You are on page 1of 18

This article was downloaded by: [Sheffield Hallam University]

On: 19 November 2008


Access details: Access Details: [subscription number 773565538]
Publisher Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,
37-41 Mortimer Street, London W1T 3JH, UK

Polycyclic Aromatic Compounds


Publication details, including instructions for authors and subscription information:
http://www.informaworld.com/smpp/title~content=t713647615

CORRECTION OF ANALYTICAL RESULTS FOR RECOVERY:


DETERMINATION OF PAH s IN AMBIENT AIR, SOIL, AND DIESEL
EMISSION CONTROL SAMPLES BY ISOTOPE DILUTION GAS
CHROMATOGRAPHY-MASS SPECTROMETRY
Andrzej Wnorowski a; Mylaine Tardif a; David Harnish a; Gary Poole a; Chung H. Chiu a
a
Air Analysis and Quality Division, Environment Canada, Ottawa, Ontario, Canada

Online Publication Date: 01 December 2006

To cite this Article Wnorowski, Andrzej, Tardif, Mylaine, Harnish, David, Poole, Gary and Chiu, Chung H.(2006)'CORRECTION OF
ANALYTICAL RESULTS FOR RECOVERY: DETERMINATION OF PAH s IN AMBIENT AIR, SOIL, AND DIESEL EMISSION
CONTROL SAMPLES BY ISOTOPE DILUTION GAS CHROMATOGRAPHY-MASS SPECTROMETRY',Polycyclic Aromatic
Compounds,26:5,313 329
To link to this Article: DOI: 10.1080/10406630601028171
URL: http://dx.doi.org/10.1080/10406630601028171

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.informaworld.com/terms-and-conditions-of-access.pdf

This article may be used for research, teaching and private study purposes. Any substantial or
systematic reproduction, re-distribution, re-selling, loan or sub-licensing, systematic supply or
distribution in any form to anyone is expressly forbidden.

The publisher does not give any warranty express or implied or make any representation that the contents
will be complete or accurate or up to date. The accuracy of any instructions, formulae and drug doses
should be independently verified with primary sources. The publisher shall not be liable for any loss,
actions, claims, proceedings, demand or costs or damages whatsoever or howsoever caused arising directly
or indirectly in connection with or arising out of the use of this material.
Polycyclic Aromatic Compounds, 26: 313329, 2006
Copyright c 2006 Taylor & Francis Group, LLC
ISSN: 1040-6638 print / 1563-5333 online
DOI: 10.1080/10406630601028171

CORRECTION OF ANALYTICAL RESULTS


FOR RECOVERY: DETERMINATION OF PAHs
IN AMBIENT AIR, SOIL, AND DIESEL
EMISSION CONTROL SAMPLES BY ISOTOPE
DILUTION GAS CHROMATOGRAPHY-MASS
SPECTROMETRY
Downloaded By: [Sheffield Hallam University] At: 15:35 19 November 2008

Andrzej Wnorowski
Mylaine Tardif
David Harnish
Gary Poole
Chung H. Chiu
Environment Canada, Air Analysis and Quality Division,
Ottawa, Ontario, Canada

Results from soil, diesel emission and ambient air control


particulate matter samples on the determination of 30 PAHs
were evaluated to establish whether the use of recovery data
would result in an improvement of the quantitation
accuracy over the uncorrected data. The performance of the
recovery-corrected technique was initially evaluated using
recovery results from PAH standard reference material
samples spiked with analogue deuterated isotopes. The
results showed an excellent correlation between recoveries
of natives and corresponding surrogates for all matrices
studied. The practical merit of the isotope dilution mass
spectrometry technique was further assessed by spiking
control samples with corresponding isotopic analogues and
comparing the measured concentration of natives obtained
with uncorrected and recovery-corrected techniques. The
data revealed that the use of recovery correction leads to

Received 10 August 2006; accepted 25 September 2006.


The authors would like to thank Dr. Ewa Dabek, of Environment Canada, for her valuable
comments and technical revision of this manuscript.
Address correspondence to Andrzej Wnorowski, Environment Canada, Air Analysis and
Quality Division, 335 River Rd., Ottawa, ON, K1A OH3 Canada. E-mail: Andrzej.Wnorowski@
ec.gc.ca

313
314 A. Wnorowski et al.

results closer to the real values, thus decreasing the


negative bias due to losses that occur during the analytical
process. The mean accuracy difference between uncorrected
and corrected data is more pronounced as the sample
matrix becomes more complex, such as soil (15 12%) or
diesel emission (8 11%), and less for simpler ambient air
matrix samples (3 16%). Precision between the two
techniques was comparable within each matrix and
relatively close between the different matrices.

Keywords PAHs quantitation, isotope dilution mass spectrometry,


Downloaded By: [Sheffield Hallam University] At: 15:35 19 November 2008

recovery-correction, GC-MS environmental analysis

INTRODUCTION

Environment Canada has been monitoring PAHs as well as other pol-


lutants since the early 80s through a joint federal-provincial program,
the National Air Pollution Surveillance Program (1, 2). Continuous de-
velopment of analytical methodologies has led Analysis and Air Quality
Division (AAQD) chemists to apply isotope dilution mass spectrome-
try for quantitation as an advancement over the uncorrected method.
Clinical, toxicological and environmental studies have already recog-
nized recovery correction as a technique which provides more accurate
quantitation of various analytes regardless of their matrices (37) and to
compensate for ion suppression due to matrix effect (8).
In isotope dilution, prior to extraction, a sample is spiked with a known
amount of isotopically labelled analytes, referred to as surrogates. The
relative response factor of the surrogate to the corresponding native an-
alyte is used to quantify analyte concentration in samples. The principal
assumption in isotope dilution is that the chemical and physical prop-
erties between the analyte and its isotopic (deuterated or 13 C-labelled)
analogue are alike. Providing that the sample is homogeneous and that the
isotopic equilibrium between matrix and the surrogate has been achieved,
the recovery of isotopically-labelled analogue and native analyte from the
analytical process should be identical (9). Consequently, the calculated
recovery of the isotope is used to correct the recovery of a correspond-
ing native present in the sample. Quantitation with the isotope dilution
recovery-correction technique does not necessitate total recovery of an-
alyte since it compensates for matrix effects, analytical process loss and
analyte transformations. Likewise, the use of isotope dilution corrects
native analyte concentration for negative bias that occurred during the
analytical process.
PAHs Quantitation by Isotope Dilution GC-MS 315

Although the isotope dilution quantitation technique has found a per-


manent and essential place in QA/QC in many laboratories, to the knowl-
edge of the authors, there has not been any published study comparing
advantages of environmental PAH recovery correction with uncorrected
quantitation. In our enquiry, both techniques have been applied to the
quantitation of PAHs in various environmental matrices.
The objective of the study reported herein was to evaluate PAH quan-
titation with recovery-corrected isotope dilution from various environ-
mental matrices, and to compare the results with the conventional non-
corrected quantification, referred to as uncorrected. In this report, we
will not discuss nor provide details on sample preparation nor analy-
Downloaded By: [Sheffield Hallam University] At: 15:35 19 November 2008

sis conditions, but rather evaluate and compare the output of the two
quantitation approaches. Consequently, we will demonstrate the ability
of surrogate recovery correction to improve quantitation of natives, re-
gardless of the method conditions or instrument used. The data used for
the study was generated in 2003 and 2004 in AAQD laboratories, as a
requirement of our Quality Control Program, and re-analyzed for the
needs of the current study. During the two-year time span, the analysis
was performed following the same methodology by numerous analysts
using various instruments, therefore, the data in a sense could be regarded
as interlaboratory generated. This fact should be taken into consideration
to understand the impact of analytical bias on the presented results and
its correction.

MATERIALS AND METHODS

Standards

NIST Standard Reference Materials (Gaithersburg, MD, USA) were


used as controls for PAH determination in diesel emissions (SRM 2975,
SRM 1650a, SRM 1650) and ambient air (SRM 1649a). Both, certified
and reference concentrations in SRMs were used for accuracy determi-
nation. Soil samples were analyzed as part of the Canadian Association
of Environmental Analytical Laboratories (CAEAL) Proficiency Test-
ing Program. The deuterated surrogates used for recovery correction
(d8 -Naphthalene, d8 -Acenaphthylene, d10 -Acenaphthene, d10 -Fluorene,
d10 -Phenanthrene, d10 -Anthracene, d10 -Pyrene, d12 -Benz[a]anthracene,
d12 -Triphenylene, d12 -Chrysene, d12 -Benzo[b]fluoranthene, d12 -
Benzo[e]pyrene, d12 -Benzo[a]pyrene, d12 -Perylene, d12 -Indeno[1,2,3-
cd]pyrene, d14 -Dibenz[a,h]anthracene, d12 -Benzo[ghi]perylene, and
d10 -Fluoranthene as internal standard), with 99% purity or greater, were
obtained from CDN Isotopes, Pointe-Claire, QC, Canada.
316 A. Wnorowski et al.

GC-MS Analysis

Sample preparation and analysis were performed using Environment


Canada-AAQD in-house methods (10,11). Briefly, each reference ma-
terial sample was spiked with a 1,000 ng of each surrogate, allowed to
equilibrate, extracted with a solvent, and passed through silica column
cleanup. Prior to GC-MS analysis, samples were additionally spiked
with d10 -Fluoranthene which was used as an internal standard to calcu-
late surrogate recoveries and to quantify natives using the uncorrected
technique. Analysis was carried out on an Agilent 6890 GC interfaced
directly to an Agilent 5973 Mass Selective Detector. A 1 L injection
Downloaded By: [Sheffield Hallam University] At: 15:35 19 November 2008

was made with an Agilent 7683 Autosampler. The cool on-column injec-
tor was set to track GC oven temperature. A DB-XLB capillary column
(30 m 0.25 mm ID, 0.25 m film thickness) from J&W Scientific
Inc. was used for the separation of PAHs. Helium was utilized as a
carrier gas with a constant flow of 2.0 mL/min. The GC oven initial
temperature was set at 90 C, after 4 minutes increased to 200 C with a
rate of 20 C/min, then increased to 250 C at 2.5 C/min, then increased
to 283 C with a rate of 1.5 C/min and held for 6 minutes. Spectro-
scopic analysis was performed using 70 eV electron impact ionization
and selected ion monitoring mode. All analytes with the same molecular
and fragment ions were chromatographically separated. A minimum of
three ions were monitored for all native analytes, two characteristic ions
for each of the surrogate analytes and one ion for the internal standard
(Tables 1 and 2).

Quantification

The target PAHs were quantified using both isotope dilution correc-
tion and uncorrected procedures. Quality control monitoring of surrogate
recoveries ensured accurate and reliable recovery-corrected quantitation
of the analytes. The acceptable range of surrogate recoveries was set
to contain within 50 and 120%, with no more than two surrogates out-
side of these limits. Target compound identification was confirmed by
the presence of the analytes characteristic ions in the right ratio at the
appropriate retention time. Daily Calibration Check and Performance
Sensitivity standards were analyzed with every sample set to monitor
system performance. Enhanced Data Analysis MSD ChemStation Soft-
ware, version D.01.00, was used for data processing. Statistical data
treatment was performed with ANOVA analysis using the Microsoft
Excel 2003 SP3. Outliers, standard deviation, mean, and significance
PAHs Quantitation by Isotope Dilution GC-MS 317

TABLE 1. PAHs Surrogates with Monitored Ions and their Intensities

Quantitation Confirmation Relative


Surrogate ion ion intensity
d8 -Naphthalene 136 134 9
d8 -Acenaphthylene 160 158 15
d10 -Acenaphthene 164 162 94
d10 -Fluorene 176 174 90
d10 -Anthracene 188 184 12
d10 -Phenanthrene 188 184 14
d10 -Pyrene 212 208 17
d12 -Benz[a]anthracene 240 236 22
Downloaded By: [Sheffield Hallam University] At: 15:35 19 November 2008

d12 -Chrysene 240 236 26


d12 -Triphenylene 240 236 28
d12 -Benzo[a]pyrene 264 260 18
d12 -Benzo[b]fluoranthene 264 260 20
d12 -Benzo[e]pyrene 264 260 24
d12 -Perylene 264 260 24
d12 -Benzo[ghi]perylene 288 284 18
d12 -Indeno[1,2,3-cd]pyrene 288 284 19
d14 -Dibenz[a,h]anthracene 292 288 26
d10 -Fluoranthene (Internal 212
and Recovery Standard)

tests with confidence limits of 95% were estimated based on the meth-
ods of Miller and Miller (12). For the purposes of this study 17 deuter-
ated surrogates were chosen (Table 1) to represent 30 native analytes
(Table 2). A mean relative response factor of surrogates which are clos-
est in chemical properties (homologues) to the native analyte is used for
natives with no corresponding isotopic analogue (Table 2). Although 30
native PAHs were studied, to simplify the data presentation, quantitation
data presented herein shows only PAHs that are common to the three
studied matrices.

RESULTS AND DISCUSSION

The results of 71 PAH control samples consisting of particulate matter


in ambient air, soil and diesel emission are presented as a comparative
study demonstrating the impact of the recovery correction technique on
the accuracy of PAH quantitation as compared to the method without
recovery correction. The results indicate unambiguously that the isotope
dilution technique provides a more accurate determination of PAH levels
Downloaded By: [Sheffield Hallam University] At: 15:35 19 November 2008

TABLE 2. Analyzed PAHs with Monitored Ions and their Corresponding Surrogates

318
Native PAH Quantitation ion Confirmation ions Relative intensities Recovery correction surrogate
Naphthalene 128 127, 102 100, 12, 7 d8 -Naphthalene
Acenaphthylene 152 151, 150 100, 20, 13 d8 -Acenapthylene
Acenaphthene 153 154, 152 100, 96, 48 d10 -Acenaphthene
2-Methylfluorene 165 180, 178 100, 90, 27 AVG (d10 -Fluorene + d10 -Phenanthrene
+ d10 -Anthracene)
Fluorene 166 165, 163 100, 91, 15 d10 -Fluorene
Anthracene 178 176, 179 100, 18, 16 d10 -Anthracene
Phenanthrene 178 176, 179 100, 19, 16 d10 -Phenanthrene
Fluoranthene 202 200, 101 100, 21, 13 AVG (d10 -Pyrene + d12 -Benz[a]anthracene
+ d12 -Triphenylene + d12 -Chrysene)
Pyrene 202 200, 101 100, 21, 13 d10 -Pyrene
1-Methylpyrene 216 215, 108 100, 68, 10 AVG (d10 -Pyrene + d12 -Benz[a]anthracene
+ d12 -Triphenylene + d12 -Chrysene)
Benzo[a]fluorene 216 215, 108 100, 72, 12 AVG (d10 -Pyrene + d12 -Benz[a]anthracene
+ d12 -Triphenylene + d12 -Chrysene)
Benzo[b]fluorene 216 215, 108 100, 89, 13 AVG (d10 -Pyrene + d12 -Benz[a]anthracene
+ d12 -Triphenylene + d12 -Chrysene)
Retene 219 234, 204 100, 70, 28 AVG (d10 -Pyrene + d12 -Benz[a]anthracene
+ d12 -Triphenylene + d12 -Chrysene)
Benzo[ghi]fluoranthene 226 224, 113 100, 22, 15 AVG (d10 -Pyrene + d12 -Benz[a]anthracene
+ d12 -Triphenylene + d12 -Chrysene)
Benz[a]anthracene 228 226, 114 100, 27, 11 d12 -Benz[a]anthracene
Chrysene 228 226, 114 100, 29, 9 d12 -Chrysene
Triphenylene 228 226, 114 100, 28, 10 d12 -Triphenylene
7-Methylbenz[a]anthracene 242 241, 239 100, 40, 28 AVG (d10 -Pyrene + d12 -Benz[a]anthracene
+ d12 -Triphenylene + d12 -Chrysene)
Downloaded By: [Sheffield Hallam University] At: 15:35 19 November 2008

Benzo[a]pyrene 252 250, 126 100, 23, 14 d12 -Benzo[a]pyrene


Benzo[b]fluoranthene 252 250, 126 100, 22, 13 d12 -Benzo[b]fluoranthene
Benzo[e]pyrene 252 250, 126 100, 28, 12 d12 -Benzo[e]pyrene
Benzo[k]fluoranthene 252 250, 126 100, 22, 13 AVG (d12 -Benzo[b]fluoranthene +
d12 -Benzo[e]pyrene +
d12 -Benzo[a]pyrene + d12 -Perylene)
Perylene 252 250, 126 100, 27, 14 d12 -Perylene
3-Methylcholanthrene 268 252, 126 100, 39, 16 AVG (d12 -Benzo[b]fluoranthene +
d12 -Benzo[e]pyrene +
d12 -Benzo[a]pyrene + d12 -Perylene)
Anthanthrene 276 274, 138 100, 20, 22 AVG (d12 -Indeno[1,2,3cd]pyrene +
d14 -Dibenz[a,h]anthracene +
d12 -Benzo[ghi]perylene)
Benzo[ghi]perylene 276 274, 138 100, 22, 20 d12 -Benzo[ghi]perylene
Indeno[1,2,3-cd]fluoranthene 276 274, 138 100, 20, 17 AVG (d12 -Indeno[1,2,3cd]pyrene +
d14 -Dibenz[a,h]anthracene +
d12 -Benzo[ghi]perylene)
Indeno[1,2,3-cd]pyrene 276 274, 138 100, 20, 18 d12 -Indeno[1,2,3cd]pyrene
Benzo[b]chrysene 278 276, 139 100, 25, 15 AVG (d12 -Indeno[1,2,3cd]pyrene +
d14 -Dibenz[a,h]anthracene +
d12 -Benzo[ghi]perylene)
Dibenz[a,h]anthracene 278 276, 139 100, 28, 15 d14 -Dibenz[a,h]anthracene

319
320 A. Wnorowski et al.

as indicated by the more consistent distribution of PAHs observed in


the recovery-corrected data. A significant difference between the non-
recovery corrected and the recovery-corrected results is observed for
samples with low recoveries. Data depicted in Figure 1 shows matrix and
analyte dependent recoveries of used surrogates. Recoveries of 14 native
PAHs and their deuterated analogues calculated with d10 -Fluoranthene
are compared in Figure 2. In Figure 3 recovery accuracies of 19 native
PAHs from three matrices calculated from uncorrected data are compared
against data corrected by recovery of deuterated analogues.

Non-Recovery Corrected Quantitation (nRC)


Downloaded By: [Sheffield Hallam University] At: 15:35 19 November 2008

A) Surrogates Recoveries
Evaluation of surrogate recoveries spiked into control samples con-
firmed that the values were within the acceptable range of 50120%
(Figure 1). The data show that the analysis of ambient air samples pro-
vides results closest to the reference values, thus the highest analytical
recovery, probably due to the relatively inert matrix of the ambient air.
Soil and diesel matrices had on average a 10% lower accuracy compared
to air samples.
Furthermore, recoveries of the presented 17 PAH surrogates differed
not only among the three matrices, but also within each matrix. There
is an evident trend in recovery with regard to volatility of PAHs. As
expected, recovery is poorer for more volatile, low molecular weight

FIGURE 1. Matrix and molecular mass effect on recovery efficiency of PAH


surrogates.
PAHs Quantitation by Isotope Dilution GC-MS 321

compounds, such as naphthalene, acenaphthylene, acenaphthene, and


fluorene, regardless of the matrix. To the contrary, pyrene and heavier
mass PAHs are characterized in our study by higher and more uniform
recoveries, especially noticeable in soil samples. The most volatile ana-
lytes are usually most difficult to quantify accurately using uncorrected
technique. Variabilities in analytical recoveries reflect, besides matrix in-
fluence, differences in analytes chemical and physical properties. There-
fore, it seems that recovery correction, in which one surrogate recovery
is used to quantify more than one native, may be introducing a bias to
the quantitation process. A sound alternative to the traditional practice
of using one surrogate for the quantitation of more than one native, is to
Downloaded By: [Sheffield Hallam University] At: 15:35 19 November 2008

use the mean relative response factors of homologues when an analogous


surrogate is unavailable, as presented and discussed later in this study.

B) Comparison of the Uncorrected Recoveries of Natives and their


Analogous Surrogates
To evaluate similarity in recoveries of natives and corresponding sur-
rogates, environmental control samples (Standard Reference Material)
were spiked with surrogates. The natives present and the added surrogates
were consequently quantified with the uncorrected method using d10 -
Fluoranthene. The correlation between recoveries of certified/reference
values of the natives and corresponding surrogates, presented in Figure 2,
confirms above 0.9 recovery ratios for all matrices (shown in Figure 2
legends). However, ambient air and diesel emission surrogate recoveries
are minimally higher than corresponding native recoveries. This fact can
suggest that perhaps equilibrium had not been reached or natives may
be more strongly bound to these matrices than spiked surrogates, conse-
quently, the corrected results could underestimate the negative bias. In
addition, relative standard deviation (RSD) of mean recoveries is higher
for diesel and air samples, 12% and 14% respectively, suggesting lower
precision of extraction from these two matrices. Quantitation of studied
PAHs from soil samples, on the other hand, is characterized by more uni-
form recoveries and significantly lower mean RSD of 6%, both natives
and surrogates, with slightly lower recoveries of surrogates than natives.
ANOVA analysis of the uncorrected results was applied to further
confirm similarity between native and corresponding surrogate recov-
eries. The test comparing two experimental means (native and surro-
gate recoveries) for each matrix confirmed strong similarity between
native and surrogate results and no evidence of systematic error at P =
0.05. Consequently, it validates the similar behaviour of isotopes and
natives in the analytical process and the reasoning behind using selected
isotopically-labelled analogues as surrogates for the studied matrices.
322 A. Wnorowski et al.
Downloaded By: [Sheffield Hallam University] At: 15:35 19 November 2008

FIGURE 2. Uncorrected recovery accuracy of surrogate and native PAHs.


PAHs Quantitation by Isotope Dilution GC-MS 323
Downloaded By: [Sheffield Hallam University] At: 15:35 19 November 2008

FIGURE 3. Corrected and uncorrected accuracy of PAHs quantitation in


environmental samples.

Accuracy and precision of our in-house methods, comprising extrac-


tion and uncorrected/corrected analytical quantitation, were evaluated
and are presented in Table 3 and will be discussed in detail in the fol-
lowing section.
Recovery-Corrected Isotope Dilution Quantitation (RC)
In the isotope dilution technique, the quantitation of the analyte con-
centration is corrected by analogue isotope recovery. In theory, no quan-
titation bias should exist if an analogue isotope is used. However, it is
324 A. Wnorowski et al.

not always possible to acquire an analogous isotope, therefore a com-


mon practice is to use the relative response factor from a compound with
the most similar chemical properties and retention time, or a compound
within the same class with similar polarity and molecular weight (13, 14),
a decision often based on a subjective judgment. This approach intro-
duces additional uncertainty since response factors of various chemicals,
even within the same chemical group, vary. Therefore, recovery correc-
tion may contribute to the overall uncertainty of quantitation depending
on the difference between the recovery of analyte and the used surrogate.
As previously discussed and shown in Figure 1 each surrogate, although
within the same chemical group, is characterized by different recovery,
Downloaded By: [Sheffield Hallam University] At: 15:35 19 November 2008

that also changes depending on matrix. Therefore, in our methodology,


in order to minimize uncertainty associated with utilization of a surro-
gate with a potentially different recovery and response factor, we use the
mean response factor of several surrogates which are closest in chemical
properties to the native analyte for those natives with no corresponding
isotopic analogue (Table 2).
Figure 3, comparing non-recovery and recovery correction quanti-
tation techniques, illustrates an improvement in overall accuracy at-
tributed to recovery correction for the three environmental matrices. The
recovery-corrected quantitation of PAHs in ambient air and diesel emis-
sion control samples provided very similar results, within 15% of the
expected value (88 15% and 87 11% respectively). Recovery cor-
rection reduces the bias in diesel emission samples to a value similar
to that of ambient air, hence it could be concluded that isotope dilution
quantification results are unaffected by type of sample matrix. Uncor-
rected quantitation of PAHs provided lower values (85 16% for air and
79 11% for diesel emission). Diesel emission samples display lower
recoveries therefore greater negative bias reflecting higher losses during
analytical process. Recovery correction, however, did not improve pre-
cision of the quantitation. As shown at the bottom of Table 3, each of the
studied matrices showed very similar uncertainty values for corrected
and uncorrected data.
Similar to ambient air and diesel emission results, recovery-corrected
quantitation of PAHs in soil shows improved accuracy of native con-
centrations compared to uncorrected data. However, in this matrix, the
difference between the two means (recovery-corrected and uncorrected)
is much more pronounced than in air or diesel. It is assumed that certain
natives have high analytical losses and their corresponding surrogates
will also have high losses. Accounting for the calculated losses of spiked
isotope material, the recovery-corrected native concentration can reflect
more accurately the true value. Consequently, for samples with low
PAHs Quantitation by Isotope Dilution GC-MS 325

recoveries it is expected that there will be a significant difference be-


tween the uncorrected and the recovery-corrected results. As a matter of
fact, the biggest difference between uncorrected concentrations (nRC)
and corrected concentrations (RC) presented in Figure 3 are for the most
volatile compounds, anthracene, fluorene and phenanthrene. ANOVA
analysis confirmed that the accuracy of recovery-corrected results from
soil samples is significantly higher than the accuracy of uncorrected data
(P = 0.05).
The recovery-corrected results of the soil samples are higher than
the consensus values of the proficiency test samples (108 14%). It
should be clarified that the consensus values originate from the uncor-
Downloaded By: [Sheffield Hallam University] At: 15:35 19 November 2008

rected mean results of 14 laboratories which participated in the CAEAL


Performance Tests, which may comprise its own negative bias. As par-
ticipants of CAEAL Performance Test, all labs test their methods, those
whose results do not meet the consensus value ranges are rejected as
outliers and not used in the calculation of the CAEAL Performance Test
mean. This illustrates again that uncorrected data will typically under-
estimate the actual value, since the consensus value is not corrected for
the recovery of surrogate.
Since uncertainty represents an important quality of the analytical
results, it is imperative to support experimental data with uncertainty es-
timation. In Figure 3, visually depicted are the quantitation results for the
PAHs in three matrices, and correspondingly their numeric uncertainties
are summarized in Table 3. Presented at the bottom of Table 3, the mean
values for uncertainties are relatively close for the three studied matrices,
regardless of the quantitation technique. The highest mean precision, re-
flected as uncertainty, is associated with diesel emission analysis and
lowest with ambient air. It would seem possible that lower association
between analytes and matrix in ambient air precludes a good precision
of their extraction, even though the mean accuracy of quantitation in air
samples is higher than in the two other matrices. To that hypothesis, closer
investigation of individual uncertainties in ambient air samples confirms
that highest uncertainty values are associated with highly volatile PAHs,
mainly fluorene and anthracene. In addition, since generally PAHs con-
centrations are lowest in air, the quantitation reflects higher analytical
uncertainty in this matrix. On the other hand, the same low precision for
high masses benzo[ghi]perylene and dibenz[a,h]anthracene may also
suggest that extraction/analysis may still require an improvement.
The presented corrected results emphasize an existing bias (although
decreased, compared to uncorrected data) between PAH quantitation
results by isotope dilution mass spectrometry and the certified values
of SRM. This could be attributed to several factors, such as an error
Downloaded By: [Sheffield Hallam University] At: 15:35 19 November 2008

TABLE 3. Quantitation Assigned Value with Uncertainty Interval (95% CL) for PAHs in Ambient Air, Soil, and Diesel
Emission Control Samples

326
Ambient air Soil Diesel emission

Recovery Accuracy Uncertainty Accuracy Uncertainty Accuracy Uncertainty


Compound correction [%] [%] [%] [%] [%] [%]
Anthracene nRC 75 46 na na 74 26
RC 91 28 na na 99 18
Benz[a]anthracene nRC 96 22 83 38 84 28
RC 93 22 94 8 92 24
Benzo[a]pyrene nRC 92 26 96 14 59 16
RC 93 32 116 24 62 22
Benzo[b]chrysene nRC 73 28 na na 79 22
RC 68 42 na na 87 22
Benzo[b]fluoranthene nRC 98 18 92 22 95 26
RC 96 20 105 34 100 34
Benzo[e]pyrene nRC 93 26 na na 87 30
RC 94 24 na na 98 32
Benzo[ghi]fluoranthene nRC 74 38 na na 81 22
RC 79 34 na na 86 18
Benzo[ghi]perylene nRC 95 48 94 32 86 24
RC 90 50 105 36 100 20
Benzo[k]fluoranthene nRC 89 20 81 24 87 14
RC 89 18 97 28 83 6
Chrysene nRC 86 28 na na 73 22
RC 95 34 na na 75 24
Dibenz[a,h]anthracene nRC 104 52 na na 87 14
RC 92 50 na na 83 6
Downloaded By: [Sheffield Hallam University] At: 15:35 19 November 2008

Fluoranthene nRC 76 20 100 16 75 14


RC 81 16 116 16 81 20
Fluorene nRC 67 60 na na na na
RC 89 50 na na na na
Indeno[1,2,3-cd]fluoranthene nRC 88 22 na na 82 26
RC 84 18 na na 92 28
Indeno[1,2,3-cd]pyrene nRC 95 28 99 28 82 32
RC 83 32 116 26 96 24
Perylene nRC 87 28 na na 79 22
RC 91 38 na na 87 22
Phenanthrene nRC 70 44 92 24 61 24
RC 85 26 112 22 81 20
Pyrene nRC 82 16 100 18 82 24
RC 88 16 115 16 91 26
Triphenylene nRC 84 26 na na 73 18
RC 91 32 na na 78 20
Mean nRC 85 31 93 24 79 22
RC 88 31 108 23 87 21

Expanded Uncertainty = % RSD x k; k is coverage factor chosen from t distribution table at P = 0.05.
naCertified/Reference values not available in SRM.

327
328 A. Wnorowski et al.

introduced by the uncertainty of the used standard materials or by the


weighing procedure, difference in calibration standards, imperfection in
manual integration of small chromatographic peaks at low concentra-
tion, or incomplete isotopic equilibrium of spikes resulting in stronger
bounding of natives to matrix than isotopes. Nonetheless, since the data
were generated in a two year time span by various GC-MS instruments,
using different standards and conducted by different analysts follow-
ing the same extraction/analytical methodology, additional uncertainty
was introduced. Variation of analytical conditions justifies considering
above data as interlaboratory generated. Consequently, it strengthens the
validity of the study.
Downloaded By: [Sheffield Hallam University] At: 15:35 19 November 2008

CONCLUSIONS

Recovery correction by isotope dilution of environmental PAHs was


tested against traditional uncorrected quantitation to compare the ac-
curacy of both approaches. The differences in recoveries and relative
response factors among the studied PAHs would strongly recommend
using of analogue surrogates for quantitation. Statistical data treatment
and recovery ratios demonstrated a similar recovery of natives to their
analogous deuterated surrogates. This confirmed their resemblance in
the analytical behaviour and the suitability of using these surrogates for
recovery-corrected quantitation. Application of isotope dilution tech-
nique for PAH quantitation enhanced the accuracy of results compared
to the uncorrected method for all matrices studied, but showed a greater
influence as the negative bias increased. Although the uncertainty values
are highest for the most volatile compounds, the application of recovery
correction improves their quantitation accuracy significantly.
Recovery-corrected compensation for analyte losses during analyti-
cal processing provides an unbiased estimate of native concentrations in
matrix. Therefore, the quantitation benefits from monitoring losses as-
sociated with volatilization, incomplete extraction or instrumental bias.
This technique provides improved results whenever losses of analytes
occur during the analytical process. For instance, it would allow for bet-
ter comparability of inter-laboratory results if all participants used this
method. It will also play an important role in method performance and
validation studies, or gaining the ability to assess the efficacy of the an-
alytical process in general. Another possible advantage emerging from
using isotope dilution is a compensation for the suppression of analyte
response. Taking into consideration all potential advantages, recovery
correction by isotope dilution should therefore be considered implicit in
analytical analysis to minimize the impact of bias in quantitation.
PAHs Quantitation by Isotope Dilution GC-MS 329

REFERENCES

1. T. Dann. Ambient Air Measurements of Polycyclic Aromatic Hydrocarbons (PAH),


Polychlorinated Dibenzo-p-Dioxins (PCDD), and Polychlorinated Dibenzofurans
in Canada (1987-1997), Environment Canada Report AAQD 97-3, Ottawa, ON,
Canada, 1997.
2. C. H. Chiu, Y. Y. Shu, T. Dann, L. Graham, G. Poole, M. Tardif, and R. Turle, Mea-
surement of PAH and Nitro-PAH in Ambient Air Samples from Selected Canadian
Sites. ISPAC19 Conference, 2125 September, Amsterdam, Netherlands (2003).
3. W. Koolstra, B. G. Wolthers, M. Hayer, H. Elzinga, Development of a Reference
Method for Determining Urinary Oxalate by Means of Isotope Dilution-Mass Spec-
trometry (ID-MS) and its Usefulness in Testing Existing Assays for Urinary Oxalate,
Downloaded By: [Sheffield Hallam University] At: 15:35 19 November 2008

Clin. Chim. Act. 170(23) (1997):227235.


4. C. von Holst, J. Stroka, and E. Anklam, Correction of Analytical Results for Re-
covery: a Comparison of the Method Performance Characteristics from Recent
Collaborative Trials Studies for Aflatoxin Quantification Using Conventional and
Robust Statistics, Food Add. and Cont. 19(8) (2002):701708.
5. B. N. Colby, and A. E. Rosecrance, Determination of Priority Pollutants in Industrial
Wastewaters by Stable-Isotope Dilution Gas Chromatography/Mass Spectrometry.
In Advances in the Identification and Analysis of Organic Pollutants in Water Vol
1. L.H. Keith (ed). Ann Arbor: Ann Arbor Science Publishers (1991), pp. 221230.
6. T. Hartman, Methodologies for the Quantification of Purge and Trap Thermal Des-
orption Analysis. Application Note No.781309. Ringoes: Scientific Instrument Ser-
vices Publications, 1999.
7. M. Thompson, S. L. R. Ellison, S. Fajgelj, P. Willetts, and R. Wood, Harmonised
Guidelines for the Use of Recovery Information in Analytical Measurement, Pure
and App. Chem. 71 (1999):337348.
8. L. L. Jessome, and D. A. Volmer, Ion suppression: A Major Concern in Mass
Spectrometry, LCGC North America 24(5) (2006):498510.
9. M. Sargent, R. Harte, and C. Harrington (eds). Guidelines for Achieving High Ac-
curacy in Inorganic and Organic Analysis by Isotope Dilution Mass Spectrometry.
London: RSC Special Publication (2002).
10. Analytical Method for the Determination of PAHs in Ambient Air and Diesel Emis-
sion Samples. Environment Canada AAQD/ETC Method 3.3/4.2/M. Ottawa: Envi-
ronment Canada AAQD/ETC Methods (2004).
11. Analytical Method for the Determination of PAHs in Soils/Sediments. Environment
Canada AAQD/ETC Method 3.8/2.3/M. Ottawa: Environment Canada AAQD/ETC
Methods (2003).
12. J. N. Miller, and J. C. Miller (eds). Statistics and Chemometrics for Analytical
Chemistry. 4th ed. Essex: Pearson Education Limited (2000).
13. M. A. Mazurek, B. R. T. Simoneit, G. R. Cass, and H.A. Gray, Quantitative High-
Resolution Gas Chromatography and High-resolution Gas Chromatography/Mass
Spectrometry Analyses of Carbonaceous Fine Aerosol Particles, Int. J. Env. Anal.
Chem. 29(1987):119139.
14. J. D. Jeremiason, J. Eisenreich, and M. J. Paterson, Accumulation and Recycling
of PCBs and PAHs in Aritificially Eutrophied Lake 227, Can. J. Fish. Aq. Sci.
56(1999):650660.

You might also like