Chemosphere 94 (2014) 199202

Contents lists available at ScienceDirect

Chemosphere
journal homepage: www.elsevier.com/locate/chemosphere

Short Communication

Comparison of solvent extraction and solid-phase extraction

for the determination of polychlorinated biphenyls in transformer oil
A.N. Mahindrakar a,, S. Chandra b, L.P. Shinde c
a
New Custom House Laboratory, Ballard Estate, Mumbai 400 001, Maharashtra, India
b
Custom House Laboratory, Vasco-Da-Gama 403 803, Goa, India
c
N.E.S. Science College, Nanded 431 605, Maharashtra, India

a r t i c l e i n f o a b s t r a c t

Article history: Solid-phase extraction (SPE) of nine polychlorinated biphenyls (PCBs) from transformer oil samples was
Received 9 March 2012 evaluated using octadecyl (CI8)-bonded porous silica. The efciency of SPE of these PCBs was compared
Received in revised form 18 August 2013 with those obtained by solvent extraction with DMSO and hexane. Average recoveries exceeding 95% for
Accepted 2 September 2013
these PCBs were obtained via the SPE method using small cartridges containing 100 mg of 40 pm CI8-
Available online 29 September 2013
bonded porous silica. The average recovery by solvent extraction with DMSO and hexane exceeded
83%. It was concluded that the recoveries and precision for the solvent extraction of PCBs were poorer
Keywords:
than those for the SPE.
PCBs
SPE
2013 Elsevier Ltd. All rights reserved.
Solvent extraction
Transformer oil
GCMS

1. Introduction
One important area of modern environmental analysis is the
determination of polychlorinated biphenyls (PCBs) and polycyclic
aromatic hydrocarbons (PAHs), both of which commonly occur in
the environment. They are highly toxic, mutagenic and persistent
in the environment, and therefore are on the US Environmental
Protection Agency (EPA) list of priority pollutants (Mahindrakar
et al., 2012). Poly chlorinated biphenyls (PCBs) are chemical com-
pounds which have been used as dielectric uids in transformers
and capacitors, as plasticizers and in hydraulic uids (Berg et al.,
1998; Bogdevich and Cadocinicov, 2004). PCBs are toxic, and they
accumulate in soil, sediment and biota when they are released into
the environment. Although most countries have strict regulations
on the use of PCBs (UNEP, 2002), enormous amounts of PCBs did
go into use before they were regulated. Indeed, they are still being
used as insulators because devices such as transformers have life-
times extending several decades. Many countries that have devel-
oped a classication system for PCB-containing uids and
materials have considered 50 mg kg1 as the benchmark level for
PCB regulation (UNEP, 1999). In accordance with PCB management,
analytical test methods for determination of PCBs in insulating oil
have been published as ASTM-D4059, EPA-600/4-81-045 and NIST
Corresponding author. Tel.: +91 7698888151; fax: +91 0265 2396080.
E-mail address: aaanil01@gmail.com (A.N. Mahindrakar).
(USEPA, 1982; ASTM D4059, 2000). In these methods, the oil,
diluted with the appropriate solvent, is treated with acid and/or
an adsorbent to remove interferences, and then analyzed by
GC-ECD. However, the method detection limits (MDLs) of PCBs
by these test methods are only 12 mg kg1 because the oil is di-
luted 50100 times to minimize the inuence of the oil during
the analytical procedure (Yun-Cheol et al., 2008). PCBs are not
easily separated from oil because the physical and chemical char-
acteristics of PCBs are very similar to those of mineral oil. The com-
ponents of oil-based liquid wastes co-elute with the PCBs during a
GC separation and the baseline shifts due to the oil matrix. The key
to lowering the MDL is to effectively remove the remaining oil dur-
ing the analytical procedure. To remove the oil, a few analytical
methods have been introduced to apply liquidliquid partition
using an appropriate solvent. Although these methods are useful
in identifying and determining the PCBs in oil, the separation of
PCBs by liquidliquid extraction (LLE) is dependent on the skill of
the analysts. To separate PCBs from oil matrix, researchers have
used partition between non-polar solvent and non-proton polar
solvent, such as dimethylsulfoxide (Orazio et al., 1989; Larsen
et al., 1991; Mahindrakar et al., 2011a,b,c), acetonitrile (Gordon
et al., 1987) and dimethylformamide (Lawn and Toffel, 1987). Solid
phase extraction (SPE) system is also reported for concentration of
sample (Kot-Wasik et al., 2004; Wolska et al., 2005; Conka et al.,
2005). This method provides a powerful technique to overcome
the drawback of LLE. A comparative study of the SPE technique

0045-6535/$ - see front matter 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.chemosphere.2013.09.006
200 A.N. Mahindrakar et al. / Chemosphere 94 (2014) 199202

2.3. Samples

The samples of transformer oil were stored in 250 mL

pre-cleaned dark glass bottles with Teon caps at a maximum tem-
perature of 4 C. Moisture determinations were made according to
ASTM D 3976 (Mahindrakar et al., 2011a,b,c). The concentration of
PCBs was calculated on dry weight basis.

2.4. Analytical procedure for Solvent extraction

The analytical procedure was used from the previous work

(Mahindrakar et al., 2011a,b,c), the sample was transferred into a
glass centrifuge tube and 1 mL of 2.5 lg of surrogate standard
pyrene-d10 was added. The rst extraction was carried out by
10 mL DMSO and water (95:5) then repeated the same with 8 mL
and nally 5 mL after keeping the tubes into centrifuge machine
(3500 rpm) for 10 min. Then 10 mL water was added in extraction
portion. This was further extracted with 10 mL hexane, repeated
the same with 8 mL and nally with 5 mL after keeping in centri-
fuge machine for 10 min. The extracted portion was reduced in
rotary evaporator and made up in 5 mL after adding 1.5 lg of inter-
nal standards and nally injected 1 lL into GCMS. The detailed
analytical procedure is shown in Fig. 1.

2.5. Sample cleanup

A variety of substances in oil can be simultaneously extracted

Fig. 1. Analytical procedure for determination of PCBs in transformer oil.
and cause interference in the determination of PCBs by GCMS.
Therefore the extraction was puried using silica gel column. The
clean-up procedure was as follows: The concentrated extracts
were applied to open glass column (0.45 m, 1 cm i.d.) lled with
(from bottom to top) glass wool, silica gel (100200 mesh) and
anhydrous sodium sulfate (2 cm). The extracts were eluted with
the mixture of hexane and acetone (1:1) from silica column, the
rst 10 mL of elute were discarded and remaining 25 mL of elute
was collected and reduced the volume in rotary evaporator and

using commercially available cartridges and the conventional sol-

vent extraction method was carried out for nine PCBs in trans-
former oils.

2. Materials and methods

2.1. Reagents

The solvent hexane was HPLC grade, DMSO and sulfuric acid
(98%) was A. R. grade used in the present work. Silica gel
(100200 mesh, SigmaAldrich) activated at 200 C for 6 h and
anhydrous sodium sulfate (A.R. grade) was used in the clean-up
stage. Water was puried with a Milli-Q plus system (Millipore).
Supelclean LC-18 SPE cartridges from Supelco, pre-packed in a
polypropylene housing cartridge containing 100 mg of C18-bonded
porous silica and with a 0.45 pm lter attached, were used.

2.2. Standard solutions

The mixture of PCBs was purchased from AccuStandard

(M-680A). Phenanthrene-d10 and chrysene-d12 from AccuStan-
dard (M-680-IS-10X) and pyrene-d10 from Sigma Aldrich were
used as an internal standards and surrogate standard respectively.
All the solutions were prepared in hexane. Solutions of 0.0156
12.50 mg kg1 were prepared from the stock solution for calibra-
tion of different congeners and 0.3 mg kg1 of internal standard
was added to each. Fig. 2. Analytical procedure for determination of PCBs in transformer oils.
 A.N. Mahindrakar et al. / Chemosphere 94 (2014) 199202 201

Table 1
Retention time of PCBs, internal standards.

Structure IUPAC No. Quantitation ion Conrmation ion RT (min) Internal standard used
2 1 188 152 10.28 Phenanthrene-d10
2,3 5 222 152 12.70 Phenanthrene-d10
2,4,5 29 256 186 15.16 Phenanthrene-d10
2,20 ,4,6 50 292 220 15.62 Phenanthrene-d10
2,20 ,3,4,50 87 326 254 23.44 Phenanthrene-d10
2,20 ,4,40 ,5,60 154 360 288 23.96 Phenanthrene-d10
2,20 ,3,40 ,5,6,60 188 394 322 26.55 Chrysene-d12
2,20 ,3,30 ,4,50 ,6,60 201 430 356 32.98 Chrysene-d12
2,20 ,3,30 4,40 ,5,50 ,6,60 209 498 428 45.60 Chrysene-d12
Phenanthrene-d10 188 187 14.16
Chrysene-d12 240 236 32.58

made up in 5 mL after adding 1.5 lg internal standards and nally
injected 1 lL into GCMS.
2.6. Analytical procedure for solid phase extraction
The sample after giving sulfuric acid treatment was loaded into
the pre-conditioned cartridge and then eluted with hexane. The
eluent was concentrated to about 1 mL on a rotary evaporator.
Then, 0.3 lg of internal standards was added to the sample, and
analysis by GCMS was performed. The detailed analytical proce-
dure is shown in Fig. 2.
2.7. Instrumental analysis of PCBs
The analysis of PCBs was carried out by an integrated system of
gas chromatography, equipped with automatic injection system
and coupled to a mass spectrometric system with quadruple ion
trap analyzer. Varian CP-3800 GC, Saturn-2200 mass spectrometer
with auto injector CP-8410 was used for analysis. The separation of
congeners was done in a 30 m length, 0.25 mm internal diameter
and 0.25 lm lm thickness coated with 5% phenyl 95% meth-
ylpolysiloxane Varian VF-5MS column. Helium was used as the
carrier gas at 9.6 psi pressure and 1 mL min1 ow. The injector
was used at constant temperature and 280 C. The initial oven tem-
perature was 80 C (4 min isothermal) to 180 C (at 20 C min1) to
Fig. 3. (A) TIC of PCBs free transformer oil with internal standards. (B) TIC of PCBs
(M-680A from AccuStandard, 2.5012.50 mg kg1 hexane) spiked in transformer
oil, Peaks: (1) PCB 1, (2) PCB 5, (3) phenanthrene-d10, (4) PCB 29, (5) PCB 50, (6) PCB
87, (7) PCB 154, (8) PCB 188 (9) chrysene-d12, (10) PCB 201 and (11) PCB 209.
250 C (at 2 C min1) to 280 C (at 10 C min1) isothermal for
4 min. The injection volume was 1 lL in split less mode. The tem-
perature of ion trap, manifold and transference line was 220 C,
50 C and 280 C respectively. The mass spectrometer was used
in SIM mode under electron impact at 70 eV and scan time 1 s.
The retention time, quantiation ions, conrmation ions, different
internal standards used for PCBs are shown in Table 1.
2.8. Efciency of extraction methods
A 2 g PCBs free transformer oil was spiked with 0.0156 mg kg1-
of each of the nine PCBs standards and subjected to the entire ana-
lytical procedure as described above for solvent extraction. This
was repeated nine times and a set of data for the efciency of
the extraction method was obtained. Similarly, the extraction ef-
ciency of the SPE method was determined by spiking PCBs free
transformer oil, followed by the experimental procedure as de-
scribed above. This was repeated nine times and a set of data for
the efciency of the SPE method was obtained.
3. Result and discussion
Before PCBs analysis, the linear range of MS detector was deter-
mined (Wells et al., 1992). The detector response was established
for each PCB containing different number of chlorine atoms. The
precision, calculated as a RSD was less than 10%. Ten replicate
injections of standard solutions (at least two concentration levels)
were injected into GCMS to check system performance and cali-
bration validity. The RSD was generally less than 5%. Limit of quan-
titation (LOQ) of each PCB was on the level 0.4 lg kg1.
Identication of PCBs in the samples was performed by comparison
of chromatograms of puried extracts with those obtained from
standard solutions. This method is based on simultaneous detec-
tion, at corresponding retention time, of chromatographic signals
of the two selected ions for each PCB (Table 1), which are chosen
according to full scan mass spectrum and chromatogram of target
compounds. Peaks were accepted as PCB if they fullled the follow-
ing criteria (Juan et al., 1999): The retention time was within
0.1 min of the retention time of the relevant PCB in the standard.
The ratio between two monitored ions was within satisfactory le-
vel (0.2) of the ratio found for the relevant PCB in the standard.
The TIC of transformer oil without PCBs along with internal stan-
dards and spiked known quantity of PCBs in transformer oil is
shown in Fig. 3. All nine PCBs can be resolved and eluted in a rea-
sonable time by employing the GC conditions stipulated above. The
efciencies of extraction of the each PCB by solvent extraction with
hexane and by the SPE technique are shown in Table 2. Average
recoveries exceeding 95% were obtained by SPE. The standard devi-
ations for the recovery of these PCBs using SPE ranged from 1.5% to
4.5%. These values are within acceptable limits for reproducible re-
sults in quantitative analysis. The recovery data using the solvent
202 A.N. Mahindrakar et al. / Chemosphere 94 (2014) 199202

Table 2 enormous time saving and less analysis cost in comparison with
Recovery of spiked PCBs from transformer oil. solvent extraction.
PCBs % Recovery RSD % Recovery solvent RSD
SPE (%) extraction (%) References
1 101.0 1.5 88.2 10.0
5 100.9 3.9 89.5 8.5 ASTM International, 2000. Analysis of PCBs in insulating liquids by gas,
29 105.0 3.5 90.1 6.0 chromatography. ASTM D4059, ASTM International, USA.
Berg, V., Birnbau, M., Bosabel, A., 1998. Toxic equivalent factors (TEFs) for PCBs,
50 98.2 3.0 87.8 4.8
PCDDs, PCDFs for humans and wildlife. Environ. Health. Persp. 106, 775792.
87 100.0 4.0 89.5 7.2
Bogdevich, O.P., Cadocinicov, O.P., 2004. PCBs determination in transformer oil by
154 97.5 3.4 86.9 6.9
SPE and gas chromatography analysis. Mold. J. Phys. Sci. 3, 226235.
188 96.9 3.2 85.5 8.0 Conka, K., Drobna, B., Kocan, A., Petrik, 2005. Simple solid-phase extraction method
201 95.0 3.8 84.2 9.0 for determination of polychlorinated biphenyls and selected organochlorine
209 95.5 2.0 83.2 4.0 pesticides in human serum. J. Chromatogr. A 1084, 3338.
Gordon, R.J., Szita, J., Faeder, E.J., 1987. Determination of polychlorinated biphenyls
in transformer oils by capillary gas chromatography. Anal. Chem. 54, 478481.
Juan, C.Y., Thomas, G.O., Semple, K.T., Jones, K.C., 1999. Methods for the analysis of
PCBs in human food faeces, and serum. Chemosphere 39, 14671476.
Kot-Wasik, A., Debska, J., Namiesnik, J., 2004. Monitoring of organic pollutants in
120 coastal waters of the Gulf of Gdansk, Southern Balti. Mar. Pollut. Bull. 49, 264
100 276.
Recovery (%)

Lawn, R.E., Toffel, S.A., 1987. Determination of polychlorinated biphenyls in waste

80
oil by gasliquid chromatography. Analyst 112, 5356.
60 Larsen, B., Tilio, R., Kapita, S., 1991. A DMSO-based cleanup procedure for
determination of PCBs in waste oil. Chemosphere 23, 10771084.
40
Mahindrakar, A.N., Chandra, S., Shinde, L.P., 2012. Gas chromatography-mass
20 spectrometry-ion trap determination of polycyclic aromatic hydrocarbons in
waste oils. Int. J. Pure. Appl. Chem. 7, 187192.
0
Mahindrakar, A.N., Chandra, S., Shinde, L.P., 2011a. Simultaneous determination of
PCB 1

PCB 5

PCB 29

PCB 50

PCB 87

PCB 154

PCB 188

PCB 201

PCB 209

polychlorinated biphenyls and polycyclic aromatic hydrocarbons in waste oils

PCBs
Fig. 4. Recoveries of PCB 1, 5, 29, 50, 87,154, 188, 201 and 209 after SPE and solvent
extraction clean-up (Black column-SPE, transparent column-solvent extraction).
extraction technique were lower than those obtained by SPE (Fig
4). Average recoveries ranging from 8795% were obtained for
the nine PCBs. The standard deviations for the recovery of these
PCBs using the solvent extraction technique, ranging from 4% to
10%, are also greater than those obtained with the SPE method.
4. Conclusion
The utilization of octadecyl (CI8)-bonded porous silica phase
exhibit a good result of the SPE cleanup and PCBs determination
by GCMS in the transformer oil than the solvent extraction with
DMSO and hexane. The SPE method has principal advantages are
better recovery and precision, less consumption of solvents,
by gas chromatography-mass spectrometry-ion trap. Anal. Chem.: Ind. J. 10,
481486.
Mahindrakar, A.N., Chandra, S., Shinde, L.P., 2011b. Gas chromatography mass
spectrometry determination of polychlorinated biphenyls in waste oils. Anal.
Chem.: Ind. J. 10, 754759.
Mahindrakar, A.N., Chandra, Nalwad, S.S., Shinde, L.P., 2011c. Capillary gas
chromatography-mass spectrometry determination of polycyclic aromatic
hydrocarbons. Alfa Univ. Int. J. Chem. 2, 104117.
Orazio, C., Meadows, J., Kapila, S., Palmer, C., Yanders, A.F., 1989. Simple clean-up
procedure for determination of polychlorinated dibenzo-p-dioxin and
polychlorinated dibenzofurans in waste oil, soil and biological tissue samples.
Chemosphere 18, 6976.
UNEP, Chemicals, 2002, PCB transformers and capacitors: from management to
reclassication and disposal. United Nation Environmental Programme.
UNEP, 1999. Guidelines for the Identication of PCBs and Materials Containing
PCBs. United Nation Environmental Programme, Geneva, Switzerland.
USEPA, 1982. The determination of polychlorinated biphenyls in transformer uid
and waste oils. EPA-600/4-81-045, USEPA, USA 1982.
Wolska, L., Adkonis, M.R., Namiesnik, J., 2005. Determining PAHs and PCBs in
aqueous samples: nding and evaluating sources of error. Anal. Bioanal. Chem.
382, 13891397.
Wells, D.E., Maier, B., Griepink, B., 1992. Calibrants and calibration for
chlorobiphenyl analysis. Int. J. Anal. Chem. 46, 255264.
Yun-Cheol, N., Kang-Jin, K., Jongki, H., Jung-Ju, S., 2008. Determination of
polychlorinated biphenyls in transformer oil using various adsorbents for
solid phase extraction. Chemosphere 73, S7S12.