Solvent Extraction Research and Development, Japan, Vol.

28, No 2, 149 – 156 (2021)

House Microwave-Assisted Solid Phase Extraction for Residual 17α-Methyltestosterone

Determination in Nile Tilapia Tissues by High-Performance Liquid Chromatography

Kawin KHACHORNSAKKUL1, Sudtida Pliankarom THANASUPSIN2 and Wijitar

DUNGCHAI1,*
1
Department of Chemistry, Faculty of Science, King Mongkut’s University of Technology Thonburi,
Pracha-U-Tit, Thungkru, Bangkok, 10140, Thailand; 2Chemistry for Green Society and Healthy Living
(ChGSH) Research Unit, Department of Chemistry, Faculty of Science, King Mongkut’s University of
Technology Thonburi, Pracha-U-Tit, Thungkru, Bangkok, 10140, Thailand.
(Received December 23, 2020; Accepted February 1, 2021)

An effective sample preparation procedure using a house microwave-assisted extraction (hMAE) procedure,
followed by cleaning with solid-phase extraction (SPE) by using HPLC with general UV-visible
spectrometry was developed for residual 17α-methyltestosterone (MT) determination in Nile tilapia tissues.
Our developed method showed a wide linear range from 25 to 800 ng g-1 with R2 as 0.9985 and presented a
low detection limit of 1.53 ng g-1 with good precision. Evaluation of accuracy for the MT determination in
real samples showed a great recovery percentage ranging from 98.41 to 100.78%. In conclusion, it can be
said that the developed method has the potential for sample preparation when compared with another method,
and can be used for the determination of MT residues in fish tissue with simplicity, rapidity, cost-
effectiveness, low solvent consumption, and minimum waste generation as well as having high accuracy and
precision.

1. Introduction
Nile tilapia, known as the “economical fish", is used to make various food owing to its low price, taste
and high nutritional value. However, male tilapia is more favored than female because it can convert energy
from feeding into muscle better than female tilapia, which provides higher economic productivity and profits,
therefore, a synthetic anabolic-androgenic steroid was administered in order to increase the male hormone
[1-3]. 17α-methyltestosterone (MT), a steroid hormone, is mixed with tilapia feed and is provided at an early
stage to selectively attain a male population. Generally, fish food containing 60 mg kg-1 of MT is given to
Nile tilapia for 21 days to produce a male Nile tilapia population [4-6], but a small amount of MT can affect
human health which is not approved for use in fish in the United States [7]. For the abovementioned, it is
necessary to develop a high sensitivity technique for monitoring the residues of MT in Nile tilapia in order
to evaluate the risk from their influence.
A sensitive assay for detecting MT at a small ng level in residues of fish tissue is desired because it is
required by the Food and Drug Administration (FDA). However, fish samples are complex in terms of
composition which causes more difficult analysis. Hence, the development of separation techniques for
residues of MT detection in fish can be challenging to attain high reproducibility and recovery. Currently,

- 149 -
numerous chromatography method combined with sample preparation processes, i.e., subcritical fluid
extraction (SFE), solid-phase extraction with salt (QuEChERS), solid-phase extraction (SPE), and solid-
phase microextraction (SPME), for determining MT content in various samples have been developed to
reduce the matrix effect and enhance measurable efficiency such as HPLC, liquid chromatography-mass
spectrometry (LC-MS), gas chromatography (GC) and GC-MS, [8-17]. Although MS detector can be widely
used to detect MT at ng level, it is expensive and needs experts for operation and analysis. Thus, our main
strategy was to develop an effective sample preparation method for measuring MT residues in Nile tilapia
tissues by using HPLC with a UV-Visible spectrometric detector (HPLC-UV), which is a general
chromatographic instrument and detector.
Nowadays, commercial home-microwave device benefits our life in many ways. It can also be used in
the science research application, especially for enhancing extraction techniques, called “Microwave-assisted
extraction; MAE” which is a popular method for solid-liquid extraction due to its rapidity, simplicity, cost-
effectiveness, and high performance. For example, Ying et al. presented MAE assisted SPE extraction for
alkaloid analysis in Stephania cepharantha sample, which demonstrated that the combination of MAE and
SPE enhanced the extraction and detection ability better than obtained the SPE method alone [18]. It can thus
be said that MAE system can assist the increasing of separation and measurement performance for analysis.
Likewise, organic compounds can be digested and extracted from solid matrix samples, such as soil, sludge
and tissue samples, by electric power and electrothermal heating from a home microwave. With respect to
green chemistry concepts, MAE is considered to be an efficiently useful tool because it can enhance
extraction performance and measurability while incorporating low solvent consumption and pre-
concentration at the same time [19-25].
Herein, we proposed the application of home microwave-assisted solid-phase extraction (hMAE-SPE)
for the determination of MT residue in Nile tilapia tissues at ng level. This developed technique performed
hMAE to increase the extractability and measurability in order to determine the trace of MT in the fish sample.
Similarly, SPE was used to clean up the interference matrix such as phospholipids or proteins and pre-
concentration of MT content. To our knowledge, this method is the first time to combine hMAE and SPE for
the determination of MT in fish samples. Furthermore, the developed approach offers great analytical
performance when compared with other methods, i.e., SPE or QuEChERS, as well as much benefit involved
in less solvent consumption and cost-effectiveness.

2. Experimental
2.1 Chemical and materials
All solvents used in extraction and analysis steps were HPLC grade. Methanol, ethanol, acetonitrile
(ACN), and dichloromethane (DCM) were purchased from Fisher Chemical. 17α-Methyltestosterone (MT)
97% HPLC grade and testosterone (T-II) 99% HPLC grade were purchased from Sigma-Aldrich. Preparation
of 5.0 mg L-1 of MT and T-II stock solution was performed by dissolving 0.50 mg of MT and T-II in methanol
and adjusting to 100 mL in a volumetric flask. A house microwave (Samsung, ME81KS-1ST) was used in
this study. Bond Elute C18 solid-phase extraction (SPE) cartridges (sorbent mass 1.0 g, volume 6.0 mL,
particles size 120 µm, and frit pore size 20 µm) were purchased from Agilent and 0.25 µm syringe
polyethylene terephthalate (PET) was purchased from Fisher Chemical.

- 150 -
2.2 Chromatographic conditions
HPLC-UV detector (Varian Pro Star, THAI UNIQUE CO., LTD, UV detector 310 ProStar and pump
230 ProStar) was performed with eclipse plus C18 column containing 4.6 mm × 150 mm and particles size
as 5 µm by the surface area of 160 m²/g and controlled pore size of 95Å (Agilent, Zorbax). Mobile phase as
methanol-ultrapure water (70:30 ratio) with isocratic eluent processing was conducted at a constant flow rate
(1.0 mL min-1) and detected at 245 nm.
2.3 Sample preparation procedures
Samples of tilapia were purchased from local markets. Fish tissue samples were mixed and
homogenized. First, 0.10 g of samples were filled into the centrifuge tube and spiked with 0.20 mL of MT
standard stock. Next, vortex was done for 2 min after that 4.0 mL of methanol was added into the mixture of
samples. Then, the extraction was conducted using the home microwave at 600 W for 3 min. Later, the
centrifuge tubes containing the mixture of samples were left at room temperature to cool down. Next, the
mixture was filtered with a 0.25 µm syringe filter. After that, the filtered solution was added to 0.10 mL of
T-II solution (which is used as an internal standard). Then, the total volume was adjusted to 5.0 mL by
methanol. In order to clean up impurities and pre-concentration, the solution was passed through the SPE
cartridge. Milli-Q water and ethanol were used as washing and eluting solvent, respectively (constant flow
rate at 1.0 mL min-1). Finally, the eluting solvent was evaporated to dryness, and the residue was dissolved
in 0.10 mL of methanol. 0.02 mL of this successful solution was injected into HPLC for analysis.
2.4 General procedure of analytical characteristics
The linearity range was obtained by various concentrations of MT in fish tissues using a spiked method
from 25.0 to 800.0 ng g-1 (n = 3), after that samples were prepared according to the proposed method, and
detection was conducted with HPLC-UV. Repeatability and reproducibility of our proposed method were
determined by analysis of MT at 50.0, 100.0, and 200.0 ng g-1 (n = 10). The limit of detection (LOD) and
limit of quantification (LOQ) were calculated from the detection of the blank signal. The blank sample was
performed without being spiked and followed the same procedure as for the analytes.

3. Results and Discussion

3.1 Optimization of sample preparation
Sample preparation is one of the crucial steps of good analytical practices. Therefore, it is necessary
to study factors affecting the extraction performance of samples from the matrix. In this study, several factors
were investigated, such as types of extractant solvent, solvent volume, microwave energy, extraction time,
and types of washing solvent, washing volume, types of eluting solvent, and eluting volume. The optimized
sample preparation factors were evaluated based on recovery percentage obtained from a spiked method of
MT standard at 200.0 ng g-1 and testosterone (T-II), which acts as an internal standard (at 100.0 ng g-1) to
reduce the tolerance in the analysis. The conditions were tested three times (n = 3). For detection, HPLC was
used with fixed conditions for separation. According to the chromatogram of the proposed method, as shown
in Figure 1, it gave a higher resolution (Rs) than 1.5. In the extraction of organic compounds in fish tissues,
the influential parameters of sample preparation were evaluated. Firstly, organic solvents for microwave
extraction are very important because they affect extracted performance of solid samples. Therefore, the
initial study was on the type of extractant solvent in order to obtain the highest extraction efficiency.

- 151 -
Methanol and water were chosen as extractant
solvents due to the high dielectric constant in which
they were transformed into heat through ionic
conduction and dipole rotation ability. The analysis
was performed with various methanol-water ratio
ranging from 0:100 to 100:0. As can be seen in
Figure 2(a), increasing the methanol ratio caused an
increase in recovery percentage. As expected, pure
methanol showed the highest percentage recovery.
Thus, pure methanol was selected as a suitable
extractant solvent. The volume of extractant solvent Figure 1. Chromatogram of MT at 200 ng g-1 and
used for extraction of a fixed amount of sample, T (internal standard) at 100 ng g-1, including their
known volume of solvent, plays an important part in structure (insert) for proposed method.
the extraction step. Various solvent
volumes (ranging from 2.0 to 10.0 mL)
were studied. As shown in Figure 2(b).
Four milliliters of methanol were selected
because the methanol volume tended to
be constant of extracted performance
after 4.0 mL of methanol. With respect to
one of the green chemistry principles
which are ‘reducing solvent’, the
minimum volume of methanol 4.0 mL
was chosen as a sufficient volume of
Figure 2. Optimum conditions for MT determination,
solvent for achieving the highest recovery
including (a) methanol: water ratio, (b) extractant volume,
percentage. The microwave energy level
(c) energy levels and (d) extraction time in microwave
and extraction time were powerful
assisted extraction process (n = 3).
parameters involved in the analysis
because methanol can evaporate when it consumes excess energy and extraction time. However, low
extraction efficiency may occur with insufficient energy and time. For this reason, suitable microwave energy
levels and extraction time were examined. We found that the highest recovery for the MT determination in
fish tissues was found at 600 W and 3 min for MAE while increasing the energy level and extraction time
related to decreasing recoveries of analysis as the result indicated in Figure 2(c-d). We found that using too
high of both conditions caused solvent evaporation because there was no liquid left in the sample container.
Therefore, 4.0 mL of pure methanol was used as extractant solvent at 600 W for 3 min in the MAE process.
In order to clean up and pre-concentration the analyte after MAE step, SPE was performed to enhance
a potential extraction of MT from complex matrices such as phospholipids and proteins. Optimization of
washing and eluting process in SPE cartridge can reduce the matrix effect and improve the detecting ability
of analysis. Therefore, the types and volume of solvents used for washing and eluting steps of MT extraction
were studied. For this method, the consequence of reversed-phase sorbent showed no interference in the

- 152 -
analysis after washing with Milli-Q
water. Figure 3(a) illustrated that the
highest recovery percentage was obtained
when using Milli-Q water as washing
solvent because the target analyte cannot
dissolve in the high polar solvent [12].
Figure 3(b) revealed that when adding
more than 3.0 mL of Milli-Q water, the
recovery percentage remained
unchanged. Therefore, 3.0 mL of pure
Milli-Q water was used as washing Figure 3. Optimum conditions for MT determination,
solvent to clean up in SPE. For the eluting including (a) washing solvent, (b) washing volume, (c)
step, various types of eluting solvents were eluting solvent, (d) eluting volume in solid phase extraction
studied. In comparison with other solvents, process (n = 3).
it was found that ethanol showed the
highest recovery percentage because MT molecules can be greatly dissolved in ethanol due to their polarity
[12]. Also, 3.0 mL of ethanol was selected as great eluent and sufficient to elute the analyte as shown in
Figure 3(c-d).
3.2 Analytical characteristics
The potential analysis of the developed approach was validated for the determination of residual MT
in Nile tilapia tissues by this appropriate sample preparation technique. Analytical performances were
determined under the optimum parameters for the preparation of fish tissue samples. Figure 4 indicated that
linearity was found in the range from 25.0 to 800.0 ng g-1 with a regression coefficient (R2) equaled to 0.9985.
The limit of detection (LOD) and limit of quantification (LOQ) were calculated from the following equations:
LOD = 3 s/m and LOQ = 10 s/m, where ‘s’ was the deviation of blank signals (n = 10) and ‘m’ was the slope
of the calibration curve. The values of LOD and LOQ were 1.53 and 5.08 ng g-1, respectively. The precision
of the developed method was evaluated by conducting repeatability and reproducibility tests for the
determination of MT standard at 50.0, 100.0,
200.0, and 400.0 ng g-1 concentration (n = 10)
and found the highest relative standard deviation
(RSD) as 1.5% as shown in Table 1.
A chromatogram was observed and no
interference compound indicated a close
position of the target compound. By comparing
the analytical performances of the proposed
method with other sample preparation methods
for detecting MT using HPLC, it was found that
the proposed method gave a lower detection
limit and a better percentage recovery as the Figure 4.  Linearity range for determination of MT
result indicated in Table 2. Concisely, it can be using proposed preparation method (n = 3).

- 153 -
observed that the SPE process alone can detect MT in Carassius tissues with adequate recovery levels,
nevertheless, a detection limit of this approach is still too high for monitoring of MT in fish tissues, i.e., 300
ng g-1. Our technique offers the recoveries values and detection limit greater than that method.
Table 1. Results for repeatability and reproducibility for developed method (n = 10).
Concentration of MT Repeatability Reproducibility (%RSD)
(ng g-1) (%RSD)
Intra-day Inter-day
50.0 1.22 1.33 1.53
100.0 1.32 1.39 1.49
200.0 1.27 1.40 1.35
400.0 1.32 1.36 1.26

Table 2. Analytical performance of the proposed method compared with other sample preparation methods
for monitoring MT using HPLC.
Separation Linearity range LOD % Recoveries Sample Reference
methods (µg g-1) (ng g-1)
SFE 0.05 – 0.25 10.0 90.0 Tilapia 8
QuEChERS 0.5 – 10.0 33.0 91.6 – 107.2 Shrimp 9
SPE 1.0 – 100.0 300.0 82.9 – 93.6 Carassius 10
LLE-SPE 15.0 – 120.0 1300.0 93.78 – 99.43 Fish feed 11
SPE 0.05 – 2.0 19.0 93.1 – 103.9 Freshwater 12
MAE-SPE 0.025 – 0.80 1.53 98.41 – 100.78 Nile tilapia This work

SFE = Supercritical fluid extraction

QuEChERS = Solid phase extraction with salt
LLE = Liquid-liquid extraction

3.3 Application for real samples

This proposed method was used to determine the amount of residual MT content in the presence of
Nile tilapia tissue samples. Six samples were collected at different areas in Thailand such as local markets
and natural ponds. Samples were prepared by our proposed method and calculate the amounts of residual
MT. Table 3 shows of amounts of MT concentration found in fish samples based on the developed method.
The different samples (A, B, C, D, E and F) found the concentration of MT equal to 202.0, 454.3, 406.2,
597.9, 100.9 and 34.9 ng g-1 respectively. Furthermore, the spiked method was implemented to validate the
measurements with the proposed sample preparation method. Fish tissue samples were spiked with three
different concentrations of MT standard solution (50.0, 100.0 and 150.0 ng g-1).
Table 3. Results of MT level in fish tissues based on the developed method (n = 3).
Samples Total found (ng g-1)
A 202.0 ± 0.10
B 454.3 ± 0.60
C 406.2 ± 0.20
D 597.9 ± 0.20
E 100.9 ± 0.15
F 34.9 ± 0.18

- 154 -
 Table 4 demonstrates the recovery percentages range from 98.4 to 100.8% with the highest RSD as
1.25%. These results indicated that the developed sample preparation method can be efficiently applied for
the determination of MT in fish samples.
Table 4. Recovery testing in fish tissue samples in the proposed method (n = 3).
Samples Found Added Total found % Recoveries % RSD
(ng g-1) standard MT (ng g-1)
(ng g-1)
50.0 248.8 98.7 0.40
A 202.0 100.0 297.2 98.4 0.24
150.0 349.7 99.4 0.31
50.0 508.2 100.8 1.05
B 454.3 100.0 545.7 98.4 0.45
150.0 598.4 99.0 0.36
50.0 452.2 99.1 0.62
C 406.2 100.0 506.6 100.1 0.97
150.0 554.8 99.7 0.62
50.0 616.6 99.8 0.53
D 597.9 100.0 658.3 98.6 1.12
150.0 718.4 100.1 0.66
50.0 149.1 98.8 0.69
E 100.9 100.0 199.8 99.4 0.63
150.0 250.1 99.7 1.25
50.0 83.6 98.5 0.58
F 34.9 100.0 134.4 99.7 0.76
150.0 184.8 99.9 0.92

4. Conclusion
We can conclude that the determination of organic compounds in a sample containing a complex
matrix requires an appropriate sample preparation technique to reduce the loss of the analyte and to increase
the measurement efficiency. This proposed technique was developed for the detection of residual steroid
hormone content, MT, in Nile tilapia tissue samples by HPLC-UV. With respect to its efficient and easy-to-
use performance, a house microwave was chosen to assist solid-phase extraction for fish sample preparation.
This technique can enhance detectability with a satisfactory recovery percentage and limit of detection. As
well, it also exhibits many benefits for analysis such as cost-efficiency, rapidity, simplicity, high sample
throughput, and low solvent consumption. More importantly, the analytical characterization of our proposed
technique can be used to screen the residual MT at ng level in fish tissues with extreme accuracy and precision.
This appropriate sample preparation can be applied as alternative sample preparation techniques to determine
MT residues in fish tissues for both routine quality control and analytical research in various applications
such as industry, agriculture, and medicine.

Acknowledgement
The authors are thankful for the financial support from Petchra Pra Jom Klao Ph.D. Research
Scholarship, King Mongkut’s University of Technology Thonburi.

- 155 -
 References
1) A. Cnaani, B. Levavi-Sivan, Sex. Dev., 3, 164-175 (2009).
2) W. L. Gale, M. S. Fitzpatrick, M. Lucero, W. M. Contreras-Sánchez, C. B. Schreck, Aquaculture, 178,
349-357 (1999).
3) J. Beardmore, G. Mair, R. Lewis, Aquaculture, 197, 283-301 (2001).
4) T. J. Pandian, S. G. Sheela, Aquaculture, 138, 1-22 (1995).
5) J. S. Abucay, G. C. Mair, Aquacult. Res., 28, 841-845 (1997).
6) L. R. Curtis, F. T. Diren, M. D. Hurley, W. K. Seim, R. A. Tubb, Aquaculture, 99, 193-201 (1991).
7) P. S. Chu, M. Lopez, S. Serfling, C. Gieseker, R. Reimschuessel, J. Agric. Food Chem., 54, 3193-3198
(2006).
8) Y. Han, Q. Ma, J. Lu, Y. Xue, C. Xue, Food Chem., 135, 2988-2993 (2012).
9) R. Fu, A. Zhai, Agilent Technologies Co. Ltd., China, 5990-6589EN (2012).
10) C.H. Zhai, Y. Zou, C. Rou, N. Jin, Agilent Technologies Co. Ltd., China, 5990-3845EN (2009).
11) A. Marwah, P. Marwah, H. Lardy, J. Chromatogr. B: Anal. Technol. Biomed. Life Sci., 824, 107-115
(2005).
12) I. R. Barbosa, S. Lopes, R. Oliveira, I. Domingues, A. M. V. M. Soares, A. J. A. Nogueira, Am. J. Anal.
Chem., 4, 207-211 (2013).
13) J. M. Storey, S. B. Clark, A. S. Johnson, W. C. Andersen, S. B. Turnipseed, J. J. Lohne, R. J. Burger, P.
R. Ayres, J. R. Carr, M. R. Madson, J. Chromatogr. B: Anal. Technol. Biomed. Life Sci., 972, 38-47
(2014).
14) D. Zeng, C. Lin, Z. Zeng, X. Huang, L. He, Agric. Sci. China, 9, 306-312 (2010).
15) P. Regal, C. Nebot, B. I. Vázquez, A. Cepeda, C. A. Fente, Meat Sci., 84, 196-201 (2010).
16) Z. Zhang, H. Duan, L. Zhang, X. Chen, W. Liu, G. Chen, Talanta, 78, 1083-1089 (2009).
17) P. Li, Z. Yan, X. Sun, S. Chen, Y. Chen, X. Zhang, J. Ocean Univ. China, 17, 1171-1177 (2018).
18) Y. Liu, D. Xie, Y. Kang, Y. Wang, P. Yang, J. Guo, J. Huang, J. Chromatogr. Sci., 54, 670-676 (2016).
19) R. Guedes-Alonso, Z. Sosa-Ferrera, J. J. Santana-Rodríguez, Food Chem., 237, 1012-1020 (2017).
20) D.-L. Su, P.-J. Li, S. Y. Quek, Z.-Q. Huang, Y.-J. Yuan, G.-Y. Li, Y. Shan, Food Chem., 286, 1-7 (2019).
21) M. Magnusson, A. K. L.Yuen, R. Zhang, J. T. Wright, R. B. Taylor, T. Maschmeyer, R. Nys, Algal Res.,
23, 28-36 (2017).
22) H. Wang, J. Ding, N. Ren, TrAC, Trends Anal. Chem., 75, 197-208 (2016).
23) A. Azzouz, E. Ballesteros, Sci. Total Environ., 419, 208-215 (2012).
24) L. Sanchez-Prado, C. Garcia-Jares, T. Dagnac, M. Llompart, TrAC, Trends Anal. Chem., 71, 119-143
(2015).
25) Y. Fan, Z. Niu, C. Xu, L. Yang, F. Chen, H. Zhang, Ind. Crops Prod., 141, 111809 (2019).

- 156 -