Materiales de referencia y comparaciones interlaboratorios
95
Sludge #1
Sludge #1
Standard solution
Congener
Median
value
(n=)
Rel. std.
dev. (%)
Median
value
(n=)
Rel. std.
dev. (%)
Median
value
(n=)
Rel. std.
dev. (%)
Prepared
concentration
(ng/ml)
PCB 81
26.0 (42)
17
11.6 (47)
35
41.0 (30)
10
40
PCB 77
658 (60)
15
290 (62)
24
41.0 (30)
10
40
PCB 123
282 (48)
36
220 (48)
44
41.3 (32)
12
40
PCB 118 18500 (66)
15
12700 (61)
14
42.5 (35)
12
40
PCB 105 7310 (65)
14
5300 (65)
16
40.9 (34)
12
40
PCB 126
55.2 (60)
20
30.6 (56)
28
41.8 (33)
13
40
PCB 167
938 (61)
16
531 (61)
19
41.0 (32)
11
40
PCB 156 2600 (60)
15
1500 (55)
13
40.3 (31)
12
40
PCB 157
534 (60)
16
331 (58)
18
39.2 (30)
11
40
PCB 169
3.14 (37)
25
2.10 (36)
39
40.4 (34)
11
40
PCB 189
156 (49)
15
63.8 (50)
15
38.1 (30)
10
40
PCB 31
7000 (42)
48
2030 (36)
39
24.6 (6)
85
PCB 28
8400 (51)
34
2370 (53)
48
34.0 (25)
20
40
PCB 52
13300 (56)
19
10000 (57)
18
38.7 (25)
8
40
PCB 101 21200 (56)
16
15000 (50)
13
38.0 (25)
7
40
PCB 149 14000 (33)
23
8600 (31)
17
0.13 (3)
6
PCB 153 17700 (52)
23
9580 (51)
19
38.2 (27)
14
40
PCB 138 19700 (58)
20
11700 (57)
15
37.7 (24)
8
40
PCB 128 4400 (36)
24
2400 (41)
36
37.0 (17)
30
40
PCB 180 8220 (57)
28
3170 (55)
26
38.5 (25)
12
40
PCB 209
226 (34)
18
210 (32)
19
38.0 (17)
19
40
Table 4: Results for the PCBs (dioxin-like and marker PCBs) (results as pg/g, dry weight
basis)
Interestingly, four laboratories reported a concentration
for PCB 31 in the standard solution while this compound
had not been added.Two of them even used a confirma-
tion column for assessing their result. The cause of this
problem cannot be deducted from the gas chromato-
graphic column used (VF-5ms, DB-XLB and DB-5ms). It
may thus be caused by a wrong peak assignation in their
own method.
Four congeners have a relative standard deviation high-
er than 50% and thus seem to be difficult to determine:
2,3,7,8-TCDD, 1,2,3,7,8,9-HxCDF, BDE 71 and BDE 206.For
2,3,7,8-TCDD and 1,2,3,7,8,9-HxCDF, the concentrations
were very close to the detection limit reported by most
of the laboratories and that may explain why the rela-
tive standard deviation is so high. But for the BDE, the
concentration in the samples were much higher than
the reported detection limits of most of the laboratories.
Co-elution of BDE 49 with BDE 71 has been reported by
a few laboratories.Many laboratories chose not to report
their data for BDE 206.
A brief summary of the methods used by the participat-
ing laboratories is shown in table 5. Soxhlet remain the
extraction technique of choice for more than 60% of
the laboratories; Accelerated Solvent Extraction (ASE) is
second with 14-32% of the laboratories using it. As ex-
pected, BDE analysis is preferably performed on short
columns (30 meters or less), while 60 meter columns are
used for PCDD/DFs, dioxin-like PCBs and marker-PCBs.
Very few laboratories used on-column or PTV injectors
(between 9 and 14%), even for the very demanding and
thermolabile PBDEs. High resolution mass spectrometry
has been used by the majority of the laboratories; a few
of them used higher resolution (usually 12 000 for 3 to
11% of the participants) while some used low resolution
or MS/MS (17-45% of the laboratories).
Interestingly, laboratories using a chromatographic col-
umn of 30 m or more for PBDE analysis do not seem to
underestimate BDE 209, neither do they overestimate
BDE 206, that would be the result of thermal degrada-
tion of BDE 209.
IV. Conclusion
Two municipal sewage sludge samples have been certi-
fied through an international interlaboratory study for
PCDD/DFs, PBDEs, dioxin-like PCBs and marker PCBs.
Relative standard deviations indicate good agreement
between laboratories, except for four congeners (two
PCDD/DF congeners and two PBDE congeners). Tar-
getted values have thus been assigned to most of the
studied congeners. Standard solutions, also analyzed in
this study, showed that most of the participating labora-
tories are able to report good results for almost all tar-
getted congeners (except for a few PCB congeners and
PBDE 206).
V. Acknowledgements
The authors would like to thank Eric Reiner,from the On-
tario Ministry of the Environment,for helpful discussions
and Brock Chittim and Colleen Tashiro from Wellington
laboratories, Ontario, for graciously providing analytical
standards to every laboratory registered in this study.
VI. References
1. Wageningen Evaluation Programmes for Analytical
Laboratories, International Sediment Exchange for Tests
on Organic Contaminants, Wageningen University, De-
partment of Environmental Sciences,The Netherlands.
PCDD/DFs
Extraction
Column type
Column length Injector type
MS resolution
Soxhlet: 67
ASE: 28
Other: 4
DB5-ms: 48
DB5: 24
Other: 30
60 m: 85
< 60 m: 15
Split/splitless: 91
Other: 9
10 000: 72
> 10 000: 11
< 10 000: 17
PBDEs
Extraction
Column type
Column length Injector type
MS resolution
Soxhlet: 77
ASE: 14
Other: 9
DB5-ms: 36
DB5: 23
Other: 41
< 30 m: 50
30 m: 36
> 30 m: 14
Split/splitless: 86
Other: 14
10 000: 50
> 10 000: 5
< 10 000: 45
Dioxin-like PCBs
Extraction
Column type
Column length Injector type
MS resolution
Soxhlet: 61
ASE: 32
Other: 7
DB5-ms: 41
DB5: 22
Other: 37
60 m: 71
Split/splitless: 88
Other: 12
10 000: 66
> 10 000: 7
< 10 000: 27
Marker PCBs
Extraction
Column type
Column length Injector type
MS resolution
Soxhlet: 59
ASE: 35
Other: 6
DB5-ms: 32
DB5: 21
Other: 47
60 m: 68
Split/splitless:91
Other: 9
10 000: 62
> 10 000: 3
< 10 000: 32
Table 5: Summary of the methods used (%)
