Anal. Chem. 2006, 78, 6260-6264
Ultrasound-Promoted Cold Vapor Generation in the Presence of Formic Acid for Determination of Mercury by Atomic Absorption Spectrometry Sandra Gil, Isela Lavilla, and Carlos Bendicho*
Departamento de Quı´mica Analı´tica y Alimentaria, Area de Quı´mica Analı´tica, Facultad de Quı´mica, Universidad de Vigo, Campus As Lagoas-Marcosende s/n, 36310 Vigo, Spain
A new cold vapor technique within the context of green chemistry is described for determination of mercury in liquid samples following high-intensity ultrasonication. Volatile Hg evolved in a sonoreactor without the use of a chemical reducing agent is carried to a quartz cell kept at room temperature for measurement of the atomic absorption. The mechanism involved lies in the reduction of Hg(II) to Hg(0) by reducing gases formed upon sonication and subsequent volatilization of Hg(0) due to the degassing effect caused by the cavitation phenomenon. Addition of a low molecular weight organic acid such as formic acid favors the process, but vapor generation also occurs from Hg solutions in ultrapure water. The detection limit of Hg was 0.1 µg/L, and the repeatability, expressed as relative standard deviation, was 4.4% (peak height). Addition of small amounts of oxidizing substances such as the permanganate or dichromate anions completely suppressed the formation of Hg(0), which confirms the above mechanism. Effect of other factors such as ultrasound irradiation time, ultrasound amplitude, and the presence of concomitants are also investigated. Some complexing anions such as chloride favored the stabilization of Hg(II) in solution, hence causing an interference effect on the ultrasound-assisted reduction/volatilization process. Nowadays, determination and monitoring of mercury has become of paramount concern, this metal being recognized as a major environmental pollution issue and health hazard for humans.1 Because of its extreme toxicity to living organisms, maximum Hg allowable concentration in many environmental and food samples have been established by regulations throughout the world. Mercury is routinely determined in a huge variety of samples by the cold vapor technique with atomic absorption spectrometry (CVAAS) or atomic fluorescence spectrometry detection.2 This technique encompasses high sensitivity, absence of spectral interferences, simplicity, and speed. Since the pioneering work by Hatch and Ott,3 CVAAS has undergone several * Corresponding author. Tel.: +34-986-812281. Fax: +34-986-812556. Email:
[email protected]. (1) Clarkson, T. W.; Magos, L.; Myers, G. J. N. Engl. J. Med. 2003, 349 (18), 1731-1737. (2) Cullen, M. Atomic Spectroscopy in Elemental Analysis; Blackwell Publishing: Oxford, 2004; p 241. (3) Hatch, W. R.; Ott, W. L. Anal. Chem. 1968, 40, 2085-2087.
6260 Analytical Chemistry, Vol. 78, No. 17, September 1, 2006
improvements concerning automation and operation mode. Mercury reduction and further release from the solution can be accomplished in both batch and on-line systems using a reducing agent (typically sodium borohydride or tin chloride) in the presence of an acid such as HCl.4 When many samples are analyzed using the CVAAS technique, operation costs and production of hazardous wastes should not be underestimated. Development of clean analytical methods within the context of green chemistry is increasingly demanded, and consequently, suppression or decrease of chemicals should be a step forward.5 In recent years, there have been significant developments in analytical methodology addressed to decrease the amount of harmful chemicals involved as well as to simplify and accelerate experimental procedures. Ultrasound have played an important role in these achievements, being a clean technology with increasing number of analytical applications such as extraction, digestion, oxidation, homogenization, emulsification, surface cleaning, and degassing.6 The aim of this work is to describe a new application of ultrasound in analytical chemistry, namely, its use for cold vapor generation from Hg(II) thus replacing the conventional chemical reaction needed for reduction of Hg(II) to Hg(0). The method proposed takes advantage of two well-known effects of ultrasound when they are applied to liquid samples containing Hg(II), i.e., production of radicals and the degassing effect. Cold vapor generation is carried out from aqueous media making use of a sonoreactor directly coupled to a quartz cell to which Hg(0) formed is transported for measurement of the mercury atomic absorption. EXPERIMENTAL SECTION Apparatus. A sonoreactor is designed for allowing highintensity probe sonication (Figure 1). The sonoreactor consists of a Pyrex glass vessel (38-mL capacity) (Barloworld Scientific) sealed up with a PTFE cap and a rubber septum. A hole (5-mm diameter) is made in the septum so that the probe (6-mm diameter) can be inserted into the vessel. The sonoreactor is (4) Welz, B.; Sperling, M. Atomic Absorption Spectrometry; Wiley-VCH Verlag GmbH: Weinheim, 1999; pp 203-219. (5) Anastas, P. T.; Warner, J. C. Green Chemistry: Theory and Practice; Oxford University Press: New York, 1998; p 30. (6) Priego-Capote, F.; Luque de Castro, M. D. Trends Anal. Chem. 2004, 23, 644-653. 10.1021/ac0606498 CCC: $33.50
© 2006 American Chemical Society Published on Web 08/03/2006
Figure 1. Schematic diagram of the ultrasound-assisted cold vapor generation system A, sonoreactor; B, ultrasonic probe; C, three-way valve; D, shut-off valve; E, generator; F, quartz cell; G, sample solution; H, soundproof chamber.
connected to a gas supply (N2) and to the quartz cell located in the optical path of the spectrometer by means of two PTFE capillary tubes. The N2 flow rate was controlled by a flowmeter (Fisher and Porter). The device is fitted with two PTFE valves manually operated. A 100-W, 20-kHz VC-100 ultrasonic processor (Sonics and Materials) equipped with a 6-mm titanium microtip was employed for ultrasound transmission. Ultrasonic vibrations of the probe microtip were fixed at any desired amplitude level using a power setting in the range 10-100% range. Warning: highintensity ultrasonication should be performed inside a soundproof chamber to prevent ear damage from ultrasound. Atomic absorption measurements of Hg were carried out with a Unicam (Cambridge, UK) atomic absorption spectrometer model Solaar 939 equipped with a Unicam EC90 quartz furnace. The quartz cell was kept at room temperature during operation. The Hg analytical line at 253.7 nm and a slit width of 0.5 nm were set up in the spectrometer. Reagents. All chemicals were of analytical-reagent grade. A Hg(II) stock solution (1000 mg/L) was prepared from HgCl2 (Carlo Erba). Working standards were made by suitable dilution of the stock solution with ultrapure water (Milli-Q, Millipore). Hydrochloric acid (37% m/m, Merck), KMnO4 (Panreac), K2Cr2O7 (Prolabo), and CH3HgCl (Riedel-de Ha¨en) were employed to investigate the mechanism for Hg(0) formation. HCOOH (Merck) was used to promote the formation of reduced mercury upon ultrasound irradiation. The following chemicals were employed for the interference study: NaCl, Na2CO3, CaCl2, MgCl2‚6H2O, Mg(NO3)2‚6H2O, Fe(NO3)3‚9H2O, CuCl2‚2H2O, and CrCl3‚6H2O from Merck; KCl from Prolabo; KNO3, NiCl2, and Fe(SO4)2Fe(NH4)2 from Probus; MnCl2‚4H2O and CoCl2‚6H2O from Scharlau; Ni(NO3)2‚6H2O, Cd(NO3)2‚4H2O, and Pb(NO3)2 from Panreac; Cr(NO3)3 from Sigma-Aldrich; SnCl2 from Carlo-Erba; humic acid from Fluka Chemie. Procedure. Ten milliliters of sample containing 0.9 mol L-1 HCOOH were dispensed into the sonoreactor for ultrasound treatment. The ultrasonic probe is immersed into the liquid at a depth of 1.5 cm, and ultrasound transmission is started. Valves C and D are closed, and ultrasound irradiation is performed for 10 min at a 50% amplitude of the probe vibration. Once Hg vapor has evolved into the gas phase of the sonoreactor, the reading is set up at the spectrometer and the valves are open so that the Hg
vapor is carried by nitrogen to the quartz cell for measurement of the Hg atomic absorption. RESULTS AND DISCUSSION Mechanism for Ultrasound-Assisted Cold Vapor Generation. Initial conditions included the presence of an organic acid, i.e., formic acid, which can form reducing radicals under sonolysis. It was seen that a 0.9 mol L-1 HCOOH concentration was sufficient to yield Hg(0) after 10-min sonication time at 50% ultrasonic amplitude. The formation of Hg(0) in the sonoreactor in the absence of an organic acid was demonstrated, but the process was much less effective. Thus, a peak absorbance of 0.210 ( 0.014 was observed with formic acid for a 5 µg/L Hg solution, while the peak absorbance dropped to 0.074 ( 0.004 in ultrapure water. When ultrasound irradiation is performed in aqueous media, a series of radicals are formed such as H• and OH• at the gas-phase interface of the cavitation bubble and to a lesser extent in the bulk solution.7 While the ability of ultrasound irradiation to promote oxidations is well known,8 reductions are more difficult to observe. However, as stated above, reducing radicals such as H• have been observed during sonication of water, which in turn, yield H2.7 The presence of an organic acid such as formic acid could provide additional reducing radicals when a solution of this acid is subjected to ultrasound irradiation. It has been reported that sonolysis of formic acid gives rise to the formation of CO, CO2 and H2.9 Therefore, Hg(II) is likely transformed into Hg(0) by the reducing gases formed upon sonolytic decomposition of formic acid. The need for a low molecular weight organic acid has also been reported elsewhere for efficient vapor generation from several elements, including Hg, using UV irradiation.10,11 Photolytic decomposition of formic acid also yields CO, H2, and CO2 as (7) Mason, T. J. Sonochemistry; Oxford University Press: Oxford, 1999; pp 10, 48. (8) Mason, T. J.; Lorimer, J. P. Applied Sonochemistry; Wiley-VCH: Weinheim, 2002; p 86. (9) Harada, H.; Nippon Kagaku Kaishi 1997, 3, 180-183. (10) Guo, X.; Sturgeon, R. E.; Mester, Z.; Gardner, G. J. Anal. Chem. 2004, 76, 2401. (11) Zheng, C.; Li, Y.; He, Y.; Ma, Q.; Hou, X. J. Anal. At. Spectrom. 2005, 20, 746-750.
Analytical Chemistry, Vol. 78, No. 17, September 1, 2006
6261
products,12 which seems to be a common feature in both UV and ultrasound-assisted vapor generation methods involving this organic acid. Ultrasound irradiation of solutions containing Hg species (i.e., methylmercury, phenylmercury, p-tholylmercury) has been attempted for the first time by the authors with the purpose of conversion of organomercurials into Hg(II) under high-intensity ultrasound irradiation.13 Efficient ultrasonic oxidation of these species occurred provided that a HCl medium was present, and besides, the Hg(II) formed was stable under the experimental conditions. Therefore, a first issue was to ascertain whether the formation of Hg(0) and its further volatilization could occur under the conditions employed for degradation of organomercurials. When 1 mol L-1 HCl was added to the liquid medium containing Hg(II) prior to ultrasonic irradiation, the atomic absorption signal was completely suppressed. This finding is in agreement with the above observations concerning the oxidation of organomercurials, in which the Hg(II) formed is not volatilized at all, but on the contrary, it remains in solution. It is clear that even small amounts of oxidants in the liquid medium should preclude the reduction of Hg(II) to Hg(0). Accordingly, no Hg(II) volatilization would occur unless reduction was facilitated. An experiment involving an oxidant was addressed to bring the reduction/volatilization mechanism to light. When 1 mg/L KMnO4 or K2Cr2O7 was added to a Hg(II) solution in ultrapure water, the atomic absorption signal was completely suppressed, thus confirming the fact that reduction of Hg(II) occurs as a result of its reaction with reducing gases formed upon sonication. Consequently, the presence of oxidants would eliminate these radicals and, in turn, would prevent Hg(II) from reduction. Discoloration of the KMnO4 solution in ultrapure water could also be observed following high-intensity ultrasonication even in the absence of Hg(II), which proved the reducing capability of ultrasound. According to these observations, the following mechanism can be proposed as being the responsible for cold vapor generation promoted by ultrasound irradiation:
2HCOOH f CO2 + H2 + CO + H2O
(1)
Hg2+ f Hg(0)sol
(2)
Hg(0)sol f Hg(0)gas
(3)
Reaction 1 occurs as a result of formic acid decomposition under ultrasound irradiation. Production of reducing gases such as CO and H2 formed upon sonication of formic acid solutions should facilitate the reduction of Hg(II) to Hg(0) according to reaction 2. The last reaction requires the absence of oxidants and complexing agents such as chloride. Once Hg(0) is formed, this species can be readily stripped from the solution as a result of the degassing effect caused by ultrasound (reaction 3). Influence of Sonochemical Parameters. Experiments were undertaken to establish the effect of the main sonochemical parameters, i.e., ultrasound irradiation time, ultrasonic amplitude, (12) Khriachtchev, L.; Mac¸ oas, E.; Petterson, M.; Ra¨sa¨nen. J. Am. Chem. Soc. 2002, 124, 10994-10995. (13) Capelo, J. L.; Lavilla, I.; Bendicho, C. Anal. Chem. 2000, 72, 4979-4984.
6262 Analytical Chemistry, Vol. 78, No. 17, September 1, 2006
Figure 2. Effect of the sonication time on the absorbance of Hg using a 10 µg/L in 1 mol L-1 formic acid. The sample volume was 10 mL; vibrational amplitude of the probe 50%. (Errors bars represent the standard deviation for N ) 3 measurements.)
composition of the liquid medium, and sample volume, on the absorbance from a 5 µg/L Hg(II) solution. A formic acid concentration of 0.9 mol L-1 was found as optimum for ultrasoundassisted cold vapor generation. Higher formic acid concentrations caused the Hg absorbance to decline. It is clear that the amounts of reducing gases, i.e., CO and H2, formed upon sonolytic decomposition of formic acid were dependent on the initial formic acid concentrations, and consequently, Hg absorbance should increase on increasing formic acid concentration. However, decomposition reactions for formic acid occur inside of the cavitation bubbles as a result of the drastic conditions existing at the moment of collapse (temperatures ∼5000 °C; 2000 atm pressure).7 The temperature inside the cavitation bubble decreases on increasing formic acid concentration in it, and in turn, this would cause the formic acid decomposition yield to worsen.14 This fact should diminish the amount of reducing gases needed for reduction of Hg(II). As can be observed in Figure 2, a steady absorbance cannot be reached at the maximum sonication time attempted. For a sonication time longer than 10 min, excessive heating of the solution took place, which could impair transmission of ultrasound from the probe. Therefore, a 10-min sonication time was chosen so that adequate sensitivity is achieved without excessive heating of the solution. To know the effectiveness of the reduction/ volatilization process, several reduction/volatilization runs were carried out with the same sample placed in the sonoreactor and atomic absorbance signals were successively obtained. For the second run, absorbance represented 45% in regard to the value obtained in the first run, while for the third run, absorbance dropped to only 23% of that value. It is estimated that an amount around 54% of the Hg present in the initial solution is passed to the gas phase of the sonoreactor at the first run. The effect of the vibrational amplitude of the probe was also studied. Absorbance increases up to a 50% amplitude and then levels off from that value. The quartz cell temperature was also seen to influence the atomic absorption signal of Hg. A decreasing absorbance was observed on increasing the cell temperature from 25 to 800 °C. It can be concluded that Hg(0) is the volatile species reaching the quartz cell since other Hg volatile compounds would need a heated quartz cell for efficient atomization. When a solution (14) Hart, E.; Henglein, A. Radiat. Phys. Chem. 1988, 32, 11-13.
Figure 3. Typical atomic absorption profile of Hg using ultrasoundassisted cold vapor generation. Table 1. Effect of Concomitants on the Ultrasound-Promoted Cold Vapor Generation.
a
interference
interference concn (mg/L)a
interference effect (%)
NaCl NaCl NaCl Na2CO3 Na2CO3 CaCl2 KCl MgCl2 MgCl2 KNO3 MnCl2 Pb(NO3)2 CoCl2 NiCl2 Ni(NO3)2 (SO4)2Fe(NH4)2 Cd(NO3)2 CuCl2 CuCl2 CrCl3 Cr(NO3)3 SnCl2 humic acid
10 100 1000 10 100 100 100 100 10 100 10 10 10 10 10 10 10 1 10 10 10 10 0.1
-7 -34 -57 -24 -50 -54 -50 -60 -7 +1 -37 -6 -21 -18 -9 -3 -3 -24 -50 -18 -7 -37 -5
Concentration of interfering agent expressed as mg/L of the metal.
of 5 µg/L Hg as methylmercury instead of Hg(II) was placed in the sonoreactor and subjected to ultrasound irradiation, no signal could be measured, meaning that methylation reactions are unlikely to occur under these conditions. A flow rate of N2 employed as carrier gas of ∼950 mL/min was found as optimum. Under these conditions, the atomic absorption profile was integrated within 20 s (Figure 3). Finally, the sample volume caused a remarkable influence on the Hg absorbance. A decreased effectiveness of ultrasound-assisted reduction/volatilization for increased sample volume is observed. This effect can be counteracted using longer irradiation times. A sample volume less than 10 mL was out of the range of the volumes recommended for the probe microtip employed (i.e.m 10-25 mL). A sample volume of 10 mL was selected for the remaining experiments. Effect of Concomitants. To evaluate the analytical applicability of the new cold vapor generation approach, the effect of some concomitants was studied (Table 1). For these experiments, the Hg(II) concentration was fixed at 5 µg/L and the effect of several alkaline and alkaline earth salts as well as some transition and heavy metals was assessed. The 100 mg/L NaCl, KCl, MgCl2, and Na2CO3 caused a depressive effect beyond 50%. This interference effect was negligible at the 10 mg/L concentration level except for Na2CO3. As explained above, chloride anion stabilizes Hg(II) in solution owing to the formation of strong complexes. In fact, chloride media have been found useful for preservation of Hg(II) in the parts per billion range.15 When KNO3 instead of KCl was
added to the solution containing Hg(II), no interference was observed for a 100 mg/L concentration. Bicarbonate and carbonate anions behave as radical scavengers in ultrasound-based degradation methods,16,17 and hence, they should prevent formic acid from degradation. This, in turn, would eliminate formation of reducing gases required for Hg(II) reduction. Reducing substances such as (SO4)2Fe(NH4)2 did not cause any interference at the 10 mg/L level. On the contrary, SnCl2, i.e., a typical reducing agent used for chemical cold vapor generation, caused a depressive effect as a consequence of Hg losses prior to ultrasonication. Transition metals caused a slight depressive effect at the 10 mg/L level with the exception of Mn(II) and Cu(II) in the chloride form. The effect of natural organic matter such as humic acid was studied. No interference was observed at the 0.1 mg/L level. Analytical Characteristics and Application. Analytical characteristics were established under optimal conditions for ultrasound-assisted cold vapor generation. The equations for the calibration lines using peak height and peak area of the atomic absorption signal were as follows:
peak height Y ) 0.0405 [Hg] + 0.0079; peak area Y ) 0.196 [Hg] - 0.0136;
r2 ) 0.9996 r2 ) 0.9992
where Y is peak or integrated absorbance and [Hg] is the Hg concentration (expressed as µg/L). The calibration curves were linear at least up to a Hg concentration of 10 µg/L. limits of detection (LODs) (3σ criterion) using peak height and peak area measurement were 0.1 and 0.3 µg/L, respectively. These LODs compare well with those reported for the use of chemical cold vapor generation in both batch and on-line systems.18,19 An improved LOD is obtained as compared with those reported in EPA methods 245.1 and 245.2 for determination of Hg by CVAAS. The repeatability expressed as relative standard deviation from N ) 15 measurements was 4.4 and 4.9% using quantitation of the atomic absorption signal in peak height and peak area, respectively. An application of ultrasound-assisted cold vapor generation to determination of Hg in CRM NWTM-27.2 (fortified water) was performed. This sample corresponded to diluted Lake Ontario water, which had been filtered and preserved in 0.2% v/v HNO3. Concentration of matrix elements such as Ca, Mg, Na, and K in this water were 6.1, 1.4, 2.2, and 0.3 mg/L, respectively. As no certified Hg content was supplied for this sample, it was spiked with 4 µg/L Hg(II). After calibration using aqueous standards prepared from Hg(II), the content found was 3.8 ( 0.3 µg/L (N ) 4), hence demonstrating the usefulness of this cold vapor generation method for determination of Hg in natural water. A model solution simulating freshwater (50 mg/L Na2CO3; 8 mg/L NaCl; 2 mg/L KCl; 50 mg/L Na2SO4) and spiked with 5 µg/L Hg was also analyzed using the standard addition method for calibration. The concentration found was 4.7 ( 0.55 µg/L (N ) 4). (15) Yu, L.-P.; Yan, X.-P. Trends Anal. Chem. 2003, 22, 245-253. (16) Camel, V.; Bermond, A. Water Res. 1998, 32, 3208-3222. (17) Olson, T. M.; Barbier, P. F. Water Res. 1994, 28, 1383-1391. (18) Landi, S.; Fagioli, F. Anal. Chim. Acta 1994, 298, 363-374. (19) Capelo, J. L.; Lavilla, I.; Bendicho, C. At. Spectrosc. 2000, 21, 229-233.
Analytical Chemistry, Vol. 78, No. 17, September 1, 2006
6263
CONCLUSIONS The work presented has proved for the first time the great potential of ultrasound irradiation to promote reduction of Hg(II) to Hg(0) and further volatilization of this species, which is the basis for a green cold vapor generation technique for mercury determination at ultratrace level. The use of typical chemical reducing agents such as sodium tetrahydroborate and stannous chloride as well as mineral acids can be eliminated. Ultrasoundassisted reduction/volatilization for cold vapor generation could also open new possibilities in field analysis since the sonoreactor can be readily adapted to portable instruments. Further research
6264
Analytical Chemistry, Vol. 78, No. 17, September 1, 2006
is needed to establish the scope of ultrasound-assisted cold vapor generation and assess its potential for application in real samples. ACKNOWLEDGMENT Financial support from the Galician government (Xunta de Galicia) (project PGIDIT05PXIB31401PR) is gratefully acknowledged. The authors thank the reviewers for helpful comments. Received for review April 7, 2006. Accepted July 11, 2006. AC0606498
