Chemosphere xxx (2014) xxx–xxx

Contents lists available at ScienceDirect

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

Critical points in the evaluation of analytical methods based on liquid

chromatography separation for the determination of doramectin
in different environmental samples
Marta Wagil, Anna Białk-Bielińska ⇑, Joanna Maszkowska, Piotr Stepnowski, Jolanta Kumirska ⇑
Department of Environmental Analysis, Faculty of Chemistry, University of Gdańsk, ul. Wita Stwosza 63, 80-308 Gdańsk, Poland

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

Article history: In recent years, the number of analytical methods of target compound residues (such as pharmaceuticals)
Received 16 January 2014 has grown rapidly. Most of them are based on high performance liquid chromatography (HPLC). From the
Received in revised form 24 March 2014 economic point of view, it is usual to apply the conditions of available HPLC methods or to design extrac-
Accepted 27 March 2014
tion and chromatographic separation conditions using HPLC and transfer them subsequently to a more
Available online xxxx
sensitive technique like liquid chromatography coupled with tandem mass spectrometry (LC–MS/MS).
Handling Editor: Klaus Kümmerer However, if such a transfer is planned, it is important to assess the quality of the newly-designed
LC–MS/MS method. The determination of parameters like matrix effects (ME), extraction efficiency
Keywords: (EE) and absolute recovery (AR) is mandatory. These parameters can visualise the weakest step in the
LC–MS/MS analytical method and enable methods based on different techniques to be compared. The aim of this
HPLC–UV work was to show how quality assessment should be carried out in order to transfer an optimised
Method transferring method from one technique to another. The representative compound used in our investigation was
Quality assessment doramectin (DOR), an anthelmintic drug used in veterinary medicine. The quality of the suggested
methods for determining this drug in three environmental matrices (water, sediment and fish tissue)
using HPLC–UV and LC–MS/MS was evaluated on the basis of known values of absolute recovery
(HPLC–UV) and matrix effect, extraction efficiency and absolute recovery (all LC–MS/MS). Finally, the
suggested methods for determining DOR in water, sediment and fish tissue based on LC–MS/MS
measurements were validated and applied to the analysis of real environmental samples.
Ó 2014 Elsevier Ltd. All rights reserved.

1. Introduction EURACHEM, a network of organisations focused on analytical

Nowadays, high performance liquid chromatography (HPLC)
techniques are commonly used for determining drugs in the envi-
ronment: liquid chromatography coupled with mass spectrometry
(LC–MS or LC–MS/MS) is the most suitable for this purpose (Silvia
Díaz-Cruz and Barceló, 2007; Fatta et al., 2007; Hernández et al.,
2007). From the economic point of view, it is standard practice to
initially optimise the analytical procedure using HPLC–UV and sub-
sequently to transfer it from HPLC–UV to LC–MS. However, many
problems can appear during such a transfer. For example, LC–MS,
unlike HPLC–UV, is very susceptible to the impact of matrix com-
ponents on the final determination of analytes (Caban et al., 2012).
⇑ Corresponding authors. Tel.: +48 58 5235208, +48 58 5235212; fax: +48 58
5235454.
E-mail addresses: a.bialk-bielinska@ug.edu.pl (A. Białk-Bielińska), jolanta.
kumirska@ug.edu.pl (J. Kumirska).
chemistry in Europe, has laid down guidelines on measurement
uncertainty in analytical measurements and holds workshops
addressing this problem (Ellison et al., 2012). The uncertainty of
results is a major problem in this kind of analysis. There are many
sources of such uncertainty, the most important of which are the
sampling procedure and the influence of matrix components on
the final analysis. Therefore, before transferring an analytical
method from one technique to another, it is crucial to make a full
quality assessment by determining the parameters affecting the
uncertainty of the analytical results. For this reason, the validation
procedure of any new method based on HPLC–UV measurements
should include the assessment of absolute recovery (AR), and any
LC–MS-based method should additionally include assessments of
matrix effects (ME) and extraction efficiency (EE).
Matrix effects occur when matrix constituents interact with the
target analytes in the ion source of the LC–MS/MS system and
affect the ionization of the target analyte (Hall et al., 2012). Such
effects can either suppress or enhance the ionisation of target

http://dx.doi.org/10.1016/j.chemosphere.2014.03.137
0045-6535/Ó 2014 Elsevier Ltd. All rights reserved.

Please cite this article in press as: Wagil, M., et al. Critical points in the evaluation of analytical methods based on liquid chromatography separation for the
determination of doramectin in different environmental samples. Chemosphere (2014), http://dx.doi.org/10.1016/j.chemosphere.2014.03.137
2 M. Wagil et al. / Chemosphere xxx (2014) xxx–xxx
analytes (Antignac et al., 2005; Patel, 2011), leading to the incor-
rect determination of analyte concentrations. Extraction efficiency
is a measure of the effectiveness of analyte isolation and enrich-
ment during extraction, whereas absolute recovery indicates the
influence of the sample components on the entire determination
method (Caban et al., 2012). It must be highlighted that HPLC–
UV technique is not susceptible to matrix effects as the detection
system differs from LC–MS technique. No ionization takes place
in HPLC–UV hence, while characterizing HPLC–UV method only
absolute recovery is necessary to be determined in contrast to
LC–MS technique – where additionally to absolute recovery (AR),
also extraction efficiency (EE) as well as matrix effects (ME) should
be determined. However, as AR value determined for LC–MS
method includes ME value, an AR obtained by HPLC–UV can be
only compared to EE obtained for LC–MS technique. For all the
above reasons, it is very important to calculate these three param-
eters for LC–MS measurements, yet many researchers do not
always do so.
Furthermore, the mobile phase composition must be appropri-
ate when one plans to transfer a HPLC method to LC–MS/MS.
Some additives can disrupt the ionisation of analytes in the ion
source (Mannuer et al., 2011). Components suitable for LC–MS
are volatile ones like ammonium acetate, formate or hydroxide,
formic or acetic acid, but not non-volatile ones, such as phosphates,
sulphates, borates and citrates, as they contaminate the electrodes
in the ion source (by precipitation). The flow rate of the mobile
phase also needs to be considered. Although theoretically,
1 mL min1 or 2 mL min1 are the respective maximum permitted
flow rates in ESI and APCI (ThermoQuest, 1999), the flow rate
should be optimised and not normally exceed 0.5 mL min1, as this
provides better ionisation conditions for the analytes.
The main aim of this work was to show how a quality assess-
ment should be performed in order to transfer an optimised
method from one technique to another. We illustrate this by taking
the determination of doramectin (DOR) in three different environ-
mental matrices as an example.
We chose doramectin (Appendix A), a macrocyclic lactone
(Danaher et al., 2006), because it is commonly administered drug
to treat parasitic infections in veterinary practice in mammalian
food-producing species. DOR is reported to be excreted almost
exclusively (almost 98%) as the parent compound (Kolar and
Kožuh Eržen, 2006). The solubility of DOR in water at 25 °C is
25 lg L1 and the pKa is 12.4 (Pfizer, 1996; Kolar and Kožuh
Eržen, 2006; Krogh et al., 2008). Owing to its soil organic car-
bon–water partitioning coefficient KOC (>1000) and octanol–water
partition coefficient (log Kow 4.44), this drug can sorbs very
strongly to solid components such as soil and sediment or be bio-
accumulated in tissue samples (e.g. in fish tissue). It can therefore
be toxic to non-target organisms when it occurs in the environ-
ment. In our previous research we established the EC50s of DOC
towards Daphnia magna (EC50 = 6.37  105 mg L1) (Kołodziejska
et al., 2013). Clearly, DOR adversely affects other organisms, even
at low concentrations. Hence, in order to protect the environment
from the negative influences of this anthelmintic drug, reliable
analytical methods must be available for monitoring its occurrence
in the environment, not only in water and sediment but also in
biota. Very little is known concerning the development of analyti-
cal methods for determining DOR in water, sediment and fish tis-
sue samples (Krogh et al., 2008). The usual technique for
extracting DOR from water samples is solid phase extraction
(Krogh et al., 2008), whereas from soil/sediment samples pressur-
ised liquid extraction coupled with solid phase extraction (Krogh
et al., 2008), accelerated solvent extraction (Gravell, 2005) or
microwave assisted extraction (Raich-Montiu et al., 2011) are used.
Furthermore, there is a lack of information concerning quality
Please cite this article in press as: Wagil, M., et al. Critical points in the evaluation of analytical methods based on liquid chromatography separation for the
determination of doramectin in different environmental samples. Chemosphere (2014), http://dx.doi.org/10.1016/j.chemosphere.2014.03.137
assessment by evaluating extraction efficiency and absolute
recovery. In this study, therefore, we present methods for deter-
mining DOR in three different matrices (water, sediment and fish
tissue) which can be employed in future environmental risk assess-
ments of this compound. The applicability of the suggested meth-
ods was validated by analysing environmental samples collected
from the River Gościcina in northern Poland.
2. Materials and methods
2.1. Chemicals
Acetonitrile (ACN) (LC–MS ChromasolvÒ) was purchased
from Sigma–Aldrich (Steinheim, Germany). Methanol (MeOH)
(HPLC-grade) was obtained from POCH S.A. (Gliwice, Poland) and
n-hexane (HPLC-grade) from J.T. Baker (Deventer, Holland).
Deionised water was produced by the HYDROLAB System (Gdańsk,
Poland). Triethylamine (TEA), ammonium acetate, sodium chloride
(anhyd.), formic acid (all analytical reagent grade) were purchased
from Chempur (Piekary Śla˛skie, Poland).
2.2. Pharmaceutical standards
Doramectin was purchased from Sigma Aldrich (Steinheim,
Germany). The standard stock solution was prepared by dissolving
25 mg of DOR in 50 mL of ACN (500 lg mL1) and stored at 18 °C.
A working solution of DOR was prepared by diluting an appropriate
amount of the stock solution in ACN. All working standards were
stored in the dark at 4 °C.
2.3. Sample preparation
Three working solutions of DOR in ACN were prepared (0.005,
0.025 and 0.05 lg mL1). Water samples were spiked with these
working solutions to the following nominal concentration of
DOR: 0.01, 0.05 and 0.1 lg L1. Each sample was prepared in
triplicate. In addition, three non-spiked (blank) samples were also
prepared.
The sediment samples were air-dried, ground in a mortar and
passed through a 2 mm sieve. Three different concentrations of
working solutions in ACN were prepared: 0.01, 0.05 and
0.1 lg mL1. 1 g of sediment was spiked with these working solu-
tions; mixed on a vortex (1 min; Vortex IKA (Vortex Genius 3),
Poland) and shaken using shaker (15 min; KS 130 basic, Bionovo,
Poland). The nominal concentrations of the respective sediments
were as follows: 0.01, 0.05 and 0.1 lg g1. Unspiked (blank)
samples were also prepared. The samples were left to dry in the
dark at room temperature for three days.
Rainbow trout were obtained from fish farms on the River
Gościcina. Finely sliced muscle tissue (1 g) was weighed into a
polypropylene tube (30 mL). The fish tissue was spiked in the same
way as the water and sediment samples using working solutions in
ACN at concentrations of 0.04, 0.2 and 0.4 lg mL1; the nominal
concentrations were 0.01, 0.05 and 0.1 lg g1. Each sample was
prepared in triplicate. Three unspiked (blank) samples were also
prepared. Samples were left in the dark for 30 min.
2.4. HPLC–UV and LC–MS/MS analysis
HPLC–UV measurements were performed on a Perkin Elmer
Series 200 with UV detector (Perkin Elmer, USA). The analyte was
separated on a Phenomenex C8 column (100  4.6 mm, 3 lm).
The mobile phase was a mixture of ACN and H2O (90:10, v/v;
isocratic; flow rate 0.5 mL min1). The injection volume was
 M. Wagil et al. / Chemosphere xxx (2014) xxx–xxx
50 lL and the analytical wavelength was 244 nm. The time of
analysis (the complete run) was 6 min.
The LC–MS/MS system consisted of an Agilent 1200 Series LC
system (Agilent Technologies Inc., Santa Clara, USA) and an HCT
Ultra ion trap mass spectrometer (Bruker Daltonics, Bremen, Ger-
many) equipped with an electrospray ionisation source. ChemSta-
tion software was used for data acquisition. The chromatographic
separations were performed under the same conditions as the
HPLC–UV measurements. MS/MS parameters such as nebuliser
gas pressure (30–50 psi), dry gas flow rate (8–10 L min1), dry
temperature (300–365 °C) and positive and negative ion mode,
were investigated in order to optimise the LC–MS/MS method.
The usefulness of the single ion monitoring mode (SIM) and the
multiple reaction monitoring (MRM) mode for the qualitative
and quantitative analysis of DOR in environmental samples analy-
sis by LC–MS/MS was also tested.
2.5. Optimisation of extraction procedures based on HPLC–UV
measurements
The extraction procedures for isolating and enriching DOR from
water,sediment and tissue samples are given in Table 1. The choice
of these methods was based on literature data provided by Raich-
Montiu et al. (2011), Rudik et al. (2002) and Hou et al. (2007), who
applied them to extract DOR from agriculture soils (Rudik et al.,
2002; Raich-Montiu et al., 2011) and from bovine liver and muscle
(Hou et al., 2007). The optimisation concerned mainly the choice of
sorbent, washing solvent and elution solvent. After extraction, the
eluates were concentrated to a volume of 1 mL on a rotary vapor-
iser, transferred to a 1.5 mL vial, evaporated to dryness under a
stream of nitrogen and reconstituted in 1 mL of mobile phase A:B
(90:10, v/v) for HPLC–UV or LC–MS/MS analysis, depending on
the purpose.
In order to verify the repeatability of the extraction of DOR from
water, sediment and tissue samples for each matrix, both the
spiked samples and the unspiked samples were prepared in tripli-
cate and subjected to the extraction procedures and HPLC–UV
measurements.
2.6. Quality assessment – calculation of matrix effect (ME), extraction
efficiency (EE) and absolute recovery (AR) for LC–MS/MS
measurements and absolute recovery for HPLC–UV analyses
In order to determine ME, EE and AR for LC–MS/MS analysis, we
applied the methodology proposed by Caban et al. (2012). These
parameters were calculated using the following equations:
  
BD
ME ¼  1  100%
A UV measurements were transferred directly to the LC–MS/MS sys-
 
CD
EE ¼  1  100%
BD
 
CD
AR ¼  1  100%
A to be optimal during the LC–MS/MS determination.
where A is the peak area of the analyte recorded for the standard
solution, B is the peak area of the analyte recorded for the sample
spiked with the target compound after extraction procedure, C is
the peak area of the analyte recorded for the sample spiked with
the target compound before extraction procedure, and D is the peak
area of the analyte recorded for the extracted unspiked sample.
The absolute recovery calculated for HPLC–UV analysis was cal-
culated in the same way as for LC–MS/MS.
Please cite this article in press as: Wagil, M., et al. Critical points in the evaluation of analytical methods based on liquid chromatography separation for the
determination of doramectin in different environmental samples. Chemosphere (2014), http://dx.doi.org/10.1016/j.chemosphere.2014.03.137
3
2.7. Validation of methods for determining DOR in water, sediment
and tissue samples based on LC–MS/MS measurements
The optimised methods were validated by analysing seven dif-
ferent concentrations of DOR in each matrix (0.5–200.0 ng L1 for
water, 2.5–100.0 ng g1 for sediment and 2.5–100.0 ng g1 for tis-
sue samples). The calibration curves for determining DOR in each
matrix were plotted from a minimum of six calibration points
obtained from six injections of each DOR extract. The linearity, cor-
relation coefficient (R2), intra-day precision (expressed by RSD),
absolute recovery (AR), accuracy, limit of detection and limit of
quantification were calculated in accordance with the strategies
described in our previous paper (Migowska et al., 2012).
2.8. Environmental samples
In order to demonstrate the applicability of the proposed meth-
ods, we collected environmental samples from the River Gościcina
in northern Poland. These were analysed for the presence of DOR
using the validated methods based on LC–MS/MS measurements.
Eight sampling points were chosen along the river for collecting
water and sediment samples (G1–G8). Rainbow trout were
obtained from four fish farms located on the river (G1, G2, G5,
G7). Samples were collected during two seasons – autumn 2012
and spring 2013.
Sample collection and preparation – depending on the matrices
– were as follows: samples of river water were collected within
2–5 m of the bank and at 50 cm depth into amber glass bottles
pre-rinsed in ultra-pure water. The samples were then frozen
and stored at 18 °C. Prior to analysis the water samples were
thawed, passed through a glass filter (pore diameter £ 47 mm)
and stored at 4 °C for no longer than 12 h. The extraction was then
carried out according to Method 3A (Table 1). Sediment samples
were collected from the upper layer (7–10 cm depth) into amber
glass bottles pre-rinsed in ultra-pure water. The sediment was left
to air-dry at room temperature, then ground in a mortar and
passed through a 2 mm sieve. The extraction was carried out
according to Method 4 (Table 1). The fish were gutted, after which
the muscle tissue was homogenised and extracted according to
Method 6 (Table 1).
3. Results and discussion
3.1. Transferring the HPLC–UV method to the LC–MS system
The chromatographic separation conditions used during HPLC–
ð1Þ
tem without any modification (the composition and flow rate of
the mobile phase were both suitable for the two techniques). The
ð2Þ following conditions – a Phenomenex C8 column (100  4.6 mm,
3 lm), a mixture of ACN and H2O (90:10, v/v) as mobile phase,
an isocratic elution programme, a mobile phase flow rate of
0.5 mL min1 and an injection volume of 50 lL – were also found
ð3Þ
The optimal conditions of the mass spectrometric analysis using
LC–MS/MS were the following: source temperature 350 °C, nitro-
gen as the nebuliser gas (30 psi), a dry gas flow rate of 10 L min1,
and helium (99.999%) as the collision gas in the ion trap. The SIM
(selected ion monitoring) mode for the qualitative and quantitative
analysis of DOR in environmental samples was selected (m/
z = 921). The MRM (multiple reaction monitoring) mode could
not be used for this purpose because the product ion intensities
were too low.
 4
determination of doramectin in different environmental samples. Chemosphere (2014), http://dx.doi.org/10.1016/j.chemosphere.2014.03.137
Please cite this article in press as: Wagil, M., et al. Critical points in the evaluation of analytical methods based on liquid chromatography separation for the

Table 1
Extraction procedures for isolating DOR from water, sediment and tissue samples.

Water
Method 1 Method 2 Method 3 Method 3A
Sorbent type Strata X-CW (200 mg/3 mL) Strata C18 (200 mg/3 mL) Strata X (200 mg/3 mL) Strata X (200 mg/3 mL)
Conditioning 3 mL ACN, 3 mL 100 mM ammonium 3 mL MeOH, 3 mL H2O 6 mL ACN, 6 mL H2O 6 mL ACN, 6 mL H2O

M. Wagil et al. / Chemosphere xxx (2014) xxx–xxx

acetate (pH = 10.5a)
Washing 1 3 mL MeOH 3 mL MeOH:H2O (5:95, v/v) 8 mL MeOH:H2O (35:65, v/ 8 mL MeOH:H2O (35:65, v/v)
v)
Drying 15 min 15 min 15 min 15 min
Washing 2 – – – 6 mL n-hexane
Elution 3 mL HCOOH:ACN (5:95, v/v) 3 mL MeOH:ACN (50:50, v/v) 6 mL ACN 6 mL ACN
Sediment Tissue
Method 4 Method 5 (Raich-Montiu et al., 2011) Method 6 (Hou et al., 2007) Method 7 (Rudik et al., 2002)
Sample pretreatment (1) Mixing: 5 mL ACN:H2O (90:10, v/v), (2) (1) Mixing: 5 mL ACN:H2O (90:10, v/v), (2) (1) Mixing: 8 mL ACN  2, 1) Mixing: 6 mL ACN + centrifugation, (2) mixing: 4 mL ACN + centrifugation, (3)
MAE (15 min, 80 °C, 400 W), (3) +5 mL H2O MAE (15 min, 80 °C, 400 W), (3) +5 mL H2O 2) + 25 mL H2O, (3) +40 ll combination of supernatant + 0.2 g NaCl + centrifugation, (4) evaporation to
TEA dryness
Sorbent type Strata X (200 mg/3 mL) Oasis HLB (200 mg/6 mL) Strata C18 (200 mg/3 mL) –
Conditioning 5 mL MeOH, 5 mL H2O 5 mL MeOH, 5 mL H2O 2.5 mL ACN, 2.5 mL –
ACN:H2O (40:60, v/v) + 0.1%
TEA
Washing 1 10 mL H2O 10 mL H2O 2.5 mL ACN:H2O (40:60, v/ –
v) + 0.1% TEA
Drying – – 15 min –
Washing 2 – – 1.5 mL n-hexane –
Elution 3 mL ACN 3 mL ACN 2 mL ACN –
a
25% NaOH.
 M. Wagil et al. / Chemosphere xxx (2014) xxx–xxx 5

Fig. 1. AR values obtained for the method of determining DOR in water, sediment and fish tissue samples using HPLC–UV analysis and ME, AR and EE values of the methods
based on combining the optimal extraction procedures with LC–MS/MS measurements; the standard deviations of these values (SD) are given in brackets.

3.2. Quality assessment of the analytical procedures
The evaluation of parameters such as matrix effects, extraction
efficiency and absolute recovery during the quality assessment of a
method is crucial for obtaining reliable data. The use of different
techniques such as LC–MS/MS or HPLC–UV to analyse the same
samples can yield different data. With regard to the level of data
deemed acceptable, we took three criteria into account (Borecka
et al., 2013): the matrix effect should be in the range between
50% and 50%, the absolute recovery should not be less than 30%
and extraction efficiency should be higher than 50%.
Fig. 1 presents the values of absolute recovery obtained during
the HPLC–UV analysis of the tested extraction procedures and the
values of absolute recovery, extraction efficiency and matrix effect
for LC–MS/MS. Due to the low level of absolute recovery (<12%)
Methods 1 and 2 were not transferred to LC–MS/MS. Methods 3, 4
and 6 for isolating DOR from water, sediment and fish tissue
samples yielded the highest absolute recovery (96.5 ± 11.2%,
105.8 ± 2.4% and 67.5 ± 4.2% for water, sediment samples and fish
tissue respectively). These methods were thus transferred to
LC–MS/MS. It turned out, however, that the transfer of the extrac-
tion procedure (Method 3) for isolating DOR from water samples
to SPE-LC–MS/MS resulted in a high level of ion suppression
(97.9 ± 13.2%); on the other hand, the extraction efficiency was
high (89.9 ± 12.9%). This implies that the purification step during
SPE did not remove interferents efficiently enough, and this ham-
pered the ionisation process in LC–MS/MS measurements seriously.
This is why the absolute recovery was very low (3.3 ± 0.8%) despite
the high extraction efficiency. Nevertheless, it should be pointed
out that the absolute recovery obtained for the HPLC–UV analysis
(96.5 ± 11.2%) of water extracts was at an acceptable level, probably
because of the above-mentioned lesser sensitivity of HPLC mea-
surements to the matrix effect (detection without analyte ionisa-
tion). This was endorsed by the high extraction efficiency of the
LC–MS/MS analysis (89.9 ± 12.9%).
In order to remove organic interferents that strongly influence
DOR detection, we decided to include in Method 3 a washing step
with 6 mL of n-hexane (Table 1). Thus, the proposed Method 3A
differed from Method 3 only in this step (Table 1). As shown in
Fig. 1, this resulted in a decrease in both matrix effect and extrac-
tion efficiency (14.5 ± 4.2% and 52.3 ± 2.2% respectively) and an
increase in the absolute recovery (from 3.3 ± 0.8% to 53.4 ± 1.1%)
of the LC–MS/MS measurements. The reduction in extraction
efficiency was probably caused by the partial elution of DOR from
the sorbent together with the contaminants. Nonetheless, the val-
ues of matrix effects, extraction efficiency and absolute recovery
obtained with Method 3A were at an acceptable level.
There is only a limited amount of data in the literature concern-
ing the determination of DOR in water samples (Gravell, 2005;
Krogh et al., 2008). Krogh et al. (2008) evaluated a method of deter-
mining DOR with recoveries between 5% and 59%, depending on
the type of sorbent used and the sample pH, but they did not give
any information on analyte losses during sample preparation.
Gravell (2005) also evaluated such methods, obtaining recoveries
of 79% and 95% for respective concentrations of 40 and 10 ng L1.
Furthermore, there is no information on quality assessment as
given by extraction efficiency and absolute recovery.
Methods 4 and 5 represent approaches for extracting DOR from
sediment samples (Table 1). The conditions in Method 5 were estab-
lished on the basis of literature data concerning the determination of
DOR in soil samples (Raich-Montiu et al., 2011). The extraction pro-
cedure in Method 4 was almost identical to that in Method 5: only
the producer of the polymeric SPE sorbent was different (Method
4 – waters, Milford, MA, USA, Oasis HLB cartridges; Method 5 – Phe-
nomenex Inc., Torrance, CA, Strata X cartridges). The extraction pro-
cedure in Method 4 combined with LC–MS/MS measurements
resulted in an analytical method characterised by a very good abso-
lute recovery (99.0 ± 4.7%), a small matrix effect (36.7 ± 2.6%) and an
acceptable extraction efficiency (114.5 ± 5.3%) (Fig. 1). These values
testify to the great efficiency of the extraction, which could have
been due to the use of MAE, and also the excellent purification of
the extract with appropriate solvents.
As already mentioned, even though DOR is easily sorbed to soil
or sediment, there are few methods for determining DOR in these
components, especially in sediments. Krogh et al. (2008) extracted
DOR from sediment and soil samples with recoveries between 20%
and 70%, depending on the solvent mixture composition used for
the extraction, the extraction temperature and sample size.
Gravell (2005) performed the same investigation: his recoveries
were between 75% and 81% for soil and 80–84% for sediment
(depending on the spiked level). However, these researchers did
not state whether the recoveries they obtained were relative or
absolute. Celestina and Kožuh Eržen (2009) proposed an analytical
method for determining abamectin and doramectin in clay and
silty clay soils, but they only calculated absolute recovery,
which was lower (75–89%) than the value we obtained in our

Please cite this article in press as: Wagil, M., et al. Critical points in the evaluation of analytical methods based on liquid chromatography separation for the
determination of doramectin in different environmental samples. Chemosphere (2014), http://dx.doi.org/10.1016/j.chemosphere.2014.03.137
6 M. Wagil et al. / Chemosphere xxx (2014) xxx–xxx
Table 2
Validation parameters of the methods for determining DOR in three different matrices: water, sediment and tissue samples based on LC–MS/MS measurements.
Environmental sample Linearity range Correlation
(ng L1)a/(ng g1)b coefficient (R2)
Water 1.9a–200.0a 0.9514
Sediment 2.5b–100.0b 1.0000
Tissue 6.9b–100.0b 0.9998
a
Water samples – concentrations of analyte in (ng L1).
b
Sediment and tissue samples – concentrations of analyte in (ng g1).
c
Three samples, five replicates.
investigation. Raich-Montiu et al. (2011) also employed MAE to
extract DOR from soil. Their AR ranged from 40% to 90%, depending
on soil type; in addition, they did not observe any ion suppression.
Park et al. (2013) described both matrix effect (1.22%) and
absolute recovery (94.7%) for the determination of DOR from
soil using supercritical fluid extraction. However, none of
these researchers calculated the three most important quality
assessment parameters.
Then we tested two extraction procedures for isolating DOR
from tissue samples (Methods 6 and 7, Table 1). Method 6 had pre-
viously been used to extract avermectins from bovine liver and
muscle (Hou et al., 2007), and Method 7 for isolating avermectins
from bovine sera and porcine liver (Rudik et al., 2002); they had
never yet been used with fish tissue samples. As higher absolute
recovery values were obtained for Method 6 (67.5 ± 4.2%), it was
this approach that was combined with LC–MS/MS measurements.
The following values of absolute recovery (AR), matrix effect (ME)
and extraction efficiency (EE) were obtained: AR = 36.3 ± 5.1%;
EE = 51.3 ± 10.7%; ME = 27.2 ± 0.3% (Fig. 1). They were considered
acceptable in the light of the previously mentioned criteria
(Borecka et al., 2013). As in the case of the water samples, the
extraction efficiency of the LC–MS/MS analysis was similar to the
absolute recovery of the HPLC–UV analysis. Noppe et al. (2009)
developed a method of determining avermectins (including DOR)
in porcine liver, meat and fish tissue. For fish tissue they obtained
recoveries between 80% and 110% but did not state whether these
were relative or absolute. If analytical data are to be reliable, then
the type of recovery must be given.
3.3. Validation of methods for determining DOR in water, sediment
and fish tissue samples based on LC–MS/MS analysis
The methods for determining DOR in water, sediment and fish
tissue samples based on LC–MS/MS measurements were validated
using working calibration standard solutions and matrix-matched
calibration solutions according to the procedure described in our
previous studies (Migowska et al., 2012). The validation parame-
ters of these methods are listed in Table 2. The linearity of these
methods was determined by analysing a minimum of six different
concentrations of each compound in each matrix with six repli-
cates. The calibration curves for LC–MS/MS measurements were
linear with a correlation coefficient (R2) P 0.9514. The accuracy
of the methods was determined by assessing the agreement
between the measured and known concentrations of the samples.
These accuracies ranged from 74.0% to 119.4% for water samples,
from 99.6% to 119.8% for sediment samples and from 97.4% to
106.1% for fish tissue samples (Table 2). The intra-day precision
was between 9.2% and 9.6% for water samples, between 1.7% and
8.6% for sediment samples and between 2.7% and 5.3% for fish tis-
sue samples. As recommended by the U.S. EPA (2003), the values
are precise when RSD is <20%. The method detection limits of
determining DOR in environmental samples were 0.6 ng L1 for
water, 0.8 ng g1 for sediment, and 2.3 ng g1 for fish tissue
samples.
Please cite this article in press as: Wagil, M., et al. Critical points in the evaluation of analytical methods based on liquid chromatography separation for the
determination of doramectin in different environmental samples. Chemosphere (2014), http://dx.doi.org/10.1016/j.chemosphere.2014.03.137
Intra-day precision Accuracy (%) MDL (ng L1)a/ MQL (ng L1)a/
(RSD (%))c (ng g1)b (ng g1)b
9.2–9.6 74.0–119.4 0.6a 1.9a
1.7–8.6 99.6–119.8 0.8b 2.5b
2.7–5.3 97.4–106.1 2.3b 6.9b
These values are in agreement with the MDLs given for methods
proposed by other researchers (Gravell, 2005; Krogh et al., 2008;
Noppe et al., 2009; Giannetti et al., 2011; Park et al., 2013). For
example, the LOQs of the methods for determining DOR are
2.4 ng L1 (Krogh et al., 2008) or 0.87 ng L1 (Gravell, 2005) in sur-
face waters, and 3.9 lg L1 for sediments (Gravell, 2005). There are
no data in the literature on determining LOQ in fish tissue. How-
ever, there are very few reports in the literature on methods for
determining DOR in environmental samples such as water, sedi-
ment and fish tissue: two methods for water samples (Gravell,
2005; Krogh et al., 2008), one for sediments (Gravell, 2005) and
one for fish tissue (Noppe et al., 2009).
3.4. Analysis of environmental samples
In order to verify the applicability of the suggested methods, we
collected environmental samples from the River Gościcina in
northern Poland. This river was chosen because eleven fish farms,
farms and dog kennels are situated along it. In order to investigate
seasonal influences, samples were collected during autumn 2012
and spring 2013. DOR residues were found in all the matrices col-
lected in spring 2013 but in none of the samples harvested in
autumn 2012. DOR was detected in five water samples, but only
in one was the concentration higher (1.92 ng L1) than the MQL.
DOR was present in eight sediment samples and in four fish tissue
samples collected in spring 2013, but the concentrations were
below the MQL.
In the literature there are hardly any data on DOR concentra-
tions in environmental water, sediment and fish tissue samples
(Pfizer, 1996; Krogh et al., 2008). For example, this drug was not
found in freshwater sediments from the River Elbe in Germany,
because the concentrations were lower than the LOD of DOR
(Krogh et al., 2008).
4. Conclusions
The main aim of this work was to show how quality assessment
should be performed in order to transfer an optimised method
from one technique to another. This assessment was based on
the evaluation of methods for determining doramectin in three
environmental samples: water, sediment and fish tissue using
HPLC–UV and LC–MS/MS. The quality of the suggested methods
for determining this drug in these matrices was evaluated on the
basis of known values of absolute recovery (HPLC–UV) and matrix
effect, extraction efficiency and absolute recovery (LC–MS/MS).
Finally, the suggested methods for determining DOR in water, sed-
iment and fish tissue based on LC–MS/MS measurements were val-
idated and applied to the analysis of real environmental samples
collected from the River Gościcina. The obtained matrix effects,
absolute recovery and extraction efficiency for the proposed LC–
MS/MS methods of determining DOR in environmental samples
were as follows: 14.5%, 53.4% and 52.3% for water; 36.7%, 99.0%
and 114.5% for sediment, 27.2%, 36.1% and 51.1% for tissue. The
data are within the established criteria. Most of the real samples
 M. Wagil et al. / Chemosphere xxx (2014) xxx–xxx
collected in spring 2013 contained DOR but the concentrations
were below MQL, whereas this drug was not found in samples har-
vested in autumn 2012. In summary, the transfer of an optimised
method from one technique to another is possible, but a proper
quality assessment of the new method based on matrix effect,
extraction efficiency and absolute recovery is mandatory.
Acknowledgements
Financial support was provided by the National Science Centre
in accordance with decision DEC-2011/03/B/NZ8/03010, Poland.
Publication is financed from the European Social Fund as a part
of the project ‘‘Educators for the elite – an integrated training pro-
gramme for PhD students, post-docs and professors as academic
teachers at the University of Gdansk’’ within the framework of
the Human Capital Operational Programme, Action IV.
Appendix A. Supplementary material
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.chemosphere.
2014.03.137.
References
Antignac, J.-P., de Wasch, K., Monteau, F., De Brabander, H., Andre, F., Le Bizec, B.,
2005. The ion suppression phenomenon in liquid chromatography–mass
spectrometry and its consequences in the field of residue analysis. Anal.
Chim. Acta 529, 129–136.
Borecka, M., Białk-Bielińska, A., Siedlewicz, G., Kornowska, K., Kumirska, J.,
Stepnowski, P., Pazdro, K., 2013. A new approach for the estimation of
expanded uncertainty of results of an analytical method developed for
determining antibiotics in seawater using solid-phase extraction disks and
liquid chromatography coupled with tandem mass spectrometry technique. J.
Chromatogr. A 1304, 138–146.
Caban, M., Migowska, N., Stepnowski, P., Kwiatkowski, M., Kumirska, J., 2012.
Matrix effects and recovery calculations in analyses of pharmaceuticals based
on the determination of b-blockers and b-agonists in environmental samples. J.
Chromatogr. A 1258, 117–127.
Celestina, T.V., Kožuh Eržen, N., 2009. Validation of analytical procedure for
simultaneous determination of two avermectins in various soils. Acta Chim.
Slov. 56, 1010–1015.
Danaher, M., Howells, L.C., Crooks, S.R.H., Cerkvenik-Flajs, V., O’Keeffe, M., 2006.
Review of methodology for the determination of macrocyclic lactone residues
in biological matrices. J. Chromatogr. B 844, 175–203.
Ellison, S. L. R., Williams, A., (Eds), 2012. Eurachem/CITAC guide: Quantifying
Uncertainty in Analytical Measurement, third ed., ISBN 978-0-948926-30-3.
Available from: <www.eurachem.org>.
Fatta, D., Nikolaou, A., Achilleos, A., Meriç, S., 2007. Analytical methods for tracing
pharmaceutical residues in water and wastewater. TrAC 26, 515–533.
Please cite this article in press as: Wagil, M., et al. Critical points in the evaluation of analytical methods based on liquid chromatography separation for the
determination of doramectin in different environmental samples. Chemosphere (2014), http://dx.doi.org/10.1016/j.chemosphere.2014.03.137
7
Giannetti, L., Giorgi, A., Necci, F., Ferretti, G., Buiarelli, F., Neri, B., 2011. Validation
study on avermectine residues in foodstuffs. Anal. Chim. Acta 700, 11–15.
Gravell, A., 2005. Agilent Technologies Inc., Application: validated method for
the determination of veterinary medicines in the environment using the agilent
LC–MSD quad and ion trap, Environment Agency, National Laboratory Services,
UK.
Hall, T.G., Smukste, I., Bresciano, K.R., Wang, Y., Mckearn, D., Savage, R.E., 2012.
Identifying and overcoming matrix effects in drug discovery and development.
In: Prasain, J.K. (Ed.), Tandem Mass Spectrometry – Applications and Principles.
InTech, Croatia.
Hernández, F., Sancho, J.V., Ibáñez, M., Guerrero, C., 2007. Antibiotic residue
determination in environmental waters by LC–MS. TrAC 26, 466–485.
Hou, X.I., Wu, Y., Shen, J., Wang, L., Ding, S., 2007. Multi-residue analysis of
avermectins in bovine liver and muscle by liquid chromatograpgy- fluorescence
detector. Chromatographia 65, 77–80.
Kolar, L., Kožuh Eržen, N., 2006. Veterinary parasiticides – are they posing an
environmental risk? Slov. Vet. Res. 43, 85–96.
Kołodziejska, M., Maszkowska, J., Białk-Bielińska, A., Steudte, S., Kumirska, J.,
Stepnowski, P., Stolte, S., 2013. Aquatic toxicity of four veterinary drugs
commonly applied in fish farming and animal husbandry. Chemosphere 92,
1253–1259.
Krogh, K.A., Björklund, E., Loeffler, D., Fink, G., Halling-Sørensen, B., Ternes, T.A.,
2008. Development of an analytical method to determine avermectins in water,
sediments and soils using liquid chromatography–tandem mass spectrometry.
J. Chromatogr. A 1211, 60–69.
Mannuer, V.S., Patel, D., Mastiholimath, V.S., Shah, G., 2011. Selection of buffers in
LC–MS/MS: an overview. Int. J. Pharm. Sci. Rev. Res. 6, 34–37.
Migowska, N., Caban, M., Stepnowski, P., Kumirska, J., 2012. Simultaneous analysis
of non-steroidal anti-inflammatory drugs and estrogenic hormones in water
and wastewater samples using gas chromatography–mass spectrometry and
gas chromatography with electron capture detection. Sci. Total Environ. 441,
77–88.
Noppe, H., Verheyden, K., Vanden Bussche, J., Wille, K., De Brabander, H., 2009.
Detection of macrocyclic lactones in porcine liver, meat and fish tissue using
LC–APCI–MS–MS. Food Addit. Contam. 26, 1232–1238.
Park, J.-H., Choi, J.-H., Abd El-Aty, A.M., Park, J.-S., Kim, B.M., Na, T.-W., Park, K.H.,
Yang, A., Md Rahman, M., Shim, J.-H., 2013. Development of an extraction
method for the determination of avermectins in soil using supercritical CO2
modified with ethanol and liquid chromatography–tandem mass spectrometry.
J. Sep. Sci. 36, 148–155.
Patel, D., 2011. Matrix effects in a view of LC–MS/MS: an overview. Int. J. Pharm. Bio.
Sci. 2, 560–564.
Pfizer Inc., 1996. Doramectin 1% injectable solution for the treatment of
parasitic infections in swine. Environmental Assessment Report, NADA
141-061.
Raich-Montiu, J., Prat, M.D., Granados, M., 2011. Extraction and analysis of
avermectines in agricultural soils by microwave assisted extraction and ultra
high performance liquid chromatography coupled to tandem mass
spectrometry. Anal. Chim. Acta 697, 32–37.
Rudik, I., Cummings, M.R., Poppenga, R.H., 2002. Isolation and multiresidue
detection of macrolide endectocides present in animal matrices. J. Vet. Diagn.
Invest. 14, 295–302.
Díaz-Cruz, M. Silvia, Barceló, D., 2007. Recent advances in LC–MS residue analysis of
veterinary medicines in the terrestrial environment. TrAC 26, 637–646.
ThermoQuest, 1999. Finnigan LCQDECA Operational Tips.
U.S. EPA recommendations 2003. Method 8000C – determinative chromatographic
separations, U.S. Environmental Protection Agency.