- Email: [email protected]
Determination of mercury in naphtha and petroleum condensate by photochemical vapor generation atomic absorption spectrometry
Microchemical Journal 110 (2013) 227–232
Contents lists available at ScienceDirect
Microchemical Journal journal homepage: www.elsevier.com/locate/microc
Determination of mercury in naphtha and petroleum condensate by photochemical vapor generation atomic absorption spectrometry☆ Alexandre de Jesus a, Ariane Vanessa Zmozinski a, Mariana Antunes Vieira b, Anderson Schwingel Ribeiro b, Márcia Messias da Silva a, c,⁎ a
Instituto de Química, Universidade Federal do Rio Grande do Sul, Av. Bento Gonçalves, 9500, 91501-970 Porto Alegre, RS, Brazil Universidade Federal de Pelotas, Centro de Ciências Químicas, Farmacêuticas e de Alimentos, Programa de Pós-Graduação em Química, Laboratório de Metrologia Química, 96160-000 Capão do Leão, RS, Brazil c Instituto Nacional de Ciência e Tecnologia do CNPq, INCT de Energia e Ambiente, UFBA, CEP 40170-115, Salvador, BA, Brazil b
a r t i c l e
i n f o
Article history: Received 15 January 2013 Received in revised form 25 March 2013 Accepted 27 March 2013 Available online 10 April 2013 Keywords: Mercury Naphtha Petroleum condensate Photochemical vapor generation Microemulsion AAS
a b s t r a c t In this work the feasibility of mercury determination in naphtha and petroleum condensate by photochemical vapor generation was investigated. The samples were pumped through the photochemical reactor as microemulsions and the volatile compounds formed were conducted to a quartz cell for the atomic absorption measurements. All the parameters of the system (sample preparation, organic precursor, sample flow rate, carrier gas flow rate and evaluation of the signal measurement mode) were investigated. The microemulsions were prepared mixing the samples with propan-1-ol and a small amount of water. The addition of low molecular weight organic acid was investigated and it was found not necessary once the propan-1-ol itself was efficient for promoting the generation of the volatile compounds of mercury. Calibration curves obtained with organic and inorganic standards showed correlation coefficients higher than 0.99 and characteristic mass of 2.0 and 2.4 ng of mercury for organic and inorganic standards, respectively, was obtained. There was no significant difference between the sensitivity of inorganic or organic standards for the calibration. Relative standard deviations ranged from 1 to 5% for three consecutive measurements. The limit of detection of 0.6 μg L−1, calculated for the amount of sample used in the microemulsion (1.0 mL), was obtained. A sample throughput was 2 samples per hour, considering triplicate analysis. Different naphtha and petroleum condensate samples were analyzed. No mercury was found in the naphtha samples, considering the limit of detection. In the petroleum condensate samples the concentrations ranged from 76 to 105 μg L−1. The accuracy was evaluated by assessing the recoveries of inorganic and organic species of mercury added to the samples, being obtained values in the range of 92 to 113%. The developed method was simple and fast, allowed direct analysis of naphtha and petroleum condensate with a reduced amount of reagents, thus contributing to green chemistry. © 2013 Elsevier B.V. All rights reserved.
1. Introduction Naphtha and petroleum condensate are derived products from crude oil and they are important raw materials for the petrochemical industry. Naphtha is a very important feedstock for the fine chemical industries, used in gasoline formulation and in the production of ethylene, propene, benzene, toluene and xylenes [1]. Petroleum condensate replaces naphtha as raw material in the petrochemical industry; it is a heavier fraction of petroleum compared to naphtha and may contain higher concentration of metals [2]. The presence of trace elements, even in μg L −1, in these petroleum-derived products can be associated
☆ Paper presented at 5th Ibero-American Congress of Analytical Chemistry 2012. ⁎ Corresponding author at: Instituto de Química, Universidade Federal do Rio Grande do Sul, Av. Bento Gonçalves, 9500, 91501-970 Porto Alegre, RS, Brazil. Tel.: +55 51 3308 6278; fax: +55 51 3308 7304. E-mail address: [email protected] (M.M. da Silva). 0026-265X/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.microc.2013.03.019
with many distinct problems such as: catalyzation of oxidative reactions, leading to a loss of the thermal stability of the products; release of some elements to the atmosphere in the refining process; corrosion problems and catalyst poisoning (in the case of naphtha); and also the presence of toxic elements remained in the distilled fractions, can be released to the environment when they are used as energy source [3,4]. Mercury is a trace component of all fossil fuels including natural gas, gas condensates, crude oil, coal, tar sands and other bitumens. The use of fossil hydrocarbons as fuels provides the main opportunity for emissions of the Hg they contain to the atmospheric environment [5,6]. Mercury is also a particular problem in the processing since it reacts with metallic surfaces causing corrosion and is a poison for noble metal catalyst used in many hydrocarbon process reactions. Many operators set tight limits on the level of Hg in naphtha feed to crackers (typically less than 5 μg L−1) [7]. Thus monitoring the presence of Hg in these products is increasingly demanded.
228
A. de Jesus et al. / Microchemical Journal 110 (2013) 227–232
The main problems of the determination of Hg in these samples are the fact that they are very complex and Hg is found in very low levels, requiring the use of highly sensitive techniques. There are few works in the literature concerning Hg determination in these samples. The direct analysis of naphtha and petroleum condensate after dilution of the samples with xylene by ICP-MS was proposed by Olsen et al. [8]. These authors reported that samples could be analyzed directly without pretreatment and the elements were determined simultaneously. However the standard deviation for Hg was very high, probably due to the instability of this element in organic solvent and possibly some memory effects. The emulsification of the samples with Triton X-100 was also proposed: it was reported that with the formation of emulsions a stability of at least one hour was obtained [9]. With the use of GF AAS, which is a technique adequate for complex matrices, and using Pd solution as chemical modifier, Santos et al. [10] used detergentless microemulsion as sample preparation and pre-concentrated Hg by multiple injections in the graphite tube obtaining a limit of detection of 0.78 μg L −1. Ceccarelli et al. [11] analyzed naphtha samples directly in the graphite tube for Hg determination obtaining a limit of detection of 32 μg kg −1. The formation and analysis of emulsions or microemulsions have been successfully applied for the preparation of oil samples, due to the homogeneous dispersion and stabilization of the oil microdroplets in the aqueous phase, which brings the viscosity close to that of an aqueous solution and reduces the organic load of the system. Besides, it allows the use of aqueous standards for calibration instead of expensive and instable organometallic standards. This technique has been applied for determination of other elements in petroleum derivatives, as naphtha and petroleum condensate [1,3,12–14]. The use of methods based on vapor generation as sample introduction can offer important advantages for analysis of complex matrices, such as petroleum derivatives. With the previous separation between the analyte and the matrix and efficiency in sample introduction, great improvement in the limit of detection can be achieved and some serious spectroscopic and/or matrix interferences eliminated [15]. The main vapor generation methods (as cold vapor and hydride generation) normally make use of expensive and instable reagents, as tetrahydroborate (NaBH4). Other disadvantages are that few elements can be converted to volatile compounds, the presence of transition metals (as Ni, Co and Cu) can cause the decomposition of the analyte hydride and the main reductor used, NaBH4, is a potential source of contamination [16]. Besides, the complete digestion of these organic samples is frequently required. There are a few works about determination of Hg in petroleum derivates (gasoline samples) using chemical vapor generation without digestion procedures [17,18]. Due to all the aspects cited the development of alternative vapor generation systems for the determination of trace elements in complex samples is still necessary in atomic spectrometry. An alternative method based on photochemical vapor generation (PVG) was proposed [15]. This method consists in the formation of analyte volatile compounds from the sample by the interaction between an organic precursor added to the sample and the ultraviolet light from the photochemical reactor. Then, the obtained volatile compounds are separated from the residual solution and carried to the detection system. In 2007, He et al. [19] published a critical evaluation of the applications of PVG in atomic spectrometry and discussion about the advances of the technique. The authors emphasized several advantages of the method such as no need of using relatively expensive and/or unstable reagents such as tetrahydroborate, that makes the method simpler, greener and in most cases cost-effective, besides the expanded number of detectable elements. Moreover, due to the higher tolerance to interferences, a drastic digestion of the sample is often unnecessary, as in the case of Hg determination in biological tissues solubilized with formic acid [20,21] or tetramethylammonium hydroxide [21]. Other examples are the direct determination of Hg in ethanol fuel [22] and wine and liquors [23].
In ten years of existence, several investigations have been done that help the understanding of the process occurring during photochemical vapor generation. However, there are still few applications to real samples. Based on the best of our knowledge, the use of PVG for determination of Hg in petroleum derivatives has not been published yet. The present work proposes the use of the photochemical vapor generation coupled with atomic absorption spectrometry for the direct determination of Hg in naphtha and petroleum condensate samples. The objective is to obtain a simple, fast and accurate method using detergentless microemulsions as sample preparation, avoiding digestion procedures. 2. Experimental 2.1. Instrumentation A photochemical reactor, consisting of spiral quartz tubing (115 cm × 1.3 mm i.d × 3.0 mm o.d, Internal volume: 2.3 mL) wrapped around of a low-pressure Hg vapor UV lamp (254 nm, 18 W, TUV PL-L, Philips, Poland), was constructed. Samples were propelled through the photochemical reactor with the assistance of a Reglo peristaltic pump (Ismatec, Switzerland). A flow of nitrogen (Air liquide, Canoas, Brazil) was introduced in the line after the photochemical reactor to carry the volatile species; the flow rate was controlled by a flow meter (Cole Parmer, Vernon Hills, IL, USA). The volatile species formed mixed with the nitrogen were introduced in a gas–liquid separator to separate the residual waste from the gaseous phase. After the separation, the volatile species were transported to a quartz tube (10 mm i.d × 10 cm length) for atomic absorption measurements. Measurements were performed with an AAS 6 Vario Atomic absorption spectrometer (Analytik Jena, Jena, Germany) equipped with a deuterium background corrector. An Hg hollow cathode lamp (Narva G.L.E., Berlin, Germany) was used as line source, operating at 5 mA and slit of 1.2 nm, and the wavelength was set at 253.7 nm. Measurements were available in peak height mode. A schematic of the system is illustrated in Fig. 1. 2.2. Reagents and samples Analytical grade reagents were used for all experiments. The inorganic working standards were daily prepared by dilution of the 1000 mg L−1 Hg stock solution (Mercury (II) nitrate, Merck, Darmstadt, Germany) with distilled, deionized water with a specific resistivity of 18 MΩ.cm obtained from a Milli-Q water purification system (Millipore, Bedford, MA, USA). The organic working standards were prepared by dilution of the 100 μg g−1 Hg stock solution (Mercury Alkyl Dithiocarbamate, Conostan, Ponca City, OK, USA) with propan-1-ol (Merck). Formic acid 85% (v/v) (Synth, Diadema, SP, Brazil) was investigated as low molecular weight organic precursor (LMWOP) to obtain the photochemical vapor generation. Propan-1-ol and butan-1-ol (J.T. Baker, USA) were used in the microemulsion preparation. Naphtha and petroleum condensates were provided by Brazilian refineries. 2.3. Sample preparation The sample preparation consisted in the formation of 2.0 mL of microemulsion. The adopted composition for the microemulsions was: sample (50% v/v), propan-1-ol (48% v/v) and water (2% v/v). In the microemulsions for calibration the inorganic or organic Hg standards (0.0, 25.0, 50.0, 75.0 and 100.0 μg L−1) were mixed just with propan1-ol. Recovery tests with inorganic and organic Hg standards have been performed for one naphtha and two petroleum condensate samples. In the microemulsion preparation, the microemulsified samples were spiked with 100 and 200 ng of inorganic or organic standards. For all measurements the microemulsions were prepared in triplicate
A. de Jesus et al. / Microchemical Journal 110 (2013) 227–232
229
Fig. 1. Schematic of the experimental PVG system.
(n = 3). The microemulsions were analyzed immediately after the preparation.
3. Results and discussion 3.1. Sample preparation aspects The most appropriate sample preparation for the naphtha and petroleum condensate samples was the formation of detergentless microemulsions. In the microemulsified medium the stabilization of hydrophilic or hydrophobic compounds is either possible, besides the facility of preparation and feasibility of addition of either inorganic or organic standards [24]. In all the detergentless microemulsions the amount of sample (naphtha or petroleum condensate) was 50% (v/v). Initially, the necessity of addition of organic precursors to the microemulsions has been investigated. Some Hg spiked microemulsions were prepared adding 20% (v/v) of organic precursor (formic acid) to 50% (v/v) of the sample spiked with 200 ng of Hg standard and the volume completed at 100% (v/v) with propan-1-ol or butan-1-ol. Formic acid was chosen because it is the most used organic precursor in the analysis by photochemical vapor generation; it is reported that this acid presents a fast decomposition under UV irradiation, which gives rise to the formation of reducing species more easily using short UV irradiation times [25,26]. Tests without addition of formic acid and also without the UV irradiation were carried out. The absorbance signals obtained for the spiked Hg added to naphtha microemulsions are presented in Fig. 2. It can be seen that without the use of formic acid (experiments C and D) there were an increase in the absorbance signals. In the literature the use of alcohols as organic precursor was already described [27] instead of the low molecular weight organic acids, as formic acid. Li et al. [23] proposed a method for determination of Hg in wine and liquor samples without the use of organic precursor, associating the generated Hg to the ethanol present in these samples. In this work, propan-1-ol and butan-1-ol, which were used in the preparation of the microemulsion, could be used as organic precursors for the photochemical vapor generation of Hg from the samples. This result was very useful because the sample preparation could be simplified. After this experiment propan-1-ol was adopted as organic precursor in the following measurements. Butan-1-ol can also be used with almost similar results. As expected, no signal can be detected without the use of UV irradiation (experiment E). The microemulsion adopted, containing 50% (v/v) sample, 48% (v/v) propan-1-ol and 2% (v/v) water, was stable for at least 24 h. The use of mineral acids, that may increase the stability of the analyte in the microemulsion [1,3,4,12,14], was not used in this work because some authors [28,29] observed that the presence of mineral acids (HNO3 and HCl), even at low concentration, drastically suppresses vapor generation by radical formation. For this reason, the stability of the analyte in the microemulsion was not investigated, and the microemulsions were analyzed immediately after the preparation, as described above.
3.2. Optimization of PVG conditions The conditions of PVG have been optimized with microemulsions of naphtha samples spiked with 200 ng Hg. The first parameter optimized was the sample flow rate. By adjusting the sample flow rate, different residence times of the sample in the reactor were achieved. As can be seen in the Fig. 3a, the lower is the flow rate the higher is the absorbance signal in the investigated range, probably due to the higher exposition time of the sample to UV radiation. However a sample flow rate of 3.5 mL min − 1 (4 min of analysis) was chosen instead of 2.4 mL min − 1, taken into account the excessive time of analysis achieved with the last condition (around of 8 min). After the optimization of the sample flow rate, the optimization of the carrier gas flow rate was carried out (Fig. 3b). The range investigated was from the minimum flow rate possible to use with the flow meter (45 mL min−1) up to 138 mL min−1. With the increase of the flow rate the absorbance signal of Hg started to decrease, because the residence time of the atomic cloud in the absorption cell decreased. Using the minimum flow rate, the highest values of absorbance were obtained but in some experiments the solution almost stopped within the transport tube because the flow rate was very low. A little increase in the carrier gas flow rate could avoid this problem. Thus the carrier gas flow rate was set at 58 mL min−1. In this last condition, a better profile of the absorbance signal was also obtained, as shown in Fig. 4. 3.3. Figures of merit All measurements were obtained in peak area and peak height. The sensitivity in peak area was higher than peak height, but the standard deviation was higher in peak area, resulting in a higher limit of detection
Fig. 2. Influence of organic precursors on the absorbance signal of 200 ng Hg added to naphtha microemulsion. All microemulsions contained 50% (v/v) of naphtha and the following reagents: A) 30% propan-1-ol + 20% formic acid; B) 30% butan-1-ol + 20% formic acid; C) 50% propan-1-ol; D) 50% butan-1-ol; E) 50% propan-1-ol (measurement carried out without UV radiation). Total volume of microemulsion: 2.0 mL.
230
A. de Jesus et al. / Microchemical Journal 110 (2013) 227–232
Fig. 3. Effect of (a) sample flow rate, and (b) carrier gas flow rate, on the absorbance signals of 200 ng Hg added to naphtha. Microemulsion composition: 50% (v/v) naphtha, 2% (v/v) water and 48% (v/v) propan-1-ol.
in peak area mode. Due to these observations in this work the evaluation of signal in peak height was adopted for all the quantification measurements. Figures of merit for the determination of Hg in naphtha and petroleum condensate by PVG-AAS using the optimized conditions are summarized in Table 1. The analytical figures of merit did not differ significantly for calibration accomplished with inorganic or organic Hg standards. This result suggests that there is no difference between the photochemical generation of Hg from different forms. The comparison of the results obtained in this work with the few works found in the literature for the determination of Hg in naphtha and petroleum condensate showed that: the LODs obtained here were better than the results obtained by GF AAS by Ceccarelli (32 μg L−1) [11]; similar to the values obtained by Santos et al. [10] using pre-concentration by multiple injections in GFAAS (0.78 μg L−1); but the LODs obtained by Kumar et al. [9] with ICP-MS (0.12 μg L−1), using semiquantitative approach, were better. In relation to the results obtained specifically by PVG for Hg determination in other kinds of samples, Vieira et al. [21] obtained LODs of 0.03 μg L−1 (instrumental LOD) and 6 ng g−1 for biological samples. Madden et al. obtained LOD of 0.12 μg L− 1 for aqueous standards using the trapping of Hg in palladium coated graphite furnace [30]. Hou et al. [26] and Li et al. [23] using detection based on atomic fluorescence spectrometry (AFS)
obtained LODs of 0.01 μg L−1 for biological samples and 0.07 μg L−1 for wines and liquors, respectively. Generation efficiency was estimated from a comparison of CV-AAS response using a conventional NaBH4 chemical reduction method with the results obtained using PVG-AAS. From a comparison of the slopes of the calibration curves (A = 0.0023 m + 0.003 for CV-AAS and A = 0.0015 m − 0.0038 for PVG-AAS), the efficiency of PVG-AAS with UV irradiation was determined to be approximately 70% of that obtained by conventional CV-AAS. 3.4. Analytical applications Due to the inexistence of certified reference materials for naphtha and petroleum condensates, or in a suitable matrix such as xylene, toluene, light oils, the accuracy of the proposed method was assessed by recovery tests, by spiking three samples with inorganic and organic Hg in two different concentrations, as shown in Table 2. Additionally, no available standard method for Hg determination in naphtha or petroleum condensate, which could be used as a comparative method, was found. Table 2 shows the results obtained for the determination of Hg in 3 naphtha and 5 petroleum condensate samples using calibration curves prepared with inorganic standard in propan-1-ol. As can be seen, the precision of the method was very good, as the mean relative
Fig. 4. Absorbance signals of 200 ng Hg in naphtha by PVG-AAS in different carrier gas flow rates: (a) 100 mL min−1; (b) 56 mL min−1; microemulsion composition: 50% (v/v) naphtha, 2% (v/v) water and 48% (v/v) propan-1-ol; total volume: 2 mL. Solid line: atomic absorption; dotted line: background absorption.
A. de Jesus et al. / Microchemical Journal 110 (2013) 227–232
231
Table 1 Figures of merit for the determination of Hg in naphtha and petroleum condensate as detergentless microemulsions by PVG-AAS. Hg standard
Linear regressiona
R
LODb (μg L−1)
LOQb (μg L−1)
m0 (ng)
Inorganic Organic
A = 0.0015 m − 0.0038 A = 0.0017 m + 0.0095
0.9971 0.9975
0.6 0.5
2.1 1.8
2.4 2.0
a b
A = absorbance (in peak height); m = mass of mercury (ng). n = 10 and calculated considering the analysis of 1.0 mL of sample in 2.0 mL of microemulsion.
standard deviations of the measurements of real and spiked samples were lower than 4%. The spike recoveries obtained (92 to 113%) indicate the absence of matrix effects in the determination of Hg in naphtha and petroleum condensate. Considering the LOD of the method, Hg was found only in the petroleum condensates. This result is in agreement with the statement discussed in the Introduction, that the petroleum condensate may contain higher concentration of metals since it is a heavier fraction of petroleum compared to naphtha. 4. Conclusion In this work the first photochemical vapor generation method applied for the determination of Hg in petroleum derivatives was developed. The formation of volatile compounds of Hg from the microemulsions without the use of organics acids was possible. The use of microemulsions as sample preparation has showed to be very useful because the propan-1-ol used for microemulsion preparation, worked also as organic precursor, besides it could solubilize the standards added for calibration. There was no difference between the photochemical vapor generation from the different species of the analyte used (inorganic and organic). An important aspect of this work was the fact that it was possible to generate analytes from microemulsified samples, avoiding the digestion procedures that are particularly difficult due to the volatile nature of both Hg and the matrix, which may lead to significant loss of the analyte. Besides, the use of a reduced amount of reagents contributed to green chemistry. It might be expected that this method could be applied for analysis of other similar petrochemical samples. Acknowledgments The authors are grateful to CNPq (Projeto Universal-Processo no. 476448/2009-3) and to FAPERGS/CNPq (Programa PRONEX-process no. 10/0012-7) for the financial support. A.J. and A.V.Z. have scholarships
Table 2 Determination of mercury in naphtha and petroleum condensate samples by PVG-AAS and recovery tests.a Sample
Found (μg L−1)
Spike with Hg
Hg added (ng)
Recovery (%)
Naphtha 01 Naphtha 02
bLOD bLOD
– Inorganic
– 100 200 100 200 – – – – 100 200 100 200 100 200 100 200
– 100 ± 4 110 ± 3 104 ± 9 101 ± 1 – – – – 92 ± 2 96 ± 4 113 ± 5 109 ± 3 101 ± 3 96 ± 3 102 ± 3 99 ± 4
Organic Naphtha 03 Petroleum condensate Petroleum condensate Petroleum condensate Petroleum condensate
01 02 03 04
bLOD 76 ± 1 78 ± 1 105 ± 5 bLOD
– – – – Inorganic Organic
Petroleum condensate 05
bLOD
Inorganic Organic
a Results expressed as mean and standard deviation of three independent microemulsions of the same sample (n = 3).
from CAPES and M.A.V., A.S.R., and M.M.S. have scholarships from CNPq. We also thank CECOM (Centro de Combustíveis, Biocombustíveis, Lubrificantes e Óleos do IQ-UFRGS) for technical support. References [1] D.M. Brum, C.F. Lima, N.F. Robaina, T.C.O. Fonseca, R.J. Cassella, Multiple response optimization for Cu, Fe and Pb determination in naphtha by graphite furnace atomic absorption spectrometry with sample injection as detergent emulsion, Spectrochim. Acta B 66 (2011) 338–344. [2] F.S. Dias, W.N.L. Santos, A.C.S. Costa, B. Welz, M.G.R. Vale, S.L.C. Ferreira, Application of multivariate techniques for optimization of direct method for determination of lead in naphtha and petroleum condensate by electrothermal atomic absorption spectrometry, Microchim. Acta 158 (2007) 321–326. [3] R.J. Cassella, B.A.R.S. Barbosa, R.E. Santelli, A.T. Rangel, Direct determination of arsenic and antimony in naphtha by electrothermal atomic absorption spectrometry with microemulsion sample introduction and iridium permanent modifier, Anal. Bioanal. Chem. 379 (2004) 66–71. [4] I.A. Silva, R.C. Campos, A.J. Curtius, S.M. Sella, Determination of lead and copper in kerosene by electrothermal atomic absorption spectrometry: stabilization of metals in organic media by a three-component solution, J. Anal. At. Spectrom. 8 (1993) 749–754. [5] S.M. Wilhelm, Mercury in petroleum and natural gas: estimation of emissions from production, processing, and combustion, Technical Report, US EPA-600/R-01/066, 2001. [6] Intergovernmental negotiating committee to prepare a global legally binding instrument on mercury, Releases of mercury from the oil and gas industry, Technical Report, UNEP (DTIE)/Hg/INC.3/5, 2011. [7] P.J.H. Carnell, A. Foster, J. Gregory, Distribution of mercury in gas processing plants, reprinted from: Hydrocarbon Engineering (2005). assessed from: http://www. jmcatalysts.com/ptd/pdfs-uploaded/Mercury%20matters%20Dec%2005.pdf. [8] S.D. Olsen, S. Werterlund, R.G. Visser, Analysis of metals in condensates and naphtha by inductively coupled plasma mass spectrometry, Analyst 122 (1997) 1229–1234. [9] S.J. Kumar, S. Gangdharran, Determination of trace elements in naphtha by inductively coupled plasma mass spectrometry using water-in-oil emulsions, J. Anal. At. Spectrom. 14 (1999) 967–971. [10] W.N.L. Santos, F.S. Dias, M.S. Fernandes, M.V. Reboucas, M.G. Pereira, V.A. Lemos, L.S.G. Teixeira, Mercury determination in petroleum products by electrothermal atomic absorption spectrometry after in situ preconcentration using multiple injections, J. Anal. At. Spectrom. 21 (2006) 1327–1330. [11] C. Ceccarelli, A.R. Picón, M. Paolini, E.D. Greaves, Total mercury determination in naphthas by either atomic fluorescence or absorption spectroscopy, Petrol. Sci. Technol. 18 (2000) 1055–1075. [12] G.P. Brandão, R.C. Campos, E.V.R. Castro, H.C. Jesus, Determination of manganese in diesel, gasoline and naphtha by graphite furnace atomic absorption spectrometry using microemulsion medium for sample stabilization, Spectrochim. Acta B 63 (2008) 880–884. [13] N.N. Meeravali, S.J. Kumar, The utility of a W–Ir permanent chemical modifier for the determination of Ni and V in emulsified fuel oils and naphtha by transverse heated electrothermal atomic absorption spectrometer, J. Anal. At. Spectrom. 16 (2001) 527–532. [14] E. Becker, R.T. Rampazzo, M.B. Dessuy, M.G.R. Vale, M.M. Silva, B. Welz, D.A. Katskov, Direct determination of arsenic in petroleum derivatives by graphite furnace atomic absorption spectrometry: a comparison between filter and platform atomizers, Spectrochim. Acta B 66 (2011) 345–351. [15] X. Guo, R.E. Sturgeon, Z. Mester, G.J. Gardner, UV Vapor generation for determination of selenium by heated quartz tube atomic absorption spectrometry, Anal. Chem. 75 (2003) 2092–2099. [16] J. Dedina, D. Tsalev, Hydride Generation Atomic Absorption Spectrometry, John Wiley & Sons, New York, 1995. [17] D.P. Torres, I.M. Dittert, H. Höhn, V.L.A. Frescura, A.J. Curtius, Determination of mercury in gasoline diluted in ethanol by GF AAS after cold vapor generation, pre-concentration in gold column and trapping on graphite tube, Microchem. J. 96 (2010) 32–36. [18] G.P. Brandão, R.C. Campos, A.S. Luna, Determination of mercury in gasoline by cold vapor atomic absorption spectrometry with direct reduction in microemulsion media, Spectrochim. Acta B 60 (2005) 625–631. [19] Y. He, X. Hou, C. Zheng, R.E. Sturgeon, Critical evaluation of the application of photochemical vapor generation in analytical atomic spectrometry, Anal. Bioanal. Chem. 388 (2007) 769–774. [20] A. López-Rouco, E. Stanisz, H. Matusiewicz, I. Lavilla, C. Bendicho, UV reduction with ultrasound-assisted gas–liquid separation for the determination of mercury in biotissues by atomic absorption spectrometry, J. Anal. At. Spectrom. 23 (2008) 1026–1029.
232
A. de Jesus et al. / Microchemical Journal 110 (2013) 227–232
[21] M.A. Vieira, A.S. Ribeiro, A.J. Curtius, R.E. Sturgeon, Determination of total mercury and methylmercury in biological samples by photochemical vapor generation, Anal. Bioanal. Chem. 388 (2007) 837–847. [22] C.S. Silva, E.Q. Oreste, A.M. Nunes, M.A. Vieira, A.S. Ribeiro, Determination of mercury in ethanol biofuel by photochemical vapor generation, J. Anal. At. Spectrom. 27 (2012) 689–694. [23] Y. Li, C.B. Zheng, Q. Ma, L. Wu, C.W. Hu, X.D. Hou, Sample matrix-assisted photoinduced chemical vapor generation: a reagent free green analytical method for ultrasensitive detection of mercury in wine or liquor samples, J. Anal. At. Spectrom. 21 (2006) 82–85. [24] A. Jesus, M.M. Silva, M.G.R. Vale, The use of microemulsion for determination of sodium and potassium in biodiesel by flame atomic absorption spectrometry, Talanta 74 (2008) 1378–1384. [25] P. Grinberg, Z. Mester, R.E. Sturgeon, A. Ferretti, Generation of volatile cobalt species by UV photoreduction and their tentative identification, J. Anal. At. Spectrom. 23 (2008) 583–587.
[26] X. Hou, X. Ai, X. Jiang, P. Deng, C. Zheng, Y. Lv, UV light-emitting-diode photochemical mercury vapor generation for atomic fluorescence spectrometry, Analyst 137 (2012) 686–690. [27] C. Han, C. Zheng, J. Wang, G. Cheng, Y. Lv, X. Hou, Photo-induced cold vapor generation with low molecular weight alcohol, aldehyde, or carboxylic acid for atomic fluorescence spectrometric determination of mercury, Anal. Bioanal. Chem. 388 (2007) 825–830. [28] A.S. Ribeiro, M.A. Vieira, S. Willie, A.J. Curtius, R.E. Sturgeon, Ultrasound-assisted vapor generation of mercury, Anal. Bioanal. Chem. 388 (2007) 849–857. [29] P. Grinberg, R.E. Sturgeon, Ultra-trace determination of iodine in sediments and biological material using UV photochemical generation-inductively coupled plasma mass spectrometry, Spectrochim. Acta B 64 (2009) 235–241. [30] J.T. Madden, N. Fitzgerald, Investigation of ultraviolet photolysis vapor generation with in-atomizer trapping graphite furnace atomic absorption spectrometry for the determination of mercury, Spectrochim. Acta B 64 (2009) 925–927.
Contents lists available at ScienceDirect
Microchemical Journal journal homepage: www.elsevier.com/locate/microc
Determination of mercury in naphtha and petroleum condensate by photochemical vapor generation atomic absorption spectrometry☆ Alexandre de Jesus a, Ariane Vanessa Zmozinski a, Mariana Antunes Vieira b, Anderson Schwingel Ribeiro b, Márcia Messias da Silva a, c,⁎ a
Instituto de Química, Universidade Federal do Rio Grande do Sul, Av. Bento Gonçalves, 9500, 91501-970 Porto Alegre, RS, Brazil Universidade Federal de Pelotas, Centro de Ciências Químicas, Farmacêuticas e de Alimentos, Programa de Pós-Graduação em Química, Laboratório de Metrologia Química, 96160-000 Capão do Leão, RS, Brazil c Instituto Nacional de Ciência e Tecnologia do CNPq, INCT de Energia e Ambiente, UFBA, CEP 40170-115, Salvador, BA, Brazil b
a r t i c l e
i n f o
Article history: Received 15 January 2013 Received in revised form 25 March 2013 Accepted 27 March 2013 Available online 10 April 2013 Keywords: Mercury Naphtha Petroleum condensate Photochemical vapor generation Microemulsion AAS
a b s t r a c t In this work the feasibility of mercury determination in naphtha and petroleum condensate by photochemical vapor generation was investigated. The samples were pumped through the photochemical reactor as microemulsions and the volatile compounds formed were conducted to a quartz cell for the atomic absorption measurements. All the parameters of the system (sample preparation, organic precursor, sample flow rate, carrier gas flow rate and evaluation of the signal measurement mode) were investigated. The microemulsions were prepared mixing the samples with propan-1-ol and a small amount of water. The addition of low molecular weight organic acid was investigated and it was found not necessary once the propan-1-ol itself was efficient for promoting the generation of the volatile compounds of mercury. Calibration curves obtained with organic and inorganic standards showed correlation coefficients higher than 0.99 and characteristic mass of 2.0 and 2.4 ng of mercury for organic and inorganic standards, respectively, was obtained. There was no significant difference between the sensitivity of inorganic or organic standards for the calibration. Relative standard deviations ranged from 1 to 5% for three consecutive measurements. The limit of detection of 0.6 μg L−1, calculated for the amount of sample used in the microemulsion (1.0 mL), was obtained. A sample throughput was 2 samples per hour, considering triplicate analysis. Different naphtha and petroleum condensate samples were analyzed. No mercury was found in the naphtha samples, considering the limit of detection. In the petroleum condensate samples the concentrations ranged from 76 to 105 μg L−1. The accuracy was evaluated by assessing the recoveries of inorganic and organic species of mercury added to the samples, being obtained values in the range of 92 to 113%. The developed method was simple and fast, allowed direct analysis of naphtha and petroleum condensate with a reduced amount of reagents, thus contributing to green chemistry. © 2013 Elsevier B.V. All rights reserved.
1. Introduction Naphtha and petroleum condensate are derived products from crude oil and they are important raw materials for the petrochemical industry. Naphtha is a very important feedstock for the fine chemical industries, used in gasoline formulation and in the production of ethylene, propene, benzene, toluene and xylenes [1]. Petroleum condensate replaces naphtha as raw material in the petrochemical industry; it is a heavier fraction of petroleum compared to naphtha and may contain higher concentration of metals [2]. The presence of trace elements, even in μg L −1, in these petroleum-derived products can be associated
☆ Paper presented at 5th Ibero-American Congress of Analytical Chemistry 2012. ⁎ Corresponding author at: Instituto de Química, Universidade Federal do Rio Grande do Sul, Av. Bento Gonçalves, 9500, 91501-970 Porto Alegre, RS, Brazil. Tel.: +55 51 3308 6278; fax: +55 51 3308 7304. E-mail address: [email protected] (M.M. da Silva). 0026-265X/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.microc.2013.03.019
with many distinct problems such as: catalyzation of oxidative reactions, leading to a loss of the thermal stability of the products; release of some elements to the atmosphere in the refining process; corrosion problems and catalyst poisoning (in the case of naphtha); and also the presence of toxic elements remained in the distilled fractions, can be released to the environment when they are used as energy source [3,4]. Mercury is a trace component of all fossil fuels including natural gas, gas condensates, crude oil, coal, tar sands and other bitumens. The use of fossil hydrocarbons as fuels provides the main opportunity for emissions of the Hg they contain to the atmospheric environment [5,6]. Mercury is also a particular problem in the processing since it reacts with metallic surfaces causing corrosion and is a poison for noble metal catalyst used in many hydrocarbon process reactions. Many operators set tight limits on the level of Hg in naphtha feed to crackers (typically less than 5 μg L−1) [7]. Thus monitoring the presence of Hg in these products is increasingly demanded.
228
A. de Jesus et al. / Microchemical Journal 110 (2013) 227–232
The main problems of the determination of Hg in these samples are the fact that they are very complex and Hg is found in very low levels, requiring the use of highly sensitive techniques. There are few works in the literature concerning Hg determination in these samples. The direct analysis of naphtha and petroleum condensate after dilution of the samples with xylene by ICP-MS was proposed by Olsen et al. [8]. These authors reported that samples could be analyzed directly without pretreatment and the elements were determined simultaneously. However the standard deviation for Hg was very high, probably due to the instability of this element in organic solvent and possibly some memory effects. The emulsification of the samples with Triton X-100 was also proposed: it was reported that with the formation of emulsions a stability of at least one hour was obtained [9]. With the use of GF AAS, which is a technique adequate for complex matrices, and using Pd solution as chemical modifier, Santos et al. [10] used detergentless microemulsion as sample preparation and pre-concentrated Hg by multiple injections in the graphite tube obtaining a limit of detection of 0.78 μg L −1. Ceccarelli et al. [11] analyzed naphtha samples directly in the graphite tube for Hg determination obtaining a limit of detection of 32 μg kg −1. The formation and analysis of emulsions or microemulsions have been successfully applied for the preparation of oil samples, due to the homogeneous dispersion and stabilization of the oil microdroplets in the aqueous phase, which brings the viscosity close to that of an aqueous solution and reduces the organic load of the system. Besides, it allows the use of aqueous standards for calibration instead of expensive and instable organometallic standards. This technique has been applied for determination of other elements in petroleum derivatives, as naphtha and petroleum condensate [1,3,12–14]. The use of methods based on vapor generation as sample introduction can offer important advantages for analysis of complex matrices, such as petroleum derivatives. With the previous separation between the analyte and the matrix and efficiency in sample introduction, great improvement in the limit of detection can be achieved and some serious spectroscopic and/or matrix interferences eliminated [15]. The main vapor generation methods (as cold vapor and hydride generation) normally make use of expensive and instable reagents, as tetrahydroborate (NaBH4). Other disadvantages are that few elements can be converted to volatile compounds, the presence of transition metals (as Ni, Co and Cu) can cause the decomposition of the analyte hydride and the main reductor used, NaBH4, is a potential source of contamination [16]. Besides, the complete digestion of these organic samples is frequently required. There are a few works about determination of Hg in petroleum derivates (gasoline samples) using chemical vapor generation without digestion procedures [17,18]. Due to all the aspects cited the development of alternative vapor generation systems for the determination of trace elements in complex samples is still necessary in atomic spectrometry. An alternative method based on photochemical vapor generation (PVG) was proposed [15]. This method consists in the formation of analyte volatile compounds from the sample by the interaction between an organic precursor added to the sample and the ultraviolet light from the photochemical reactor. Then, the obtained volatile compounds are separated from the residual solution and carried to the detection system. In 2007, He et al. [19] published a critical evaluation of the applications of PVG in atomic spectrometry and discussion about the advances of the technique. The authors emphasized several advantages of the method such as no need of using relatively expensive and/or unstable reagents such as tetrahydroborate, that makes the method simpler, greener and in most cases cost-effective, besides the expanded number of detectable elements. Moreover, due to the higher tolerance to interferences, a drastic digestion of the sample is often unnecessary, as in the case of Hg determination in biological tissues solubilized with formic acid [20,21] or tetramethylammonium hydroxide [21]. Other examples are the direct determination of Hg in ethanol fuel [22] and wine and liquors [23].
In ten years of existence, several investigations have been done that help the understanding of the process occurring during photochemical vapor generation. However, there are still few applications to real samples. Based on the best of our knowledge, the use of PVG for determination of Hg in petroleum derivatives has not been published yet. The present work proposes the use of the photochemical vapor generation coupled with atomic absorption spectrometry for the direct determination of Hg in naphtha and petroleum condensate samples. The objective is to obtain a simple, fast and accurate method using detergentless microemulsions as sample preparation, avoiding digestion procedures. 2. Experimental 2.1. Instrumentation A photochemical reactor, consisting of spiral quartz tubing (115 cm × 1.3 mm i.d × 3.0 mm o.d, Internal volume: 2.3 mL) wrapped around of a low-pressure Hg vapor UV lamp (254 nm, 18 W, TUV PL-L, Philips, Poland), was constructed. Samples were propelled through the photochemical reactor with the assistance of a Reglo peristaltic pump (Ismatec, Switzerland). A flow of nitrogen (Air liquide, Canoas, Brazil) was introduced in the line after the photochemical reactor to carry the volatile species; the flow rate was controlled by a flow meter (Cole Parmer, Vernon Hills, IL, USA). The volatile species formed mixed with the nitrogen were introduced in a gas–liquid separator to separate the residual waste from the gaseous phase. After the separation, the volatile species were transported to a quartz tube (10 mm i.d × 10 cm length) for atomic absorption measurements. Measurements were performed with an AAS 6 Vario Atomic absorption spectrometer (Analytik Jena, Jena, Germany) equipped with a deuterium background corrector. An Hg hollow cathode lamp (Narva G.L.E., Berlin, Germany) was used as line source, operating at 5 mA and slit of 1.2 nm, and the wavelength was set at 253.7 nm. Measurements were available in peak height mode. A schematic of the system is illustrated in Fig. 1. 2.2. Reagents and samples Analytical grade reagents were used for all experiments. The inorganic working standards were daily prepared by dilution of the 1000 mg L−1 Hg stock solution (Mercury (II) nitrate, Merck, Darmstadt, Germany) with distilled, deionized water with a specific resistivity of 18 MΩ.cm obtained from a Milli-Q water purification system (Millipore, Bedford, MA, USA). The organic working standards were prepared by dilution of the 100 μg g−1 Hg stock solution (Mercury Alkyl Dithiocarbamate, Conostan, Ponca City, OK, USA) with propan-1-ol (Merck). Formic acid 85% (v/v) (Synth, Diadema, SP, Brazil) was investigated as low molecular weight organic precursor (LMWOP) to obtain the photochemical vapor generation. Propan-1-ol and butan-1-ol (J.T. Baker, USA) were used in the microemulsion preparation. Naphtha and petroleum condensates were provided by Brazilian refineries. 2.3. Sample preparation The sample preparation consisted in the formation of 2.0 mL of microemulsion. The adopted composition for the microemulsions was: sample (50% v/v), propan-1-ol (48% v/v) and water (2% v/v). In the microemulsions for calibration the inorganic or organic Hg standards (0.0, 25.0, 50.0, 75.0 and 100.0 μg L−1) were mixed just with propan1-ol. Recovery tests with inorganic and organic Hg standards have been performed for one naphtha and two petroleum condensate samples. In the microemulsion preparation, the microemulsified samples were spiked with 100 and 200 ng of inorganic or organic standards. For all measurements the microemulsions were prepared in triplicate
A. de Jesus et al. / Microchemical Journal 110 (2013) 227–232
229
Fig. 1. Schematic of the experimental PVG system.
(n = 3). The microemulsions were analyzed immediately after the preparation.
3. Results and discussion 3.1. Sample preparation aspects The most appropriate sample preparation for the naphtha and petroleum condensate samples was the formation of detergentless microemulsions. In the microemulsified medium the stabilization of hydrophilic or hydrophobic compounds is either possible, besides the facility of preparation and feasibility of addition of either inorganic or organic standards [24]. In all the detergentless microemulsions the amount of sample (naphtha or petroleum condensate) was 50% (v/v). Initially, the necessity of addition of organic precursors to the microemulsions has been investigated. Some Hg spiked microemulsions were prepared adding 20% (v/v) of organic precursor (formic acid) to 50% (v/v) of the sample spiked with 200 ng of Hg standard and the volume completed at 100% (v/v) with propan-1-ol or butan-1-ol. Formic acid was chosen because it is the most used organic precursor in the analysis by photochemical vapor generation; it is reported that this acid presents a fast decomposition under UV irradiation, which gives rise to the formation of reducing species more easily using short UV irradiation times [25,26]. Tests without addition of formic acid and also without the UV irradiation were carried out. The absorbance signals obtained for the spiked Hg added to naphtha microemulsions are presented in Fig. 2. It can be seen that without the use of formic acid (experiments C and D) there were an increase in the absorbance signals. In the literature the use of alcohols as organic precursor was already described [27] instead of the low molecular weight organic acids, as formic acid. Li et al. [23] proposed a method for determination of Hg in wine and liquor samples without the use of organic precursor, associating the generated Hg to the ethanol present in these samples. In this work, propan-1-ol and butan-1-ol, which were used in the preparation of the microemulsion, could be used as organic precursors for the photochemical vapor generation of Hg from the samples. This result was very useful because the sample preparation could be simplified. After this experiment propan-1-ol was adopted as organic precursor in the following measurements. Butan-1-ol can also be used with almost similar results. As expected, no signal can be detected without the use of UV irradiation (experiment E). The microemulsion adopted, containing 50% (v/v) sample, 48% (v/v) propan-1-ol and 2% (v/v) water, was stable for at least 24 h. The use of mineral acids, that may increase the stability of the analyte in the microemulsion [1,3,4,12,14], was not used in this work because some authors [28,29] observed that the presence of mineral acids (HNO3 and HCl), even at low concentration, drastically suppresses vapor generation by radical formation. For this reason, the stability of the analyte in the microemulsion was not investigated, and the microemulsions were analyzed immediately after the preparation, as described above.
3.2. Optimization of PVG conditions The conditions of PVG have been optimized with microemulsions of naphtha samples spiked with 200 ng Hg. The first parameter optimized was the sample flow rate. By adjusting the sample flow rate, different residence times of the sample in the reactor were achieved. As can be seen in the Fig. 3a, the lower is the flow rate the higher is the absorbance signal in the investigated range, probably due to the higher exposition time of the sample to UV radiation. However a sample flow rate of 3.5 mL min − 1 (4 min of analysis) was chosen instead of 2.4 mL min − 1, taken into account the excessive time of analysis achieved with the last condition (around of 8 min). After the optimization of the sample flow rate, the optimization of the carrier gas flow rate was carried out (Fig. 3b). The range investigated was from the minimum flow rate possible to use with the flow meter (45 mL min−1) up to 138 mL min−1. With the increase of the flow rate the absorbance signal of Hg started to decrease, because the residence time of the atomic cloud in the absorption cell decreased. Using the minimum flow rate, the highest values of absorbance were obtained but in some experiments the solution almost stopped within the transport tube because the flow rate was very low. A little increase in the carrier gas flow rate could avoid this problem. Thus the carrier gas flow rate was set at 58 mL min−1. In this last condition, a better profile of the absorbance signal was also obtained, as shown in Fig. 4. 3.3. Figures of merit All measurements were obtained in peak area and peak height. The sensitivity in peak area was higher than peak height, but the standard deviation was higher in peak area, resulting in a higher limit of detection
Fig. 2. Influence of organic precursors on the absorbance signal of 200 ng Hg added to naphtha microemulsion. All microemulsions contained 50% (v/v) of naphtha and the following reagents: A) 30% propan-1-ol + 20% formic acid; B) 30% butan-1-ol + 20% formic acid; C) 50% propan-1-ol; D) 50% butan-1-ol; E) 50% propan-1-ol (measurement carried out without UV radiation). Total volume of microemulsion: 2.0 mL.
230
A. de Jesus et al. / Microchemical Journal 110 (2013) 227–232
Fig. 3. Effect of (a) sample flow rate, and (b) carrier gas flow rate, on the absorbance signals of 200 ng Hg added to naphtha. Microemulsion composition: 50% (v/v) naphtha, 2% (v/v) water and 48% (v/v) propan-1-ol.
in peak area mode. Due to these observations in this work the evaluation of signal in peak height was adopted for all the quantification measurements. Figures of merit for the determination of Hg in naphtha and petroleum condensate by PVG-AAS using the optimized conditions are summarized in Table 1. The analytical figures of merit did not differ significantly for calibration accomplished with inorganic or organic Hg standards. This result suggests that there is no difference between the photochemical generation of Hg from different forms. The comparison of the results obtained in this work with the few works found in the literature for the determination of Hg in naphtha and petroleum condensate showed that: the LODs obtained here were better than the results obtained by GF AAS by Ceccarelli (32 μg L−1) [11]; similar to the values obtained by Santos et al. [10] using pre-concentration by multiple injections in GFAAS (0.78 μg L−1); but the LODs obtained by Kumar et al. [9] with ICP-MS (0.12 μg L−1), using semiquantitative approach, were better. In relation to the results obtained specifically by PVG for Hg determination in other kinds of samples, Vieira et al. [21] obtained LODs of 0.03 μg L−1 (instrumental LOD) and 6 ng g−1 for biological samples. Madden et al. obtained LOD of 0.12 μg L− 1 for aqueous standards using the trapping of Hg in palladium coated graphite furnace [30]. Hou et al. [26] and Li et al. [23] using detection based on atomic fluorescence spectrometry (AFS)
obtained LODs of 0.01 μg L−1 for biological samples and 0.07 μg L−1 for wines and liquors, respectively. Generation efficiency was estimated from a comparison of CV-AAS response using a conventional NaBH4 chemical reduction method with the results obtained using PVG-AAS. From a comparison of the slopes of the calibration curves (A = 0.0023 m + 0.003 for CV-AAS and A = 0.0015 m − 0.0038 for PVG-AAS), the efficiency of PVG-AAS with UV irradiation was determined to be approximately 70% of that obtained by conventional CV-AAS. 3.4. Analytical applications Due to the inexistence of certified reference materials for naphtha and petroleum condensates, or in a suitable matrix such as xylene, toluene, light oils, the accuracy of the proposed method was assessed by recovery tests, by spiking three samples with inorganic and organic Hg in two different concentrations, as shown in Table 2. Additionally, no available standard method for Hg determination in naphtha or petroleum condensate, which could be used as a comparative method, was found. Table 2 shows the results obtained for the determination of Hg in 3 naphtha and 5 petroleum condensate samples using calibration curves prepared with inorganic standard in propan-1-ol. As can be seen, the precision of the method was very good, as the mean relative
Fig. 4. Absorbance signals of 200 ng Hg in naphtha by PVG-AAS in different carrier gas flow rates: (a) 100 mL min−1; (b) 56 mL min−1; microemulsion composition: 50% (v/v) naphtha, 2% (v/v) water and 48% (v/v) propan-1-ol; total volume: 2 mL. Solid line: atomic absorption; dotted line: background absorption.
A. de Jesus et al. / Microchemical Journal 110 (2013) 227–232
231
Table 1 Figures of merit for the determination of Hg in naphtha and petroleum condensate as detergentless microemulsions by PVG-AAS. Hg standard
Linear regressiona
R
LODb (μg L−1)
LOQb (μg L−1)
m0 (ng)
Inorganic Organic
A = 0.0015 m − 0.0038 A = 0.0017 m + 0.0095
0.9971 0.9975
0.6 0.5
2.1 1.8
2.4 2.0
a b
A = absorbance (in peak height); m = mass of mercury (ng). n = 10 and calculated considering the analysis of 1.0 mL of sample in 2.0 mL of microemulsion.
standard deviations of the measurements of real and spiked samples were lower than 4%. The spike recoveries obtained (92 to 113%) indicate the absence of matrix effects in the determination of Hg in naphtha and petroleum condensate. Considering the LOD of the method, Hg was found only in the petroleum condensates. This result is in agreement with the statement discussed in the Introduction, that the petroleum condensate may contain higher concentration of metals since it is a heavier fraction of petroleum compared to naphtha. 4. Conclusion In this work the first photochemical vapor generation method applied for the determination of Hg in petroleum derivatives was developed. The formation of volatile compounds of Hg from the microemulsions without the use of organics acids was possible. The use of microemulsions as sample preparation has showed to be very useful because the propan-1-ol used for microemulsion preparation, worked also as organic precursor, besides it could solubilize the standards added for calibration. There was no difference between the photochemical vapor generation from the different species of the analyte used (inorganic and organic). An important aspect of this work was the fact that it was possible to generate analytes from microemulsified samples, avoiding the digestion procedures that are particularly difficult due to the volatile nature of both Hg and the matrix, which may lead to significant loss of the analyte. Besides, the use of a reduced amount of reagents contributed to green chemistry. It might be expected that this method could be applied for analysis of other similar petrochemical samples. Acknowledgments The authors are grateful to CNPq (Projeto Universal-Processo no. 476448/2009-3) and to FAPERGS/CNPq (Programa PRONEX-process no. 10/0012-7) for the financial support. A.J. and A.V.Z. have scholarships
Table 2 Determination of mercury in naphtha and petroleum condensate samples by PVG-AAS and recovery tests.a Sample
Found (μg L−1)
Spike with Hg
Hg added (ng)
Recovery (%)
Naphtha 01 Naphtha 02
bLOD bLOD
– Inorganic
– 100 200 100 200 – – – – 100 200 100 200 100 200 100 200
– 100 ± 4 110 ± 3 104 ± 9 101 ± 1 – – – – 92 ± 2 96 ± 4 113 ± 5 109 ± 3 101 ± 3 96 ± 3 102 ± 3 99 ± 4
Organic Naphtha 03 Petroleum condensate Petroleum condensate Petroleum condensate Petroleum condensate
01 02 03 04
bLOD 76 ± 1 78 ± 1 105 ± 5 bLOD
– – – – Inorganic Organic
Petroleum condensate 05
bLOD
Inorganic Organic
a Results expressed as mean and standard deviation of three independent microemulsions of the same sample (n = 3).
from CAPES and M.A.V., A.S.R., and M.M.S. have scholarships from CNPq. We also thank CECOM (Centro de Combustíveis, Biocombustíveis, Lubrificantes e Óleos do IQ-UFRGS) for technical support. References [1] D.M. Brum, C.F. Lima, N.F. Robaina, T.C.O. Fonseca, R.J. Cassella, Multiple response optimization for Cu, Fe and Pb determination in naphtha by graphite furnace atomic absorption spectrometry with sample injection as detergent emulsion, Spectrochim. Acta B 66 (2011) 338–344. [2] F.S. Dias, W.N.L. Santos, A.C.S. Costa, B. Welz, M.G.R. Vale, S.L.C. Ferreira, Application of multivariate techniques for optimization of direct method for determination of lead in naphtha and petroleum condensate by electrothermal atomic absorption spectrometry, Microchim. Acta 158 (2007) 321–326. [3] R.J. Cassella, B.A.R.S. Barbosa, R.E. Santelli, A.T. Rangel, Direct determination of arsenic and antimony in naphtha by electrothermal atomic absorption spectrometry with microemulsion sample introduction and iridium permanent modifier, Anal. Bioanal. Chem. 379 (2004) 66–71. [4] I.A. Silva, R.C. Campos, A.J. Curtius, S.M. Sella, Determination of lead and copper in kerosene by electrothermal atomic absorption spectrometry: stabilization of metals in organic media by a three-component solution, J. Anal. At. Spectrom. 8 (1993) 749–754. [5] S.M. Wilhelm, Mercury in petroleum and natural gas: estimation of emissions from production, processing, and combustion, Technical Report, US EPA-600/R-01/066, 2001. [6] Intergovernmental negotiating committee to prepare a global legally binding instrument on mercury, Releases of mercury from the oil and gas industry, Technical Report, UNEP (DTIE)/Hg/INC.3/5, 2011. [7] P.J.H. Carnell, A. Foster, J. Gregory, Distribution of mercury in gas processing plants, reprinted from: Hydrocarbon Engineering (2005). assessed from: http://www. jmcatalysts.com/ptd/pdfs-uploaded/Mercury%20matters%20Dec%2005.pdf. [8] S.D. Olsen, S. Werterlund, R.G. Visser, Analysis of metals in condensates and naphtha by inductively coupled plasma mass spectrometry, Analyst 122 (1997) 1229–1234. [9] S.J. Kumar, S. Gangdharran, Determination of trace elements in naphtha by inductively coupled plasma mass spectrometry using water-in-oil emulsions, J. Anal. At. Spectrom. 14 (1999) 967–971. [10] W.N.L. Santos, F.S. Dias, M.S. Fernandes, M.V. Reboucas, M.G. Pereira, V.A. Lemos, L.S.G. Teixeira, Mercury determination in petroleum products by electrothermal atomic absorption spectrometry after in situ preconcentration using multiple injections, J. Anal. At. Spectrom. 21 (2006) 1327–1330. [11] C. Ceccarelli, A.R. Picón, M. Paolini, E.D. Greaves, Total mercury determination in naphthas by either atomic fluorescence or absorption spectroscopy, Petrol. Sci. Technol. 18 (2000) 1055–1075. [12] G.P. Brandão, R.C. Campos, E.V.R. Castro, H.C. Jesus, Determination of manganese in diesel, gasoline and naphtha by graphite furnace atomic absorption spectrometry using microemulsion medium for sample stabilization, Spectrochim. Acta B 63 (2008) 880–884. [13] N.N. Meeravali, S.J. Kumar, The utility of a W–Ir permanent chemical modifier for the determination of Ni and V in emulsified fuel oils and naphtha by transverse heated electrothermal atomic absorption spectrometer, J. Anal. At. Spectrom. 16 (2001) 527–532. [14] E. Becker, R.T. Rampazzo, M.B. Dessuy, M.G.R. Vale, M.M. Silva, B. Welz, D.A. Katskov, Direct determination of arsenic in petroleum derivatives by graphite furnace atomic absorption spectrometry: a comparison between filter and platform atomizers, Spectrochim. Acta B 66 (2011) 345–351. [15] X. Guo, R.E. Sturgeon, Z. Mester, G.J. Gardner, UV Vapor generation for determination of selenium by heated quartz tube atomic absorption spectrometry, Anal. Chem. 75 (2003) 2092–2099. [16] J. Dedina, D. Tsalev, Hydride Generation Atomic Absorption Spectrometry, John Wiley & Sons, New York, 1995. [17] D.P. Torres, I.M. Dittert, H. Höhn, V.L.A. Frescura, A.J. Curtius, Determination of mercury in gasoline diluted in ethanol by GF AAS after cold vapor generation, pre-concentration in gold column and trapping on graphite tube, Microchem. J. 96 (2010) 32–36. [18] G.P. Brandão, R.C. Campos, A.S. Luna, Determination of mercury in gasoline by cold vapor atomic absorption spectrometry with direct reduction in microemulsion media, Spectrochim. Acta B 60 (2005) 625–631. [19] Y. He, X. Hou, C. Zheng, R.E. Sturgeon, Critical evaluation of the application of photochemical vapor generation in analytical atomic spectrometry, Anal. Bioanal. Chem. 388 (2007) 769–774. [20] A. López-Rouco, E. Stanisz, H. Matusiewicz, I. Lavilla, C. Bendicho, UV reduction with ultrasound-assisted gas–liquid separation for the determination of mercury in biotissues by atomic absorption spectrometry, J. Anal. At. Spectrom. 23 (2008) 1026–1029.
232
A. de Jesus et al. / Microchemical Journal 110 (2013) 227–232
[21] M.A. Vieira, A.S. Ribeiro, A.J. Curtius, R.E. Sturgeon, Determination of total mercury and methylmercury in biological samples by photochemical vapor generation, Anal. Bioanal. Chem. 388 (2007) 837–847. [22] C.S. Silva, E.Q. Oreste, A.M. Nunes, M.A. Vieira, A.S. Ribeiro, Determination of mercury in ethanol biofuel by photochemical vapor generation, J. Anal. At. Spectrom. 27 (2012) 689–694. [23] Y. Li, C.B. Zheng, Q. Ma, L. Wu, C.W. Hu, X.D. Hou, Sample matrix-assisted photoinduced chemical vapor generation: a reagent free green analytical method for ultrasensitive detection of mercury in wine or liquor samples, J. Anal. At. Spectrom. 21 (2006) 82–85. [24] A. Jesus, M.M. Silva, M.G.R. Vale, The use of microemulsion for determination of sodium and potassium in biodiesel by flame atomic absorption spectrometry, Talanta 74 (2008) 1378–1384. [25] P. Grinberg, Z. Mester, R.E. Sturgeon, A. Ferretti, Generation of volatile cobalt species by UV photoreduction and their tentative identification, J. Anal. At. Spectrom. 23 (2008) 583–587.
[26] X. Hou, X. Ai, X. Jiang, P. Deng, C. Zheng, Y. Lv, UV light-emitting-diode photochemical mercury vapor generation for atomic fluorescence spectrometry, Analyst 137 (2012) 686–690. [27] C. Han, C. Zheng, J. Wang, G. Cheng, Y. Lv, X. Hou, Photo-induced cold vapor generation with low molecular weight alcohol, aldehyde, or carboxylic acid for atomic fluorescence spectrometric determination of mercury, Anal. Bioanal. Chem. 388 (2007) 825–830. [28] A.S. Ribeiro, M.A. Vieira, S. Willie, A.J. Curtius, R.E. Sturgeon, Ultrasound-assisted vapor generation of mercury, Anal. Bioanal. Chem. 388 (2007) 849–857. [29] P. Grinberg, R.E. Sturgeon, Ultra-trace determination of iodine in sediments and biological material using UV photochemical generation-inductively coupled plasma mass spectrometry, Spectrochim. Acta B 64 (2009) 235–241. [30] J.T. Madden, N. Fitzgerald, Investigation of ultraviolet photolysis vapor generation with in-atomizer trapping graphite furnace atomic absorption spectrometry for the determination of mercury, Spectrochim. Acta B 64 (2009) 925–927.