Sequential Flow Injection Analysis System On-Line Coupled to High

Speciation of Inorganic- and Methyl-Mercury in Biological Matrixes by Electrochemical ..... There and back again: The tale of 2 asteroid sample-return...
2 downloads 0 Views 300KB Size
Anal. Chem. 2006, 78, 2494-2499

Sequential Flow Injection Analysis System On-Line Coupled to High Intensity Focused Ultrasound: Green Methodology for Trace Analysis Applications As Demonstrated for the Determination of Inorganic and Total Mercury in Waters and Urine by Cold Vapor Atomic Absorption Spectrometry C. Fernandez,† Antonio C. L. Conceic¸ a˜o,‡ R. Rial-Otero,§ C. Vaz,† and J. L. Capelo*,§

Laborato´ rio de Ana´ lises do IST and Centro de Quı´mica Estrutural, Instituto Superior Te´ cnico, Av. Rovisco Pais, 1049-001, Lisboa, Portugal, and REQUIMTE, Departamento de Quı´mica, Faculdade de Cieˆ ncias e Tecnologia, Universidade Nova de Lisboa, 2829-516, Monte de Caparica, Portugal

A new concept is presented for green analytical applications based on coupling on-line high-intensity focused ultrasound (HIFU) with a sequential injection/flow injection analysis (SIA/FIA) system. The potential of the SIA/ HIFU/FIA scheme is demonstrated by taking mercury as a model analyte. Using inorganic mercury, methylmercury, phenylmercury, and diphenylmercury, the usefulness of the proposed methodology for the determination of inorganic and total mercury in waters and urine was demonstrated. The procedure requires low sample volumes and reagents and can be further applied to all chemical reactions involving HIFU. The inherent advantages of SIA, FIA, and HIFU applications in terms of high throughput, automation, low reagent consumption, and green chemistry are accomplished together for the first time in the present work. Sample pretreatments are, generally, carried out manually with time-consuming approaches. In addition, the handling operation in the laboratory might cause sample contamination and also give rise to large sample/reagent consumption. The aforementioned drawbacks can be partially or totally reduced using, when possible, on-line sample pretreatments.1-3 Sequential injection (SIA) and flow injection (FIA) analysis are well-established and powerful analytical tools, and presently, wet chemical analysis can be fully automated.4 * Corresponding author. E-mail: [email protected]. Phone: + 351 212 949 649. Fax: + 351 212 948 550. † Laborato´rio de Ana´lises do IST, Instituto Superior Te´cnico. ‡ Centro de Quı´mica Estrutural, Instituto Superior Te´cnico. § Universidade Nova de Lisboa. (1) Wang J.; Hansen E. H.; Miro, M. Anal. Chim. Acta 2003, 499, 139-147. (2) Chomchoei, R.; Miro, M.; Hansen, E. H.; Shiowatana, J. Anal. Chem. 2005, 77, 2720-2726. (3) Hansen, E. H. Talanta 2004, 64, 1076-1083. (4) Wang, J. H.; Hansen, E. H. TrAC-Trends Anal. Chem. 2005, 24, 1-8.

2494 Analytical Chemistry, Vol. 78, No. 8, April 15, 2006

The use of high-intensity focused ultrasound (HIFU) in the analytical laboratory has increased during the past decade as a consequence of its properties and simplicity of handling.5,6 HIFU can be used in a wide variety of approaches for sample pretreatment, such as the following: (i) solid-liquid extraction of organic species,7 (ii) lowering quantification limits in electroanalytical applications,8 (iii) total or partial solid-liquid elemental extraction,9 (iv) shortening of solid-liquid sequential extraction schemes,10 and (v) elemental speciation.11-13 Recently, the ultrafast digestion of proteins for proteome research has also been reported.14 On-line ultrasonic applications have been rarely used in continuous systems despite their advantages of automation and low sample and reagent consumption.5,15 In addition, the attempts made to date for on-line sonication are, generally, related to ultrasonic baths in closed or open systems.15 Direct on-line application of an ultrasonic probe in a closed system accomplished with a SIA/FIA system for analytical applications, to the best of our knowledge, has never been attempted before.5,15 The reasons may lay in the inherent difficulties to setting up an ultrasonic probe in a closed system due to probe vibrations, as explained below. (5) Capelo, J. L.; Mota, A. M. Curr. Anal. Chem. 2005, 2, 193-201. (6) Capelo, J. L.; Ximenez-Embun, P.; Madrid-Albarra´n, Y.; Camara, C. TrACTrends Anal. Chem. 2004, 23, 331-340. (7) Capelo, J. L.; Galesio, M. M.; Felisberto, G. M.; Vaz C.; Costa-Pessoa, J. Talanta 2005, 66, 1272-1280. (8) Davis, J.; Vaughan, D. H.; Stirling, D.; Nei, L.; Compton, R. G. Talanta 2002, 57, 1045-1051. (9) Capelo, J. L.; Maduro, C.; Vilhena, C. Ultrason. Sonochem. 2005, 12, 225232. (10) Filgueiras, A. V.; Lavilla, I.; Bendicho, C. Anal. Bioanal. Chem. 2002, 374, 103-108. (11) Capelo, J. L.; Maduro, C.; Mota, A. M. J. Anal. At. Spectrom. 2004, 19, 414-416. (12) Capelo, J. L.; Ximenez-Embun, P.; Madrid-Albarra´n, Y.; Camara, C. Anal. Chem. 2004, 76, 233-237 (13) Capelo J. L.; Lavilla I.; Bendicho, C. Anal. Chem. 2000, 72, 4979-4984. (14) Lopez-Ferrer, D.; Capelo, J. L.; Vazquez, J. J. Proteome Res. 2005, 4, 15691574. (15) Luque-Garcia, J. L.; Luque de Castro, M. D. TrAC-Trends Anal. Chem. 2003, 22, 41-47. 10.1021/ac0516685 CCC: $33.50

© 2006 American Chemical Society Published on Web 03/15/2006

Figure 1. Dimensions of the chemical reactor for on-line high-intensity focused ultrasound applications.

EXPERIMENTAL SECTION Apparatus. A Branson Sonifier 150 ultrasonic cell disruptorhomogenizer (100 W, 20 kHz, Branson Ultrasonics Corp.) equipped with a 6-mm titanium microtip was used. Ultrasonic energy irradiation was fixed at any desired level using a power setting in the 50% range with the 6-mm microtip. The Sonifier 150 had a digital LCD display, which provides a continuous readout of the watts delivered to the end of the probe (range 8-10 W in this work. Warning: ultrasonication must be performed with care; manufactured safety conditions must be followed; experiments must be carried out in a fume cupboard; and hearing protection must be used). The SIA/HIFU/FIA system used for cold vapor generation consisted of a six-position, type 50, Teflon rotary valve model 5011 from Rheodyne (Cotati, CA); a closed reactor as described in Figure 1 was attached to a Branson Sonifier 150 ultrasonic cell disruptor-homogenizer, as depicted in Supporting Information, Figure 1, and consisted of a four-channel Gilson (Villiers le Bel, France) Minipuls 2 peristaltic pump, a four-channel Ismatec (Glattbrugg, Switzerland) programmable peristaltic pump model Reglo Digital MS-4/12, a Perkin-Elmer (Uberlingen, Germany) membrane gas-liquid separator, a six-port injection valve (Supelco, Bellefonte, PA) with a 500-µL loop, and a Fisher and Porter (Warminster, PA) flow meter (0-100% N2, 200 mL min-1). Ismatec Tygon tubing type R3607 of different internal diameters, 2.06- and 3.15-mm i,d,, was used for carrying the reducing agent, carrier solution, and waste solution. The initial conditions for cold vapor generation using NaBH4 as a reducing agent were established in previous work11,13 in which a similar FIA system was used, and these initial conditions were as follows: 0.15% mass v-1 NaBH4 solution stabilized in 0.5% mass v-1 NaOH at 3 mL min-1; 3% v v-1 HCl solution used as carrier at 10 mL min-1. Mercury atomic

absorbance was measured with a Thermo (Cambridge, UK) atomic absorption spectrometer model Solar S2 equipped with a homemade quartz tube. Inorganic mercury determination was performed with SnCl2;13 a 5% mass v-1 SnCl2 solution in 1% mass mass-1 HCl was used as reducing stream with a 1 mL min-1 flow rate; a 0.01% mass mass-1 HCl solution was used as carrier with a 12 mL min-1 flow rate. The quartz tube was kept at room temperature during operation. A mercury hollow-cathode lamp (Thermo) operated at 4 mA was used as a radiation source. The mercury line at 253.7 nm and a slit width of 0.5 nm were used for measurements. Reagents and Standards. All chemicals were of analytical reagent grade. An inorganic mercury stock standard solution (Merck, Darmstat, Germany, 1 g L-1) was used. A methylmercury (MeHg+) stock standard solution (100 mg L-1) was prepared from methylmercury chloride (Riedel-de Ha¨en, Seelze, Germany), by dissolving the appropriate amount of the solid in pure ethanol (Merck), and made up to volume with ultrapure water. Stock standard solutions (100 mg L-1) of phenylmercury (PhHg+) and diphenylmercury (d-PhMg+) were prepared from the corresponding chloride salts (Riedel-de Ha¨en). Phenylmercury was dissolved in pure methanol and made up to volume with ultrapure water. Magnetic agitation for 4 h was necessary. Diphenylmercury was dissolved in pure methanol (Merck). All stock standard solutions were stored in a refrigerator at 4 °C and protected from light. Working standard solutions were prepared just before use by appropriate dilution of the stock standard solution. Sodium hypochlorite solution was purchased from Aldrich (Milwaukee, WI). Hydrochloric acid, 37% mass/mass, from Panreac (Barcelona, Spain) packed in a black plastic bottle was used. Tin(II) chloride used as reducing agent was prepared by dissolving the appropriate Analytical Chemistry, Vol. 78, No. 8, April 15, 2006

2495

Table 1. Operating Sequence of the SIA/HIFU/FIA Manifold

step

injection valve position

selection valve position

pump P1 on/off

pump P2 on/off (5 mL min-1)

pump P2 volume handling (mL)

ultrasonic probe

1 2 3 4 5 6

load load load load load inject

4 5 5 6 1 1

on on on on on on

on, inverse flow on, direct flow off on, inverse flow on, direct flow off

8 8 0 8 8 8

off off 1 min off off off

mass of tin(II) chloride dehydrate (Panreac) in concentrated hydrochloric acid and diluting with ultrapure water. Sodium tetrahydroborate(III) (Merck) was prepared fresh daily by dissolving the solid in a sodium hydroxide solution. Diluted hydrochloric acid was used as carrier. The interference study was carried out by dissolving the appropriate amounts of the following compounds: iron(III) nitrate (Merck); sodium chloride (Panreac); and calcium carbonate (Aldrich). Estuarine water reference material for trace metals NRC-SLEW-3 (National Research Council from Canada, http://www.nrc-cnrc.gc.ca/) and urine mercury reference material INSPQ/H-02-04 (Institut National de Sante Publique, http://www.inspq.qc.ca/) were used in order to assess proof of the methodology. The influence of each compound was determined at two concentration levels, 103 and 104 mg L-1. Simulated waste effluents spiked with different mercury compounds were prepared by dissolving the appropriate amounts of sodium oxalate (Merck) or potassium hydrogen phthalate (Panreac) to produce a chemical oxygen demand (COD) of up to 104 mg L-1. Description of the SIA/HIFU/FIA System. The SIA/HIFU/ FIA system for on-line determination of mercury is summarized in Table 1 and is shown in Figure II, Supporting Information, and is also depicted in Figure 2. In the first step, Figure 2A, the holding coil, HC (Ismatec Tygon tube type R3607, 3.15 mm i.d., 600 mm long) is filled with 4 mL of sample since the central communication channel of the selection valve, SV, is directed to port 4 and pump 2 is in reverse flow. In the second step, the central communication channel of the SV is directed to port 5 and pump 2 is in direct flow; now the chemical reactor is filled with 4 mL of sample. In the third step, pump 2 is off and the sonication probe is turned on; high-intensity focused ultrasound is now applied to the sample for 1 or 3 min. In the fourth step, when the sonication process is finished, the central communication channel of the SV is directed to port 6, pump 2 is in reverse flow, and the holding coil is filled again, this time with the treated sample. In the fifth step, the central communication channel of the SV is directed to port 1 and pump 2 is in direct flow, which allows filling the loop of the injection valve (IV). In the last step, Figure 2B, pump 2 is off and the injection valve is positioned at “inject”. Now the treated sample is in the FIA system where it is fluxed toward the gas-liquid separator (GLS), merging with the NaBH4. Hg0 is formed and fluxed with the N2 stream toward the quartz T, where the mercury measurement takes place. Now replicates of the treated sample can be done by refilling the loop or the excess of sample can be send to waste using the FIA system. The system is ready to start with a new sample again. HIFU Oxidation Procedure. Solutions were prepared by mixing different amounts of ultrapure water, HCl, standard 2496 Analytical Chemistry, Vol. 78, No. 8, April 15, 2006

solutions, and NaClO immediately before the sample treatment. Once the chemical reactor (Figure 1) was filled with the sample solution, ultrasonication was applied for 1 min at 50% of sonication amplitude. Sonication was not applied to control solutions. The mercury concentration chosen in this work was 30 ppb. Sonicated solutions were pumped by the SIA system to the FIA system through port 1 of the six-port injection valve.

Figure 2. SIA/HIFU/FIA manifold for on-line determination of mercury by cold vapor atomic absorption spectroscopy. Key: NaBH4, NaBH4 solution; HCl, HCl solution; sample, sample solution mix with HCl and HClO; P1, P2, peristaltic pumps; INV, inverse flow; DIR, direct flow; HC, holding coil; IV, injection valve with a 500-µL sample loop; SV, selection valve; W, waste; AAC, atomic absorption cell; GLS, gas-liquid separator; N2, stream of N2; US, ultrasonic probe. (A) Injection valve in load position. (B) Injection valve in injection position. CR, chemical reactor.

RESULTS AND DISCUSSION There are two main difficulties dealing with fitting a probe system into a closed vessel.16 First, the size of the ultrasonic probe can be too short or too large for the volumes to be treated. Since the performance of the ultrasonic probe changes as a function of its design, a compromise between probe size and sample volume must be achieved to obtain reliable results. Second, and more important, the probe is undergoing ultrasonic vibrations at most points along its length. Nevertheless, there is a null point, sometimes with a screw thread, on most commercial probe sonicators. Anything attached onto the probe at this point will not be subject to vibration and will not affect the performance of the ultrasonication. The description of the HIFU closed system used here is shown in Figure 1 and also in the Supporting Information, Figure I. Once the HIFU oxidation was complete, the measurements were performed immediately in order to minimize potential degradation. The data reported in the literature about degradation of mercury compounds are inconclusive. For instance, the photolytic degradation of methylmercury was cited,17 whereas other authors stressed the high stability of this organomercurial under drastic conditions.18 Although chlorine can absorb in the UV region (e.g., 253.7 nm), in our experimental conditions, such interference was not observed. Influence of Sodium Hypochlorite Concentration. The organomercurial degradation based on the application of sonolysis through a probe sonicator (HIFU) was described for first time by Capelo et al.13 The proposed methodology showed (i) efficient decomposition of organomercurials within a few minutes using mild conditions, e.g., diluted HCl at room temperature, and (ii) high tolerance to the presence of salts and oxidizable organic mater. Nevertheless, recent findings have stressed the importance of the presence of the hypochlorite ion in solution when the above referred methodology is applied.19 In these sets of experiments, the concentration of HCl 1 M as given in ref 13 was chosen, and the variable optimized was the ClO-concentration. Results are shown in Figure 3 where recovery was expressed as the ratio of the peak absorbance corresponding to a 30 µg L-1 concentration of organomercurial (expressed as Hg) and a 30 µg L-1 Hg2+ solution prepared in the same liquid medium as the organomercurial. It should be noted that the reducing agent used in this study was sodium tetrahydroborate(III). This agent does not provide the same sensitivity for all mercury species20and, in the absence of any oxidation caused by the application of HIFU, may give a measurable absorbance due to the partial conversion of organic mercury into Hg0. As can be seen in Figure 3, the 4-mL sample volumes containing variable concentrations of ClO- in the range 0-3%, v/v, and a fixed concentration of HCl 1 M were sonicated at a 50% amplitude for 60 s. In the case of methylmercury, the minimum ClO- concentration needed to obtain its degradation in a ratio higher than 90% with the aid of an ultrasonic (16) Mason, T. J. Sonochemistry; Oxford University Press: New York, 1999; pp 58-67. (17) Luna, J.; Brunetto M. R.; Burguera, J. M. L.; Burguera, M.; Galignani, M. Fresenius J. Anal. Chem. 1996, 354, 367-369. (18) Ahmed, R.; Stoeppler, M. Analyst 1986, 111, 1371-1374. (19) Capelo, J. L.; Rivas, M. G.; Oliveira, L. G.; Vilhena, C.; Santos, A. C.; Valada, T.; Galesio, M.; Oliveira, P.; Gomes da Silva, M. D. R.; Gaspar, E. M.; Alves, S.; Fernandez, C.; Vaz, C. Talanta 2006, 68, 813-818. (20) Baxter, D. C.; Frech, W. Anal. Chim. Acta 1990, 236, 377-284.

Figure 3. Influence of NaClO concentration (% v v-1) on mercury recovery. Solutions were prepared in 1 M HCl (b) without sonication and (9) with sonication. HIFU conditions: 50% sonication amplitude, 1-min sonication time.

field was 0.001% v/v. For the same concentration but without sonication, the degradation yield was 58%. Without sonication, the maximum methylmercury degradation observed was ∼80% for the highest hypochlorite concentration. In all the hypochlorite concentrations assessed, the methylmercury degradation was higher when the HIFU was applied. The lower the ClO- concentration, the higher the influence of the sonication process. Figure 3 also shows that phenylmercury was the most readily degradable species with the lower ClO- concentration; the organic conversion into inorganic mercury was higher than 80%, for 1-min sonication time. Another interesting finding is that, for high hypochlorite concentrations, phenylmercury and diphenylmercury were completely decomposed without the aid of ultrasonication. Nevertheless, high concentrations of toxic chemicals, such as hypochlorite. should be, when possible, avoided. Finally, Figure 3 also shows that relative standard deviations were, in general, lower than 10% (n ) 3). During the treatment, the proper performance of the Analytical Chemistry, Vol. 78, No. 8, April 15, 2006

2497

Figure 4. Influence of HCl concentration (M) on mercury recovery. Solutions were prepared in 0.01% v v-1 NaClO (b) without sonication and (9) with sonication. HIFU conditions: 50% sonication amplitude, 1-min sonication time.

ultrasonic probe was verified in the closed vessel by visually observing the characteristic cone of bubbles below the head of the probe, caused by the cavitation phenomenon.21 The optimum hypochlorite concentration was chosen as 0.01% v/v. Hence, although this procedure makes use of hypochlorite, it only does so in very modest quantities and therefore contributes to the green quality of the proposed procedure. Influence of Hydrochloric Acid Concentration. Figure 4 shows the effect of hydrochloric acid in the degradation process. The hypochlorite concentration used in these experiments was 0.01% v/v. The HCl concentration range studied was 0.001-1 M. As can be observed in Figure 4, the methylmercury was completely decomposed when ultrasonication was used in the presence of a 0.5 M hydrochloric acid concentration. To obtain the same result for the other two organomercurials, the hydrochloric acid concentration needed was lower, of the order of ∼0.01 M. In these experiments, the same trend was observed when the influence of (21) Didenko Y. T.; Suslick, K. S. Nature 2002, 418, 394-397.

2498 Analytical Chemistry, Vol. 78, No. 8, April 15, 2006

the hypochlorite ion was studied; that is, the methylmercury was the most refractory species destroyed by the treatment. During the optimization process, clogging of the system was not observed. The hydrochloric acid concentration chosen as optimum was 0.05 M. Influence of Solution Composition in the Degradation Process. Different substances can be present in solution and act as scavengers of the hydroxyl and others radicals generated by the sonication process, yielding lower degradation rates. The presence of these scavengers has been mentioned to be an important drawback of most advanced oxidation processes.22,23 Hence, oxidation experiments were done involving concentration levels of 103 and 104 mg L-1 Ca(CO3)2, Fe(NO)3, and NaCl. These substances were chosen because they have been related to hydroxyl radical scavengers or because they are involved in redox processes.24 Results are shown in Table 2. As can be noted, recoveries were always higher than 90% when HIFU plus NaClO was used. In addition, experiments were carried out with model wastewaters containing different COD concentrations, from 103 to 104 mg of O2 L-1, and prepared from sodium oxalate and potassium phthalate, as described elsewhere.25 Results are also shown in Table 2. The higher interference for the degradation process was due to the oxalate ion. In this solution, the recoveries were lower than 70% for all the species tested when 1-min sonication time was used (data not shown). The ion oxalate is a refractory compound that can act as scavenger and is also the final product of many degradation processes involving sonication.26 However, when the sonication time was increased up to 3 min, the recuperation yields were higher than 90%. Low recoveries were achieved for methylmercury when sonication was not used. Analytical Figures of Merit. Calibration was performed with a series of Hg(II) standards (2-50 mg L-1). Sensitivity (m) was the slope value obtained by least-squares regression analysis of calibration curves based on peak height measurements. The equation (n ) 5) for the calibration curve was as follows: Y ) (0.0035 ( 0002) [Hg] ( (0.001 ( 001), where Y is peak absorbance and [Hg] is the mercury concentration expressed as micrograms per liter. The limit of detection, equal to 0.5 µg L-1, was defined as 3 × s × m-1, with s being the standard deviation corresponding to 10 blank injections and m the slope of the calibration graph. The quantification limit, defined as 10 × s × m-1, was 1.7 µg L-1. The relative standard deviations, estimated from 10 standard replicates and calculated at concentrations of 5 and 50 µg L-1 were, respectively, 6 and 3%. Determination of Inorganic Mercury and Total Mercury in Spiked Reference Materials. The estuarine water reference material for trace metals (NRC-SLEW-3) was used as the matrix for model solutions containing 30 ppb total spiked mercury. The solutions were prepared with different concentrations of the inorganic and organic species studied. Inorganic mercury was determined in solutions acidified up to 0.5 M HCl, without HIFU and with SnCl2 as reducing agent. This reagent was chosen for inorganic determinations because NaBH4 can also partially reduce (22) Camel, V.; Bermond, A. Water Res. 1998, 32, 3208-3222. (23) Olson, T. M.; Barbier, P. F. Water Res. 1994, 28, 1383-1391. (24) Hua, L.; Hoffmann, M. R. Environ. Sci. Technol. 1996, 30, 864-871. (25) Capelo J. L.; Lavilla I.; Bendicho, C. Anal. Chem. 2001, 73, 3732-3736. (26) Weavers, L. K.; Malmstadt, N.; Hoffmann, M. R. Water Res. 2000, 34, 12801285

Table 2. Influence of Foreign Substances (mg L-1) on Mercury Recovery (30 µg L-1) from Spiked Reference Material Estuarine Water NRC-SLEW-3a phenylmercury (30 µg L-1)

methylmercury (30 µg L-1) with US (mg L-1) reagent

103

KC8H5O4 H2C2O4 CaCO3 Fe2(NO3)2 NaCl

99 ( 98 ( 3b 92 ( 2 95 ( 3 97 ( 1

a

without US (mg L-1)

104 3b

91 ( 91 ( 1b 88 ( 1 91 ( 3 94 ( 4

2a

with US (mg L-1)

103

104

103

63 ( 8 9(1 70 ( 9 74 ( 4 77 ( 2

45 ( 3 8(1 59 ( 18 69 ( 2 75 ( 9

102 ( 94 ( 4b 101 ( 3 99 ( 4 99 ( 4

without US (mg L-1)

104 2b

diphenylmercury (30 µg L-1)

97 ( 101 ( 2b 99 ( 4 101 ( 3 97 ( 6 2b

with US (mg L-1)

103

104

103

96 ( 3 15 ( 3 98 ( 6 95 ( 5 98 ( 2

49 ( 1 16 ( 1 100 ( 3 98 ( 3 95 ( 2

101 ( 93 ( 7b 97 ( 4 94 ( 2 99 ( 5

without US (mg L-1)

104 3b

102 ( 91 ( 3b 98 ( 3 99 ( 3 99 ( 3 1b

103

104

98 ( 8 11 ( 1 94 ( 4 97 ( 3 94 ( 3

81 ( 3 12 ( 1 99 ( 4 99 ( 6 101 ( 6

HIFU conditions: 50% sonication amplitude, 1-min sonication time. b Sonication time was 3 min.

Table 3. Mercury Determination (µg L-1) by SIA/HIFU/FIAa In-Hg+

MeHg+

PhHg+

d-PhHg+

Hg found

30 10 10 0 10

0 10 10 10 0

0 10 10 10 10

0 10 0 10 10

27 ( 3a 11 ( 3a 28 ( 2b 27 ( 2b 31 ( 3b

a HIFU was off and no hypochlorite ion was added. b HIFU was on and hypochlorite ion was added. See text for details.

organic mercury species to inorganic mercury, thus yielding inconsistent results for inorganic determinations. Total mercury was determined after the previously described HIFU oxidation process, with the optimized concentrations for hydrochloric acid and hypochlorite ion. The results are shown in Table 3. Recovery was expressed as the ratio of the peak absorbance corresponding to a 30 µg L-1 concentration of organomercurial (expressed as Hg) and a 30 µg L-1 Hg2+ solution prepared in the same liquid medium as the organomercurial. As can be noted, inorganic mercury and total mercury can be separately determined in the same solution. When HIFU is off and the hypochlorite ion is not added to solution, the signal corresponds only to the inorganic mercury present in solution. On the contrary, when the sonication process takes place in the presence of the hypochlorite ion, the signal corresponds to all mercury species in solution. Further experiences were developed using INSPQ H-02-04 urine mercury reference material. The urine was diluted 5 times and then was fortified with Hg2+ and CH3Hg+ to a final concentration of 30 ng g-1 in both mercury species. Finally, the sample treatment with sonication was applied. As mentioned above for water, when no

HIFU was applied, the mercury recovered was only the inorganic one (95 ( 7%, n ) 3). After treatment, the total mercury added, including the methylmercury, was completely recovered (97 (8%, n ) 3). CONCLUSIONS This work has demonstrated the value of the on-line hyphenation of HIFU with a SIA/FIA system. At present, HIFU can be fully automated since commercially available software allows the implementation of the sonication process in on-line systems. These findings can reduce the amount of chemicals used as well as their handling in the laboratory, minimizing the risks of contamination. Since HIFU is used for many different purposes, this new concept of the SIA/HIFU/FIA is expected to stimulate new experiments in the analytical laboratory. For instance, columns allowing preconcentration can be easily added to the system. Furthermore, the proposed methodology can be also used with ICP, ICPMS, AFS and, as a general role, in any FIA or SIA system. ACKNOWLEDGMENT J.L.C. acknowledges Professor Dr. Frausto da Silva, Technical University of Lisbon, Portugal, for fruitful discussions and Dr. Adrian Oehmen, from the New University of Lisbon, Portugal, for carefully reading of the manuscript. SUPPORTING INFORMATION AVAILABLE Additional information as noted in text. This material is available free of charge via the Internet at http://pubs.acs.org. Received for review September 19, 2005. Accepted February 20, 2006. AC0516685

Analytical Chemistry, Vol. 78, No. 8, April 15, 2006

2499