SciELO - Scientific Electronic Library Online

 
vol.11 issue2Synthesis of platinum complexes from N-benzyl ethylenediamine derivativesCriterious preparation and characterization of earthworm-composts in view of animal waste recycling. Part I. Correlation between chemical, thermal and FTIR spectroscopic analyses of four humic acids from earthworm-composted animal manure author indexsubject indexarticles search
Home Pagealphabetic serial listing  

Services on Demand

Journal

Article

Indicators

Related links

Share


Journal of the Brazilian Chemical Society

Print version ISSN 0103-5053On-line version ISSN 1678-4790

J. Braz. Chem. Soc. vol.11 n.2 São Paulo Mar./Apr. 2000

http://dx.doi.org/10.1590/S0103-50532000000200010 

Article

 

Development of a coulometric method for the determination of gaseous sulfur compounds in urban atmospheres

 

Magali Silva and Luis F. P. Dick*

Departamento de Metalurgia, Universidade Federal do Rio Grande do Sul, Av. Osvaldo Aranha, 99/6° andar, 90035-190, Porto Alegre - RS, Brazil

 

 

Uma nova metodologia para a determinação de compostos gasosos de enxofre presentes em atmosferas urbanas foi desenvolvida e aplicada durante três meses na região central da cidade de Porto Alegre- RS e em uma refinaria de petróleo, localizada próxima à essa cidade. O método se baseia na determinação coulométrica da taxa de sulfetação da prata pura. Os resultados foram comparados com o método da pararosanilina, que está sendo atualmente utilizado para a determinação de SO2 em alguns dos locais aqui estudados. Os resultados obtidos mostraram que uma boa correlação entre o método coulométrico e o método da pararosanilina é observada quando somente SO2 está presente, como normalmente é esperado para atmosferas urbanas. Quando H2S está também presente, como no caso das áreas industriais, a nova metodologia tende fortemente a super estimar derivados gasosos de enxofre.

 

A new procedure for the determination of atmospheric sulfur compounds was developed and applied during three months in the central area of the City of Porto Alegre- RS, and one month near an oil refinery located close to this city. The method is based on the coulometric determination of the tarnishing rate of pure silver. The results were compared with the method of pararosaniline, which is being currently used in some of the studied sites for the determination of SO2. The results obtained showed a good correlation between the coulometric method and the pararosaniline method, when only SO2 is present, as usually is the case of urban atmospheres. When H2S is also present, as in the case of industrial areas, the methodology tends to strongly overestimate SO2 concentration.

Keywords: coulometric method, tarnishing of Ag, urban pollution, SO2, H2S

 

 

Introduction

Besides suspended particles, gaseous compounds, such as carbon monoxide (CO), carbon dioxide (CO2) and sulfur dioxide (SO2), are normally the main atmospheric pollutants in urban areas1, 2. Previous studies showed that in Porto Alegre- RS, approximatly 3.7 tons of SOx are daily emitted. Further 3.4 tons of SOx per day are produced in the industrial metropolitan area around the city3.

The main SO2 source in urban areas is commonly accepted as resulting from the use of fossil fuels. In warm regions like Southern Brazil, the main source should be attributed to the transport activity and to the use of Diesel oil. Due to local variations in the traffic density and in the wind velocity in a city, a local monitoring of SO2 is more adequate to identify critical areas, which are harmful to the human health and are also very corrosive to metals, stones and concrete structures.

The aim of the present study was to develop a method for the determination of atmospheric SO2 content. The method should be inexpensive and simple, making possible the local determination of the mean concentration in several points in an urban area. As showed below, the tarnishing rate of silver shows the desired characteristics for an indicator of the SO2 concentration.

Pure silver, when exposed to H2S or SO2, forms a tarnishing layer consisting only of achantite (alpha Ag2S). This sulfide has an extremely low solubility product in water, Kps=1 10-50 mol L-1 3,4 and is very conductive and adherent to the silver surface, enabling its coulometric determination after the atmospheric exposure of silver3, 5. Moreover, due to the high mobility of Ag+ ions in the Ag2S salt, the tarnishing rate is time independent for tarnishing layers up to several microns thick, being directly proportional to the H2S and SO2 concentrations5. As an example, the tarnishing rate of Ag in 1ppm H2S, which is higher than the one in SO2, remains controlled by surface reaction for several days6, 7.

The results presented here were measured during the southern hemisphere summer of 1999 in three urban points of Porto Alegre- RS, showing heavy traffic, where only SO2 is expected, and also in two sites near an oil refinery, with different distances from a H2S source. An increasing SO2/H2S concentration ratio is expected here for greater distances from the H2S source, resulting from the oxidation of H2S to SO2 during its transport from the source to the measuring site. A long-term study is now under way to determine the influence of traffic intensity and climate parameters in the correlation between the coulometric and pararosaniline methods.

 

Experimental

For the coulometric determination of tarnishing rates, coupons of pure silver (Sigma-Aldrich, 99.99% Ag) with geometric areas between 5 and 6 cm2 were cut and provided with a 0.5 mm diameter hole. After vacuum annealing (650 °C h-110-2 atm) the coupons were polished on both sides with emery paper up to 2400 mesh and degreased in ethanol. For the atmospheric exposure, the coupons were hanged by a PTFE thread in the vertical position at five "monitoring stations", where a good ventilation and rain protection were provided, and also other atmospheric data were being collected. Three monitoring stations were located in downtown sites, with dense traffic, Rodoviária, Borges and Azenha. Among these sites, Rodoviária, where the Central Bus Station is located, shows usually the highest SO2 emissions. The other two monitoring stations were located at 19 km far from downtown of Porto Alegre- RS, at the oil refinery "Refinaria Alberto Pasqualine-Petrobras", namely REFAP1 and REFAP2. The station REFAP1 lied 750 m far from an H2S-SO2-source, and REFAP2, respectively 1,500 m. At three of these stations, Rodoviária, REFAP1 and REFAP2, SO2 was also collected and the concentration determined by the pararosaniline method as described in the appropriate norm8. The exposures were proceeded from January to March 1999. The mean relative humidity and temperature during this period were 79.4% and 24.9 ºC respectively, as determined for Porto Alegre- RS by the local state climate station.

After exposure for different times, the Ag coupons were rinsed with water and acetone and reduced potentiodynamically in a standard three electrodes cell. The electrolyte was a borate buffer of pH 10 (3.092 g L-1 H3BO3, 3.728 g L-1 KCl, 1.756 g L-1 NaOH), with addition of small quantities of Na2S and purged previously for 8 h with N2. Good results were obtained for sweep velocities in the negative direction of 0.1 mV s-1, starting from the Ag2S/Ag equilibrium potential. This low sweeping rate was necessary to avoid the superposition of the Ag2S reduction peak and the current related to the hydrogen evolution reaction. The equilibrium potential (Ag2S/Ag) was determined by a previous potentiodynamic sulfidation and reduction of a blank Ag probe in the same solution. The potentials were measured against an AgCl/Ag reference electrode in 3.5 mol L-1 KCl, but are referred in the text to the normal hydrogen electrode (NHE). After data acquisition by computer, the mass of Ag2S per area was calculated from the charge density under the observed cathodic peak, assuming Ag2S as the only tarnishing product and subtracting the charge related to the background current density.

 

Results and Discussion

The voltammograms of the reduction of the exposed Ag coupons are shown in Figures from 1 to 5. A sharp reduction peak in the range between E(Ag2S/Ag) and –200 mV more negative than this, i.e (from –350 to 550 mV) is always observed. The peak current density grows with the exposure time, and it can be attributed undoubtedly to the reduction of the environmentally formed Ag2S. No reduction peaks could be observed for unexposed Ag samples. The background current density, measured on unexposed samples, or estimated on exposed samples as the current minimum between the Ag2S peak and the hydrogen evolution region, ranged between 0.2 and 1µA cm-2, indicating the reduction of residual O2 dissolved in the electrolyte. The coupons exposed at urban sites, i.e. Rodoviária, Borges and Azenha, (Figures 1-3, respectively) showed less pronounced reduction peaks, comparing to the ones exposed at the oil refinery sites, REFAP1 and REFAP2 (Figures 4 and 5). For exposures at the less aggressive sites Borges and Azenha, a clear identification of the reduction peaks, at a potencial sweep rate of 0.1 mV s-1 was only possible for exposure times of ca. 10 days or longer. For the more aggressive sites at the refinery, clear peaks were already identified after seven days of exposure.

 

2a09f1.gif (11086 bytes)

 

2a09f2.gif (11058 bytes)

 

 

2a09f3.gif (11414 bytes)

 

 

2a09f4.gif (12884 bytes)

 

 

2a09f5.gif (13039 bytes)

 

Samples exposed for more then 20 days at the most aggressive sites showed a second reduction peak (Figure 1, 4, 5). Its current density, subtracting the background current, grows with time of exposure and with the aggressivity of the atmosphere, and its current was allways less than 3% of the main Ag2S peak. A second reduction peak was also observed by the reduction of Ag samples tarnished strongly in aqueous sulfide solutions of 0.01mol L-1 Na2S6,7. Only Ag2S could be detected in that case by X-ray diffraction and thus, it is here assumed that only Ag2S is formed atmospherically.

The mass of Ag2S formed vs. the exposure time is shown for all sites in Figure 6. For the determination of the respective tarnishing rates, a linear correlation was found for the mass vs. time curves. The correlation coefficients were normally higher then 0.93, with exception of the Borges site, where R = 0.70 was measured. The aggressivity of the sites and the respective Ag tarnishing rates, lie in the following decrescent order: m/tREFAP1 = 1.00, m/tREFAP2 = 0.34, m/tRodoviária = 0.08, m/tAzenha = 0.06 and m/tBorges = 0.007µg Ag2S cm2 day-1. These results indicate that there are differences of more than one order of magnitude in the tarnishing rates and SO2 concentrations of sites lying only 1.1km apart in downtown Porto Alegre (Borges and Rodoviária). Moreover, for a distance of 2.5 km from the H2S-S02 source at the refinery, the atmosphere is still 4.25 times more aggressive than the most polluted site identified in downtown Porto Alegre, i.e. Rodoviária.

 

2a09f6.gif (7491 bytes)

 

The comparison between the Ag tarnishing rate and the mean SO2 concentration as determined by daily measurements during the same period with the pararosaniline method is presented in Figure 7. Using the data from the two sites where only SO2 is expected (Rodoviária and REFAP2) a ratio between the SO2 concentration and the Ag tarnishing rate of 86±4g SO2 m-3/g Ag2S cm-2 day-1. For the site REFAP1, where H2S is also present, a 4 times smaller ratio is found. This is certainly due to a much higher Ag tarnishing rate in H2S than in SO2. The coulometric method suffers from H2S interference, and no quantitative results can be expected in this case. Even so, it points out that the site REFAP1 has a much more aggressive atmosphere than REFAP2. The pararosaniline method, currently in use, indicated a SO2 concentration 1.5 times higher at the most distant point from the source than at a closer one. This is probably due to the oxidation of H2S to SO2 in the way between the source and REFAP2.

 

2a09f7.gif (7139 bytes)

 

A comparison between tarnishing rates measured in an environmental chamber6, 7 containing 1 ppm H2S at T= 25ºC and the ones determined in this work at a mean relative humidity of 79.4% and temperature of 24.9°C is presented in Figure 8. Assuming that the tarnishing rate is directly proportional to the H2S concentration, an equivalent H2S concentration can be calculated. The values obtained are 0.36 ppm for REFAP1, 0.12 ppm for REFAP2 and 0.03 ppm for Rodoviária.

 

2a09f8.gif (9722 bytes)

 

Conclusions

The coulometric method, based on the determination of the tarnishing rate of Ag, can be used for the determination of SO2 concentrations in urban atmospheres. When H2S is also present, as in industrial areas, an over estimation of the SO2 concentration is expected, as the method can not separate the effect of these two species. At least one week is necessary till the presence of SO2 can be detected in urban sites. Due to its low cost and simplicity, the method is more adequate for the local determination of long term average concentrations of SO2.

 

Acknowledgments

The Secretaria Municipal do Meio Ambiente from the City of Porto Alegre and the company REFAP-PETROBRAS are gratefully acknowledged for technical support. M. Silva thanks CAPES for a fellowship.

 

References

1. Pereira, P. M.; Saldiva, P. H. N.; Sakae, R. S.; Bohm, G. M.; Martins, M. A. Environ. Res. 1995, 69, 96.         [ Links ]

2. Darlington, T. L.; Kahlbaum, D. F.; Heuss, J. M.; Wolff, G. T. J. Air & Waste Manage. Assoc. 1997, 47, 1070.         [ Links ]

3. Refinaria Alberto Pasqualine Reports- Levantamento sobre a contribuição do setor automotivo na geração de poluentes atmosféricos na região metropolitana de Porto Alegre 1996.         [ Links ]

4. Licht, S. J. Electrochem. Soc. , 1988, 135, 2971.         [ Links ]

5. Franey, J. P.; Kammlott, G. W.; Graedel, T. E. Corros. Sci. 1985, 25, 1163.         [ Links ]

6. Dick, L. F. P.; Kaiser, H. Kaesche, The Electrochem. Soc. Softbound Proc. Series, Pennington NJ, 1993, p 406.         [ Links ]

7. Dick, L. F. P.; Kaiser, H. Kaesche, The Electrochem. Soc. Softbound Proc. Series, Pennington NJ, 1991, p 273.         [ Links ]

8. Annual Book of ASTM Standards, 1972, 23, 843.         [ Links ]

 

Received: June 01, 1999

 

* e-mail: lfdick@vortex.ufrgs.br

Presented at the XI Simpósio Brasileiro de Eletroquímica e Eletroanalítica, Maragogi - AL, Brazil, April 5-9, 1999. Guest editor: Luis Alberto Avaca.

Creative Commons License All the contents of this journal, except where otherwise noted, is licensed under a Creative Commons Attribution License