figure 7). At -200 V, the fraction of molecules
passing the ion focus into the mass analyzer were the same,
causing the RPA125o/12.5 to be 100%.
160.0 t
in
oJ
120.0
40.0
t
0,0 1
0
.
-50
.
.
.
-100
-150 -200 -250
Entrance lens potential (V)
-300
Figure 8. RPA12s0/12.s(m/z 289 ion) of 12.5 ng of diazepam-ds with in-
creasingentrancelenspotential.Otherdetectorelementsheldconstantas
follows: electron multiplier,-2000 V; repeller, 10.2 V; ion focus, 0.0 V.
4.0
--
D5(12.5)- 4l- -D5(1250)
3.0t~-
~,,_~
.~ 2.0
"~ 1.0
0.0
=
i
I
9
-50
-100
-150
-200
9
-2~
-300
Entrance lens potential (V)
Figure 9. Ion current response (m/z 289 ion) of 12.5 ng of diazepam-ds
in the presence of 12.5 ng and 1250 ng diazepam with increasing
entrance lens potential. Electron multiplier, -2000 V; repeller, 10.2 V; ion
focus, 0.0 V.
The entrance lens potential setting was the potential established at the entrance lens when m/z 502 was being monitored. The actual entrance lens potential varied relative to this
value as each mass-to-charge ratio was being detected. The
entrance lens value was established to regulate the velocity of
each ion entering the mass analyzer, the value being directly
proportional to the mass-to-charge ratio of the ion being monitored. With the ion focus potential set to 0 V, increasing the
entrance lens potential resulted in an initial increase in the
response of D512.s and D51250 (Figure 9). However, as the
entrance lens potential was increased further, the peak-area
response of D512.5and D51250decreased. Initially, the increased
energy was extracting a greater fraction of the total ions from
the source; however, because the entrance lens potential also
controls the velocity of the ions entering the mass analyzer, as
the potential was increased over -50 V, the efficiency of
detecting the m/z 284 or 289 ion decreased because the velocity
through the mass analyzer was too high. Although the
RPA125o/12.5 was near 100% (Figure 8) at an entrance lens
potential of-100 V, this was not the optimum setting for maximum efficiency of detecting these ions.
The data suggested that the mechanism of concentrationdependent ion current reduction involved a decrease in ion
extraction and ion focusing efficiencyof charged particles from
the ion source into the mass analyzer. The source potentials set
during an autotune calibration were not sufficient to extract a
constant fraction of the 289 ions generated by 12.5 ng of
diazepam-ds when the amount of coeluting diazepam exceeded
25 ng (Figure 2).
This matrix phenomenon has serious effects on quantitative
and qualitative analysis and therefore compromises the validity
of data obtained for detection and confirmation of drugs in
urine. The ability to detect matrix effects was affected by the
mode of analysis (full scan or SIM). Unknown substances that
coelute and interfere with the target compound can be readily
detected in the full scan mode by the presence of additional
ions contaminating the mass spectrum of the target drug. In an
SIM analysis, where at least three ions are monitored, interference can be recognized by unacceptable ion ratios when
the interfering compound has ions in common with the compound of interest. If a coeluting deuterated analogue of the
compound of interest, present at the threshold concentration
for the assay, is used as an internal standard, interference can
further be detected by a reduction in or complete absence of
ions monitored for the internal standard. This means of identifying an interfering compound is particularly important in
cases in which the ion ratios are not altered by the coeluting
compound.
In an SIM analysis in which the internal standard does not
coelute with the compound of interest, detection of an interfering compound with either the compound of interest or the
internal standard may be impossible. Suppression of the ion
response of the analyte below the threshold concentration or
the presence of altered ion ratios most likely would, under the
standard operating procedures used by many drug testing laboratories, result in a negative report without further investigation.
Stringent sample cleanup procedures are needed to remove
21
Journal of Analytical Toxicology,Vol. 21, January/February1997
matrix interferences to improve the specificity of the analysis.
Because of the complex nature of a urine sample, it is almost
impossible to design a cleanup procedure that would account
for all of the possible exogenous substances that may be present. To prevent lowered or elevated results from occurring,
analysts must be aware of this type of interference and modify
their procedures to incorporate strategies to detect it. This
may involve performing the analysis in the full scan mode or
using a deuterated coeluting internal standard at the threshold
concentration and monitoring at least three ions when performing the analysis in the SIM mode. If a coeluting substance
is detected, the problem can be eliminated by identifying the
substance and processing the sample to remove the interferant.
Conclusion
Concentration-dependent decreases in ion current in the
mass selective detector appeared to be due to insufficient energy
for extraction of the ions from the source at the source element
potentials established by the autotune algorithm. Increasing the
potential of the source elements to aid in the extraction of ions
at high concentrations caused inefficient focusing of ions into
the mass analyzer. The data indicated that there was a limited
dynamic range of total solute concentration in the ion source at
which efficient detection of compounds could be obtained. To
prevent false negatives results for the detection of target compounds, one must be aware of the presence of coeluting substances. The most effective way to determine the presence of
interferants is to perform the analysis in the full scan mode.
Alternatively, when performing the analysis in the SIM mode,
the recovery of a coeluting deuterated anatogue should be critically evaluated of a coeluting deuterated analogue should be
critically evaluated to detect possible detrimental effects of a
coeluting substance.
Acknowledgment
The authors would like to thank Albert J. Kind of the Microchemistry Laboratory, University of Connecticut and Paul
Goodly of Hewlett-Packard, Inc. for reviewing and discussing
interpretation of the data presented in this study.
22
References
1. E. O'Connor, D. Ostheimer, and A.H.B. Wu. Limitations of
forensic urine drug testing by methodology and adulteration.
AACC Ther. Drug Monit. Toxicol. Update 14:275-96 (1993).
2. D. Anderson, M. Cremese, E. Forte, A. Wu, and D. Hill. False negative results for benzoylecgonine in urine using GC/MS due to fluconazole [Abstract]. Clin. Chem. 39:1233 (1993).
3. D. Ostheimer, M. Cremese, E. Forte, A. Wu, and D. Hill. Characterization of one mechanism of negative interference in the
GC/MS analysis of drugs of abuse in urine. S.O.F.T./C.A.T. Joint
Meeting, Phoenix, AZ, October 1993.
4. A.H.8. Wu, D. Ostheimer, M. Cremese, E. Forte, and D. Hill.
Characterization of drug interferences caused by coelution of
substances in gas chromatography/mass spectrometry confirmation of targeted drugs in full-scan and selected-ion monitoring
modes. Clin. Chem. 40:216-20 (1994).
5. Y. Tondeur, RW. Albro, J.R. Hass, D.J. Harvan, and J.L. Schroeder.
Matrix effect in determination of 2,3,7,8-tetrachlorodibenzodioxin
by mass spectrometry. Anal. Chem. 56:1344-47 (1984).
6. G.D. Marbury, J.M. Turlington, F.R. Tuschall, and E.J. Kantor.
Unusual nonlinearity effects observed in the analysis of 2, 3, 7,
8-TCCC: Implications for quantitative accuracy: Chemosphere
18:819-26 (1989).
7. R.D. Voyksner, J.R. Hass, G.W. Sovocool, and M.M. Bursey. Comparison of gas chromatography/high resolution massspectrometry
and mass spectrometry/mass spectrometry for the detection of
polychlorinated biphenyls and tetrachlorodibenzofuran. Anal.
Chem. 55:744-49 (1983).
8. B.W. Patterson and R.R. Wolfe. Concentration dependence of
methylpalmitate isotope ratios by electron impact ionization gas
chromatography/mass spectrometry. Bio. MassSpec.22:481-86
(1993).
9. HP 59940A MS ChemStation (HP-UX series): MS Chemstation
Manual. Publication No. 59940-90004, 1989.
10. J.R. Donnelly, G.W. Sovocool, and R.K. Mitchum. Ion abundance criteria for GC/MS environmental analysis. J. Assoc. Off.
Anal. Chem. 71 : 434-39 (1988).
11. B. Hu and J. Qiu. The nonlinearity of upper limiting pressure for
quadrupole mass spectrometers. J. Vac. Sci. TechnoL5:2652-60
(1987).
12. E. Constantin and A. Schnell. MassSpectrometry. Ellis Horwood,
New York, NY, 1990.
13. R.I. Reed. Ion Production by Electron Impact. Academic Press,
New York, NY, 1962.
14. F.W. Karasek and R.E. Clement. Basic GasChromatography/Mass
Spectrometry: Principles and Techniques. Elsevier, New York,
NY, 1988.
15. G.F. Poole and S.A. Schuette. Contemporary Practice of Chromatography. Elsevier, New York, NY, 1984.
16. HP 59808 Mass Selective Detector: Hardware Manual. Publication No. 05970-90049. Hewlett-Packard Company, 1988.
Manuscript received March 18, 1996;
revision received July 8, 1996.