Basalt experiments with

Figure 9. Percentage of initial concentrations of tracers left in solution as a function of time for the sorption experiments with the basalt samples. ( ), U Np f AX Pu fax Am (O), Cm.

These quantities are both higher for experiment 4 and 5 than for the three experiments with unaltered repository components The loss of Pu through the core for experiments 1 and 2, for example, was 0 4 and 0 1 dpm/mL, respectively in experiments 4 and 5 the loss of Pu through the core appears to be 20 to 40 times greater This exercise is intended to show that there is no legitimate way to compare the behavior of altered and unaltered basalt from these data Experiment 6, yet to begin, should clarify this situation. What is clear from comparing the data in Table VI is that actinide behavior in altered and unaltered repository situations will be quite different ... 

The shock experiments with synthetic basaltic eucrite at room temperature reveal a similar pressure-dependent sequence of characteristic shock features in pyroxene and plagioclase to those seen in samples recovered from previous experiments (see, e.g., ]9,10]). Also, tliere is no change in the sequence of progressive shock effects at elevated initial temperatures. Amorphization of plagioclase... 

Results of the leaching experiments at 25°C are more difficult to interpret. Basalt ground water is not the most effective leaching agent nor is shale the least effective. It appears that at the lower temperature fluoride plays a minor role and other, less obvious, factors predominate. Some of the differences could result from kinetic effects. This is a continuing study with further sampling scheduled at still longer time periods, so it is possible that later results will help clarify the 25°C data. 

During the last two decades, many experimental studies on the seawater-rock interaction at elevated temperatures (100-400°C) have been conducted. Particularly, detailed seawater-basalt interaction experiments have been done. Several experimental studies on seawater-rhyolite interaction and seawater-sedimentary rock interaction are also available (Bischoff et al., 1981). Examples of chemical compositions of modified seawater experimentally interacted with various kinds of rocks are shown in Table 1.9. 

Fig. 3.23 Left-. Calculated relationship between the thickness of an alteration rind and/or dust coating on a rock and the amount of 15.0-keV radiation absorbed in the rind/coating for densities of 0.4, 2.4, and 4.0 g cm [57]. The bulk chemical composition of basaltic rock was used in the calculations, and the 15.0 keV energy is approximately the energy of the 14.4 keV y-ray used in the Mossbauer experiment. The stippled area between densities of 2.4 and 4.0 g cm is the region for dry bulk densities of terrestrial andesitic and basaltic rocks [58]. The stippled area between densities of 0.1 and 0.4 g cm approximates the range of densities possible for Martian dust. The density of 0.1 g cm is the density of basaltic dust deposited by air fall in laboratory experiments [59]. Right Measured spectra obtained on layered laboratory samples and the corresponding simulated spectra, from top to bottom 14.4 keV measured (m) 14.4 keV simulated (s) 6.4 keV measured (m) and 6.4 keV simulated (s). All measurements were performed at room temperature. Zero velocity is referenced with respect to metallic iron foil. Simulation was performed using a Monte Carlo-based program (see [56])...

E. L. Shock (1990) provides a different interpretation of these results he criticizes that the redox state of the reaction mixture was not checked in the Miller/Bada experiments. Shock also states that simple thermodynamic calculations show that the Miller/Bada theory does not stand up. To use terms like instability and decomposition is not correct when chemical compounds (here amino acids) are present in aqueous solution under extreme conditions and are aiming at a metastable equilibrium. Shock considers that oxidized and metastable carbon and nitrogen compounds are of greater importance in hydrothermal systems than are reduced compounds. In the interior of the Earth, CO2 and N2 are in stable redox equilibrium with substances such as amino acids and carboxylic acids, while reduced compounds such as CH4 and NH3 are not. The explanation lies in the oxidation state of the lithosphere. Shock considers the two mineral systems FMQ and PPM discussed above as particularly important for the system seawater/basalt rock. The FMQ system acts as a buffer in the oceanic crust. At depths of around 1.3 km, the PPM system probably becomes active, i.e., N2 and CO2 are the dominant species in stable equilibrium conditions at temperatures above 548 K. When the temperature of hydrothermal solutions falls (below about 548 K), they probably pass through a stability field in which CH4 and NII3 predominate. If kinetic factors block the achievement of equilibrium, metastable compounds such as alkanes, carboxylic acids, alkyl benzenes and amino acids are formed between 423 and 293 K. 

In a basalt-rhyolite interdiffusion experiment (Alibert and Carron, 1980), potassium concentrations CK were measured in a basalt at a given arbitrary distance y in pm between rhyolitic and basaltic liquids experimentally heated for 5000 seconds (Table 5.5 and Figure 5.4). In order to determine the diffusion coefficients, a fit of the experimental points with a polynomial is requested. Use the reduced concentration u, (the fractional deviation of the concentration at a, from the concentrations in the original liquids) given by... 

Figure 5.4 Least-square fit of the K concentration data in Alibert and Carron (1980) experiment of diffusion at basalt-rhyolite interface by a polynomial of degree (n — 1) with n = 6 (top) and n = 10 (bottom). When n increases from 6 to 10, the solution begins oscillating between the data.

In the laboratory experiments of Seyfried et al. (1998), naturally altered sea floor basalt (5 Li = +7.4) was reacted with Li-free alkali-chloride aqueous fluid at 350°C for 890 hours (initial fluid/solid mass ratio 2). Samples of the fluid were taken throughout the experiment, and showed initial rapid influx of isotopically heavy-enriched Li released by early-dissolving alteration minerals. However, with progressive reaction, isotopic composition of the fluid decreased and Li concentration reaehed apparent steady state. Although an equilibrium model applies best to the synthetic results, Rayleigh distillation was considered most likely to apply in hydrothermal reactions occurring in nature. 

In order to obtain information about the exchange processes between the basalt and the salt, leaching experiments were performed with 1.5 m HC1 on basalt powder samples that had previously been washed in bi-distilled water in order to remove salt minerals. The H20-washed whole rock samples, as well as the resulting leachate and residue fractions have been analysed for major elements, REE, and for Sr isotopes. 

Rock Samples. Three rock types were selected as substrates basalt from the Umtanum unit in the Pasco Basin in Washington state, quartz monzonite from the Climax Stock of the Nevada Test Site, and shale (metashale) from the Eleana Formation of the Nevada Test Site. Since both the basalt and the quartz monzonite exhibited different kinds and amounts of alteration within the same rock type, two samples from each rock type were used in the experiments. However, there was insufficient material to study the interaction of the more altered of these rock types with all five actinides, therefore, only the interaction with Pu was studied. 

In order to produce an aqueous solution which fulfills these criteria, 120 g each of basalt, quartz monzonite, and shale were ground to powders less than 37 m in diameter. Each of the samples was placed in two liters of distilled-deionized water which had been pre-equilibrated with an atmosphere containing 10 percent CO2, 90 percent Ar, and 10 ppm Og. The experiment was carried out in an inert atmosphere box at room temperature (26 + 2°C). Samples of the fluid (10 ml) were extracted at various times over a 35-day period and filtered (O.OSpm). Analyses for Na, K, Mg, Ca, Fe, Al, Si02 (aq). Eh, and pH were made on each sample. The experiment was terminated at the end of 846 hours and analyses for HCO3, SO4, and Cl were made on each of the fluids. 

Final composition of the aqueous phase. The final compositions of the waters resulting from the three dissolution experiments have been summarized and listed together with compositions of waters from natural systems (Table VI). The experimental and natural basalt waters have very similar compositions. However, the experimental quartz monzonite water has a higher than natural K content while the shale water has higher than natural K and Na contents. The HCO3 content of each of the experimental waters is higher than the content of its natural counterpart while the oposite is true for 04. 

In one experiment basalt was used which contained the following materials (the minerals which were predominantly present are indicated with a ) ... 

Since crushed basalt has been recommended as a major backfill component (1), experiments were completed to evaluate the rate of dissolved oxygen consumption and the redox conditions that develop in basalt-water systems under conditions similar to those expected in the near-field environment of a waste package. Two approaches to this problem were used in this study (l)the As(III)/As(V) redox couple as an indirect method of monitoring Eh and (2) the measurement of dissolved oxygen levels in solutions from hydrothermal experiments as a function of time. The first approach involves oxidation state determinations on trace levels of arsenic in solution (4-5) and provides an estimate of redox conditions over restricted intervals of time, depending on reaction rates and sensitivities of the analyses. The arsenic oxidation state approach also provides data at conditions that are more reducing than in solutions with detectable levels of dissolved oxygen. 

Arsenic Oxidation States. A solution sample was taken 257 hr after initiation of the 300°C basalt + arsenic-doped deionized water experiment (Run D2-8, Table II). The data from arsenic oxidation state AAS analysis of the initial As(V)-doped water (0-hr sample) and of the 257-hr solution sample are given in Table HI. All detectable arsenic was in the +3 oxidation state [As(V) <15pg/L] in the 257-hr sample. Standard additions of AsGD) and As(V) to the 257-hr sample were quantitatively recovered. To desorb arsenic from particulates in this sample, an aliquot of the solution was treated with 5% hydrofluoric acid. The higher As(III) content of the treated 257-hr sample aliquot (110 vs. 61pg/L, Table HI) demonstrates that sorption occurred. Scanning transmission electron microscopic (STEM) analysis of the particulates indicated the presence of poorly crystallized high-iron illite . 

An apparent first order rate constant of 1.5 x 10 2 hr 1 was derived from the DO data from the 150°C experiment. For the 100°C experiment, the rate constant is about 4.5 x 10 4 hr"1. Further experiments are needed to determine the full rate law for oxygen consumption. Because reaction mechanisms and/or rates can change with time, extrapolation to conditions under which basalt controls Eh may not be justified. 

The relationship of the stirring rate in these experiments to the rates of hydrolysis reactions of basalt phases is indicative of surface-reaction controlled dissolution (21). First order kinetics are not inconsistent with certain rate-determining surface processes (22). Approximate first order kinetics with respect to dissolved oxygen concentration have been reported for the oxidation of aqueous ferrous iron (23) and sulfide (24), and in oxygen consumption studies with roll-type uranium deposits(25). 

Applications. In the following paragraphs, the conditions (temperature, time, water/rock mass ratio, surface area) and the results on closed system oxygen consumption and redox conditions of the basalt-water experiments are compared to expected conditions in the open system backfill and near-field environment of an NWRB. Crushing of basalt for pneumatically emplaced backfill could result in a substantial fraction of finegrained basalt with a variety of active surface sites for reaction similar to the crushed basalt used in the experiments. The effects of crushing on rates of mineral-fluid reactions are well documented (10,26). 

With the addition of bentonite to a crushed basalt backfill, aqueous diffusion would be the most effective mass transfer process (31). Aagaard and Helgeson (32) state that at temperatures <200°C, aqueous diffusion rates are orders of magnitude greater than rates of silicate hydrolysis even in acid solutions. Therefore, the dissolution rate of backfill phases and the overall mass transfer process could be controlled by reactions at the mineral-fluid interface. As stated earlier, dissolution of basalt phases in the sampling autoclave experiments may also be controlled by interface reactions. 

Sufficient DO data were not obtained from basalt-synthetic Grande Ronde groundwater experiments to allow determination of a definitive rate law. A first order kinetic model with respect to DO concentration was assumed. Rate control by diffusion kinetics and by surface-reaction mechanisms result in solution composition cnanges with different surface area and time dependencies (32,39). Therefore, by varying reactant surface area, determination of the proper functional form of the integrated rate equation for basalt-water redox reactions is possible. 

---