Evaluation of acute toxicity and genotoxicity of liquid products from pyrolysis of Eucalyptus grandis wood

Article abstract of DOI:10.1007/s002449910022

Slow pyrolysis of Eucalyptus grandis wood was performed in an oven laboratory, and smoke was trapped and condensed to yield liquid products. Polycyclic aromatic hydrocarbons (PAHs) and phenolic fractions were isolated from the former liquid products using adsorption column chromatography (ACC) and identified by GC/MS. Concentrations of PAH and phenolic fractions in total pyrolysis liquids were respectively 48.9 μg/g and 8.59% (w/w). Acute toxicity of total samples of pyrolysis liquids and the phenolic fraction was evaluated by means of two bioassays, namely, 24-h immobilization bioassay with Daphnia magna and Microtox(TM) bioassays, the latter employing the luminescent bacteria Photobacterium phosphoreum. Total pyrolysis liquids and the PAH fraction were evaluated for genotoxicity by the Microtox(TM) bioassay conducted using rehydrated freeze-dried dark mutant of the luminescent bacteria Vibrio fisheri strain M169. Total pyrolysis liquids and the phenolic fraction, respectively, in concentrations of 170 and 68 mg/L were able to immobilize 50% (EC₅₀) of the D. magna population following 24-h exposure. Concentrations of 19 and 6 mg/L, respectively, for total pyrolysis liquids and phenolic fraction were the effective concentrations that resulted in a 50% (EC₅₀) reduction in light produced by bacteria in the Microtox(TM) bioassay. Accordingly, the Microtox(TM) bioassay was more sensitive to toxic effects of both kind of samples than the D. magna bioassay, particularly for the phenolic fraction. Regarding to the genotoxicity evaluation, the results achieved by Microtox(TM) bioassay showed that total pyrolysis liquids had no genotoxic effects with and without exogenous metabolic activation using rat liver homogenate (S9). However, the PAH fraction showed toxic effects with rat liver activation and had a dose-response number (DRN) equal to 1.6, being in this way suspected genotoxic. The lowest detected concentration (LDC) of the PAH fraction able to cause genotoxic effects was 375 μg/L.

Full text of DOI:10.1007/s002449910022

Arch. Environ. Contam. Toxicol. 38, 169–175 (2000)
DOI: 10.1007/s002449910022
A R C H I V E S O F
E
Cnovnitraomnmineanttioanl
a n d
T
oxicology
2000 Springer-Verlag New York Inc.
௠
Evaluation of Acute Toxicity and Genotoxicity of Liquid Products from Pyrolysis
of Eucalyptus grandis Wood
A. S. Pimenta,1 J. M. Bayona,2 M. T. Garc´ıa,3 A. M. Solanas4
1
Department of Forestry Engineering, Laboratory of Wood Energy and Technology, Federal University of Vic¸osa, Campus Universita´rio, 36571-000,
Vic¸osa (MG), Brazil
2
Department of Environmental Chemistry, IIQAB-CID-CSIC, Jordi Girona 18, E-08034 Barcelona, Spain
Department of Surfactant Technology, IIQAB-CID-CSIC, Jordi Girona 18, E-08034 Barcelona, Spain
Department of Microbiology, University of Barcelona, Avda, Diagonal, 645, 08028-Barcelona, Spain
3
4
Received: 20 April 1999/Accepted: 27 July 1999
Abstract. Slow pyrolysis of Eucalyptus grandis wood was
performed in an oven laboratory, and smoke was trapped and
condensed to yield liquid products. Polycyclic aromatic hydro-
carbons (PAHs) and phenolic fractions were isolated from the
former liquid products using adsorption column chromatogra-
phy (ACC) and identified by GC/MS. Concentrations of PAH
and phenolic fractions in total pyrolysis liquids were respec-
tively 48.9 µg/g and 8.59% (w/w). Acute toxicity of total
samples of pyrolysis liquids and the phenolic fraction was
evaluated by means of two bioassays, namely, 24-h immobiliza-
tion bioassay with Daphnia magna and Microtox௢ bioassays,
the latter employing the luminescent bacteria Photobacterium
phosphoreum. Total pyrolysis liquids and the PAH fraction
were evaluated for genotoxicity by the Microtox௢ bioassay
conducted using rehydrated freeze-dried dark mutant of the
luminescent bacteria Vibrio fisheri strain M169. Total pyrolysis
liquids and the phenolic fraction, respectively, in concentrations
of 170 and 68 mg/L were able to immobilize 50% (EC50) of the
D. magna population following 24-h exposure. Concentrations
of 19 and 6 mg/L, respectively, for total pyrolysis liquids and
phenolic fraction were the effective concentrations that resulted
in a 50% (EC50) reduction in light produced by bacteria in the
Microtox௢ bioassay. Accordingly, the Microtox௢ bioassay
was more sensitive to toxic effects of both kind of samples than
the D. magna bioassay, particularly for the phenolic fraction.
Regarding to the genotoxicity evaluation, the results achieved
by Microtox௢ bioassay showed that total pyrolysis liquids had
no genotoxic effects with and without exogenous metabolic
activation using rat liver homogenate (S9). However, the PAH
fraction showed toxic effects with rat liver activation and had a
dose-response number (DRN) equal to 1.6, being in this way
suspected genotoxic. The lowest detected concentration (LDC)
of the PAH fraction able to cause genotoxic effects was 375
µg/L.
Near 6.8 million tons of charcoal are consumed every year in
Brazil, and 54% of this overall amount is produced by using
wood from Eucalyptus planted forests as raw material (ABRA-
CAVE 1997), basically from the species E. grandis. Metalwork
sector, including pig iron, steel, and iron alloy industries,
consume 85% of overall charcoal, constituting the so-called
tropical metallurgy. Most industrial wood carbonization plants
currently operating in Brazil are oriented to charcoal production
by means of partial-combustion masonry kilns that do not
recover the by-product gases and condensable compounds. In
fact, mass and energy yields are very poor in the conventional
kilns, about 33 and 50%, respectively, especially when by-
products are not recovered for recycling (Resende et al. 1994).
Normally, in the slow pyrolysis process, the losses as smoke
based on initial wood mass can reach about 50–55% of the
initial carbon, 75–52% of the initial hydrogen, and 87–90% of
the initial oxygen as smoke (Wenzl 1970;Almeida 1982). When
released from industrial plants to the environment, pyrolysis
volatiles promote substantial air, ground, and water pollution.
Contaminants occurring in the pyrolysis smoke are divided into
three main groups: suspended solids, noncondensable gases,
and condensable organic compounds. The chemical nature of
the compounds found in the last group of pollutants has been
increasingly studied because a better understanding of the
characteristics and properties of such effluents is important to
develop suitable devices for reducing pollutant emissions, such
as condensers, burners, or filtering systems.
During thermal decomposition of woody biomass, part of the
smoke from the pyrolysis bed can be condensed, yielding a
liquid that after standing separates to two phases: an aqueous
phase containing alcohols, ketones, and other low-molecular-
weight volatile compounds, and an oily phase–denominated
wood tar that is composed mainly by moisture, wood creosote
(phenols mixture), and 30–50% of a polymeric pitch (Carazza
et al. 1991). Apart from water, wood tar contains condensable
semivolatile and nonvolatile compounds with molecular weights
varying from acetic acid (MW ϭ 60) to tar pitch (MW Ͼ 500).
More than 400 different compounds have been identified in
pyrolysis tars (Pakdel and Roy 1988, 1991; Desbe`ne et al.
1991), and several of them, such as the polycyclic aromatic
Correspondence to: J. M. Bayona

170
A. S. Pimenta et al.
450°C (about 8 h). Pyrolysis liquids were obtained by trapping the
smoke from the pyrolysis bed through a metal condenser at room
temperature. Three replications, with sample size about 200 g, were
performed, and the resulting liquid products were mixed and stored at
Ϫ20°C. Except as indicated, all solvents and reagents (pesticide-
analysis grade) used in this work were purchased from Merck
(Darmstadt, Germany).
hydrocarbons (PAHs), have a strong environmental impact
owing to their mutagenic and carcinogenic character (Kaden et
al. 1979; Gold et al. 1989). In addition, some PAHs have high
acute toxicity and can be detected by means of the photobacte-
ria assays (Kagan et al. 1985; Newsted and Giesy 1987). In the
literature, there are several reports on hazardous and toxic
effects of wood smoke to respiratory human health, reflected as
different kinds of airway diseases and caused by exposure to
different sources, e.g., woodstoves, fireplaces, oven smoke
(EHD 1997).
Isolation of PAH and Phenolic Fractions
In Brazil, workers involved in industrial charcoal production
are continuously exposed to smoke; despite the reasonable
knowledge available on chemicals found in Eucalyptus pyroly-
sis liquids, there is no report on evaluating either toxic or
genotoxic effects of such chemicals on living organisms.
Identification of specific chemical compounds able to promote
deleterious effects on human health or ecological effects is a
major research goal in environmental and health sciences.
However, the high complexity of the wood tars limits the
possibility of direct analysis and identification of components
using a single technique; thus, analytical procedures commonly
used frequently comprise the need of sample fractionation
before chromatographic or spectroscopic analysis (Brage and
Sjo¨stro¨m 1991; Pimenta et al. 1998). Isolation and identifica-
tion of wood tar–derived compounds have been accomplished
by fractionation techniques, commonly adsorption column
chromatography or solid-phase extraction followed by GC,
GC/MS, LC, or NMR characterization. Recently, supercritical
fluid extraction followed by GC/MS allowed a fast determina-
tion and quantification of PAHs in liquid products from
pyrolysis of E. grandis wood (Pimenta et al. 1998).
An ACC procedure for isolating both PAH and phenolic fractions from
pyrolysis liquids was performed by using a pair of glass columns.
Figure 1 shows the experimental scheme employed to isolate PAH and
phenolic fractions. After ACC isolation, GC/MS analysis and acute
toxicity and genotoxicity evaluations were carried out. The first
column, packed with anhydrous sodium sulfate, was used to dehydrate
the samples and retain part of the tar pitch. A second column, packed
with neutral aluminum oxide, was used to retain the remaining pitch tar
and fractionate the extract from the first column. Four replications of
theACC procedure were carried out to assess the method precision, and
after partitioning, total samples were spiked with 0.5 µg/g of perdeuter-
ated pyrene from CIL (Woburn, MA) as internal standard. However,
the fractions evaluated by bioassays were obtained without previous
addition of the internal standard.
For column chromatography fractionation, initially the sample
spiked with the internal standard was quantitatively transferred in a
glass column filled with 6 g anhydrous sodium sulfate (previously
activated at 400°C for 12 h). After transferring the sample, the column
was eluted with 35 ml methylene chloride. The extract from the first
column was evaporated to 1 ml in a rotary evaporator at 35°C and
Coupling short-time bioassays and chemical analysis has
been successfully used for the characterization of toxic compo-
nents in environmental samples (Grifoll et al. 1990, 1992).
Short-term bioassays constitute valuable tools to evaluate toxic
effects of many classes of pollutants to living organisms.
Among the most used bioassays, Microtox௢ and the 24-h
immobilization bioassay with Daphnia magna show good
sensitivity and are useful to evaluate the toxicity of many
classes of chemical contaminants (Dutka et al. 1983; Giesy et
al. 1988). Microtox௢ is a bioassay with luminescent bacteria,
which is a rapid screening alternative to standard acute toxicity
bioassays using fishes or invertebrates (Giesy et al. 1988). In
the present work, total samples of liquids from pyrolysis of E.
grandis wood and some chemical fractions isolated by ACC
have been evaluated for the acute toxicity and genotoxicity.
Specifically, the objectives of this work were: (1) to determine
the PAH and phenolic content in liquids from pyrolysis of E.
grandis wood by using adsorption column chromatography
(ACC) procedures followed by GC/MS determination; (2) to
evaluate the acute toxicity of the total pyrolysis liquids and the
phenolic fraction by means of two bioassays; and (3) to
evaluate the genotoxicity of the total pyrolysis liquids and the
PAH fraction using the Microtox௢ bioassay.
Materials and Methods
Debarked 7-year-old E. grandis wood was obtained from plantations
located at Vic¸osa (MG, Brazil). Wood chips were pyrolyzed in a
laboratory oven (Fabbe௢, Sa˜o Paulo, Brazil) with an average heating
rate of 56.3°C/h from room temperature reaching a final temperature of
Fig. 1. Isolation of chemical fractions from pyrolysis liquids followed
by GC/MS determination and acute toxicity and genotoxicity evalua-
tion assays

Acute Toxicity and Genotoxicity of Liquid Pyrolysis of E. grandis
171
loaded onto a second glass column filled with 9 g of neutral aluminium
oxide (70–230 mesh, previously activated for 12 h at 400°C). The
second column was conditioned prior to use by gravity feed with 10 ml
of n-hexane and after addition of the extract from the first column, a
15-ml n-hexane fraction was eluted and discarded. A second fraction
was eluted with 30 ml of a methylene chloride/n-hexane mixture (95:5
v/v). This fraction was collected and evaporated almost to dryness in a
rotary evaporator. A third fraction containing the phenolic fraction was
eluted from the column with 30 ml of a methylene chloride/methanol
mixture (1:1 v/v). After solvent evaporation, the PAH and phenolic
containing fractions were redissolved to 1 ml volume with ethyl
acetate. Aliquots were collected from these solutions and stored for
further chromatographic analysis.
This bacterial bioassay is based on the reduction of light emitted by a
strain of luminescent marine bacterium, P. phosphoreum, on exposure
to a toxic sample. The system has been developed commercially under
the trademark Microtox௢ by Microbics Corporation (Carlsbad, CA).
Intensity of light output depends on several external factors, including
temperature, pH, salinity, nature and concentration of the toxic
compound, and so on. As with most other bioassays for aquatic toxicity,
the bioassay is nonspecific and determines the combined toxic effects
of all components in a mixture. Lyophilized P. phosphoreum cells were
reconstituted for testing. Microtox௢ toxicity data are based on a
30-min exposure of the bacteria to the chemical solution at 15°C.
Toxicity results have been reported as the effective concentration
promoting a 50% (EC50) reduction in light emitted by the bacteria.
Genotoxicity Bioassay
Analysis of PAH and Phenolic Fractions
Genotoxicity assay, Mutatox௢ (Microbics Corporation, Carlsbad, CA)
was conducted according to standard procedures (Mutatox 1993), using
rehydrated freeze-dried dark mutant of the luminescent bacteria Vibrio
fisheri strain M 169 (Bulich 1993). Different genotoxic mechanisms,
such as base substitution, frameshift, DNA synthesis inhibition, DNA
damage, and DNA intercalation agents, result in appearance of light in
Mutatox௢ strain M 169. In this way, genotoxic effects can be detected
by measuring the ability of a specific chemical to restore the
luminescent state in the bacterial cells. This bioassay can be conducted
with and without exogenous metabolic activation using a rat liver
homogenate (S9). Total pyrolysis liquids and PAH fraction isolated
from the former by ACC were evaluated for genotoxicity. Microtox௢
bioassay procedure included an initial dilution of the sample to 50% in
Microtox௢ medium followed by eight additional dilutions using the
method for environmental samples developed and validated by Micro-
bics Corporation (Mutatox 1993). Light emitted from the bacteria in
the media controls, solvent controls, and sample dilutions were
measured after 16, 20, and 25 h of incubation at 27°C. Changes in
luminescence were monitored by the Microbics 500 Analyzer. All
samples were assayed with and without exogenous metabolic activa-
tion using a rat liver homogenate (S9). Positive and negative controls
were used to assess of the assay sensitivity and consisted of a 200 mg/L
phenol solution in methanol as a control of direct genotoxins without
S9 fraction, and 50 µg/ml 2-nitrosoguanidine solution in dimethylsulf-
oxide (DMSO) as a control for progenotoxins with S9 fraction.
The degree of light increase in Microtox௢ bioassay indicated the
relative genotoxicity, which was evaluated according to the criteria
established by Johnson (1992). The response was considered positive
when the light output was at least 100 and at least three times the
average light output of the negative control. Dose-response number
(DRN) was defined as the number of positive responses recorded at
different concentrations per dilution series and the mean DRN from the
triplicate study (16, 20, and 24 h) was also calculated. The lowest
detected concentration (LDC) was defined as the lowest positive
genotoxic response per dilution series. Taking into account these
values, a sample was classified genotoxic when the mean DRN in
triplicate series was more or equal to 3 and the total number of different
Analysis of the PAH fraction was performed by GC/MS with a Fisons
MD 800 instrument (Milan, Italy) in the electron impact (EI) mode (70
eV) operating in the selected ion recording (SIR) mode (mass range:
50–400 D). The following PAH molecular ions were monitored: 166
(fluorene), 168 (dibenzofuran), 178 (phenanthrene and anthracene),
180 (methylfluorenes), 192 (methylphenanthrenes), 202 (fluoranthene
and pyrene), 212 (perdeuterated pyrene), 216 (methylfluoranthenes or
pyrenes), 228 (benz[a]anthracene and chrysene ϩ triphenylene), 252
(benzo[b]fluoranthene, benzo[k]fluoranthene, benzo[e]pyrene, ben-
zo[a]pyrene, and perylene), 276 (benzo[ ghi]perylene), and 278
(dibenz[ah]anthracene).
Injector, transfer line, and ion source temperatures were held at 280,
280, and 200°C, respectively. The analytical column was a 30 m length
and 0.25 mm internal diameter HP-5 (Avondale, PA) with 0.25 µm film
thickness. Column temperature was programmed from 80 to 100°C at
15°C/min and then to 310°C at 4°C/min, holding the final temperature
for 10 min. For quantification, the external standard method was used
with the PAHs included in the EPA priority pollutants list as calibrants.
The precision studies were performed with at least four independent
replicates. The standard PAHs mixture was purchased from Accu
Standard (New Haven, CT).
The phenolic fraction characterization was achieved by using the
same analytical column employed for the analysis of the PAH fraction.
However, a different temperature program was used to optimize the
separation of phenols. After an isothermal period of 1 min at 60°C, the
column temperature was increased at 2°C/min to 80°C. From this
temperature to 200°C, a heating rate of 4°C/min was used, holding the
final temperature for 10 min. For quantification, the external standard
method was used with o- and p-cresol as calibrants. The creosote
components were identified by correlating mass spectra and relative
retention times with data available in the literature (Carazza et al.
1991).
Acute Toxicity Bioassays
Two acute toxicity bioassays described in detail elsewhere were
undertaken: the 24-h immobilization bioassay with D. magna (freshwa-
ter crustacea) (OECD 1984) and the Microtox௢ bioassay (Ribo and
Kaiser 1987; Mutatox 1993), which employs the luminescent bacte-
rium Photobacterium phosphoreum. In the D. magna acute immobiliza-
tion bioassay, toxic effects on the swimming capability of the crustacea
caused by the investigated substance are determined. Bioassays were
carried out at 20°C in complete darkness. The percentage immobility at
24 h was plotted against concentration on logarithmic-probability
paper. Toxicity results have been expressed as the estimated concentra-
tion to immobilize 50% (EC50) of the D. magna population after 24 h
exposure. The luminescent bacterial bioassay, known commercially as
the Microtox௢ bioassay, is widely used in aquatic toxicity assessment.
Table 1. Products and yields obtained from pyrolysis of Eucalyptus
grandis wood
Yields
Products
(% w/w)*
Charcoal
38.5
45.5
16.0
Liquid products
Noncondensable gases
Total
100.0
* (%) w/w based on initial dry wood mass

172
A. S. Pimenta et al.
Fig. 2. Reconstructed ion chromatogram of the PAH
fraction (see analytical conditions in text). Numbers on
peak apexes correspond to compounds listed in Table 2
concentrations with positive response was at least 2; suspected
genotoxic when the mean DRN was higher than 0, but less than 3; and
finally, nongenotoxic when there was no positive genotoxic responses
in at least three triplicate series.
Table 2. Concentrations and yields of the polycyclic aromatic
hydrocarbons isolated from pyrolysis liquids of Eucalyptus grandis
wood
Concentrations
(µg/g)a
Yields
(g/ton)b
Compound
Results and Discussion
Dibenzofuran
Fluorene
⌺Methylfluorenes
Phenanthrene
Anthracene
3-methylphenanthrene
2-methylphenanthrene
4-methylphenanthrene
1-methylphenanthrene
Fluoranthene
4.52
7.03
16.1
3.92
2.10
1.43
1.88
2.33
1.14
1.05
1.21
3.16
0.53
0.51
0.24
0.18
0.19
0.24
0.14
0.32
0.31
0.33
48.9
9.6
1.85
2.88
6.58
1.60
0.86
0.58
0.77
0.95
0.47
0.43
0.49
1.29
0.22
0.21
0.10
0.07
0.08
0.10
0.06
0.13
0.13
0.13
19.98
Table 1 displays the average yields from the pyrolysis of the
wood chips as charcoal, liquid, and gaseous products (% w/w
based on initial dry wood mass). The noncondensable gases
yield was obtained by difference.
Analysis of PAH and Phenolic Fractions
Pyrene
The ACC recovery of the PAH fraction was calculated based on
the average concentration of perdeuterated pyrene in the
extracts after extraction and fractionation steps. A recovery of
115% was obtained for PAHs, with a coefficient of variation of
11.3% (four replicates). Figure 2 shows the reconstructed ion
chromatogram obtained from the PAH fraction. Table 2 shows
the identified PAHs and their respective concentrations and
yields. The concentrations of PAHs were listed in µg/g of PAH
per L of total pyrolysis liquids, and the yields listed in g of
yielded PAH per ton of pyrolyzed dry wood mass. By using the
ACC technique, a total PAH concentration of 48.9 µg/g could
be isolated from pyrolysis liquids. Together with PAHs, several
other compounds in concentrations in the range of 1 to 5 µg/g
were extracted. Among these compounds, some fatty acid esters
were identified according to their mass spectra.
⌺methylfluoranthenes or pyrenes
Benz[a]anthracene
Crysene ϩ triphenylene
Benzo[b]fluoranthene
Benzo[k]fluoranthene
Benzo[e]pyrene
Benzo[a]pyrene
Perylene
Indeno[1,2,3-cd]pyrene
Dibenzo[a]anthracene
Benzo[ghi]perylene
Total
Coefficient of variation (%) [n ϭ 4]
a µg of PAH per g of total pyrolysis liquids
bg of yielded PAH per ton of pyrolyzed dry wood mass
Figure 3 shows the total ion chromatogram obtained from the
phenolic fraction. Table 3 shows the phenolic compounds and
their respective concentrations and yields. The concentrations
were listed in g of phenolic compound per L of total pyrolysis
liquids, and the yields listed in kg of yielded phenolic com-
pound per ton of pyrolyzed wood dry mass. Phenolic com-
pounds from smoke pyrolysis of E. grandis are mainly deriva-
tives of phenol, guayacol (2-methoxyphenol), and syringol
(2,6-dimethoxyphenol) besides their methyl, ethyl, and propyl-
C4-substituted derivatives. Figure 4 shows the molecular struc-
tures of the main phenolic compounds found in the pyrolysis
liquids. As it can be pointed out, from total phenolic fraction,
about 50% corresponds to syringol (2,6-dimethoxyphenol) and
its C4-substituted derivatives.
(EC50) of the crustacea population after 24 h exposure. The total
pyrolysis liquids and isolated phenolic compounds caused such
effect, respectively, in concentrations of 170 and 68 mg/L (mg
of added wood creosote per L of water). P. phosphoreum
toxicity bioassays results also shown in Table 4 were expressed
as the effective concentration resulting in a 50% (EC50)
reduction in light produced by the bacteria after 30 min
exposure. The concentrations of 19 and 6 mg/L (mg of added
wood creosote per L of Microtox௢ medium) were obtained,
respectively, for total pyrolysis liquids and phenolic fraction.
Both bioassays showed a similar trend being the isolated
phenolic fraction the most toxic against organisms. Accord-
ingly, D. magna is more sensitive to toxic effects of the phenolic
fraction than to total pyrolysis liquids, since the concentration
of the isolated phenols is 2.5 times lesser than total liquids was
able to immobilize 50% of the crustacea population. Similarly,
P. phosphoreum in Microtox௢ bioassay was more sensitive to
phenolic fraction, of which a concentration about three times
Acute Toxicity Bioassays
D. magna toxicity test results shown in Table 4 were expressed
as the estimated concentration able to immobilise the 50%

Acute Toxicity and Genotoxicity of Liquid Pyrolysis of E. grandis
173
doses of PAHs at concentrations of 1/400 after 16, 20, and 24 h
exposure and also at the concentration of 1/800. Since the
calculated DRN was 1.6, it can be concluded that PAH mixture
isolated from E. grandis pyrolysis liquids is suspected as
genotoxic. The LDC of the PAH fraction able to cause
genotoxic effects was 375 µg/L (1/800) as µg of added PAH
fraction per L of Mutatox௢ medium. In wood pyrolysis liquids,
the PAHs and phenolic compounds are closely mixed with the
tar pitch, and only traces of PAHs, e.g., are present in the
aqueous phase (Pakdel and Roy 1991). It probably could
explain the negative genotoxic results found in evaluating total
pyrolysis liquids, since the PAHs are strongly adsorbed to the
pitch, being less bioavailable to bacteria. Another possible
explanation about the negative genotoxicity of the liquid
condensates despite PAHs present at the ppm level may be the
antagonistic interactions with other compounds. Similar effects
have been described in the Salmonella/Microsome bioassay of
complex environmental samples (Grifoll et al. 1990, 1992).
On the other hand, the genotoxicity of tars released from a
vacuum pyrolysis process has been evaluated by the Ames
bioassay, with naphthalene and anthracene as the predominant
hydrocarbons, which showed a low mutagenic activity unless a
large quantity of oils is tested (Pakdel and Roy 1991). Most of
the studies demonstrated genotoxic effects of wood smoke
associated to emissions from combustion, gasification, or
pyrolysis systems, but none of them has evaluated chemical
compounds found in the liquids from slow pyrolysis of E.
grandis wood. According to Mezerette and Girard (1991), at
least 30 out of the 77 biologically active contaminants occur-
ring in the atmosphere have been found in pyrolysis tars. In
charcoal-making processes where smoke is not recycled, all of
these pollutants are released to the atmosphere.
Fig. 3. Total ion chromatogram (TIC) of the phenolic fraction (see
analytical conditions in text). Numbers on peak apex correspond to
compounds listed in Table 3
Table 3. Composition and yields of phenolic compounds isolated from
pyrolysis liquids of Eucalyptus grandis wood
Concentrations
(g/L)a
Yields
(kg/ton)b
Compound
Butyrolactone
Phenol
Cyclotene
o-cresol
m- and p-cresol
Guayacol (2-methoxyphenol)
Maltol
0.22
0.33
0.15
0.51
1.50
1.24
0.22
0.23
1.81
2.59
14.25
8.59
8.56
13.31
9.45
0.20
3.57
0.58
13.1
17.44
0.10
0.15
0.07
0.23
0.68
0.56
0.10
0.10
0.82
1.18
6.48
3.91
3.89
6.05
4.30
0.09
1.62
0.26
5.96
7.94
2,4-xylenol
4-methylguayacol
4-ethylguayacol
Syringol (2,6-dimethoxyphenol)
4-propylguayacol
1,2,3-trimethylbenzene
4-methylsyringol
4-ethylsyringol
Combining the data from Tables 1 and 2, it was estimated that
the pyrolysis of 1 ton of E. grandis wood emits about 20 g of the
PAH mixture into the atmosphere. In Brazil, about 23 millions
of tons of wood are consumed every year by charcoal makers,
and 54% of this overall amount, about 12.5 millions of tons, is
wood from Eucalyptus plantations. Thus, it can be estimated
that at least 250 tons of the PAH mixture suspected as genotoxic
is released to the environment every year from industrial
carbonization plants.
4-allylsyringol
4-propylsyringol
3,4-dimethoxybenzoic acid
3,4,5-trimethoxybenzoic acid
Other compounds
a (%) g of compound per L of total pyrolysis liquids
b (%) kg of yielded compound per ton of pyrolyzed dry wood mass
Conclusions
PAH and phenolic fractions could be efficiently isolated from
total pyrolysis liquids employing anACC procedure, which was
simple and rapid to perform. The ACC procedure used in the
present work was able to isolate target fractions from the matrix
samples without extracting the polymeric pitch. Thus, the
extracts could be analyzed by GC/MS without any further
purification and directly used to perform acute toxicity and
genotoxicity bioassays.
lesser concentration that total pyrolysis liquids was enough to
reduce to 50% of the light emitted by bacteria. These results
suggest that the phenolic fraction is responsible for most of the
acute toxicity found in the wood smoke condensates, where it is
found in a concentration of about 8% (w/w).
Total pyrolysis liquids and isolated phenolic fraction showed
moderate toxicity by using both D. magna and Microtox௢ tests,
but the latter was more sensitive in detecting acute toxicity of
the studied samples. P. phosphoreum cells in Microtox௢ test
had more sensitivity to phenolic fraction, of which a concentra-
tion about three times lower than total pyrolysis liquids one was
enough to reduce 50% of the light emitted by bacteria. In brief,
the Microtox௢ test was more sensitive to toxic effects of both
kind of samples than the D. magna test.
Genotoxicity Bioassay
According to the criteria described in the experimental section,
it was verified that total pyrolysis liquids had no genotoxic
effects with and without exogenous metabolic activation using
rat liver homogenate (S9). Table 5 shows that genotoxic effects
could be detected for the PAH fraction with rat liver activation.
Clearly, positive response was obtained in the dilution series at

174
A. S. Pimenta et al.
Fig. 4. Molecular structures of the main phenolic com-
pounds found in pyrolysis liquids
Table 4. Acute toxicity tests resultsa
Acknowledgments. A.S.P. would like to thank the CNPq (Brazilian
Research Agency) for fellowship funding. We are indebted to Rosi
Alonso, Roser Chaler, and Roberto Alzaga for technical support,
suggestions, and comments, respectively. This work was sponsored by
the CICYT-Spain-AMB 98-0913.
Daphnia magna
24-h EC50
Microtox௢ Test
30-min EC50
(mg/L)c
Sample
(mg/L)b
Total pyrolysis liquids
Phenolic fraction
170 (105–235)
68 (33–103)
19 (15–21)
6 (5–7)
a Bracketed figures show 95% confidence ranges on EC50 values
b mg of added wood creosote per L of water
c mg of added wood creosote per L of Microtox௢ medium
References
ABRACAVE (Brazilizan Association for Renewable Forests) (1997)
Statistical yearbook. Belo Horizonte-MG, Brazil
Almeida MR (1982) Recovery of tar from masonry kilns. In: Produc¸a˜o
e utilizac¸a˜o de carva˜o vegetal. CETEC Press, Belo Horizonte,
Brazil
Table 5. Mutatox௢ test results obtained in evaluating the PAH
fraction
Brage C, Sjo¨stro¨m K (1991) Separation of phenols and aromatic
hydrocarbons from biomass tar using aminopropylsilane normal-
phase liquid chromatography. J Chromatogr 538:303–310
Bulich AA (1993) The mode of action by genotoxic agents in the
restoration of light in Mutatox௢ test system. Microbics Corpora-
tion, Carlsbad, CA
Carazza F, Rezende EA, Pasa VMD, Lessa A (1991) Fractionation of
wood tar. Proceedings of the Conference on Advances in Thermo-
chemical Biomass Conversion, Interlaken, Switzerland
Desbe`ne PL, Essayegh M, Desmazieres B, Basselier JJ (1991) Contri-
bution to the analytical study of biomass pyrolysis oils. In:
Bridgewater AV, Grassi G (eds) Biomass pyrolysis liquids up-
grading and utilisation. Elsevier Applied Science, London, pp
155–175
After S9
Activation
Parameter
Direct
0
Dose response number (ϮSD)
Lowest detected concentration
(µg/L)a
Time (h)b
1.6 Ϯ 0.58
0
0
375
16.2
Evaluation
Nongenotoxic Suspected genotoxic
a µg of added PAH fraction per L of Mutatox௢ medium
bExposure time showing the maximum DRN is reported
In evaluating the genotoxicity, Mutatox௢ test with rat liver
activation could detect genotoxic effects of the PAH fraction,
while the total pyrolysis liquids did not show any effect with or
without activation. The obtained LDC of the PAH fraction able
to cause genotoxic effects was 375 µg/L.
Dutka BJ, Nyholm N, Petersen J (1983) Comparison of several
microbiological toxicity screening tests. Wat Res 17:1363–1368
EHD (Environmental Health Division, US Agency for International
Development) (1997) An annotated bibliography on acute respira-
tory infections (ARI) and indoor air pollution. Washington, DC

Acute Toxicity and Genotoxicity of Liquid Pyrolysis of E. grandis
175
Giesy JR, Graney RL, Newsted JL, Rosiu CL, Benda A, Kreis RG Jr,
Horvath FJ (1988) Comparison of three sediment bioassay meth-
ods using Detroit river sediments. Environ Toxicol Chem 7:483–
498
Gold LS, Slone TS, Bernstein L (1989). Summary of carcinogenic
potency and positivity for 492 rodent carcinogens in the carcino-
genic potency database. Environ Health Persp 79:259–272
Grifoll M, Solanas AM, Bayona JM (1990) Characterization of
genotoxic components in sediments by mass spectrometric tech-
niques combined with Salmonella/Microsome Test. Arch Environ
Contam Toxicol 19:175–184
Grifoll M, Solanas AM, Bayona JM (1992) Bioassay directed chemical
characterization of genotoxic agents in the dissolved and particu-
late water phases of the Besos and Llobregat Rivers (Barcelona,
Spain). Arch Environ Contam Toxicol 23:19–25
Johnson BT (1992) An evaluation of a genotoxic assay with liver S-9
activation and luminescent bacteria for detection. Environ Toxicol
Chem 11:473–480
Kaden DA, Hites RA, Thilly WG (1979) Mutagenicity of soot and
associated polycyclic aromatic hydrocarbons to Salmonella typhi-
murium. Cancer Res 39:4152–4159
Kagan JA, Kagan ED, Kagan IA, Kagan PA, Quigley S (1985) The
photoxicity of non-carcinogenic PAHs in aquatic organisms.
Chemosphere 14:1829–1834
Mezerette C, Girard P (1991) Environmental aspects of gaseous
emissions from wood carbonisation and pyrolysis processes. In:
Bridgewater AV, Grassi G (eds) Biomass pyrolysis liquids upgrad-
ing and utilisation. Elsevier Applied Science, London, pp 263–287
Mutatox௢ (1993) Mutatox manual. Microbics Corporation, Carlsbad,
CA, pp 1–17
Newsted JL, Giesy JP (1987) Predictive models for photoinduced acute
toxicity of PAHs to Daphnia magna. Env Tox Chem 6:445–461
OECD (1984) Guidelines for testing of chemicals, vol. 1, section 2:
Effects on biotic systems, Daphnia magna acute immobilisation
Test 202. OECD, Paris
Pakdel H, Roy C (1988) Chemical characterization of wood pyrolysis
oils obtained in a vacuum pyrolysis multiple-hearth reactor.
American Chemical Society Symposium Series, chapter 19, 203–
219
Pakdel H, Roy C (1991) Hydrocarbon content of liquid products and tar
from pyrolysis and gasification of wood. Energy & Fuels 5:427–
436
Pimenta AS, Vital BR, Bayona JM, Alzaga R (1998) Characterisation
of polycyclic aromatic hydrocarbons in liquid products from
pyrolysis of Eucalyptus grandis by supercritical fluid extraction
and GC/MS determination. Fuel 77:1133–1139
Resende MEA, Pasa VMD, Lessa A (1994) Continuous charcoal
production—a Brazilian experience. In: Bridgewater AV (ed)
Advances in thermochemical conversion of biomass, vol. 2.
Blackie Academic & Professional, London, pp 1289–1298
Ribo LM, Kaiser KLE (1987) Phobacterium phosphoreum toxicity
bioassay: 1. Test procedures and applications. Tox Assess Int Q
2:305–323
Wenzl HFJ (1970) The chemical technology of wood. John Wiley &
Sons, New York