Page 10 - Read Online

Page 4 of 17 Hammel et al. J Environ Expo Assess 2024;3:8 https://dx.doi.org/10.20517/jeea.2023.51

13
BB-153 in this analysis. Octa to deca PBDEs were quantified using C-BDE-209, whereas PCB-198 was used
for quantification of tri-hepta PBDEs. The dust samples (0.5 g) were extracted as the milk samples and
cleaned up on a column consisting of 2 g activated aluminum oxide, 2 g sulfuric acid impregnated silica,
and some Na SO , eluted with 60 mL hexane and analyzed by GC-ECNI-MS for the PBDEs given above.
2
4
The chemical analyses of HBCDDs and NBFRs included α-, β-, and γ-HBCDD, EH-TBB, 2,3-
dibromopropyl 2,4,6-tribromophenyl ether (DPTE, also known as TBP-DBPE), bis(2-ethylhexyl)
tetrabromophthalate (BEH-TEBP), decabromodiphenyl ethane (DBDPE), as well as syn- and anti-DP. Their
analysis in breast milk and dust followed the method described in Vorkamp et al., with some
modifications . In brief, sub-samples of approximately 0.5 g of house dust or 20 mL breast milk (after
[45]
13
heating to 37 °C) were dried with diatomaceous earth, spiked with isomer-specific internal standards of C-
13
HBCDD as well as C-BEH-TEBP (where relevant) and extracted by Pressurized Liquid Extraction (dust)
or Soxhlet (milk) using hexane:dichloromethane (1:1). Dust samples were cleaned up on a column
consisting of 2 g activated aluminum oxide, 2 g deactivated silica and some Na SO and eluted with 60 mL
4
2
hexane:dichloromethane (1:1). For the milk extracts, the column clean-up was the same described in
Vorkamp et al. [44,45] , but applied twice and eluted with 250 mL hexane:dichloromethane (1:1). For both dust
and milk extracts, the eluates were split equally in two for analysis of NBFRs and HBCDDs, respectively.
[45]
Each of the milk extracts was evaporated to dryness, using silicone-coated vials for NBFRs and
reconstituted in 200 µL isooctane or methanol for analysis of NBFR and HBCDD, respectively. For NBFRs
in dust, the extracts were reduced in volume by rotary evaporation and under nitrogen, using 300 µL
isooctane as a keeper, and adjusted to a precise volume of 1 mL in isooctane. The second dust extract was
evaporated to dryness and reconstituted in 500 µL methanol for HBCDD analysis. NBFRs were analyzed by
GC-ECNI-MS using the parameters described in Vorkamp et al., but combining all NBFRs in one analytical
run on a 15 m DB-1 GC column (0.25 mm i.d., 0,25 µm film thickness, J&W Scientific, Folsom CA,
USA) . HBCDD-isomers were analyzed by LC-MS-MS as previously described . The NBFR BEH-TEBP
[45]
[46]
was analyzed separately in a subset of milk samples (n = 5), using Gel Permeation Chromatography for
clean-up as described in Vorkamp et al. because BEH-TEBP is not stable to the acid treatment used for
clean-up of the milk samples . BEH-TEBP was included in all dust analyses.
[45]
Each analytical batch included QA/QC measures such as procedural blanks, internal or standard reference
material (NIST SRM 2585), as described elsewhere [15,41] . For PBDEs in breast milk, a sample of human breast
milk from the ring test “Dioxins in Food, 2006” was used (with consensus values for BDE-28, 47, 99, 100,
[47]
153, 154, 209) . The obtained results were within ±20% of the consensus values for all congeners except for
BDE-209 in one batch, which was 37% above the consensus value, and BDE-154, which was generally
overestimated due to the co-elution with BB-153 described above. NIST SRM 2582 was also used for quality
control of HBCCDs and NBFRs, based on indicative values published in the literature [22,48] . For the NBFR
13
analyses, PCB-198 [and, in some batches, C-polychlorinated naphthalene (PCN)-27] were added for
recovery determination prior to extraction. The recovery rate for dust was 91.5% on average, with a range of
68%-116%. The recovery standards were difficult to integrate in the milk samples, leading to an average of
85% (range 52%-120%). However, the low numbers are likely caused by chromatographic challenges due to
matrix effects, rather than indicating losses in the process.

Statistical analyses
Analyses were conducted using SAS statistical software (version 9.4; SAS Institute Inc., Cary, NC, USA).
Based on examination of distributions of all PBDEs and NBFRs in breast milk and the biological matrices
previously analyzed (placenta, fetal and maternal blood), the majority of which were right-skewed, non-
parametric statistics or log -transformed concentrations were used for all analyses. Statistical analyses were
10
conducted only for chemicals with detection frequencies > 65% overall. For any concentrations below the