Hollow-fiber liquid phase microextraction for lignin pyrolysis acids in aerosol samples and gas chromatography-mass spectrometry analysis.

Abstract

A method based on three-phase hollow fiber liquid phase microextraction was developed and successfully applied to aerosols for the analysis of lignin pyrolysis acids such as syringic acid, vanillic acid and p-salicylic acid. Important parameters related to extraction process like organic solvent for membrane phase, tri-n-octylphosphine (TOPO) oxide contents in organic solvent, stirring speed, extraction time etc. were optimized. 6-Undecanone with 15% TOPO contents (w/v) was found a suitable solvent for organic liquid membrane, 900 rpm was the optimum stirring speed and time of 4h was found optimum extraction time. Donor phase pH was 1.3 while acceptor phase pH was adjusted to 9.5. The optimized extraction method was used for the extraction of real aerosol samples. Analytes were derivatized using BSTFA containing 1% trimethylsilyl chloride and gas chromatography mass spectrometry was used for analysis. Very low limits of detection in the range 0.2-1.0 ng L(-1) were found, corresponding to 10-50 pg m(-3) of analytes in aerosols. Extraction efficiency obtained ranged 60.3-71.7% and enrichment factors ranged 3015-3585 times. The optimized method was successfully applied to aerosol samples and all of the selected analytes were detected in the analyzed samples.

Full text

Page 1

Journal
of
Chromatography
A,
1249 (2012) 48–
53
Contents
lists
available
at SciVerse
ScienceDirect
Journal
of
Chromatography
A
j our
na
l ho
me
p ag e:
www.elsevier.com/locate/chroma
Hollow-fiber
samples
liquid
phase
microextraction
for
lignin
pyrolysis
acids
in
aerosol
and
gas
chromatography–mass
spectrometry
analysis
Murtaza
Hyder∗,
Jan
Åke
Jönsson
Centre
for
Analysis
and
Synthesis,
Department
of
Chemistry,
Lund
University,
Sweden
a
r
t
i
c
l
e

i
n
f
o
Article
Received
Received
Accepted
Available online 21 June 2012
history:
6 February
2012
in revised
form
11 June
2012
14 June
2012
Keywords:
Lignin
Aerosols
Enrichment
Hollow
pyrolysis
acids
fiber
liquid
phase
microextraction
a
b
s
t
r
a
c
t
A
applied
salicylic
phase,
were
liquid
time.
method
containing
Very
analytes
3015–3585
selected
method
based
on
three-phase
hollow
fiber
liquid
phase
microextraction
was
developed
and
successfully
to
aerosols
for
the
analysis
of lignin
pyrolysis
acids
such
as
syringic
acid,
vanillic
acid
and
p-
acid.
Important
parameters
related
to
extraction
process
like
organic
solvent
for
membrane
tri-n-octylphosphine
(TOPO)
oxide
contents
in
organic
solvent,
stirring
speed,
extraction
time
etc.
optimized.
6-Undecanone
with
15%
TOPO
contents
(w/v)
was
found
a
suitable
solvent
for
organic
membrane,
900
rpm
was
the
optimum
stirring
speed
and
time
of
4
h
was
found
optimum
extraction
Donor
phase
pH
was
1.3
while
acceptor
phase
pH
was
adjusted
to
9.5.
The
optimized
extraction
was
used
for
the
extraction
of
real
aerosol
samples.
Analytes
were
derivatized
using
BSTFA
1%
trimethylsilyl
chloride
and
gas
chromatography
mass
spectrometry
was
used
for
analysis.
low
limits
of
detection
in
the
range
0.2–1.0
ng
L−1were
found,
corresponding
to
10–50
pg
m−3of
in
aerosols.
Extraction
efficiency
obtained
ranged
60.3–71.7%
and
enrichment
factors
ranged
times.
The
optimized
method
was
successfully
applied
to
aerosol
samples
and
all
of
the
analytes
were
detected
in
the
analyzed
samples.
© 2012 Elsevier B.V. All rights reserved.
1.
Introduction
Biomass
substances
sources
to
cellulose
ucts
particulate
organic
as
in
Lignin
cellulose
to
an
sors
and
lignin
ing
products
burning
is an important
primary
source
of
many
trace
to the
atmosphere.
The
chemical
signature
of
emission
is
important
for
knowing
the
input
from
different
sources
aerosol
mass
[1].
Biomass,
mainly
wood,
consists
of
biopolymers,
and
lignin
and
its
thermal
degradation
(pyrolysis)
prod-
have
been
studied
as
tracers
of
biomass
burning
[2].
The
smoke
matter
composition
has
been
extensively
studied
for
its
contents
[3,4].
Levoglucosan
and
other
anhydrous
sugars
cellulose
pyrolysis
products
have
been
considered
main
tracers
atmospheric
aerosols
for
biomass
burning.
is
the
third
most
abundant
biopolymer
in
wood
after
and
hemicellulose
[5]. Lignin
contents
range
from
25%
35%
for
softwood
and
18%
to 25%
for
hardwood
[6,7].
Lignin
is
amorphous
polymer
of
phenylpropane
units,
where
the
precur-
aromatic
alcohols
(monolignols)
namely
p-coumaryl,
coniferyl
sinapyl
alcohols
polymerize
in a random
way
[8].
Burning
of
biopolymer
produces
a variety
of
pyrolysis
products
includ-
phenols,
aldehyde,
ketones,
and
alcohols.
These
breakdown
generally
retain
the
original
substituents
(OH,
OCH3) on
∗Corresponding
E-mail
smsmurtaza@yahoo.com.au
author.
Tel.:
+46
765646632.
addresses:
Murtaza.Hyder@organic.lu.se,
(M.
Hyder).
the
4-hydroxybenzoic
OH
particulate
These
wood
tracers
Information
helpful
biomass
atmospheric
The
phase
level
acids
solvents
extraction
These
ple,
organic
large
ment
methods
of
requirements
lytes
phenyl
ring
[9,10].
Syringic
acid,
vanillic
acid,
p-anisic
acid
and
acid
are
some
aromatic
acids
with
OCH3and
substituent
on phenyl
ring
that
have
been
detected
in smoke
matter
in various
studies
[11–13].
lignin
pyrolysis
products
have
been
used
as
tracers
for
burning
in urban
areas
[9,14].
The
concentrations
of these
are
helpful
in the
source
apportionment
of organic
aerosols.
about
contents
of lignin
pyrolysis
products
is
further
in distinguishing
the
contribution
from
different
kinds
of
(softwood,
hardwood
and
grasses)
burning
toward
the
aerosols
[2].
aim
of
this
study
was
to develop
a hollow
fiber
liquid
microextraction
based
analytical
method
to meet
the
trace
detections
and
quantification
requirement
of lignin
pyrolysis
from
aerosols.
Commonly
extraction
methods
using
organic
or
water
with
ultrasonic
assistance,
etc.
are
used
for
of
polar
analytes
like
acids
from
aerosol
samples
[15,16].
conventional
methods
have
several
disadvantages;
for exam-
those
methods
not
only
use
large
amounts
of
pure,
costly
solvents
but
also
they
are
not
environment
friendly,
as
amounts
of
organic
solvents
are
evaporated
to the
environ-
during
pre-concentration
step.
Furthermore,
these
extraction
are
not
selective
and
do not
provide
high
enrichment
analytes.
Therefore
these
extraction
methods
do
not
meet
the
for
trace
level
detection
and
quantification
of
ana-
such
as
lignin
pyrolysis
acids.
0021-9673/$
http://dx.doi.org/10.1016/j.chroma.2012.06.039
– see
front
matter ©

2012 Elsevier B.V. All rights reserved.

Page 2

M.
Hyder,
J.Å.
Jönsson
/ J. Chromatogr.
A 1249 (2012) 48–
53
49
In present
microextraction
overcome
aerosol
inside
donor
of
three-phase.
of
ous,
phase.
gas
provides
phase
with
the
of
in
better
its
level
Liquid
for
samples.
ticulate
first
to aerosols
study
we
developed
a hollow
fiber
liquid
phase
(HF-LPME)
method
for
lignin
pyrolysis
acids
that
all
the
disadvantages
associated
with
conventional
extraction
techniques.
In HF-LPME
the
accepter
phase
is
the
lumen
of
hollow
fiber
and
it is
completely
separated
from
phase
by
an organic
membrane
phase
held
in the
porous
wall
hollow
fiber.
There
are
two
variants
of
HF-LPME,
two-phase
and
In two-phase
HF-LPME
the
acceptor
phase
in lumen
hollow
fiber
is an organic
solvent
while
the
donor
phase
is aque-
and
membrane
phase
in the
walls
of
fiber
is same
as
acceptor
Two-phase
HF-LPME
is
generally
used
where
a
subsequent
chromatographic
mass
spectrometry
analysis
is desired,
as
it
analytes
in organic
solvent
in the
extract
[17]. In three-
HF-LPME
the
acceptor
and
donor
phases
are
aqueous
usually
the
pH values
adjusted
to some
specific
values
depending
on
nature
of
analytes
and
an
organic
membrane
phase,
in the
wall
hollow
fiber,
separates
them.
We
preferred
three-phase
HF-LPME
the
present
study
because
it provides
more
selective
extraction,
clean
up and
high
enrichment
factors
allowing
very
low
lim-
of
detection
and
quantification,
suitable
for
trace
and
ultra
trace
analysis
[18].
membrane
extraction
techniques
are
generally
applied
extraction
and
pre-concentration
of
analytes
from
liquid
phase
These
techniques
have
not
been
applied
to airborne
par-
matter
for
extraction
and
enrichment
of
analytes.
For
the
time,
we
applied
liquid
membrane
based
extraction
method
in our
recent
studies
[19,20].
2.
Materials
and
methods
2.1.
Reagents
and
standards
The
standards
of
syringic
acid,
vanillic
acid,
p-salicylic
acid
(St.
(TOPO)
Organics
bis(trimethylsilyl)trifluoroacetamide
trimethylsilyl
also
was
gas
enborg,
gradient
Q3/2
(200
were
The
obtained
Individual
anisic
of
acids
from
500
1-phenyldodecane
hexane.
The
directcalibration
standard
solutions
(0.020–2.000
derivatization
analyzed
In the
LPME
acid,
and
p-anisic
acid
were
purchased
from
Sigma–Aldrich
Louis,
MO,
USA);
n-hexane,
tri-n-octylphosphine
oxide
and
1-phenyldodecane
(97%)
were
provided
by Acros
(Geel,
Belgium).
Acetone
(HPLC
(BSTFA
grade)
and
excluding
N,O-
≥ 98%,

chloride)
containing
1% trimethylsilyl
chloride
were
purchased
from
Sigma–Aldrich.
Methanol
(99.9%
HPLC
grade)
provided
by
Fisher
Scientific
(Waltham,
MA,
USA).
Helium
(99.9995%)
was
provided
by
Strandmøllen
Lab
Line
(Klamp-
Denmark).
Ultrapure
reagent
water
purified
by
a Milli-Q
system
(Millipore,
Bedford,
MA,
USA)
was
also
used.
The
Accurel
PP
polypropylene
hollow-fiber
membranes
(HF)
?m
wall
thickness,
600
?m inner
diameter,
0.2
?m
pore
size)
obtained
from
Membrana
GmbH
(Wuppertal,
Germany).
47 mm
quartz
filters
(Pall
TissuquartzTM, binder
free)
were
from
Sigma–Aldrich.
standard
solutions
of
syringic
acid,
vanillic
acid,
p-
acid
and
p-salicylic
acid
were
prepared
at
concentrations
100
?g mL−1in methanol.
A mixture
solution
containing
these
was
prepared
at a concentration
of
2 ?g mL−1in methanol
each
individual
solution.
An
internal
standard
solution
of
?g mL−1of
1-phenyldodecane
was
prepared
in n-hexane.
Then
solution
was
diluted
to get
2 ?g mL−1in n-
All
the
solutions
were
stored
curves
different
in the
were
refrigerator
established
concentrations
at
4◦C.

from
levels

at

six

?g mL−1) by injecting
them
into
the
GC–MS
after
with
BSTFA.
Each
level
of
concentration
was
in triplicate.
present
study
we
have
applied
the
three-phase
HF-
to aerosol
samples
for
extraction
of
syringic
acid,
vanillic
p-salicylic
acid
and
p-anisic
acid
and
obtained
very
low
limits
immersed
level
fiber
brane
in
was
chromatography
of
detection
in range
of
a few
pg m−3. Aerosols
samples
were
in aqueous
donor
phase
with
an adjusted
pH to a low
and
stirred
at 900
rpm
for
15
min.
Then
a prepared
hollow
with
6-undecanone
containing
15%
TOPO
as
organic
mem-
phase
in the
porous
wall
of
hollow
fiber
and
acceptor
phase
the
lumen
of
the
hollow
fiber
was
used
for
extraction.
The
extract
evaporated,
derivatized
with
BSTFA
and
analyzed
using
gas
mass
spectrometry.
2.2.
Sample
collection
The
sampling
site
Vavihill
is a
EUSAAR
and
EMEP
(European
Monitoring
ated
located
tances
Helsingborg
25
Samples
TissuquartzTM, binder
by
with
4
Petri
freezer
and
Evaluation
Programme)
background
station
situ-
in Southern
Sweden
(56◦01?N,
13◦09?E,
172).
The
station
is
not
imminently
to any
local
pollution
sources,
although
the
dis-
to the
densely
populated
areas
of Malmö,
Copenhagen
and
west
to southwest
of
the
station
are
only
45,
50
and
km,
respectively.

were

collected

on

47
mm

quartz

filters

(Pall
free,
PALL
Life
Sciences,
Ann
Arbor,
MI)
an aerosol
flow
of
38
L min−1from
a PM10(particulate
matter
diameter
≤10
?m)
inlet.
Filters
were
baked
in 900◦C for
h prior
to sampling.
After
sampling
the
filters
were
stored
in a
dish,
wrapped
in aluminum
foil
in the
refrigerator
(+8◦C)
or
(−30◦C).
2.3.
microextraction
Preparations
for
3-phase
hollow
fiber
liquid
phase
Three-phase
hollow
fiber
liquid
phase
microextraction
is
based
on
This
donor
an
hydrophobic
hollow-fiber
brane.
The
mately
10
acceptor
pH
the
end
that
a few
the
fiber
20
at
with
system
(in
the
tion
flask.
stirring
filter
in
configuration
and
under
After
the
the
the
principle
of
supported
liquid
membrane
(SLM)
extraction.
technique
involves
a process
of analyte
enrichment
from
a
phase
(aqueous)
to an acceptor
phase
(aqueous)
through
organic
membrane
phase
that
is immobilized
in pores
of some
support.
We
used
a polypropylene
micro-porous
as
a
hydrophobic
support
for
the
organic
liquid
mem-
hollow
fiber
was
cut
into
6 cm long
pieces
(with
approxi-
20
?L lumen
volume).
Those
were
washed
with
acetone
for
min
in a sonicator
and
then
dried
at room
temperature.
50 ?L of
solution
(0.1
M (NH4)2CO3solution
in reagent
water
with
adjusted
to 9.5)
was
withdrawn
into
a 100
?L syringe.
The
tip
of
syringe
was
inserted
into
the
lumen
of
the
hollow
fiber
from
one
of
fiber.
The
fiber
was
immersed
into
liquid
membrane
phase
is an
organic
solvent
(6-undecanone
containing
15%
TOPO)
for
seconds.
Then
the
acceptor
from
the
syringe
was
pushed
into
lumen
of the
fiber.
First
30
?L of
acceptor
was
flushed
out
of
the
removing
any
organic
solvent
inside
the
lumen.
The
remaining
?L of
acceptor
was
left
in the
lumen.
The
fiber
was
then
sealed
the
open
end
with
the
help
of
a
small
piece
of
aluminum
foil
syringe
inserted
into
the
other
end
of
fiber.
In this
way
an SLM
was
obtained
in which
a column
of
20
?L acceptor
solution
the
lumen
of
fiber)
was
covered
with
an
organic
solvent
held
in
pores
of
the
hollow
fiber
wall.
100
mL
of
donor
(sample)
solu-
(0.05
M H2SO4in reagent
water)
was
taken
into
a volumetric
A magnet
bar
was
put
in the
flask,
which
was
placed
on
a
plate
and
either
spiked
with
analyte
solutions
or pieces
of
containing
analytes
for
extraction.
The
syringe
was
clamped
such
a
way
that
fiber
was
dipped
into
the
donor
solution.
This
is
similar
to the
one
described
by Zhao
and
Lee
[21]
Hyder
et
al.
[20]. Extraction
was
carried
out
for
a
suitable
time
a selected
stirring
speed.
extraction
the
fiber
was
taken
out
of
donor
phase
and
then
acceptor
phase
from
the
lumen
of
the
fiber
was
pulled
back
into
syringe
and
the
extract
was
transferred
to a GC
vial
equipped

Page 3

50
M.
Hyder,
J.Å.
Jönsson
/ J. Chromatogr.
A 1249 (2012) 48–
53
Table 1
Parameters
for
chromatographic
identification,
calibrations
and
method
validation
parameters.
Syringic
acid
(pKa= 4.33)

Vanillic
acid
(pKa= 4.45)

p-Anisic
acid
(pKa= 4.47)

p-Salicylic
acid
(pKa=
4.48)
Retention
Identification
Linearity
R2
LOD
LOD
Repeatability
Reproducibility
Enrichment
Extraction
time
(min)

29.1
297
10–400
0.9964
1.0
50
10.2
12.6
3015
60.3

30.6
297
10–400
0.9987
0.5
25
7.6
8.7
3305
66.1

25.05
209
10–400
0.9994
0.2
10
5.2
6.4
3230
64.4

27.15
223,
10–400
0.9992
0.2
10
5.0
7.3
3585
71.7
and
quantification
ions
267
range
(ng
L−1)

(ng
L−1)
(pg
m−3)a

(RSD
%),
n = 5
(RSD
%),
n = 10
factor

efficiency
(%)

aConcentration
ng
in terms
of
ng m−3was
calculated
by
supposing
that
filter
was
exposed
to 250
m3of
air
to collect
particulate
matter
giving
analyte
mass
equal
to LOD
in
L−1on
0.5
cm2of filter.
with
formed
establish
in
under
(IS)
of
trimethylsilyl
sealed
allowed
1 h.
to
a 300
?L glass
insert.
Derivatization
using
BSTFA
was
per-
using
the
method
described
by Pietrogrande
et
al.
[22]. To
calibration
curves,
10 ?L of
standard
solutions
was
taken
a GC
vial
with
a 300
?L glass
insert.
It was
evaporated
to dryness
a stream
of
nitrogen
at 60◦C.
Then
15 ?L of
internal
standard
1-phenyldodecane
solution
(2 ?g mL−1in n-hexane)
and
10 ?L
N,O-bis(trimethylsilyl)trifluoroacetamide
(BSTFA)
containing
1%
chloride
was
added
to each
vial.
The
vials
were
air
with
a screw
cap
with
a
Teflon
septum.
Derivatization
was
by putting
the
vials
in an
oven
at a temperature
of
80◦C for
Then
the
vial
was
cooled
to room
temperature
and
transferred
the
autosampler
on
the
GC–MS
for
analysis.
2.4.
GC–MS
analysis
All the
matograph
sampler
gies,
HP-5ms
of
The
increased
220◦C and
(total
(99.9995%)
ture
injection
source
The
70
tion
For
itoring
for
were
analyte.
each
analyses
were
performed
using
a 6890
series
gas
chro-
(GC)
equipped
with
Agilent
7683
series
automatic
liquid
and
a 5973-N
mass
selective
detector
(Agilent
Technolo-
Palo
Alto,
USA).
Analytes
were
separated
using
an
Agilent
capillary
GC
column
(Agilent
Technologies
19091S-433)
30 m
× 0.25
mm
with
a phase
thickness
of
0.25
?m.
optimized
temperature
program
was:
70◦C,
hold
2 min,
at 2.5◦C min−1to 120◦C,
increased
at 10◦C min−1to
increased
at
20◦C min−1to a final
temperature
of 300◦C
analysis
time:
36 min).
The
carrier
gas
used
was
helium
with
a flow
rate
of
1.5
mL
min−1. The
injector
tempera-
was
290◦C, and
the
injection
was
made
in splitless
mode.
The
volume
was
1 ?L. The
transfer
line,
quadrupole
and
ion
temperatures
were
280◦C, 180◦C and
250◦C respectively.
MS
operated
in the
electron
impact
ionization
mode
(EI)
at
eV.
Scan
mode
was
used
for
the
standards
and
for
the
identifica-
of each
compound.
The
scan
mass
range
was
set
to 50–500
m/z.
the
quantification
of
analytes
in the
samples,
selective
ion
mon-
(SIM)
mode
was
used.
The
most
abundant
ion
was
used
quantification
in each
case,
while
second
most
abundant
ions
usually
used
for
the
identification
and
confirmation
of
the
The
ions
used
for
the
identification
and
quantification
of
compound
are
shown
in Table
1.
3.
Results
and
discussion
3.1.
Method
optimization
In
three-phase
important
the
[23].
position
extraction
on
brane
HF-LPME
there
are
many
parameters
that
are
and
need
to be
considered.
These
parameters
determine
extraction
efficiency,
enrichment
and
selectivity
of
analytes
This
includes
the
pH of
donor
and
acceptor
phases,
com-
of
organic
liquid
membrane
phase,
stirring
speed
and
time.
In three
phase
HF-LPME
two
extractions
are
going
at
the
same
time;
from
donor
phase
to organic
liquid
mem-
phase
and
from
membrane
phase
to acceptor
phase
[24].
So
the
uid
into
extraction
a careful
selection
of
above-mentioned
parameters
can
lead
to
extraction
of target
analytes
from
donor
phase
to organic
liq-
membrane
and
then
a complete
trapping
of target
analytes
acceptor
phase
and
this
provides
high
enrichment
factor
and
efficiency.
3.2.
The
donor
and
acceptor
phase
composition
For
getting
high
enrichment
factors
and
selective
extraction
of
analytes
phase
transfer
must
brane
molecules
[23,25].
pKavalue
a
of
role
acceptor
greater
phase
In
acceptor
acceptor
of
values
pH
pH
tor
experiments
to interferences
in the
analytical
process,
pH of
both
donor
and
acceptor
phase
is of
utmost
importance.
For
an easy
of
the
analytes
from
donor
to acceptor
phase,
the
analytes
be
in a form
that
is easily
dissolved
into
the
organic
mem-
phase.
Generally,
organic
phases
liquid
membranes
allow
in uncharged
form
(non
ionized)
to diffuse
through
them
An
important
parameter
in selection
of the
donor
pH is the
of
target
analytes.
Thus,
acidic
pH of
the
donor
phase
with
value
less
than
pKaof
target
analytes
will
facilitate
the
extraction
acidic
analytes
[25]. The
pH of
the
acceptor
phase
plays
a vital
in trapping
of
target
analytes.
For
acidic
analytes
the
pH of
phase
is adjusted
to high
enough
value
(about
3.3
units
than
pKavalue)
to completely
trap
the
analytes
in acceptor
[26].
the
present
study
we
did
not
optimize
the
pH of
donor
and
phase
as
it is well
established
in many
studies
that
pH
of
and
donor
phase
can
be
selected
based
on the
pKavalues
target
analytes
as
described
above.
In our
case
the
analyte
pKa
ranged
from
4.33
to 4.48.
So we
selected
the
donor
phase
value
at
1.3
to have
analytes
in an
uncharged
form
and
acceptor
was
adjusted
to 9.5
to completely
trap
the
analytes
in accep-
phase.
We
used
0.05
M H2SO4solution
as
donor
phase
for
our
and
0.1
M (NH4)2CO3solution
acceptor
phase.
3.3.
Selection
of
organic
liquid
for membrane
phase
Selection
another
be
ubility
the
give
ious
in
ble
extraction
vents
We
n-hexyl
our
donor
vations
of
a suitable
organic
solvent
as
a
liquid
membrane
is
critical
parameter.
It is desired
that
the
organic
solvent
to
selected
should
have
high
affinity
for
target
analytes
and
its
sol-
to donor
(aqueous)
phase
should
very
low.
The
viscosity
of
solvent
should
also
be
medium,
as
highly
viscous
solvent
will
a thick
solvent
membrane
and
slower
mass
transfer
[26]. Var-
organic
solvents
have
been
used
as
organic
liquid
membrane
different
studies.
Organic
solvents
that
have
polarity
compara-
to that
of
the
target
analytes
have
been
shown
to give
optimum
efficiency
and
enrichment.
Sometimes
a
mixture
of
sol-
can
be
used
as
well
[27].
compared
three
different
organic
solvents,
n-undecane,
di-
ether
(DHE)
and
6-undecanone
to find
a suitable
one
for
analytes.
Extraction
was
carried
out
for
2 h at 900
rpm
with
phase
pH at
1.3
and
acceptor
phase
pH at 9.5.
Three
obser-
were
made
for
every
solvent.
100
mL
of
donor
phase
was

Page 4

M.
Hyder,
J.Å.
Jönsson
/ J. Chromatogr.
A 1249 (2012) 48–
53
51
Fig.
tion
phase
9.5
1. Effect
of
organic
membrane
phase
solvent
on
enrichment
of
analytes,
extrac-
parameters
were
as,
stirring
speed
900
rpm,
extraction
time
2 h,
100
mL
of
donor
with
pH at 1.3
and
spiked
to give
400
ng L−1of
analytes,
acceptor
phase
pH
and
one
solvent
at
a time.
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
16 14 121086420
Analyte peak area / IS peak area
TOPO contents (%)
Vanillic Acid
p-Anisic Acidp-Salicylic AcidSyringic Acid
Fig.
conditions
and
multiplied
2. Effect
of
TOPO
contents
(%,
w/v)
in organic
liquid
membrane
phase,
extraction
were
900
rpm,
2 h extraction
time,
100
mL
of donor
phase
with
pH at 1.3
spiked
at 400
ng
L−1and
acceptor
phase
pH 9.5
(syringic
acid
response
was
by
20 to get
it into
scale).
spiked
to give
lytes
internal
low
while
slightly
low.
as
molecule
all
we
study.
with
20 ?L of
all
acids
mixture
of
concentration
2 ?g mL−1
400
ng
L−1of
analytes.
Fig.
1 shows
the
enrichment
of
ana-
in terms
of
the
ratio
of analytes
peak
area
to that
of the
standard
(IS).
The
non-polar
solvent,
n-undecane
gave
very
enrichment
for
vanillic
acid,
p-anisic
acid
and
p-salicylic
acid
we
could
not
get
any
extraction
of
syringic
acid.
DHE
gave
better
enrichment
for
all
studied
analytes
but
still
very
6-Undecanone
gives
high
enrichment
for
all
studied
analytes
compare
to the
solvents
discussed
before.
The
6-undecanone
contains
carbonyl
( C O)
group,
which
is also
present
in
target
analytes
so its
polarity
is
comparable
to analytes.
Hence
selected
6-undecanone
as
optimum
organic
solvent
for
this
3.4.
organic
Influence
of tri-n-octylphosphine
oxide
contents
in the
membrane
phase
TOPO
enhances
the
flux
of
analytes
from
donor
to acceptor
phase
through
atom
compounds
gen
We
organic
tried
ing
Section
Fig.
brane
the
membrane.
It has
a
lone
electron
pair
on
the
oxygen
in the
molecule
that
enables
it to form
complex
with
polar
such
as
carboxylic
acids
and
alcohols
through
hydro-
bonding,
and
it acts
as
an
efficient
extractant
in LPME
[28].
tried
TOPO
contents
in a range
of
0–15%
in 6-undecanone
as
liquid
membrane
phase.
Higher
contents
of
TOPO
were
not
because
of
its
limited
solubility
in 6-undecanone.
For
study-
the
effect
of
TOPO,
extraction
was
carried
out
as
described
in
3.3.
2 shows
the
effect
of
TOPO
contents
in organic
liquid
mem-
phase
on
enrichment
of analytes
expressed
in terms
of
ratio
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
1000900800 700 600500400
Stirring Speed (rpm)
300 2001000
Analyte peak area / IS peak area
Vanillic AcidSyringic Acidp-Anisic Acidp-Salicylic Acid
Fig.
2
1.3)
pH
3. Effect
of
stirring
speed
on
analytes
enrichment,
extraction
carried
out
for
h,
15%
TOPO
in 6-undecanone
as liquid
membrane,
100
mL
of
donor
phase
(pH
was
spiked
with
analytes
to give
400
ng L−1concentration
and
acceptor
phase
was
9.5.
of
was
membrane
case
for
groups
methoxy
as
neighborhood
analytes
response
to that
of
internal
standard.
The
enrichment
found
to increase
with
the
increase
of
TOPO
contents
in organic
phase.
The
influence
of
TOPO
was
most
OH
prominent
in
of
p-salicylic
acid
because
it has
two
groups
available
H-bonding.
Syringic
acid
and
OH group
vanillic
acids
also
have
two

OH
but,
one
of
the
is hindered
by neighboring
( OCH3) groups.
Vanillic
acid
is having
better
enrichment
compared
to syringic
acid
as
it has
just
one
methoxy
group
in
of

OH.
3.5.
Influence
of stirring
speed
Extraction
analytes
reduces
improved
speed,
the
observed
the
relative
We
ing
terms
The
stirring
was
experiments.
efficiency
depends
on
the
time
of
contact
between
and
the
organic
membrane
[26,29]. Donor
phase
stirring
the
extraction
time
to reach
optimum
enrichment
due
to
mass
transfer
of
analytes.
However,
at
too
high
a stirring
the
mass
transfer
may
become
limited
by
the
dissolution
of
membrane
liquid
[26,30]. At
too
high
stirring
speed,
it has
been
that
bubbles
are
produced
in donor
phase
that
attach
to
fiber
surface
resulting
in lower
extraction
efficiency
and
high
standard
deviation
(RSD)
[26].
performed
the
extraction
at different
stirring
speed
rang-
from
300
to 900
rpm.
Fig.
3 shows
enrichment
of
analytes
in
of
ratio
of
analytes
peak
area
to internal
standard
peak
area.
enrichment
for
all
the
analytes
was
found
to increase
with
the
speed.
Maximum
enrichment
was
found
at 900
rpm
so it
selected
as
optimum
stirring
speed
and
was
used
for
further
3.6.
Influence
to extraction
time
Optimization
good
put.
with
efficiency
continuously
performed
lytes
of
standard
studied
4
vanillic
further
of
extraction
time
is
important
not
only
to achieve
extraction,
but
also
to obtain
a
reasonable
sample
through-
The
higher
the
time
allowed
for
analytes
to come
in contact
organic
liquid
membrane
the
higher
will
be
the
extraction
provided
that
the
organic
solvent
of
membrane
phase
is
depleted
for
analytes
by acceptor
phase
[31,32].
We
extractions
in order
to observe
the
enrichment
of
ana-
as
a
function
of extraction
time.
Fig.
4 shows
the
enrichment
analytes
in term
of
ratio
of
peak
areas
of analytes
to internal
peak
area
at different
extraction
times.
All
the
analytes
showed
an
increase
in the
enrichment
with
the
time.
After
h there
was
slight
decrease
in the
enrichment
of
p-anisic
acid
and
acid.
So we
selected
4 h as
optimum
extraction
time
for
experiments.

Page 5

52
M.
Hyder,
J.Å.
Jönsson
/ J. Chromatogr.
A 1249 (2012) 48–
53
0
0.2
0.4
0.6
0.8
1
1.2
76543210
Analyte peak area / IS peak area
Time Hours
Vanillic Acid Syringic Acd p-Anisic Acidp-Salicylic Acid
Fig.
tion
donor
and
4.
Enrichment
of analytes
as a function
of
extraction
time,
extraction
condi-
were,
15%
TOPO
in 6-undecanone
as liquid
membrane,
900
rpm,
100
mL
of
phase
(pH
1.3)
was
spiked
with
analytes
to give
400
ng
L−1concentration
acceptor
phase
pH
was
9.5.
3.7.
Method
validation
The
practical
applicability
of
HF-LPME
to aerosols
is
determined
by
(LOD),
tigating
filters
spiked
tion
24
Then
stirred
using
were
ratios
used
of
of standard
finding
the
repeatability,
reproducibility,
limits
of
detection
limits
of quantification
(LOQ)
and
linearity
range.
For
inves-
the
repeatability
and
reproducibility,
pieces
(n = 10)
of
(similar
to those
used
for
sampling)
with
area
0.5
cm2were
with
20
?L of
analyte
mixture
(standards)
with
concentra-
2 ?g mL−1of
each
analyte.
Filters
were
kept
in a
refrigerator
for
h to get
maximum
possible
adsorption
of
analytes
to the
filters.
5 filter
pieces
were
immersed,
each
in 100
mL
of
donor
phase,
vigorously
for
15 min
and
then
extracted
as
described
above
the
optimum
conditions
and
analyzed,
other
5 filter
pieces
extracted
and
analyzed
in a similar
way
on the
next
day.
The
of
peak
areas
of
analytes
to internal
standard
peak
area
were
for
calculations
of
RSD.
For
finding
the
linearity
range,
pieces
filters
with
area
0.5
cm2(in
triplicates)
were
spiked
with
20 ?L
solutions
at different
concentrations
ranging
from
0.05
to
After
and
under
tized
minimum
imum
parameters
2 ?g mL−1, and
then
the
filters
were
kept
in refrigerator
for
24
h.
that,
the
filter
pieces
were
immersed
in 100
mL
of
donor
phase
stirred
vigorously
for
15
min.
All
extractions
were
carried
out
optimum
conditions
using
HF-LPME,
extracts
were
deriva-
and
analyzed
by
GC–MS.
Limits
of
detection
were
taken
as
the
concentration
providing
chromatographic
signals
min-
3 times
higher
than
background
noise.
Table
1 shows
the
estimated
for
method
validation.
3.8.
Enrichment
factor
and
extraction
efficiency
The
enrichment
factor
(Ee) is the
measure
of
accumulation
of
analytes
given
to
expressed
in the
acceptor
phase
as
compared
to donor
phase
and
is
by
the
ratio
of
concentration
of an
analyte
in acceptor
phase
the
initial
concentration
of
the
analyte
in donor
phase.
It can
be
as
Ee=Ca
Ci
where
after
donor
of
lyte
Cais the
concentration
of an analyte
in the
acceptor
phase
extraction
and
Ciis the
initial
concentration
of the
analyte
in
phase.
The
extraction
efficiency
(E)
is defined
as
the
amount
an analytes
extracted
in the
acceptor
phase
out
of
the
total
ana-
amount
in donor
phase
and
is expressed
as
E
=na
nt
× 100
where
and
in
In order
tors
sampling)
spiked
tion
piece.
nais the
number
of
moles
of an analyte
in the
acceptor
phase
ntis the
total
number
of
moles
of
the
analyte
present
initially
donor
phase.
to obtain
the
extraction
efficiency
and
enrichment
fac-
we
took
12
pieces
of
filters
(similar
to those
used
for
aerosol
each
with
an area
of
0.5
cm2. Every
piece
of filter
was
with
20 ?L of
mixture
of
analytes
solution
with
concentra-
2 ?g mL−1giving
an
amount
of
40 ng
of each
analyte
on
each
These
filter
pieces
were
kept
in refrigerator
for
24
h to allow
Fig.
5.
Total
ion
chromatograms
obtained
by injecting
standard
solution
and
sample
extract,
while
selected
ion
chromatogram
is shown
in windows
for
studied
analytes.

Page 6

M.
Hyder,
J.Å.
Jönsson
/ J. Chromatogr.
A 1249 (2012) 48–
53
53
Table 2
Concentrations
of
analytes
for
aerosol
samples
analyzed
expressed
in terms
of
ng m−3.
Syringic
acid
(ng
m−3)
Vanillic
acid
(ng
m−3)
p-Anisic
acid
(ng
m−3)
p-Salicylic
acid
(ng
m−3)
11/02/2009
17/02/2009
25/02/2009
06/03/2009
22/03/2009
09/04/2009

1.21
0.92
0.95
1.07
0.70
0.44

1.82
1.97
1.21
1.58
1.06
0.94

0.96
1.09
1.25
0.80
0.63
0.69

2.05
1.83
2.44
1.92
1.47
0.85





analytes
sample
in
lytes
Extracts
analyzed
were
the
enrichment
to adsorb
well
and
to make
filters
similar
to the
aerosol
as
much
as
possible.
Then
every
piece
of
the
filter
was
placed
100
mL
of
donor
phase
and
stirred
vigorously
for
15 min
and
ana-
were
extracted
using
HF-LPME
under
optimum
conditions.
were
dried
under
a gentle
flow
of
nitrogen,
derivatized
and
using
GC–MS.
The
peak
areas
for
the
analytes
obtained
compared
with
direct
calibrations
with
standards
to estimate
amounts
of
analytes
extracted
and
extraction
efficiency
and
factors
were
calculated
(Table
1).
3.9.
Aerosol
samples
analysis
The
optimized
method
was
also
used
for
real
sample
analysis.
The
piece
orously
using
derivitized
son
sample
Totally
detected
the
which
year.
aerosol
sample
filters
were
punched
into
0.5
cm2pieces.
Each
was
immersed
into
100
mL
of
donor
phase
and
stirred
vig-
for
15
min
and
then
extracted
under
optimum
conditions
HF-LPME.
Extracts
were
dried
under
gentle
flow
of
nitrogen,
and
analyzed
using
GC–MS.
Fig.
5 shows
a compari-
of chromatograms
obtained
by
injecting
standard
solutions
and
extract.
6 aerosol
samples
were
analyzed.
All
the
analytes
were
in all
the
samples
analyzed.
All
the
analytes
were
found
in
lowest
concentrations
for
sample
collected
on
9th
of
April
2009,
may
be associated
to less
biomass
burning
than
earlier
in the
All
the
concentrations
are
shown
in Table
2.
4.
Conclusion
We developed
and
be
lignin
high
of
trace
samples.
a three-phase
HF-LPME
method
for
extraction
pre-concentration
of
lignin
pyrolysis
acids.
The
method
can
successfully
applied
to aerosols
for
analytes
such
as
acids
from
pyrolysis
to give
very
low
limits
of
detection
and
very
enrichment
factors,
using
only
a small
(few
?L) amount
organic
solvent.
Thus,
three-phase
HF-LPME
is very
useful
in
level
detection
and
quantification
of
analytes
in aerosol
References
[1]
[2] B.R.T.
[3] S.B.
(1988)
[4] S.B.
23
[5] A.U.
[6] R.M.
(Ed.),
York,
[7]
Fuel
[8]
[9] S.B.
26 (1992)
[10]
[11]
[12] P.M.
[13]
[14]
Sci.
[15] A.
Chem.
[16]
[17]
[18] J.Å.
[19]
(2011) 919.
[20]
[21] L.
[22] M.C.
[23]
(1998) 3906.
[24] J.Å.
[25] J.Å.
[26] L.
2318.
[27] J.
[28] R.
[29]
[30] J.-F.
[31] D.K.
matogr.
[32] L.
B.R.T.
Simoneit,
V.O.
Elias,
Mar.
Pollut.
Bull.
42
(2001)
805.
Simoneit,
Appl.
Geochem.
17 (2002)
129.
Hawthorne,
D.J.
Miller,
R.M.
Barkley,
M.S.
Krieger,
Environ.
Sci.
Technol.
22
1191.
Hawthorne,
M.S.
Krieger,
D.J.
Miller,
M.B.
Mathiason,
Environ.
Sci.
Technol.
(1989)
470.
Buranov,
G.
Mazza,
Ind.
Crop.
Prod.
28
(2008)
237.
Rowell,
R.
Pettersen,
J.S.
Han,
J.S.
Rowell,
M.A.
Tshabalala,
in:
R.M.
Rowell
Handbook
of
Wood
Chemistry
and
Wood
Composites,
CRC
Press,
New
2005.
M.E.
Himmel,
J.O.
Baker,
R.P.
Overend,
Enzymatic
Conversion
of
Biomass
for
Production,
American
Chemical
Society,
Washington,
DC,
1994.
S. Liu,
Biotechnol.
Adv.
28
(2010)
563.
Hawthorne,
D.J.
Miller,
J.J.
Langenfeld,
M.S.
Krieger,
Environ.
Sci.
Technol.
2251.
D. Hoffmann,
Y.
Iinuma,
H.
Herrmann,
J. Chromatogr.
A
1143
(2007)
168.
P.M.
Fine,
G.R.
Cass,
B.R.T.
Simoneit,
Environ.
Sci.
Technol.
35
(2001)
2665.
Fine,
G.R.
Cass,
B.R.T.
Simoneit,
Environ.
Eng.
Sci.
21
(2004)
705.
O. Faix,
D.
Meier,
I. Fortmann,
Eur.
J. Wood
Wood
Prod.
48
(1990)
281.
W.F.
Rogge,
L.M.
Hildemann,
M.A.
Mazurek,
G.R.
Cass,
B.R.T.
Simoneit,
Environ.
Technol.
32
(1998)
13.
Hoffer,
A.
Gelencer,
M.
Blazsó,
P.
Guyon,
P.
Artaxo,
M.O.
Andreae,
Atmos.
Phys.
6 (2006)
3505.
D.R.
Oros,
B.R.T.
Simoneit,
Aerosol
Sci.
Technol.
31
(1999)
433.
J.Å.
Jönsson,
L. Mathiasson,
J. Sep.
Sci.
24
(2001)
495.
Jönsson,
L. Mathiasson,
Trends
Anal.
Chem.
18
(1999)
325.
M.
Hyder,
L.L.
Aguilar,
J. Genberg,
M.
Sandahl,
C.
Wesén,
J.Å.
Jönsson,
Talanta
85
M.
Hyder,
J. Genberg,
J.Å.
Jönsson,
Anal.
Chim.
Acta
713
(2012)
79.
Zhao,
H.K.
Lee,
Anal.
Chem.
74 (2002)
2486.
Pietrogrande,
D.
Bacco,
M.
Mercuriali,
Anal.
Bioanal.
Chem.
396
(2010)
877.
L.
Chimuka,
N.
Megersa,
J. Norberg,
L.
Mathiasson,
J.Å.
Jönsson,
Anal.
Chem.
70
Jönsson,
L. Mathiasson,
Trends
Anal.
Chem.
11
(1992)
106.
Jönsson,
P.
Lövkvist,
G.
Audunsson,
G.
Nilvé,
Anal.
Chim.
Acta
277
(1993)
9.
Chimuka,
T.A.M.
Msagati,
E.
Cukrowska,
H.
Tutu,
J. Chromatogr.
A 1217
(2010)
Lezamiz,
J.Å.
Jönsson,
J. Chromatogr.
A 1152
(2007)
226.
Wolny,
A.
Koll,
L.
Sobczyk,
J.
Phys.
Chem.
89
(1985)
2053.
J.-F.
Peng,
J.-F.
Liu,
X.-L.
Hu,
G.-B.
Jiang,
J. Chromatogr.
A
1139
(2007)
165.
Liu,
L.
Toräng,
P.
Mayer,
J.Å.
Jönsson,
J. Chromatogr.
A
1160
(2007)
56.
Dubey,
D.
Pardasani,
A.K.
Gupta,
M.
Palit,
P.K.
Kanaujia,
V. Tak,
J. Chro-
A 1107
(2006)
29.
Zhu,
L.
Zhu,
H.K.
Lee,
J. Chromatogr.
A 924
(2001)
407.