19890

Environ  Sci  Pollui  Res  (2015)  22:19887-19895

Ö  Springer

solvent  and  weighted.  Gravimetric  results  were  used  to  quantify ­
  swelling  as  the  volume  increases  after  extraction  relatively
to  the  initial  polymer  volume,  taking  the  loss  of  oligomers  into
account.  Finally,  strips  were  air  dried  in  a  fume  hood  to  evaporate ­
  absorbed  solvent  and  re-weighted.  The  mass  of  released
oligomers  was  calculated  from  the  initial  mass  of  the  strip  and
the  mass  after  extraction  and  drying.
In  addition,  time  and  temperature  of  extraction  were  optimized ­
  to  achieve  the  highest  release  rates  of  oligomers.  Therefore, ­
  (i)  time  series  extraction  experiments  («=3)  where  the
static  extraction  step  increased  in  10  min  steps  from  10  up  to
90  min,  at  constant  temperature  (100  °C)  and  (ii)  extraction
experiments  at  different  temperatures  (75,  100  and  125  °C)
but  constant  extraction  time  (50  min)  were  performed  with  nhexane/acetone
  (l:lv/v).
For  a  reliable  evaluation  of  the  clean-up  efficiency  with
ASE,  a  comparison  was  made  with  classical  clean-up  procedures, ­
  e.g.  Soxhlet  extraction  or  extraction  by  shaking
(Table  3).  For  comparable  results,  all  clean-up  methods  were
performed  two  times  with  three  strips  each  cut  from  the  same
PDMS  sheet.
Sample  extraction
The  main  aim  during  extraction  and  purification  of  PDMS
samplers  is  to  achieve  good  recovery  rates  of  analytes,  as  well
as  to  completely  remove  non-crosslinked  silicone  oligomers.
In  order  to  gain  optimized  recovery  rates  of  CHCs  and  PAFls,
the  same  organic  solvent  should  be  used  for  the  entire  extraction ­
  and  clean-up  process  (Fig.  1)  to  avoid  target  compound
losses  due  to  solvent  exchange  during  the  extraction  process.
In  contrast  to  the  pre-cleaning  step  (s.a.),  where  the  solvent
should  have  a  high  oligomer  release  capacity,  organic  solvent
for  sampler  extraction  should  yield  a  minimum  oligomer
release.
Therefore,  in  a  first  step,  different  solvents  («-hexane/acetone
  (l:lv/v),  dichloromethane,  acetonitrile  and  acetonitrile/
methanol  (2:1  v/v))  typically  used  for  non-polar  contaminants
extraction  from  passive  sampler  (Rusina  2009;  Schafer  et  al.
2010;  Smedes  and  Booij  2012;  Shahpoury  and  Flageman
2013)  were  tested  twice  for  their  extraction  efficiency  for  analytical ­
  target  compounds  by  using  ASE.  Briefly,  ASE  cells
were  filled  with  sea  sand  as  filling  matrix  and  spiked  with  IS.

Each  sample  was  extracted  in  3  cycles  of  5-min  static  time
(100  °C)  to  find  the  optimal  extraction  time  for  a  complete
extraction  of  analytes.  An  azeotropic  solvent  exchange  from
the  more  polar  solvents  acetonitrile  and  methanol  to  hexane
was  performed  with  an  excess  of  hexane  according  to  Smedes
and  Booij  (2012).
In  a  second  step,  extraction  tests  were  performed  on  precleaned ­
  PDMS  sheets  spiked  with  PRCs  (Table  1)  in  a
methanol/water  mixture  (90:10v/v)  according  to  Rusina
(2009)  and  shaken  for  14  days,  while  the  water  content  was
increased  to  methanol/water  (1:1  v/v)  after  1  week.  Spiked
PDMS  sheets  were  ASE  extracted,  whereas  each  ASE  cell
(100  mL)  was  filled  with  six  spiked  PDMS  sheets  (1  sample),
filled  up  with  pre-combusted  sea  sand  (Merck,  Darmstadt,
Germany)  and  IS.  Extraction  was  performed  with  optimized
solvent,  temperature  and  time  (1x10  min,  100  °C,  hexane/
dichloromethane  (1:1  v/v)).  These  PDMS  sheets  represent  the
fabrication  blank  with  no  further  transportation  or  deployment
in  water.
In  a  third  step,  the  newly  developed  method  (Fig.  1)  was
applied  on  field  samples  which  have  been  deployed  in  marine
waters  of  the  German  Bight  (Fleligoland  waters)  for  43  days
and  in  brackish  waters  of  the  Baltic  Sea  (Fehmam  waters)  for
63  days.  Field  samples  comprised  each  of  a  set  of  two  deployed ­
  sampler  and  a  transport  blank  and  enabled  the  direct
comparison  of  real  sampler  matrix  with  laboratory  sampler
blanks.
All  extracts  were  evaporated  to  1  ml  by  parallel  solvent
reduction  and  further  purified  (“Extract  purification  and  analysis”) ­
  prior  to  GC-MS  analysis.
Extract  purification  and  analysis
The  ASE  sample  extracts  need  additional  purification  steps  to
remove  last  traces  of  silicone  oligomers  and  co-extracted  material ­
  (e.g.  organic  matter)  from  field  samples.  Purification
from  co-extracted  material  was  performed  by  SPE  using
500-mg  silica  gel.  Target  compounds  were  eluted  with
hexane/dichloromethane  (70:30,  5  mL)  and  evaporated  to
1  ml  by  a  gentle  stream  of  nitrogen.  Extracts  were  further
purified  by  HPLC-SEC.  Co-extracted  silicone  oligomers  were
separated  by  F1PLC-SEC  (injection  volume  0.5  mL)  with
hexane/dichloromethane  (80:20),  whereby  the  first  fraction

Table  3  Comparison  of  different  silicone  rubber  pre-clean-up  methods  regarding  solvent  and  time

Clean-up  method

According  to  reference

Total  solvent  (mL)

Time  (h)

Swelling  (%)

Release  of  oligomers  (%)

Soxhlet  extraction  ethylacetate

Smedes  and  Booij  (2012)

400

100

92

2.5

Extraction  ethylacetate

Shahpoury  and  Hageman  (2013)

400

48

76

2.4

Extraction  n-hexane/acetone  (3:1)

Schafer  et  al.(2010)

800

96

124

2.5

ASE  n-hexane/acetone  (1:1)

This  study

169

1.2

42

2.2

Results  of  experimental  approach  using  different  clean-up  methods  in  regard  to  swelling  and  the  associated  silicone  oligomer  release  in  %