Adsorbed Chemical Elements of River Runoff of Solids and Their Role in the Transformation of Dissolved Matter Runoff into the Ocean

Abstract

A procedure for experimental modeling of sorption–desorption processes in the mixing zone of river and sea waters, which excludes the determination of the absolute concentrations of adsorbed chemical elements, has been proposed. Based on experimental data, quantitative characteristics of the ion-exchange transformation of dissolved matter runoff during the penetration of terrigenous material into the marine environment were obtained. The real input of calcium into the ocean as a result of desorption from the solid substances of river runoff increases by 8.3–8.7%, while input of sodium, potassium, and magnesium decreases by 14.0–14.6, 22.2–23.3, and 3.0–3.2% of their dissolved river runoff. For trace elements, sorption–desorption processes lead to an increase in the runoff of dissolved manganese, cobalt, nickel, cadmium, thallium, barium, and ammonium by 98.6–103.5, 20.6–21.6, 3.8–4.0, 15.6–16.5, 4.7–4.9, 20.3–21.3, and 0.8% and to a decrease in the runoff of dissolved lead and cesium by 9.1–9.4 and 2.6–2.8%.

1. Introduction

One of the main inputs in the geochemical balance of the ocean is the river runoff of dissolved and solid matter, the chemical composition of which has been studied quite well [1,2,3,4,5]. It is generally accepted that dissolved substances are directly involved in the formation of the seawater salt composition, while solid phases affect seawater mainly indirectly through reactions of authigenic mineral formation (reverse weathering and diagenesis). At the same time, there are many facts indicating that due to sorption–desorption processes in the mixing zone of river and sea waters, the real amounts of dissolved forms of chemical elements entering the ocean can differ significantly from the calculated values based on data on the chemical composition and water discharges in the river’s outlet [4,6].

The adsorbed and dissolved forms of chemical elements are in a mobile dynamic equilibrium, and composition of the adsorbed complex, which is understood as the sum of adsorbed components, is easily and relatively quickly transformed when the composition of the aqueous medium in contact with the solid phase changes. According to experimental modeling data, the transition of terrigenous material from freshwater to marine environment leads to the replacement of most of the calcium by sodium, potassium, and magnesium [7,8,9]. This markedly increases the estimated river runoff for dissolved calcium and reduces the corresponding estimates for potassium, sodium, and magnesium [10]. However, the existing methods for determining the adsorbed complex composition have a number of irremediable disadvantages that do not allow considering the results obtained with their help to be sufficiently reliable. Such disadvantages include the distortion of the adsorbed complex composition during the preliminary removal of soluble substances, the possibility of chemical interaction of desorbing solutions with the studied solid phases, etc. Quantitative estimates of the sorption–desorption transformation of the river runoff of dissolved trace elements are based mainly on field observations on the distribution of their concentrations in the mixing zone and the form of the concentration–salinity relationship, which allow for ambiguous interpretation due to the simultaneous occurrence of various physical, chemical, and biological processes in nature [4].

Recently, the authors have developed an original procedure for experimentally determining the balance of sorption-desorption processes in the mixing zone of river and sea waters, which is maximally similar to natural conditions. This manuscript describes this procedure and presents the first results of its application.

2. Materials and Methods

The procedure for determining the balance of sorption–desorption processes in the mixing zone of river and sea waters is as follows.

A weighted portion of the solid phase (terrigenous material) is brought to the state of sorption equilibrium with fresh water, an analog of the river runoff, after which 1/3 of settled equilibrium solution is filtered and separated into two aliquots. In the first aliquot, the content of macro and trace elements is analyzed, which corresponds to the initial equilibrium concentration of their dissolved forms in suspension solid phase–fresh water. Artificial seawater with a salinity of 105‰ is added to the second aliquot in the amount of 1/2 of its volume (to prepare 105‰ seawater, the three-fold amounts of salts contained in normal 35‰ seawater are taken, minus the masses of salts contained in the two-fold volume of fresh water used in the experiment, so that when mixing one volume of 105‰ seawater and two volumes of fresh water, the composition of the mixed solution exactly corresponded to normal seawater with a salinity of 35‰). The obtained solution “A” has a salinity of 35‰, and its composition reflects the integral, not related to sorption–desorption, result of the chemical transformation of the river runoff of dissolved substances at their penetration into the sea basin (in this case, the factors of chemical transformation will be, for example, the processes of coagulation and flocculation).

To the remaining equilibrium suspension, from which 1/3 of liquid phase has been extracted, an appropriate amount of 105‰ artificial seawater is added, and with continuous stirring in the presence of air, the system is brought to the state of sorption equilibrium. The salinity of the obtained solution “B” is 35‰, and its composition is determined by both changes in the composition of the river runoff of dissolved matter and the transformation of the composition of the adsorbed complex. It is obvious that the difference between the compositions of solutions “B” and “A” corresponds to the integral balance of the sorption–desorption transformation of the adsorbed complex of the solids of river runoff.

This procedure was applied to determine the balance of macro and trace elements (Na, K, Mg, Ca, N-NH₄, Cs, Ba, Mn, Co, Ni, Cd, Tl, and Pb) in the mixing zone of river and sea waters.

Samples of freshwater terrigenous material were selected according to the principle of similarity to the mineralogical composition of the global runoff of solids. Five samples of freshwater bottom sediments prewashed by distilled water from pore solutions were used in the experiments: dark-gray silt (creek, Nizhny Novgorod Region, Russia), two samples of silty sand (Remna and Seima Rives in the same place), and two samples of sandy brown silt (Lake Pes’vo, Tver Region, Russia). According to the data of X-ray phase analysis, clay minerals in the samples are mainly represented by hydromicas (40–60%), kaolinite (20–30%), and chlorite (10–15%). This is close to the relative abundance of clay minerals in the suspended matter of world rivers, where the proportion of hydromicas, smectite minerals, kaolinite, and chlorite averages 50, 25, 15, and 10% [11].

Plastic vessels with 15 g of air-dry sediments were poured with 150 mL of water from the Mozhaisk Reservoir (Moscow Region, Russia), filtered through an autoclaved dense paper filter and containing 0.43, 0.05, 0.61, and 1.50 mg-eq/L of sodium, potassium, magnesium, and calcium, as well as 0.12, 0.28, and 2.05 mg-eq/L of chlorides, sulfates, and hydrocarbonates. The suspensions were bubbled with air for 8 h and kept without stirring for 16 h for settling of the fine fraction. This was enough time to establish sorption equilibrium between Mozhaisk Reservoir water and freshwater sediments. After that, 50 mL of settled solution was taken from each plastic vessel and filtered through a membrane filter 0.22 μm. To 40 mL of the filtrate was added 20 mL of 105‰ artificial seawater containing 1436.7, 31.3, 326.9, and 61.6 mg-eq/L of sodium, potassium, magnesium, and calcium, as well as 1683.4 and 173.2 mg-eq/L of chlorides and sulfates, so that at mixing the latter with water from the Mozhaisk Reservoir in proportion 1:2, the composition corresponded to normal seawater with a salinity of 35‰. The obtained solutions “A” were filtered once again through a membrane filter 0.22 μm into two polypropylene flasks each: (1) without preservation for the determination of major cations and ammonium, and (2) with preliminarily added aliquots of 5 N high-purity nitric acid (0.25 mL per 10 mL solution) for trace element analysis. The remaining 100 mL of water from the Mozhaisk Reservoir with 15 g of sediments were poured with 50 mL of 105‰ seawater, and the suspensions were brought to an equilibrium state with continuous air bubbling (also ~8 h). The obtained solutions “B”, similarly to solutions “A”, were filtered through a membrane filter 0.22 μm into polypropylene flasks without preservation and with preservation by 5 N nitric acid.

The concentrations of major cations and ammonium were measured by capillary electrophoresis. The concentrations of dissolved trace elements (Cs, Ba, Mn, Co, Ni, Cd, Tl, and Pb) were determined by mass-spectrometry with inductively coupled plasma. The relative measurement error was ±3%. The trueness of the analyses was assessed using the international standards of river water SLRS-4 and SLRS-5 and the standard of estuarine water with a salinity of 15‰ SLEW-3, for which the discrepancy between the measured and certified concentrations of the studied elements did not exceed 20%.

3. Results and Discussion

3.1. Major Cations

According to the experimental data summarized in

Table 1. Changes in the composition of the adsorbed complex of freshwater terrigenous material at the interaction with seawater ¹.

Solid Phase	Equilibrium Concentrations [i], mg-eq/L
	Solution “A” (Mixing in the Absence of Solid Phase)	Solution “B” (Mixing in the Presence of Solid Phase)	Difference ∆[i] = [i]B−[i]A
Na
Dark-gray silt	480.5	466.8	−13.7
Sandy brown silt, sample 1	479.6	467.0	−12.6
The same, sample 2	479.1	467.3	−11.8
Silty sand, sample 1	479.6	472.1	−7.5
The same, sample 2	479.9	472.1	−7.8
Mean			−10.68
K
Dark-gray silt	10.55	8.60	−1.95
Sandy brown silt, sample 1	10.48	8.76	−1.72
The same, sample 2	10.29	8.68	−1.61
Silty sand, sample 1	10.39	9.42	−0.97
The same, sample 2	10.19	9.20	−0.99
Mean			−1.45
Mg
Dark-gray silt	110.6	107.4	−3.2
Sandy brown silt, sample 1	109.0	106.4	−2.6
The same, sample 2	110.1	107.5	−2.6
Silty sand, sample 1	109.5	107.5	−2.0
The same, sample 2	111.6	109.3	−2.3
Mean			−2.54
Ca
Dark-gray silt	20.56	38.85	18.28
Sandy brown silt, sample 1	21.15	36.83	15.68
The same, sample 2	20.76	37.44	16.68
Silty sand, sample 1	20.82	31.27	10.45
The same, sample 2	20.38	32.01	11.63
Mean			14.54

¹ Under the experimental conditions (mass ratio solid phase/solution 1:10), the difference in the concentrations of ion i in the solution in mg-eq/L with the opposite sign corresponds to the change in the content of this ion in the adsorbed complex in mg-eq/100 g.

, the ion-exchange balance of major cations during the interaction of freshwater terrigenous material with seawater is maintained with a high accuracy. The mean sum of the amounts of sodium, potassium, and magnesium removed from the solution (–14.67 mg-eq/100 g) almost exactly corresponds to the amount of calcium entering the solution (14.54 mg-eq/100 g): the residual is only 0.9%.

For correct comparison of the results of experiments and field observations, it is better to use not absolute values but their equivalent amounts, because changes in the absolute concentrations of exchangeable ions depend on the total exchange capacity of mineral phases, which is subject to strong variability [10]. To eliminate the influence of this factor, the changes in the concentrations of adsorbed ions ∆[i] were normalized to the change in the content of adsorbed calcium ∆[Ca] (

Table 2. Calcium-normalized changes in the composition of the adsorbed complex of terrigenous material during the transition from freshwater to marine environment.

Material	Ratio of the Equivalent Concentrations Difference in Solution ∆[i]/∆[Ca]				Reference
	Na	K	Mg	Ca
Terrigenous material of river runoff, natural observations	−0.67	−0.19	−0.15	1.00	[10]
Soil and clay minerals, experimental modeling	−0.77	−0.09	−0.14	1.00	“
Freshwater bottom sediments, experimental modeling	−0.73	−0.10	−0.17	1.00	This study
Mean	−0.72	−0.13	−0.15	1.00	“

). The ratios ∆[i]/∆[Ca] do not depend on the dispersion degree and other properties of solid phases, which affect the value of their total exchange capacity, and unambiguously characterize the process of chemical transformation of the adsorbed complex.

Comparison of calcium-normalized changes in the composition of the adsorbed complex of terrigenous material (Table 2) shows good agreement between field and experimental data. The dissolution of 1.0 mg-eq of calcium from the adsorbed complex of terrigenous material is accompanied, on average, by the uptake of 0.72, 0.13, and 0.15 mg-eq of sodium, potassium, and magnesium from seawater. This relation can be used to calculate the global ion-exchange balance at the river–sea geochemical barrier.

According to [10], the average specific value of the exchange capacity of solids of river runoff is 28 mg-eq/100 g (280 g-eq/t), and the calcium proportion in the adsorbed complex decreases from 74.8 to 14.5% when freshwater conditions change to marine environment. This leads to the input from the adsorbed complex to the dissolved state of 168.8 g-eq, or 3.38 kg of calcium from each ton of terrigenous material entering the ocean. With the most probable mass of river runoff of solid substances being 15.0–15.7 Gt/yr [12,13], desorption of calcium is equal to 50.7–53.1 Mt/yr. The ratios of the difference between equivalent concentrations of dissolved sodium, potassium, and magnesium to the similar difference in calcium concentrations ∆[i]/∆[Ca], given in Table 2, make it possible to determine the global ion-exchange balance of cations of the main salt composition at the river–sea geochemical barrier (

Table 3. Input of dissolved cations of the main salt composition with river runoff into the ocean with correction for the ion exchange in the adsorbed complex of solid substances.

Ion	Input into the Ocean, Mt/yr			Contribution of the Ion Exchange to the Input into the Ocean (% of the Input with the River Runoff) 1
	River Runoff [14]	Ion Exchange in the Adsorbed Complex of the Solids of River Runoff 1	Input with Correction for the Ion Exchange 1
Na	300	$-\frac{41.9}{43.9}$	$\frac{258.1}{256.1}$	$-\frac{14.0}{14.6}$
K	58	$-\frac{12.9}{13.5}$	$\frac{45.1}{44.5}$	$-\frac{22.2}{23.3}$
Mg	152	$-\frac{4.6}{4.8}$	$\frac{147.4}{147.2}$	$-\frac{3.0}{3.2}$
Ca	613	$\frac{50.7}{53.1}$	$\frac{663.7}{666.1}$	$\frac{8.3}{8.7}$

¹ The numerator and denominator are the values calculated for the global river runoff of solid substances, equal to 15.0 Gt/yr [13] and 15.7 Gt/yr [12].

). The input of calcium increases by 8.3–8.7%, while the runoff of other cations decreases by 14.0–14.6% for sodium, by 22.2–23.3% for potassium, and by 3.0–3.2% for magnesium. It follows that the sorption–desorption transformation noticeably changes the runoff of the major cations, which must be taken into account in the geochemical balance of the ocean.

3.2. Trace Elements

Experimental determination of the sorption–desorption balance of dissolved trace elements at the interaction of freshwater terrigenous material with seawater is shown in

Table 4. Sorption–desorption balance of dissolved trace elements at the interaction of freshwater terrigenous material with seawater ¹.

Solid Phase	Equilibrium Concentrations [i], μg/L			Specific Desorption (+) or Sorption (−), μg/g (g/t)
	Solution “A” (Mixing in the Absence of Solid Phase)	Solution “B” (Mixing in the Presence of Solid Phase)	Difference ∆[i] = [i]B − [i]A
N-NH4
Dark-gray silt	257	305	48	0.48
Sandy brown silt, sample 1	179	211	32	0.32
The same, sample 2	185	217	32	0.32
Silty sand, sample 1	157	177	20	0.20
The same, sample 2	162	182	20	0.20
Mean			30	0.30
Cs
Dark-gray silt	0.198	0.044	−0.154	−0.0015
Sandy brown silt, sample 1	0.063	0.005	−0.058	−0.0006
The same, sample 2	0.085	0.007	−0.078	−0.0008
Silty sand, sample 1	0.099	0.039	−0.060	−0.0006
The same, sample 2	0.080	0.024	−0.056	−0.0006
Mean			−0.081	−0.0008
Ba
Dark-gray silt	32	1650	1618	16.2
Sandy brown silt, sample 1	66	1450	1384	13.8
The same, sample 2	87	1510	1423	14.2
Silty sand, sample 1	32	1090	1058	10.6
The same, sample 2	64	1100	1036	10.4
Mean			1304	13.0
Mn
Dark-gray silt	1540	17780	16240	162.4
Sandy brown silt, sample 1	680	13140	12460	124.6
The same, sample 2	880	17110	16230	162.3
Silty sand, sample 1	120	1770	1650	16.5
The same, sample 2	80	260	180	1.8
Mean			9350	93.5
Co
Dark-gray silt	1.91	22.15	20.24	0.202
Sandy brown silt, sample 1	1.25	8.81	7.56	0.076
The same, sample 2	2.13	8.86	6.73	0.067
Silty sand, sample 1	0.51	3.70	3.19	0.032
The same, sample 2	0.94	6.47	5.53	0.055
Mean			8.65	0.086
Ni
Dark-gray silt	13.92	35.28	21.36	0.214
Sandy brown silt, sample 1	7.21	17.59	10.38	0.104
The same, sample 2	16.59	24.95	8.36	0.084
Silty sand, sample 1	0.71	2.27	1.56	0.016
The same, sample 2	1.97	2.88	0.91	0.009
Mean			8.51	0.085
Cd
Dark-gray silt	0.14	7.54	7.40	0.074
Sandy brown silt, sample 1	0.22	2.61	2.39	0.024
The same, sample 2	0.40	5.65	5.25	0.052
Silty sand, sample 1	0.13	2.03	1.90	0.019
The same, sample 2	0.32	0.72	0.40	0.004
Mean			3.47	0.035
Tl
Dark-gray silt	0.07	0.50	0.43	0.0043
Sandy brown silt, sample 1	0.03	0.40	0.37	0.0037
The same, sample 2	0.01	0.33	0.32	0.0032
Silty sand, sample 1	0.01	0.10	0.09	0.0009
The same, sample 2	<0.01	0.07	0.07	0.0007
Mean			0.26	0.0026
Pb
Dark-gray silt	5.05	0.92	−4.13	−0.041
Sandy brown silt, sample 1	3.16	0.92	−2.24	−0.022
The same, sample 2	3.39	1.45	−1.94	−0.019
Silty sand, sample 1	3.08	2.44	−0.64	−0.006
The same, sample 2	3.22	2.17	−1.05	−0.010
Mean			−2.00	−0.020

¹ Mass ratio solid phase/solution 1:10.

. For all trace elements, the measured concentrations of dissolved forms both in fresh and sea water are significantly lower than those in equilibrium with the least soluble solid phases [15,16,17,18,19,20]. Due to this, the processes of dissolution–precipitation of solid phases cannot take part in the chemical transformation of the studied elements runoff in the mixing zone of river and sea waters, and biogeochemical and sorption–desorption processes come to the fore.

The obtained data indicate that the penetration of freshwater terrigenous material into the marine environment causes the desorption of ammonium, barium, manganese, cobalt, nickel, cadmium, and thallium, as well as the sorption of cesium and lead. In absolute terms, the effect of sorption-desorption processes in the first approximation is proportional to the relative abundance of trace elements (mean values of desorption (+) or sorption (−) in μg/g or g/t are denoted in brackets):

Mn (93.5) > Ba (13.0) >> Ni (0.085) ≈ Co (0.086) >> Tl (0.0026),

Pb (–0.020) > Cs (–0.0008).

Table 4 also shows that the absolute values of sorption and desorption increase in the passage from coarse-grained (silty sands) to fine-grained (silts) sediments. For manganese and nickel, the differences are maximal (17–18 times); for cobalt, cadmium, thallium, and lead they are equal to 5–6 times; and for ammonium, cesium, and barium, as well as for the major cations, they do not exceed 2.5 times. This is in agreement with the well-known regularity of an increase in the specific sorption capacity of solid phases with an increase in their dispersion degree and, consequently, the specific surface area.

Based on experimental data, it is possible to make approximate estimates of the transformation of the dissolved trace elements runoff under the influence of exchange processes in the adsorbed complex of freshwater terrigenous material at the river–sea geochemical barrier. Multiplying the most probable mass of river runoff of solid substances equal to 15.0–15.7 Gt/yr [12,13] by the mean values of specific desorption (sorption) of components i (Table 4), we calculated the amounts of the latter that pass into the solution, or, on the contrary, are removed from it as a result of sorption–desorption processes in the mixing zone of river and sea waters. Then, the contribution of ion exchange to the input of the considered components into the ocean was estimated as a percentage of their input with river runoff (

Table 5. Input of dissolved trace elements with river runoff into the ocean with correction for the ion exchange in the adsorbed complex of solid substances.

Ion	Concentration in the River Runoff, μg/L 1	Input into the Ocean, Thous. t/yr			Contribution of the Ion Exchange to the Input into the Ocean (% of the Input with the River Runoff) 3
		River Runoff 2	Ion Exchange in the Adsorbed Complex of the Solids of River Runoff 3	Input with Correction for the Ion Exchange 3
N-NH4	14	584	$\frac{4.5}{4.7}$	$\frac{588.5}{588.7}$	0.8
Cs	0.011	0.46	$-\frac{0.012}{0.013}$	$\frac{0.448}{0.447}$	$-\frac{2.6}{2.8}$
Ba	23	959	$\frac{195}{204}$	$\frac{1154}{1163}$	$\frac{20.3}{21.3}$
Mn	34	1420	$\frac{1400}{1470}$	$\frac{2820}{2890}$	$\frac{98.6}{103.5}$
Co	0.15	6.26	$\frac{1.29}{1.35}$	$\frac{7.55}{7.61}$	$\frac{20.6}{21.6}$
Ni	0.80	33.4	$\frac{1.28}{1.33}$	$\frac{34.68}{34.73}$	$\frac{3.8}{4.0}$
Cd	0.08	3.34	$\frac{0.52}{0.55}$	$\frac{3.86}{3.89}$	$\frac{15.6}{16.5}$
Tl	(0.02)	0.83	$\frac{0.039}{0.041}$	$\frac{0.869}{0.871}$	$\frac{4.7}{4.9}$
Pb	0.079	3.29	$-\frac{0.30}{0.31}$	$\frac{2.99}{2.98}$	$-\frac{9.1}{9.4}$

¹ N-NH₄ [21], other trace elements [5]. ² The volume of water runoff into the World Ocean taken to be equal to 41,700 km³/yr [22]. ³ The numerator and denominator are the values calculated for the global river runoff of solid substances, equal to 15.0 Gt/yr [13] and 15.7 Gt/yr [12].

).

As can be seen in Table 5, the change from freshwater to marine conditions is accompanied by desorption of manganese, cobalt, nickel, cadmium, thallium, barium, and ammonium in the amounts of 98.6–103.5, 20.6–21.6, 3.8–4.0, 15.6–16.5, 4.7–4.9, 20.3–21.3, and 0.8% of the input of dissolved forms of these elements into the ocean without taking into account their transformation on the river–sea geochemical barrier. Lead and cesium, conversely, are adsorbed and removed from the solution, resulting in that the global runoff of their dissolved forms decreasing by 9.1–9.4 and 2.6–2.8%, respectively. This suggests that sorption–desorption processes in the mixing zone of river and sea waters lead to a significant transformation of the runoff of dissolved trace elements.

Thus, the presented experimental results in comparison with the data of natural observations for major cations allows for establishing that the described procedure of experimental modeling of sorption–desorption processes at the river–sea geochemical barrier is suitable for estimating their real contribution to the transformation of dissolved matter runoff into the ocean.

4. Conclusions

The proposed procedure for experimental modeling of the transformation of the adsorbed complex of terrigenous material in the mixing zone of river and sea waters makes it possible to study this process under conditions maximally similar to those in nature.

As a result of sorption-desorption processes at the river–sea geochemical barrier, the input of dissolved sodium, potassium, and magnesium into the ocean decreases by 41.9–43.9, 12.9–13.5, and 4.6–4.8 Mt/yr, and the input of dissolved calcium increases by 50.7–53.1 Mt/yr, which is equal to −(14.0÷14.6), −(22.2÷23.3), −(3.0÷3.2), and 8.3÷8.7% of their river runoff, respectively.

Penetration of solid substances of river runoff into the marine environment causes desorption of manganese, cobalt, nickel, cadmium, thallium, barium, and ammonium equal to 98.6–103.5, 20.6–21.6, 3.8–4.0, 15.6–16.5, 4.7–4.9, 20.3–21.3, and 0.8%, respectively, of the input of dissolved forms of these elements into the ocean without taking into account their transformation on the river–sea geochemical barrier. Lead and cesium, on the contrary, passed into the adsorbed complex and were removed from the solution, which reduces the global runoff of their dissolved forms by 9.1–9.4 and 2.6–2.8%.