In the Russian Far East (Khabarovsky krai) the Mukhen spa located in the southern part of the Khabarovsk krai, 100 km east of the town of Khabarovsk (Figure 1). This spa is unique due chemical composition of groundwater: two absolutely different type of cold (7˚C) CO₂-rich mineral waters (Ca-Mg-HCO₃ and Na-HCO₃) issue from drilled wells in the same area. Both types of water are used by local spas. Ca-Mg-HCO₃ type is shallow water (100 m deep), whereas Na-HCO₃ is relatively deep (150 m deep). The calcium-magnesium type demonstrates a relatively low TDS (up to 1.7 g/l) and elevated contents of Fe, Mn, Ba, and SiO₂, while the sodium water exhibits high TDS (up to 14 g/l) and enrichment in Li, B, Sr, Br, and I.

The purpose of this study is to elucidate a genesis of water and gas phase of this spa and identify processes that control the chemical evolutions of relatively deep Na-HCO₃ groundwater from shallow Ca-Mg-HCO₃ groundwater in the Punchi River catchment area of the northern Sikhote-Alin mountains in Russia’s Far East.

The Muhen River catchment lies in the northern part of the Sikhote-Alin mountains system on the boundary zone of the Sikhote-Alin faulted belt and the Middle-Amur Cenozoic depression, at the north-eastern end of the Pereslavian graben.

The rocks exposed in the study region range from Mesozoic to modern. The Mesozoic sedimentary rocks outcrop to the east of the study area and include sandstone, siltstone and shale. The sandstone is generally quite porous and permeable but often pores space has been infilled with clay minerals and carbonate minerals. The deep aquifer is hosted in the upper fractured zone of this sediment. The Oligocene-Miocene rocks consist of approximately 40 m of moderate sorted sand, gravel and rabble with coal-beds and bodies of dolerite and basalt. The shallow stratal water is hosted here. The main recharge zone of this water occurs in the north of basin where the sediment complex outcrops (Thepkaia et al., 2005) [1].

The sedimentary rocks are overlapped by Miocene-Quaternary basalts with thickness of 50 - 400 m. The basalts exhibit slight weathering and are characterized by very low permeability, with limited and superficial water circulation. The main minerals are plagioclase, pyroxene and olivine.

Alluviation associated with the Muhen River system is widespread in this area. It consists of terrestrial poorly sorted sand and gravel with stone, siltstone and clay and reaches a thickness up to 15 m.

Numerous tectonic faults are spread across the study area. Punchi fault is the largest one. It is located under the cap of the Miocene-Quaternary basalts in southeast trend. The width of the fault zone is not less then 150 - 200 m.

The study of the mineral waters and gases was conducted from 2000 to 2013 yy. Water samples were analyzed for major, trace and rare-earth elements (REE). The variable parameters were measured in situ, and the samples were filtered through 0.45 m membrane filters. The samples for isotopic analysis (δ¹⁸O) and [δ²H) were not filtered and collected in a glass. The samples for tritium determination were taken in plastic bottles without preliminary oxidation and filtration.

Major cations and anions were analyzed by ion chromatography. Trace element and REE concentrations in groundwater and bedrock samples were determined by ICP-MS analysis (Agilent 4500 and 7500c). Analytical precision for the REEs was better than 5% RSD. The gas studies were carried out using gas chromatography. The stable isotopes of δ²H, δ¹⁸O, and δ¹³C in aqueous and gas phases were analyzed using an MI-1201 M mass spectrometer.

The tritium was measured using a proportional gas counter with an operating volume of 4 liter without enrichment (concentration) at the Pacific Oceanological Institute of the Far East Branch of the Russian Academy of Sciences. Tritium has been measured at the Pacific Ocean Institute of FEB RAS by counting β-decay in a liquid scintillation counter. The preliminary enrichment in two stages was carried out before the measurement. This allowed increasing the sensitivity limit to 0.03 TU.

Representative results of the chemical analyses of the studied waters are shown in Table 1. Two chemically distinct groundwater types occur in the study area: Ca-Mg-HCO₃ and Na-HCO₃ waters (Figure 2). Shallow samples are Ca-Mg-HCO₃ type water and the deeper samples are Na-HCO₃ type water. The shallow water typically has a total dissolved solid (TDS) content less than 400 mg/L, pH 4.9 - 5.8 and a molar Ca/Mg ratio of about 0.94. Na-HCO₃ groundwater occurs beneath Ca-Mg-HCO₃ water and has a very high TDS up to 14,000 mg/L, pH 7.2 - 7.5.

n.a.-not analysed.

The analysis of the chemical composition of the ground and surface waters (Punchi River) showed that their TDS increases from the river waters through the Ca-Mg-HCO₃ waters to the HCO₃-Na waters. Obviously, the main increase in the salinity was caused by the influx of two components: hydrocarbonate ions and sodium cations. The calculated correlation coefficients (r) are 0.99 for the hydrocarbonate ion and 0.98 for sodium.

Note that the sodium waters have elevated contents of the following components: Cl⁻ (109 - 120 mg/l), Li⁺ (up to 3.5 mg/l), B (50 - 60 mg/l), Sr²⁺ (up to 0.3 mg/l), and Br⁻ (0.15 mg/l), as well as I⁻ (about 0.04 mg/l). The calcium-magnesium waters show high contents of Fe_tot (up to 13 mg/l), Mn²⁺ (up to 0.6 mg/l), Ba²⁺ (up to 0.2 mg/l), and Si (SiO₂ about 80 mg/l). Both the waters show a positive correlation between the potassium and lithium contents with a correlation coefficient of 0.993, between the calcium and magnesium with r = 0.97, between the silica and iron with r = 0.92, and between the potassium and calcium with r = 0.98. This indicates that, in spite of such strong differences in the concentrations, these elements are supplied both in the Ca-Mg-HCO₃ and Na-HCO₃ groundwaters from water-bedrock interaction. The CO₂ influx in the water system shifts the pH toward lower values, thus producing a more aggressive environment and, correspondingly, more intense dissolution of the host rocks. The negative correlation of silica with calcium, sodium, and potassium presumably indicates that the dissolution of the primary aluminosilicates and the influx of the major components in the water were accompanied by intense removal of silicon.

The enrichment of the Ca-Mg-HCO₃ waters in trace components (Sr, Ba, Mn, Fe, Co, and Ni) was caused by the leaching of the elements from silicates and aluminosilicates that compose the basaltic sequence. The elevated contents of Li, B, and Sr in the Na-HCO₃ waters were related to the intense leaching of these components from the sandstones and shales of the water-bearing complex in the presence of CO₂. The analysis of the chemical composition of both the groundwaters during the past dozen years indicates stable contents of all the major components. However, the waters of the Na-HCO₃ type show a steady increase in their B and Sr contents.

The proportions of the main components (Table 2) indicate that the Na-HCO₃ waters were not contributed by marine sedimentation waters and that both the types are infiltration meteoric waters.

The molar Sr²⁺/Ca²⁺ ratio varies from 0.006 in the calcium-magnesium waters to 0.024 in the sodium waters. The obtained data are close to the molar Sr²⁺/Ca²⁺ ratio in shales (0.007) (Krauskopf, 1979) [2] and sodium feldspars (0.01 - 0.02) (Smith, 1974) [3]. The increase in the molar Sr²⁺/Ca²⁺ ratio with increasing TDS presumably resulted from the precipitation of calcite and dissolution of sodium feldspars.

Concentration of rare earth elements for both types of groundwater and surface water were determined from filtered samples and then as unfiltered samples. The REE abundance in the all studied waters is listed in Table 2. Shale-normalized plots of the REEs in these waters are shown in Figure 3. For comparison the earlier publicized data on concentration REE in the Miocene basalts (Martynov, 1999) [4] was also added on this plot. In this

Note: 1-unfiltered samples; 2-filtered samples.

paper, we normalize to North American Shale Composite (NASC) data taken from Hannigton & Shelkovitch, 2001. Shale-normalized REE plots for all water samples are flat to slightly enriched in the HREEs.

The unfiltered samples general indicate higher concentrations than filtered ones, with larger variations for the HREE. In all samples the REE concentration becomes greater with increasing atomic number. The observed enrichment in the HREEs compared to the LREEs is believed to result from formation of stronger-carbonate complexes with increasing atomic number (Lee and Berne, 1993) [5]. Cerium anomalies were clearly found in the REE patterns of the filtered samples (Figure 4), whereas slight negative Ce anomalies are present in surface and unfiltered samples. In the samples of basalts samples (Martynov, 1999) [4] Ce anomaly is also absent. This reflects the reducing conditions of the underground environment where the both types of groundwater evolved. Positive Eu anomalies are evident for all water and basalt samples.

The REE concentration and shale-normalized signature for both types of groundwater are alike and closely resemble the REE signature of basalt rocks from northern Sikhote-Alin. This is likely related to a common path

of groundwater evolution via basalts.

The gas of both boreholes is practically identical in its chemical composition and represents pure CO₂ with an admixture of other gases of no more than 0.8% (Table 3). It was found that the gas composition at the Mukhen deposit is similar to that from other Far East deposits with high pressure CO₂.

The gas factor, i.e., the gas/water volume ratio, varies for different aquifers. It accounts for approximately 3 in the low-TDS waters (Ca-Mg-HCO₃) with an average gas saturation of about 1.3 l/s, whereas Na-HCO₃ waters have high gas saturation of about 6.5 l/s and a gas factor of 32 - 50. In the 1990s, the associated gas from borehole 30 was utilized for food and technical purposes. The partial pressure of CO₂ in the water was calculated to be 0.24 bar for borehole 3 and 0.47 bar for borehole 30.

The equilibrium-disequilibrium conditions in the water-rock system were determined using Phreeqc (Parkhust, 1995) [6], Waterq 4f (Ball and Nordstrom, 1991) [7], and Aquachem 5.1 software (User Guide, 2005) [8]. The thermodynamic modeling showed that the Ca-Mg-HCO₃ waters are undersaturated with respect to all the carbonate minerals, except for siderite (the saturation indexes are −2.5 for calcite, −2.6 for aragonite, −4.4 for dolomite, 0.7 for siderite, andvv −3.5 for strontianite), as well as with respect to sulfates (the saturation indexes are −3.6 for gypsum, −3.8 for anhydrite, and −3.6 for celestine), plagioclases (the saturation indexes are −5.7 for anorthite and −2.0 for albite), potassium feldspar (1.4), olivine (−16.5), pyroxenes (−10.7), and amorphous silicon (−0.2). They are also oversaturated with respect to quartz (1.1) and clay minerals (7.9 and 7.0 are the saturation indexes for kaolinite and smectite, respectively). The insignificant oversaturation with respect to siderite is related to the high content of divalent iron in this type of water. The HCO₃-Na waters are in equilibrium with quartz (0.5) and chalcedony (0.3); oversaturated with respect to all the carbonate minerals (the saturation indexes are 1.6 for calcite, 1.5 for aragonite, 3.2 for dolomite, 0.7 for siderite, and 0.9 for strontianite) and clay minerals (the saturation indexes are 7.0 for kaolinite and 6.1 for smectite); and undersaturated with respect to sulfate minerals, albite (−1.6), anorthite (−2.8), and chlorite (−7.5).

The thermodynamic calculations show that both type of mineral waters are strongly unsaturated to the primary rocks and are in equilibrium with the clay minerals. However, the Ca-Mg-HCO₃ waters have entered the stage of primary accumulation of calcium, which is supplied from Ca-bearing minerals of the bedrocks. Silicon is removed from the water by precipitation of secondary quartz, which fills fissures in the host rocks. In contrast, sodium groundwaters passed through the carbonate barrier and became oversaturated with respect to all the carbonate minerals. Hence, calcium is not accumulated in this type of water, being removed from the solution in the form of carbonates. At the same time, sodium is accumulated in the solution by leaching from primary aluminosilicates. The constant CO₂ gradient provides a significant influx of this element in the water, because, as follows from the equation of albite dissolution in the presence of CO₂, the Na⁺ concentration in the solution posi-

tively correlates with the gas pressure in the system.

The calculated saturation indexes are consistent with the mineral composition of the bedrocks. Secondary alterations of basalts are expressed in the replacement of the olivine and glass of the groundmass by iddingsite and clay minerals.

In the diagrams of the mineral stability, the surface and hydrocarbonate magnesium-calcium groundwaters are plotted in the kaolinite field, while the hydrocarbonate sodium groundwaters are plotted in the montmorillonite field.

Detailed description of the geological and hydrogeological setting of the area together with the chemical data of surface and groundwater provided information about the origin and evolution of two types of groundwater found in the Muhen River catchment.

1) The intense supply of CO₂ along deep-seated faults in the subsurface aquifers is the main factor responsible for the formation of the Mukhen spa. The obtained δ13C (gas) data indicate that the gas in both the boreholes has a common mantle source, which is well consistent with the data on other high p CO₂ spas of the Far East.

2) The mineral waters of both the geochemical types are of meteoric origin, and their geochemical composition was formed by water-rock-gas interaction. The formation of the Ca-Mg-HCO₃ groundwater occurred under intense water exchange owing to the interaction of meteoric waters with the host rocks. The sodium hydrocarbonate (Na-HCO₃) waters were formed under conditions of retarded water exchange during the movement of meteoric waters at deep levels, where they interacted with host rocks (shales, sandstones, and granites) to change their chemical composition. During the chemical transformations, the waters passed through a carbonate barrier with the removal of calcium from the solution by precipitation of carbonates. This process was accompanied by the formation of clay minerals (kaolinite, montmorillonite), which provided sodium accumulation in the waters. The strong difference in TDS of the mineral waters was caused by the difference (by more than 10 times) in the partial gas pressure during the reactions.

3) Negative shift in δ¹⁸O of the hydrocarbonate sodium waters results from the CO₂-H₂O isotopic exchange at CO₂ gas predominance over H₂O (water).

This study was supported by grant of Russian Foundation for Basic Research (Project No 14-05-00171А).

N. A. Kharitonova,G. A. Chelnokov,I. V. Bragin,O. V. Chudaev, (2015) Chemical and Isotopic Composition of Water and Gas Phases from Mukhen Spa (Far East of Russia). Journal of Geoscience and Environment Protection,03,6-13. doi: 10.4236/gep.2015.35002