Short communication Automated on-line preconcentration of trace aqueous mercury with gold trap focusing for cold vapor atomic absorption spectrometry Mahitti Puanngam a , Purnendu K. Dasgupta b,n , Fuangfa Unob a,n a Department of Chemistry, Faculty of Science, Chulalongkorn University, Payathai Road, Bangkok 10330, Thailand b Department of Chemistry and Biochemistry, The University of Texas at Arlington, Arlington 76019-0065, United States article info Article history: Received 1 March 2012 Received in revised form 24 May 2012 Accepted 24 May 2012 Available online 4 June 2012 Keywords: Mercury Cold vapor atomic absorption spectrometry Solid phase preconcentration Gold trap abstract A fully automated system for the determination of trace mercury in water by cold vapor atomic absorption spectrometry (CVAAS) is reported. The system uses preconcentration on a novel sorbent followed by liberation of the mercury and focusing by a gold trap. Mercury ions were extracted from water samples by passage through a solid phase sorbent column containing 2-(3-(2-aminoethylthio)- propylthio)ethanamine modified silica gel. The captured mercury is released by thiourea and then elemental Hg is liberated by sodium borohydride. The vapor phase Hg is recaptured on a gold-plated tungsten filament. This is liberated as a sharp pulse (half-width o2 s) by directly electrically heating the tungsten filament in a dry argon stream. The mercury is measured by CVAAS; no moisture removal is needed. The effects of chloride and selected interfering ions were studied. The sample loading flow rate and argon flow rates for solution purging and filament sweeping were optimized. An overall 50- fold improvement in the limit of detection was observed relative to direct measurement by CVAAS. With a relatively modest multi-user instrument we attained a limit of detection of 35 ng L 1 with 12% RSD at 0.20 mgL 1 Hg level. The method was successfully applied to accurately determine sub-mgL 1 level Hg in standard reference water samples. & 2012 Elsevier B.V. All rights reserved. 1. Introduction Mercury is one of the most toxic elements but its unique attributes have made it difficult to find an alternative. The popularity of mercury-based energy efficient lamps and a variety of non-rechargeable batteries almost guarantee that even under the best of circumstances some of this mercury will find its way into the natural environment. There is also increasing use of coal as an energy source in a power-hungry planet, especially in developing countries. Widespread distribution and eventual deposition of mercury occurs via this route. There are many reviews on mercury in the environment and on the importance of its trace determination [1–5]. Lower and lower concentrations are necessary to be measured as our awareness of the omnipre- sence of the metal grows. Mercury is one of the only two metals that can be measured by atomic absorption spectrometry (AAS) or atomic fluorescence spectrometry [6] as the gaseous element at room temperature. This provides both matrix isolation and good sensitivity. The sensitivity is not enough, however, in AAS, the more affordable and the more widely used technique in the developing world, to measure levels of mercury in most water samples of interest. Preconcentration has thus become the sine qua non for trace mercury determination. A comprehensive list of all papers that have reported on some form of mercury precon- centration is prohibitive: at the time of this writing, Web of Science returns 178 entries with mercury and preconcentration in the title and some 41,300 entries with these in the ‘‘topic’’ coverage. Kerstin et al. [7] reviewed the preconcentration of mercury from natural waters in 2009. A non-exhaustive list of reported recent preconcentration media utilize itaconic acid [8], diphenylthiocarbazone [9], triisobutylphosphine sulfide [10], diphenylcarbazide [11] and gold [12] that are coated or bonded on various sorbents; just cation exchange resin can also be used [13]. Ionic liquids (IL’s) increasingly find use for unrelated diverse problems; mercury preconcentration is no exception. IL’s have been used for liquid–liquid extraction of mercury after forming a chelate [14] and in a more ingenious manner, as a single drop headspace microextractant [15]. There are several early reports of preconcentration by chelating mercury and preconcentrating the same on a standard hydrophobic sorbent, e.g., C-18 functionalized silica [16]; an attractive variant of this now simply uses a knotted PTFE coil, instead of a packed sorbent bed [17]. Functionalized nanosized sorbents [18,19] disperse in solution rapidly and pro- vide for an interesting means of field preservation of samples without agitation; however, automated analysis of a large number Contents lists available at SciVerse ScienceDirect journal homepage: www.elsevier.com/locate/talanta Talanta 0039-9140/$ - see front matter & 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.talanta.2012.05.055 n Corresponding authors. E-mail addresses: Dasgupta@uta.edu (P.K. Dasgupta), Fuangfa.U@chula.ac.th (F. Unob). Talanta 99 (2012) 1040–1045