magnetic solid-phase extraction using magnetic hypercrosslinked polymer for rapid determination of...

Research Article

Magnetic solid-phase extraction usingmagnetic hypercrosslinked polymer for rapiddetermination of illegal drugs in urine

A novel magnetic material Fe3O4/SiO2/P(MAA-co-VBC-co-DVB) was prepared via the

hypercrosslinking of its precursor which was produced via precipitation polymerization of

methacrylic acid (MAA), vinylbenzyl chloride (VBC), and divinylbenzene (DVB) in the

presence of Fe3O4/SiO2 submicrospheres with the surface containing abundant reactive

double bonds. The resultant sorbent was characterized by scan electron microscopy, N2

adsorption, and Fourier transform infrared spectroscopy. It was found that this material

had remarkable features such as large surface area (500 m2/g) and pore volume (0.32 cm3/

g), as well as desirable chemical composition (including hydrophobic and ion-exchange

moieties). Taking advantages of the Fe3O4/SiO2/P(MAA-co-VBC-co-DVB), a magnetic SPE

(MSPE) coupled with capillary electrophoresis (CE) method was developed for the

determination of illegal drugs in urine samples. The extraction time could be clearly

shortened up to 3 min. The recoveries of these drug compounds were in the range of

84.0–123% with relative standard deviations ranging between 1.7 and 10.5%; the limit of

detection was in the range of 4.0–6.0 mg/L. The proposed method is simple, effective, and

low-cost, and provides an accurate and sensitive detection platform for abused drug

analysis.

Keywords: Capillary electrophoresis / Fe3O4/SiO2/P(MAA-co-VBC-co-DVB) /Hypercrosslinking / Illegal drugs / Magnetic SPEDOI 10.1002/jssc.201100634

1 Introduction

Solid-phase extraction (SPE), by far, is the most widely used

method for the preconcentration and purification of a wide

range of analytes from various matrices. Advantages of

these solid-sorbent-based solid–liquid extraction techniques

include the simplicity, flexible selection of sorbents, and low

consumption of organic solvents [1–5]. Of all the different

types of sorbents available (i.e. silica-, carbon-, and polymer-

based sorbents), polymeric materials are one group of the

most important sorbents used in SPE, since they offer

distinct advantages such as sorbent stability under a very

broad pH range of analysis conditions [6].

As a new generation of highly porous polymers,

hypercrosslinked polymer (HCP) possesses high micropore

contents and correspondingly high-specific surface areas

(>1000 m2/g), suggesting its great application potential in

sorption processes [7–10]. So far, HCPs (with or without

enhanced hydrophilicity) have delivered satisfactory results

when used as sorbents in SPE to extract several groups of

compounds from different samples matrices [6, 11–14].

However, almost all these reported SPE processes needed to

perform a time-costing step of column passing and filtration

operation because HCP sorbents were generally packed into

SPE cartridges. On the other hand, current researches on

SPE materials mainly focus on mixed-mode sorbents that

combine both properties, i.e. capacity and selectivity, in a

single material [15]. In such type of sorbent, the matrix

components (or interferences) and analytes can be eluted

separately during the washing and the elution steps,

respectively.

Magnetic SPE (MSPE) has received considerable atten-

tion in recent years due to its great potential applications in

separation science [16–20]. The adsorbent need not be

packed into cartridges, instead dispersed in a sample solu-

tion or suspension. The powdery magnetic adsorbent can be

reversibly agglomerated and redispersed in solution or

Qiang Gao1,2�

Cai-Yong Lin1�

Dan Luo3

Li-Li Suo1

Jian-Li Chen1

Yu-Qi Feng1

1Key Laboratory of AnalyticalChemistry for Biology andMedicine (Ministry ofEducation), Department ofChemistry, Wuhan University,Wuhan, P. R. China

2Engineering Research Center ofNano-Geomaterials of Ministryof Education, Department ofMaterial Science and ChemistryEngineering, China University ofGeosciences, Wuhan,P. R. China

3Shimadzu Global COE forApplication and TechnicalDevelopment, ShimadzuInternational Trading (Shanghai)Co., Ltd., Shanghai, P. R. China

Received July 18, 2011Revised August 6, 2011Accepted August 6, 2011

Abbreviations: AIBN, 2,2-azobis(2-methyl-propionitrile);AMP, amphetamine; DVB, divinylbenzene; EGDMA,ethylene glycol dimethacrylate; HCP, hypercrosslinkedpolymer; KET, ketamine; MAA, methacrylic acid; MPS, 3-(methacryloxy) propyl trimethoxysilane; MSPE, magneticSPE; VBC, vinylbenzyl chloride �These authors have contributed equally to this study.

Correspondence: Yu-Qi Feng, Key Laboratory of AnalyticalChemistry for Biology and Medicine (Ministry of Education),Department of Chemistry, Wuhan University, Wuhan 430072,P. R. ChinaE-mail: [email protected]: 186-27-68755595

& 2011 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.jss-journal.com

J. Sep. Sci. 2011, 34, 3083–3091 3083

suspensions by the application and removal of an appro-

priate magnetic field, and thus the phase separation could

be conveniently realized. Previously, we prepared a MSPE

material by distillation-precipitation polymerization of

methacrylic acid (MAA) and ethylene glycol dimethacrylate

(EGDMA) in the presence of Fe3O4/SiO2 microspheres with

the surface containing abundant reactive double bonds [18].

The resultant Fe3O4/SiO2/P(MAA-co-EGDMA) microspheres

served well as mixed-mode sorbents being capable of selec-

tively enriching sulfonamides from complex milk matrix

through hydrophobic and ion-exchange interactions. More-

over, the fast mass-transfer enabled the extraction to finish

within 0.5 min [18]. In this present study, we tried to combine

the advantages of Davankov-type HCP [7] and Fe3O4/SiO2/

P(MAA-co-EGDMA) in a single sorbent by a facile two-step

strategy. First, a precursor material was prepared by precipi-

tation polymerization of MAA, vinylbenzyl chloride (VBC),

and divinylbenzene (DVB) in the presence of Fe3O4/SiO2 with

its surface containing vinyl groups; then, hypercrosslinking

reaction (i.e. Friedel–Crafts reaction [7]) of the as-prepared

precursor was carried out to produce desired porous product

(denoted as Fe3O4/SiO2/P(MAA-co-VBC-co-DVB)). To the best

of our knowledge, this is the first time that a magnetic core-

shell HCP with weak cation-exchange character has been

exploited as the sorbent for MSPE.

Amphetamines (AMPs) are psychostimulant drugs of the

phenethylamine class, and have been abused enormously

during the last few years [21]; besides, ketamine (KET), a

minor anesthetic that is frequently used in many minor

medical operations, but it is also abused as a drug (commonly

known as ‘‘special K’’) [22]. The escalation of the abuse of

these drugs has caused serious health and social problems

[23]. In response to a growing demand for reliable evidence of

AMPs and/or KET use, qualitative and quantitative detection

of AMPs and/or KET in biological matrices has been devel-

oped using a variety of techniques such as gas chromato-

graphy (GC), gas chromatography-mass spectrometry (GC-

MS), high-performance liquid chromatography (HPLC), and

liquid chromatography-mass spectrometry (LC-MS) [23–31].

To meet the growing demands of separation ability and

reductions in organic solvent consumption, capillary elec-

trophoresis (CE) has recently become more and more popular

and has been extensively applied in the analysis of illegal

drugs in court and clinical diagnosis [32, 33]. In this study, a

simple, rapid, and efficient analytical method for the simul-

taneous monitoring of AMPs and KET in urine was devel-

oped by combining Fe3O4/SiO2/P(MAA-co-VBC-co-DVB)-

based MSPE with CE analysis.

2 Materials and methods

2.1 Chemicals and reagents

Ferric chloride (FeCl3 � 6H2O), ferrous chloride

(FeCl2 � 4H2O), oleic acid, toluene, isopropanol, methanol,

ammonia water (wt, 25%), anhydrous disodium phosphate,

dichloroethane, anhydrous ferric chloride, and 2,2-azobis(2-

methyl-propionitrile) (AIBN) were all purchased from

Sinopharm Chemical Reagent (Shanghai, China). Tetra-

ethyl-orthosilicate (TEOS) and 3-(methacryloxy) propyl

trimethoxysilane (MPS) were obtained from the Chemical

Plant of Wuhan University (Wuhan, China). MAA (98%)

was obtained from Acros (NJ, USA). VBC (90%) and DVB

(80%) were purchased from Sigma-Aldrich (St. Louis, MO,

USA). AIBN was recrystallized from ethanol, and other

reagents were of analytical grade and used directly without

further purification.

Trifluoroacetic acid (TFA, LC-MS grade) was supplied

by Fluka (Steinheim, Germany). Acetonitrile (ACN, analy-

tical grade) was purchased from J&K Chemical (Tianjin,

China). Acetone (pesticide grade) was obtained from Fisher

(Fair Lawn, NJ, USA). Purified water was obtained with a

Millipore Milli-Q apparatus (Bedford, MA, USA).

AMP, methamphetamine (MA), 3,4-methylenedioxy-

amphetamine (MDA), 3,4-methylenedioxyl-methampheta-

mine (MDMA), and KET were purchased from Beijing

Municipal Public Security Bureau (Beijing, China), Berbine

was obtained from J&K Chemical. Standard solutions at

1000 mg/mL in Milli-Q water were prepared for each

compound. The mixed stock solution of 50 mg/mL for each

compound was prepared by diluting the standard solutions

and stored at 01C in darkness. Berberine was used as the

internal standard, and its concentration was fixed at 500 mg/

mL with methanol.

2.2 Urine sample

The urines taken from the healthy volunteers were checked

to be free of any of the selected drugs and used for

calibration and validation purposes. These urines were

filtered through a 0.22-mm nylon membrane first. After

being added certain amounts of standard drug solutions and

then diluted (1:4 v/v) by phosphate buffers (PBS), the urine

samples (10 mM PBS; pH 9.2) with the drug concentrations

over a range of 0.02–10 mg/mL were obtained. After

standing several minutes to intermix completely, 1 mL

urine sample was supplied to MSPE procedure. Blank

samples were prepared in the same way as described above

but without the analyte-spiking step.

The urines of suspected addicts were provided by the

Beijing Municipal Public Security Bureau (Beijing, China),

and they were prepared without spiking step as those

of blank samples. All samples were kept at 01C before

analysis.

2.3 Preparation of Fe3O4/SiO2/P(MAA-co-VBC-co-

DVB)

The preparation procedure of Fe3O4/SiO2/P(MAA-co-VBC-

co-DVB) consisted of the following steps.

J. Sep. Sci. 2011, 34, 3083–30913084 Q. Gao et al.


First, Fe3O4/SiO2 submicrospheres with surface

containing abundant reactive double bonds (i.e. MPS-

modified Fe3O4/SiO2, where MPS represents the siloxane

MPS) were prepared according to our previously used

method [18].

Then, the precursor was prepared by a precipitation

polymerization method [6] with some modification. MPS-

modified Fe3O4/SiO2 (2.0 g), MAA (1.0 g), DVB (4.0 g), VBC

(5.0 g), AIBN (100 mg), and ACN (400 mL) were succes-

sively added into a 500-mL three-necked round-bottom flask

equipped with a condenser pipe and a stirring device. After

deoxygenated by N2 for several minutes, the mixture was

heated from room temperature to 601C, and then kept at

this temperature for 46 h. After the mixture was cooled to

room temperature, the resulting precursor particles were

separated by means of external magnetic filed, washed

several times by redispersion in ACN and magnetic collec-

tion, and overnight dried in vacuo at 601C.

At last, the hypercrosslinked reaction (i.e. Friedel–Crafts

reaction) was carried out similarly to the reported method

[7]. Briefly, anhydrous FeCl3 (5.3 g) was dissolved in

dichloromethane (150 mL). After filtration, the filtrate (i.e.

FeCl3 saturated solution) was introduced into a 200-mL

round-bottom flask. Then, as-prepared precursor particles

(5.0 g) were added. The hypercrosslinked reaction was

allowed to proceed at 801C for 18 h under nitrogen. After the

mixture was cooled to room temperature, the resulting

product Fe3O4/SiO2/P(MAA-co-VBC-co-DVB) was separated

by means of external magnetic filed, washed successively by

methanol and diluted nitric acid solution (1 M), and Soxhlet-

extracted with acetone for 12 h. After washing successively

by methanol and ether, the product was dried under

reduced pressure at 601C till constant weight.

The as-prepared materials were characterized by Quanta

200 scanning electron microscopy (SEM, FEI, Holland),

AVATAR 360 FT-IR (Thermo, USA), and PPMS-9 vibrating

sample magnetometer (QUANTOM, USA).

2.4 MSPE procedure

Typically, 1.0 mg of Fe3O4/SiO2/P(MAA-co-VBC-co-DVB)

composites was put into a 1.5-mL vial; after activating

with methanol, 1.0 mL diluted urine was added into the

vial. The mixture was vortexed for 3 min to form a

homogeneous dispersion solution, and then drug-absorbed

magnetic composites were separated rapidly from the

solution by the application of an external magnet. After

discarding the supernatant solution, the magnetic compo-

site was washed with 1 mL Milli-Q water. Then, analytes

were eluted from the magnetic composite by 50 mL of

acetone containing 2% TFA v/v under vortex (2 min). After

addition of 50 mg/mL internal solution (6 mL), the eluted

solution was evaporated to dryness under a mild N2 stream

at 351C. The residue was dissolved by 20 mL of Milli-Q water,

and the resultant solution was supplied to CE system for

analysis.

2.5 Instrumentation and analytical conditions

CE analysis was performed on a Beckman Coulter MDQ

equipped with a UV–Vis detector instrument (Fullerton,

CA, USA). Separations were carried out in a 60 cm (effective

50 cm, 75 mm id) fused silica capillary (Yongnian Fiber

Plant, Hebei, China). Before use, the capillary was rinsed at

50 psi sequentially with 1 mol/L HCl (5 min), and

water (5 min), 1 mol/L NaOH (5 min), water (5 min),

followed by conditioning with background electrolyte

(BGE) for 10 min. Between runs, the capillary was rinsed

at 50 psi sequentially with BGE (2 min) with 1 mol/L NaOH

(2 min), water (2 min), and BGE (2 min). The UV absor-

bance detection was performed at 214 nm. The CE system

was operated using normal polarity (the cathode was located

at the outlet).

Sample injections were performed hydrodynamically

for 15 s at 0.5 psi. Injection volumes were determined using

the CE expert program from Beckman Coulter. The applied

voltage was 20 kV and the capillary temperature was 251C.

Before use, the BGE, 30 mM PBS (pH 2.0) containing

15% v/v ACN, was filtered through a 0.45-mm microfilter

and degassed in an ultrasonic bath for 5 min.

3 Results and discussion

3.1 Characterization of sorbents

The morphologies and sizes of MPS-modified Fe3O4/SiO2

and Fe3O4/SiO2/P(MAA-co-VBC-co-DVB) were investigated

by SEM (Fig. 1A and B). Obviously, the two magnetic

sorbents consist of monodisperse and spherical submicro-

particles. Moreover, Fe3O4/SiO2/P(MAA-co-VBC-co-DVB)

has a mean size of about 800 nm (Fig. 1B) that is much

larger than that (�600 nm) of MPS-modified Fe3O4/SiO2

(Fig. 1A), demonstrating the successful encapsulation of

P(MAA-co-VBC-co-DVB) on the surface of Fe3O4/SiO2 and

the formation of core-shell architecture.

At room temperature, typical magnetization curves of

MPS-modified Fe3O4/SiO2 and Fe3O4/SiO2/P(MAA-co-

VBC-co-DVB) were studied (Fig. 1C). Clearly, both magnetic

curves have no magnetic hysteresis loops, suggesting they

are superparamagnetic. Moreover, Fe3O4/SiO2/P(MAA-co-

VBC-co-DVB) has a saturation magnetization (Ms) of

1.35 emu/g. On the other hand, the Ms of Fe3O4/SiO2/

P(MAA-co-VBC-co-DVB) is obviously lower than that of

MPS-modified Fe3O4/SiO2 (3.32 emu/g), implying the

existence of a significant amount of P(MAA-co-VBC-co-DVB)

in Fe3O4/SiO2/P(MAA-co-VBC-co-DVB).

Figure 2 shows the dispersion and agglomeration

process of the magnetic Fe3O4/SiO2/P(MAA-co-VBC-co-

DVB) microspheres. The homogeneously dispersed

magnetic microspheres could go straight toward the magnet

and adhered to the side wall of vials when the external

magnetic field was applied, and the aqueous solution

became clear and transparent immediately.

J. Sep. Sci. 2011, 34, 3083–3091 Sample Preparation 3085


FT-IR spectroscopy was employed to further verify the

P(MAA-co-VBC-co-DVB) formation on the surface of the

Fe3O4/SiO2 core. As shown in Fig. 3A, the MPS component

of MPS-modified Fe3O4/SiO2 can be confirmed by the peaks

at 1638 and 1710 cm�1 corresponding to the stretching

vibration of the vinyl groups and carbonyl units of MPS,

respectively [18]. The successful coating of P(MAA-co-VBC-

co-DVB) onto the surface of Fe3O4/SiO2 is also proven by the

FT-IR spectrum in Fig. 3B with the presence of these peaks

at 1610, 1510, and 1453 cm�1 assigned to the carbon–carbon

stretching vibrations in the aromatic ring [34]. In addition, it

can be found that the absorption peaks of vinyl groups of

MPS disappeared after P(MAA-co-VBC-co-DVB) formation

(Fig. 3B). The reason may be due to an in-depth reaction of

the reactive surface MPS groups with MAA, VBC, and DVB

[18].

The highly porous polymer shell of Fe3O4/SiO2/

P(MAA-co-VBC-co-DVB) was also characterized by measur-

ing the texture parameters of samples using a BET treat-

ment of N2 sorption isotherm data (Table 1). It can be found

that the surface area and pore volume of MPS-modified

Fe3O4/SiO2 are merely 10 m2/g and 0.017 cm3/g, respec-

tively. After precipitation polymerization, the as-prepared

magnetic material (i.e. precursor of Fe3O4/SiO2/P(MAA-co-

VBC-co-DVB)) has also a low surface area (35 m2/g) and a

low pore volume (0.048 cm3/g). Interestingly, the two

parameters of Fe3O4/SiO2/P(MAA-co-VBC-co-DVB) reach

up to 500 m2/g and 0.32 cm3/g after Friedel–Crafts reaction,

indicating the hypercrosslinked network of P(MAA-co-VBC-

co-DVB) has been formed on Fe3O4/SiO2 core. The highly

porous polymer shell should be beneficial to enhance

extraction capacity of magnetic sorbent.

3.2 Optimization of conditions for MSPE

To illustrate the role of hypercrosslinked P(MAA-co-VBC-co-

DVB) shell in extraction, the extraction efficiencies between

MPS-modified Fe3O4/SiO2, Fe3O4/SiO2/P(MAA-co-VBC-co-

DVB) and its precursor have been compared beforehand. As

shown in Fig. 4, it can be found that MPS-modified Fe3O4/

SiO2 shows poor adsorption capacity to drugs in urine. In

contrast, the adsorption of drugs on the precursor of Fe3O4/

SiO2/P(MAA-co-VBC-co-DVB)) is significantly improved.

Moreover, the Fe3O4/SiO2/P(MAA-co-VBC-co-DVB)) reaches

the highest extraction efficiency for all drugs. These results

indicate that both the chemical composition (including

hydrophobic and ion-exchange moieties) of P(MAA-co-VBC-

co-DVB) and its highly porous structure are the key factors

in the extraction of drugs.

In order to evaluate the applicability of Fe3O4/SiO2/

P(MAA-co-VBC-co-DVB) for the extraction of five abused

Figure 1. SEM images of MPS-modified Fe3O4/SiO2 (A) andFe3O4/SiO2/P(MAA-co-VBC-co-DVB) (B), and magnetizationcurves of samples (C).

Figure 2. The dispersion (A) and separation (B) process ofmagnetic Fe3O4/SiO2/P(MAA-co-VBC-co-DVB) microspheres.

Figure 3. FT-IR spectra of MPS-modified magnetite/SiO2 (A) andFe3O4/SiO2/P(MAA-co-VBC-co-DVB) (B).



drugs (i.e. AM, MA, MDA, MDMA, and KET) from urine

samples, the parameters that might affect the performance

of the extraction, such as sorbent amount, pH value, and

ionic strength etc., were optimized. When one parameter

was changed, the others were fixed at their optimized

values.

3.2.1 Sorbent amount

Different amounts of Fe3O4/SiO2/P(MAA-co-VBC-co-DVB)

ranging from 1.0 to 5.0 mg were applied to extract the drugs

from urine sample (Fig. 5). The results show that within the

investigated range (1.0–5.0 mg) the sorbent amount has no

significant effect on the extraction efficiency, and the

increase of sorbent amount seems to make a slight decrease

in extraction recovery. To understand this phenomenon, two

aspects (extraction and elution) should be considered

comprehensively. In our study, the extractions were carried

out under the same solution condition. Thus, the distribu-

tion coefficient (K) was constant and increasing the amount

of sorbent would enhance extraction capacity. However, in

the elution stage (K and volume of desorption solution were

constant), the more the sorbent was involved, the more it

retained analytes. The latter fact was perhaps more

prominent, which led to the result that increasing the

amount of sorbent reduced the recovery of analytes. To

achieve a good recovery, 1.0 mg magnetic sorbent was

employed in the following experiment.

3.2.2 Extraction time

In addition to facilitate the phase separation, another

obvious advantage of MSPE is its capability of rapid

enrichment. To illustrate this, the investigation on the

extraction time was carried out. The extraction time profiles

were conducted by increasing the vortex time from 0.5 to

4 min (Fig. 6). It can be found that all the analytes roughly

reach their extraction plateaus even within 0.5 min. The

adsorption reaches equilibrium rapidly since this extraction

mode (i.e. dispersion extraction) can facilitate mass transfer

by drastically increasing the interfacial area between the

solid sorbent and the sample solution. In this study, the

extraction time was set at 3 min.

3.2.3 pH value and ionic strength

The sample solution pH and ionic strength not only

influenced the degree of ionization of target analytes, but

also determined the surface charge of magnetic sorbent.

The pH optimization was performed in 10 mM phos-

phate matrix solution over the pH range of 3–10. The

P(MAA-co-VBC-co-DVB) layer contains many carboxyl

Table 1. The textural parameters of magnetic materials

Samplea) Surface

area (m2/g)

Pore volume

(cm3/g)

Pore size

(nm)

MPS-modified Fe3O4/SiO2 10 0.017 –

Precursor material 35 0.048 –

Fe3O4/SiO2/P(MAA-co-

VBC-co-DVB)

500 0.32 2.6

a) The precursor of Fe3O4/SiO2/P(MAA-co-VBC-co-DVB).

Figure 4. The CE response of drugs extracted by the Fe3O4/SiO2/P(MAA-co-VBC-co-DVB) (a), the precursor of Fe3O4/SiO2/P(MAA-co-VBC-co-DVB) (b) and the MPS-modified Fe3O4/SiO2 (c).

Figure 6. The effect of extraction time on extraction efficiency.

Figure 5. The effect of sorbent amount on extraction efficiency.



groups in its backbone and thus becomes more and more

negatively charged with the increase of pH; on the other

hand, the basic drug molecules are protonated at low pH (i.e.

pH 3–6) and appear basically as neutral state at relatively high

pH (i.e. pH 6–10) [35]. Hence, the extraction efficiency for all

these analytes increases with the increase of the sample

solution pH from 3 to 6, and almost remains unchanged at

pH 6–10 (Fig. 7). Thereby, the sample pH was fixed at 9.2 for

the following experiments. Additionally, it is worth to

mention that the extraction efficiency for all these analytes is

considerably high throughout the whole pH range (either

high or low pH), indicating that the hydrophobic interaction

should be the critical driving force in our extraction.

The effect of ionic strength was investigated by gradu-

ally increasing the phosphate concentration of matrix solu-

tion from 10 to 50 mM at pH 9.2. We found that the

extraction amount for all these analytes has showed a

declining trend with the increase of ionic strength. To

achieve good recovery and accurately control matrix pH, the

phosphate concentration was fixed at 10 mM for the

following experiments.

3.2.4 Desorption conditions

The desorption time was optimized by increasing the vortex

time from 0.5 to 4 min. It was found that 2.0 min was

enough to elute the extracted drugs from the magnetic

sorbent. Therefore, the elution time was fixed at 2 min for

the following experiments.

The desorption solvent and desorption volume were also

investigated. The results showed that 50 mL of acetone

containing 2% TFA could elute the extracted drugs

completely from the Fe3O4/SiO2/P(MAA-co-VBC-co-DVB).

3.3 Reproducibility of Fe3O4/SiO2/P(MAA-co-VBC-

co-DVB)

At present, unsatisfactory repeatability is the common

limitation of sorbent when different batches of materials

are used. In the present study, the batch-to-batch reprodu-

cibility of the Fe3O4/SiO2/P(MAA-co-VBC-co-DVB) was

investigated. Three batches of Fe3O4/SiO2/P(MAA-co-VBC-

co-DVB) prepared under the same conditions were used for

the extraction of drugs. Table 2 summarizes the response

signal and RSD values of five drugs with these three batch

sorbents. All these RSD values are lower than 11.0%,

indicating that the laboratory-made sorbent Fe3O4/SiO2/

P(MAA-co-VBC-co-DVB) has good reproducibility.

3.4 Methodological validation

Five abused drugs were analyzed on MSPE-CE under the

above optimal conditions. The calibration curves were

constructed by plotting peak area ratio of analyte to internal

standard (A/Ai) versus concentration ratio (C/Ci) with a

linear least-squares regression (R2) analysis. As summarized

in Table 3, satisfactory correlation coefficients ranging from

0.9918 to 0.9999 are obtained for all the analytes in the

concentration range of 0.02–10 mg/mL. The limit of detec-

tion (LOD) and limit of quantification (LOQ) were calculated

by the concentrations at which signal-to-noise (S/N) ratios

were equal to 3 and 10, respectively. LOD and LOQ data are

in the range of 4.0–6.0 and 13.2–19.8 mg/L, respectively

(Table 3).

The recoveries were calculated by comparing the

extracted amounts of drugs from those of the samples with

the corresponding spiking amounts on calibration curves.

The recovery was measured at three different concentra-

tions, and the spiking levels ranged from 0.05 to 5 mg/mL.

The recoveries and RSDs are summarized in Table 4; mean

recoveries are in the range of 84.0–123%.

The intra-assay precision was determined on the same

day and consisted of three series and four replicates at each

of three concentration levels. Interassay precision was

calculated with three replicates at the three fortification

levels on three different days. The numerical value used was

the RSD of triplicate measurements of the analytes. The

results obtained are summarized in Table 5. RSDs of intra-

and interday ranging from 2.0 to 10.0% and from 5.3 to

12.1% are obtained, respectively.

Figure 7. The effect of matrix pH on extraction efficiency.

Table 2. The extraction response signal and RSD value of five

illegal drugs with three different batches of Fe3O4/SiO2/

P(MAA-co-VBC-co-DVB) sorbent

Compound Response (A/Ai)

Batch 1 Batch 2 Batch 3 RSD (%)

AM 0.83 0.80 0.70 9.2

MA 1.19 1.13 1.05 6.4

MDA 0.76 0.72 0.66 6.8

MDMA 1.17 1.24 1.00 11.0

KET 2.02 2.21 1.98 5.9



Comparative study of our developed method with other

reported sample preparation procedures [27–30] was

performed and the results are summarized in Table 6. It can

be seen that the developed method is very sensitive, and it is

surely much less complicated and time consuming.

3.5 Application to real samples

To demonstrate the applicability of above-optimized MSPE-

CE method, two urine samples from two suspected addicts

were analyzed. Figure 8 shows typical electropherograms

Table 3. Calibration curves, LOD, and LOQ data of abused drugs in urine samples

Compound Linear range (mg/mL) Regression line LOD (mg/L) LOQ (mg/L)

Slope Intercept R2

AM 0.02–5 0.5840 0.0935 0.9999 4.3 14.2

MA 0.02–5 0.8295 0.1842 0.9918 4.1 13.6

MDA 0.02–5 0.7267 0.0659 0.9967 6.0 19.8

MDMA 0.02–5 1.0247 0.0217 0.9971 4.0 13.2

KET 0.02–10 1.2820 0.2384 0.9924 5.0 16.5

Table 4. Extraction recoveries and RSDs obtained by MSPE-CE from urine samples spiked with five drugs at different concentrations

Compound 0.05 mg/mL 0.5 mg/mL 5 mg/mL

Recovery (%) RSD (%) Recovery (%) RSD (%) Recovery (%) RSD (%)

AM 84.8 2.9 97.9 2.6 84.0 6.1

MA 99.4 4.8 84.0 2.0 94.4 8.3

MDA 98.6 1.9 93.6 1.7 86.3 8.2

MDMA 98.7 4.7 102 2.2 91.6 10.5

KET 88.3 2.6 115 8.0 123 7.2

Table 5. Precisions (intra- and interassay) of relative peak areas at three different concentrations by MSPE-CE from urine samples spiked

with five drugs

Compound Intraday precision (RSD, %, n 5 4) Interday precision (RSD, %, n 5 3)

0.05 mg/mL 0.5 mg/mL 5 mg/mL 0.05 mg/mL 0.5 mg/mL 5 mg/mL

AM 3.3 2.0 7.3 5.3 6.5 10.0

MA 7.8 3.7 8.3 6.5 7.7 9.2

MDA 2.1 4.3 8.5 8.0 9.4 9.5

MDMA 7.4 3.5 8.9 8.5 7.6 8.0

KET 3.9 2.9 10.0 6.5 3.8 12.1

Table 6. Comparison of the sample preparation procedures and LODs between different methods for their application in urine samples

Abused drugs Matrix Extraction Determination LODs (mg/L) Ref.

AM, MA, MDA, MDMA Urine Monolithic silica disk-packed spin column extraction HPLC 100 [27]

AM, MA, MDA, MDMA Urine Monolithic silica spin column extraction GC-MS 5–10 [28]

AM, MA, MDA, MDMA Urine Poly(MAA-EGDMA) monolith in-tube SPME HPLC 1.4–4.0 [29]

AM, MA, MDA, MDMA Urine Poly(MAA-EGDMA) monolith in-tube SPME CE 25–34 [30]

AM, MA, MDMA, MDEA, MBDB Urine Electromembrane extraction HPLC 5–10 [31]

AM, MA, MDMA Urine Varian ethyl (C2) cartridges HPLC 12–100 [32]

AM, MA, MDA, MDMA, KET Urine Fe3O4/SiO2/P(MAA-co-VBC-co-DVB)-based MSPE CE 4.0–6.0 This study



obtained by MSPE-CE of two positive urine samples. Peak

identification of target analytes in real samples was based on

the migration of standard samples, and was confirmed by

spiking known standard compounds into the samples. The

quantitative results are summarized in Table 7. In the

sample No.1, AM (0.12 mg/mL), MA (0.33 mg/mL), and

MDA (0.34 mg/mL) are detected; in the sample No.2, AM

(1.78 mg/mL), MA (15.30 mg/mL), MDMA (0.68 mg/mL), and

KET (0.15 mg/mL) are found. This result indicates that both

the selected suspects are drug abusers, and the No.2 should

be an ‘‘Ice’’ abuser.

4 Concluding remarks

In summary, a novel magnetic material Fe3O4/SiO2/

P(MAA-co-VBC-co-DVB) was synthesized, and a detailed

investigation was carried out with respect to the application

of this sorbent to the MSPE of basic illegal drugs from

complex urine samples. Due to the advantages of large

surface area (500 m2/g) and desirable chemical composition,

this adsorbent showed high extraction capacity and adequate

extraction selectivity toward illegal drugs in urine matrix. In

addition, when used in MSPE, the powdery magnetic

adsorbent can disperse in sample solution uniformly and

can be separated magnetically, thus a fast extraction (within

3 min) and a convenient phase separation was realized.

Based on these merits, a rapid, simple, and effective method

for the analysis of illegal drugs in urine samples was

established by coupling Fe3O4/SiO2/P(MAA-co-VBC-co-

DVB)-based MSPE with CE. The LODs and LOQs of illegal

drugs were in the range of 4.0–6.0 and 13.2–19.8 mg/L,

respectively, with inter- and intra-day precisions o12.1% at

three different concentrations and recoveries between 84.0

and 123%. Our results confirmed that the MSPE-CE method

based on Fe3O4/SiO2/P(MAA-co-VBC-co-DVB) could be

employed for routine analysis.

This work is partly supported by grants from the NationalNature Science Foundation (No. 91017013; No. 31070327),the Fundamental Research Funds for the Central Universities,State Key Laboratory of Coal Conversion Foundation (No. 10-11-610), the Special Fund for Basic Scientific Research ofCentral Colleges, China University of Geosciences (Wuhan)(No. CUGL090307) NSFC.

The authors have declared no conflict of interest.

5 References

[1] Berrueta, L. A., Gallo, B., Vicente, F., Chromatographia1995, 40, 474�483.

[2] Hennion, M. C., J. Chromatogr. A 1999, 856,3–54.

[3] Lemos, V. A., Teixeira, L. S., Bezerra, M. A., Costa, A. C.S., Castro, J. T., Cardoso, L. A. M., Jesus, D. S., Santos,E. S., Baliza, P. X., Santos, L. N., Appl. Spectrosc. Rev.2008, 43, 303–334.

[4] Richardson, S. D., Anal. Chem. 2009, 81, 4645–4677.

[5] Rubio, S., Perez-Bendito, D., Anal. Chem. 2009, 81,4601–4622.

[6] Bratkowska, D., Marce, R. M., Cormack, P. A., Sher-rington, D. C., Borrull, F., Fontanals, N., J. Chromatogr.A 2010, 1217, 1575–1582.

[7] Davankov, V. A., Tsyurupa, M. P., React. Polym. 1990,13, 27–42.

[8] Sychov, C. S., Ilyin, M. M., Davankov, V. A., Sochilina, K.O., J. Chromatogr. A 2004, 1030, 17–24.

[9] Martin, C. F., Stockel, E., Clowes, R., Adams, D. J.,Cooper, A. I., Pis, J. J., Rubiera, F., Pevida, C., J. Mater.Chem. 2011, 21, 5475–5483.

[10] Fontanals, N., Cortes, J., Galia, M., Marce, R. M.,Cormack, P. A., Borrull, F., Sherrington, D. C., J. Polym.Sci. A Polym. Chem. 2005, 43, 1718–1728.

[11] Fontanals, N., Marce, R. M., Cormack, P. A., Sherring-ton, D. C., Borrull, F., J. Chromatogr. A 2008, 1191,118–124.

[12] Fontanals, N., Galia, M., Cormack, P. A., Marce, R. M.,Sherrington, D. C., Borrull, F., J. Chromatogr. A 2005,1075, 51–56.

Figure 8. Electropherograms of illegal drugs in standard solu-tion at each drug concentration of 5 mg/mL (a) and in two positiveurine samples (b and c). Orders of peaks: (1) AM, (2) MA, (3)MDA, (4) MDMA, (5) KET, and I.S.

Table 7. Determination of two urine samples from two

suspected addictsa)

AM MA MDA MDMA KET

Sample No.1 (mg/mL) 0.12 0.33 0.34 n.d. n.q.

Sample No.2 (mg/mL) 1.78 15.30 n.d. 0.68 0.15

a) n.d., not detected; n.q., not quantified.



[13] Fontanals, N., Cormack, P. A., Sherrington, D. C., J.Chromatogr. A 2008, 1215, 21–29.

[14] Bratkowska, D., Fontanals, N., Borrull, F., Cormack, P.A., Sherrington, D. C., Marce, R. M., J. Chromatogr. A2010, 1217, 3238–3243.

[15] Fontanals, N., Cormack, P. A., Sherrington, D. C., Marce,R. M., Borrull, F., J. Chromatogr. A 2010, 1217,2855–2861.

[16] Mirka, S., Ivo, S., J. Magn. Magn. Mater. 1999, 194,108–112.

[17] Meng, J. R., Bu, J., Deng, C. H., Zhang, X. M., J. Chro-matogr. A 2011, 1218, 1585–1591.

[18] Gao, Q., Luo, D., Ding, J., Feng, Y. Q., J. Chromatogr. A2010, 1217, 5602–5609.

[19] Chen, L. G., Zhang, X. P., Sun, L., Xu, Y., Zeng, Q. L.,Wang, H., Xu, H. Y., Yu, A. M., Zhang, H. Q., Ding, L., J.Agric. Food Chem. 2009, 57, 10073–10080.

[20] Ding, J., Gao, Q., Luo, D., Shi, Z. G., Feng, Y. Q., J.Chromatogr. A 2010, 1217, 7351–7358.

[21] Soares, M. E., Carvalho, M., Carmo, H., Remiao, F.,Carvalho, F., Bastos, M. L., Biomed. Chromatogr. 2004,18, 125–131.

[22] Morgan, M., Loh, L., Singer, L., Moore, P. H., Anaes-thesia 1971, 26, 158–159.

[23] Meng, P. J., Fang, N., Wang, W., Liu, H. W., Chen, D. Y.,Electrophoesis 2006, 27, 3210–3217.

[24] Dowling, G., Regan, L., Tierney, J., Nangle, M., J.Chromatogr. A 2010, 1217, 6857–6866.

[25] Wang, I. T., Feng, Y. T., Chen, C. Y., J. Chromatogr. B2010, 878, 3095–3105.

[26] Meyer, M. R., Wilhelm, J., Peters, F. T., Maurer, H. H.,Anal. Bioanal. Chem. 2010, 397, 1225–1233.

[27] Namera, A., Nakamoto, A., Nishida, M., Saito, T., Kish-iyama, I., Miyazaki, S., Yahata, M., Yashiki, M., Nagao,M., J. Chromatogr. A 2008, 1208, 71–75.

[28] Nakamoto, A., Nishida, M., Saito, T., Kishiyama, I.,Miyazaki, S., Murakami, K., Nagao, M., Namura, A.,Anal. Chim. Acta 2010, 661, 42–46.

[29] Fan, Y., Feng, Y. Q., Zhang, J. T., Da, S. L., Zhang, M., J.Chromatogr. A 2005, 1074, 9–16.

[30] Seidi, S., Yamini, Y., Baheri, T., Feizbakhsh, R., J.Chromatogr. A 2011, 1218, 3958–3965.

[31] Bugamelli, F., Mandrioli, R., Cavallini, A., Baccini, C.,Conti, M., Raggi, M. A., J. Sep. Sci. 2006, 29, 2322–2329.

[32] Wei, F., Fan, Y., Zhang, M., Feng, Y. Q., Electrophoresis2005, 26, 3141–3150.

[33] Maurer, H. H., Anal. Bioanal. Chem. 2005, 381, 110–118.

[34] Quillard, S., Louarn, G., Lefrant, S., Phys. Rev. B 1994,50, 12496–12508.

[35] Sadeghipour, F., Giroud, C., Rivier, L., Veuthey, J. L., J.Chromatogr. A 1997, 761, 71–78.



magnetic solid-phase extraction using magnetic hypercrosslinked polymer for rapid determination of...

Documents