7th International Conference on Mercury as a Global Pollutant_
1921
Optimisation of cold vapour atomic absorption spectrometry for determination of high levels of total mercury in activated carbon
Darija Gibičar1, Vesna Jereb1, Radojko Jačimovič1, Tatjana Dizdarevič2, Elisabeth C. Paiva3, Ricardo Melamed3, Luis Gonzaga S. Sobral3, Zdravko Špirič4, & Milena Horvat1
'Jožef Stefan Institute, Jamova 39, SI-1000 Ljubljana, Slovenia, E-mail: darija.gibicar@ijs.si 2 Idrija Mercury Mine, Idrija, Slovenia, E-mail: tatjana.rzs.idrija@s5.net QCentre for Mineral Technology, Avenida Ip^ 900, Rio de Janeiro, Brazil, E-mail: rmelamed@cetem.gov.br 4INA Industria nafte, Naftaplin, Zagreb, Croatia, E-mail: zdravko.spiric@ina.hr
Abstract: The relative efficiency of digestion/leaching procedures for the determination of high mercury concentrations in activated carbon obtained from natural gas treatment facilities was investigated. The method is based on acid digestion/leaching, reduction by SnClp, gold amalgamation and detection by cold vapour atomic spectrometry. Sample decomposition was carried out in sealed Pyrex ampoules and closed Teflon vials, as well as in "cold finger" vessels. We used various combinations of acids (HNOq/H2SO4, HNOq, HNOq/HC1, HN03/HF/HCl and HCl), digestion time, temperature and filtration. The efficiency of decomposition was obtained by comparison with results obtained by the Eschka method, radiochemical neutron activation analysis (RNAA) and ^-instrumental neutron activation analysis (k0-INAA). Our results showed that for samples of activated carbon containing up to 6 % of mercury, good results could be obtained using various digestion procedures (HNOq/HC1, HNOq/HF/HC1 and HCl) with the exception of HNOq or HNOq/H2SO4, which resulted in lower values. In addition, for samples of activated carbon with more than 6 % of mercury, efficient digestion/leaching can be obtained by treating the activated carbon in a "cold" finger vessel with 20 mL of concentrated HCl or aqua regia (HC1:HNOq (3:1)).
Key words: atomic absorption spectrometry, mercury, activated carbon
Introduction
Many analytical techniques have been developed for the determination of total mercury, especially for low mercury levels in environmental samples, using CV AAS. When samples with very high mercury concentrations (e.g., in the mg/g concentration range) are to be analysed by such a sensitive technique the major source of error may be
due to large dilution factors. Also, preparation of such contaminated samples in trace analysis laboratory may lead to contamination of the laboratory and/or the measurement system. The activated carbon studied in this work originates from natural gas treatment facilities, where it is used for removal of mercury from the natural gas stream. On average, the concentrations of Hg in such carbon can reach up to 30 % (w/w) of mer-
RMZ-M&G 2004, SI
1976
Part 3 - Analytical
cury. The second problem is the preparation of a homogeneous and representative sample of activated carbon, and the third important problem to be mentioned is related to the efficiency of digestion/leaching procedures. In order to overcome these difficulties the granules of activated carbon were ground in a stainless steel mixer.
The classical approach to determine Hg at high concentration is based on the gravimetric principle that was for centuries used in the Hg mining industry (the Eschka method). There are a number of alternative methods that can also be used, such as radiochemical neutron activation analyses (RNAA) or instrumental NAA (INAA), but these methods are only available in specialized institutions. The goal of this study was to investigate the relative efficiency of digestion/leaching procedures for subsequent determination of high mercury levels in activated carbon using gold amalgamation and detection by cold vapour atomic absorption spectrometry (CV AAS)[12]. Sample decomposition was carried out in sealed Pyrex ampoules and closed Teflon vials, as well as in "cold" finger vessels. We used various combinations of acids (HN03/ H2S04, HN03, HN03/HCl, HN03/HF/HCl and HCl), digestion time, temperature and filtration. The efficiency of decomposition was compared with results obtained by Eschka method'51, radiochemical neutron activation analysis (RNAA)[3,4] and k0-instru-mental neutron activation analysis (k0-INAA)[4] as reference methods.
Experimental
The activated carbon studied in this work originated from natural gas treatment facili-
ties, and was received in granulated form. Since the efficiency of wet digestion and leaching of mercury from the sample also depends on the size of the particles, the granules of activated carbon were ground in a stainless steel mixer. About 50-100 mg of ground activated carbon was weighed directly in a Teflon vial, Pyrex ampoule or "cold" finger digestion vessel. After addition of different acids the vessel was closed, Pyrex ampoules were sealed and the mixture was left to react at room temperature for 1 hour. Digestion/leaching was finished by heating the mixtures in an Al block at 100 (135) °C for 3 or 12 hours on a hot plate. Some digests were filtrated through Whatman GF/C filter prior dilution. To the digest was added 10 mL of BrCl and Milli-Q water to the desired volume. An aliquot of the digest was added to the reduction vessel and after reduction with SnCl2, mercury was swept from the solution by aeration and concentrated on a gold trap. Mercury was then released from the gold trap by heating and measured by cold vapor atomic absorption spectrometry (CV AAS).
The Eschka method is used for fast quantitative determination of Hg in the ores and is based on amalgamation process on a pure-golden pot. Ground and dried sample was weighed in a porcelain dish and 3 g of iron chips and 30 g of Zn0 were added. The porcelain dish was covered with a previously weighed pure-golden pot and heated on an electric furnace at 600-800 °C for 30 minutes. Meanwhile fresh deionized water was constantly added on the surface of the golden pot, so the temperature of water did not exceed 40 °C. When the reaction was finished, the porcelain dish was left to cool down for 10 minutes. The golden pot was dried with a
RMZ-M&G 2004, SI
7th International Conference on Mercury as a Global Pollutant
filter paper and weighed again. The % of Hg was calculated from the difference between the mass of a golden pot before and after the amalgamation process, and divided with the mass of the sample.
Results and discussion
Samples of activated carbon were analyzed in both granulated form and in powder form,
1923
obtained by grinding. The acid digestion/ leaching involved 6 mL of an HN03/H2S0R (4:2) mixture and heating in an Al-block on a hot plate at 100 °C for 3 hours. Before dilution with Milli-Q water 1.0 ml of BrCl was added. The measurement proceeded as described above. The differences in homogeneity and digestion/leaching efficiency between the granulated and powder form of activated carbon in 9 parallel aliquots are presented in Table 1. Table 2 presents results
Table I. Comparison of results for total mercury (%) in granulated and powder forms of activated carbon, determined by CV AAS
Hg(%)
Parallel 1 2 3 4 5 6 7 8 9 Sample form:
1.	Granulated 1.24 0.916 0.666 0.463 0.536 0.421 - - -
2.	Powder 4.23 3.42 3.56 3.59 3.87 3.82 4.22 3.84 4.41
Table 2. Results for total mercury (%) in four samples of activated carbon (A, B, C, D) obtained by the Eschka method, RNAA, ko-INAA, and CV AAS after different digestion/leaching procedures
		Sample		
	A	B	C	D
Eschka method	30.7	5.5	5.7	18.2
RNAA13'4]	25.6	5.2	6.1	15.0
ko-INAA[41	31.9	5.9	-	18.9
6mLHN03:H2S04(2:l) 100 °C, 3 hrs, Teflon vial	19.8	4.0	2.7	5.7
6mLHN03:H2S04(2:l) 100 °C, 3 hrs, Teflon vial, filtered (GFC)	18.0	4.9	-	14.3
lOmLHNOj: HF (2:1) + 1.5 mL HCl 135 °C, 12hrs, Teflon vial	15.7	-	5.8	13.1
3mLHCl:HN03(4:l) 100 °C, 12hrs, Pyrex tube, filtered (GFC)	24.6	-	5.3	12.6
^ 20 mL HC1:HNC>3 (3:1) O 100 °C, 3hrs, cold finger vessel	33.5	-	-	-
20 mL HCl 100 °C, 3 hrs, cold finger vessel	31.4	-	-	-
5.0 mL HCl 100 °C, 12 hrs, Teflon vial, filtered (GFC)	22.4	-	6.1	17.6
20 mL HN03 100 °C, 3 hrs, cold finger vessel	3.9	-	-	-
3.0mLHN03 100 °C, 12 hrs, Pyrex tube, filtered (GFC)	-	-	3.5	1.9
RMZ-M&G 2004, SI
1976
Part 3 - Analytical
for total mercury (%) in four samples of activated carbon obtained by the Eschka method, RNAA, k0-INAA, and CV AAS with different digestion/leaching procedures.
In the granulated form of activated carbon concentrations of mercury were of only about 10-35 % of those found for the powdered form of the sample. The granulated sample was also non-homogeneous. Results from Table 1 show that the size of the particles of activated carbon is extremely important for more efficient acid digestion and leaching of mercury from the sample. Samples of activated carbon with high Hg concentrations of the order of 300,000 mg/kg (30 %), digested with HNOq as the major acid, showed about 40 % - 90 % lower results compared to the results obtained by RNAA, k0-INAA, and the Eschka method. After digestion/leaching in sealed Pyrex ampoules or Teflon vials using aqua regia (HCl/HNOQ (4:1)) or concentrated HCl for 12 hours mercury concentrations were underestimated by about 25 %. These results may reflect inefficient decomposition of the sample and losses with Hg volatilisation during longer extraction. Slightly higher Hg concentrations compared to results obtained by RNAA, k0-INAA and the Eschka method were found after digestion/leaching in a "cold" finger vessel using 20 mL of aqua regia (HCl/HNOQ (3:1)) for only 3 hours. Samples of activated carbon with Hg concentrations in the order of 60,000 mg/kg (6 %) digested with HNOQ as the major acid showed
about 50 % lower results than those obtained by RNAA, k0-INAA and the Eschka method. Results obtained by digestion/leaching with HN03/HF/HCl, aqua regia or HCl alone showed good agreement with the results obtained by these other methods.
CONCLUSIONS
For determination of mercury in very contaminated samples of activated carbon containing more than 6 % of mercury the CV AAS method can be applied, if appropriate wet digestion/leaching of the sample is used. The efficiency of wet digestion/leaching depends on the size of the particles. Grinding before wet digestion/leaching is therefore a very important step. However, by far the best method in laboratories routinely analyzing very contaminated samples is based on gravimetry (Eschka method), while radio-chemical methods are useful as reference methods for occasional quality control.
Acknowledgment
The research was supported by the Slovene Ministry of Education, Science and Sport through the SL0-BRA 3 project entitled "Remediation of Mercury Contaminated Sites". The authors are grateful to INA-Industrija nafte d.d. for providing the samples of activated carbon.
RMZ-M&G 2004, SI
7th International Conference on Mercury as a Global Pollutant_
1925
References
[1]	Horvat, M., Lupsina, V., Pihlar, B. (1991): Deter-
mination of total mercury in coal fly ash by gold amalgamation cold vapor atomic absorption spectrometry; Analytica Chimica Acta 243, pp. 71-79.
[2]	Horvat, M., Zvonaric T., Stegnar, P. (1986): Opti-
mization of a wet digestion method for the determination of mercury in blood by cold vapor absorption spectrometry (CV AAS); Vestn. Slov. Kem. Drus. 33, pp. 47S-486.
[3]	Kosta, L., Byrne A. R. (1969): Activation analysis
for mercury in biological samples at nanogram level; Talanta 16, pp. 1297-1303.
[4]	Jacimovic, R., Horvat, M. (2004): Determination of
total mercury in environment and biological samples using k0-INAA, RNAA and CV AAS/ AFS techniques: Advantages and disadvantages; Journal of Radioanalytical and Nuclear Chemistry 2S9, pp. 38S-390.
[5]	ISO 334:1992 Solid mineral fuels — Determination
of total sulfur — Eschka method.
RMZ-M&G 2004, SI