Polymer pyrolysis and oxidation studies in a continuous feed and flow reactor: Cellulose and polystyrene

Article abstract of DOI:10.1021/es980796q

A dual-zone, continuous feed tubular reactor is developed to assess the potential for formation of products from incomplete combustion in thermal oxidation of common polymers. Solid polymer (cellulose or polystyrene) is fed continuously into a volatilization oven where it fragments and vaporizes. The gas-phase polymer fragments flow directly into a second, main flow reactor to undergo further reaction. Temperatures in the main flow reactor are varied independently to observe conditions needed to convert the initial polymer fragments to CO₂ and H₂O. Combustion products are monitored at main reactor temperatures from 400 to 850 °C and at 2.0-s total residence time with four on-line GC/FIDs; polymer reaction products and intermediates are further identified by GC/MS analysis. Analysis of polymer decomposition fragments at 400 °C encompasses complex oxygenated and aromatic hydrocarbon species, which range from high-molecular-weight intermediates of ca. 300 amu, through intermediate mass ranges down to C₁ and C₂ hydrocarbons, CO, and CO₂. Approximately 41 of these species are positively identified for cellulose and 52 for polystyrene. Products from thermal oxidation of cellulose and polystyrene are shown to achieve complete combustion to CO₂ and H₂O at a main reactor temperature of 850 °C under fuel-lean equivalence ratio and 2.0-s reaction time. A dual-zone, continuous feed tubular reactor is developed to assess the potential for formation of products from incomplete combustion in thermal oxidation of common polymers. Solid polymer (cellulose or polystyrene) is fed continuously into a volatilization oven where it fragments and vaporizes. The gas-phase polymer fragments flow directly into a second, main flow reactor to undergo further reaction. Temperatures in the main flow reactor are varied independently to observe conditions needed to convert the initial polymer fragments to CO₂ and H₂O. Combustion products are monitored at main reactor temperatures from 400 to 850°C and at 2.0-s total residence time with four on-line GC/FIDs; polymer reaction products and intermediates are further identified by GC/MS analysis. Analysis of polymer decomposition fragments at 400°C encompasses complex oxygenated and aromatic hydrocarbon species, which range from high-molecular-weight intermediates of ca. 300 amu, through intermediate mass ranges down to C₁ and C₂ hydrocarbons, CO, and CO₂. Approximately 41 of these species are positively identified for cellulose and 52 for polystyrene. Products from thermal oxidation of cellulose and polystyrene are shown to achieve complete combustion to CO₂ and H₂O at a main reactor temperature of 850°C under fuel-lean equivalence ratio and 2.0-s reaction time.

Full text of DOI:10.1021/es980796q

Environ. Sci. Technol. 1999, 33, 2584-2592
exist over possible products resulting from incomplete
combustion in municipal waste incinerators (2-5).
Polymer Pyrolysis and Oxidation
Studies in a Continuous Feed and
Flow Reactor: Cellulose and
Polystyrene
A number of studies on thermal and oxidative decom-
position of polymers have reported valuable data on products
or overall rates of polymer volatilization and reaction under
non-steady-state operating conditions. Many of these studies
are based on experiments that use a batch or a semibatch
reactor with continuous gas flow and batch polymer inlet,
drop tube furnace, or temperature-programmed gravimetric
analysis. The batch reactors report valuable product forma-
tion data but do not determine simultaneous time of
formation and fuel equivalence in the vapor phase and are
often operated at high initial fuel equivalence. A drop tube
furnace monitors mass loss and ignition characteristics of a
particle, in addition to intermediate product formation, but
time of vapor-phase species reaction can vary from flow
through the reactor to formation on exit. Global kinetics of
polymer volatilization is determined by thermal gravimetric
analysis, which can monitor mass loss and products as the
initial mass is exposed to a temperature ramp (6-9). Thermal
gravimetric studies are valuable in determining temperature
regimes of polymer reaction, transformation, and volatiliza-
tion, in addition to intermediate product identification.
Previous Studies: 1. Cellulose. Studies on pure cellulose
pyrolysis and oxidation fall into general categories of low
and high temperatures. There are few studies on pure
cellulose at temperatures over 700 °C in the literature due
to its low heating value and oxygenated structure. Pastorova
et al. (10) report an extensive list of products from the pyrolysis
of cellulose, which is initially charred at 190-390 °C and
then further reacted at temperatures of 610 and 770 °C. They
report that aromatic compounds increase with an increase
in char temperature. Keating and Gupta (11) investigated
the oxidative pyrolysis of cellulose at temperatures between
1500 and 2900 K at identical fuel equivalence ratios in air
and in O2. They observe larger quantities of the identified
products from pyrolysis and oxidation in air, relative to that
in O2. They recommend a cellulose combustion process of
initial pyrolysis and then stages of oxidation and pyrolysis
for minimization of intermediate products and NO2 forma-
tion.
B YU N G - I K P A R K ,
J O S E P H W . B O Z Z E L L I , * A N D
M I C H A E L R . B O O T Y
Department of Chemical Engineering, Chemistry, and
Environmental Science, New Jersey Institute of Technology,
Newark, New Jersey 07102
M A R Y J . B E R N H A R D , K A R E L M E S U E R E ,
C H A R L E S A . P E T T I G R E W ,
J I - C H U N S H I , A N D S T A C I L . S I M O N I C H
The Procter and Gamble Company, Cincinnati, Ohio
A dual-zone, continuous feed tubular reactor is developed
to assess the potential for formation of products from
incomplete combustion in thermal oxidation of common
polymers. Solid polymer (cellulose or polystyrene) is fed
continuously into a volatilization oven where it fragments and
vaporizes. The gas-phase polymer fragments flow directly
into a second, main flow reactor to undergo further
reaction. Temperatures in the main flow reactor are varied
independently to observe conditions needed to convert
the initial polymer fragments to CO and H O. Combustion
2
2
products are monitored at main reactor temperatures
from 400 to 850 °C and at 2.0-s total residence time with
four on-line GC/FIDs; polymer reaction products and
intermediates are further identified by GC/MS analysis.
Analysis of polymer decomposition fragments at 400 °C
encompasses complex oxygenated and aromatic hydrocarbon
species, which range from high-molecular-weight
intermediates of ca. 300 amu, through intermediate mass
ranges down to C and C hydrocarbons, CO, and CO .
A number of studies have been reported on cellulose
pyrolysis and combustion at temperatures below 700 °C, and
these consistently show similar intermediate product dis-
tributions (12-15). Pouwels et al. (12) identified a wide range
of gas-phase products from Curie-point pyrolysis coupled
with gas GC/ MS at 510 °C under batch conditions. Jakab et
al. (13) focus on thermal decomposition products from
cellulose and wood effected by the presence of solvents when
their reactor is heated to 400 °C at 20 °C/ min. Shafizadeh
(14) studied cellulose pyrolysis from below 300 to over 500
°C. He reports acceptable conversion to desired products
(light hydrocarbons, levoglucosan, anhydro sugars, oligosac-
charides, pyrans, and furan dehydration species) is obtained
between 300 and 500 °C.
Suuberg et al. (6) and Milosavljevic and Suuberg (7)
conducted a number of studies on cellulose pyrolysis. They
conclude from thermal gravimetric analysis (TGA) experi-
ments and literature data that cellulose exhibits two different
activation energies in its decomposition. One value reported
for overall decomposition is 140-155 kJ/ mol, which is
observed with both low and high heating rates and T-ramps
to temperatures above 327 °C. A second value of about 218
kJ/ mol applies for decomposition with lower heating rates
to temperatures below 327 °C. They conclude that the
occurrence of different activation energies is due, in part, to
1
2
2
Approximately 41 of these species are positively identified
for cellulose and 52 for polystyrene. Products from
thermal oxidation of cellulose and polystyrene are shown
to achieve complete combustion to CO and H O at a
2
2
main reactor temperature of 850 °C under fuel-lean
equivalence ratio and 2.0-s reaction time.
Introduction
Oxyhydrocarbon and hydrocarbon polymers such as cel-
lulose, polystyrene, and polyethylene constitute a significant
proportion of municipal solid waste. In the United States,
cellulose as the main component of paper contributes
approximately 39% by mass and plastics contribute ap-
proximately 9% by mass to municipal solid waste (1).
Cellulose is also a major component of the various forms of
biomass. Although these materials are useful as fuels in
incineration processes and refuse-to-energy conversion due
to their relatively high heat value, environmental concerns
* Author for correspondence: phone: (973) 596-3459; fax: (973)
596-5854; e-mail: Bozzelli@tesla.Njit.edu.
9
2 5 8 4 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 33, NO. 15, 1999
10.1021/es980796q CCC: $18.00
1999 Am erican Chem ical Society
Published on Web 06/24/1999

FIGURE 1. Two-zone oven, continuous feed experimental apparatus. TC denotes a thermocouple and proportional controller pair providing
temperature control for the electrical oven heaters.
the role of relatively high-molecular-weight products, or tars,
and that the actual kinetics is more complex than can be
described by an overall process.
particulate effluent decrease with increasing temperature
and with decreasing polymer particle size. They indicate that
polystyrene produces more soot and PAHs than other
polymer plastics and recommend polystyrene be burned
under conditions of high excess air to minimize emission of
these species.
The research group of Karasek (26) identified a number
of polyaromatic hydrocarbons from their high-temperature
(850-950 °C) batch feed (2 g) pyrolysis and oxidation flow
tube experiments. The gas velocity through their reactor was
slow, ca. 12 cm/ s, with products collected in cryogenic traps
and on glass-wool filters. Elomaa and Saharinen (27)
performed similar experiments with smaller samples, col-
lecting soot in addition to analyzing the soot extracts for
PAHs.
Many investigators have examined the pyrolysis and
oxidation of polystyrene at temperatures between 200 and
550 °C (29-35). Brauman et al. (29) investigated the thermal
degradation of vertically mounted polystyrene rods under
radiant heating and nonflaming conditions at 446 °C in air
and in nitrogen. They report average styrene monomer yields
of 33% from reaction under nitrogen and relative concentra-
tions of high-boiling-point products: styrene > trimer >
dimer > tetramer > pentamer > hexamer. Shapi and Hesso
(32) studied thermal decomposition of polystyrene at 300 °C
using a Pyrex pyrolysis tube under synthetic air or nitrogen.
They extracted vapor-phase products in a hexane solvent
trap and report 164 volatile species from reaction in air and
71 in nitrogen. Relative product yields were 86.8% in air and
90.3% in nitrogen, with relative yield percents of styrene
(27.3%) > benzaldehyde (14.7%) > dimer (8.2%) > trimer
(3.1%) in air. Carniti et al. (35) studied thermal degradation
of polystyrene and resulting products from batch reaction
in a vacuum-sealed glass vial, at temperatures to 420 °C. The
thermal degradation of polystyrene was reported to follow
Several research groups investigated thermal decomposi-
tion of biopolymers, of which cellulose is an important
fraction, over a range of temperatures (16-21). Funazukuri
et al. (16) measured weight loss and product yields (C6 species
and below) from the flash pyrolysis of cellulose powder,
cellulose particles, and filter paper under batch conditions.
Reaction temperatures and heating interval ranged from 310
to 770 °C and from 0.5 to 20 s, respectively. They report that
CO yields increase with weight loss and pyrolysis temperature.
CO2 yields are proportional to weight loss but are independent
of pyrolysis temperature. Encinar et al. (17) used a tubular
reactor containing a batch sample of polymer to study
gaseous effluent from olive and grape bagasse. Gas con-
centrations were measured at temperatures from 300 to 900
°C. The gaseous products and solids from combustion were
observed to decrease with increasing temperature. Studies
by McKenzie, Lipari, Williams, and Pakdel all report that
products from thermal degradation of biomass are similar
to those observed in cellulose pyrolysis and oxidation (18-
21).
2. Polystyrene. A number of studies have been reported
on the thermal decomposition of polystyrene at temperatures
from 700 to 1200 °C (22-28). Levendis’s group (22-24) report
data on decomposition of polystyrene in horizontal batch
reactors and in drop tube furnaces, with gas-phase and
particulate samples collected on glass fiber filters or XAD
adsorbents. Polymer particles allowed to burn under flame
conditions show formation of soot and high levels of PAHs.
Durlak et al. (25) studied the combustion of polystyrene
spheres in an upflow tubular reactor with 0.5-s residence
time over a temperature range of 800-1200 °C. They reported
that the total PAHs and mass of the combined gas-phase and
9
VOL. 33, NO. 15, 1999 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 2 5 8 5

FIGURE 2. Reactor temperature profiles at 2.0-s residence time.
first-order kinetics in the initial part of the reaction up to 300
min.
Present Study. Results on conversion and product
formation from thermal oxidation of cellulose and polysty-
rene as a function of temperature using a continuous flow
reactor at known reaction time and fuel equivalence ratio
(φ) are described in this study. The reactor has two isothermal
reaction zones in series: a volatilization oven into which
polymer and air are fed and a higher temperature zone to
study reaction (oxidation/ pyrolysis) of the volatilization
products. Analysis of initial, intermediate, and final products
as a function of temperature, at known φ and reaction time,
allows one to identify conditions for complete conversion of
polymer fragments to CO2 and H2O. The apparatus includes
a continuous monitor of vapor-phase products to ensure
uniform, steady-state operation. The continuous feed, tem-
perature, φ, and residence time parameters are all relevant
to operation of practical municipal solid waste incinerators.
FIGURE 3. GC/FID signal from the total hydrocarbon versus time
indicating closeness to steady-state operation of the reactor. Arrows
indicate inlet of sample to GCs.
channel tube of about 20-cm length made of quartz or
stainless steel; it is filled with a uniform distribution of
polymer powder at a known bulk density. This U-channel
“boat” is fed slowly into the volatilization oven by connection
to a variable drive syringe-pump (Sage Industries/ Orion
Research Inc.) with a 2-mm diameter stainless steel rod. The
feed rod passes through a septum at the upstream end of the
reactor tube in order to seal the reaction zones from outside
air. The feed rates are around 55 mg/ min for cellulose and
8 mg/ min for polystyrene and are sufficient to allow steady-
state conditions. The volatilization oven heaters are adjusted
to maintain a uniform temperature between 350 and 450 °C.
Reagents. The polymer powder reagents used were
microcrystalline cellulose (Sigma Sigmacell type 101); poly-
styrene (Aldrich Chemical Co.) with average molecular weight
230 000, melt index 8.5, particle density 1.04 gm/ cm3, and
particle size in the range 0.1-0.3 mm.
Gas Flow. Dry air was inlet to the reactor tube 30 cm
upstream of the initial volatilization oven heater and flowed
over the polymer within the U-channel boat and through
the reactor at a known flux. With a fixed 2.0-s gas residence
time in the main reactor, the flow Reynold’s number was in
the range 70-170. This ensured that the flow was within the
laminar regime throughout the reactor and that there was
no tendency for the flow to lift unvaporized polymer particles
into the reactant stream.
A second method of gas flow incorporated a small nitrogen
flow over the feed polymer, with the main air plus supple-
mental oxygen inlet separately via a Tee. These streams met
and mixed in the volatilization oven, and the initial O2:N2
ratio was always 21:79. The N2 stream prevents back-diffusion
of the oxidation or flame radical pool to unfed polymer and
preignition of the feed polymer, which was observed in
polymer mixture reaction studies where the solids exhibited
lower melting points.
Analysis. The analytical train consists of three on-line
gas chromatographs using flame ionization detectors (GC/
FIDs). The first GC (Varian 3700) uses two 3-mm diameter
packed columns with two gas sample injectors and two FIDs.
CO and CO2 are methanated using a ruthenium catalyst and
hydrogen-reducing agent to provide high sensitivity by a FID
after separation on a Carbosphere packed column (Alltech
Associates Inc.). The second column in this GC is a Super-Q
packed column (Alltech) which separates the light hydro-
carbon species from C1 to C4. The second GC (Hewlett-
Experimental Section
Reactor. The apparatus consists of a horizontal tubular flow
reactor with on-line analytical train, as shown in Figure 1.
The quartz reactor tube is 2.0 m in length, is 1.05 cm in inner
diameter, and is surrounded by cylindrical electrical heaters
(Thermcraft Co.) distributed over the two adjacent oven
zones. The first zone, or volatilization oven, has a 7.6-cm
long heater on the inlet side followed by a 20.3-cm long heater.
The shorter heater is used to provide higher heat flux to
compensate for heat loss at the reactor inlet and brings the
incoming air and polymer feed up to temperature quickly.
The longer heater serves to maintain a uniform temperature
throughout the volatilization oven and allows the initial vapor
fragments to pass into the second zone without condensation.
The second oven zone is the main reactor; it has two
central 46-cm long heaters and two 20.3-cm end heaters.
Effluents pass from the main reactor to the analytical train
through Pyrex and stainless steel transfer lines that are heated
to 220 °C. The volatilization oven temperature is fixed at 400
°C. The main reactor temperature is held constant ((10 °C)
and is controlled in the range 400-850 °C, with 2.0-s vapor
residence time.
The volatilization oven and main reactor heaters are
controlled by proportional controllers (Jenco Co.) with K-type
thermocouples (Omega Engineering Inc.). Uniform temper-
ature profiles, in both time and distance along each reactor
zone, are readily achieved within (10 °C (see Figure 2). The
heaters are usually operated at about one-half of their
maximum rated power to limit temperature over-run and
oscillation. This limits the main reactor temperature to a
maximum of 1100 °C, which is in the range of the flue gas
temperatures of municipal solid waste incinerators (3, 4).
Polym er Feed. The continuous polymer powder feed
mechanism consists of a small diameter open-top or U-
9
2 5 8 6 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 33, NO. 15, 1999

TABLE 1. Distribution and Fate of Vapor-Phase Products in the Oxidation of Cellulosea
main reactortemperature (°C)
500 550 600 650
no. MW
products
400
450
700
750
1
2
3
4
5
6
7
8
16 m ethane, CH4
28 ethene, C2H4
30 ethane, C2H6
42 propene, C3H6
44 acetaldehyde, C2H4O
56 butene, C4H8
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
68 furan, C4H4O
58 acetone, C3H6O
72 2-butanone, C4H8O
82 2-m ethylfuran, C5H6O
78 benzene, C6H6
96 2-ethylfuran, C6H8O
96 2,5-dim ethylfuran, C6H8O
84 3-penten-2-one, C5H8O
92 toluene, C7H8
9
10
11
12
13
14
15
16
x
x
x
x
x
x
84 3H-furan-2-one, C4H4O2
17 100 3-hexen-1-ol, C6H12
O
18
19
20
96 3-furfural, C5H4O2
96 2-furfural, C5H4O2
98 2-furanm ethanol, C5H6O2
x
x
x
x
21 114 3-hexyne-2,5-diol, C6H10O2
+
22
23
98 5-m ethyl-2(3H)-furanone, C5H6O2
96 cyclopent-2-ene-1,4-dione, C5H4O2
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
24 104 styrene, C8H8
x
+
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
25
26
27
96 2-m ethyl-2-cyclopenten-1-one, C6H8O
84 5H-furan-2-one, C4H4O2
96 2-cyclohexen-1-one, C6H8O
28 112 3-m ethyl-2,5-furandione, C5H4O3
29 110 5-m ethyl-2-furfural, C6H6O2
x
x
x
x
x
x
30
94 phenol, C6H8O
31 118 2,3-benzofuran, C8H6O
x
32 112 3,4-dihydro-6-m ethyl-2H-pyran-2-one, C6H8O2
33 112 2-hydroxy-3-m ethyl-2-cyclopenten-1-one, C6H8O2
34 124 2-m ethyl-1,4-benzenediol, C7H8O2
35 126 3-(hydroxy)-2-m ethyl-4-pyrone, C6H6O3
36 144 5-(hydroxym ethyl)-2-tetrahydrofurfural-3-one, C6H8O4
37 144 1,4:3,6-dianhydro-R-D-glucopyranose, C6H8O4
38 126 5-hydroxym ethyl-2-furfural, C6H6O3
39 162 1,6-anhydro-â-D-glucose, C6H10O5
x
x
x
x
x
x
x
x
40
41
28 CO
44 CO2
x
x
x
x
x
x
x
x
x
x
x
fuel equivalence ratio (φ)
0.21
0.21
0.21
0.22
0.22
0.25
0.27
0.28
carbon m ass balance (%)
61.76 69.17 73.68 82.40 88.18 88.71 95.53 93.12
a
Volatilization oven tem perature 400 °C, m ain reactor tem perature as indicated; x signifies detected by GC and GC/MS analysis, + signifies
detected by GC/MS and not detected by GC.
Packard 5890 series II) is fitted with a capillary column for
separation of gas-phase products over C5. For studies on
cellulose the capillary column is an Alltech Associates Inc.
AT-5 (SE-54) 1.2-µm, 30-m × 0.53-mm column, while for
studies on polystyrene a J & W Scientific DB-1 (SE-30) 1.2-
µm, 15-m × 0.53-mm column is used. A capillary column
with the same liquid phase is used for identification of gas-
phase products in the GC/ MS (Varian 3400 gas chromato-
graph with Varian Saturn II ion trap mass spectrometer).
The third GC (Varian 3700) operates without column
packing and uses reactor effluent for carrier at 250 °C. Its FID
is used to measure total hydrocarbon and oxyhydrocarbon
products. This signal serves as an important monitor on
uniform, steady-state operation. It monitors continuity of
total operation: gas flow, polymer feed, volatilization rate,
absence of preignition, reaction, product effluent, and
product transfer to analytical equipment. Figure 3 shows a
typical total hydrocarbon signal from this FID against time
for the approximate 15-30-min duration of an experimental
run with cellulose and polystyrene. This continuous moni-
toring of total hydrocarbons also serves as a second mea-
surement on destruction efficiency and supports quantitative
results obtained by the other three GC/ FIDs.
Identification of reaction products includes matching of
retention times against known standard compounds under
identical GC conditions and analysis of GC mass spectra
against spectral libraries and known compound (class)
fragmentation pathways (36). Gas products for GC/ MS
analysis were collected in an impinger containing 30 mL of
methylene chloride at 0 °C (ice bath). The extracts were
subsequently concentrated to 1 mL by purge with nitrogen,
with analysis via 1.0-µL autoinjection. The reaction product
spectrum was compared with standard spectra of the NIST
92 Library (37) and with spectra of purchased standard
compounds.
Quantitation from identified polymer oxidation or py-
rolysis products is from response factors determined by
injection of an external standard, whenever possible. Re-
sponse factors for products without external standards were
assigned according to their carbon number and structure.
Detection limits of gas-phase products except CO and CO2
were at a level of 10-100 ppb.
Results and Discussion
Reactor Operation. Volatilization oven heaters are set to
maintain a constant, uniform temperature between 350 and
9
VOL. 33, NO. 15, 1999 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 2 5 8 7

450 °C. This is below typical grate and fuel-bed temperatures
of municipal solid waste incinerators, which are in the
approximate ranges 425-540 and 750-1100 °C, respectively
(3, 4). This temperature range is chosen to ensure complete
and uniform volatilization of polymer feed, with limited
further degradation of the initial polymer fragments within
the volatilization zone.
The main reactor temperature was initially held at 400
°C. With both the volatilization oven and main reactor at this
moderately low temperature, it is assumed that initial polymer
fragments leaving the volatilization oven do not undergo
significant further reaction or condensation downstream of
the volatilization oven and can be analyzed using the on-
line GCs. Once this initial product data is obtained, the
volatilization oven is maintained at 400 °C and separate
experiments are run at main reactor temperatures between
400 and 850 °C in 50 °C intervals.
Exothermicity of the reaction alters the reactor temper-
ature profile slightly. This change is small, less than 10 °C,
for the polymers and feed rates in this study. Measured
temperature increases due to reaction are below 10 °C in the
first 3 cm of the volatilization oven stage and below 3-5 °C
in the first 2 cm of the main reactor. There are no other
observed changes in the temperature profiles.
Cellulose. Cellulose is a natural biopolymer, with its
molecule (C6H10O5) consisting of approximately 50% oxygen
by mass. For this series of runs the carrier gas was dry air,
the volatilization oven was maintained at 400 °C throughout,
the residence time in the main reactor was a constant 2.0 s,
and the fuel equivalence ratio was kept at a fuel-lean value
in the range 0.20-0.30. The fuel equivalence ratio for a given
run is calculated as the ratio of moles of cellulose (fuel) to
oxygen divided by the stoichiometric ratio for the overall
reaction: C6H10O5 + 6O2 f 6CO2 + 5H2O. Consideration of
the fuel equivalence range and complete reaction for cellulose
imply a decrease in the mole percentage of oxygen from
21.6% in inlet air to a value between 14.38% and 16.72% for
the reactor effluent.
Initial products from reaction with both the volatilization
oven and main reactor temperatures at 400 °C were identified,
and 41 species are listed in Table 1. The main pyrolysis
products are anhydro sugars, pyran derivatives, alkylfurans,
hydrocarbons, and carbonyl compounds. Cellulose in the
volatilization oven undergoes dehydration, decarboxylation,
transglycosylation, and fission to produce major anhydro
sugars such as levoglucosan as well as low-molecular-weight
products such as CO, CO2, methane, ethene, acetaldehyde,
furans, acetone, and 2-furfural. These product distributions
are in agreement with the thermal decomposition mechanism
for cellulose proposed by Pouwels et al. (12) and Shafizadeh
(14).
Levoglucosan is one of the primary products and is present
at 2% at 400 °C and 2-s reaction time; its concentration is
reduced to below its detection limit as the main reactor
temperature is increased to 500 °C. It is difficult to analyze
levoglucosan with a GC due to its low volatility and
interference with other anhydro sugars and oligosaccharides
(38-40). The mass balance deficit in Table 1 is attributed to
the production of these sugars, which are believed to be
major pyrolysis products (41) and accumulate in the reactor’s
condensation trap (see Figure 1) at low temperatures.
5-(Hydroxymethyl)-2-tetrahydrofurfural-3-one, 5-(hydroxy-
methyl)-2-furfural, and 1,4:3,6-dihydro-R-D-glucopyranose
are initial products from dehydration and transglycosylation
reactions. These C6 oxygenated products are destroyed at
temperatures of 500 to 600 °C as indicated by increased
carbon mass balance above 500 °C and destruction of the C5
and higher-molecular-weight products at 600 °C (C1 to C4
species, methylfuran, 2-furfural, and aromatic compounds
are exceptions). Carbon balance is 88.18% at 600 °C, relative
FIGURE 4. Cellulose decomposition product data from the on-line
GCs at main reactor temperatures of 500, 650, and 850 °C at 2.0-s
residence time: (a) Super-Q column, (b) AT-5 capillary column, (c)
Carbosphere column.
to 82.40% at 550 °C and 73.68% at 500 °C. All aromatic
compounds disappear at 650 °C, with the exception of
benzene, which is present at trace levels at 700 °C. No
intermediate or incomplete products of cellulose combustion
were detected above 700 °C.
Gas chromatograms from the three different columns are
shown at main reactor temperatures of 500, 650, and 850 °C
and 2.0-s residence time in Figure 4. This provides a clear
picture of the overall product concentrations and their change
with temperature. At 400 and 500 °C light hydrocarbons from
C1 to C4 (Super-Q column) and heavy organic products (AT-5
capillary column) were relatively abundant, while CO and
CO2 (Carbosphere column) were low. At temperatures above
650 °C higher-molecular-weight organic compounds were
low and their levels were barely detectable. Light hydrocarbon
concentrations began to show rapid falloff, and CO and CO2
levels increased, as expected. An overall conversion of higher-
molecular-weight cellulose polymer fragments to lower-
molecular-weight fragments is observed with increasing
reaction temperature. The volatilization oven temperature,
inlet carrier gas composition, and main reactor residence
times were all held fixed, and fuel equivalence was also near-
constant. Initial polymer fragment concentrations and
compositions entering the main reactor were thus the same
throughout.
Figure 5a shows product yields versus temperature for
three intermediate molecules: 2-butanone, 2-methylfuran,
and 2-furfural. Data for these species show that falloff in
concentration begins near 450 °C for 2-butanone and
2-methylfuran, which is lower than the fall-off temperature
for C1 to C4 species. A major stable product is 2-furfural
produced from ring cleavage of levoglucosan via transgly-
cosylation and dehydration. The data show 2-furfural remains
as one of the higher level initial products with molecular
weight above C5, from 500 to 600 °C, and that it decreases
to near-zero detection level at 650 °C.
Figure 5b illustrates product yields for the C1 to C4
species: methane, ethene, acetaldehyde, acetone, and furan
versus temperature. Increase in temperature to 650 °C in the
9
2 5 8 8 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 33, NO. 15, 1999

FIGURE 6. Polystyrene decomposition product data from the on-
line GCs at main reactor temperatures of 500, 650, and 850 °C at 2.0-s
residence time: (a) Super-Q column, (b) DB-1 capillary column, (c)
Carbosphere column.
occurs rapidly between 650 and 700 °C at which CO is oxidized
to CO2.
Polystyrene. Polystyrene is a nonoxygenated vinylphenyl
polymer with monomer formula C8H8. The experiments were
performed at lean fuel equivalence ratios over the range
0.07-0.13, based on the overall reaction: C8H8 + 10O2
f
8CO2 + 4H2O. At these equivalence ratios and assuming
complete reaction, the mole percentage of oxygen in inlet
air decreases from 21.6% to a value between 18.73% and
20.07% in the reactor effluent; 52 species are positively
identified by GC and GC/ MS analysis, and data are presented
in Table 2. Most products show structures related to aromatic
monomer, dimer, and trimer of styrene. Initial primary
products at 400 °C are predominantly styrene: monomer,
dimer, trimer, and benzaldehyde. Minor initial products at
400 °C are CO2, benzene, toluene, ethylbenzene, R-meth-
ylstyrene, 2,3-dihydro-1H-inden-1-one, 3-phenyl-2-propen-
1-ol, 1,1′-(1,3-propanediyl)bisbenzene, (E)-stilbene, 1,1′-(1-
butene-1,4-diyl)bis-(Z)-benzene, 1,1′-(1,2-ethanediyl)bis-4-
methylbenzene, 1,3-diphenyl-(E)-2-propen-1-one, trans-1,2-
dibenzoylethylene, and quaterphenyl. Secondary products
such as phenylethyne, benzofuran, 1,3-benzodioxole, naph-
thalene, and biphenyl result from oxidation and pyrolysis of
initial products in the main reactor zone. These species are
relatively low in concentration at 400 °C, where they are
detected by GC/ MS analysis only, but their concentrations
increase with increasing main reactor temperature.
Figure 6 shows GC chromatographic data for temperatures
of 500, 650, and 850 °C at 2.0-s residence time. The trends
show a clear increase in conversion of high-molecular-weight
intermediates to lower-molecular-weight products (similar
to results found for cellulose) as main reactor temperature
is increased. The change in the distribution of product
concentrations with temperature is dramatic and demon-
strates that the aromatic products formed in the thermal
oxidation of polystyrene can be degraded under these fuel-
lean 700-850 °C conditions.
FIGURE 5. Product yields versus main reactor temperature in the
thermal oxidation of cellulose: (a) C and C major products, (b) C
4
5
1
to C light hydrocarbons, (c) CO and CO .
4
2
main reactor results in increased concentrations of light
hydrocarbons methane and ethene, followed by a rapid falloff
at higher temperatures. This trend is consistent throughout
our data for cellulose. A similar trend, but with concentration
maxima being reached at lower temperatures of 550 °C, is
seen for two of the heavier species within this group,
acetaldehyde and furan. Concentrations of all species in this
group are below detection limits at and above 700 °C.
Product yields for CO and CO2 are illustrated in Figure 5c
as percentage of moles of carbon in the product relative to
moles of carbon in the cellulose feed. We note that CO is
used as an indicator for complete destruction of PICs in
municipal solid waste incinerators, since CO is a stable
compound. In this study CO persists at 1.99% at 700 °C, but
at our fuel-lean equivalence ratios the concentration of CO
is always below that of CO2. This implies that CO2 may be
produced directly from unimolecular decomposition of the
cellulose structure and/ or anhydro sugars at low temperature.
Complete conversion of other organics through CO to CO2
Polystyrl radical is formed by chain-end and random
â-scission (elimination reactions) at weak bonds (adjacent
to radical sites) in the polymer chain, as reported by several
researchers (42-45). Styrene is produced from â-scission of
9
VOL. 33, NO. 15, 1999 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 2 5 8 9

TABLE 2. Distribution and Fate of Vapor-Phase Products in the Oxidation of Polystyrenea
main reactortemperature (°C)
500 550 600 650 675
no. MW
products
400
450
700
750
1
2
3
4
5
6
7
8
9
16 m ethane, CH4
28 ethene, C2H4
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
42 propene, C3H6
44 acetaldehyde, C2H4O
56 butene, C4H8
56 acrolein, C3H4O
58 acetone, C3H6O
78 benzene, C6H6
92 toluene, C7H8
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
+
x
x
x
+
+
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
10 106 ethylbenzene, C8H10
11 102 phenylethyne, C8H6
12 104 styrene, C8H8
13 106 benzaldehyde, C7H6O
14 118 R-m ethylstyrene, C9H10
15 118 benzofuran, C8H10
16 122 1,3-benzodioxole, C7H6O
17 120 benzeneacetaldehyde, C8H8O
18 116 indene, C9H10
x
x
x
x
x
O
x
x
x
x
x
19 132 3-butenylbenzene, C10H12
20 148 1-phenyl-1,2-propanedione, C9H8O2
21 132 1-(m ethylenepropyl)benzene, C10H12
22 132 2,3-dihydro-1H-inden-1-one, C9H8O
23 122 benzoic acid, C7H6O2
+
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
+
x
+
+
+
x
+
+
+
x
24 128 naphthalene, C10H8
25 132 2-m ethylbenzofuran, C9H8O
x
x
x
26 134 3-phenyl-2-propen-1-ol, C9H10
27 132 3-phenyl-2-propanal, C9H8O
28 154 biphenyl, C12H10
O
x
x
x
x
x
x
29 168 diphenylm ethane, C13H12
30 180 1,1′-diphenylethylene, C14H12
31 182 bibenzyl, C14H14
32 194 R-m ethyl-(E)-stilbene, C15H14
33 180 4-ethenyl-1,1′-biphenyl, C14H12
+
x
x
x
x
x
x
x
x
x
x
x
34 182 benzophenone, C13H10
O
+
35 196 1,1′-(1,3-propanediyl)bisbenzene, C15H16
36 180 9,10-dihydrophenanthrene, C14H12
37 180 (E)-stilbene, C14H12
x
x
x
x
x
+
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
38 208 styrene, dim er, C16H16
39 194 1-m ethyl-2-(2-phenylethenyl)benzene, C15H14
40 178 anthracene, C14H10
41 206 1,4-diphenyl-1,3-butadiene, C16H14
42 208 1,1′-(1-butene-1,4-diyl)bis-(Z)-benzene, C16H16
43 210 1,1′-(1,2-ethanediyl)bis-4-m ethylbenzene, C16H18
44 234 cis-4b,5,6,10b,11,12-hexahydrochrysene, C18H18
45 204 2-phenylnaphthalene, C16H12
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
46 208 1,3-diphenyl-(E)-2-propen-1-one, C15H12
O
47 222 1-(4-m ethylphenyl)-3-phenyl-2-propen-1-one, C16H14
48 236 trans-1,2-dibenzoylethylene, C16H12O2
49 312 styrene, trim er, C24H24
O
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
50 306 quaterphenyl, C24H18
x
x
x
x
51
52
28 CO
44 CO2
x
x
x
x
x
x
fuel equivalence ratio (φ)
carbon m ass balance (%)
0.07 0.07 0.07 0.08 0.09 0.10 0.10 0.10 0.11
77.56 78.02 91.17 95.36 92.41 88.59 85.95 88.32 93.73
a
Volatilization oven tem perature 400 °C, m ain reactor tem perature as indicated; x signifies detected by GC and GC/MS analysis, + signifies
detected by GC/MS and not detected by GC.
chain-end polystryl radical in propagation reactions. Dimer,
trimer, and oligomers of styrene result from intramolecular
hydrogen transfers and â-scission reactions. Relative con-
centrations observed in initial vapor-phase products are
monomer >trimer >dimer at 400 °C, as shown quantitatively
in Figure 7a. These data are consistent with trends reported
by previous researchers (29-31). Maximum product yields
for styrene (46.40%) and benzaldehyde (11.89%) are observed
at 550 °C. The maxima for styrene trimer (17.51%) and dimer
(4.27%) occurred at 400 °C. These major volatilization
products of polystyrene are reduced to levels of 0.03% at 600
°C for the trimer, while at 675 °C styrene monomer is 0.96%
and the dimer is 0.15%.
2,3-Dihydro-1H-inden-1-one is an oxygenated species that
appears in relatively high levels at low temperatures and
then decreases in intensity at 650 °C. Stilbene, 1-methyl-2-
(2-phenylethenyl)benzene, 1,3-diphenyl-2-propen-1-one, and
quaterphenyl are also present in the temperature range 400-
600 °C but are below detection limits by 675 °C.
Benzene and ethene follow patterns shown in Figure 7b,c;
maximum benzene levels of 5.71% occur at 650 °C. C2 to C4
compounds show similar results to that of benzene, while
ethene is 2.24% at 675 °C. Benzofuran persists at 675 °C,
along with naphthalene, 1,1-diphenylethylene, and 2-phen-
ylnaphthalene; these species survive through 675 °C, but only
at barely detectable levels. Benzene level decreased rapidly
9
2 5 9 0 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 33, NO. 15, 1999

FIGURE 7. Product yields versus main reactor temperature in the thermal oxidation of polystyrene: (a) styrene monomer, dimer, and trimer
and benzaldehyde, (b) C to C aromatic hydrocarbons, (c) C to C light hydrocarbons, (d) CO and CO .
6
9
1
4
2
above 650 °C and was below detection limits at 700 °C and
above.
polystyrene volatilization and oxidation result in aromatic
and substituted aromatics. Volatilization products from
thermal pyrolysis or oxidation of cellulose and polystyrene
are shown to achieve complete combustion to CO2 and H2O
at temperature of 750 °C under the reaction conditions of
this study.
Continuous feed, steady-state reactor operation enables
the monitoring of polymer volatilization (initial fragmentation
and decomposition), initial and intermediate polymer reac-
tion products, and their concentrations as a function of
temperature. These data supplement knowledge of the
reaction process provided by batch or drop-tube furnace
reactors. It is of value in interpretation of initial volatilization
species, both qualitative distribution and concentrations. It
also provides quantitative data on fate or conversion versus
temperature.
Carbon mass balance at 650 °C is 88.59% and decreases
to 85.95% at 675 °C, which is an important temperature for
polystyrene oxidation in this study. At this temperature, the
levels of more common polyaromatic hydrocarbons, such as
indene, naphthalene, biphenyl, benzofuran, and anthracene,
are diminished, i.e., present only in very low concentration.
High-molecular-weight PAHs with fused ring structures are
not observed above 675 °C.
Carbon monoxide and carbon dioxide yields versus main
reactor temperature are shown for polystyrene in Figure 7d.
CO and CO2 yields are lower for polystyrene than for cellulose
up to reactor temperatures near 650 °C. This is a consequence
of higher stability in the polystyrene volatilization products
due to conjugation in the styrene and the absence of oxygen
in the rings and side chains. Although CO converts to CO2
at temperatures in the range 650-700 °C for both cellulose
and polystyrene, the conversion is less rapid for polystyrene,
suggesting a less oxidative radical pool environment. Com-
plete conversion of CO to CO2 occurs at a temperature of 750
°C for both polystyrene and cellulose.
We suggest the use of total hydrocarbon and CO or CO2
as a continuous monitor relevant to the attainment of uniform
operation in these studies. Total HC is a valid monitor of the
steady state when operating at low conversion, and CO or
CO2, especially if calibrated for mass balance, is appropriate
to higher conversion.
Overall Polym er Oxidation: Cellulose and Polystyrene.
The overall effect of increasing temperature from 400 to 850
°C with other parameters fixed is to reduce higher-molecular-
weight intermediates from polymer decomposition for both
cellulose and polystyrene, under conditions of fuel-lean
equivalence ratio and 2.0-s reaction time. An initial increase
in levels of lower-molecular-weight products is observed,
followed by their conversion to CO2 and H2O as the
temperature of complete combustion is approached.
The decomposition products from the two polymers are
complex and are representative of the different polymer
structures and their respective pathways of decomposition.
Cellulose decomposition results in a large variety of complex
organic hydrocarbons and oxyhydrocarbon species, while
Acknowledgments
This work was supported in part by a grant from the Procter
and Gamble Co.
Literature Cited
(1) U.S. EPA. Characterization of Municipal Solid Waste in the United
States: 1996 Update; EPA530-R-97-015.
(2) Serio, M. A.; Wojtowicz, M. A.; Charpenay, S. Pyrolysis. Ency-
clopedia of Energy, Technology, and the Environment; Wiley:
New York, 1995; pp 2281-2308.
(3) Brunner, C. R. Handbook of Incineration Systems; McGraw-
Hill: New York, 1991.
(4) Kilgroe, J. D.; Nelson, L. P.; Schindler, P. J.; Lanier, W. S. Combust.
Sci. Technol. 1990, 74, 223-244.
9
VOL. 33, NO. 15, 1999 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 2 5 9 1

(5) Gurman, J. L.; Baier, L.; Levin, B. C. Fire Mater. 1987, 11, 109-
130.
(27) Elomaa, M.; Saharinen, E. J. Appl. Polym. Sci. 1991, 42, 2819-
2824.
(6) Suuberg, E. M.; Milosavljevic, I.; Oja, V. Twenty-Sixth Symposium
(International) on Combustion; The Combustion Institute:
Pittsburgh, PA, 1996; pp 1515-1521.
(7) Milosavljevic, I.; Suuberg, E. M. Ind. Eng. Chem. Res. 1995, 34,
1081-1091.
(8) Serio, M. A.; Charpenay, S.; Bassilakis, R.; Solomon, P. R. Biomass
Bioenergy 1994, 7, 107-124.
(9) Green, A. E. S.; Zanardi, M. Int. J. Quant. Chem. 1998, 66, 219-
227.
(10) Pastorova, I.; Botto, R. E.; Arisz, P. W.; Boon, J. J. Carbohydr. Res.
1994, 262, 27-47.
(11) Keating, E. L.; Gupta, A. K. Combustion Modeling, Cofiring and
Nox Control; FACT-Vol. 17, ASME 1993; pp 33-40.
(12) Pouwels, A. D.; Eijkel, G. B.; Boon, J. J. J. Anal. Appl. Pyrolysis
1989, 14, 237-280.
(28) You, J.; Chiang, P.; Chang, S.; Wang-Wuu, S. J. Hazard. Mater.
1996, 48, 69-82.
(29) Brauman, S. K.; Chen, I. J.; Matzinger, D. P. J. Appl. Polym. Sci.
1983, 21, 1831-1845.
(30) Ide, S.; Ogawa, T.; Kuroki, T.; Ikemura, T. J. Appl. Polym. Sci.
1984, 29, 2561-2571.
(31) Cascaval, C. N.; Straus, S.; Brown, D. W.; Florin, R. E. J. Polym.
Sci. 1976, 57, 81-88.
(32) Shapi, M. M.; Hesso, A. J. Anal. Appl. Pyrolysis 1990, 18, 143-
161.
(33) Carniti, P.; Gervasini, A.; Beltrame, P. L.; Audisio, G.; Bertini, F.
Appl. Catal. A 1995, 127, 139-155.
(34) Beltrame, P. L.; Bergamasco, L.; Carniti, P.; Castelli, A.; Bertini,
F.; Audisio, G. J. Anal. Appl. Pyrolysis 1997, 40-41, 451-461.
(13) Jakab, E.; Liu, K.; Meuzelaar, H. L. C. Ind. Eng. Chem. Res. 1997,
36, 2087-2095.
(14) Shafizadeh, F. J. Anal. Appl. Pyrolysis 1982, 3, 283-305.
(15) Piskorz, J.; Radlein, D.; Scott, D. S. J. Anal. Appl. Pyrolysis 1986,
9, 121-137.
(16) Funazukuri, T.; Hudgins, R. R.; Silveston, P. L. Ind. Eng. Chem.
Process Des. Dev. 1986, 25, 172-181.
(17) Encinar, J. M.; Beltran, F. J.; Bernalte, A.; Ramiro, A.; Gonzalez,
J. F. Biomass Bioenergy 1996, 11, 397-409.
(18) McKenzie, L. M.; Hao, W. M.; Richards, G. N.; Ward, D. E. Environ.
Sci. Technol. 1995, 29, 2047-2054.
(19) Lipari, F.; Dasch, J. M.; Scruggs, W. F. Environ. Sci. Technol.
1984, 18, 326-330.
(20) Williams, P. T.; Besler, S. Fuel 1993, 72, 151-159.
(21) Pakdel, H.; Roy, C. In Pyrolysis Oils From Biomass; Soltes, E. J.,
Milne, T. A., Eds.; ACS Symposium Series No. 376; American
Chemical Society: Washington, DC, 1988; pp 203-219.
(22) Wheatley, L.; Levendis, Y. A.; Vouros, P. Environ. Sci. Technol.
1993, 27, 2885-2895.
(23) Panagiotou, T.; Levendis, Y. A.; Carlson, J.; Dunayevskiy, Y. M.;
Vouros, P. Combust. Sci. Technol. 1996, 116-117, 91-128.
(24) Panagiotou, T.; Levendis, Y. A.; Carlson, J.; Vouros, P. Twenty-
Sixth Symposium (International) on Combustion; The Combus-
tion Institute: Pittsburgh, PA, 1996; pp 2421-2430.
(25) Durlak, S. K.; Biswas, P.; Shi, J.; Bernhard, M. J. Environ. Sci.
Technol. 1998, 32, 2301-2307.
(35) Carniti, P.; Beltrame, P. L.; Armada, M.; Gervasini, A.; Audisio,
G. Ind. Eng. Chem. Res. 1991, 30, 1624-1629.
(36) McLafferty, F. W.; Turecek, F. Interpretation of Mass Spectra,
4th ed.; University Science: Sausalito, CA, 1993.
(37) NIST. NIST 92 Library; National Institute of Standards and
Technology: Gaithersburg, MD.
(38) Pouwels, A. D.; Tom, A.; Eijkel, G. B.; Boon, J. J. J. Anal. Appl.
Pyrolysis 1987, 11, 417-436.
(39) Helleur, R. J. J. Anal. Appl. Pyrolysis 1987, 11, 297-311.
(40) Arisz, P. W.; Lomax, J. A.; Boon, J. J. Anal. Chem. 1990, 62, 1519-
1522.
(41) Radlein, D.; Grinshpun, A.; Piskorz, J.; Scott, D. S. J. Anal. Appl.
Pyrolysis 1987, 12, 39-49.
(42) Bouster, C.; Vermande, P.; Veron, J. J. Anal. Appl. Pyrolysis 1989,
15, 249-259.
(43) Audisio, G.; Bertini, F. J. Anal. Appl. Pyrolysis 1992, 24, 61-74.
(44) Schroder, U. K. O.; Ebert, K. H.; Hamielec, A. W. Makromol.
Chem. 1984, 185, 991-1001.
(45) Bockhorn, H.; Hornung, A.; Hornung, U. Twenty-Seventh
(International) Symposium on Combustion; The Combustion
Institute: pp 1343-1350.
Received for review August 4, 1998. Revised manuscript
received April 1, 1999. Accepted April 19, 1999.
(26) Hawley-Fedder, R. A.; Parsons, M. L.; Karasek, F. W. J. Chro-
matogr. 1984, 315, 201-210.
ES980796Q
9
2 5 9 2 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 33, NO. 15, 1999