Abstract: ............................................................................................................................ 2

Narrative............................................................................................................................. 3

Goals: .................................................................................................................... 3

2.0 Scientific Significance........................................................................................... 3

3.0 Relevance of Project to the EM Cleanup Mission: ............................................ 4

4.0 Background............................................................................................................ 4

5.0 Research Plan: ...................................................................................................... 8

5.1 Preliminary Studies: ................................................................................. 8

5.2 Research Design and Methodologies....................................................... 9

Task I - Sludge Synthesis and Characterization: .......................................... 9

Task II - Phase Chemistry and Ageing Studies: ........................................... 11

Task III - Sorption by Tank Sludge Components: ....................................... 13

Task IV - Desorption and Dissolution Processes:......................................... 15

Task V - Atomistic Aspects of Sorption: ..................................................... 18

5.3 Summary: ................................................................................................. 21

5.4 Collaborative Arrangements...................................................................... 21

5.5 Literature Cited: ...................................................................................... 22

.

Appendix I - Biographical Sketches................................................................................ 26

Appendix II - Facilities and Resources: .......................................................................... 30

Appendix III - Budget and Budget Explanation: ............................................................ 31

Appendix IV - Current and Pending Support................................................................. 31

Abstract

It will not be possible to recover all the sludge from decommissioned waste storage tanks. To provide believable performance assessments for decommissioned tanks the DOE must predict the leach rates of radionuclides from the remaining sludge. The current methodologies for doing this are, generally, not realistic representations of the chemical processes that will operate in a decommissioned tank. The proposed research will develop more appropriate release models. It is likely that the sludge components primarily responsible for retaining radionuclides will be impure hydrous iron oxides, aluminum phosphates and hydrous iron bismuth silicates. Both the sorption and desorption of surrogate radionuclides by these materials will be addressed experimentally. A key aspect of this proposal is to age the sludge by mild thermal treatments to obtain a picture of how the sludge will release (or retain) its radionuclides over the full period of regulatory concern (typically 10,000 years). In addition to monitoring changes in the bulk chemistry of sludge-fluid systems the sludge phases will be characterized at an atomic level using transmission electron, atomic force and scanning-tunneling microscopy. Documenting changes in bulk chemistry while at the same time characterizing the materials at an atomic level will significantly advance in our general understanding of how trace elements are mobilized (and immobilized) in a variety of near surface environments.

Narrative

1.0 Goals:

This proposal is intended to provide a firm scientific basis for predicting the long-term fate of radionuclides remaining in decommissioned waste tanks. The source terms presently being used in tank closure performance assessments are scientifically indefensible. Costly schedule delays and serious embarrassment to the DOE are inevitable if such studies are ever subjected to a rigorous scientific review. It is our intent to obtain fundamental scientific insights into the ageing processes of freshly precipitated tank sludge components, and then to document how these changes effect the radionuclides associated with them in trace amounts.

2.0 Scientific Significance:

The study of mineral surface - fluid interactions has seen explosive growth in the last two decades. At this juncture the general rules describing such processes have been worked out for most common soil materials. However, because of their great diversity it, can truly be said that devil lies in the details. This proposal would take the next step in understanding how the most significant sorber in nature, iron hydroxide, actually works at various stages along its crystallization path to hematite. In particular, we will focus on correlating changes in surface chemistry (e.g. the sorptive and desorptive properties of the solid), with how the phase chemistry changes as it ages. Current models describing sorption by hydrous iron oxides (Dzombak and Morel, 1990) rely on two sites (strong and weak) that occur in a fixed proportion and concentration on the mineral surfaces. In reality, the situation may be considerably more complicated. The proposed research should provide a check on the validity of the two site model for the common types of iron oxide encountered in nature.

A second contribution will be to extend our understanding of sorption on hydrous iron oxide to include impure precipitates containing significant admixtures of elements such Al, Cr(III), and Ni. Understanding the behavior of these mixed hydroxides would significantly improve our ability to model pollutant migration in numerous settings beyond just those related to nuclear waste management concerns.

Finally, this study will also examine the sorptive behavior of several less studied materials that are of particular interest to the radwaste community. Almost nothing is known of the phases likely to form in the iron-bismuth-silicate-hydroxide system. Thus, the proposed research will provide basic scientific information on a new class of materials, both in terms of overall phase chemistry and ion exchange properties. For aluminum phosphates a great deal is known about the microporous crystalline phases that can be formed, but significantly less is known about the ion exchange of the less crystalline materials that form as initial precipitates.

To summarize, the overall scientific significance of this project lies in three areas: (1) gaining fundamental insights into the ageing of hydrous inorganic precipitates, (2) correlating changes in the solid phase chemistry of these materials with changes in how they interact with fluids that surround them, and (3) understanding the fate of trace components as the ageing of the precipitates occurs.

3.0 Relevance of Project to the EM Cleanup Mission:

Five decades of DOE nuclear fuel reprocessing have generated immense volumes of liquid and semi-solid wastes. The eventual treatment of these wastes will leave a large number of contaminated, though virtually empty tanks. The preferred option is to decommission these structures in place. This, in turn, will require performance assessments from DOE to demonstrate the safety of such practices.

Most of the residual contamination will reside in sludge adhering to the inner surfaces of the tank. A complete performance assessment would evaluate: (1) drainage of fluid from the pores within the sludge - which will remove much of the Cs and Tc inventory due to their relatively high solubilities, (2) colloidal transport of sludge particles, (3) the sorption/desorption behavior of sludge components and, (4) radioactive releases due to the dissolution of the sludge components. The first two aspects of the problem lie in the province of fluid flow modeling, while the last two topics are addressed by this proposal. Insights gained from this research will impact the DOE clean-up mission in two ways, by making DOE closure plans defensible and by suggesting less costly strategies for tank closure.

4.0 Background:

Broadly speaking, tank wastes can be subdivided into three categories; salt cake, liquid supernatant, and sludge. Current Hanford tank remediation strategies (Rapko, et al, 1996) call for using dilute NaOH/NaNO₃ solutions to dissolve the salt cake and sluice out most of the sludge along with the components already dissolved in the liquid supernatant. Plans for other installations are less well defined and may involve rinses with other reagents such as oxalic acid (Ebra, 1985) or just fresh water. In any case, all of these processes will leave a residue of sludge that cannot be readily removed from the inner surfaces of the tank. It is the radionuclides associated with these materials that are the focus of this proposal.

Sludge is a catch-all term covering pore-fluid rich sediment layers that accumulate over the years on the bottoms of most waste tanks. Sludge generally consists of fine grained precipitates (3-100 nm particle sizes, Liu, J. et al., 1995) formed during mixing and ageing of the aqueous tank contents. Bulk chemical analyses of the sludge solids are readily available. Iron, and then aluminum, are typically the most abundant components (DiCenso, et al., 1995, Colton, 1994; Rykken, et al., 1985; Eibling and Fowler, 1983; Fowler, 1982). Beyond that, the viscidities of varying reprocessing technologies preclude further rankings regarding relative abundances. However, in one setting or another the following may be abundant enough to form separate phases in the sludge: Mn, Cr, U, Ca, P, Bi, Ni, Hg, Si, Pb, Zr, Mg and ferrocyanide. It also needs to be emphasized that sludges contain appreciable amounts of pore fluid. However, both the rinses use to initially empty the tanks and the influx of groundwaters after closure will significantly lower the ionic strength, pH, and radionuclide inventory of these pore fluids from that of the indigenous pre-closure liquid supernatant.

Only a limited amount of information is available on actual phases present in the sludge. The most common components in Hanford tank sludges seem to be (see Liu, et al., 1995, Rapko, et al., 1996) sodium phosphate, amorphous or poorly crystalline (e.g. ferrihydrite to goethite) ferric hydroxide, iron-bismuth silicate hydroxide, crystalline (tentatively identified as cancrinite) and amorphous aluminum silicates, amorphous and crystalline aluminum hydroxide (boehmite and gibbsite) and chromium (III) hydroxide. These assessments were, however, made on tanks with relatively high loadings of iron and phosphate. In other tanks with proportionally higher aluminum loadings amorphous aluminum phosphate is a major constituent (J. Liu, personal communication). Phases occasionally found in minor amounts include bismuth oxide, bismuth chromate, iron-bismuth phosphate, Fe-Mn oxides, apatite (calcium hydroxy-phosphate) and lanthanum pyrophosphate. It has also been suggested that a separate uranium phase similar to schoepite may exist where the bulk U content is high. Carbonates of Mg and Ca may exist in aged wastes exposed to the air. Finally, because of the differing processing technologies the proportions of the different phases are likely to be highly variable.

Although the Hanford Tanks have received the most attention, sufficient information is available to hint at some important differences between these sludges and those at Savannah River. Much waste was processed at Hanford prior to development of the PUREX process. In these early waste streams bismuth, phosphate, lanthanum and - somewhat later - chromate were integral part of the plutonium recovery process. The PUREX process avoided these steps. Thus, on average, Savannah River wastes contain less of these elements. Another difference is that at Savannah River appreciable volumes of zeolite (principally Linde AW-500, reputedly with chabazite structure) were disposed of in the tanks along with the liquid wastes. These zeolites have degraded to an unknown (but presumably large) extent to form a phase similar to the mineral natrodavyne (Fowler and Wallace, 1980). Goslen (1986) describes the sludge from one Savannah River tank as consisting of "insoluble calcium carbonate and sulfate, low solubility calcium nitrate, zeolite which contains large amounts of aluminum and aluminum and iron oxides".

Not all sludge components are equally likely to retain radionuclides for long enough to be of concern from a performance assessment standpoint. As mentioned earlier, much of the dissolved inventory in the pore fluids will be removed the early stages of tank closure. The same can be said for other modestly soluble components in the sludge such as sodium phosphate. If a caustic or oxalic acid rinse is used then a portion of the aluminum hydroxide and aluminum phosphate phases may be dissolved as well (J. Liu, personal communication). In short, some components of the sludge will be removed shortly after groundwater gains access to the tanks. Thus, their long term fate will not be an issue in performance assessment. It also follows that their removal from the sludge can adequately be understood by performing short term leach studies on actual sludge.

At the other end of the spectrum are compounds such as apatite and the anhydrous rare earth phosphates. The geologic record clearly shows that these compounds retain trace levels of many elements (including a number of actinides as well as their decay products) for billions of years. Since radionuclides contained in these phases are effectively immobile it is reasonable to ignore them in performance assessment calculations. The low solubility of anhydrous Bi₂O₃ in basic fluids suggests that the radionuclides in this phase would also be immobile for the time frame of interest, though the absence of Bi₂O₃ as a common mineral precludes using the geologic record to support this assessment.

In the case of the degraded zeolite one can infer that it should not carry much other than Cs in the solid and that pore fluids may well resemble those initially present in the tank. Thus, the solids present in these materials will probably only be important from the standpoint of Cs releases so they will not be considered further in this proposal.

In short, although sludges may contain numerous phases only a few will probably play a role as long term sources for radionuclides: hydrous aluminum hydroxide, hydrous iron (III) hydroxide, iron bismuth silicate hydroxide, and hydrous aluminum phosphate. The range of information available about these of compounds is quite variable. Iron hydroxides are so heavily studied that computer programs are available to predict the competitive sorption of different ions (including the actinides) in various environments (Bethke, 1994). Models are also available to at least semiquantitatively predict the ageing of largely amorphous iron oxide ("ferrihydrite") to goethite (Smith, et al, 1994). However, little is known about the effects of ageing on the sorption properties of these materials. In addition, a detailed reading of the documentation on tank chemistry strongly suggests that the term "iron hydroxide" includes materials containing appreciable admixtures of Cr(III) and Ni (where Ni-ferrocyanide was added to the tanks). The sorptive and ageing behaviors of such mixed hydroxides is poorly studied.

Aluminum hydroxide sorption is less well studied than iron hydroxide, but sufficient data is available to demonstrate that its sorptive capacity is significantly less than that of iron hydroxide. Thus, along with the hydroxides of Ni and Cr, Al will only be considered with regard to the role it plays as an impurity in the iron oxides.

Iron bismuth silicate hydroxide lies at the other end of the spectrum. We were unable to locate any published studies regarding the phase chemistry, behavior during ageing, or sorptive properties of artificially synthesized materials. However, the geologic literature regarding weathered ore deposits rich in Bi (Ridkosil, et al, 1996) suggests that the mineral bismuthoferrite (Fe₂Bi(SiO₄)₂(OH)) is a stable phase over the long term. This tentative identification is supported by some fragmentary evidence from TEM analysis of materials in actual tank wastes: both the few diffuse electron diffraction bands and the elemental proportions deduced from the EDS spectra on this material were consistent with such a phase assignment (J. Liu, personal communication). Finally, although the literature on crystalline aluminate phosphate molecular sieves is immense, the potential sorption and desorption of ions by gel-like materials likely to occur in the tanks is apparently unstudied.

Understanding various aspects of colloid chemistry lies at the heart of an immense variety of natural and industrial processes. In nature the migration of both pollutants and vital nutrients is often governed the surface chemistry of the relatively small amounts of colloidal sized clays or hydrous iron oxides present in soils. Colloid chemistry is also vital to many industrial processes: paints, ceramics, ore dressing, cosmetics, pharmaceutical and food processing to name but a few. In general, all of these applications rest on one single fact: colloids have immense surface areas. However, because of their high surface areas colloids are often singularly susceptible to change.

Of particular relevance here is the propensity for poorly crystallized or amorphous hydrous phases to spontaneously invert to crystalline materials while simultaneously undergoing a significant degree of dehydration. This changes both surface properties (e.g. density and dissociation constants of sorptive sites) and decreases the surface areas of the materials. A second topic of equal concern are the kinetic constraints on such transitions.

To maintain close ties with the tank remediation issue, three classes of materials will be studied: Fe>>Al-Ni-Cr hydroxides, iron bismuth silicate hydroxide, and hydrous aluminum phosphate. Of these, only the iron bismuth silicate hydroxide is so specialized as to be relevant only to the tank remediation issue. Research on the general iron hydroxide system will allow for more realistic modeling of a variety of natural ground water systems since seldom is pure iron hydroxide found in nature. The current status of this topic is that rudimentary predictions can be made on the rate ferrihydrite transforms to goethite (Smith, et al, 1994), and a significant body of sorption data exists on various forms of iron hydr(oxide). But, very little is known of the fate of mixed hydroxides containing small admixtures of Al, Cr(III) or Ni. Studies on the sorptive properties of amorphous aluminum phosphate appear to be absent, though they have the potential for impacting a number of natural and industrial processes. For all three cases the main thrust will be to correlate physical changes in the materials brought on by artificially induced ageing with the sorptive (and desorptive) properties of the materials.

This study takes an experimental approach in addressing these questions. Two general classes of experiments will be carried out: (a)"simple" systems where Fe (with Ni, Cr, Al) hydroxides, aluminum phosphates, or variations on the theme of Fe+Bi+Si+OH can be studied individually and (b) studies on simulated sludges. The former will be needed to improve our general understanding of colloid ageing as well as provide more accurate mechanistic models for use in performance assessment calculations. The latter will provide a level of "ground truth" to the exercise (e.g. verify that the P.A. models address processes that dominate in complex sludge-like systems).

With regard to using simulated sludges, it has been suggested that leach tests on actual "hot" sludge samples would be more appropriate. Several arguments mitigate against this. First, such tests would only give information on the leach behavior of sludge at its present age, and give no insight into how (or if) the sludge properties will change many years after the tanks are decommissioned. Typically, predictions regarding the fate of buried nuclear wastes are required to extend to at least 10,000 years.

Aside from the ageing issue, there are a number of other arguments against using actual "hot" sludge for such test. The cost of such an exercise is beyond the scope of this proposal, as witnessed by the sizeable teams assembled to carry out the few existing characterization studies on actual waste (DiCenso et al., 1995; Rapko et al., 1996). If leaching is to be done on actual wastes this clearly lies in the province of site managers who "own" these wastes, and have been allocated the sizeable budgets commensurate with such ownership. Secondly, in the words of an anonymous, though knowledgeable, reviewer leach tests on actual wastes are "not as elegant or amenable to modeling manipulations". Presumably, this latter problem could eventually be overcome with sufficient characterization of the waste - but, given the absence of even a single study where this was actually done, one concludes that currently the problem is just too poorly constrained to be approachable. We submit that basic research into the topics proposed here will provide the focus needed to make such characterizations both technically and economically feasible.

Finally, over the past year the reality of actually having to do something with these tanks has swept across the DOE complex. In that period the number of citations describing tests that use "simulated" wastes have also skyrocketed. Thus, there is an ever growing precedent for carrying out the "discovery" phase of such research using simulants rather than actual waste. Actual wastes are employed only when "proof of principal" is already in hand.

5.0 Research Plan:

Four basic steps will be needed to carry out the proposed research. First, we will fabricate artificial sludge and artificial sludge components (Task I). This will be followed by characterizing the phase chemistry of the solid components as ageing occurs (Task II). The third and fourth research objectives are directed at the sorption (Task III) and desorption (Task IV) behavior of these materials using radionuclide surrogates. Finally (Task V), scanning force microscopy will be employed to gain an atomistic understanding of how changes in the solid surfaces influence the sorption/desorption behavior of sludge.

5.1 Preliminary Studies:

A variety of prior and ongoing activities by the authors have direct bearing on the work scope of this proposal. A program to develop getter materials for Hanford LLW-glass involved assessing the sorption of a number of relevant radionuclides on iron oxides as well as hydrotalcites of various compositions (Balsley et al., 1996). Past performance assessment activities at Sandia have identified the following radionuclides as being of concern for the long-term closure of Savannah River's waste tanks (Kuehne and Krumhansl, 1996):⁹⁹Tc, ⁷⁹Se (as SeO₃⁼), ¹⁵⁴Eu, ¹³⁵Cs, ¹³⁷Cs, ¹²⁶Sn, ¹⁵¹Sm, ⁹⁰Sr, and the actinides. This list is not remarkable in that it parallels many others compiled over the last thirty years by various investigators. It does, however, provide a useful basis for planning our experimental program. Unpublished work recently completed for the WIPP (J.L. Krumhansl) assessed the actinide sorption potential of the mineral apatite in brines, thus demonstrating the potential for actinide sorption on phosphates even at very high ionic strengths. K.L. Nagy (Nagy, 1995), has considerable experience in using AFM - STM techniques to quantifying the growth and dissolution kinetics of kaolinite. As will become apparent later (Task V), the ability to provide atomistic insights into mineral surface processes may hold the key to predicting radionuclide release rates from sludge components. Finally, J. Liu has performed most of the existing TEM characterization on actual "hot" tank waste sludge components from the Hanford Reservation.

5.2 Research Design and Methodologies:

Task I - Sludge Synthesis and Characterization

Technical Discussion: This task provides the sludge samples needed for the remainder of the study. Only recently has the artificial preparation of chemically relevant synthetic sludges received much attention. Hobbs (1995) addressed the problem by dissolving appropriate amounts of the nitrate salts of Fe, Al, Mn, and Ni in 2M nitric acid. Then, radionuclide simulants were added in trace amounts, after which the mix was titrated to the desired (strongly basic) pH using a second stock solution of 50% sodium hydroxide. Finally, the mix was shaken for a week, filtered through a 0.2 micron filter, and analyzed for trace elements. It is noteworthy that Hobbs (1995) draws a distinction between high activity wastes, where Al predominates over iron, and low activity wastes where iron dominates and Al is absent from the recipe.

Norton (1994) provides a more elaborate mix which, like the high activity waste of Hobbs(1995), has more Al than Fe. However, since this results in a final solution that is strongly basic, and aluminum hydroxide is appreciably soluble in such fluids, it is not clear that the precipitates formed will actually contain Al-containing phases. Their mix also incorporates small amounts of Cr(III), silica, sulfate, zirconium, fluoride, and (following addition of excess NaOH) potassium carbonate. After standing for several years, Norton (1994) notes that the particle sizes in his mix matched closely that of at least one sample of actual waste. No attempt was made to assess whether during its earlier history the match was as good.

Finally, a recent and quite elaborate formulation has been proposed by Russell and Smith (1996b) Basic steps are first using NaOH to neutralize nitrate solutions of the major components of this particular formulation (Mn, Fe, Zr, Ni, Nd), and then the addition of other components to mimic the complete minor and trace element make up of the particular tank waste in question. The procedure differs significantly from Hobbs (1995) in that the dopants are added after the main phases are precipitated. While the procedure proposed by Russell and Smith (1996b) is probably too elaborate for the purposes of this proposal it will serve as a guide line in estimating the relative amounts of various trace components to be added to the sludge formulations.

The most significant trace components are those needed to account for the radioisotope inventory of the sludge. As previously noted, ⁹⁹Tc, ⁷⁹Se (as SeO₃⁼), ¹⁵⁴Eu, ¹³⁵Cs, ¹³⁷Cs, ¹²⁶Sn, ¹⁵¹Sm, ⁹⁰Sr, and the actinides are the radionuclides of greatest concern. Except for Tc and the actinides it is possible to substitute naturally occurring non-radioactive isotopes. Tc and the actinides, however, present more of a challenge. Of the actinides, it is our intention to restrict this study to U and Th. However, Nd is a reasonable substitute for Am and will be included as well. This leaves unaddressed the pentavalent actinides (principally Np and Pu), for which no good substitutes exist. Finally, Re will be substituted for Tc. ReO₄^- and TcO₄^- have similar chemistries and differ significantly only in that TcO₄^- is a somewhat stronger oxidizing agent. Sandia does have operational radiochemical facilities able to handle tracer levels of radioactive isotopes. Thus, if concerns develop regarding this substitution, a limited number of well constrained experiments using tracer levels of Tc could be carried out toward the end of the program.

In addition to radioisotopes, the fate of some RCRA metals are of concern from a performance assessment standpoint. For our study we will include tracer amounts of Cd, Pb, and chromate (as well as the Cr(III) necessary to formulate the sludge). Hg, the other element of concern from the standpoint of RCRA, presents significant analytic difficulties in the presence of Fe so it will be omitted from our matrix. For brevity, the list of elements (radioactive and RCRA) to be added at trace levels to the various sludge formulations will be referred to as "dopants".

Experimental Approach: The following activities will provide the materials for use in the remainder of the study:

This activity focuses on producing "representative" artificial sludge. At the onset, Norton's recipe will be followed as it seems to be reasonably comprehensive, although in some cases it will be augmented to allow for the inclusion of the Bi, phosphate, and possibly silica needed to simulate some early sludges generated at Hanford. Thus, the initial experimental matrix will consist of four batches: no Bi+PO₄ and no dopant, Bi+PO₄ and no dopant, Bi+PO₄ and dopant, no Bi+PO₄ and dopant. Where Bi, phosphate and dopants are needed they will be added to the Fe-Al containing stock solutions prior to the addition of base responsible for precipitating the solid "sludge" phase(s). Dopant-free batches will provide information on whether the dopants influence the rate of phase changes, and also provide an aged dopant free material for later sorption tests. Initially the relative abundance of the dopants will be adjusted to approximate the ratios in actual sludge (Georgeton and Hester, 1995, Russel and Smith, 1996). However, this position may need to be reassessed if this plan results in concentrations which are below the detection limits of the analytic tools used to characterize the solids.

In addition to the complex mixtures described above, it will also be necessary to synthesize individual sludge solid components. Again, parallel batches of each - with and without dopant - will be prepared. In the case of the hydrous oxide mixes an experimental matrix with variable ratios Fe to impurity (Cr(III) Ni(II) and Al(III)) will be prepared by neutralizing acidified mixtures of the nitrate salts of these materials. An upper limit of one part impurity (Ni, Al, Cr) to nine parts Fe will be used unless progress in characterizing actual sludges over the next few years indicates that this is grossly non-representative. In the case of the iron bismuth silicate hydrate the problem is less will constrained but a matrix of mixes with bismuthoferrite (Fe₂Bi(SiO₄)₂(OH) at the compositional center appears to be a reasonable starting position. Aluminum phosphates will be prepared by starting with acidified mixtures of aluminum nitrate and phosphoric acid in the ratios of 1:0.3, 1:1 and 1:3. The pH will then be brought to nearly neutral so that the solids precipitate.

Once the solids have been prepared they will be allowed to sit for at least a month at room temperature before being rinsed free of residual synthesis fluids. Next, they will be resuspended in a 0.05M NaHCO₃- NaNO₃ solution to simulate what might be expected over the long term inside a decommissioned tank. They will remain stored in this fluid until needed for later tests.

Task II Phase Chemistry and Ageing Studies

Technical discussion: The central issue of this proposal is how a select group of hydrous colloidal materials will evolve over time, and what effects the changes noted will have on the ability of these materials to retain radionuclides. This section addresses ageing and characterization of the solids. In the case of simple iron oxides we will extend the basic scientific understanding to include the behavior of impure mixtures involving Al, Cr, and Ni.

For aluminum phosphate systems the it appears that information relevant to the ageing issue is restricted to studies carried out in the hydrothermal regime (Popolitov, 1992). Activation energies for a three step process [dissolution of an Al-phosphate glass, transport down a temperature gradient, and finally the nucleation and precipitation of anhydrous AlPO₄] were in the range 2.6 to 3.4 kJ/mol. At 200^o C, production rates were in the range of 0.1 to 0.3 g/day of crystals and about a two order of magnitude decrease in rate would be expected in going from 200^oC to room temperature. Unfortunately the weights of the solid feed stock and the size of the autoclave were not given but one can infer it was a small bench-top set up. Thus, if - as anticipated - our tests last for many months we should be able to observe some ageing even in the room temperature runs.

Finally, there is no published literature on the phase chemistry or transformation rates in the Fe-Bi-Si-OH systems. However, since the synthesis of clays even under hydrothermal conditions is a several month proposition we assumed that ageing will be a relatively slow process when compared to the other materials being evaluated.

Experimental Approach: Once the materials are prepared they will be subjected to an accelerated ageing process. This will create materials that represent what will exist in a decommissioned tank over the full time span of regulatory concern (roughly 10,000 years). The issue of how to accelerated ageing is essentially moot as there is just one option - the application of heat. Hydrothermal conditions should not be required, though autoclaves are available if needed. We anticipate treating the sludge simulants at 60^o and 90^o C as well as keeping part of the mix at room temperature. At least two elevated temperatures are required because an Arrhenius formalism may be needed to extrapolate kinetic data to room temperature. Also, the empirical activation energy determined from such plots can provide insights into the reaction mechanism(s) responsible for the ageing process. At various times splits from each preparation will be taken and analyzed as follows:

Because of the largely amorphous nature of the precipitates it is likely that electron microscopy will be more informative than X-ray diffraction (XRD. Initial screening of precipitates will be performed using a scanning electron microscope equipped with an energy dispersive unit (SEM/EDS). Based on the findings of Liu et al., 1994 it is likely that much of the material will be too fine grained to be analyzed fully using a SEM. Thus, a subset of the materials will also be examined by transmission electron microscopy (TEM). All three capabilities of the TEM will be needed; general morphology, composition (as determined by the instrument's X-ray EDS capacity) and crystallinity, as assessed by electron diffraction patterns. Particular items to be addressed are: phase identification, the relative abundance of the phases, and their morphology.

Nuclear magnetic resonance (NMR) will also be used to study the chemical speciation in those instances where the iron content is very low or it is absent. The peak position and shape, as well as the line width, are very sensitive to the coordination number and chemical environment of certain elements abundant in the tank sludge (e.g. Al and P), as well as certain radioisotopes such as Cs. Therefore, it is possible to document the formation of various compounds and species even for noncrystalline materials.

BET surface areas will be determined using a micrometrics ASAP-2000 unit. This information is a prerequisite to making quantifying sorption and desorption data to be gathered in other parts of this study, as well as in monitoring changes in the solids responsible for the sorption. Our laboratory routinely performs such measurements so obtaining the raw data presents little challenge. However, in a multicomponent mix interpreting these data can be problematic. To accomplish this we will have to rely heavily on the TEM results. Where this shows that a single phase comprises the bulk of the sample it is expected that the raw specific surface area data will approximate that of the phase of interest. In other cases, the specific surface areas of one particular phase in a mix may be obtained from BET measurements made on a similar pure materials (fabricated in an analogous manner), and an estimate of the proportion of that material in the mix - assuming that the TEM can provide the needed proportional information as well as verify that the particles have similar shapes and sizes in both cases.

For iron and aluminum oxides, it is also possible to estimate sorption site densities within about a factor of roughly three (Davis and Kent, 1990). Thus, the surface area can be estimated by saturating the surface sites with a component that has a particular affinity for just the phase in question. A bulk analysis for that component can then be used to calculate a specific surface area provided one has a knowledge of the relative abundance of the various phases in the solid. The latter piece of data can generally be deduced from the phase identification done on the TEM or a bulk chemical analysis of the sample.

The final metric of solid phase evolution will be the evolution of particle size distributions using a Particle Measuring System, Inc. PMS/100 (0.1 to 1 microns) and a Coulter Multisizer (3 to 100 microns). Particle size distributions will be compared with those available for actual waste sludges (Russell and Smith, 1996) to ensure that the samples are a reasonable match to the real waste. Also, it again needs to be stressed that changes over time (or the lack there of) are the variable of prime interest.

Task III Sorption by Tank Sludge Components

Technical Discussion: It is a common practice in performance assessments (P.A.) to employ the Kd of the source material to describe the initial release of radionuclides from the contaminated site (Shaman, 1992, Maheras et al., 1994, Immes, 1994). The approach is open to criticism since it fails to consider the effects of changing bulk fluid chemistry over time, mass balance constraints (e.g. the total amount of a radionuclide present and the ion exchange capacity of the material), solubility limits, or kinetic aspects of the problem such as dissolution and (re)crystallization of the sludge - which may liberate occluded radionuclides to the bulk solution. On the other hand, the approach is computational simple, and can be justified by the argument that the last thing a radionuclide would be exposed to as it leaves the tank are the surface sorption sites on the sludge components.

This task seeks to provide a thermodynamic basis for predicting the nature and extent of dopant sorption onto tank sludge components. For pure iron oxide (see Dzombak and Morel, 1990, for example) and - to a lesser degree - pure aluminum hydroxide this already exists (Huang and Stumm, 1973; Balistrieri, et al., 1981; Brady, 1994; Berger, 1996). But, no data base deals with the sorption onto amorphous hydrous oxides of mixed composition.

With regard to aluminum phosphate, the published exchange studies focus on neutral hydrocarbon molecules and the molecular sieve aspects of neutrally charged microporous zeolitic lattices. Lattices with a net charge (and hence able to engage in ion exchange processes) have received little attention. And, in fact, a search of the literature failed to locate any reports that dealt with actinide sorption. There are, however, published accounts suggesting that actinide sorption may occur on the phosphate mineral apatite in low ionic strength solutions (Morrison and Spangler, 1992; Allard, 1984). Unpublished work recently completed at Sandia demonstrates actinide sorption at high ionic strengths as well. Actinide-apatite Kd values were measured in WIPP brines equilibrated with MgO. Hydroxyapatite demonstrated a strong sorptive capacity for Am⁺³ and NpO₂^- , but significantly smaller affinity for Pu⁺⁵ species. Carbonate complexing of hexavalent uranium prevented its sorption to any appreciable degree. Because hydroxyapatite is an important component of teeth and bones there is additional information in the medical literature (Christoffersen, 1980) that confirms that apatite can sorb a wide variety of relevant elements

Finally, because Bi-Fe-Si-OH systems are not widely encountered in either industry or nature there is no published record on its exchange properties. However, the apparent structure of the mineral bismuthoferrite (Zhukhlistov and Zvyagin, 1977) is suggestive of some ion exchange potential. That is, in this mineral the Fe, O, OH, and Si atoms make up a distorted kaolinite structure with Si in tetrahedral coordination and Fe in the octahedral sites. However, unlike the octahedral Al in kaolinite, the iron in bismuthoferrite does not completely satisfy the negative charge of the lattice. The remaining charge deficiency is satisfied by Bi atoms located in interlayer positions. Thus, although it has a TO-type clay lattice like kaolinite, it may have ion exchange properties that resemble TOT-type clays such as montmorillonite. Such a lattice arrangement also bodes well for being able to artificially synthesize the phase in reasonable periods of time. That is, only two natural clays have been observed to directly precipitate from solution under laboratory conditions: kaolinite ( Nagy, 1995), and sepiolite (Brady, 1992).

Experimental Approach: We propose an experimental program to provide the thermodynamic and kinetic constants needed to constrain the role of dopant sorption in tank sludges at various stages during the ageing process. These results will ultimately provide us with some "rules of thumb" describing sorption potential for a range of ions on various sludge components. These insights can then be broadened to predict the behavior of most radionuclides for most post-closure tank environments. Because our concern is with sorption of dopants this task employs the undoped fraction of the synthetic sludge mixes prepared in Task I. The issue of hysteresis in sorption - desorption experiments is dealt with in Task IV. This section is constrained to addressing relatively rapid surface processes indicative of true exchange reactions. What follows is a tabulation of characterization procedures to be applied at different stages of in the ageing process.

Measure pH-dependent surface charge on the single-phase samples. First the NaNO₃- NaHCO₃ solution will be replaced with a simple NaNO₃ ( 0.01, 0.1, 1.0 M) and NaCl (0.0001M) solutions, and then titrations carried out between pH 6 and 13 as the sludge is stirred. For each titration a proton mass balance calculation will be done to quantify mineral surface charge as a function of pH and roughly identify the reactive surface groups. By varying the NaNO₃ levels we will be able to establish how various levels of groundwater dilution will alter the surface charge. The 0.0001 M NaCl titration will be done to establish a baseline and to approximate surface charge in more dilute, soil solutions that seem likely to result after groundwater has been in contact with the sludge for centuries.

The purpose of the work above is to quantify the baseline potential for electrostatic interaction between the sludge mineral surfaces and the relevant dopants. From the titration data thermodynamic constants describing proton and hydroxyl adsorption to the various minerals will be derived. Binding constants will be fit with a modified 2-layer model of the near surface potential field (Dzombak and Morel, 1990). Unraveling sorption on a thermodynamic basis is critical as it will allow our results to be portable, and applied in a general sense to tank sludges which may differ somewhat in chemistry.

Measure pH-dependent binding of dopants at a variety of ionic strengths in nitrate solutions. Using an automatic titrator technique (Balsley, et al., 1996), dopant adsorption studies can be rapidly carried out over a range of pH values. These tests will employ the same solutions as in the surface charge titration section (4.3.1). At each stable pH along the titration curve an aliquot of solution will be extracted for analysis of dopant(s). Initially the analyses will be carried out by ion chromatographic (anions) or direct current plasma emission spectroscopy (cations) techniques. If these methods lack sufficient sensitivity then an inductively coupled plasma mass spectrograph will be employed.

Interpreting the data will depend on the identification of sorption edges. Sorption edges are then used as input into FITEQL, a linear least squares fitting program which helps identify the most reasonable adsorption stoichiometry and the optimal thermodynamic constants describing the process Westall and Herbelin, 1994, This task builds on the preceding activity in that protonation/deprotonation constants are needed to extract binding constants. The resultant thermodynamic framework will allow the sorbed fraction of the various dopants in the tanks to be roughly estimated as a function of solution composition and sludge mineralogy.

Sorption onto iron and aluminum (hydr)oxides is well enough understood that data interpretation becomes just a matter of evaluating the relevant constants, using an appropriate solution chemistry, for models that already exist. For the other sludge components the course is less clear since our understanding of the basic mechanism(s) responsible for surface charge development on such phases is still being debated. Thus, a much greater interpretative effort will be required to deal with these materials.

The ability to model sorption by actual complex sludge mixes will be the final research activity in task. The experimental procedure will be essentially the same as for the pure materials, except this time complex sludge mixtures rather than individual phases will be employed. The metric for success will be that the theoretical values derived from data on individual materials lead to predictions that are within a factor of two of the Kd 's determined on sludge simulants.

Task IV Desorption and Dissolution Processes

Technical Discussion: As already noted, a number of formal problems are associated with using a Kd to define a P.A. source term. However, there is also one practical matter that can seriously impact the validity of this approach. Formally, a Kd defines the ratio of the concentration of a contaminant sorbed onto a solid to the amount left in solution - assuming the system has equilibrated. A common, though essentially fallacious, approach is to relate the Kd to the bulk contaminant concentration of the solid without first verifying that the contaminant is actually sorbed onto the surface. Transport models using this approach also assume that contaminants desorb instantaneously from mineral surfaces. This is seldom the case (see for example Barney, 1984, and Davis and Kent, 1990). In particular, problems arise from the fact that much of the contaminant content may be coprecipitated rather than sorbed (Bruno, et al., 1995) so the bulk composition may have nothing to do with the amount available for desorption.

The other part of the problem is that other processes with nothing to do with sorption may release metals that are often mistakenly identified with kinetically inhibited sorption reactions. Axe and Anderson, 1995, find that Sr sorption onto ferrihydrite, in fact, is best modeled by a two step process. The first step is the expected rapid sorption by ion exchange, and the second is modeled by simple diffusion into the microporosity of the particle (with and effective diffusion coefficient of 4x10^-13 cm²/sec). The release of components in the latter category will have nothing to do with sorption / desorption reactions. In another case Bruno et al., 1995, finds that the release of uranium from iron oxide can be modeled assuming equilibrium with a solid schoepite-like phase if its activity is diminished in proportion to the mole fraction of the coprecipitated uranium in the iron oxide. Again, using such data to generate Kd values is a meaningless exercise since sorption and desorption had nothing to do with the process.

The negative ramifications of improper Kd evaluation are significant. Literature Kd values used in conjunction with a bulk concentrations may predict an unrealistically high release rates since only a small fraction of the contaminant on the solid will actually desorb. In this instance expensive and unnecessary over-engineering may result. Alternatively, using a bulk analyses together with an analysis of the fluid with which it presumably equilibrated will produce an unrealistically high Kd if part of the contaminant is actually coprecipitated. This would, in turn, allow for predicting unrealistically low release rates. In either case the resulting performance assessment fails to provide realistic criteria for decision making.

Experimental Approach: The basis for this research lies in comparing doped and undoped materials. The basic experimental strategy is similar to that of Phase II, namely to evolve a theoretical understanding of the process based on the behavior of the individual phases, and then use this to generate a formalism that works for complex mixtures. For each material there are three approaches: (a) theoretically predicted dopant desorption based on the Task III sorption test results, (b) measure actual dopant desorption on solids that were loaded during the Task III sorption tests, and (c) measure dopant desorption from materials prepared initially were dopants was added prior to the precipitation of that sludge component(s). It is the comparison between these three approaches that provides the backbone of our analysis of desorption phenomenon.

The following four subtasks will address these issues on sequentially aged materials:

First, a relatively simple measure will be made of the relative effects of ageing. This will come from analyses of the fluids coexist with the solids at various stages in the ageing process. In the event that ageing seriously degrades or enhances the ability of the solid(s) to retain dopants there should be a large change in these concentrations. At the same time, the concentrations of the principal sludge components will also be monitored as a further indication of whether recrystallization reactions are proceeding.

Computing theoretical desorption isotherms for each sludge (or individual sludge component) will come next. This will not be a major undertaking since all the basic tools needed were developed during the sorption tests (Task III). To apply this data all we will have to do is take into account the actual solid/water ratios being employed in the experiments described below. These calculations will provide a baseline against which the results of the desorption experiments can be compared. This comparison will also provide a quantitative basis for estimating how large an error will be incurred when sorption data is used as a basis for generating a P.A. source term for radionuclide releases.

Evaluating dopant exchange capacity will be the next step in this procedure. One split of each sample will be dissolved for a complete chemical analysis. The other split will be rinsed three times with pH 5 buffered 1M LiBr solutions to exchange sorbed dopant ions from sample surfaces. Equilibration times will be only long enough to allow for the separation of the solid precipitate by centrifugation since true ion exchange is rapid enough to be completed in but a few minutes. The final rinse will be with 0.01 M LiOH to replace any H that happened to be sorbed with Li ion. Solids treated in this manner will then be washed with deionized water until an AgBr precipitate fails to form in the rinse water. These solids will then be dissolved for analysis (Li, Br, and dopants), and the amount of dopant displaced from the solid will also be checked by analyzing the spent LiBr -LiOH rinse solutions for displaced dopants. These data will address the issue of the amount of dopant that is actually available to participate in sorption/desorption processes, as opposed to the amount of a contaminant that is present in bulk.

Measuring desorption is the focus of the next subtask. Padmanabham (1983 a,b) found that desorption rates depended on both the pH and the time over which sorption continued prior to initiating the desorption part of the investigation. Axe and Anderson, 1995, present data suggesting that one underlying factor is that, at least for ferrihydrite, is that some of the process would not be related to desorption at all. It is likely that this same lack of determinacy will be observed during our studies.

The experiments will involve taking a split of various doped materials, washing them free of the fluids they have been aged in and then suspending them in dopant-free NaNO₃-NaOH solutions. The build-up in dopant concentration will be monitored until a (virtually) steady state concentration has been achieved. It is anticipated that at least two weeks will be needed for this to occur (Bruno, et al., 1995). These experiments will employ both the single phase solids and the complex sludge mixes. Solution compositions will be monitored with one of the reaction path codes (EQ3/6, REACT) to ensure that when a dopant reaches a steady state it reflects desorption equilibrium and not the saturation state of a particular compound.

Experimental desorption data will first be compared with the expected concentrations computed from sorption experiments (Task III). We anticipate that this comparison will not be very favorable and that further modeling of the desorption process will be needed. The next approach tried will be to determine if desorption from the complex sludge preparations can be modeled by using the desorption data from the individual phases mixed in the appropriate proportions. This will not explain the disagreement, but it will at least suggest a simple, though empirical, procedure whereby a predictive framework can be generated for P.A. purposes.

This subtask will determine bulk dissolution rates for the sludge components in the study. These data will be compared with the dopant release rates from the last subtask in order to determine if the anticipate long term tail on the release processes simply reflects bulk dissolution of the sludge, or the slow release of dopants from the microporosity of the materials without any significant dissolution of the matrix proper. Initially, the attempt will be made to extract this information from the experiments in the preceding task. Initially a DCP, ICP-AES or an ICP-MS will be employed to analyze the solutions. However, for the Fe and P-containing components solubilities may be so low that these compounds will not be detectable on any of these instruments. In this circumstance it will be necessary to prepare duplicate sludge mixes with radiotracer amounts of iron and phosphorous. With liquid scintillation counting the detection limits are low enough that the primary limitation on the experimental technique is in filtering the sample rather than in the analytic technique used for the analysis.

Task V Atomistic Aspects of Sorption

Technical Discussion: It is unlikely that the desorption experiments will agree closely with predictions based on the sorption study results because not all of the sorbed dopants will remain permanently exposed on the surfaces they sorbed onto initially. Thus, resolution of this discrepancy will require a truly atomistic understanding of the surface chemistry of these materials. Experimental studies by Bruno et al. (1995) illustrate the types of problems that can be anticipated. These authors found that virtually all the tracer quantities of uranium were scavenged during the rapid precipitation of Fe(OH)₃ from solution. However, as the precipitate aged some uranium was released back into solution. Concentrations of re-released uranium appeared to stabilize in a matter of weeks. However, the authors also reported that the iron hydroxide continued to recrystallize for three years, the entire duration of their study. Smith et al.(1994) provided a mathematical model for the evolution of iron oxides that indicates that the conversion of amorphous to crystalline a-FeOOH (goethite) should actually continue for about a 100 years, at which point the solid would be primarily well-crystallized a-FeOOH. By modeling the decrease in surface area accompanying this change they predicted a 10-fold decrease in the iron oxide Kd for U(VI) in the first 100 years. Thus, the chemistry of simulated as well as tank sludges clearly will evolve over time due to the evolution of mineralogy and surface area available for adsorption.

In order to obtain an atomistic picture of processes occurring at the surfaces of the sludge particles, we propose to apply atomic force and scanning tunneling microscopy. Both techniques have been successfully used to characterize surface morphology (AFM) and surface composition (STM) of iron oxide phases (Eggleston, 1993, Maurice, et al., 1995). In fact, Eggleston and Stumm (1993) were able to image sorbed chromium on hematite surfaces in aqueous solutions and derive an estimate of surface diffusion rate of Cr. Junta-Rosso and Hochella (1996) recently were able to identify surface contaminants on hematite such as 10 Å thick layers of sorbed carbon.

Experimental Approach: We propose to assess the underlying controls on sorption/desorption mechanisms by monitoring the surface roughness and possibly even image sorbed dopant atoms on the sludge components as they age. Interpretation will be aided by comparing data on sludge components that have received different pretreatments including: (a) no exposure to dopants, (b) exposure to dopants, (c) exchange with LiBr at pH 5, and (d) coprecipitation from aqueous solutions containing the dopants. AFM and STM will be the primary surface analytical techniques used. SEM/EDS and TEM will be applied to check for secondary phase formation. In planning the following research strategy we have assumed that hydrous iron oxide will be the phase primarily responsible for the uptake and release of radionuclides. If this proves to be incorrect then essentially the same activities will be carried out; but redirected at the more appropriate material(s).

The following subtasks outline the research to be carried out in this section:

Initially we will investigate how changes in surface morphology over time affect sorption/desorption processes. As the Fe-oxyhydroxide phases recrystallize, the surface area exposed to solution will likely change. Recrystallization to more stable phases may either increase or decrease the surface area depending on the number and size of crystals that form. Recrystallization due to Ostwald-ripening will reduce the total surface area as small particles dissolve to provide material for forming larger crystals.

We will address subtask 1 by examining surface processes on a-FeO(OH) (goethite) and/or g-FeO(OH) (lepidocrocite) using a Digital Instruments Nanoscope IIIa Multimode AFM. Both minerals can be obtained as euhedral multi-micron-sized crystals. In this crystal size range goethite typically exhibits a prismatic or fibrous habit, whereas lepidocrocite forms tabular or blady crystals. Although the grain sizes are small, techniques are available for mounting such grains for imaging both in air (using contact and Tapping ModeÔ AFM) and in fluids ( Dove and Chermak, 1994.) The euhedral crystals will insure that crystallographically-identifiable relative flat surfaces can be examined. Most imaging will be done in air on samples that have been reacted in ex-situ experiments. Microscope software will be used to quantify changes in surface morphology, e.g., step height, surface roughness, fractal dimension.

The pH variations that will exist over time in the tanks may accelerate or retard the recrystallization process. Therefore, we will attempt to examine the recrystallization process in simulated high pH solutions using the fluid cell and TMAFMÔ. Care will be taken to consistently identify the mineral face under study. Recrystallization of metastable to stable phases will be conducted in solutions at pH 7, 9, and 11 containing (1) no Fe, and (2) Fe at two supersaturation concentrations. We will also attempt an examination of Ostwald ripening at the three pH values. We will mount two sizes of crystals, one group will be submicron to 1 micron, and the second group will be 5 to 10 microns. The small sized-crystals will be chosen to have surface energies higher than that attained at the equilibrium solubility. Therefore, these crystals should dissolve and material should migrate to the larger crystals and reprecipitate. In the closed fluid cell, diffusion will be the mode of transport of material through solution. Therefore, grains will be placed close enough together to allow diffusion on the time-scale of an in-situ experiment (up to a few hours.) In this way we can assess the affects on surface morphology due to two recrystallization processes.

The fate of the dopants themselves is the second topic of interest. As aging progresses do dopants behave as molecular sorbates or do they segregate into distinct phases? If the latter process dominates is there any indication that the surfaces of other phases (i.e., the Fe-oxyhydroxides) will be covered by the precipitates? Surface coverage by secondary precipitates may have an effect on recrystallization by reducing the active surface area.

To address the second subtask, we will select optimal experimental conditions for obtaining images of surface morphological changes during recrystallization by either aging or Ostwald ripening. Then we will apply STM to the selected mineral surface in an attempt to mimic the approach of Eggleston and Stumm (1994) used to measure Cr-adsorption. We will adjust the concentration of contaminant and pH of solution as in the above experiments for subtask 1 in order to evaluate the range of sorption processes from the molecular level up to the coprecipitate level. If precipitates are observed to form, we will compare their morphology in parallel AFM experiments.

Our approach to quantifying the underlying causes for sorption-desorption rests on monitoring the surface chemistry of the sludge components as they age. Interpretation will be aided by the comparison of sludge components that have received different treatments: e.g. (a) prior to being exposed to various dopants (b) after exposure to dopants, (c) after LiBr-pH5 exchange, and (d) and finally iron (hydr)oxides that was precipitated with the dopant already in solution; before and after treatment with the LiBr-pH5 solution. This work will depend heavily on AFM and STM techniques, though some TEM support may also be needed. The results from Phase V will be combined with the macroscopic system studies (Tasks III and IV) to generate models that predict the long-term release behavior of radionuclides that reside with the solid sludge components.

5.3 Summary:

It is not surprising that performance assessment source terms have typically relied on overly simplistic representations of the near field chemistry. These are very complex issues that will require many years of intensive research before they are resolved. Such commitments of time and resources are simply incompatible with the tight schedules of most project-driven performance assessments. However, unless the DOE develops credible source terms for such models it runs the risk of eventually having its entire performance assessment methodology called into question.

This proposal presents a graded approach to resolving the source term

problem. The first two tasks are directed at fabricating and characterizing sludge phases. The next two tasks will identify how in bulk the phases will retain and release radionuclides over time. The last task address the more difficult issue of providing the "smoking gun" needed to prove that the models used in earlier tasks do, in fact, have some basis at an atomistic level.

The body of this proposal is occupied by a rather detailed accounting of the individual steps. This rigor is necessitated by the fact that: (1) sludges are complex chemical systems, and (2) to produce programmatically useful results requires a multidisciplinary approach. Without this degree of planning and organization it would be difficult to argue that anything but chaos would be the outcome of funding the proposal. However, In spite of this structure one should not loose sight of the fact that the same types of activities are required for any experimental program, whether the objective is "applied" or "pure" research. The two basic scientific themes which form the core of this proposal are the needs to:(1)gain fundamental insights into the early ageing processes of hydrous inorganic precipitates, and (2) correlate the solid phase chemistry of these materials with changes in their surface properties and how they interact with surrounding fluids. Success in either venture will have impact that extends far beyond the boundaries set by nuclear waste disposal issues.

5.4 Collaborative Arrangements:

Most of the principals involved are at Sandia National Laboratories, in Albuquerque, N.M. However, by virtue of his experience in the TEM and NMR characterization of tank waste phases, we have arranged with Dr. J. Liu at PNNL to collaborate with us.

5.5 Literature Cited:

1. Axe, L., Anderson, P.R., 1995, Sr Diffusion and Reaction within Fe Oxides: Evaluation of the Rate Limiting Mechanism for Sorption, Jour. of Colloid and Interface Science, v. 175(#1), p. 157-165.

2. Balsley S. D., Brady P. V. and J. L. Krumhansl. 1996a. Iodide Retention by Cinnabar (HgS) and Chalcocite. Environ. Sci. Technol. In Press.

3. Balsley, S.D., Brady, P. V., and Krumhansl, J.L., 1996b, Anion Scavengers for Low Level Radioactive Waste Backfills, Jour. of Env. Eng. (in review).

4. Bethke C.M., 1994, The Geochemist's Workbench, A Users guide to Rxn, Act2, Tact, React, and Gtplot, 213 pp.

5. Berger, A., 1996, Controls on Heavy Metal Mobility at the Pecos Mine Operable Unit San Miguel County, N.M., Ph.D Thesis, University of Illinois at Urbana-Champaign (Draft Copy).

6. Barney, G.S., 1984, Radionuclide Sorption and Desorption Reactions with Interbed Materials from the Columbia River Basalt Formation, in: Barney, G.S., Navratil, J.D., and Schulz, W.W.(eds), Geochemical Behavior of Disposed Radioactive Waste, American Chemical Society Symposium Series 246, Washington, D.C., p. 3-23.

7. Brady, P.V., 1994, Alumina Surface Chemistry at 25, 40 and 60^oC, Geochimica et Cosmochimica Acta, vol. 58(3), p. 1213-1217.

8. Brady P. V., Cygan R. T. and Nagy K. L. 1996. Molecular Controls on Kaolinite Surface Charge and Metal Sorption. in Sorption of Metals on Earth Materials Ed. E. A. Jenne. Academic Press. In Press.

9. Brady P. V. 1996. Surface-Controlled Reactivity of Metal Carbonates. in Aqueous Chemistry and Geochemistry of Metal Oxides and Hydroxides and Related Materials Ed. J. Voight, W. Casey, B. Bunker and L. Crossey). Materials Research Society. In Press.

10. Brady P. V. 1992, Surface Complexation and Mineral Growth, in Water-Rock Interaction, Kharaka, Y., and Maest, A.S., eds, Proc. 7th International Symposium on Water-Rock Interaction WR1-7, Park City, Utah, 13-18 July, 1992, p. 85-88.

11. Bruno, J., de Pablo, J., Duro, L., and Figuerolia, 1995, Experimental Study and Modeling of the U(VI)-Fe(OH)₃ Surface Precipitation/Coprecipitation Equilibria, Geochimica et Cosmochimica Acta, vol. 59(20), p. 4113-4123.

12. Carroll, S.A., 1993, Precipitation of Nd-Ca Carbonate Solid Solutions, at 25^oC, Geochimica et Cosmochimica Acta, vol. 57, 3383-3393.

13. Colton, N.G., 1994, Sludge Pretreatment Chemistry Evaluation: Enhanced Sludge Washing Separation Factors, TWRSPP-94-053, Pacific Northwest Laboratory.

14. Christoffersen, J., 1980 Kinetics of Dissolution of Calcium Hydroxyapatite III, Nucleation-Controlled Dissolution of a Polydispersed Sample of Crystals; Journal of Crystal Growth, v.49, p. 29-44.

15. Davis, J.A., and Kent, D.B., 1990, Surface Complex Modeling in Aqueous Geochemistry, Mineral-Water Interface Geochemistry, in Reviews in Mineralogy, MSA vol. 23, p.203.

16. DiCenso, A.T., Amato, L.C., Lambie, R.W., Franklin, J.D., Seymour, B.J., Johnson, K.W., Stevens, R.H., Remund, K.M., Sasaki, L.M., Simpson, B.C., 1995, Tank Characterization Report for Single-Shell Tank 241-C-109, WHC-SD-WM-ER-402 REV 0.

17. Dove, P.M., and Chermak, J.A., 1994, Mineral Water Interactions: Fluid Cell Applications of Scanning Force Microscopy, in Scanning Probe Microscopy of Clay Minerals, K.L. Nagy and A.E. Blum eds, CAMS Workshop Lectures Series, vol. 7, p. 3-90.

18. Dzombak, D.A., and Morel, F.F.M., 1990, Surface Complexation Modeling, Hydrous Ferric Oxide, Wiley, New York, 393 p.

19. Ebra, M.A., Oxalic Acid Cleaning of Tank 24H, Technical Division, Savannah River Laboratory Memorandum Sept. 9, 1985, DDPST-85-782 (Acc. No 189389).

20. Eggleston, C.M., and Stumm, W., 1994. Scanning Tunneling Microscopy of Cr(III) Chemisorbed on a-Fe₂O₃ (001) Surfaces from Aqueous Solutions: Direct Observation of Surface Mobility and Clustering, Geochimica et Cosmochimica Acta, vol. 57, p. 4843-4850.

21. Eibling, R.E., and Fowler, J.R., 1983, Updated Waste Composition at the Savannah River Plant, Technical Division, Savannah River Laboratory Memorandum DPST-83-313, Feb. 16, 1983.

22. Fowler, J.R., 1982, Waste Sludge Composition at the Savannah River Plant, American Nuclear Society Transactions, v. 41, p. 159-160.

23. Fowler, J.R., and Wallace, M., 1980, CRC Zeolite in SRP Waste, Technical Division, Savannah River Laboratory Memorandum DPST-80-488, Dec. 2, 1980.

24. Georgeton, G.K., and Hester, J.R.,1995, Characterization of Radionuclides in HLW Sludge Based on Isotopic Distribution in Irradiated Assemblies (U), WSRC-TR-94-0562 (Savannah River High Level Waste Engineering, HLW Engineering Support), 39 pp.

25. Goslen, A.Q., 1986, Tank 19 Salt Removal, DPSP-84-17-7.

26. Huang, C.P., and Stumm, W., 1973, Specific Adsorption of Cations on Hydrous, g-Al₂O₃, J. Coll., Interface Sci., vol. 43, p. 409-420.

27. Hobbs. D.T., 1995, Concentrations of Metals and Non-Metals in Alkaline Waste Slurries, Westinghouse Savannah River Company, Savannah River Technology Center Memorandum WSRC-TR-96-0058, Feb. 28, 1995.

28. Junta-Rosso. J.L., Hochella, M.F., 1996, The Chemistry of Hematite {001} Surfaces, Geochimica et Cosmochimica Acta, vol. 60(2), p. 305-314.

29. Liu, J., Virden, J.W., Bunker, B.C., Lin Song, 1994, Hanford Tank Sludges: Investigation of Physical and Chemical Properties Using Transmission Electron Microscopy (abs.), 208 ACS national meeting, Aug. 21-26, paper # I&EC-96.

30. Liu, J., Thomas, L.E., Chen, Y.L., and Wang. L-Q., 1995, "C 4.0 Sludge Characterization Studies", in Tank Waste Treatment Science Task Quarterly Report for April-June 1995 (J.P. LaFemina Task Leader), p. 4.1-4.15.

31. Maheras, S.J., Rood, A.S., Magunson, S.O., Sussman, M.E., and Bhatt, R.N., 1994, Radioactive Waste Management Complex Low Level Waste Radiological Performance Assessment, EEG-WM-8773, EG&G Idaho, Idaho Falls, May, pp. 3-31 to 3-35.

32 Maurice, P.A., Hochella, M.F., Parks, G.A., Sposito, G., and Schwertmann, U., 1995, Evolution of Hematite Surface Microtopography Upon Dissolution by Simple Organic Acids, Clays and Clay Minerals, v. 43(#1), P. 29-38.

33. Morrison, S.J., and Spangler, R.R., 1992, Extraction of Uranium and Molybdenum from Aqueous Solutions: A Survey of Industrial Materials for Use in Chemical Barriers for Uranium Mill Tailings Remediation, Environ. Sci. Technol. vol. 26, p. 1922-1931.

34. MIMES, EG&G Idaho, and WSRC 1944, Radiological Performance Assessment for the E-Area Vaults Disposal Facility, WARC-RP-94-218, Rev. ). Aiken, S.C.:Westinghouse Savannah River Company, Appendix D.

35. Nagy, K.L., 1994, Application of Morphological Data Obtained using Scanning Force Microscopy to Quantification of Fibrous Illite Growth Rates, Clay Minerals Society Workshop Lectures, v. 6, Scanning Probe Microscopy of Clay Minerals, (eds. K.L. Nagy and A.E. Blum).

36. Nagy, K.L., 1995, Dissolution and Precipitation Kinetics of Sheet Silicates, In "Chemical Weathering Rates of Silicate Minerals", Reviews in Mineralogy, V. 31, Min. Soc. Amer., 173-233.

37. Norton, M.V., 1994, Laboratory Testing In-Tank Sludge Washing, PNL-10153, p.3-4.

38. Padmanabham, M, (1983a) Adsorption-Desorption Behavior of Copper(II) at the Goethite-Solution Interface, Aust. J. Soil Res., vol. 21, p. 309-320.

39. Padmanabham, M, (1983b) Comparative Study of the Adsorption-Desorption Behavior of Copper(II,), Zinc(ii), Cobalt(II) and Lead(II) at the Goethite-Solution Interface, Aust. J. Soil Res., vol. 21, p. 515-525.

40 Popolitov, V.I., 1993, Crystallization of Aluminum Phosphate in Aqueous Solutions of HCl, HNO₃, KF and NaOH, Inorganic Materials, v. 29(5), p. 663-666.

41. Rapko, B.M., Blanchard, D.L., Colton, N.G., Felmy, A.R., Liu, J., Lumetta, G.J., 1996, The Chemistry of Sludge Washing, and Caustic Leaching Processes for Selected Hanford Tank Wastes, PNNL-11089, UC-721.

42. Ridkosil, T., Sejkora, J., and Srein, V.L., Smrkovecite, 1996, Monoclinic Bi2O(OH)(PO4), a new mineral of the Atelestite Group, Neues Jahrbuch fur Mineralogie-Monatshefte, v. 33(#3), pp. 97-102.

43. Russell, R.L., and Smith, H.D., 1996, Simulation and characterization of a Hanford high-level waste slurry, PNNL-11293, 50pp.

44. Russell, R.L., and Smith, H.D., 1996b, Simulant specification for the blended privatization tank waste for the high level waste vitrification process specified in the RFP, PNNL 11293, 22 pp.

45. Rykken, L.E., Wilson, J.A., and Hardt, T.L., 1985, Analytic Characterization of West Valley High-Level Waste Sludge, Waste Management 85, p. 603-609.

46. Shaman, R. Chau, N., and Jennrich, E.A., 1992, The Source Computer Codes: Models for Evaluating the Long-Term Performance of SWSA Disposal Unites, Version 1.0: User's Manual, RA-9005/8-1 Rogers and Associated Engineering Corporation, Salt Lake Cit, Utah, April, 1992.

47. Smith, R.W., Walton, J.C., and Rahman, M., 1994, Effects of Recrystallization on Time Variant Sorption of Radionuclides onto Steel Corrosion Products, Scientific Basis for Nuclear Waste Management, Mat. Res. Soc. Symp. Proc., vol. 333, p. 713-718.

48. Stone, A.T., and Morgan, J.J., 1987, Reductive Dissolution of Metal Oxides, in: Aquatic Surface Chemistry, W. Stumm (ed.) Wiley-Interscience, 221-254.

49. Stumm, W., 1992, Chemistry of the Solid-water Interface, Wiley, New York, 428.

50. Stumm, W., and Furrer, 1987, The Dissolution of Oxides and Aluminum Silicates; Examples of Surface-Coordination-Controlled kinetics, in: Aquatic Surface Chemistry, W. Stumm (ed.) Wiley-Interscience, 197-220.

51. Van Cappellen, P., Charlet, L., Stumm, W., and Wersin, P., 1993, A Surface Complexation Model of the Carbonate Mineral-Aqueous Solution Interface, Geochimica et Cosmochimica Acta, vol. 37, p. 3505-3518.

52. Westall, J., and Herbelin, A., 1994, FITEQL, A Computer Program For Determination of Chemical Equilibrium Constants from Experimental Data, Department of Chemistry, Oregon State University, Chervils, Oregon, Oct. 4, 1994, Report 94-01.

53. Zhukhlistov, A.P., Zvyagin, B.B., 1976, Determination of the crystal structures of chapmanite and bismuthferrite by high-voltage electron diffraction, Sov. Phys. Crystallogr. v. 22(#4), p. 419-423.

Appendix I - Biographical Sketches:

Jim Krumhansl joined the Sandia Staff in October, 1976. His first three years were spent managing the Conasauga Near Surface Heater Test. This multi­million dollar field exercise was conducted at Oak Ridge, and involved collecting and synthesizing thermal, mechanical, and chemical data on complex thermally induced processes. His next task was managing the geochemical and near­field interactions facet of the U.S Subseabed Nuclear Waste Disposal Program. In this capacity he integrated programmatic research in metallurgy, radionuclide chemistry and migration, near­field hydrothermal geochemistry, and radiation effects. Somewhat later he was heavily involved with both the WIPP and the Yucca Mountain repository programs. Here he addressed problems in metallurgy, brine chemistry, clay mineral diagenesis, in situ borosilicate glass wasteform tests, cement/concrete durability and development of backfills. He was also involved to a significant degree in a NRC­sponsored near field natural analogue program to address the large scale effects of heating tuffaceous rocks. His current involvement with the WIPP is a continuation of earlier studies in clay mineralogy. Recently, he also has undertaken an experimental program to measure actinide Kd values on the mineral apatite in WIPP brines. Recent activities in the area of material science include involvement in Sandia's crystalline silicotitanate program, where he is involved in crystallographic and phase stability studies. He was part of the team that received the R&D 100 award for developing this material. He is currently involved in: (1) evaluating MgO pH-buffer backfills for the WIPP, (2) in providing chemical inputs for a generic P.A. activity on tank closure funded through Tank Focus Area program, (3) is carrying out experimental studies on the irreversible sorption of Cs by various clays , and (4) performing research to identify and test inorganic anion exchange materials for use in reactive barriers and in-tank radionuclide "getters".

Selected Publications:

1. Krumhansl, J.L., Stein, C.L., Jarrell, G.D., and Kimball, K.M., 1991, Summary of WIPP Room B Heater Test: brine and backfill material data, SAND90­06267, 40 pp.

2. Krumhansl, J.L., Kimball, K.M. and Stein, C.L., 1991, A Review of WIPP Repository Clays and their Relationship to Clays of Adjacent Strata, SAND90­0549, 65 pp.

3. Stein, C.L., Kimball, K.M., Westrich, H.R. and Krumhansl, J.L., 1991, Chromium in Groundwater Monitoring Wells at the Chemical Waste Landfill, SNL Technical Area III, SAND91­0131, 51 pp.

4. Krumhansl, J.L., Hinkebein, T.E., and Myers, J., 1991, The Hydrothermal Stability of Cement Sealing Materials in the Potential Yucca Mountain High Level Nuclear Waste Repository, in Advanced Cementitious Systems: Mechanisms and Properties, F.P. Glasser et al., Editors, Materials Research Society Symposium Proceedings, V. 245, Boston, Massachusetts, Dec. 2­4, 1991, 105 ­110).

5. Krumhansl, J.L., and Lambert, S.J., 1991, Degradation of Portland Cements Exposed to Evaporite Brine at Hydrothermal Temperatures, in Advanced Cementitious Systems: Mechanisms and Properties, F.P. Glasser et al., Editors, Materials Research Society Symposium Proceedings, V. 245, Boston, Massachusetts, Dec. 2­4, 1991, p. 123 ­ 128.

6. Krumhansl, J.L., 1993, The Accelerated Testing of Cements in Brines, Proceedings of the American Ceramic Society, April 19­22, 1993, Cincinnati, Ohio, "Cement-Based Materials", p. 153-164.

7. Stockman, H.W., Krumhansl, J.L., Ho, C., and McConnell, V.S., 1994,The Valles Natural Analogue Project, NUREG/CR­6221 (SAND94­0650, FIN# A1824), 112 pp.

8. Molecke, M. A., Sorensen, N.R., and Krumhansl, J.L., 1994, Results from Simulated Contact-Handling Tansuranic Waste Experiments at the Waste Isolation Pilot Plant, Scientific Basis for Nuclear Waste Management XVII, Mat. Res. Soc. Symp. Proc., v. 133, p. 681-686.

9. Balsley, S.D., Brady, P.V., Krumhansl, J.L., and Anderson, H.L., 1996, Iodide Retention by Metal Sulfide Surfaces: Cinnabar and Chalcocite, Environmental Science and Technology, v. 36(10), p. 3025 - 3027.

10. Balsley, S.D., Brady, P.V., and Krumhansl, J.L., 1996, Anion Scavengers for Low Level Radioactive Waste Backfills, Jour. of Env. Eng. (in review).

11. Zheng, Z., Philip, C.V., Anthony, R.G., Krumhansl, J.L., Trudell, D.E., and Miller, J.E., 1996, Ion Exchange of Group I Metals by Hydrous Crystalline Silicotitantates, I&EC Research, in press.

Patrick V. Brady became a member of the technical staff in the Geochemistry Department in 1993 after two and half years as an Assistant Professor in the Dept. of Geological Sciences at Southern Methodist University in Dallas, Texas. Prior to joining Sandia he received a B.S. degree in Geology from the University of California at Berkeley, an M.S. and Ph.D. from Northwestern University, and spent a year working as a post-doctoral appointee for Werner Stumm at the Swiss Federal Institute of Technology - Institute for Water Resources and Water Pollution Control (ETH-EAWAG). His primary interests are soil and mineral surface chemistry, and global environmental change. In the past 7 years he has edited one book and been first or second author on over twenty peer-reviewed journal articles. In the past two years waste-containment work done for the WIPP, the US-NRC and Westinghouse-Hanford has resulted in numerous publications of programmatic relevance.

Selected Publications:

1. Brady P. V. (Editor) 1996. Physics and Chemistry of Mineral Surfaces. CRC Press 352p. BOOK.

2. Brady P. V., Papenguth H. W. and Kelly J. W. 1996 .Ca, Mg and Nd Adsorption to Dolomite Surfaces in DOE/SNL review.

3. Brady P. V., Cygan R. T. and Nagy K. L., 1996, Molecular Controls on Kaolinite Surface Charge, J. Coll. Interf. Sci. In Press.

4. Brady P. V., Cygan R. T. and Nagy K. L. 1996. Molecular Controls on Kaolinite Surface Charge and Metal Sorption, in Sorption of Metals on Earth Materials Ed. E. A. Jenne. Academic Press. In Press.

5. Brady P. V. 1996. Surface-Controlled Reactivity of Metal Carbonates. in Aqueous Chemistry and Geochemistry of Metal Oxides and Hydroxides and Related Materials Ed. J. Voight, W. Casey, B. Bunker and L. Crossey). Materials Research Society. In Press.

6. Brady P. V., J. L. Krumhansl and H. W. Papenguth. 1996. Surface Complexation Clues to

Dolomite Growth, Geochim. Cosmochim. Acta (60) 727-731.

7. Brady P. V. and W. A. House 1996. Surface­Controlled Dissolution and Growth of Minerals (Chapter 4 in Physics and Chemistry of Mineral Surfaces, CRC Press Ed. P.V. Brady) p. 221­302.

8. P. V. and J. M. Zachara. 1996. Geochemical Applications of Mineral Surface Science. (Chapter Brady 5 in Physics and Chemistry of Mineral Surfaces, CRC Press Ed. P.V. Brady) p. 303­352.

9. Brady P. V. and M. W. Kozak. 1995. Geochemical Engineering of Low Level Radioactive

Waste in Cementitious Environments, Waste Management (15) 293-301, 1995.

10. Brady P. V., 1994, Alumina Surface Chemistry at Elevated Temperatures. Geochim. Cosmochim. Acta GCA (58) 1213-1217.

11. Brady P. V. and Walther J. V., 1992, Surface Chemistry and Silicate Dissolution at Elevated Temperatures. Amer. J. Sci. (292) pp. 639-658.

12. Brady P. V. 1992, Silica Surface Chemistry at Elevated Temperatures, Geochim. Cosmochim. Acta (56) 2941-2946.

13. Brady, P.V., Cygan, R.T., and Nagy, K.L., 1996. Molecular Controls of Kaolinite Surface Charge, J. Colloid Interface Sci (in press).

Kathyrn L. Nagy has been a Senior Member of the Sandia Technical Staff since 1994. Her current research interests are the kinetics of water/rock interactions, mineral surface chemistry, clay chemistry, scanning probe (atomic force) microscopy, isotopic exchange in hydrothermal systems, experimental hydrothermal geochemistry, sedimentary diagenesis and theoretical (computational) geochemistry. Her current Sandia research is on WIPP clay backfill longevity and AFM studies relating to Cs-sorption on kaolinite clay. Kathy is a widely recognized expert on the application of scanning probe microscopy, and was coeditor of the leading reference work on the geologic applications of this class of techniques. Publications and abstracts since 1992 number 11. Prior to joining the Sandia Staff Kathy was a research geologist with Exxon Production Research Company. In this capacity she employed AFM technology to research growth rates of kaolinite and illite. Between the time when she received her Ph.D. from Texas A&M University (1988) and when she joined the staff of Exxon (1991) she was associated with Yale University, first as a Post-Doctoral fellow and then as an Associate Research Scientist.

Selected Publications:

1. Nagy K. L., Cygan, R. T., Sturchio N. C., Chiarello, R. P., 1996, Heterogeneous nucleation and growth of clays: gibbsite and brucite on muscovite. Invited abstract for symposium on "Structure and Reactivity of Mineral Surfaces", 6th V. M. Goldschmidt Conference, Heidelberg.

2. Nagy, K.L., 1995, Dissolution and Precipitation of Sheet Silicates, in Chemical Weathering Rates of Silicate Minerals, A.F. White and S.L. Brindle (ed) , Reviews in Mineralogy, vol. 31, Min. Sac. Amer., p. 173-233.

3. Nagy, K.L., 1994, Application of MorphologicaL Data Obtained Using Scanning Force Microscopy to Quantification of Fiberous Illite Growth Rates, Clay Minerals Society Workshop Lectures Vol. 6, Scanning Probe Microscopy of Clay Minerals, ed., K.L.Nag and A.E. Blum.

4. Nagy, K.L., and Lasaga, A.C., 1993, Simultaneous Precipitation Kinetics of Gibbsite and Kaolinite at 80^oC and pH3, Geochimica et Cosmochimica Acta, vol. 57, 4329-4335.

5. Nagy, K.L., and Lasaga, A.C., 1992, Dissolution and Precipitation Kinetics of Gibbsite at 80^oC and pH 3: The Dependence on Solution Saturation State, Geochimica et Cosmochimica Acta, vol. 56, 3093-3111.

6. Nagy, K.L., Cygan, R.T., Sturchio,N.C., Chiarello, R.P., 1996, Heterogeneous Nucleation and Growth of Clays: Gibbsite and Brucite on Muscovite, Invited abstract for Symposium "Structure and Reactivity of Mineral Surfaces", the V.M. Goldschmidt Conference, Heidelberg.

7. Nagy, K.L., 1995, Dissolution and Precipitation of Sheet Silicates, in Chemical Weathering Rates of Silicate Minerals, A.F. White and S.L. Brindle (ed) , Reviews in Mineralogy, vol. 31, Min. Sac. Amer., p. 173-233.

8. Nagy, K.L., 1994, Application of Morphological Data Obtained Using Scanning Force Microscopy to Quantification of Fiberous Illite Growth Rates, Clay Minerals Society Workshop Lectures Vol. 6, Scanning Probe Microscopy of Clay Minerals, ed., K.L.Nag and A.E. Blum.

Dr. Jun Liu has a background in materials synthesis, colloidal and surface chemistry, and materials characterization, He initiated the project at PNNL to characterize the solid phases in tank waste materials using electron microscopy, and has been principal investigator for projects in this area since he joined PNNL in 1992. He also actively participates in other tank waste projects on the surface and colloidal preoperties of tank sludges. He has published widely in the area of materials synthesis and colloidal chemistry in professional journals.

Selected Publications:

1. Liu, J., Shih, W.Y., Sarikaya, M., and Aksay, I.A., 1990, Fractal Colloidal Aggregates with Finite Interactions, Physical Review A, v. 41(4), p. 3206-3213.

2. Liu, J., Shih, W.Y., Kikuchi, R., and Aksay, A.I., 1991, Clustering of Binary Colloidal Systems: Theory, Journal of Colloidal and Interface Science, v. 142(2), p. 357-368.

3. Rieke, P.C., Marsh, B.D., Wood, L.L., Tarasevich, B.J., Campbell, A.A., Gryxell, G.E., Graff, G.L., Song, L., Liu, J., Fryxell, , 1994, Aqueous Solution Deposition Kinetics of Iron Oxyhydroxide in Sulfonic Acid Terminated Self Assembly Monolayers, Langmuir, v. 11, p. 689-692.

4. Liu, J., Kim, A.Y., Virden, J.W., and Bunker, B.C., 1995, Effect of Colloidal Particles on the Formation of Ordered Mesoporous Materials, Langmuir, v. 11, p. 689-692.

5. Liu, J., Wang, L.Q., Bunker, B.C., Virden, J.W., 1995, Effect of Hydrolysis on the Colloidal Stability of Fine Alumina Suspensions, J. Materials Science and Engineering-A, v. 35, p. 435-439.

6. Liu, J., Kim, A.Y., and Virden, J., 1996, Preparation of Mesoporous Spherulite in Surfacant Solutions, J. Porous Materials, v. 2., p. 201-205.

7. Bruinsma, P., Wang, Y., Li, X., Liu, J., and Bunker, J., 1997, Rheological and Solid State Liquid-Separation Properties of Mixed Suspensions of Gibbsite and Boehmite, J. Coll. Interface Sci., in press.

Appendix II - Facilities and Resources:

Laboratory facilities at Sandia are well suited to supporting this study. In-house equipment belonging to the geochemistry department includes: a Spectraspan VIII direct coupled emission spectrometer (DCP) for cation analysis, a 2000i Dionix ion chromatograph for anion analysis, a Philips X-ray diffractometer (XRD), a JEOL T300 scanning electron microscope equipped with X-ray energy dispersive capability (SEM/EDS), a recently purchased Digital Instruments Nanoscope IIIa Multimode AFM (AFM/STM), several autotitrators and the normal assortment of analytic balances, centrifuges, pH meters etc. that are needed to make an analytic facility functional. In addition, because of the Geoscience connection, we have a well stocked lapidary facility for sectioning and preparing ceramic materials. We are also currently collaborating with other groups to bring on line a radiochemical facility able to handle tracer amounts of various radioactive isotopes. Computer software needed to support this research is also on hand and operational. This includes the React and EQ3/6 codes for reaction path and solution equilibria calculations. FITEQL, the state of the art code for fitting surface chemical results is set up to run, as is HYDRAQL, another surface speciation code. Finally the Molecular Systems, Inc (formerly BIOSYM) package is available for molecular modeling and crustal structure applications. Further afield, over the years Sandia's analytic support groups have routinely made available ICP/AES, ICP/MS, TEM, IR, NMR, and XPS support to the geochemistry department.

PNNL laboratory is similarly well equipped to carry out its mission having a wide range of analytic tools available. These have already been used to characterize actual "hot" wastes from various on-site tanks. in particular Dr. Liu has direct access to the TEM which will be used in characterizing the waste simulants from this program.

Appendix III - Budget and Budget Justification:

Amounts in the sum of $287, $299K, and $277K are requested for Sandia National Laboratories for each successive year of this grant. The bulk of this money will cover salaries for the principal investigators. J.L. Krumhansl will perform much of the experimental work and chemical analyses. P.V. Brady will oversee the acquisition of data relating to surface chemical processes, as well interpret the laboratory measurements. K.L. Nagy will carry out the AFM and STM studies In addition, funds are included for a full time technician (or postdoctoral appointee). Because much of the work involves experimental activities funds are requested for both the purchase of laboratory supplies and the maintenance of equipment. No significant equipment purchases are anticipate. Travel funds have been included to cover the travel needed to coordinate the program (year 1) and attendance of professional society meetings (years 2 and 3).

A separate request for $103K, $120K, and $103 is made to cover PNNL activities for the three year program. This includes both salary and equipment maintenance. J. Liu will carry out TEM and NMR characterization studies on waste form simulants and aid in the synthesis and interpretation of the data gained by other researchers on the program.

Appendix IV - Current and Pending Support:

Preparation of this proposal has obviously benefited from a variety of programs, past and presently active. The Hanford Getter Development Program was active in the past - and may again receive funding in FY98. A continuation of the CST program is currently directed at identifying inorganic anion exchangers, though funding for FY97 was only $50K. Funding of performance assessment activities related to Tank Focus Area concerns continues, though these are principally paper studies. NRC support continues for examining the sorption of Cs onto clays. A Sandia-funded LDRD provides support for studying the irreversible sorption of radionuclides onto a variety of soil components, notably clays and iron oxides.

Acknowledgment and Disclaimer