Determination of methoxyphenols in ambient atmospheric particulate matter: Tracers for wood combustion

Article abstract of DOI:10.1021/es0486871

Combustion of wood and other biomass fuels produces source-specific organic compounds arising from pyrolysis of lignin, including substantial amounts of 4-substituted methoxylated phenolic compounds (methoxyphenols). These compounds have been used as atmospheric markers to determine the contribution of wood smoke to ambient atmospheric fine particulate matter (PM). However, reliable quantification of methoxyphenols represents an analytical challenge because these compounds are polar, semivolatile, and somewhat reactive. We report herein an improved gas chromatographic-mass spectrometric (GC/MS) method for the sensitive and reliable determination of methoxyphenols in low-volume ambient PM samples. Deuterated standard compounds are added to the environmental samples prior to extraction to determine analyte recoveries in each sample. Analytical figures of merit for the assay, as applied to ambient PM_2.5 and PM₁₀ samples are as follows: recovery = 63-100%; precision = 2-6%; analytical limit of detection (S/N 2) = 0.002 μg/mL; limit of quantitation = 0.07-0.45 ng/m³ (assuming a 14 m³ sample). The improved method was applied to ambient PM samples collected between 1999 and 2000 in Seattle, WA. Particle-bound methoxyphenol concentrations in the range <0.1 to 22 ng/m³ were observed and the methoxyphenols were present almost exclusively in the fine (PM_2.5) size fraction. We also demonstrated that XRF analysis of samples of atmospheric PM collected on Teflon filters significantly reduced the levels of methoxyphenols measured in the PM samples in subsequent assay of the same filters. Therefore, XRF analysis of filters, commonly undertaken to obtain trace element concentrations for use in source apportionment analyses, would preclude the subsequent analysis of those filters for methoxyphenols and other similarly semivolatile or reactive organic chemicals.

Full text of DOI:10.1021/es0486871

Environ. Sci. Technol. 2005, 39, 631-637
responsible for occasional severe episodes of air pollution
Determination of Methoxyphenols in
Ambient Atmospheric Particulate
Matter: Tracers for Wood
Combustion
(4-7). In the developing world, biomass fuels are still the
major source of energy for cooking and space heating, and
indoor exposures of up to several mg/m3 have been reported
(4, 8, 9).
Several toxic or mutagenic chemicals are present in wood
smoke, including polycyclic aromatic hydrocarbons, alde-
hydes, and free radicals (10-12). Several epidemiological
studies have reported associations between wood smoke
exposure and adverse health effects including eye, nose, and
throat irritation, decrements in lung function, and increased
respiratory infections (4, 8, 9, 13). Children with asthma
appear to be particularly susceptible to the effects of wood
smoke; reported symptoms included increased respiratory
symptoms, lower respiratory infections, and decreased
pulmonary function (13).
C H R I S T O P H E R D . S I M P S O N , *
M I C H A E L P A U L S E N , R U S S E L L L . D I L L S ,
L . - J . S A L L Y L I U , A N D D A V I D A . K A L M A N
Department of Environmental and Occupational Health
Sciences, School of Public Health and Community Medicine,
University of Washington, Box 357234,
Seattle, Washington 98195-7234
A wealth of recent publications have reported on the
chemical composition of wood smoke (12, 14-16) from
source samples. Of the hundreds of organic compounds
present in wood smoke, 4-substituted methoxylated phenolic
compounds (methoxyphenols) and levoglucosan (a sugar
anhydride) have been suggested as potential molecular
markers (3, 17). Methoxyphenols are abundant in wood
smoke, their presence in atmospheric PM is unique to
biomass combustion, and they are relatively stable tracers.
Reliable molecular markers for wood smoke would facilitate
source apportionment studies of atmospheric PM and would
be useful for studying exposure-health effect relationships
for wood smoke. Emission rates for total methoxyphenols
were in the range 900-4200 mg/kg fuel, and the compounds
were present both in the vapor phase and bound to particles
(12, 14, 15).
In contrast, relatively few reports exist describing meth-
oxyphenol levels in ambient PM samples. Hawthorne et al.
reported total methoxyphenol levels in the range 354-3510
ng/m3 (median ) 907 ng/m3) in ambient samples collected
in wintertime 1988-1989 in Minneapolis, MN, and Salt Lake
City, UT (17). They used a high-volume PM10 sampler
containing a quartz filter backed with polyurethane foam
cartridges; thus, their measurements represent a sum of both
particle-bound and vapor-phase methoxyphenols. Schauer
et al. described 1995 wintertime methoxyphenol levels at
three locations in the San Joaquin Valley, CA: Fresno,
Bakersfield, and the Kern Wildlife Refuge. The sum of
methoxyphenols ranged from 0.4 to 876 ng/m3 (median )
97 ng/m3) (3), with the highest concentrations occurring
during a wood smoke-dominated, relatively severe air
pollution event in Fresno (fine PM levels ) 55 µg/m3). The
samples were collected on quartz filters using a high-volume
(10 m3/h) dichotomous virtual impactor and represent only
the particle-bound methoxyphenols.
Previous measurements of methoxyphenols in wood
smoke and ambient PM have used GC/MS methodology (3,
14, 15, 18, 19). However, we discovered several important
analytical issues related to these methods. Many of these
studies did not employ authentic standard compounds to
calibrate the GC retention times and the MS detector
response, and stable-isotope-labeled compounds were not
used to validate analyte recoveries. In seeking to validate a
sensitive and reliable analytical method for the determination
of particle-associated methoxyphenols in ambient PM
samples, we encountered unexpected problems including
chemical reactivity and volatilization of the methoxyphenols.
We describe herein an improved quantitative GC/MS assay
for the determination of particle-bound methoxyphenols in
low-volume ambient PM samples, in which multiple iso-
Combustion of wood and other biomass fuels produces
source-specific organic compounds arising from pyrolysis
of lignin, including substantial amounts of 4-substituted
methoxylated phenolic compounds (methoxyphenols). These
compounds have been used as atmospheric markers to
determine the contribution of wood smoke to ambient
atmospheric fine particulate matter (PM). However, reliable
quantification of methoxyphenols represents an analytical
challenge because these compounds are polar, semi-
volatile, and somewhat reactive. We report herein an improved
gas chromatographic-mass spectrometric (GC/MS)
method for the sensitive and reliable determination of
methoxyphenols in low-volume ambient PM samples.
Deuterated standard compounds are added to the
environmental samples prior to extraction to determine
analyte recoveries in each sample. Analytical figures of merit
for the assay, as applied to ambient PM2.5 and PM10
samples are as follows: recovery ) 63-100%; precision
) 2-6%; analytical limit of detection (S/N 2) ) 0.002 µg/
mL; limit of quantitation ) 0.07-0.45 ng/m3 (assuming a 14
m3 sample). The improved method was applied to ambient
PM samples collected between 1999 and 2000 in Seattle,
WA. Particle-bound methoxyphenol concentrations in the
range <0.1 to 22 ng/m3 were observed and the methoxyphenols
were present almost exclusively in the fine (PM2.5) size
fraction. We also demonstrated that XRF analysis of samples
of atmospheric PM collected on Teflon filters significantly
reduced the levels of methoxyphenols measured in the
PM samples in subsequent assay of the same filters. Therefore,
XRF analysis of filters, commonly undertaken to obtain
trace element concentrations for use in source apportionment
analyses, would preclude the subsequent analysis of
those filters for methoxyphenols and other similarly
semivolatile or reactive organic chemicals.
Introduction
Wood smoke emissions from residential fireplaces are a major
source of atmospheric fine particulate matter (PM) in several
communities in the United States (1-3). In addition, wood
smoke emissions from wildfires and prescribed burns are
* Corresponding author phone: (206)543-3222; fax: (206)616-2687;
e-mail: simpson1@u.washington.edu.
9
10.1021/es0486871 CCC: $30.25
Published on Web 12/15/2004
2005 American Chemical Society
VOL. 39, NO. 2, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 631

a silica column (dichloromethane). The syringones were
recrystallized from methanol-water. Butylsyringone: purity
93.5%; mass spectra: m/z (%)181 (100), 224 (24), 153 (11),
196 (8); mp 80-82°. Propylsyringone: purity 99.1%; mass
spectra: m/z (%) 181 (100), 210 (28), 153 (18), 123 (5); mp
113-115°.
Acetates of each analyte were synthesized using the
procedure of Chau and Coburn (23) with some modification;
three to four washes with 5% HCl were necessary to remove
pyridine from the product. Column chromatography (silica
with dichloromethane) was used to purify acetates, whose
purity after isolation was >97%.
TABLE 1. Purity and Deuteration Conditions for Synthesis of
Recovery Standards
deuteration deuteration chemical isotopic
reaction
temp (°C)
reaction
time (h)
purity
(%)
purity
(%)
chemical
d6-acetovanillone
d2-4-methylsyringol
d2-4-ethylsyringol
d2-4-propylsyringol
150
150
150
150
14.9
9.9
9.9
6
100
91
99
51a
97b
91b
92b
97
a 40% d7 and 9% d5. b Balance was d1.
Sinapylaldehyde propionate was synthesized analogously
to the acetates but using propionic anhydride. The compound
was recrystallized from toluene (mp 132°; purity 99.6% trans,
0.4% cis; mass spectra: m/z (%) 208 (100), 177 (38), 180 (32),
165 (27), 137 (14)).
topically labeled surrogate compounds are used to monitor
analyte recovery.
Experimental Section
Chemicals. 2,3-Dimethoxyphenol, acetovanillone, guaiacyl-
acetone (4-hydroxy-3-methoxyphenylacetone), coniferyl-
aldehyde (4-hydroxy-3-methoxycinnamaldehyde), sinapyl-
aldehdye (trans-3,5-dimethoxy-4-hydroxycinnamaldehdye),
acetosyringone (3′,5′-dimethoxy-4′-hydroxyacetophenone),
trimethoxybenzene, 2-chloro-4-methoxyphenol, and 2,6-
dimethoxyphenol were purchased from Aldrich (Milwaukee,
WI). 4-Ethylguaiacol and 4-propylguaiacol were obtained
from Lancaster Synthesis (Windham, NH). Ethyl acetate (ACS
grade) and triethylamine (HPLC grade) were purchased from
Fisher Scientific.
Chemical Synthesis. Eight deuterated methoxyphenols
were prepared as described in our previous study (20). Four
additional deuterated compounds, listed in Table 1, were
prepared from the proteo-analogue by deuterium exchange
(D2O) under acid conditions (DCl and CrCl2) (20). Ring
protons were exchanged during the reaction except for
acetovanillone, which appears to have also exchanged side
chain protons. Hydroxyl deuterium was exchanged back to
hydrogen during workup. Compounds were purified by silica
gel chromatography using dichloromethane as eluent. Table
1 gives reaction conditions and purities. Isotopic purity was
determined from the isotope ratios of the molecular ion
cluster. Chemical purity was assessed by gas chromatography
with flame ionization detection.
Hydrogenation (5% palladium on charcoal) of syringal-
dehyde and acetosyringone gave 4-methylsyringol (mp 39.5-
41°, lit. mp 40.5-41.5; purity 99.2%) and 4-ethylsyringol
(purity 91.7%), respectively (21). 4-Allylsyringol was synthe-
sized from syringol and allyl bromide via the ether, 1-allyl-
syringol (21), and was purified by fractional vacuum distil-
lation followed by recrystallization from methanol (mp 28-
31°; purity 99.1%). 4-n-Propylsyringol was made by hydro-
genation of 4-allylsyringol (21) and purified by vacuum
distillation (purity 99.9%).
Propionylsyringol (propylsyringone; 1-(4-hydroxy-3,5-
dimethoxyphenyl)-propan-1-one) and butyrylsyringol (bu-
tylsyringone; 1-(4-hydroxy-3,5-dimethoxyphenyl)-butan-1-
one) were synthesized by condensing syringol and the
corresponding carboxylic acid in the presence of BF3 (22).
Syringol (32.4 mmol) and the acid (propionic or butyric; 25.9
mmol) were heated with stirring to 65-85 °C under nitrogen.
BF3 was slowly bubbled through the solution to saturate it.
The black-brown solution was stirred for 1.5 h, removed from
the heat, and kept at room temperature overnight. Equal
volumes of diethyl ether and water (50 mL) were added. The
ether layer was separated, washed twice with 100 mL water,
twice with 100 mL 5% NaHCO3, and finally three times with
80 mL 5% NaOH. The NaOH solution was made acidic with
concentrated HCl and extracted three times with 200 mL
dichloromethane. The dichloromethane was dried with
sodium sulfate and then evaporated at reduced pressure.
The residue (syringol and the syringone) was separated on
Aerosol Collections. This study is part of the larger Seattle,
WA, exposure and health effect panel study (24). This study
used 24-h integrated samples collected from subjects’
residences and one centrally located outdoor site between
Oct 25, 1999 and January 2001. PM10 and PM2.5 samples were
collected over 24 h (4 PM-4 PM) using the single-stage inertial
Harvard Impactors (HI) (Air Diagnostics and Engineering,
Inc., Naples, ME) at 10 LPM and 37-mm Teflon filters (25).
One pair of HI2.5 and HI10 was located in the main activity
room of each subject’s home, while a second pair was located
outside the home concurrently. The limit of detection and
precision for the 24-h integrated HI is 1 and 1.2 µg/m3,
respectively (24). Two particulate matter reference materials,
SRM 1649a (urban dust) and SRM 2975 (diesel PM), available
from NIST (Gaithersburg, MD), were also used in method
development studies.
Sample Preparation. All filters were equilibrated at 35%
relative humidity and 22 °C for at least 24 h prior to weighing
(26). All filter weights were measured in either duplicate or
triplicate using a seven-place electronic ultramicrobalance
(Mettler Toledo, Model UMT2, Greifensee, Switzerland). After
weighing, filters were stored at -4 °C until further chemical
analysis. All glassware coming in contact with the sample or
sample extracts was silylated with hexamethyldisilazane prior
to use by a vapor-phase method (27). Prior to extraction, the
polyolefin support ring must be separated from the Teflon
filter membrane. This is necessary because chemical con-
taminants are present in the support ring that interfere with
subsequent GC/MS analyses. The Teflon filter membranes
were separated from the polyolefin support ring by using a
custom-built Teflon/stainless steel cutter, described in detail
in an earlier report (28). The filter was placed in a 40-mL
glass headspace vial and deuterated recovery standards were
spiked onto the filter (∼40 ng/compound in 20 µL ethyl
acetate). The vial was sealed and the spike allowed to age for
1 h. Extraction solvent (30 mL ethyl acetate containing 3.6
mM triethylamine) was injected into the headspace vial
through a Teflon-faced silicone septa, and the samples were
then placed in an ultrasonic bath at ambient temperature
for 1 h. After sonication, the solvent was decanted into 50-
mL Turbovap tubes and was reduced in volume under
nitrogen to approximately 0.5 mL at 45 °C and atmospheric
pressure. Extracts were filtered through 0.45-µm PTFE syringe
filters, into amber glass autosampler vials, and internal
standard mix (2,3-dimethoxyphenol, trimethoxybenzene, and
2-chloro-4-methoxyphenol, 0.32 µg in 20 µL ethyl acetate)
was added.
Samples of the PM reference materials were prepared
and extracted as described for the filter samples, with the
following exceptions. An aliquot of the PM reference material
was weighed into a centrifuge tube. After the PM sample had
been extracted, it was centrifuged at 1800g for 30 min, and
the supernatant was decanted into the Turbovap tubes.
9
632 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 2, 2005

quantitation for each analyte as the greater of the average
concentration of analyte detected in blank filter extracts, or
the analytical limit of detection. Μethoxyphenols in envi-
ronmental samples were identified by the presence of a
chromatographic peak with the expected m/z ratio and
comparison of retention time relative to the appropriate
labeled standard compound. The relative retention time data
were obtained from the calibrants. This approach accounted
for systematic shifts in absolute retention time between
samples and calibrant solutions and facilitated more reliable
analyte identification in the environmental samples.
Selected filter extracts were also analyzed for levoglucosan,
as described previously (28). The levoglucosan was converted
to a pertrimethylsilyl derivative, followed by GC/MS analysis.
TABLE 2. Methoxyphenol Analytes, Quantitation Ions, and
Retention Times
compd compd
retention quantitation
no.
typea
compd name
time (min)b
ion
1
2
3
4
5
6
7
8
9
ISTD
trimethoxybenzene
7.43
6.92
6.93
7.91
7.91
8.48
8.68
8.68
8.71
8.90
8.91
9.34
9.34
9.41
9.42
9.44
9.44
168
113
109
126
123
154
140
137
143
157
154
167
164
140
137
154
151
164
170
168
164
154
151
169
167
137
194
169
167
184
182
181
178
181
181
208
208
rec. std d4-guaiacol
analyte guaiacol
rec. std d3-methylguaiacol
analyte methylguaiacol
ISTD
rec. std d3-ethylguaiacol
analyte ethylguaiacol
ISTD
2,3-dimethoxyphenol
2-chloro-4-methoxyphenol
10 rec. std d3-syringol
11 analyte syringol
12 rec. std d3-eugenol
13 analyte eugenol
14 rec. std d3-propylguaiacol
15 analyte propylguaiacol
16 rec. std d2-vanillin
Results and Discussion
Initially, we evaluated recoveries of a mixture of deuterium-
labeled methoxyphenols in ethyl acetate, amended onto
Teflon filter membranes loaded with PM2.5 (see Table 2 for
list of deuterated compounds). The solvent was allowed to
evaporate for 1 h, and then the filter was extracted with 30
mL methylene chloride/acetone (1:1) in an ultrasonic bath.
The extracts were analyzed by GC/MS-SIM after reducing
the extract volume to 0.5 mL. Recoveries of all compounds
except syringaldehyde were low (median 7%, range 2-45%)
because of evaporative losses. Therefore, in all subsequent
experiments, the recovery standards were amended onto the
filters in glass headspace vials, which were immediately
sealed.
The ability of several solvents and solvent mixtures to
extract methoxyphenols from clean Teflon filter membranes
was evaluated. Deuterium-labeled methoxyphenols in ethyl
acetate were amended onto clean Teflon filter membranes
and the solvent was allowed to evaporate in glass headspace
vials for 1 h, before extraction and analysis. Good recoveries
for all compounds were obtained when ethyl acetate was
used as the solvent, (median 97%, range 92-111%) but
recoveries for the alkyl-substituted syringols were poor when
acetone (median 13%, range 8-15%) or methylene chloride:
acetone (1:1) (median 31%, range 23-33%) was used as the
solvent.
17 analyte vanillin
18 analyte cis-isoeugenol
19 rec. std d2-methylsyringol
20 analyte methylsyringol
21 analyte trans-isoeugenol
22 rec. std d6-acetovanillone
23 analyte acetovanillone
24 rec. std d2-ethylsyringol
25 analyte ethylsyringol
26 analyte guaiacylacetone
27 analyte allylsyringol
28 rec. std d2-propylsyringol
29 analyte propylsyringol
30 rec. std d2-syringaldehyde
31 analyte syringaldehyde
32 analyte acetosyringone
33 analyte coniferylaldehyde
34 analyte propionylsyringol
35 analyte butyrylsyringol
9.70
9.70
9.70
10.13
10.23
10.26
10.30
10.30
10.76
10.86
10.91
10.91
11.00
11.00
11.63
12.06
12.27
12.81
13.31
36 ISTD
sinapylaldehyde
37 analyte sinapylaldehyde propionate 13.90
a ISTD, internal standard; rec. std, recovery standard. b These data
pertain to the acetate derivatives of the methoxyphenols.
We compared assay performance for GC/MS analysis of
the methoxyphenols as free phenols and after conversion to
their acetate derivatives. To form the acetate derivatives, the
extracts were treated with 50 µL of a 4:3 solution of acetic
anhydride in triethylamine, the vial was capped, and then
it was heated at 70 °C for 3 h.
Organic Analysis. The extracts were analyzed for meth-
oxyphenols by using GC/MS with selected ion monitoring
(SIM). The GC/MS system consisted of an HP-7673 au-
tosampler, HP5890 series II gas chromatograph, and HP
5989A MS engine mass spectrometer (Agilent Technologies,
Wilmington, DE). The sample (2 µL) was injected splitless
into the injector at 250 °C. The column (RTX-5 Sil-MS, Restek,
Bellefonte, PA; 30 m, 0.25 mm i.d., 0.25-µm film) was held
at 55 °C for 1 min, then ramped to 280 °C at 15 °C/min.
Helium was used as the carrier gas, at a constant pressure
of 18 psi. The transfer line was maintained at 280 °C, the
mass spectrometer source at 280 °C, and the manifold at 110
°C. The mass spectrometer was operated in EI mode, with
an ionization energy of 70 eV.
The major fragment ions listed in Table 2 were used for
quantification. To enhance sensitivity, these ions were
acquired in six time-programmed acquisition windows, with
a maximum of eight specific ions per window. The standard
curve for the methoxyphenols was linear between 0.01 µg/
mL and 2.0 µg/mL. The analytical limit of detection (S/N )
2) was 0.002 µg/mL. Low and variable amounts of some
analytes were detected in blank filter extracts, in the range
0.002-0.013 µg/mL. Therefore, we defined the limit of
On the basis of the above observations, ethyl acetate was
selected as a solvent of choice for extraction of methoxy-
phenols from PM. Ten outdoor PM2.5 samples were analyzed
for methoxyphenols. Deuterium-labeled compounds were
amended onto the filters prior to analysis. Recoveries of the
R-carbonyl substituted methoxyphenols were greater than
100% (median 202%, range 118-302%). The high recovery
values may be due to formation of these compounds from
the breakdown of other methoxyphenols, as the recoveries
of most of the other methoxyphenols were low and variable
(median 14%, range 0-162%). In an attempt to identify causes
of the variable recoveries, we further analyzed the extracts
with the mass spectrometer operated in scan mode. We
identified substantial amounts of o- and p-nitrophenol and
a putatative NO2-adduct of the 2-chloro-4-methoxyphenol
ISTD in many of these extracts. The levels of both of the NO2-
derivatives detected in the extracts were inversely correlated
with the recoveries of the guaiacol-type and syringol-type
deuterated methoxyphenols (Figure 1 and Table 3), whereas
recoveries of the R-carbonyl-substituted compounds showed
a positive association with levels of the nitrophenol derivatives
(r ) 0.67, Table 3). The syringol-type compounds were more
readily degraded than the guaiacol-type compounds. These
data imply that some PM components are capable of nitrating
the methoxyphenols under relatively mild conditions.
We also investigated the recovery of methoxyphenols from
PM standard reference materials. Deuterium-labeled meth-
oxyphenols in ethyl acetate were amended onto urban dust
PM samples (SRM1649a) and the solvent was allowed to
9
VOL. 39, NO. 2, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 633

via a ketol (29). Dilute nitric acid readily nitrates meth-
ylguaiacol and produces catechols, benzoquinone, and oxalic
acid through oxidation (30). Products from incomplete
oxidation, vanillin and syringaldehyde, were observed and
likely came from oxidation of the alkene-substituted guaiacols
and syringols: isoeugenol, eugenol, coniferylaldehyde, and
sinapylaldehyde. The same products were found after nitric
acid oxidation of wood (31); lignin in wood contains alkenyl
polymers of guaiacol and syringol. Our instrumental analysis
conditions would not efficiently detect carboxylic acids (e.g.,
vanillic acid). The relative reactivity of the alkene substituents
can be expected to increase with conjugation and explains
the rapid oxidation of R,â-unsaturated aldehydes (e.g.,
coniferylaldehyde), an observation also noted in wood lignin
oxidation by nitric acid (32). Also, the rate of degradation
was fastest for syringol and its alkyl-substituted analogues,
intermediate for guaiacol and alkyl-substituted guaiacols,
and slowest for R-carbonyl substituted compounds. This
order of reactivity is consistent with the results we observed
for the ambient PM samples and can be rationalized by the
electronic effects of the ring substituents on ring substitution
reactions: methoxy substituents moderately activate arene
substitution reactions, the effect being greater when two
methoxy groups are present ortho and para to the unsub-
stituted ring positions, as in the syringols. In contrast,
R-carbonyls are strongly deactivating.
On the basis of our understanding of the chemistry
described above, we sought to improve the recovery of the
methoxyphenols by interfering with the chemical reactions
leading to their breakdown. Therefore, we evaluated the
ability of several solvent additives to inhibit breakdown of
the methoxyphenols. Compounds evaluated included acids,
bases, antioxidants, and radical scavengers. Triethylamine
(3.6 mM in ethyl acetate) was effective at preventing
degradation of the methoxyphenols, and it did not interfere
with subsequent GC/MS analysis. Using this modified
extraction solvent, good recoveries for deuterium-labeled
methoxyphenols amended onto SRM1649a were obtained,
(median 95% range 65-96%).
We compared GC/MS analysis of methoxyphenols both
as free phenols and as the acetate derivatives. We found that
acetylation of the phenols improved chromatographic peak
shape, increased sensitivity, and reduced the frequency with
which the front end of the GC column and the injector had
to be maintained. However, the acetylation reaction was
relatively slow for the syringol-type compounds and required
heating at 70 °C for ∼3 h for complete reaction. Typical ion
chromatograms for our standard mixture and a representative
(outdoor) environmental sample are illustrated in Figure 2.
The improved procedure for determination of methoxy-
phenols in ambient PM was evaluated when applied to
samples of PM collected from an urban air monitoring site
(Beacon Hill, Seattle, WA). Results are presented in Table 4
along with recovery data and limits of quantitation. The limit
of quantitation, defined for each analyte as the greater of the
average concentration of analyte detected in blank filter
extracts or the analytical limit of detection at s/n ) 2, was
calculated assuming a 14.4 m3 sample (10 L/min × 24 h).
The limits of quantitation for guaiacol, vanillin, and aceto-
vanillone were higher than the other compounds because of
the presence of variable levels of these analytes in blank
filter extracts. The precision of this assay, on the basis of
replicate measurements of the deuterated recovery standards
(at an effective concentration of ∼3 ng/m3) was 2-6%. The
data have not been adjusted on the basis of recovery of the
deuterated compounds, because deuterated surrogates were
not available for all analytes.
FIGURE 1. Association between methoxyphenol recovery and
presence of nitrophenol derivatives in extracts of ambient PM. (A)
o-nitrophenol; (B) NO2-adduct of 2-chloro-4-methoxyphenol.
TABLE 3. Correlations between Deuterated Methoxyphenol
Recoveries and Presence of Nitrophenol Derivatives in
Extracts of Ambient PM Samples
pearson r
avg recovery avg recovery avg recovery
of guaiacolsa of syringolsb of r-carbonylsc
o-nitrophenol
NO2-adduct of 2-chloro-
4-methoxyphenol
-0.74d
-0.37
-0.89e
-0.64d
0.363
0.674d
a Sum of d4-guaiacol, d3-4-methylguaiacol, d3-4-ethylguaiacol, d3-
eugenol, and d3-4-propylguaiacol. b Sum of d3-syringol, d2-4-methyl-
syringol, d2-4-ethylsyringol, and d2-4-propylsyringol. c Sum of d2-
vanillin, d6-acetovanillone, and d2-syringaldehyde. d p<0.05. e p<0.01.
evaporate for 1 h before extraction and analysis. As was
observed with the ambient fine PM samples, recoveries of
the R-carbonyl-substituted methoxyphenols were somewhat
elevated (median 104%, range 103-142%), whereas the
recoveries for the other methoxyphenols were poor (guai-
acols: median 27%, range 24-72%; syringols: median 3%,
range 2-6%). Similar results were obtained using the
following solvent mixtures: methylene chloride:acetone (1:
1), methylene chloride:ethyl acetate (1:1), methylene chloride:
methanol:hexane (1:1:1). In contrast, when deuterated
methoxyphenols were amended onto diesel PM samples
(SRM2975), recoveries of most of the deuterated compounds
from diesel PM were good (median 80%, range 73-130%),
except for the alkyl substituted syringols (median 14%, range
11-49%).
Together, these initial studies consistently showed that,
in the presence of PM, the methoxyphenols were susceptible
to chemical degradation, through nitration and oxidation,
during the extraction procedure. The specific components
of PM responsible for this degradation were not identified.
We were able to reproduce this degradation by treating ethyl
acetate solutions containing methoxyphenols with HNO3
(0.005%) or NO2 (50 mL, 40 ppm v/v). In these experiments,
we observed degradation products consistent with nitration
of the aromatic ring, oxidation, and ring opening. Oxidative
cleavage of alkenes by nitric acid produces carboxylic acids
Using the modified extraction solvent, acceptable recov-
eries were obtained for all the recovery standards, in all the
samples (median 74%, range 65-100%). Multiple methox-
9
634 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 2, 2005

TABLE 4. Analytical Performance and Methoxyphenol
Concentrations in Ambient PM2.5 (n ) 10) from Seattle, WA
recovery of
deuterated
limit of
median and
range
standards quantification
(% ( 1 SD)
compd
guaiacol
methylguaiacol
ethylguaiacol
syringol
eugenol
propylguaiacol
vanillin
methylsyringol
acetovanillone
ethylsyringol
allylsyringol
propylsyringol
syringaldehyde
acetosyringone
coniferylaldehyde
4-propylsyringone
4-butylsyringone
sinapylaldehyde
(ng/m3)
(ng/ m3)
65 ( 2
79 ( 4
70 ( 2
70 ( 3
73 ( 4
75 ( 3
63 ( 2
83 ( 5
79 ( 4
100 ( 6
0.45
0.06
0.07
0.08
0.13
0.09
0.25
0.07
0.17
0.03
0.08
0.08
0.08
0.07
0.10
0.07
0.07
0.12
0.64 (0.55-0.67)
0.06 (<0.06-0.07)
0.13 (<0.07-0.15)
0.13 (<0.08-0.14)
0.34 (0.32-0.43)
<0.09
0.96 (0.52-1.45)
0.23 (<0.07-0.35)
0.22 (<0.17-0.25)
0.27 (<0.13-0.32)
<0.08
89 ( 4
87 ( 5
0.19 (<0.08-0.23)
0.61 (0.08-1.52)
0.47 (<0.07-1.06)
0.42 (<0.10-1.16)
0.31 (<0.07-0.54)
<0.07
0.56 (<0.12-1.43)
regulations which restrict residential wood burning when
unfavorable atmospheric conditions exist. (1). The median
PM2.5 concentration for the ambient PM samples in Table 4
was 18 µg/m3 (range 7-23 µg/m3); Schauer’s samples were
collected in part during a relatively severe air pollution
episode in 1995 in Fresno, CA (PM2.5 levels up to 55 µg/m3)
(3). The methoxyphenols more volatile than syringaldehyde
exist predominantly in the vapor phase at ambient temper-
atures (3). Hence, the concentrations of lower molecular
weight methoxyphenols (e.g., guaiacol, syringol) observed
in the current study only account for the amount bound to
particles and retained on the filter. Our efforts to capture
these compounds on polyurethane foam (PUF) at 10 LPM
over 24 h, adapted from Hawthorne et al. (14, 17), were not
successful. Breakthrough of guaiacol and syringol with the
PUF sorbent occurred at very low sample volumes at ambient
temperature (<15 m3 for a 30-mL PUF cartridge when
sampled for 24 h at 10 L/min.). We are continuing research
toward identifying a sorbent with appropriate higher affinity
for the vapor-phase methoxyphenols.
FIGURE 2. GC/MS selected ion chromatograms for (1) methoxyphenol
standard mixture, (2) extract of wood smoke source sample from
open fire, and (3) extract of ambient PM2.5 sample. Compounds are
acetates unless otherwise stated. Annotation key: (A) guaiacol,
(B) 4-methylguaiacol, (C) 2,3-dimethoxyphenol, (D) 4-ethylguaiacol,
(E) 2-chloro-4-methoxyphenol, (F) syringol, (G) eugenol, (H) 4-pro-
pylguaiacol, (I) vanillin, (J) cis-isoeugenol, (K) 4-methylsyringol, (L)
trans-isoeugenol, (M) acetovanillone, (N) 4-ethylsyringol, (O) guai-
acylacetone, (P) 4-allylsyringol, (Q) 4-propylsyringol, (R) syringal-
dehyde, (S) 4-acetosyringone, (T) coniferyladehyde, (U) 4-propyl-
syringone, (V) 4-butylsyringone, (W) sinapylaldehyde, and (X)
sinapylaldehyde propionate.
Using the modified extraction solvent, nitration or oxida-
tion of the methoxyphenols was not observed during sample
extraction and analysis. Although nitration of other aromatic
compounds in the atmosphere is well established (33), we
are not aware of reports of nitrated methoxyphenols in
ambient PM samples. Clearly, the ease with which several of
the methoxyphenols were degraded under relatively mild
conditions in the laboratory has important implications for
the use of those compounds as tracers for wood smoke in
environmental PM samples. Future research is required to
identify the PM components responsible for the degradation
of the methoxyphenols and to determine whether the
degradation takes place in the air prior to sample collection,
on the filter media during sample collection or subsequent
storage, or only during sample extraction and analysis.
Hawthorne et al. reported that methoxyphenols bearing
alkenyl substituents (e.g., isoeugenol, allylsyringol) were
depleted in ambient samples relative to source samples
collected from residential chimneys, whereas the R-carbonyl-
substituted compounds vanillin and syringaldehyde were
proportionally enriched in the ambient samples (17). Thus,
Hawthorne’s data suggest that specific methoxyphenols, in
particular those bearing alkenyl substituents, chemically
degrade in (outdoor) air. Until quantitative measures of the
stability of methoxyphenols, both in the air and adsorbed on
yphenols were detected in the low ng/m3 range with vanillin,
syringaldehyde, coniferylaldehyde, and sinapylaldehyde be-
ing the most abundant compounds. Comparison of the
current data with previously reported measurements of
methoxyphenols in ambient samples is difficult because
differences in sample collection methods may influence the
comparability. Our observed methoxyphenol levels are lower
than 1988/1989 wintertime ambient levels in Minneapolis
(e.g., syringaldehyde: median 65 ng/m3, range 6.1-142 ng/
m3) and Salt Lake City (syringaldehyde: median 6.8 ng/m3,
range 0.1-23 ng/m3) reported by Hawthorne et al. (17),
although Hawthorne’s data include both particle-bound and
vapor-phase chemical (our measurements only capture
particle-bound methoxyphenols). Our reported methoxy-
phenol levels are also lower than wintertime (1995) levels of
particle-bound methoxyphenols reported by Schauer et al.
at urban sites (Fresno and Bakersfield) in the San Joaquin
Valley (e.g., 11-135 ng/m3 for syringaldehyde) but are similar
to levels reported for a rural site in the Kern Wildlife Refuge
(syringaldehyde 0.3 ng/m3) (3). Combustion of wood for
residential heating is common in Minneapolis, Salt Lake City,
and Fresno (2, 17). While wood burning is also a major source
of PM in Seattle, peak wintertime excursions in PM have
been greatly reduced in recent years as a consequence of
9
VOL. 39, NO. 2, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 635

sample collection media for analysis of organic tracers in
PM samples. The use of quartz filters to collect PM samples
may compromise subsequent quantitative analysis for spe-
cific methoxyphenols and potentially for other reactive
organic chemicals.
TABLE 5. Concentrations of Collocated PM2.5 and PM10 (in
µg/m3) and Methoxyphenols and Levoglucosan (in ng/m3)
sample sample
location type
PM
mass
levoglu- van- syring-
aceto- sinapyl-
cosan illin aldehyde syringone aldehyde
Determination of Methoxyphenols in PM. The newly
developed and validated analytical procedure for determi-
nation of methoxyphenols in PM was applied to ambient
PM samples collected in Seattle, WA. As part of the exposure
panel study, co-located PM2.5 and PM10 samples were
collected and PM2.5 samples were subject to trace element
analysis with XRF (24). This raises two issues when multiple
analyses are needed from the same filter samples: Do the
PM2.5 samples give similar concentrations of methoxyphenols
before and after the XRF? If not, do the co-located PM10 and
PM2.5 samples give similar methoxyphenols measurements
such that the co-located PM10 samples not subject to XRF
could be used for analysis of methoxyphenols? We hypoth-
esized that PM2.5 and PM10 samples give similar methox-
yphenol levels because the vast majority of wood smoke
particle mass is associated with particles with aerodynamic
diameters <1 µm (PM1) (34). Nine pairs of co-located PM2.5
and PM10 samples that had not been subjected to XRF analysis
were analyzed for methoxyphenols. The results are sum-
marized in Table 5. The anhydro-sugar levoglucosan, a
product of pyrolysis of cellulose and a putatative marker for
wood smoke PM, has previously been reported in this same
set of samples (28), and the data are included in Table 5 for
comparison with the methoxyphenols.
Only the relatively nonvolatile carbonyl-substituted meth-
oxyphenols were reliably detected in these environmental
samples. Syringol and guaiacol, and their methyl, ethyl
ethenyl, and propyl substituted analogues exist to a large
extent in the gas phase and were rarely present above the
quantification limit in PM samples. There is no significant
difference between methoxyphenol concentrations (ex-
pressed as ng/m3) as determined from co-located PM10 and
PM2.5 samplers using either a paired t-test or a nonparametric
sign test.
outdoor PM10
PM2.5
32
26
40
25
35
26
36
32
39
26
38
27
37
25
29
24
30
23
39
51
0.24 0.11
0.25 0.10
0.29 0.27
0.29 0.22
0.18 0.09
0.22 0.13
0.26 0.19
0.22 0.17
0.40 0.48
0.48 0.45
0.19 0.19
0.29 0.25
0.33 0.24
0.25 0.19
1.60 1.66
1.45 1.47
0.96 0.58
0.79 0.48
0.02
ndb
0.05
0.02
nd
0.02
0.04
0.04
0.11
0.15
nd
nd
0.02
nd
1.05
0.88
0.17
0.11
1.47
1.30
1.61
1.56
0.88
0.75
1.65
1.89
1.22
1.27
0.97
0.83
1.09
0.77
5.13
4.62
1.44
0.73
outdoor PM10
PM2.5
35
37
outdoor PM10
PM2.5
13
15
indoor PM10
PM2.5
33
26
indoor PM10
PM2.5
63
63
indoor PM10
PM2.5
15
20
indoor PM10
PM2.5
28
18
indoor PM10
PM2.5
762
700
141
138
indoor PM10
PM2.5
mean
PM10
PM2.5
35
26
125
119
0.49 0.42
0.47 0.38
0.21
0.20
1.72
1.52
p-valuea
<0.0001
0.77 0.48 0.16
0.30
0.08
a Paired t-test. b nd, not detected.
filter media, are established, reported ambient concentrations
of methoxyphenols should be interpreted with caution.
Various filter media, most commonly Teflon and quartz,
are used to collect samples of atmospheric PM. To determine
the general applicability of this assay, we determined
recoveries of deuterated methoxyphenols (∼40 ng) amended
onto clean Teflon and quartz filters (n ) 5). Median
methoxyphenol recoveries from Teflon were 92% (range 83-
132%). Recoveries of most of the deuterated methoxyphenols
from quartz filters were also acceptable (median 98%, range
87-131%), except for the syringol-type compounds (median
46%, range 33-72%). During development of the method,
we observed that syringol and its alkyl-substituted analogues
are more susceptible to chemical transformation via ring-
substitution reactions than other methoxyphenols. These
data indicate that Teflon membranes should be the preferred
On the basis of the above findings, we analyzed 10 pairs
of co-located ambient PM2.5 and PM10 samples in which one
sample from each pair had already undergone XRF analysis.
TABLE 6. Effect of XRF Analysis on Subsequent Determination of Methoxyphenol Concentrations in Co-Located Ambient PMa
site no.
XRF treatment
vanillin
acetovanillone
syringaldehyde
acetosyringone
coniferylaldehyde
sinapylaldehyde
1
+
-
+
-
+
-
+
-
+
-
+
-
+
-
+
-
+
-
+
-
0.99
1.10
1.62
1.81
3.21
3.66
2.60
3.13
2.73
3.33
3.07
3.74
2.54
2.38
1.08
1.35
0.68
0.77
1.46
1.50
0.52
0.49
0.49
0.50
0.85
0.76
0.63
0.77
0.82
0.77
0.66
1.69
0.63
0.66
0.55
0.51
0.45
0.40
0.52
0.51
1.14
1.41
1.62
1.96
6.67
8.64
6.31
8.51
5.62
7.48
6.69
9.45
7.28
6.93
1.17
1.56
0.70
0.77
1.14
1.21
0.57
0.72
0.62
0.74
3.92
5.05
3.96
5.13
3.07
4.48
3.96
6.69
5.05
4.97
0.52
0.69
0.32
0.35
0.52
0.55
1.63
1.81
1.75
1.86
5.01
6.13
5.01
5.99
5.08
6.82
5.96
9.49
5.22
5.05
1.64
1.81
nd
1.04
0.99
2
3
2.39
2.63
12.23
14.97
12.99
15.54
15.75
21.87
15.62
22.29
12.90
11.97
3.74
4
5
6
7
8
4.29
9
1.10
nd
1.22
10
1.80
1.84
2.01
1.89
mean
+
-
2.00
2.28
0.61
0.71
3.83
4.79
2.25
2.94
3.68
4.53
7.98
9.77
p-value
0.01
0.39
0.02
0.04
0.06
0.07
a Concentrations in ng/m3.
9
636 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 2, 2005

Although XRF analysis is considered “nondestructive”, we
were concerned that the XRF analysis may cause degradation
or volatilization of organic chemical species on the filter,
thus compromising subsequent analysis of methoxyphenols
in the PM samples. XRF analysis was performed per batch
of 12 filters placed in the XRF spectrometer under vacuum
(0.15 Torr) for 12 h. During the 12-h period, each filter was
subjected to X-ray bombardment for a total of 45 min, with
beam energies of 7.5-55 kV. Filters were stored at -4 °C
prior to and immediately after the XRF analysis. The results
are summarized in Table 6.
Acknowledgments
This work was funded in part by a U.S. EPA cooperative
agreement (CR82717701), the Northwest Center for Particu-
late Air Pollution and Health (U.S. EPA Grant CR827355),
and the National Institute of Occupational Safety and Health
(R03-OH007656).
Literature Cited
(1) Maykut, N. N.; Lewtas, J.; Kim, E.; Larson, T. V. Environ. Sci.
Technol. 2003, 37, 5135-5142.
(2) Poore, M. W. J. Air Waste Manage. Assoc. 2002, 52, 3-4.
(3) Schauer, J. J.; Cass, G. R. Environ. Sci. Technol. 2000, 34, 1821-
1832.
The co-located samples were either both PM2.5 samples
(sites 1-3) or co-located PM10 and PM2.5 samples (sites 4-10).
For sites 4-7, the PM2.5 samples received XRF analysis,
whereas for sites 8-10, it was the PM10 samples that received
XRF analysis. On the basis of our conclusions from Table 5,
co-located PM10 and PM2.5 samples were treated as equivalent
samples for the paired t-test. As shown in Table 6, concen-
trations of vanillin, syringaldehyde, and acetosyringone were
significantly lower in the filter subjected to XRF analysis when
compared to the corresponding co-located sample that was
not subjected to XRF analysis. Levels of coniferylaldehyde
and sinapylaldehyde were also decreased in the XRF treated
samples, and this difference approached statistical signifi-
cance. Similar results were obtained when nonparametric
methods (sign test) were used to analyze the data in Table
6. These observations are in contrast to our recent report
that XRF analysis of ambient PM samples did not affect
subsequent determination of levoglucosan (28). It is possible
that the energy of the X-ray beam causes chemical trans-
formation of the methoxyphenols. Alternatively, the reduction
in methoxyphenol levels we observed following XRF treat-
ment may be due to evaporation of the semivolatile meth-
oxyphenols either from direct heating by the X-ray beam or
during the 12 h the filters are under vacuum inside the XRF
spectrometer. The vacuum inside the XRF spectrometer is
0.15 Torr. In comparison, the pressure drop across the HI10
and HI2.5 used to collect the samples in the present study
was 27 and 29 Torr, respectively, and the pressure drop across
a high-volume sampler (e.g., Chem-vol) may be as high as
100 Torr. Although the pressure drop across typical PM
samplers does not approach the vacuum inside the XRF
spectrometer, it is nevertheless probable that the accuracy
of measurements of particle-bound methoxyphenols that
are collected by typical ambient PM samplers may be
compromised by a significant desorption artifact.
In conclusion, we developed an improved analytical
method that allows for quantitative determination of meth-
oxyphenols in samples of ambient PM, with good recoveries
and low limits of detection. We observed that ambient PM
and PM standard reference materials could cause oxidation
and nitration of methoxyphenols during sample extraction;
however, this degradation could be prevented by using a
solvent additive (triethylamine). The use of deuterium-labeled
authentic standard compounds to monitor methoxyphenol
recovery is a critical feature of this method; our recovery
data demonstrate acceptable reproducibility and precision
for determination of methoxyphenols in environmental
samples collected on PTFE membrane filters. We also
demonstrated that methoxyphenols are present predomi-
nantly in the fine (PM2.5) particle size fraction and that XRF
analysis of samples compromised subsequent measurement
of methoxyphenol levels. Our study underscores the fact that
many methoxyphenols are chemically reactive and exhibit
appreciable volatilitystwo features that could limit the utility
of these chemicals in source attribution studies. Future
studies are called for to measure the atmospheric stability
of the methoxyphenols and to assess the influence of
adsorption and desorption artifacts on measurements of
methoxyphenols.
(4) Brauer, M.; Hisham-Hashim, J. Environ. Sci. Technol. 1998, 32,
404A-407A.
(5) Mott, J. A.; Meyer, P.; Mannino, D.; Redd, S. C.; Smith, E. M.;
Gotway-Crawford, C.; Chase, E. West J. Med. 2002. 176, 157-
162.
(6) Osterman, K.; Brauer, M. In Forest Fires and Regional Haze in
Southeast Asia; Eaton, P., Radojevic, M., Eds.; Nova Science
Publishers: Hauppauge, NY, 2001; Chapter 10.
(7) Smith, T.; Ward, G. Air sampling of the 2000 Montana wildfire
season. Presented at the Air and Waste Management Association
Annual Conference, 2001; Paper 1131.
(8) Ezzati, M.; Kammen, D. M. Environ. Health Perspect. 2001, 109,
481-488.
(9) Smith, K. R. Proc. Natl. Acad. Sci. U.S.A. 2000, 97, 13286-13293.
(10) Leonard, S. S.; Wang, S.; Shi, X.; Jordan, B. S.; Castranova, V.;
Dubick, M. A. Toxicology 2000, 150, 147-157.
(11) Pryor, W. A. Free Radical Biol. Med. 1992, 13, 659-676.
(12) Schauer, J. J.; Kleeman, M. J.; Cass, G. R.; Simoneit, B. R. Environ.
Sci. Technol. 2001, 35, 1716-1728.
(13) Larson, T. V.; Koenig, J. Q. Annu. Rev. Public Health 1994, 15,
133-156.
(14) Hawthorne, S. B.; Krieger, M. S.; Miller, D. J.; Mathiason, M. B.
Environ. Sci. Technol. 1989, 23, 470-475.
(15) Rogge, W. F.; Hildemann, L. M.; Mazurek, M.; Cass, G. R.;
Simoneit, B. R. T. Environ. Sci. Technol. 1998, 32, 13-22.
(16) Fine, P. M.; Cass, G. R.; Simoneit, B. R. Environ. Sci. Technol.
2002, 36, 1442-1451.
(17) Hawthorne, S. B.; Miller, D. J.; Langenfeld, J. J.; Krieger, M. S.
Environ. Sci. Technol. 1992, 26, 2251-2262.
(18) Nolte, C. G.; Schauer, J. J.; Cass, G. R.; Simoneit, B. R. Environ.
Sci. Technol. 2001, 35, 1912-1919.
(19) Kja¨llstrand, J.; Ramnas, O.; Petersson, G. J. Chromatogr., A 1998,
824, 205-210.
(20) Dills, R. L.; Zhu, X.; Kalman, D. A. Environ. Res. 2001, 85, 145-
158.
(21) Coscia, C. J.; Schubert, W. J.; Nord, F. F. J. Org. Chem. 1961, 26,
5085-5091.
(22) Buu-Ho¨ı, N. P.; Se´alles, J., Jr. J. Org. Chem. 1955, 20, 606-609.
(23) Chau, A. S. Y.; Coburn, J. A. J. Assoc. Off. Anal. Chem. 1974, 57,
389-393.
(24) Liu, L. J.; Box, M.; Kalman, D.; Kaufman, J.; Koenig, J.; Larson,
T.; Lumley, T.; Sheppard, L.; Wallace, L. Environ. Health Perspect.
2003, 111, 909-918.
(25) Marple, V. A.; Rubow, K. L.; Turner, W.; Spengler, J. D. J. Air
Pollut. Control Assoc. 1987, 37, 1303-1307.
(26) Allen, R.; Box, M.; Liu, L.-J. S.; Larson, T. J. Air Waste Manage.
Assoc. 2001, 51, 1650-1653.
(27) Fenimore, D. C.; Davis, C. M.; Whitford, J. H.; Harrington, C. A.
Anal. Chem. 1976, 48, 2289-2290.
(28) Simpson, C. D.; Dills, R. L.; Katz, B. S.; Kalman, D. A. J. Air Waste
Manage. Assoc. 2004, 54, 689-694.
(29) Franz, J. E.; Herber, J. F.; Knowles, W. S. J. Org. Chem. 1965, 30,
1488-1491.
(30) Sobolev, I. J. Org. Chem. 1961, 26, 5080-5085.
(31) Brink, D. L.; Bicho, J. G.; Merriman, M. M. Adv. Chem. Ser. 1966,
59, 177-204.
(32) Polcin, J.; Rapson, W. H. Pulp Pap. Can. 1971, 72, T114-T125.
(33) Atkinson, R.; Arey, J. Environ. Health Perspect. 1994, 102 (Suppl
4), 117-126.
(34) Kleeman, M. J.; Schauer, J. J.; Cass, G. R. Environ. Sci. Technol.
1999, 33, 3516-3523.
Received for review August 20, 2004. Revised manuscript
received October 20, 2004. Accepted October 22, 2004.
ES0486871
9
VOL. 39, NO. 2, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 637