Microscale optical fibre sensor for BTEX monitoring in landfill leachate

Abstract

Leachates derived from landfill constitute a serious environmental concern as a variety of pollutants can leak from the landfill sites with potential risk of groundwater pollution. Benzene, toluene, ethylbenzene, p-xylene, m-xylene and o-xylene (BTEX) are pollutants widely found in landfill leachate and therefore the development of analytical methodologies which can provide a rapid and simple measurement of these compounds is of great interest. This paper reports an optical fibre (OF) sensor for environmental applications, namely for BTEX determination in landfill leachate samples.Firstly the analytical signal of the developed OF sensor was evaluated regarding some operational variables affecting the deposition of the sensitive siloxane film (type of siloxane cladding, deposition technique, number of cycles for deposition, cure temperature and experimental conditions for stabilization) in order to achieve a high sensitivity of the analytical system. Secondly, the OF based sensor was applied to landfill leachate samples for detection and quantification of BTEX. The obtained results have been compared to the ones obtained by the method based on a gas chromatography-flame ionization detector (GC-FID).The developed OF methodology showed notable analytical advantages, such as high analytical sensitivity and accuracy, good linearity and stability of the analytical signal, and detection limits in the order of a few nanograms (raging from 0.9 to 1.4 ng) for BTEX determination.

---

Introduction

Leachate from landfills is a potential environmental hazard as it can contaminate surface and groundwater. The composition of landfill leachate varies widely as a function of the landfill age and waste sources. Some specific organic compounds identified in the leachate are phthalates, pesticides, furans, polycyclic aromatic hydrocarbons (PAHs) and benzene, toluene, ethylbenzene, p-xylene, m-xylene and o-xylene (BTEX).^1–3

Different instrumental methodologies have been implemented to landfill leachate analysis, regarding the determination and characterization of the organic composition of this environmental matrix. Some of the most widely used systems rely on gas chromatography (GC) coupled to: mass spectrometry (MS), electron capture detector (ECD), flame ionization detector (FID) and photoionisation detector (PID).¹ These methodologies usually require a sample pre-treatment or analyte pre-concentration. An alternative to the classical liquid–liquid techniques, which requires the use of large quantities of extraction solvents, solid phase extraction (SPE),⁴purge and trap⁵ and solid phase microextraction (SPME)⁶ have been used for analyte extraction of landfill leachates.

The investment in analytical methodologies which provides in situ measurements is imperative for landfill pollution control, since conventional analytical methods can lead to volatile losses during sampling and transport procedures.

Optical fibre (OF) sensors have a great potential for detection and monitoring of volatile organic compounds (VOCs), including BTEX, since they show several analytical advantages, such as low cost compared with the conventional laboratory apparatus, immunity to electromagnetic fields, high sensitivity and accuracy, multiplexing capabilities, ruggedness and portability in a harsh and hazardous environment.⁷ Additionally, they can also be suitable and advantageous for remote detection, continuous measurements, in situ and real-time monitoring of BTEX at a very high safety level. In this field Silva et al.⁸ have developed an OF sensor for remote monitoring of BTEX in indoor atmospheres of industrial environments. In summary, OF sensors constitute a very important and useful tool in many technological fields, namely environmental control, with remarkable analytical advantages.^9–14

This work aims to develop the application of an OF sensor for direct measurements of BTEX in landfill leachates, with an expected analytical performance comparable to a classical methodology for BTEX measurement based on a gas chromatography-flame ionization detector (GC-FID).

Experimental

Chemicals

Preparation of the polymeric coating solutions in dichloromethane (DCM). Commercial polymerspoly[methyl(3,3,3-trifluoropropyl)siloxane] (PMTFPS), polydimethylsiloxane (PDMS) and poly[dimethylsiloxane-co-[3-(2-(2-hydroxyethoxy)ethoxy)propyl]methylsiloxane] (PDMSHEPMS) available from Aldrich (481645, DMPSV and 480320, respectively) were tested as OF sensitive claddings. The sensitive cladding was achieved by coating the OF end with a solution of each polymer at 0.01% in DCM.

Preparation of calibrants for BTEX determination. The standards injected were aromatic volatiles from an analytical grade standard mix solution of benzene, toluene, ethylbenzene, m-xylene, p-xylene and o-xylene, 100 µg mL⁻¹ of each compound in methanol (Cat no. 47504, Supelco, Spain).

Deposition of the polymeric films

The coating solution was deposited on the OF surface by three different techniques: dip-coating (DC), constant rate (CR) and spray (SP). DC is a low cost and easy to operate deposition technique where no sophisticated equipment is required. The fibre end was dipped into the coating solution for 5 s and then cured at 70 °C for 1 h. Three coating cycles were performed and after the last one the curing time was 24 h to achieve a hard porous film. The CR technique consists of immersing and withdrawing the OF end into the coating solution at constant velocity. In the CRdeposition technique the OF immersion/withdrawing occurs at 10 cm min⁻¹, providing more efficient control and easier standardization of the experimental conditions concerning the polymeric material deposition. The polymeric material was also deposited by a SP technique with the aim of further increasing the polymeric film uniformity with consequent improvements on the analytical signal and standardization of the deposition process operational conditions. In this technique the coating solution is introduced into the aerographer reservoir and sprayed onto the OF surface. A rotation system was also included in the home-made deposition chamber, allowing a uniform, constant and circular movement of the cylindrical and uncladded OF surface during the spraying process of the coating solution.

Experimental set-up of the optical fibre sensor

Fig. 1 shows the analytical apparatus and experimental lay-out used for calibration of the OF sensor and for analysis of BTEX in landfill leachate, highlighting a detailed view of the sensor head. The sensor system is generically constituted of three main components: an injection/sampling system, a miniaturized column for analyte separation, and an OF with associated electronics, constituting the detection system component. The different sensor system components were connected by three way valves (V1, V2 and V3).

Fig. 1 Experimental set-up of the OF sensor system (F = flowmeter; IC = injection cell; NR = narrowed region; OC = optical coupler; OF = optical fibre; PC = laptop for data acquisition and processing; PF = fused silica fibre coated with PDMS; TH = tape heater; TT = Teflon tube; V1, V2 and V3 = Valves).

For the sensor system calibration, the BTEX liquid mixture was injected with a micro-syringe (Hamilton) at the top of a glass cell (injection cell) with a controlled temperature of 150 °C by a coiled tape heater (TH, Cole Parmer). After vaporization the analytes were carried by a continuous stream (25 mL min⁻¹) of nitrogen (N45, Praxair, Portugal), controlled with a flowmeter (Sigma, Spain) to the column and finally to the analytical tube. The miniaturized glass column containing a fused silica fibre coated with a film of polydimethylsiloxane (PDMS) was surrounded by a second tape heater and constitutes the separation or adsorption/desorption system component. At the adsorption step the V1 is open for injection cell-column direction, while V3 is closed for the direction of column-analytical tube way. The valves position allows the adsorption of analytes at the separation component preventing the flow of analytes into the detection component, at this stage. The analyte molecules are thermally desorbed by increasing the temperature of the tape heater, which surrounds the column, by a temperature program controlled with home-made software. The tape heater starts at 25 °C with a program rate of 30 °C min⁻¹ until 75 °C and 10 °C min⁻¹ until 150 °C. The temperature program rate used was previously selected taking into account the best analytical performance (sensitivity and selectivity) of the OF sensor vs. temperature program rate or heating pattern. At this stage V2 is closed and V3 is open in the column-analytical tube direction, allowing the flow of analytes to the detection component of the analytical device.

The detection system component includes:

(a) the sensor head which constitutes a 15 mm long uncladded optical fibre section, onto which a nanometric polymeric film was deposited on the end. The sensor head preparation consisted of: firstly, removing the silicone passive OF cladding on a few millimetres length using a buffer stripper, and then, on the immersion of this OF section in a solvent (dichloromethane); secondly, adjusting the uncladded OF section to 15 mm using a Cleaver V6 (from Future Instrument) precision fibre cleaver and checking the cleanliness of the OF surface by optical microscopy; and finally, on covering the OF with the sensitive film of siloxane. This sensitized OF section was introduced through a Teflon plug inside the 7.7 cm long glass tube with a narrowed region of a few millimetres (analytical tube) centred with the sensor head. The choice of the sensor head length was based on the high analytical signal obtained for a sensor head length of 15 mm, during a study of the sensor behaviour for an analytical system constituting of sensor heads with different lengths (5, 15, 20 and 30 mm). A monomode optical fibre pigtail, core and cladding diameters of 9 and 125 µm, of a directional 50:50 Y coupler were used;

(b) a laser diode (AC/DC power or 9 V battery power, 1 mW output) set at a wavelength of 1550 nm (Oz Optics, Canada) for interrogation signal generation; and

(c) a photodiode (110/220V AC 50/60 Hz Universal power supply or battery power) also from Oz Optics for measurements of the reflected optical power changes.

The optical power guided through the fibre is reflected at the fibre/polymer interface and the exposure of the polymeric film to organic vapours inside the analytical tube causes a change in its refractive index leading to a variation of the light power.

For calibration of the OF device, five repeated measurements were performed for each different amount (40, 50, 60, 70 and 80 ng) of the BTEX standard mixture.

For landfill leachate sampling, the injection cell was replaced by a pump which provides a continuous stream of sample through the separation system component in which adsorption and desorption phases took place, being the experimental lay-out similar to the one described for the calibration step. Thus the analyte molecules are firstly adsorbed onto the PDMS fibre and then thermally adsorbed by increasing the tape heater temperature at the glass column. At the same time the V3 is open in the column-analytical tube direction allowing analyte detection.

For the assessment of the influence of some operational conditions regarding sensitive film deposition in the analytical signal, the experimental set-up was simplified by directly connecting the injection cell to the analytical tube with the main propose of reducing the total time of analysis. Since these studies were carried out using toluene as a proxy of aromatic compound, no selective detection was required.

GC-FID methodology

In order to compare the analytical performance of the developed methodology based on OF detection for actual samples of landfill leachates analysis, the concentrations of BTEX were also measured by GC-FID methodology using the purge and trap extraction procedure, according to US EPA method 5030B.¹⁵ The GC-FID methodology was implemented in a Gow-Mac Series 600 with a capillary column (fused silica-Supelcowax, 30 m × 0.32 mm ID × 1.0 µm, 100% PEG, Cat no. 24211, Supelco, Spain). The operation conditions of GC-FID were based on method 1501 by NIOSH;¹⁶ the column temperature starts at 40 °C (10 min) and rises up to 230 °C with a program rate of 10 °C min⁻¹; the temperature of the injector was kept at 250 °C during the analysis, and the FID temperature was set at 300 °C, the carrier gas was helium at constant flow of 2.6 mL min⁻¹.

Results and discussion

In order to achieve an analytical signal as high as possible, some operational conditions such as, the type of siloxane cladding, deposition technique, number of cycles for deposition, cure temperature and experimental conditions for stabilization of the polymeric film, have been assessed prior to the BTEX analysis in landfill leachate.

Effects of different siloxane polymeric coatings

Siloxane polymers exhibit some particular optical and chemical properties such as, high gas permeability, high thermal and oxidative stabilities and hydrophobic behaviour, which make them suitable for sensitive cladding in OF sensor applications.^17,18 Besides, this class of polymers allows extremely thin films to be obtained which are an excellent chemical interface for organic compound detection due to their high sensitivity, fast and reversible analytical response and adequate sensing performance at room temperature.

In this section the OF sensor was tested for three different sensitive coatings: PMTFPS, PDMSHEPMS and PDMS; since different transducing materials promote different sensitive responses to organic compounds. Fig. 2a shows the sensor response obtained for toluene, using an OF sensitized with the three different polymeric coatings. The analytical signal amplitude decreases in the following sequence: PMTFPS > PDMSHEPMS > PDMS. The detection limits were estimated¹⁹ as 0.8 ng for a sensor with PMTFPS, and 0.9 ng for a sensor coated with either PDMSHEPMS or PDMS film.

Fig. 2 Analytical signal (a) and SEM images (b) of an OF surface coated with polymeric films of PMTFPS, PDMSHEPMS and PDMS (OF = coated optical fibre; SS = SEM stage).

The different polymers used were characterized by scanning electron microscopy (SEM) regarding their film's morphology. The SEM analyses were performed in an Aspex SEM 1200024. The results obtained by SEM analysis, as shown in Fig. 2b, suggest that the OF surface shows different morphological characteristics for the three different polymers under study. The most homogeneous film was obtained for PMTFPS showing a very uniform and regular texture. The PDMSHEPMS film shows a more irregular structure with small beads (dark areas with spherical shape) distributed by the whole covering area. The PDMS film shows some large beads, especially on the outlying OF surface.

The highest analytical signal amplitude of the OF sensor coated with PMTFPS, could be due to the polarizability of this polymeric material. The fluorine atoms in PMTFPS promote polarizability and consequently enhance interactions with polarizable compounds such as toluene, resulting in a higher change of optical power and therefore higher analytical signals. This increase of the analytical signal and sensitivity, due to the use of transducer materials with a high potential for polarizability, has already been reported by Abdelmalek et al.²⁰ during a comparison of the performance of OF sensors sensitized with porous silica films and phenyl-modified porous silica films.

Comparison of techniques for deposition of polymeric films

The analytical signal achieved for toluene detection with an OF sensor coated with a polymeric film obtained by SP technique is considerably higher than the one obtained using either CR or DC as deposition techniques, as it can be inferred from Fig. 3. This could be due to the highly uniform morphology of the polymeric film produced by SP, which avoids any light scattering phenomena from structural inhomogeneities occurring at the films produced by the other two deposition techniques (DC and CR). The SEM images in Fig. 3 of the OF surfaces coated with a PMTFPS film deposited by DC, CR and SP techniques show that surface texture after polymeric film deposition depends on the coating technique. The OF surface covered by DC showed an irregular texture with some ruggedness. The OF sensitized surface obtained by CR or SP was clearly more homogeneous and smooth. However the CR-OF surface showed a fractured cover structure which may be due to an irregular deposition of the polymeric film. The OF surface coated by SP, yielded a smooth and homogeneous coating, which covered the OF surface without any cracks or holes; therefore the SP technique becomes potentially the most adequate for OF sensor applications.

Fig. 3 Sensor response after injection of 0.05 μg of toluene using an OF coated with a polymeric film deposited by different coating techniques (dip-coating, constant rate, spray) and SEM images of an OF coated with a polymeric film deposited by the three coating techniques (OF = coated optical fibre; SS = SEM stage).

In summary, from the SEM images shown in Fig. 3 it is possible to infer that the film obtained with the SP technique would lead to an uniform distribution of the polymeric material and consequently to the best analytical signal of the OF sensor.

Effect of the number of cycles for deposition of the polymeric film

The control of the sensitive film thickness constitutes a very important factor regarding analytical performance of the sensor since it affects the response time considerably, and the sensitivity of the sensing device. The effect of this factor on the analytical signal of the developed sensor for toluene detection was assessed by using an OF sensitized with PMTFPS (at a coating concentration of 0.01% in DCM) obtained with 1 and 3 cycles of deposition of polymeric material. The different polymeric film thicknesses achieved were estimated by Rutherford backscattering spectrometry (RBS). The main measures obtained for the two sensor systems evaluated are systematized in Table 1.

Table 1 Parameters of the sensor system using an OF sensitized with PMTFPS performing 1 and 3 cycles of deposition of the polymeric film

	1 cycle	3 cycles
Film thickness (nm)	2	8
Sensitivity (dB µg^−1)	891	305
Analytical signal for 0.06 µg of toluene (dB) ± standard deviation of five replicates	28.6 ± 0.2	16.6 ± 0.4
Response time (s)	18	22
Recovery time (s)	5	8
Detection limit (ng)	0.8	1.4

---

The analytical system constituted an OF coated with a polymeric film deposited in 1 coating cycle, shows a higher sensitivity (891 dB µg⁻¹) and lower detection limit (0.8 ng) compared to the values obtained for an analytical system with a polymeric film deposited in 3 coating cycles (sensitivity and detection limit of 305 dB µg⁻¹ and 1.4 ng, respectively). Moreover, the OF sensor coated with a 2 nm polymeric film (1 coating cycle) shows a faster response, and recovery time than the OF sensitized with 3 coating cycles. The results could be explained taking into account how the sensitive film thickness affects the analyte molecules adsorption and interaction which take place at the organic vapour/sensitive film interface.²¹

The sensing mechanism of the OF sensor is based on the sensitive polymeric film deposited onto the uncladded section of the OF end surface, which suffers a reversible change in its optical properties (refractive index) when exposed to the analyte molecules, leading to a detectable modulation of the light power guided through the OF. Thus the exposure of the polymeric film to the organic vapour inside the analytical tube causes a change in the reflected light power, which was measured as the analytical signal. The reflected optical power is determined by both Fresnel's law²¹ at the interface fibre core/sensing layer, and evanescent wave^20,22 of the light on the 15 mm sensitized fibre.

Cure temperature and operational conditions for polymeric film stabilization

Improvements in the analytical signal of an OF polymer based sensor, when the film cure temperature increases from 25 to 70 °C, is already reported elsewhere.¹⁷ This study reports the sensor behaviour for more extreme polymeric film cure temperatures, namely 140 °C, besides laboratory (25 °C) and moderated (70 °C) cure temperatures. In this set of experiments the OF analytical system was characterized in terms of the produced optical signal and sensitivity, and the OF surface morphology by SEM observation. The highest homogeneity of the coating films was achieved for a polymeric film cured at 70 °C. This analytical system, besides showing the highest analytical sensitivity, produced the lowest detection limit and analytical error, as shown in Table 2.

Table 2 Parameters of the sensor system during toluene detection using an OF coated with a polymeric film cured at 25 °C, 70 °C and 140 °C

	T = 25 °C	T = 70 °C	T = 140 °C
Determination coefficient (r^2)	0.9968 (p < 7.72 × 10^−5)	0.9998 (p < 1.33 × 10^−6)	0.9937 (p < 2.11 × 10^−4)
Standard error	0.787	0.238	1.28
Sensitivity (dB µg^−1)	760	891	884
Linear working range (µg)	0.06–0.10	0.04–0.08	0.08–0.12
Detection limit (ng)	3.1	0.8	4.3

---

The OF system constituted by a polymeric film cured at 140 °C shows a higher detection limit and standard error, thus it is not adequate for OF sensor purposes. Moreover the optical patch cord showed structural damages maybe due to partial charring after exposure to 140 °C during 24 h, which could lead to light scattering and decrease the patch cord resistance. On the other hand, low temperatures (25 °C) for polymeric film curing, are also not appropriate since this analytical system shows a poor overall analytical performance when compared to a system with a polymeric film cured at moderated temperatures.

A set of experiments were performed in order to establish the operational conditions for stabilization of the polymeric film, which allows the best analytical signal. Firstly, an OF coated with a polymeric film (PMTFPS at 0.01% in DCM) deposited by SP, cured at 70 °C during 24 h was tested immediately for toluene detection, without any stabilization phase. Secondly, an OF also coated with a polymeric film and cured at 70 °C during 24 h was stabilized for another 24 h at room temperature; in this experiment the stabilization phase was performed without gas flow. Thirdly, an OF was prepared in the above mentioned operational conditions for deposition, cure and stabilization, but with a constant flow (14 mL min⁻¹) of nitrogen during the stabilization phase. Fig. 4 shows the OF sensor baseline (optical signal recorded without organic vapours injection) and the SEM images collected for the OF surface produced for each of the above described experiments. It is possible to conclude that smooth polymeric films with high homogeneity—achieved in this experiment by increasing the time of the stabilization phase and introducing a constant flow of an inert gas in this phase—reduces the sensor system noise. In fact the sensor baseline becomes more stable, that is, with no significant optical power changes, as the stabilization conditions have been appropriately chosen.

Fig. 4 Optical signal and SEM images obtained for an OF surface coated with a polymeric film stabilized at different operational conditions: (a) polymeric film without stabilization; (b) polymeric film stabilized during 24 h without gas flow; (c) polymeric film stabilized during 24 h with a constant gas flow (OF = coated optical fibre; SS = SEM stage).

The SEM images of the OF surface suggest that: (a) the fibre stabilization time is an important parameter concerning polymeric film uniformity and; (b) the stabilization stage with a nitrogen flow provides a smoother layer at the OF surface with a uniform and regular texture. The OF surface coated with a non stabilized polymeric film showed regions with different topographical contours and spot heights (Fig. 4a), which become less apparent in an OF with a polymeric film stabilized during 24 h without and with gas flow (Fig. 4b, and Fig. 4c, respectively).

Analytical performance of the OF sensor for BTEX detection

The sensor performance was also evaluated for detection and quantification of BTEX. This set of experiments was performed following the operational conditions found earlier to be fit for the purpose of obtaining adequate analytical performance of the developed OF sensor regarding BTEX detection: a sensor head with 15 mm length sensitized with a polymeric film of PMTFPS at 0.01% in DCM, deposited in 1 coating cycle by the SP technique, the polymeric film cured at 70 °C and stabilized during 24 h with a constant flow of nitrogen.

The results displayed in Fig. 5 show an analytical system with high sensitivity and linearity (Fig. 5a) in an adequate range of concentrations of BTEX compounds.

Fig. 5 Sensor response obtained during BTEX analysis: (a) calibration curves obtained with the OF sensor for injections of BTEX between 0.04 and 0.08 µg; (b) optical power decrease obtained for a BTEX mixture of 0.05 µg; (c) sensor response (mean and standard deviation of five repeated evaluations) for 0.06 µg of BTEX during four weeks.

The total analytical time was found to be 9 min for the seven completely separated compounds detected, as shown in Fig. 5b. The OF system provides a faster BTEX analysis than the one performed by the GC-FID method, which requires an analytical time of around 30 min for each single run. Besides, the GC-FID method requires a large time to perform the sample pre-treatment, as recommended in method 5030B¹⁵ by the US EPA.

The developed OF sensor shows different sensitivity for the six aromatic hydrocarbons analysed, increasing in the following order: 751 dB µg⁻¹ for benzene < 892 dB µg⁻¹ for toluene < 933 dB µg⁻¹ for ethylbenzene < 941 dB µg⁻¹ for p-xylene < 949 dB µg⁻¹ for m-xylene < 964 dB µg⁻¹ for o-xylene. The analytical signal depends on the chemical interactions which can take place between the sensitive cladding (polymeric film) and the analyte molecules, modifying the refractive index of the polymeric film and consequently the reflected optical power. The intensity of the modulated reflected signal is proportional to the amount of analyte present at the analytical tube and also depends on the interaction between different physical and chemical factors, such as the waveguide, sensitive cladding and analyte properties, and also the molecular interactions that could occur between the analyte molecules and the sensitive OF cladding. In this way, the different analyte properties, such as the boiling temperature (80.1 °C for benzene < 110.6 °C for toluene < 136.2 °C for ethylbenzene < 138.4 °C for p-xylene < 139.1 °C for m-xylene < 144.4 °C for o-xylene) and vapour pressure (95.2 for benzene > 28.4 for toluene > 9.6 for ethylbenzene > 8.8 for p-xylene > 8.4 for m-xylene > 6.7 for o-xylene, in mmHg at 25 °C) are sufficiently different to explain the significant differences of sensitivities observed for each organic compound analyzed. This observation is in good agreement with the behaviour of the OF polymer based sensors reported by Silva et al.^8,17 and Abdelmalek et al.²⁰ for different VOCs detection.

The estimated detection limit values obtained with the reported sensor for benzene, toluene, ethylbenzene, p-xylene, m-xylene and o-xylene were 1.4, 0.8, 1.0, 0.9, 1.0 and 0.9 ng, respectively. These values highlight the high potential of the OF sensor for BTEX monitoring at trace levels.

In order to evaluate the stability of the developed OF methodology an ANOVA was applied, using SigmaStat 3.0²³ to all data (displayed in Fig. 5c) on benzene, toluene, ethylbenzene, m-xylene, p-xylene and o-xylene, and it showed that there is not a statistically significant difference between weeks (p = 0.267, 0.448, 0.966, 0.658, 0.753 and 0.938 for benzene, toluene, ethylbenzene, p-xylene, m-xylene and o-xylene, respectively), that is, the difference in the mean values among the different weeks is not great enough to exclude the possibility that the difference is just due to random variability in BTEX analysis.

BTEX monitoring in landfill leachate

Fig. 6 shows the results obtained by the OF and the GC-FID method when applied to five actual samples of landfill leachate. Quantification of samples was performed by linear regression of the calibration data, and every assay was repeated five times.

Fig. 6 Results obtained for five actual samples of landfill leachate analyzed by the OF sensor and the GC-FID method.

The comparison of the means of the five actual samples analyzed by both the OF sensor and the GC-FID method, shows that there is no statistically significant difference between the results obtained for each sample (p = 0.335, 0.880, 0.810, 0.660, 0.255 and 0.489 for benzene, toluene, ethylbenzene, p-xylene, m-xylene and o-xylene, respectively) by the two analytical systems.

The total BTEX concentration is below 2 µg L⁻¹ for the five analyzed actual samples. The landfill leachate samples were collected in different places at the same Portuguese landfill cell site and therefore it is understandable that there are similar average values of BTEX in the different analyzed samples.

The aromatic compound concentrations found in the analyzed landfill leachate are included in the lower range of BTEX concentrations reported in other studies regarding landfill leachate analysis and chemical characterization.^24,25 According to the studies referred to above, the BTEX concentration in landfill leachate can range from centesimal values to thousands of µg L⁻¹, depending on the waste disposal practices, management and landfill technologies, general waste composition, and waste age.²⁵

In order to test the developed OF analyzer for interferences that could be present in the leachate matrix from landfill, the effects of naphthalene and n-propylbenzene, used as an example of interference compounds, on the analytical signal were also evaluated. The compounds tested produce peaks, but in different analytical time windows to the compounds of interest, and therefore did not exhibit any interference in the analytical signal.

Conclusions

The analytical signal of the developed OF sensor for BTEX monitoring was optimized regarding some polymeric film parameters, such as, type of polymeric film, deposition technique, number of cycles for deposition, cure temperature and experimental conditions for stabilization of the polymeric film. The higher analytical signal obtained for the sensor system was achieved using a sensor head consisting of an OF sensitized with a thin film of PMTFPS at 0.01% in DCM (over a 15 mm length), deposited by SP using a constant flow of N₂ for stabilization during 24 h using a 1 coating cycle and cured 70 °C. The sensor system shows a high sensitivity concerning BTEX detection with an estimated detection limit in the order of a few nanograms.

The developed methodology showed an analytical performance for BTEX detection in landfill leachates comparable to the GC-FID method. Additionally, the ease of assembling and compact design of the optical device confers high potential for continuous and in situ measuring of BTEX compounds to the developed analytical system.

References

1. L. Jiménez, R. Alzaga and J. M. Bayona, Int. J. Environ. Anal. Chem., 2002, 82, 415–430.
2. E. Benfenati, G. Facchini, P. Pierucci and R. Fanelli, TrAC, Trends Anal. Chem., 1996, 15, 305–310.
3. E. Klimiuk and D. Kulikowska, Pol. J. Environ. Stud., 2004, 13, 525–532.
4. A. Baun, S. D. Jensen, P. L. Bjerg, T. H. Christensen and N. Nyholm, Environ. Sci. Technol., 2000, 34, 1647–1652.
5. A. Yashura, H. Shiraishi, M. Nishikawa, T. Yamamoto, T. Uehiro, O. Nakasugi, T. Okumura, K. Kenmotsu, H. Fukui, M. Nagase, Y. Ono, Y. Kawagoshi, K. Baba and Y. Noma, J. Chromatogr., A, 1997, 774, 321–332.
6. A. Ribeiro, M. H. Neves, M. F. Almeida, A. Alves and L. Santos, J. Chromatogr., A, 2002, 975, 267–274.
7. M. Campbell, Sensor Systems for Environmental Monitoring, Blackie Academic & Professional, London, 1997, vol. 1.
8. L. Silva, T. Rocha-Santos and A. C. Duarte, Talanta, 2009, 78, 548–552.
9. B. Mizaikoff, R. Göbel, R. Krska, K. Taga, R. Kellner, M. Tacke and A. Katzir, Sens. Actuators, B, 1995, 29, 58–63.
10. R. Jakusch, B. Mizaikoff, R. Kellner and A. Katzir, Sens. Actuators, B, 1997, 38, 83–87.
11. R. P. McCue, J. E. Walsh, F. Walsh and F. Regan, Sens. Actuators, B, 2006, 114, 438–444.
12. C. Young, S.-S. Kim, Y. Luzinova, M. Weida, D. Arnone, E. Takeuchi, T. Day and B. Mizaikoff, Sens. Actuators, B, 2009, 140, 24–28.
13. C. Elosua, C. Bariain, I. R. Matias, F. J. Arregui, E. Vergara and M. Laguna, Sensors, 2009, 137, 139–146.
14. L. Silva, T. Rocha-Santos and A. C. Duarte, Global NEST J., 2008, 10, 217–225.
15. U.S. EPA, Method 5030B, Purge and trap for aqueous samples, 1996; http://www.epa.gov/waste/hazard/test%20methods/sw846/pdfs/5030b.pdf, accessed on 27/04/2009.
16. NIOSH Manual of Analytical Methods (NMAM), Hydrocarbons aromatic: Method 1501, DHHS (NIOSH) Publication, 4th ed., Issue 3, 2003.
17. L. Silva, T. Rocha-Santos and A. C. Duarte, Sens. Actuators, B, 2008, 132, 280–289.
18. B. Zimmermann, J. Bürck and H.-J. Ache, Sens. Actuators, B, 1997, 41, 45–54.
19. J. N. Miller, J. C. Miller, Statistics and Chemometrics for Analytical Chemistry, Pearson Prentice Hall, New York, 5th ed., 2005.
20. F. Abdelmalek, J. M. Chevelon, M. Lacroix, N. Jaffrezic-Renault and V. Matejec, Sens. Actuators, B, 1999, 56, 234–242.
21. C. Elosua, C. Bariain, I. R. Matias, A. Rodriguez, E. Colacio, A. Salinas-Castillo, A. Segura-Carretero and A. Fernandez-Guttiérrez, Sensors, 2008, 8, 847–859.
22. M. Benounis, N. Jaffrezic-Renault, J.-P. Dutasta, K. Cherif and A. Abdelghani, Sens. Actuators, B, 2005, 107, 32–39.
23. SigmaStat 3.0 (Statistic Software for Windows), Jandel Scientific, Erkrath, Germany, 1994.
24. P. Kjeldsen, P. L. Bjerg, P. Winther, K. Rugge, J. K. Pedersen, B. Skov, A. Foverskov, T. H. Christensen, Proceedings of the Prague Conference, Groundwater Quality: Remediation and Protection, 1995.
25. T. H. Christensen, P. Kjeldsen, P. L. Bjerg, D. L. Jensen, J. B. Christensen, A. Baun, H.-J. Albrechtsen and G. Heron, Appl. Geochem., 2001, 16, 659–718.

---