Introduction

Mercury (Hg) differs from other heavy metals in that it continuously goes through the deposition and re-emission cycle in the environment because of its high vapor pressure (Poissant et al. 2000). Some recent models of its movement through the environment suggest that Hg emissions and re-emissions from soil and vegetation account for between 19 and 51 % of the estimated 5,500–8,900 tons of Hg currently being emitted and re-emitted to the atmosphere from all sources (UNEP 2013); therefore, air–soil exchange is an important component of the Hg biogeochemical cycle. Soil conditions are typically favorable for the formation of inorganic Hg(II) compounds such as HgCl2, Hg(OH)2 and inorganic Hg(II) compounds complexed with organic anions (Schuster 1991). Hg(II) in soil can be reduced via both biotic and abiotic processes (Schlüter 2000; Poissant and Casimir 1998) after which it can be emitted to the atmosphere in the form of elemental mercury vapor (Hg0).

Many previous studies agree that evasion of Hg from soils is affected by various factors including climate parameters (temperature, solar radiation, relative humidity, wind speed), soil properties (organic matter, pH, cation exchange capacity), and Hg content and speciation in soil (Yang et al. 2007; Xin et al. 2007; Choi and Holsen 2009a; Kocman and Horvat 2010; Corbett-Hains et al. 2012; Fu et al. 2012); however, results of how these factors influence the Hg emissions are often contradictory. While many studies found a significant positive relationship between solar radiation and Hg emissions (Carpi and Lindberg 1997; Poissant and Casimir 1998; Engle et al. 2001; Zhang and Lindberg 2001), higher Hg emissions were often observed during the night time or on cloudy days (Engle et al. 2005; Zhang et al. 2008). The effect of radiation wave length on Hg emissions from soil surfaces, even in controlled environments, is not conclusive (Moore and Carpi 2005; Carpi et al. 2007; Xin et al. 2007; Choi and Holsen 2009b). In addition, some studies found a positive correlation between Hg emissions and soil humidity (Scholtz et al. 2003; Bahlmann et al. 2004; Gustin and Stamenkovic 2005), while others observed the opposite (Leonard et al. 1998; Wallschläger et al. 2000; Gabriel et al. 2006; Choi and Holsen 2009a).

These contradictory results suggest that air–soil Hg exchange is concurrently affected by various factors and cannot be explained by a single parameter. To date, there have been only a limited number of studies carried out to determine the combined effects of the parameters that influence Hg emissions from soil. Lin et al. (2010) observed the synergistic effects between air temperature and soil moisture and between air temperature and light, but more laboratory data are needed to verify these hypothesis. In this study, Hg fluxes were measured in a controlled environment to identify the individual and combined effects of the most important parameters influencing the emissions including, temperature, UV radiation, and soil humidity.

Materials and methods

Soil samples

This study was conducted using lawn soil which was located on the campus of Kangwon National University in Chuncheon, South Korea, where no large industrial Hg sources existed. The lawn was completely covered by 3 cm tall grasses before about 2.5 kg of soil was collected approximately 5–7 cm below the surface using a plastic shovel. The soil samples without any vegetation were then freeze-dried and homogenized before analysis.

The total Hg content in soil was measured using a DMA 80 automatic Hg analyzer (Milestone, Inc.). A standard curve was obtained (0.5 ppb–2 ppm) and R 2 was >0.999. Standard reference materials (SRM) (MESS 3, marine sediment, NIS) were used to calculate the accuracy and precision, and recovery ranged from 102 to 108 %. The method detection limit calculated as three times the standard deviation of seven sequential analyses of SRMs was 0.029 μg kg−1.

Soil organic matter and organic carbon were determined by loss on ignition (Nelson and Sommers 1996). Soil pH was measured by soaking in distilled water. Soil particle size distribution was measured using the hydrometer method (Bohn and Gebhardt 1989; Bouyoucos 1962).

Flux measurement

A cylindrical polycarbonate dynamic flux chamber (DFC) (3.78 L) was used, similar to that used by Choi and Holsen (2009a, b). The diameter of the DFC opening over the soil surface was 18 cm, and four inlet holes (1 cm diameter) were evenly distributed around the chamber wall to insure that the chamber was well mixed. A detailed description and diagram can be found in Choi and Holsen (2009a). In these experiments, the homogenized soils were evenly distributed on stainless steel tray (about 5 cm depth), and the edges of chambers were placed 3 cm under the soil surface. Mercury concentrations in inlet (C in) and outlet (C out) were alternatively measured using Tekran 1110 synchronized two-port sampling unit and Tekran 2537B analyzer, and 10 min average of C in and C out was measured to calculate the Hg emission flux every 20 min using Eq. (1).

$$F \, = {{ \, \left( {C_{\text{out}} {-} \, C_{\text{in}} } \right) \times Q} \mathord{\left/ {\vphantom {{ \, \left( {C_{\text{out}} {-} \, C_{\text{in}} } \right) \times Q} A}} \right. \kern-0pt} A}$$
(1)

where F is Hg emission flux in ng m−2 h−1, C out and C in are the concentrations of Hg at the outlet and inlet in ng m−3, respectively, Q is the flushing flow rate through the chamber in m3 h−1, and A is the surface area of the soil exposed in the chamber in m2. The optimum flushing rate, Q of 5 Lpm was chosen since it yielded the best and most consistent recovery (98 %) of standard Hg vapor in previous study using same size and shape of DFC (Choi and Holsen 2009a, b). A positive value of F indicates a net Hg emission from the soil surface, while a negative value indicates net Hg deposition to the soil surface.

The Tekran 2537B was calibrated with its internal permeation source before sampling. It was also manually calibrated using injection of Hg0 saturated vapor. R 2 was >0.9995 using five point manual injection, and the recoveries were 108.1 ± 1.6 and 102.5 ± 2.7 % for cartridge A and B, respectively. Relative standard deviation was 3.3 ± 1.3 % from seven injections of Hg0 saturated vapor. The concentration difference between inlet and outlet was <5 % when a DFC was placed on the baked aluminum foil with no light.

Effects of parameters

Three environmental parameters were controlled, including soil temperature, relative humidity in the soil pore (RHsp), and UV radiation. Soil temperature was increased by placing the stainless steel tray containing the soil on a hotplate. To control RHsp, collected soil was first freeze-dried, and distilled water was sprayed onto the soil surface to reach the desired RHsp value. Soil surface temperature and RHsp were measured every 5 min using a HOBO data logger (U23-001, onset, INC., USA) placed 1–2 cm under the soil surface.

UV lights with three UV tubes (UV-A 320–380 nm, 0.368 mW cm−2, UV-B 280–320 nm, 0.259 mW cm−2, UV-C 200–280 nm, 0.115 mW cm−2) were installed inside the DFC to allow 100 % of the generated UV light to reach the soil surface, while the outside of DFC was wrapped in aluminum foil. It should be noted that the strength of the UV-A and UV-B radiation used in these experiments was higher than found in the natural environment (UV = ~3 mW cm−2, Park et al. 2013), and radiation below 290 nm rarely reaches the Earth’s surface due to the absorption by stratospheric ozone (Seinfeld and Pandis 2006).

Statistical analysis

Most of the data in this study are not normally distributed; however, if the number of data exceeded 30, it was assumed to be a normal distribution based on the central limit theorem (Ott 1995). Otherwise, non-parametric tests were used. To statistically compare two datasets, the Mann–Whitney U test was used when the number of data was <30. All the statistical analysis was conducted using the SPSS (Statistical Package for the Social Sciences, Ver. 12).

Results and discussion

Basic properties of soils

Lawn soil was collected in February and May, 2011 Hg content 32.3 ± 2.4 (n = 6) and 37.2 ± 4.2 μg kg−1 (n = 6), respectively (Table 1). Mercury levels were similar to those found in bare soils and grassland in Canada and the USA (Poissant and Casimir 1998; Stamerkovic et al. 2008); however, they were significantly lower than those measured in forested and agricultural land (Table 1). Mercury in bare soil in Seoul, South Korea measured in 1990 was significantly higher (240 μg kg−1) than that found in this study, possibly because atmospheric Hg concentration has been significantly decreased since that time (Kim and Kim 2000, 2001; Nguyen et al. 2007; Kim et al. 2009) and/or because Seoul is a much more urbanized city with more Hg sources than the city where the sampling site was located in this study. Organic matter (OM) and organic carbon (OC) contents were relatively lower than those found in forested and agricultural soils in previous studies (Table 1). Soil OM has been suggested to contain effective binding sites for Hg(II), resulting in a positive correlation between Hg content and OM content in soil (Meili 1991; Yang et al. 2007). The contribution of sand was higher than silt and clay, and the average pH was 6.6.

Table 1 Selected soil properties and total Hg content
Full size table

Effect of soil temperature

The impact of soil temperature on Hg emissions was determined under the dark conditions (the outside of chamber was wrapped in aluminum foil). At an initial soil surface temperature of 28 °C, the flux was negative (−3.2 ± 2.1 ng m−2 h−1). However, it was positive and significantly enhanced when the soil temperature was increased to 37 °C (44.2 ng m−2 h−1) and then to 40 °C (245.2 n m−2 h−1) (Fig. 1). This result is in agreement with previous research (Choi and Holsen 2009a; Kocman and Horvat 2010; Lin et al. 2010), and may be due to an increased Hg reduction rate, enhanced bacterial activity, the increased vapor pressure of Hg0, and/or to a decrease in the sorption of Hg by soil.

Fig. 1
figure 1

Effect of soil temperature on Hg emission flux

Full size image

When the temperature was increased by 3 °C (from 28 to 31 °C) at around 400 min, the flux became positive but there was no significant increase in the magnitude of the flux. However, when the temperature increments were 6.6 and 11 °C at around 1,200 and 2,000 min, respectively, the Hg emission flux increased exponentially (Fig. 1), following the Arrhenius equation. The activation energy (E a) was calculated from the temperature dependence of the Hg emission flux as ln(F) = ln(A) − E a /RT, where F was the Hg emission flux, A was the frequency factor, T was the absolute soil surface temperature and R was the universal gas constant, and was 109 kJ mol−1 (r = 0.85, p value < 0.01), which is similar to, or slightly higher than that found previously under dark conditions: 82–109 kJ mol−1 (Kocman and Horvat 2010), 49–63 kJ mol−1 (Bahlmann et al. 2006) and 68.9–107.9 kJ mol−1 (Gustin et al. 1997). These differences in E a may be due, at least in part, to differences in Hg concentrations in the soil as E a typically was higher for background soil (Carpi and Lindberg 1998; Poissant and Casimir 1998). Bahlmann et al. (2006) suggested that E a decreases logarithmically with increasing Hg concentration in soils (E a = −5.972 ln(Hg concentration in μg g−1) + 71.857). Based on this equation, the predicted E a for the soil sample in this study was 91.9 kJ mol−1. Since the enthalpy of vaporization of elemental Hg (Hg0) is 59 kJ/mol, the Hg emissions from soils in this study and other previous studies cannot be explained by the direct emission of Hg0, indicating that the type of Hg species and their bindings in soils affect Hg emissions. Choi and Holsen (2009b) found different dependency of Hg fluxes on temperature between sterilized and non-sterilized soils, indicating that E a can also be affected by microbial activity.

UV effect

At constant temperature, when soil was exposed to UV-A and UV-B radiation, Hg emissions initially increased approximately 13 (313.1 ± 26.21 ng m−2 h−1) and 25 times (611.5 ± 143.6 ng m−2 h−1), relative to the dark condition (24.3 ± 24.1 ng m−2 h−1), and then decreased as time increased (Fig. 2), suggesting that both UV radiation wavelengths directly reduced Hg(II) to Hg0. The decreasing trend in Hg emissions as exposure time increased was possibly caused by a decrease in the available Hg for reduction, because light energy has the greatest impact at upper 2 mm of the surface (Xin et al. 2007; Herbert and Miller 1990).

Fig. 2
figure 2

Effects of UV-A and UV-B (top panel), and UV-C on Hg emissions (bottom panel)

Full size image

Hg emissions due to UV-B exposure were enhanced from initially 15 to approximately 1,000 ng m−2 h−1 when the soil temperature was 24–27 °C (Fig. 2); however, the increment was significantly decreased when the soil was at 13–14.5 °C (−1.6 to 6.2 ng m−2 h−1) (Fig. 3). According to the relationship between Hg emission flux and soil temperature obtained in this study, lnF = −13125(1/T) + 46.547 (Pearson r = 0.85), the Hg emission flux was predicted to be 10.5–16.4 ng m−2 h−1 for soil temperatures of 24–27 °C. The observed Hg emission during exposure to UV-B at a soil temperature of 24–27 °C (454–989 ng m−2 h−1) was much higher than the expected emission; however, at 13–14.5 °C, Hg emissions (0–6 ng m−2 h−1) were just a little higher than that predicted by the temperature alone (2.0–2.5 ng m−2 h−1). This result indicates that interactions between UV-B and temperature are enhanced at higher temperatures.

Fig. 3
figure 3

Effect of UV-B on Hg emissions at the soil temperature of 13–14.5 °C

Full size image

When soil was exposed to UV-C, the flux was negative, indicating that Hg was deposited onto the soil surface (Fig. 2). As the duration of UV-C radiation exposure increased, the magnitude of the negative flux also increased (Pearson r = −0.78, p value < 0.001). The four core oxidation reactions for Hg are with O3, Cl2, H2O2 and OH in gas phase, and there are six core oxidation reactions in the aqueous phase including with O3, OH, HO2 and HOCl (Pal and Ariya 2004; Gbor et al. 2006; Holloway et al. 2012). To find out whether O3 was a major component in the oxidation of Hg, O3 concentrations were measured under the UV-C radiation, and they dropped nearly to zero as soon as UV-C was radiated (Fig. 4), indicating that Hg oxidation and O3 destruction concurrently occurred. O3 can react with Hg0, producing 0.5 mol of reactive gaseous mercury (Holloway et al. 2012); however, the reaction is not fast enough to cause instant deposition of Hg based on the reaction constant of k = 2.11 × 10−18 cm3 molec−1 s−1 as used in CMAQ-Hg model (Holloway et al. 2012). Hg0 can also be oxidized by OH, which has a much higher reaction constant ranging from 7.7 × 10−14–1.6 × 10−11 cm3 molec−1 s−1 (Sommar et al. 2001; Miller et al. 2001; Pal and Ariya 2004; Holloway et al. 2012) than the reaction with O3. High OH can be produced with higher humidity and strong actinic fluxes through the reaction with the excited oxygen atom, O(1D) within Hartley bands (<320 nm) (Seinfeld and Pandis 2006), followed by reaction with Hg0. The removal rate of Hg0 by UV-C exposure was not compared with a wide range of relative humidity; however, the Hg deposition flux generally increased as RHsp decreased (Figs. 2, 5, p value < 0.001 in Pearson correlation test), possibly indicating that evaporated water could play a role in Hg oxidation reactions. Other compounds including H2O, HCl, NO2, CO2, and SO3 in the flux chamber might also oxidize Hg under the influence of <254 nm light (McLarnon et al. 2005).

Fig. 4
figure 4

Decreasing O3 concentration with UV-C exposure in DFC

Full size image
Fig. 5
figure 5

Statistically significant correlation between RHsp and Hg deposition flux under the UV-C exposure

Full size image

Additional experiments with UV-exposure and higher soil temperature (24–27 °C) resulted in much higher deposition fluxes than at lower temperature (Fig. 6). This result indicates that the temperature was also an important factor for oxidation of Hg in the presence of UV-C. Up to 23 % of Hg (average = 11.7 %) from the inlet was removed when the soil temperature ranged from 10.4 to 13.7 °C (Fig. 2), while a maximum of 70 % was removed from the inlet with the soil temperature of 24–27 °C (Fig. 6). Choi and Holsen (2009b) found that approximately 20 % of inlet Hg was removed with UV-C exposure. On the other hand, Xin et al. (2007) measured an increased Hg emission from a soil surface with UV-C exposure; however, incoming air containing no oxidants was used in their experiments. The result of Xin et al. (2007) indicates that UV-C can also result in conversion of Hg(II) to Hg0 in the soil as well as UV-A and UV-B. Therefore, the deposition flux during UV-C exposure shown in this study was tempered by an increased emission flux due to the reduction of Hg in soil by UV-C.

Fig. 6
figure 6

Hg deposition occurring during UV-C exposure in DFC

Full size image

Soil water content

Four groups of dry (RHsp = 2.3–8.0 %) and wet (RHsp = 51.0–99.8 %) soils were tested under dark conditions, and Hg emissions from dry soils were consistently higher than from wet soils at similar soil temperature (Fig. 7), indicating that water inhibited Hg emissions when no light was present (p value < 0.001). When RHsp varied from 58 to 100 %, no relationship was found between RHsp and Hg emission, contrary to the findings of Gustin (2003) and Kocman and Horvat (2010) suggesting that more polar water molecules could displace Hg from binding sites on the soil, facilitating its release. It has also been suggested in previous studies, that an increase of soil water often results in a decrease of soil redox potential, enhancing the reduction of Hg(II) to Hg0 (Ponnamperum 1972; Zarate-Valdez et al. 2006). These hypotheses cannot support the results of this study, in which Hg emissions from dry soils were significantly higher than from wet soils under the dark condition at constant temperature (Fig. 7). Gustin and Stamenkovic (2005) suggested the possibility that at high water volumes, the water will fill the interstitial space in the soil matrix and inhibit the gaseous exchange with the atmosphere, suppressing the Hg emissions from the saturated soil although this phenomenon is not likely to significantly impact these results as water contents were below field saturation.

Fig. 7
figure 7

Hg emission flux from dry and wet soil surfaces under the dark condition. Soil temperature was constantly maintained within each group

Full size image

The effect of temperature on Hg emissions was also investigated for wet and dry soils. Hg emissions dramatically increased as soil temperature increased for both wet and dry soils, but the magnitude of the increase was approximately two times larger for wet soil (Avg. 84.9 ng m−2 h−1) than for dry soil (Avg. 42.2 ng m−2 h−1) (p value < 0.001, Fig. 8). The calculated E a from the temperature dependence of Hg emission flux was 78.7 kJ mol−1 and 69.5 kJ mol−2 for dry and wet soils (Analysis of Covariance, p value < 0.001), respectively, indicating that Hg from wet soil was more easily volatilized as temperature increased. This result indicates that water in soil promotes the Hg emissions by decreasing binding energy.

Fig. 8
figure 8

Interacting effect of soil temperature and RHsp on Hg emissions using dry soil (left) and wet soil (right)

Full size image

It is widely accepted that the temperature is an important factor for Hg emissions from soil surfaces, and the importance of temperature was shown to become much greater for soil with high water content than for dry soil in this study. When water is volatilized from the soil surface, the associated Hg in soil water may also be volatilized. This finding can explain the result of previous research in which there was observed a negative correlation between atmospheric RH and Hg emission from soil surface (Poissant and Casimir 1998; Ericksen et al. 2006). Besides, Lin et al. (2010) found the synergistic effect (20–30 % of additional flux enhancement) for interactions with air temperature (15 and 30 °C) and soil moisture (2.5 and 27.5 %) in their experiments. In this study, a much more synergistic effect (54–210 % of additional flux enhancement) was observed.

It should also be noted that even after the RHsp dropped nearly to 0, Hg emissions were still higher for the initially wet soil than for dry soil (Fig. 8), which cannot be explained only using the known physical processes and suggests that the Hg reduction by bacteria which became active with initially high water content and high soil temperature, continued to reduce Hg. Furthermore, an increase in soil water may result in a decrease of the soil redox potential, enabling Hg(II) to be more easily reduced to Hg(0) due to the enhanced availability of electrons (Zarate-Valdez et al. 2006; Moore and Castro 2012), reducing Hg(II) even after the RHsp dropped.

When UV-B was radiated to both dry and wet soils, it clearly enhanced Hg emissions from both soils; however, the effect of UV-B on the Hg emission flux was statistically larger for dry soil than for wet soil (p value < 0.001) (Fig. 9), indicating that the water content in soil suppressed the reduction of Hg(II) by UV-B exposure. This result suggests that water in soil negatively impacts Hg emissions unless water is evaporated from soil to air due to the temperature increase or significant concentration gradient of water between soil and air.

Fig. 9
figure 9

Interacting effect of UV-B radiation and RHsp on Hg emissions for dry soil (left) and wet soil (right). UV-B was radiated since 100 min

Full size image

Conclusion

In this study, the individual and interacting effects of major environmental parameters including soil temperature, soil pore RH (RHsp), and UV radiation on Hg emission flux were investigated in laboratory experiments. Hg emission significantly increased with increases in soil temperature and UV-B exposure. While it was clear that soil temperature positively affected the Hg emissions, the increment in Hg emissions was much higher for wet soil than for dry soil, indicating that water content synergistically affected Hg emissions. It was also observed that even after the RHsp dropped nearly to 0, the Hg emissions remained still higher for initially wet soil than for dry soil, suggesting that initially high water content continued to promote Hg reduction mechanisms for an extended period. On the other hand, without a temperature change under the dark conditions and UV-B exposure, the Hg emission flux was higher for dry soil than for wet soil. These results indicate that high RHsp inhibited Hg emissions at constant temperature (or simultaneous RHsp). The overall results of this study suggest that water in soil impacts Hg emissions differently under different conditions. The interacting effect of soil water content with other parameters has possibly resulted in the many contradicting findings in previous studies investigating how water content impacts Hg emissions. In addition, it may partially explain a wide range of E a of Hg emission (59–109 kJ mol−1) reported by previous researchers although this parameter can also be affected by other factors including Hg species, Hg content in soil, and their binding in soil.

The effect of UV wavelength on Hg emission was also investigated, and the increase of Hg emissions under UV-B exposure was the greatest, followed by UV-A exposure. Under UV-C exposure, Hg emissions were negative and atmospheric O3 concentrations significantly decreased, indicating that atmospheric Hg oxidation and O3 destruction simultaneously occurred. OH is a possible component of Hg oxidation and can concurrently destroy O3; however, it should be further investigated whether other compounds can oxidize Hg under the influence of <254 nm light.