A new versatile electrochemical sensor based on poly(styrene-co-acrylic acid) PSA/SiO2/Fe3O4/AuNPs/lignin (L-MMS) modified glassy carbon electrode (GCE) was developed for the selective detection of trace trinitrotoluene (TNT) from aqueous media with high sensitivity. The fabricated magnetic microspheres were characterized by transmission electron microscopy (TEM), energy-dispersive X-ray spectroscopy (EDS), and X-ray photoelectron spectroscopy (XPS). L-MMS films were cast on the GCE surface to fabricate the TNT sensing electrode. The limit of detection (LOD) of TNT determined by the amperometric i-t curve reached 35 pM. The lignin film and well packed Fe3O4/AuNPs facilitated the pre-concentration of trace TNT on the electrode surface resulting in a fast amperometric response of 3 seconds near the detection limit. The high sensitivity and excellent catalytic activity of the modified electrode could be attributed to the lignin layer and highly packed Fe3O4/AuNPs on the electrode surface. The total recovery of TNT from tapwater and seawater matrices was 98% and 96%, respectively. The electrode film was highly stable after five repeated adsorption/desorption cycles. The new electrochemical sensing scheme provides a highly selective, sensitive and versatile assay for the in-situ detection of TNT in complex water media.

INTRODUCTION

Water security is one of the most persistent issues due to the inadequate access to clean water and sanitation in many regions worldwide especially in third world countries. Also, environmental, health, and potential security concerns, including anti-terrorism, have generated high demands for portable, rapid, sensitive, and low cost detection of threatening contaminations in soil effluents, ground and surface waters. Nitroaromatic explosives, including 2,4,6-trinitrotoluene (TNT), have contaminated soils and aquifers worldwide because of their extensive utilization in military activities and improper waste management (McQuade et al. 2000; Thomas et al. 2007; Tenhaeff et al. 2010; Zhou et al. 2011). Due to their mutagenic and toxic properties, the presence of TNT in surrounding soils, waterways, and reservoirs must be monitored at ultratrace levels to mitigate the contamination risk. Among widely employed methods for detecting trace TNT (Caygill et al. 2012), electrochemical detection methods are attractive because of the inherent redox activity of TNT and the portability of the electrochemical sensors (Blue et al. 2013; O'Mahony & Wang 2013). Surface modification of electrodes has been used to enhance the sensitivity and selectivity of electrochemical measurement including functionalized graphene (Tang et al. 2010; Lu et al. 2011), mesoporous SiO2 (Zhang et al. 2006), AgNPs (Maduraiveeran & Ramaraj 2009; Chen et al. 2012), metal nanoparticles/carbon nanotubes (Hrapovic et al. 2006) and TiO2 (Filanovsky et al. 2007). Among these nanostructures, noble nanoparticles such as AuNPs and AgNPs can provide a facile and versatile electrode surface with excellent electron transfer kinetics and a high signal to noise ratio (Li et al. 2010; Liu et al. 2011). An electrochemical sensor was established as a fast and inexpensive analytical alternative for on-site monitoring of nitroaromatic explosives in groundwater, including TNT, using a carbon fibre electrode (Fu et al. 2005). The method was attractive because it can be integrated within a microfluidic platform for simultaneous sample handling and processing. A recent report has demonstrated that lignin-modified magnetic microspheres (L-MMS) can act as sensitive surface enhanced Raman (SERS) substrate for the selective detection of TNT from soil samples below 1 × 10−8 μg/g (Mahmoud & Zourob 2013). The urge for more cost-effective methods for on-site detection of explosives has led us to develop a new electrochemical sensing material comprising (polystyrene-co-acrylic acid) PSA/SiO2 microsphere modified with highly packed Fe3O4/AuNPs and a lignin film coated on the glassy carbon electrode (GCE) for the ultra-trace detection of TNT. The modified electrode will be used for the ultrasensitive detection of TNT from water samples using voltammetric and amperometric electrochemical detection methods. Well-packed Fe3O4/AuNPs together with lignin coating is adopted as an excellent magnetic collector for pre-concentration of trace TNT in contaminated waters for subsequent electrochemical detection.

MATERIALS AND METHODS

Materials

Ferric chloride (FeCl3•6H2O), ferrous chloride (FeCl2•4H2O), sodium chloride (NaCl), triethylamine, hydrogen tetrachloroaurate(III) hydrate (99.9%), alkaline lignin, potassium persulfate (PPS), toluene, anhydrous ethanol, thioctic acid (Thc), aqueous ammonia (NH3•H2O, 28%), 2-(N-morpholino)ethanesulfonic acid (MES), N-Hydroxysuccinimide (NHS), N-(3-dimethylaminopropyl)-N'-ethylcarbodiimide (EDC), tetraethoxysilane (TEOS > 98%), and 3-aminopropyltriethoxysilane (APTES), were obtained from Sigma-Aldrich. 2,4,6-Trinitrotoluene (TNT), 2,4-dinitrotoluene (DNT), and dinitrobenzene (DNB) (1000 μg/mL) were obtained from Cerilliant Corporation, Texas.

Characterization tools

The size and morphology of the nanoparticles were examined by field emission transmission electron microscope (TEM) using a Philips CM200 microscope operating at 200 kV. Ten microliters of a dilute solution containing the nanoparticles was drop-casted on a carbon-coated copper grid (400 mesh) and dried under ambient conditions. EDX (Oxford) has software with a database of reference spectra for elemental analysis, compositional nano-analysis and mapping. The magnetic properties of the composite were investigated using a vibrating sampling magnetometer (VSM) 7,307 (Lakeshore Cryotronic, Westerville, Ohio). Electrochemical measurements were performed with a CHI workstation 760D (CH instruments, Austin, Texas). Glassy carbon electrodes (d = 2 mm) were pre-polished to a mirror finish with 1.0 and 0.3 μm alumina slurries and then rinsed with Milli-Q water, followed by electrochemical polishing in 0.5M H2SO4 by repeated cycling (0.15–1.0 V, scan rate 50 mVs−1) until a stable voltammogram was obtained.

Preparation of PSA/SiO2/Fe3O4/AuNPs microspheres (magnetic microspheres; MMS)

The PSA/SiO2 was synthesized with an average diameter of 600 nm following the detailed procedure described elsewhere (Li et al. 2011; Mahmoud & Zourob 2013). The PSA/SiO2/Fe3O4 microspheres were prepared according to the common co-precipitation method as described elsewhere with slight modification (Mahmoud & Zourob 2013). Briefly, FeCl3·6H2O (6.5 mg/mL) and FeCl2·4H2O (3.2 mg/mL) were mixed in 1% PSA/SiO2 aqueous suspension. The suspension was vigorously stirred under nitrogen for 30 minutes. prior to the dropwise addition of ammonium hydroxide (0.65 mL, 28%). After heating at 80 °C for 2 hours, the resulting black precipitate was purified by three washing cycles with water/ethanol and was recovered by magnetic separation. The obtained PSA/SiO2/Fe3O4 nanoparticles were then suspended in water to a final concentration (23 mg/mL, 4 mL) prior to the addition of HAuCl4 (3.9 mg/mL, 1.6 mL) solution containing β- cyclodextrin (2.2 mg/mL, 2 mL). Ice-cold NaBH4 (20 μL, 0.1 M) was then added dropwise to the slowly mixing solution until a stable reddish black colloid was obtained. The obtained PSA/SiO2/Fe3O4/AuNPs colloid was purified by magnetic separation and washed (with water-ethanol) three times to remove any unbound AuNPs.

Preparation of lignin modified PSA/SiO2/Fe3O4/AuNPs (L-MMS)

The PSA/SiO2/Fe3O4/AuNPs was modified with lignin using a three step method. First, carboxylic acid groups were introduced to the PSA/SiO2/Fe3O4/AuNPs by coupling Thc (1.5 mg/mL) to the AuNP surface via Au-S bonding. Second, aminated lignin was prepared by reacting lignin (100 mg) with epichlorohydrin (55 mg/g lignin) in 1 M NaOH for 2 hours at 60 °C to introduce the epoxy groups onto the lignin surface, followed by heating with 50% (w/v) NaOH, 29.4% NH4OH (5 mL/g lignin) and maintaining the pH at 12 for 2 hours at 60 °C. The suspension was then neutralized with HCl (1 mL, 1 M) and the resulting (brown) precipitate was washed with water three times.

Third, aminated lignin was re-suspended in water and conjugated onto the carboxylated magnetic nanocomposite surface using a typical carbodiimide coupling protocol. In brief, 10 mM EDC and 20 mM NHS were added to PSA/SiO2/Fe3O4/AuNPs/Thc (20 mg/mL) suspended in 0.05 M MES and 0.5 M NaCl buffer, pH 5.7 (5 mL). The suspension was stirred for 2 hours at room temperature. The activated MMS conjugate (5 mL, 50 mg) was recovered by magnetic separation and transferred to another vial containing aminated lignin (10 mg/ml) and triethylamine (pH 9, room temperature). The reaction was allowed to stir for 12 hours, the samples were then separated with a magnet and subjected to three washing cycles (water/methanol). The produced L-MMS was suspended in ethanol and was kept refrigerated for the electrochemical measurements.

Electrode modification and electrochemical detection experiments

To modify the glassy carbon electrode (GCE), a drop of a well dispersed ethanol suspension of L-MMS (20 μL, 1 mg/ml) was cast directly onto the surface of the electrode and the solvent was allowed to evaporate to form a stable film. The supporting electrolyte was purged with nitrogen gas prior to each experiment. The modified electrode was placed in an electrochemical cell containing 0.5 M NaCl solution as supporting electrolyte. The three-electrode electrochemical cell consisted of the modified GC working electrode, a Pt wire counter electrode and an Ag/AgCl (3 M NaCl) reference electrode (Bioanalytical Systems, West Lafayette, Indiana). Increasing concentrations of TNT were spiked into the electrolyte solution and the electrode was kept in contact with the stirred solution for 1 minute. prior to each run. The potential was scanned linearly between (−1.2 to 0) V, at 100 mVs−1.

TNT pre-concentration and electrochemical assay

Solutions of fresh and seawater samples spiked with TNT were prepared in triplicate to give 1 μM concentrations of standard analytical reference TNT. Tapwater samples were collected from the municipal water of Doha (Qatar), and seawater was collected from the Arabian Gulf near Doha (Qatar). Samples were refrigerated immediately after collection for further use. Tapwater and seawater measured salinity were 430 ppm and 38,000 ppm, respectively. Seawater was filtered with 0.2 μm nylon membrane prior to each experiment. Tapwater was used without further purification. TNT spiked water samples (1 mL, 1 μM) were shaken with 1 mg of L-MMS in 1.5 mL Eppendorf tube for 15 minutes at room temperature. The magnetic microspheres were then separated from the solution with a strong magnet, washed three times with deionized water, and finally re-suspended in ethanol (300 μL) in the same container. The collected magnetic microspheres were then drop-casted on per-polished GCE and allowed to dry to form a stable film for the electrochemical detection following the same protocol described above.

RESULTS AND DISCUSSION

Synthesis and characterization of L-MMS

Figure 1 describes the fabrication steps of L-MMS. The PSA/SiO2 was synthesized as described elsewhere (Li et al. 2011). The Fe3O4NPs were formed on the surface of PSA/SiO2 using the typical co-precipitation of FeCl3·6H2O and FeCl2·4H2O by ammonium hydroxide. AuNPs was grafted onto the surface of PSA/SiO2/Fe3O4 using the modified Stöber method (Stöber et al. 1968). Thioctic acid (Thc) was used to introduce terminal carboxylic acid groups onto the surface of the AuNPs. Finally, aminated lignin was covalently attached to the surface of MMS using a two-step carbodiimide coupling protocol to impart high selectivity to the substrate due to the proven affinity of lignin towards TNT (Zhang et al. 2011). An initial lignin concentration of 60 mM was used to give the optimum surface coverage on MMS (Mahmoud & Zourob 2013). The stable and uniform dispersion of Fe3O4NPs and AuNPs on PSA/SiO2 microspheres (600 nm) was then confirmed by the HRTEM as shown in Figure 2(a). AuNPs (3–7 nm in diameter) appeared to be darker than the 10–20 nm Fe3O4NPs as observed in the HRTEM micrograph (Figure 2(b)) due to the higher electron density of Au as compared with Fe3O4 (Robinson et al. 2010). The elemental composition was determined by energy-dispersive X-ray spectroscopy (EDS) as described in Figure 2(c). The metallic Fe represented by 0.75, 6.5, and 7 KeV signals confirmed the presence of Fe3O4NPs. The peaks at 2, 9.5 and 11.5 MeV represent AuNPs. The XPS pattern of the L-MMS showed well defined Fe binding energies for Fe2p3/2 at 711 eV and Fe2p1/2 at 724.4 eV due to the coexistence of Fe+2 and Fe+3 ions which are comparable to the literature values (Prakash et al. 2007). After lignin modification, a shoulder peak corresponding to N1s was observed at 400.0 eV (Figure S1, Supporting Information, available online at http://www.iwaponline.com/wst/072/399.pdf). Table S1 (online at http://www.iwaponline.com/wst/072/399.pdf) describes the elemental composition during the formation steps of the microspheres as extracted from XPS data. The Fe and Au atomic % confirmed the high packing of both Fe3O4NPs and AuNPs. After the conjugation of lignin in L-MMS, the atomic % of Au and Fe were decreased most likely due to the formation of the lignin layer. The calculated molecular composition closely matched the composition of the starting materials within 5% standard deviation. The magnetic permeability of L-MMS was evaluated by a vibrating sampling magnetometer at room temperature. The magnetic saturation value was 32.14 ± 0.09 emu g−1 (Figure S2, Supporting Information, online at http://www.iwaponline.com/wst/072/399.pdf).

Figure 1

Synthesis scheme of the PSA/SiO2/Fe3O4/AuNPs/lignin microspheres (L-MMS).

Figure 1

Synthesis scheme of the PSA/SiO2/Fe3O4/AuNPs/lignin microspheres (L-MMS).

Figure 2

(a) HRTEM of Fe3O4/AuNPs on the surface of PSA/SiO2 microsphere; (b) high magnification TEM shows the dispersion of Fe3O4/AuNPs on a single microsphere; (c) EDS spectra from the surface of a single microsphere. The scale bars are 300 nm and 10 nm for A and B, respectively.

Figure 2

(a) HRTEM of Fe3O4/AuNPs on the surface of PSA/SiO2 microsphere; (b) high magnification TEM shows the dispersion of Fe3O4/AuNPs on a single microsphere; (c) EDS spectra from the surface of a single microsphere. The scale bars are 300 nm and 10 nm for A and B, respectively.

Electrochemical characterization

Subsequently, the electrochemical detection scheme of TNT was constructed by modifying the GCE surface with L-MMS films. Surface modification is used to enhance the sensitivity and selectivity of the electrochemical measurement (Hrapovic et al. 2006; Filanovsky et al. 2007; Maduraiveeran & Ramaraj 2009; Tang et al. 2010; Lu et al. 2011; Chen et al. 2012). L-MMS was drop-casted onto the electrode surface to form a stable film. Then, the electrochemical response of the system was used to investigate the ability of L-MMS modified GCE to detect TNT in aqueous media. Figure 3(a) and 3(b) show the cyclic voltammograms (CV) of the bare and L-MMS modified GCE in 0.5 M NaCl containing increasing concentrations of TNT at a scan rate of 100 mV/s. In the presence of 50 nM TNT, the bare GCE produced weak and broad reduction peaks at −522, −700, and −824 mV corresponding to the sequential reduction of nitro groups to the corresponding amines (Zhang et al. 2006). On the other hand, the L-MMS modified GCE produced well-defined reduction peaks with about a 10-fold increase in peak current and a negative potential shift at the same TNT concentration (Figure 3(b), red curve). The reduction peaks appeared at −541, −724, and −878 mV. When TNT concentration was reduced to 5 nM, the three reduction peaks are still clearly identified with L-MMS modified GCE while they were undetectable in the case of the bare electrode. Furthermore, the electrochemical response was compared for each preparation step of L-MMS in the presence of 30 nM TNT. A significant enhancement of the reduction peaks were observed for GCE modified with L-MMS as compared with PSA/SiO2/Fe3O4/AuNPs and PSA/SiO2/Fe3O4, respectively (Figure S3, Supporting Information, online at http://www.iwaponline.com/wst/072/399.pdf). The higher sensitivity of L-MMS towards TNT could be explained by: (i) the ability of the lignin modified magnetic nanoparticles to selectively capture and pre-concentrate the target TNT molecules; (ii) the high packing, and the electromagnetic coupling between Fe3O4NPs and AuNPs which enhanced the localized electrochemical reduction of TNT. A typical amperometric (i-t) curve was used to further illustrate the detection sensitivity of the L-MMS modified GCE towards TNT. As shown in Figure 4(a), when microlitres of TNT with a concentration range from 0.01–10 nM were spiked into 0.5 NaCl buffer solutions, the oxidation current responded steeply and a steady state was reached. The inset of Figure 4(a) represents a magnification of the first three concentrations of TNT. The electrode response time varies from 5–10 seconds. This excellent sensitivity of the modified electrode could be attributed to the lignin layer and highly packed Fe3O4/AuNPs on the electrode surface. A linear relationship was observed between current response and TNT concentration in the range of 0.05–10 nM (Figure 4(b)). The linear regression equation was {(I) μA = 0.10206 + 0.0077 (C) nM}, with a sensitivity of 4.81 ± 0.07 μA and a limit of detection (LOD) of 35 pM; R2 = 0.9987. LOD was estimated from 3(Sb/m), where Sb is the standard deviation of the measurement signal for the blank and m is the slope of the analytical curve in the linear region. The response signal for 10 nM TNT was 1.8-fold higher than 5 nM. The LOD was significantly lower than values reported using other nanomaterial-modified GC electrodes in the range of 1.5–44 nM (Wang et al. 2004; Zhang et al. 2006; Fernández et al. 2014). The results provide the new hybrid material as a rapid and selective sorbent of TNT. Although a lower detection limit of TNT was obtained previously by surface enhanced Raman scattering (SERS) using L-MMS (Mahmoud & Zourob 2013), electrochemical detection provides a cheaper and more versatile assay for the on-site detection and subsequent remediation of TNT in complex water samples.

Figure 3

(a) Cathodic voltammograms for bare glassy carbon electrode (GCE); and (b) L-MMS -modified GCE in 0.5 M NaCl solution containing increasing concentrations of TNT {0 (black), 5 (blue), 30 (green), 50 (red) nM, respectively}.

Figure 3

(a) Cathodic voltammograms for bare glassy carbon electrode (GCE); and (b) L-MMS -modified GCE in 0.5 M NaCl solution containing increasing concentrations of TNT {0 (black), 5 (blue), 30 (green), 50 (red) nM, respectively}.

Figure 4

(a) The current–time (i-t) response of L-MMS/GCE measured at −541 mV in 0.5 M NaCl solution with successive additions of 0.01–10 nM TNT; (b) linear relationship between peak current and TNT concentration. Scan rate = 0.1 Vs−1, Pt counter, Ag/AgCl reference electrodes.

Figure 4

(a) The current–time (i-t) response of L-MMS/GCE measured at −541 mV in 0.5 M NaCl solution with successive additions of 0.01–10 nM TNT; (b) linear relationship between peak current and TNT concentration. Scan rate = 0.1 Vs−1, Pt counter, Ag/AgCl reference electrodes.

Electrochemical detection assay

Next, it was important to evaluate the ability of L-MMS/GCE to detect trace TNT from water samples. Deionized water was used as the reference water source and each experiment was done in triplicate. Figure 5 describes the detection and magnetic separation scheme. Typically, 1 mg of L-MMS is added to water samples spiked with different concentrations of TNT. After the extraction experiment is complete, magnetic separation is used to collect the magnetic microspheres bearing TNT. Different TNT-bearing L-MMS/GCEs were evaluated with differential pulse voltammetry (DPV) in 0.5 M NaCl buffer solution using a 3 electrode cell as described in the experimental section. DPV was used in this case because it can provide sensitive detection and diminishes the background current, especially in complex media. Figure 6(a) shows DPV curves with increasing current density corresponding to the increasing initial concentration of TNT. The results from Figure 6(b) reveal that the representative peak current of TNT at −541 mV had a linear relationship with the increase of TNT concentration in the range of 10–400 nM with a sensitivity of 30.2 ± 0.5 μA and an LOD of 1.8 nM; R2 = 0.9997. The low detection limit could be attributed to the high ability of L-MMS to absorb TNT from complex media, excellent electrochemical activity of L-MMS hybrid material, the large surface area, and the synergistic effect of Fe3O4/AuNPs. Total recovery of the initial TNT concentration was 92%, quantified by comparing DPV current signals of the recovered TNT from the water samples with those of the corresponding initial concentrations.

Figure 5

The detection scheme for TNT from water samples using L-MMS.

Figure 5

The detection scheme for TNT from water samples using L-MMS.

Figure 6

(a) Differential pulse voltammetry (DPV) for different concentrations of TNT in 0.5 M NaCl on L-MMS /GCE. TNT concentration range: 0–400 nM; (b) linear relationship between peak current ratio and TNT concentration from DPV (ΔI = I/Io, I, Io are the current intensity in presence and absence of TNT, respectively).

Figure 6

(a) Differential pulse voltammetry (DPV) for different concentrations of TNT in 0.5 M NaCl on L-MMS /GCE. TNT concentration range: 0–400 nM; (b) linear relationship between peak current ratio and TNT concentration from DPV (ΔI = I/Io, I, Io are the current intensity in presence and absence of TNT, respectively).

Selectivity and reproducibility of the electrochemical assay

Selectivity of the electrochemical method towards TNT has been tested by comparing the response of L-MMS/GCE to other structure-like analytes. An efficient modified electrode should be able to discriminate the target analyte from common interferences in ambient conditions (Blue et al. 2013); 0.1 μM of TNT and structure-like molecules (toluene, DNB, and DNT) were compared following the above described detection scheme. As described in Figure 7(a), TNT showed much higher electrochemical response as compared with toluene, DNB, and DNT. As expected, toluene showed negligible change of current intensity. The slight current enhancement was observed by the addition of DNB, and DNT. Although DNT has the most similar structure to TNT, the electrochemical sensor demonstrated far better selectivity towards TNT as compared with DNT. This is most likely due to the high affinity of the lignin layer to TNT which improved the selective capture and pre-concentration of the TNT molecules on L-MMS/GCE surface and therefore greatly enhanced the electrochemical signal of TNT as compared with structure-like nitroaromatics (Zhang et al. 2011; Mahmoud & Zourob 2013). Lignin is known to have a strong adsorption capacity for TNT, at neutral pH, through hydrogen bonding formation (Zhuo et al. 2009; Zhang et al. 2011). Previous reports indicate that CV can differentiate explosive compounds such as DNT/TNT from others such as nitrobenzene (Chen et al. 2006). Electrochemical detection of nitro-explosives has also been previously demonstrated in environmental samples using mediated electrode systems and exhibited excellent differentiation between differing nitroaromatic species (Chen et al. 2006; Guo et al. 2011). The reusability of the L-MMS composite was investigated by regenerating the microspheres after desorbing TNT with ethanol (2 ml/mg) for 30 minutes. after each run. More than 88% recovery was obtained after five adsorption/desorption cycles (Figure 7(b)).

Figure 7

(a) DPV peak current ratio of L-MMS /GCE in 0.5 M NaCl solution containing 1 μM TNT, DNT, DNB, and Toluene (ΔI = I/Io, I, Io are the current intensity in presence and absence of the analyte, respectively); (b) representative current response of L-MMS /GCE after repeated cycles of magnetic separation of TNT and reuse.

Figure 7

(a) DPV peak current ratio of L-MMS /GCE in 0.5 M NaCl solution containing 1 μM TNT, DNT, DNB, and Toluene (ΔI = I/Io, I, Io are the current intensity in presence and absence of the analyte, respectively); (b) representative current response of L-MMS /GCE after repeated cycles of magnetic separation of TNT and reuse.

TNT detection from real water samples

The electrochemical detection scheme was further investigated in real water samples. Tapwater and seawater samples were used with salinity 430 ppm and 38,000 ppm, respectively. Under optimum conditions, L-MMS was exposed to the tapwater and seawater samples containing 1 μM of TNT. The recovery assay and electrochemical test were performed as described above in Figure 5. The relative recovery was estimated from the ratio of the DPV signals obtained for 1 μM TNT in real water samples and ultrapure water. The relative recoveries for seawater and tapwater were 96% and 98%, respectively. This insignificant difference of the total recovery and the negligible effect of water salinity have proven the efficiency of the L-MMS to detect TNT in complex water matrices. Selectivity and applicability of sensors to ‘real’ environmental samples is an important factor in their deployment in field settings. Due to their high selectivity and sensitivity, the electrochemical sensor presented here forms an excellent basis for the quantification of trace TNT in real samples. This was comparable with other electrochemical sensing platforms which have been used for TNT detection, with LOD as low as 1–17 μM in sea, lake or tap water (Guo et al. 2011; Caygill et al. 2012; Caygill et al. 2013).

CONCLUSIONS

The current study presented a simple and selective electrochemical detection method for TNT in water media using lignin-modified magnetic microspheres as an active layer on the GCE. The LOD determined by amperometric i-t curve was 35 pM. An excellent catalytic activity of the modified electrode represents response time from one second or less for concentrations above 10 nM to 3 seconds near the detection limit. Also, TNT was successfully detected and removed from both tapwater and seawater with about 96% total recovery without significant impact on the water salinity. The high efficiency of the L-MMS was attributed to the ability of the lignin-modified magnetic microspheres to selectively capture and pre-concentrate the target TNT molecules which can be removed from water with a magnet. This detection scheme can be easily adopted for developing a versatile, cheap, and yet very sensitive portable sensor for TNT detection in waters. The proposed electrochemical assay can be scalable and may provide practical solutions for security and environmental applications.

ACKNOWLEDGEMENTS

This work was partially supported by the Natural Sciences and Engineering Research Council of Canada (NSERC). The authors thank Christophe Chabanier of the Centre Énergie, Matériaux et Télécommunications Canada for XPS measurement, and McGill University for HRTEM measurement.

REFERENCES

REFERENCES
Blue
R.
Vobecka
Z.
Skabara
P. J.
Uttamchandani
D.
2013
The development of sensors for volatile nitro-containing compounds as models for explosives detection
.
Sens. Actuators B: Chem.
176
(
0
),
534
542
.
Caygill
J. S.
Davis
F.
Higson
S. P. J.
2012
Current trends in explosive detection techniques
.
Talanta
88
(
0
),
14
29
.
Caygill
J. S.
Collyer
S. D.
Holmes
J. L.
Davis
F.
Higson
S. P. J.
2013
Disposable screen-printed sensors for the electrochemical detection of TNT and DNT
.
Analyst
138
(
1
),
346
352
.
Filanovsky
B.
Markovsky
B.
Bourenko
T.
Perkas
N.
Persky
R.
Gedanken
A.
Aurbach
D.
2007
Carbon electrodes modified with TiO2/metal nanoparticles and their application for the detection of trinitrotoluene
.
Adv. Funct. Mater.
17
,
1487
1492
.
Fu
X.
Benson
R. F.
Wang
J.
Fries
D.
2005
Remote underwater electrochemical sensing system for detecting explosive residues in the field
.
Sens. Actuators B: Chem.
B106
,
296
301
.
Li
H.
Sun
Z. Y.
Zhong
W. Y.
Hao
N.
Xu
D. K.
Chen
H. Y.
2010
Ultrasensitive electrochemical detection for DNA arrays based on silver nanoparticle aggregates
.
Anal. Chem.
82
,
5477
5483
.
McQuade
D. T.
Pullen
A. E.
Swager
T. M.
2000
Conjugated polymer-based sensory materials
.
Chem. Rev.
100
,
2537
2574
.
Prakash
R.
Choudhary
R. J.
Chandra
L. S. S.
Lakshmi
N.
Phase
D. M.
2007
Electrical and magnetic transport properties of Fe3O4 thin films on a GaAs (100) substrate
.
J. Phys.-Condes. Matter
19
,
486212
.
Robinson
I.
Tung
L. D.
Maenosono
S.
Walti
C.
Thanh
N. T. K.
2010
Synthesis of core-shell gold coated magnetic nanoparticles and their interaction with thiolated DNA
.
Nanoscale
2
(
12
),
2624
2630
.
Stöber
W.
Fink
A.
Bohn
E.
1968
Controlled growth of monodispersed spheres in the micron size range
.
J. Colloid Interface Sci.
26
(
1
),
62
69
.
Thomas
S. W.
Joly
G. D.
Swager
T. M.
2007
Chemical sensors based on amplifying fluorescent conjugated polymers
.
Chem. Rev.
107
,
1339
1386
.
Zhang
J.
Lin
X.
Luo
X.
Zhang
C.
Zhu
H.
2011
A modified lignin adsorbent for the removal of 2,4,6-trinitrotoluene
.
Chem. Eng. J.
168
(
3
),
1055
1063
.

Supplementary data