Mercury isotope fractionation during liquid-vapor evaporation experiments

 Anthro- pogenic emissions, mostly due to stationary combustion (notably coal), non-ferrous metal refining, cement produc- tion and gold mining, therefore represent approximately 58–71% of the total primary emissions. Evaporation (distil- lation) of liquid mercury can take place in a variety of industrial applications such as refining processes. Mercury has a high-saturated vapor pressure for a metal and is easily subject to liquid–vapor changes. The study of liquid mer- cury evaporation through a simple experimental setup should provide a first basis for isotopic fractionation in the liquid–vapor system.Up to now, evaporation and condensation laboratory experiments have mainly been used to understand elemental and isotopic fractionation of extraterrestrial material in the context of solar nebula processes. High temperature evapo- rations have been conducted on Mg (Wang et al., 2001; Richter et al., 2002, 2007; Young et al., 2002) and on K iso- topes (Yu et al., 2003), while moderate temperature evapo- ration (Wombacher et al., 2004) has been undertaken with cadmium metal. All of these studies showed substantial iso- topic fractionation during evaporation and condensation, and in particular an enrichment of the lighter isotopes in the vapor phase.¨nsted published with G. von Hevesy a short communication (Bro¨nsted and von Hevesy, 1920), reporting that they had achieved the partial separa- tion of Hg isotopes by vacuum evaporation of liquid mer- cury. A 14% Rayleigh evaporation yielded a light evaporated fraction with a density of 0.999980 times the original density, and a 75% evaporation yielded a residue that was heavier by a factor of 1.000031. These measure- ments were made by pycnometry, a highly precise method for determining relative density differences, with typical precision of  (density). Shortly afterwards, Bro2008; Biswas et al., 2008; Ghosh et al. 2008; Jackson et al., 2008). In contrast to MDF, which is governed by chemical energy of the starting and transition states of reactant molecules, NMF reported on Hg isotopes in experimental and natural processes have been suggested to result from either magnetic interactions (magnetic iso- tope effect) (Buchachenko et al., 2004; Bergquist and Blum, 2007; Buchachenko et al., 2007) or from nuclear volume effects (also known as nuclear field shift effect) (Schauble, 2007).

Experimental evaporation

Two different types of evaporation were conducted.The first one, representing equilibrium evaporation (Fig. 1a), consisted of introducing 30 g of liquid mercury under atmospheric pressure into a 40 ml vessel, hermeti- cally sealed by a Teflon lined septum. The temperatureHamilton Gastight syringe instantaneously injected and bubbled into a second hermetically sealed vessel which contained 1 ml concentrated HNO3, then shaken for 10 h. An estimation of the amount of sampled gaseous Hg was done assuming saturation vapor pressure laws for liquid mercury (log (P) = 10.122–3190/T(K)), Weast (1999). The collected vapor was adjusted to reachsamples in the 2–22 C temperature range presented yields of dissolution more than 90 ± 10%. Due to the large volume of the syringe and the fact that several up- takes were necessary, a 10% concentration uncertainty is assumed. These yields allowed us to justify that equilib- rium evaporation conditions were reached.Fig. 1. A schematic of experimental evaporation systems. (a) Equilibrium evaporation: 30 g liquid Hg was introduced into a 40 ml glass vessel. Saturated vapor was taken up after 24 h equilibration at various temperatures (0–22 C) and dissolved into nitric acid. (b) Dynamic evaporation: liquid Hg was evaporated under a vacuum of 105 bar and vapor was condensed onto the wall of a cold trap. Reactor A contained a ca. 10 mg liquid Hg droplet which evaporated at 22 C between 6 and 24 h. Reactor B contained 250 g (or 50 g, see text) liquid Hg fractions which evaporated between 22 and 100 C for a few minutes. A heating tape surrounded the glass tubing in order to maintain thequid mercury was continuously evaporated and its vapor condensed onto the wall of the cold trap (diameter 5 mm, length 10 cm). The condensed mercury vapor, which repre- sents the cumulated evaporated fraction, was collected and dissolved in concentrated nitric acid. The evaporation glass line was cleaned in concentrated nitric acid bath after each run to prevent any contamination due to possible adsorp- tion of mercury onto the tube wall.