Journal of Chromatography A, 1108 (2006) 176–182

Pressurized liquid extraction of polychlorinated dibenzo-p-dioxins,

polychlorinated dibenzofurans and coplanar polychlorinated
biphenyls from contaminated soil
Osamu Kiguchi a,b,∗ , Takashi Kobayashi b , Katsumi Saitoh b , Nobuaki Ogawa a
aFaculty of Engineering and Resource Science, Akita University, 1-1 Tegata Gakuen-cho, Akita 010-8502, Japan
b Chemical Substance Section, Environmental Research and Information Center of Akita Prefecture, 3-1-1 San-no, Akita 010-8572, Japan
Received 30 October 2005; received in revised form 3 January 2006; accepted 6 January 2006

Abstract
Extraction solvents for pressurized liquid extraction (PLE) used to extract polychlorinated dibenzo-p-dioxins, polychlorinated dibenzofurans
(PCDD/PCDFs), and coplanar polychlorinated biphenyls (Co-PCBs) from contaminated soil were investigated. The PCDD/PCDFs and Co-PCBs
in Certified Reference Material: CRM 0422 (Forest soil) were extracted using toluene, n-hexane, acetone, acetone/toluene and acetone/n-hexane
(1:1, v/v). Soxhlet extraction was the reference method. Results demonstrated that PLE using mixed solvents produced better analyte recoveries
than the single solvents. However, these results were lower than those for Soxhlet extraction. Additional extraction cycles using mixed solvents
achieved better recovery results. Mixed solvents and several extraction cycles were necessary for satisfactory extraction of more tightly bound
PCDD/PCDFs and Co-PCBs from soil.
© 2006 Elsevier B.V. All rights reserved.

Keywords: Extraction method; Pressurized liquid extraction (PLE); Soil; Extraction solvent; Extraction efficiency; Persistent organic pollutant

1. Introduction One developed extraction method is pressurized liquid

Polychlorinated dibenzo-p-dioxins (PCDDs), polychlori-
nated dibenzofurans (PCDFs), and polychlorinated biphenyls
(PCBs) are well known as persistent and highly toxic organic
pollutants (POPs) [1]. Chemical analyses of these compounds
are complicated and time-consuming processes that include
extraction, multi-step cleanup and fractionation, and analysis
using gas chromatography/mass spectrometry (GC/MS).
Automation or semi-automation of analytical operations
has attracted great interest to facilitate analytical processes.
Various extraction methods to improve extraction times and
solvent volumes have been developed because the most com-
monly used extraction method for POPs, Soxhlet extraction,
typically requires 16–36 h and large amounts of organic
solvents (200–500 ml), which are also hazardous to human
health.
∗ Corresponding author. Tel.: +81 18 860 4010; fax: +81 18 860 4016.
E-mail address: bxc00120@nifty.com (O. Kiguchi).
extraction (PLE), which has been used as an effective extrac-
tion technique and is comparable to Soxhlet extraction [2–5].
The PLE method uses extraction solvents at high temperatures
(40–200 ◦ C) and pressures (3.4–21 MPa) to increase the extrac-
tion process kinetics. Extraction solvents of PLE are commonly
used at temperatures that are higher than their respective boiling
points and at high pressures to increase the analyte solubilities
in solvents. These conditions allow extraction to take place in a
short period (15 min) with smaller solvent volumes (15 ml) than
those for Soxhlet extraction [6].
Recently, extractions of POPs from solid environmental sam-
ples have been performed using PLE [7–14]. Optimization
studies of PLE conditions, such as solvents and temperatures,
have also been undertaken [7–14] because recoveries of POPs
from environmental samples depend strongly on these condi-
tions. However, studies on simultaneous extractions of both
PCDD/PCDF and PCB from solid environmental samples have
never been performed sufficiently. The main reason is that PLE
conditions for each compound are indicated separately by United
States Environmental Protection Agency Method 3545A (EPA

0021-9673/$ – see front matter © 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.chroma.2006.01.013
 O. Kiguchi et al. / J. Chromatogr. A 1108 (2006) 176–182
Method 3545A) [15]. For that reason, the study of extraction
solvents of PLE for quantitative and simultaneous extractions
of PCDD/PCDF and PCB should facilitate not only extraction
processes, but also the overall analytical systems.
This paper reports the recovery results obtained from using
the PLE that were investigated for quantitative and simultaneous
extractions of PCDD/PCDF and Coplanar PCB (Co-PCB) from
Certified Reference Material: CRM 0422 (Forest soil) [16] using
various solvents. The results were compared to those achieved
using Soxhlet extraction.
2. Experimental
2.1. Sample
A Certified Reference Material (CRM 0422; Japan Soci-
ety for Analytical Chemistry, Tokyo, Japan) was used in this
study [16]. It contains high parts-per-trillion levels (ng/kg) of
PCDD/PCDF and Co-PCB in soil (see Table 1). The respective
water content of CRM soil sample at 110 ± 2 ◦ C for 2 h and igni-
tion loss at 600 ± 25 ◦ C for 2 h were about 13% and 41%. The
certified values were based on Soxhlet extraction using toluene.
2.2. Standards and reagents
The PCDD/PCDF and Co-PCB standards (DF-CVS-B10 and
PCB-CVS-A10) and their 13 C12 -labeled cleanup and injection
internal compounds (DF-LCS-A, PCB-LCS-A, DF-IS-I and
PCB-IS-B) in nonane solution were purchased from Wellington
Laboratories (Ontario, Canada). The n-hexane, toluene, acetone
and dichloromethane used for this study were purchased from
Wako Pure Chemical Industries (dioxin-analysis grade; Osaka,
Japan).
2.3. Extraction
Pressurized liquid extraction was carried out using an accel-
erated solvent extraction system (ASE 200; Dionex, Sunnyvale,
CA, USA) equipped with 33 ml stainless-steel extraction cells.
The CRM sample of 5 g was placed into the stainless-steel
extraction cell prepared for PLE by inserting a cellulose filter and
thimble (Dionex) to prevent clogging of metal frits by extracts at
the cap. The dead volume was filled with pre-cleaned glass beads
(BZ-1, ␾ 0.991–1.397 mm; AsONE, Tokyo, Japan). Then the
cell was sealed with the top cell cap. The PLE conditions used for
this study conformed to standard conditions of PCDD/PCDFs
recommended by the EPA Method 3545A [15]. They are
13.8 MPa system pressure, two static cycles and 10 min static
times per cycle (2 × 10 min static time), with oven temperature
of 150 ◦ C. We selected toluene, n-hexane and acetone as single
solvents, and acetone/n-hexane (1:1, v/v) and acetone/toluene
(1:1, v/v) as mixed solvents. Reasons for the choice of these sol-
vents are as follows: the EPA Method 3545A for PCDD/PCDFs
and PCBs, respectively, recommended single solvents such as
toluene and n-hexane. In addition, this method explained that
the application of water-miscible solvents, such as acetone,
facilitates the extraction of wet solids by allowing the mixed
177
solvent to penetrate the layer of water on the solid particles’
surface. For this reason, toluene, n-hexane and acetone were
selected as single solvents for this study. Furthermore, regarding
mixed solvents as described above, the manuscript also recom-
mended acetone/n-hexane and acetone/dichloromethane (1:1,
v/v) for PCBs. However, dichloromethane has more-toxic prop-
erties than the selected single solvents. Consequently, acetone/n-
hexane and acetone/toluene (1:1, v/v) were selected for this
study instead of acetone/dichloromethane (1:1, v/v). Triplicate
extractions were performed using each of the selected extraction
solvents under identical extraction conditions.
Soxhlet apparatus with a Dean-Stark water separator (Kimura
Rikagaku, Tokyo, Japan) and each of the same extraction sol-
vents as described in PLE were used for Soxhlet extraction of
the CRM sample. The Soxhlet extraction was accomplished in
a pre-baked cellulose thimble containing 5 g sample aliquots.
Quintuplicate extractions were performed using ca. 400 ml of
toluene for >16 h with ca. 30 min of reflux cycle times. Tripli-
cate extractions were performed using other extraction solvents
under the same extraction conditions as those for toluene.
2.4. Cleanup
Cleanup procedures were accomplished according to the
Manual for the Survey and Measurement of Dioxins in Soil
(Japanese Official Test Method) [17]. The outline is as follows:
each diluted 13 C12 -labeled PCDD/PCDF and Co-PCB internal
standard was added to extracts before cleanup. The extracts were
treated with concentrated sulfuric acid and multi-layered silica-
gel column chromatography. Thereafter, the extracts were frac-
tionated using an active carbon-dispersed silica-gel reversible
column (Kanto Kagaku, Tokyo, Japan). The column was first
eluted with n-hexane, n-hexane/dichloromethane (4:1, v/v).
Subsequently, it was eluted using toluene from the opposite
flow direction. The monoortho-PCBs were eluted in the second
fraction. The nonortho-PCBs and PCDD/PCDFs were eluted
in the third fraction. Each macro-concentrated fraction was
concentrated to near-dryness using a nitrogen blowdown appa-
ratus and a gentle stream of purified nitrogen gas. Finally,
the diluted 13 C12 -labeled injection internal standards were
added to each sample to calculate the recoveries of the 13 C12 -
labeled cleanup internal standards that had been added before
cleanup.
2.5. Determination of PCDD/PCDFs and Co-PCBs
Determination procedures were also accomplished according
to the Japanese Official Test Methods described in the previous
section. Analyses using GC/MS were performed with a high-
resolution gas chromatograph (HRGC) and a high-resolution
mass spectrometer (HRMS) (JMS-700D; JEOL, Tokyo, Japan).
We achieved GC separation for TetraCDDs/CDFs to Hex-
aCDDs/CDFs, for HeptaCDDs/CDFs to OctaCDD/CDFs, and
for TetraCBs to HeptaCBs, respectively, using a capillary col-
umn SP2331 (60 m × 0.32 mm i.d.; Supelco, Bellefonte, PA,
USA), DB17 (30 m × 0.25 mm i.d.; J&W Scientific, Folsom,
CA, USA), and HT8 (50 m × 0.22 mm i.d.; SGE, Tokyo, Japan).
178 O. Kiguchi et al. / J. Chromatogr. A 1108 (2006) 176–182

Table 1
Soxhlet extraction of Certified Reference Material: CRM 0422 (Forest soil) using various solvents
Compound Certified valuea Toluene n-Hexane Acetone Acetone/toluene, Acetone/n-
1:1 (v/v) hexane, 1:1
(v/v)
Recoveryb RSDc Recovery RSD Recovery RSD Recovery RSD Recovery RSD
(%) (%) (%) (%) (%) (%) (%) (%) (%) (%)

PCDD/PCDFs isomers
2,3,7,8-TetraCDD 4.51 ± 0.62 96 1.3 27 22.0 80 5.6 85 19.2 66 15.2
1,2,3,7,8-PentaCDD 25.3 ± 3.3 109 5.5 36 0.6 84 2.7 100 8.2 79 10.0
1,2,3,4,7,8-HexaCDD 26.6 ± 2.6 101 4.9 40 5.4 81 2.7 89 13.6 81 7.1
1,2,3,6,7,8-HexaCDD 54.1 ± 4.5 105 5.0 45 4.7 83 4.7 94 11.9 82 3.4
1,2,3,7,8,9-HexaCDD 49.6 ± 4.5 104 3.3 42 2.8 78 4.0 96 7.4 79 2.6
1,2,3,4,6,7,8-HeptaCDD 384 ± 39 124 1.8 60 8.7 100 4.0 109 7.7 100 6.0
OctaCDD 1721 ± 148 92 2.8 74 9.1 105 9.6 99 11.8 101 6.7
2,3,7,8-TetraCDF 35.2 ± 4.0 103 6.3 37 7.7 82 0.0 92 6.4 76 2.2
1,2,3,7,8-PentaCDF 56.5 ± 7.9 124 3.1 52 3.9 106 5.1 119 11.7 102 1.0
2,3,4,7,8-PentaCDF 59.3 ± 6.6 114 6.0 49 2.0 99 4.5 106 10.9 94 6.8
1,2,3,4,7,8-HexaCDF 67.6 ± 9.2 116 3.7 55 1.5 103 0.8 106 14.3 97 3.2
1,2,3,6,7,8-HexaCDF 75.1 ± 7.7 104 4.3 47 0.0 91 3.0 96 10.8 84 1.6
1,2,3,7,8,9-HexaCDF 5.04 ± 0.59 118 15.0 65 6.1 97 8.3 114 8.1 114 2.6
2,3,4,6,7,8-HexaCDF 107 ± 10 107 4.8 50 0.0 93 0.6 98 16.6 93 1.0
1,2,3,4,6,7,8-HeptaCDF 320 ± 32 105 3.4 56 0.0 90 4.0 93 14.0 90 2.0
1,2,3,4,7,8,9-HeptaCDF 38.8 ± 3.7 114 5.2 62 4.2 107 5.0 121 20.5 102 5.2
OctaCDF 221 ± 23 117 5.7 63 0.0 92 7.5 98 22.8 90 5.0
Average 109 4.8 51 4.6 92 4.2 101 12.7 90 4.8
PCDD/PCDF homologue
TetraCDDs 615 ± 94 106 3.6 41 2.3 94 2.6 98 10.0 90 2.7
PentaCDDs 612 ± 86 104 4.6 41 0.9 89 1.2 98 7.0 84 4.6
HexaCDDs 761 ± 72 104 3.5 46 3.6 88 3.3 96 11.2 85 3.0
HeptaCDDs 774 ± 80 135 2.2 69 5.7 113 2.4 118 8.5 112 5.5
OctaCDD 1721 ± 148 92 2.8 74 9.1 105 9.6 99 11.8 101 6.7
TetraCDFs 925 ± 140 113 3.6 44 1.3 99 2.1 107 9.1 94 3.6
PentaCDFs 911 ± 130 112 3.5 49 1.5 102 1.7 110 13.1 98 2.4
HexaCDFs 794 ± 78 104 3.1 50 1.9 95 1.0 100 13.8 91 1.9
HeptaCDFs 517 ± 49 109 3.2 60 1.0 95 3.3 99 15.1 95 1.4
OctaCDF 221 ± 23 117 5.7 63 0.0 92 7.5 98 22.8 90 5.0
PCDDs total 4458 ± 440 106 2.5 59 5.7 100 4.5 102 9.9 97 4.1
PCDFs total 3365 ± 370 110 2.8 51 0.4 98 1.1 104 12.5 94 1.8
PCDD/PCDF total 7823 ± 780 108 2.5 56 3.5 99 2.8 103 10.9 96 3.1
Average 109 3.4 54 2.8 98 3.3 103 12.0 94 3.5
Co-PCBs
3,4,4 ,5-TetraCB (#81)d 31.4 ± 3.1 105 9.5 44 6.2 93 5.6 89 12.7 87 2.8
3,3 ,4,4 -TetraCB (#77) 266 ± 28 110 3.9 44 5.5 92 1.7 98 11.5 89 2.6
2 ,3,4,4 ,5-PentaCB (#123) 45.6 ± 4.7 116 4.1 67 1.9 105 7.8 115 2.5 110 2.3
2,3 ,4,4 ,5-PentaCB (#118) 1249 ± 84 107 5.2 60 4.3 103 2.4 107 0.9 102 2.7
2,3,4,4 ,5-PentaCB (#114) 24.0 ± 3.6 123 10.0 71 1.8 120 11.6 123 2.8 119 5.0
2,3,3 ,4,4 -PentaCB (#105) 520 ± 48 107 5.5 59 6.2 99 1.4 111 4.4 103 1.5
3,3 ,4,4 ,5-PentaCB (#126) 110 ± 11 120 4.8 56 2.0 111 5.2 115 12.5 109 0.6
2,3 ,4,4 ,5,5 -HexaCB (#167) 130.6 ± 8.3 111 4.4 73 2.6 110 2.7 118 3.7 113 1.0
2,3,3 ,4,4 ,5-HexaCB (#156) 261 ± 15 113 2.1 71 4.4 114 4.8 114 3.4 106 4.7
2,3,3 ,4,4 ,5 -HexaCB (#157) 100.4 ± 6.8 109 4.3 75 1.8 115 4.1 118 1.0 111 0.3
3,3 ,4,4 ,5,5 -HexaCB (#169) 34.5 ± 1.5 111 1.4 69 2.8 117 5.5 120 16.5 110 4.9
2,3,3 ,4,4 ,5,5 -HeptaCB (#189) 60.3 ± 5.0 109 4.6 78 3.0 118 1.2 116 3.8 117 6.8
Average 112 5.0 64 3.6 108 4.5 112 6.3 106 2.9

Experiments were made in quintuplicate for toluene and triplicate for the others. CRM 0422 was purchased from Japan Society for Analytical Chemistry.
a Certified values are based on Soxhlet extraction using toluene.
b Recovery is the ratio of each isomer concentration to the certified values.
c RSD = relative standard deviation.
d Numbers in parentheses = International Union of Pure and Applied Chemistry compound numbers.
 O. Kiguchi et al. / J. Chromatogr. A 1108 (2006) 176–182
The HRGC/HRMS analyses were performed according to the
conditions described in a previous report [18].
Recoveries of the 13 C12 -labeled cleanup internal standards
were good; the respective recoveries of those compounds were
60–117% for Soxhlet samples and 61–121% for PLE samples.
3. Results and discussion
Prior to pressurized liquid extraction using toluene, n-hexane,
acetone, acetone/toluene (1:1, v/v), and acetone/n-hexane (1:1,
v/v), the Soxhlet extractions of CRM soil (CRM 0422) were
performed using these solvents. Table 1 summarizes these com-
pounds’ recoveries. The recoveries are ratios of each isomer’s
concentration to the certified values.
For Soxhlet extraction using single solvents, Soxhlet extrac-
tions using toluene and acetone normally yielded good recov-
eries of these compounds between 80 and 120%. Results for
toluene and acetone agreed well with the certified concentra-
tions. Only Soxhlet extraction using n-hexane produced con-
siderably low recoveries of PCDD/PCDF and Co-PCB (<80%).
Regarding Soxhlet extraction using mixed solvents, the results
were comparable to those using toluene and acetone. Par-
ticularly, the results obtained from Soxhlet extraction using
acetone/n-hexane (1:1, v/v) yielded better recovery results
than those using n-hexane. These results indicate that a polar
solvent, such as acetone, achieved better recoveries of both
PCDD/PCDFs and Co-PCBs than did the n-hexane, a non-polar
solvent. However, for acetone/toluene (1:1, v/v), no apprecia-
ble effects were recognized even though acetone was mixed
with toluene. This lack of effect might have resulted from the
difference in the solvent composition in the Soxhlet chamber
during Soxhlet extraction because the difference in the boiling
points of acetone (56 ◦ C) and n-hexane (69 ◦ C) is small: about
13 ◦ C. The difference in boiling points of acetone (56 ◦ C) and
toluene (111 ◦ C) is large: about 55 ◦ C. Accordingly, condensed
acetone during Soxhlet extraction for acetone/toluene might not
come into direct contact with the sample in the Soxhlet chamber,
which thereby, causes ineffectual extraction. Consequently, the
choice of acetone or toluene, or their mixtures, used for Soxh-
let extraction definitely affects quantitative recoveries of both
PCDD/PCDFs and Co-PCBs from contaminated soil.
Examinations by PLE using the same extraction solvents as
those used for Soxhlet extraction were performed. Tables 2 and 3
summarize those results. Recovery results of the “first step”
show recoveries of PCDD/PCDF and Co-PCB obtained using
the selected solvents and the standard conditions described in
Section 2.3 (first step PLE). Recovery results of “first + second
step” show summary recoveries of those compounds obtained
from the first step PLE and further extraction using the same
solvents and conditions once again (second step PLE).
The first step PLE using single solvents produced unsatis-
factory recovery results for both PCDD/PCDFs and Co-PCBs
and for PCDD/PCDFs alone. Contrary to our expectations, the
first step PLE using toluene produced unsatisfactory results even
though the recommended PLE conditions [15] were used for
PCDD/PCDFs. Comparable results were obtained for n-hexane.
Among the single solvents, first step PLE using acetone exhib-
179
ited slightly higher recoveries of PCDD/PCDF and Co-PCB than
those using toluene or n-hexane. However, the recoveries of
some higher chlorinated PCDD/PCDFs (HeptaCDD/CDFs to
OctaCDD/CDFs) were lower than those of lower chlorinated
PCDD/PCDFs (TetraCDD/CDFs to HexaCDD/CDFs). Further-
more, comparison of the recovery results of PCDD/PCDFs for
PLE and Soxhlet were clearly different from the previous report
[3]. The result for first step PLE suggests that poor recoveries
of higher chlorinated PCDD/PCDFs are attributable to strong
hydrophobic interactions between these compounds and the
soil matrix [12] because the quantity of soil organic matter in
the CRM sample was relatively abundant: 41% (see Section
2.1). In addition, PCDD/PCDFs’ high hydrophobic properties
(octanol–water partition coefficients) increase with increasing
chlorine numbers of these compounds [19]. For mixed solvents,
good recovery results were found as a result of the addition of
acetone. Recoveries of PCDD/PCDFs and Co-PCBs were better
than those for single solvents. These results suggest that PLE
using n-hexane or toluene mixed with acetone are effective in
extracting both PCDD/PCDFs and Co-PCBs. Similar results to
those for Soxhlet extraction were also obtained for the first step
PLE. Furthermore, it is noteworthy that poor PLE recoveries
of higher chlorinated PCDD/PCDFs for toluene or n-hexane
were markedly improved by the addition of acetone to their
single solvents. This fact demonstrates that acetone, which is
highly capable of building hydrogen bonding, can disrupt the
strong analyte–matrix interaction resulting from hydrogen bond-
ing of the analyte molecules and active sites on the soil matrix
[2,20]. Consequently, less-polar solvents mixed with acetone
contributed to the release of analytes from highly contaminated
soil. Actually, a non-polar solvent mixed with acetone is most
suitable for disrupting analyte (e.g. poly aromatic hydrocarbons,
PAHs) interaction with the matrix, as others have also found
[20,21].
However, recovery results for PCDD/PCDFs and Co-PCBs
obtained from the first step PLEs demonstrated clearly that
they are incomplete because these recovery results were slightly
lower than those for Soxhlet extraction (Tables 1–3). Fur-
thermore, the relative standard deviations (RSDs) of both
PCDD/PCDFs and Co-PCBs for the first step PLE were
high, generally near 10%, or even higher than 10%. There-
fore, further extraction using identical solvents and conditions
were performed once again to achieve satisfactory extrac-
tions of PCDD/PCDF and Co-PCB (Tables 2 and 3). Aver-
age recoveries of both PCDD/PCDF and Co-PCB obtained
from the first + second step PLE were equal to or greater
than those obtained from Soxhlet extractions. In particular, the
first + second step PLE using mixed solvents were about 1.0–1.7
times those obtained from Soxhlet extractions, except for that
of n-hexane. The RSDs of both PCDD/PCDFs and Co-PCBs
for the first + second step PLE were lower than those of the first
step PLE. The good recovery results of second step PLE might
be attributable to pre-wetting of the sample matrix by the sol-
vents used for the first step PLE because it will facilitate better
contact of the analytes with the solvent and enhance extraction
compared to the first step PLE [2]. Consequently, analytes might
be more quickly dissolved in the solvent.
180 O. Kiguchi et al. / J. Chromatogr. A 1108 (2006) 176–182

Table 2
Pressurized liquid extraction (PLE) of Certified Reference Material: CRM 0422 (Forest soil) using single solvents
Compound Certified valuea Toluene n-Hexane Acetone

First stepb First + second First step First + second First step First + second
stepc step step
Recoveryd RSDe (%) Recovery RSD Recovery RSD Recovery RSD Recovery RSD Recovery RSD
(%) (%) (%) (%) (%) (%) (%) (%) (%) (%)

PCDD/PCDFs isomers
2,3,7,8-TetraCDD 4.51 ± 0.62 84 11.1 113 3.4 64 15.8 87 11.5 87 5.9 119 5.7
1,2,3,7,8-PentaCDD 25.3 ± 3.3 70 11.8 107 9.0 70 6.5 92 6.4 79 8.7 112 13.6
1,2,3,4,7,8-HexaCDD 26.6 ± 2.6 78 12.2 125 4.6 74 5.9 102 5.5 90 15.0 129 3.4
1,2,3,6,7,8-HexaCDD 54.1 ± 4.5 62 13.5 95 12.5 60 12.9 80 4.3 76 11.2 102 1.8
1,2,3,7,8,9-HexaCDD 49.6 ± 4.5 66 16.9 105 13.5 60 5.1 83 10.1 77 8.0 108 1.1
1,2,3,4,6,7,8-HeptaCDD 384 ± 39 71 14.8 113 9.3 70 9.8 95 3.7 80 12.3 115 8.2
OctaCDD 1721 ± 148 59 6.7 95 2.8 74 12.1 96 10.3 74 4.6 105 5.6
2,3,7,8-TetraCDF 35.2 ± 4.0 61 2.7 93 7.1 63 13.6 84 8.4 70 10.2 100 5.3
1,2,3,7,8-PentaCDF 56.5 ± 7.9 88 4.6 124 5.7 87 13.5 113 4.5 97 5.9 133 5.4
2,3,4,7,8-PentaCDF 59.3 ± 6.6 83 9.4 108 16.7 79 3.7 103 5.0 93 10.3 129 2.0
1,2,3,4,7,8-HexaCDF 67.6 ± 9.2 75 8.5 107 21.6 69 11.8 94 1.8 88 11.3 123 1.4
1,2,3,6,7,8-HexaCDF 75.1 ± 7.7 70 7.2 101 21.4 72 13.4 94 4.3 79 10.3 112 8.2
1,2,3,7,8,9-HexaCDF 5.04 ± 0.59 71 13.8 101 35.5 73 11.0 106 9.4 73 12.4 122 10.8
2,3,4,6,7,8-HexaCDF 107 ± 10 65 11.7 105 5.8 71 15.3 93 5.1 74 11.0 110 1.0
1,2,3,4,6,7,8-HeptaCDF 320 ± 32 66 8.2 102 1.8 67 9.8 88 4.7 73 6.5 104 5.3
1,2,3,4,7,8,9-HeptaCDF 38.8 ± 3.7 72 16.4 116 5.9 72 9.4 100 1.5 76 3.9 113 7.0
OctaCDF 221 ± 23 65 4.0 104 6.0 75 12.5 .01 2.7 77 11.8 111 4.5
Average 71 10.2 107 10.7 71 10.7 95 5.8 80 9.4 114 5.3
PCDD/PCDF homologue
TetraCDDs 615 ± 94 80 7.6 118 3.3 81 10.1 104 1.7 97 8.1 129 4.2
PentaCDDs 612 ± 86 70 7.7 108 6.5 74 11.7 98 4.6 83 12.2 117 4.2
HexaCDDs 761 ± 72 68 9.4 105 6.2 71 11.1 93 1.4 83 9.2 115 1.8
HeptaCDDs 774 ± 80 73 11.2 115 7.5 76 10.3 102 4.6 84 11.9 120 8.4
OctaCDD 1721 ± 148 59 6.7 95 2.8 74 12.1 96 10.3 74 4.6 105 5.6
TetraCDFs 925 ± 140 80 5.4 118 2.5 81 9.1 108 2.7 89 11.0 123 3.9
PentaCDFs 911 ± 130 84 3.0 125 4.6 88 9.9 114 0.6 97 11.1 133 0.4
HexaCDFs 794 ± 78 74 7.7 112 6.6 78 12.9 102 3.4 84 10.2 120 0.8
HeptaCDFs 517 ± 49 67 7.8 105 1.4 69 11.2 91 4.0 74 6.9 106 5.1
OctaCDF 221 ± 23 65 4.0 104 6.0 75 12.5 101 2.7 77 11.8 111 4.5
PCDDs total 4458 ± 440 68 7.9 106 4.1 75 9.1 99 3.7 82 10.1 115 3.9
PCDFs total 3365 ± 370 77 4.6 115 1.4 80 10.2 105 0.6 87 0.6 122 1.9
PCDD/PCDF total 7823 ± 780 72 6.4 110 2.9 77 9.4 101 2.1 84 8.5 118 2.9
Average 72 6.9 110 4.3 77 10.7 101 3.3 84 9.5 118 3.7
Co-PCBs
3,4,4 ,5-TetraCB (#81)f 31.4 ± 3.1 75 6.3 111 3.8 82 16.1 104 6.0 96 6.1 125 7.4
3,3 ,4,4 -TetraCB (#77) 266 ± 28 76 2.7 108 5.3 78 8.7 101 6.5 88 6.5 113 3.3
2 ,3,4,4 ,5-PentaCB (#123) 45.6 ± 4.7 77 1.9 109 2.4 94 6.0 115 3.7 88 5.8 115 4.0
2,3 ,4,4 ,5-PentaCB (#118) 1249 ± 84 74 4.6 105 5.3 86 13.6 107 4.7 89 11.4 114 2.2
2,3,4,4 ,5-PentaCB (#114) 24.0 ± 3.6 82 5.8 123 8.0 98 10.4 120 1.0 92 7.6 120 10.0
2,3,3 ,4,4 -PentaCB (#105) 520 ± 48 76 2.0 107 2.8 90 9.7 111 2.0 88 8.4 113 3.5
3,3 ,4,4 ,5-PentaCB (#126) 110 ± 11 87 4.6 128 6.7 94 10.4 120 2.1 99 4.6 130 4.9
2,3 ,4,4 ,5,5 -HexaCB (#167) 130.6 ± 8.3 78 4.4 112 3.7 95 10.8 116 2.6 100 9.2 128 2.0
2,3,3 ,4,4 ,5-HexaCB (#156) 261 ± 15 80 3.4 115 6.0 93 10.9 116 2.7 99 4.4 128 3.8
2,3,3 ,4,4 ,5 -HexaCB (#157) 100.4 ± 6.8 73 1.6 109 7.3 92 4.5 114 3.6 92 9.3 121 4.9
3,3 ,4,4 ,5,5 -HexaCB (#169) 34.5 ± 1.5 87 9.1 130 2.3 96 7.3 123 3.3 101 11.5 135 8.3
2,3,3 ,4,4 ,5,5 -HeptaCB (#189) 60.3 ± 5.0 82 10.6 117 8.6 95 10.5 119 2.8 92 8.4 124 11.0
Average 79 4.7 114 5.2 91 9.9 114 2.9 94 7.8 122 5.4

Experiments were made in triplicate. CRM 0422 was purchased from Japan Society for Analytical Chemistry.
a Certified values were based on Soxhlet extraction using toluene.
b First step = recoveries for first step PLE under standard conditions (system pressure of 13.8 MPa, extraction temperature of 150 ◦ C, two static PLE cycles and

10 min static PLE time/cycle).

c First + second step = sum of recoveries for first and second step PLE under standard conditions.
d Recovery = ratio of each isomer concentration to the certified values.
e RSD = relative standard deviation.
f Number in parentheses = International Union of Pure and Applied Chemistry compound numbers.
 O. Kiguchi et al. / J. Chromatogr. A 1108 (2006) 176–182 181

Table 3
Pressurized liquid extraction (PLE) of Certified Reference Material: CRM 0422 (Forest soil) using mixed solvents
Compound Certified valuea Acetone/toluene, 1:1 (v/v) Acetone/n-hexane, 1:1 (v/v)

First stepb First + second stepc First step First + second step

Recoveryd RSDe Recovery RSD Recovery RSD Recovery RSD

(%) (%) (%) (%) (%) (%) (%) (%)

PCDD/PCDFs isomers
2,3,7,8-TetraCDD 4.51 ± 0.62 81 12.3 112 12.1 83 5.6 113 3.4
1,2,3,7,8-PentaCDD 25.3 ± 3.3 80 17.3 117 17.4 86 18.7 119 11.3
1,2,3,4,7,8-HexaCDD 26.6 ± 2.6 81 7.1 127 6.2 91 13.2 134 8.1
1,2,3,6,7,8-HexaCDD 54.1 ± 4.5 74 13.2 113 13.0 81 6.8 112 5.8
1,2,3,7,8,9-HexaCDD 49.6 ± 4.5 80 17.7 118 12.9 85 9.5 126 7.9
1,2,3,4,6,7,8-HeptaCDD 384 ± 39 93 11.3 135 6.9 96 5.4 136 4.8
OctaCDD 1721 ± 148 76 7.7 114 2.6 83 10.7 114 7.7
2,3,7,8-TetraCDF 35.2 ± 4.0 76 4.3 106 7.9 77 3.7 108 0.0
1,2,3,7,8-PentaCDF 56.5 ± 7.9 98 12.8 135 9.5 108 4.3 145 4.6
2,3,4,7,8-PentaCDF 59.3 ± 6.6 96 10.0 133 6.7 83 6.1 116 3.8
1,2,3,4,7,8-HexaCDF 67.6 ± 9.2 86 8.8 127 2.0 99 4.5 136 2.2
1,2,3,6,7,8-HexaCDF 75.1 ± 7.7 82 6.2 119 5.0 89 6.0 123 6.1
1,2,3,7,8,9-HexaCDF 5.04 ± 0.59 99 13.9 145 8.6 95 1.2 130 2.3
2,3,4,6,7,8-HexaCDF 107 ± 10 77 15.1 111 9.7 87 6.5 118 1.6
1,2,3,4,6,7,8-HeptaCDF 320 ± 32 74 10.6 108 10.0 72 4.3 102 3.1
1,2,3,4,7,8,9-HeptaCDF 38.8 ± 3.7 95 18.2 135 8.8 107 13.6 145 10.7
OctaCDF 221 ± 23 81 14.7 123 6.7 81 5.6 119 8.2
Average 84 11.8 122 8.6 88 7.4 123 5.4
PCDD/PCDF homologue
TetraCDDs 615 ± 94 92 7.8 126 4.9 92 5.8 124 2.9
PentaCDDs 612 ± 86 87 13.8 123 8.8 101 11.6 136 6.8
HexaCDDs 761 ± 72 81 14.6 120 10.5 94 1.8 132 1.5
HeptaCDDs 774 ± 80 93 9.1 136 6.0 107 6.4 150 5.2
OctaCDD 1721 ± 148 76 7.7 114 2.6 83 10.7 114 7.7
TetraCDFs 925 ± 140 98 8.6 134 5.1 103 0.9 139 0.6
PentaCDFs 911 ± 130 97 12.2 135 7.3 104 0.9 141 0.9
HexaCDFs 794 ± 78 88 9.7 126 4.7 104 4.1 141 3.2
HeptaCDFs 517 ± 49 78 10.7 114 8.7 79 3.6 111 2.3
OctaCDF 221 ± 23 81 14.7 123 6.7 81 5.6 119 8.2
PCDDs total 4458 ± 440 84 9.9 122 5.5 93 5.1 128 3.2
PCDFs total 3365 ± 370 91 10.3 129 6.1 98 2.1 134 1.6
PCDD/PCDF total 7823 ± 780 87 10.1 125 5.8 96 3.6 131 2.1
Average 87 10.7 125 6.4 95 4.8 131 3.6
Co-PCBs
3,4,4 ,5-TetraCB (#81)f 31.4 ± 3.1 87 7.5 118 3.4 92 3.5 120 2.2
3,3 ,4,4 -TetraCB (#77) 266 ± 28 82 9.8 112 9.0 92 6.3 120 5.5
2 ,3,4,4 ,5-PentaCB (#123) 45.6 ± 4.7 86 8.5 121 6.4 107 8.1 135 6.7
2,3 ,4,4 ,5-PentaCB (#118) 1249 ± 84 80 5.9 110 4.0 86 7.3 113 6.4
2,3,4,4 ,5-PentaCB (#114) 24.0 ± 3.6 96 7.0 131 6.1 110 2.9 137 1.7
2,3,3 ,4,4 -PentaCB (#105) 520 ± 48 84 9.0 114 7.3 88 1.2 116 0.7
3,3 ,4,4 ,5-PentaCB (#126) 110 ± 11 97 12.2 133 6.3 97 1.0 135 2.8
2,3 ,4,4 ,5,5 -HexaCB (#167) 130.6 ± 8.3 83 10.8 118 7.3 94 8.8 128 7.9
2,3,3 ,4,4 ,5-HexaCB (#156) 261 ± 15 84 13.4 121 8.9 95 8.0 124 4.8
2,3,3 ,4,4 ,5 -HexaCB (#157) 100.4 ± 6.8 78 6.7 112 10.2 95 2.5 125 4.2
3,3 ,4,4 ,5,5 -HexaCB (#169) 34.5 ± 1.5 98 16.1 141 12.5 110 9.4 152 6.5
2,3,3 ,4,4 ,5,5 -HeptaCB (#189) 60.3 ± 5.0 83 7.6 117 5.6 107 4.8 136 7.6
Average 87 9.6 121 7.2 98 5.3 128 4.8

Experiments were made in triplicate. CRM 0422 was purchased from Japan Society for Analytical Chemistry.
a Certified values are based on Soxhlet extraction using toluene.
b First step = first step PLE under standard conditions (system pressure of 13.8 MPa, extraction temperature of 150 ◦ C, two static PLE cycles and 10 min static PLE

time/cycle).
c First + second step = sum of recoveries for first and second step PLE under standard conditions.
d Recovery = ratio of each isomer concentration to certified values.
e RSD = relative standard deviation.
f Numbers in parentheses = International Union of Pure and Applied Chemistry compound numbers.
182 O. Kiguchi et al. / J. Chromatogr. A 1108 (2006) 176–182
The low recoveries and the larger RSDs of these compounds
for the first step PLE might have resulted from the dependence
on the sample’s particle size distribution. It has been reported
that the PLE of PCBs from solid samples with a larger par-
ticle size distribution (>15 ␮m) engendered lower recoveries
and larger RSDs [9]. Regarding the soil sample (CRM 0422)
used for this study, the particle size distribution was <16 ␮m for
about 60%, but was >16 ␮m (16–96 ␮m) for about 40% [16].
Therefore, the obtained results used for this study might result
from dependence on the sample’s particle size distribution. How-
ever, the recovery results of PCDD/PCDFs and Co-PCBs for the
second step PLE seem to show that consideration of only the
dependence on the sample’s particle size distribution is inad-
equate to explain the low recoveries of those compounds for
the first step PLE because the difference of recoveries between
the first and second step PLE for dependence on the sample’s
particle size distribution were less than 10% [9], but those for
this study were greater than 20% (Tables 2 and 3). Therefore,
the difference might occur because most of PCDD/PCDFs and
Co-PCBs obtained using the second step PLE exist in a strong
analyte–matrix interaction; they exist as more tightly bound in
the soil matrix. Hawthorne et al. [14] reported that many pol-
lutant molecules become “sequestered” as they age for decades
in the environment. Therefore, it is likely that lower recoveries
and the larger RSDs for the first step PLE are also attributable to
existence of PCDD/PCDFs and Co-PCBs that are more tightly
bound in the soil matrix.
Accordingly, the first step PLE might not always be sufficient
to assure completely exhaustive and simultaneous extractions
of PCDD/PCDFs and Co-PCBs that were more tightly bound
to the soil matrix. We suggest that the choice of mixed sol-
vents such as n-hexane and toluene with acetone is important,
along with several extraction cycles, is required for quantitative
and simultaneous extraction of PCDD/PCDF and Co-PCB from
soil. Numerous researchers have reported that several extrac-
tion cycles are necessary to achieve satisfactory extraction of
PCDD/PCDFs [7,8] or some PCBs [9,13]. Our results, which
support those findings, were obtained both simultaneously and
quantitatively.
4. Conclusion
The use of mixed solvents, such as n-hexane and toluene
with acetone for PLE, produced good recoveries of both
PCDD/PCDFs and Co-PCBs that were tightly bound to the soil
matrix. However, the mere use of mixed solvents was insufficient
to obtain satisfactory results. Additional extraction cycles using
the mixed solvents were required to achieve equivalent or better
recovery results than those obtained through Soxhlet extraction.
Consequently, we conclude that the use of mixed solvents such
as n-hexane, toluene with acetone and several extraction cycles
are essential factors of PLE for quantitative and simultaneous
extraction of PCDD/PCDFs and Co-PCBs that were more tightly
bound to the soil matrix.
References
[1] US Environmental Protection Agency, Persistent Organic Pollutants: A
Global Issue, A Global Response, EPA 160-F-02-001, US Environmental
Protection Agency, Washington, DC, 2002.
[2] B.E. Richter, B.A. Jones, J.L. Ezzell, N.L. Porter, N. Avdalovic, C. Pohl,
Anal. Chem. 68 (1996) 1033.
[3] B.E. Richter, J.L. Ezzell, D.E. Knowles, F. Hoefler, A.K.R. Mattulat, M.
Scheutwinkel, D.S. Waddell, T. Gobran, V. Khurana, Chemosphere 34
(1997) 975.
[4] X. Lou, H.G. Janssen, C.A. Cramers, Anal. Chem. 69 (1997) 1598.
[5] H.J. Vandenburg, A.A. Clifford, K.D. Bartle, S.A. Zhu, J. Carroll, I.D.
Newton, L.M. Garden, Anal. Chem. 70 (1998) 1943.
[6] Extraction of polychlorinated dibenzo-p-dioxins and polychlorinated
dibenzofurans from environmental samples using accelerated solvent
extraction (ASE® ), Application note 323, Dionex, Sunnyvale, CA,
1999.
[7] P. Popp, P. Keil, M. Möder, A. Paschke, U. Thuss, J. Chromatogr. A
774 (1997) 203.
[8] H. Bautz, J. Polzer, L. Stieglitz, J. Chromatogr. A 815 (1998) 231.
[9] E. Björklund, S. Bøwadt, T. Nilsson, L. Mathiasson, J. Chromatogr. A
836 (1999) 285.
[10] E. Björklund, T. Nilsson, S. Bøwadt, Trends Anal. Chem. 19 (2000)
434.
[11] I. Windal, D.J. Miller, E. De Pauw, S.B. Hawthorne, Anal. Chem. 72
(2000) 3916.
[12] A. Hubert, K.D. Wenzel, M. Manz, L. Weissflog, W. Engewald, G.
Schüürmann, Anal. Chem. 72 (2000) 1294.
[13] C. Bandh, E. Björklund, L. Mathiasson, C. Näf, Y. Zebühr, Environ.
Sci. Technol. 34 (2000) 4995.
[14] S.B. Hawthorne, C.B. Grabanski, E. Martin, D.J. Miller, J. Chromatogr.
A 892 (2000) 421.
[15] US Environmental Protection Agency Method 3545A, Pressurized fluid
extraction (PFE), EPA SW-846, US Government Printing Office, Wash-
ington, DC, November 2000.
[16] K. Okamoto, T. Imagawa, H. Ito, M. Takeuchi, S. Yamasaki, A. Ochi,
N. Ito, T. Kida, S. Tsuruta, K. Kakita, A. Ono, M. Sakata, Bunseki
Kagaku 53 (2004) 1335.
[17] Environment Agency, Government of Japan, Manual for the survey and
measurement of dioxins in soil, Soil and Agricultural Chemicals Divi-
sion, Water Quality Bureau, Tokyo, 2000.
[18] O. Kiguchi, T. Kobayashi, K. Saitoh, N. Ogawa, Anal. Chim. Acta 546
(2005) 102.
[19] D. Mackay, W.Y. Shiu, K.C. Ma, Illustrated Handbook of Physical–
Chemical Properties and Environmental Fate for Organic Chemicals,
vol. II, CRC Press, Boca Raton, FL, 1991, pp. 368–566.
[20] J. Hollender, B. Koch, C. Lutermann, W. Dott, Int. J. Environ. Anal.
Chem. 83 (2002) 21.
[21] J. Pörschmann, J. Plugge, Fresenius J. Anal. Chem. 364 (1999) 643.