SRT-LWP-2000-00189

This report was prepared as an account of work sponsored by an agency of the United States Government. Neither the United States Government nor any agency thereof, nor any of their employees, makes any warranty, express or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product or process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific commercial product, process or service by trade name, trademark, manufacturer, or otherwise does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United States Government or any agency thereof. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government or any agency thereof.

This report has been reproduced directly from the best available copy.

Available for sale to the public, in paper, from:  U.S. Department of Commerce, National Technical Information Service, 5285 Port Royal Road, Springfield, VA 22161,  phone: (800) 553-6847,  fax: (703) 605-6900,  email:  orders@ntis.fedworld.gov   online ordering:  http://www.ntis.gov/support/ordering.htm

Available electronically at  http://www.osti.gov/bridge/

Available for a processing fee to U.S. Department of Energy and its contractors, in paper, from: U.S. Department of Energy, Office of Scientific and Technical Information, P.O. Box 62, Oak Ridge, TN 37831-0062,  phone: (865 ) 576-8401,  fax: (865) 576-5728,  email:  reports@adonis.osti.gov

The current version of crystalline silicotitanate (TAM5) is commercially available from UOP under the trade name IONSIV^â IE-911. TAM5 was extensively tested by several researchers and was determined as the best currently available material for removing radioisotopes from various types of nuclear wastes salt solutions stored at various D. O. E. sites. The studies at Savannah River Technology Center (SRTC) indicated that the CST granules tend to leach into the nuclear waste simulants as it is processed by the ion exchange columns that is packed with CST granules from UOP. We, at Texas A&M University, agreed to conduct research to compliment the efforts at SRTC so that IONSIV^â IE-911 could be used for the treatment of nuclear waste stored at the D. O. E. Savannah River facility. After consultation, we developed a Task Plan in January 2000. According to the agreement between Westinghouse Savannah River Company, Savannah River Technology Center, Aiken SC 29808 and, College Station, TX 77843, synthesis and the performance evaluations of crystalline silicotitanates (CST) were performed the during period of April 1 – September 30, 2000. Our main goals were delivery of a kilogram of CST (TAM5-4) synthesized at Texas A&M University in July to SRTC, performance evaluation of CST in nuclear waste simulants, and consultation mainly by telephone.

We have conducted several series of ion exchange experiments with eight simulants and samples from four batches of CST comprising of: powder from the UOP and our CST samples, DG141 and TAM5-4 and TAM5-5. We conducted duplicate experiments with each source of CST. We also analyzed the nuclear simulant after the ion exchange experiments for Ti, Si, and Nb that might have leached from the CST. Each simulant was prepared based on our standard simulant (5.1 M NaNO₃, 0.6 M NaOH) but adding other ingredients such as oxalic acid, sodium carbonate, and hydrogen peroxide.

One of the simulants with oxalic acid contained excess oxalic acid, which did not dissolve. Another simulant contained a concentration of 0.0016M oxalic acid. It was observed from experiments involving both simulants, that oxalic acid did not have any effect on the performance of CST. Sodium carbonate showed some effect on the Kd values, which is noticeable especially in one experiment involving 0.48 M Na₂CO₃ and 5 days of equilibration. The observed carbonate effect is much smaller than what is predicted using our models. We used simulants with hydrogen peroxide at three concentration levels 1M, 0.1M, and 0.0025M. We have observed that CST could easily be dissolved by a simulant with 1M H₂O₂ and moderately by a simulant with 0.1M H₂O₂. However, simulant with 0.0025 M H₂O₂ does not have any effect.

About ten years ago at Texas A&M University, a new crystalline silicotitanate, (TAM5) was synthesized by hydrothermal synthesis using sodium hydroxide and the alkoxides of titanium and silicon.^1,2 The crystalline structure of TAM5 is similar to that of Sitinakite, a rare Russian mineral with the pharmacosiderite structure.^3,4 The structures of Sitikanite and pharmacosiderite are well established. TAM5 has straight uniform channels of 3.5 Å dia. and belongs to a new class of materials similar to Molecular Sieves. Typical crystal sizes are in the range of 0.1 – 0.4 µm. Sodium ions are ion exchangeable with protons and also with other alkali metal ions without affecting the crystal structure. Since the channel size of TAM5 and the diameter of Cs ions are very close, ion exchange with Cs ions is almost irreversible.

Replacing about 25% of titanium atoms with niobium atoms modifies TAM5. The resulting product was tested extensively to enable its use as an inorganic ion exchanger for the removal of radioisotopes, especially ¹³⁷Cs, from highly alkaline nuclear waste solutions. IONSIV^â IE-911 is a granular form of modified TAM5 (MTAM5) that is commercially available from UOP. During the last decade, we have acquired unique expertise in the synthesis of crystalline silicotitanates (CST). We have also synthesized silicotitanates from inorganic precursors in aqueous solutions without the use of any organics, such as alkoxides and chelates. Additionally, alternative alkali metal hydroxides can replace the use of sodium hydroxide. By using other alkali metal ions, different products were produced; yet, the basic crystal structure of TAM5 remained intact. The alkali metal ion used controls the channel diameter, as well as whether the channel network is parallel as in TAM5 or intercepting at right angles as in TAM4.

Silicotitanate Crystal Structure

Extensive search by J. Krumhansl of Sandia National Laboratory revealed that TAM5 is structurally related to a rare Russian mineral, Sitinakite, from Kibinskii alkaline massif. The idealized formula for the mineral is Na₂(H₂O)₂Ti₄O₅(OH)(SiO₄)₂K(H₂O)_1.7. Similarly the formula for TAM5 is Na₂(H₂O)₂Ti₄O₅(OH)(SiO₄)₂Na(H₂O)_1.7. The presence of potassium is the major distinction between Sitinakite and TAM5. The Sitinakite also contains about 5 wt% of niobium, as well as small amounts of other elements.

The crystal structure of Sitinakite was established by the single crystal XRD work by E. V. Sokolova et al. They pictured that the crystal consists of (a) a column built with clusters of four octahedral Ti atoms linked with oxygen, and (b) a string of alternating Si tetrahedra and Na octahedra. When four columns are attached with four strings, a straight channel of 3.5 A is created, where columns and strands are parallel to each other. Potassium ions and water molecules occupy the channels in Sitinakite while sodium ions along with water molecules occupy the channels in TAM5. A drawing of TAM5 emphasizing the channels is shown in Figure i. The molecular structure of the Ti₄ cluster bridges by silicate groups and the channel are illustrated in Figure ii.

During the hydrothermal synthesis, high alkalinity with sodium hydroxide is extremely critical for the formation of the TAM5 structure. In the XY projection, bridging the cubes with silicate groups forms the TAM5 structure. Within the cubes Ti and O occupy alternate corners. In the Z-axis direction silicate groups are linked by sodium ions forming a straight (–O-Si-O-Na-)_n strand. In TAM5, the cubes were linked by oxygen in the Z-axis direction. TAM5 has parallel channels (3.5 A dia.) perpendicular to the XY projection.

The alkali hydroxide plays a ‘template role’ during the synthesis of TAM5. Once the crystals are synthesized the alkali metal ions are easily ion exchanged with protons using a strong acid e.g. (2 M HCl) while keeping the crystal structure intact. H-form TAM5 can be loaded with other alkali metal ions, which are larger than a Na ion. Larger ions can only occupy the channels.

The performance of a CST as an ion exchanger is very difficult to predict. The TEM of CST indicates a major phase with high crystallinity that is structurally similar to Sitinakite, a second phase rich in silicate and a third phase of mostly of niobate. The powder XRD patterns of CST and elemental compositions of CST with good Cs loading capacity are similar, but the reverse is not true.

According to the agreement between Westinghouse Savannah River Company (Savannah River Technology Center, Aiken SC 29808) and Texas A&M University (College Station, TX 77843) the synthesis and the performance evaluations of crystalline silicotitanates (CST) were performed the during period of April 1 – September 30, 2000. IONSIV^â IE-911 is the UOP Trade name for the CST which was originally prepared at Texas A&M University in 1990 and modified later by adding Niobium to the synthesis recipe, for improved performance in highly alkaline Nuclear Waste Simulants. When the modified CST with niobium was prepared at Texas A&M University, we named it TAM5. Whenever we receive a sample from UOP, we use their trade name IONSIV^â IE-910 for the powder form and IONSIV^â IE-911 for the granular form. IONSIV^â IE-910 is very similar to TAM5 and IONSIV^â IE-911 contains a binder.

Using our original synthesis conditions we synthesized five batches of CST using a five-gallon reactor vessel and produced about a kilogram of CST in each attempt. TAM5-1, TAM5-2, TAM5-3, TAM5-4 and TAM5-5 are our ids for these CST. To evaluate these five batches of CST, we used powder X-ray diffraction patterns, elemental analysis by Inductively Coupled Plasma (ICP) spectrometry, and the cesium distribution coefficient (Kd) in our standard nuclear simulant. We have determined that TAM5-4 and TAM5-5 were good materials. A kilogram of TAM5-4 was shipped to Savannah River Technology Center, Aiken SC 29808 fulfilling the first part of our contract agreement in July, 2000. The other efforts at Texas A&M University were to compliment the works of researchers at Savannah River Technology Center. We had several conference-telephone calls redefining our two other tasks, testing CST in nuclear simulants and modeling the performance of CST in nuclear simulants.

Synthesis of the Crystalline Silicotitanates

A five-gallon, stirred reactor (1963 model from High Pressure Industries, Hatboro, Pennsylvania) was refitted with a new seal (grafoil sandwiched between two Teflon discs) that was developed at Texas A&M University. The reactor was tested extensively for leaks at 230°C under a pressure ranging up to 2000 psi prior to using it for the CST synthesis. Titanium isopropoxide and tetraethylorthosilicate from Aldrich, along with 50% NaOH from VWR scientific were used. Hydrous Niobium oxide (filter cake with 50% water) was acquired from a commercial source and used for the synthesis of TAM5-4.

For the synthesis of TAM5-5 hydrous niobium oxide was prepared from niobium chloride (Alfa Aesar). For the synthesis TAM5-4, first titanium isopropoxide (1080.2 g) and tetraethylorthosilicate (893.3 g) were mixed in a five-gallon plastic bucket. Then sodium hydroxide solution, which was prepared by mixing 747.84g of 50% NaOH sodium hydroxide with 6682.7ml water, was added to the plastic bucket and mixed well by stirring the contents with a plastic spatula to complete the gel formation. Next, 458.8 g of hydrous niobium oxide (50% H₂O) was added and mixed well. The mixture (gel, Ti:Si:Nb:NaOH = 1:1.13:0.4:2.6) was then transferred to a 5 gallon stainless steel reactor. Additionally, 4 liters of water were used to wash out the residues on the bucket into the reactor.

The reactor load weighed about 13,862.8g and the reactor was approximately 75% full. The reactor was capped and heated while stirring. It took about 2 hours for the inside reactor temperature to reach 230°C. Temperature controlled heating continued at 230°C for another 2 hours and then the heat was turned off. The reactor was cooled overnight and the solids in the reactor were transferred into two 1 liter Buchner funnels with glass fritted disc. The solid material in each funnel was first washed three times, with 2 liters of water per wash, and then three times with acetone, 1 liter of acetone was used for each washing. The solid products were subsequently air-dried. The solid had 25% volatiles (water and acetone). The solid was then scanned by a Scintag model XDS 2000 X-ray diffractometer. The X-ray powder diffraction pattern of the solid material identified it as a good quality crystalline sodium silicotitanate (TAM5). Cesium ion exchange experiments in our standard nuclear waste simulants (5.1 M NaNO₃, 0.6 M NaOH , 100 PPM Cs⁺) were also used to confirm quality of the material.

Evaluation of TAM5 using Cs distribution coefficient (K_d) experiments

The ion exchange capacity of a CST is initially used to determine the quality of the material from each batch of synthesis. For that purpose we used a standard nuclear waste simulant (5.1 M NaNO₃, 0.6 M NaOH, 100 ppm Cs). Generally 100 mg of silicotitanate and 10 ml of standard simulant was transferred to a plastic vial, closed, and then shaken continuously for 48 hours using Burrel Wrist-Action shaker. The vial was then centrifuged using a IEC EXD Centrifuge (Damion/IEC Division) at 60% power for 10 minutes. The clear supernatant liquid was separated from the vial. Then the Cs levels in the supernatant liquid and a control (basic salt solution without the silicotitanate shaken for 48 hours) were determined using a Varian AA 30 Atomic Absorption Spectrometer. The Cs distribution coefficient ( K_d) was calculated. The quality of TAM5 is determined using the Kd. Frequently K_d in neutral, as well as acidic salt solutions, were also determined using similar procedures. The amount of Ti, Si, and Nb in the supernatant liquid was determined by elemental analysis using an Inductively Coupled Argon Plasma Spectrometer (ICAP, Thermo Jarrel Ash Model Poly Scan 61E). The very low level of Ti, barely above the background, indicated that no TAM5 is present in the supernatant liquid. We have also tried membrane filters with 0.2-micron pores for separation of the CST from the supernantant liquid and observed that centrifuging is as good as filtration using membrane filters. Si and Nb were typically at higher levels. The source of Si and Nb could be from phases other than TAM5 crystals.

Ion exchange experiments for performance evaluation a comparative study of CST samples from four different batches

IONSIV^â IE-910 from UOP (UOP), TAM5 prepared by Ding Gu (DG141), TAM5 from the fourth batch (TAM5-4), and TAM5 from the fifth batch (TAM5-5) are the four CST samples used in the study. UOP and DG141 were 6 years old samples. Both TAM5-4 and TAM5-5 were synthesized in June 2000. The samples heated at 400°C overnight and weight loss was determined. Since TAM5-4 and TAM5-5 were air-dried samples, they experienced 25% weight loss. They lost water and acetone during heating. UOP and DG141 experienced 8% and 12% weight loss respectively. We prepared seven simulants by varying the composition of our standard simulants by adding carbonate, oxalate and hydrogen peroxide. These studies were used to evaluate the effect of carbonate, oxalate and hydrogen peroxide on Kd and leaching of Si, Ti and Nb from the CST during the equilibration experiments. The experiments were conducted in duplicates. Kd and leaching of Si, Ti and Nb were determined as previously discussed.

Analysis Methods

Cesium was analyzed by atomic absorption using a Varian AA 30 Atomic Absorption spectrometer. The nuclear waste simulants were diluted 1:3 by adding water so that the capillary and the burner would not become plugged. The instrument was calibrated with cesium standard solutions and a blank with similar matrix composition. The elemental analysis of Si, Ti and Nb in the nuclear waste simulant samples was performed using an inductively coupled argon plasma (ICAP) spectrometer (Thermo Jarrel Ash Polyscan 61E). The simulants were diluted 1:3 by adding water in order to enhance the air sol formation. A Scintag XDS 2000 X ray spectrometer was used to obtain the powder X ray diffraction (XRD) pattern of solid samples. The CST solids after equilibration with nuclear simulants were separated by centrifugation, washed with water and acetone, air–dried and scanned for the XRD patterns. One hundred-mg sample of CST was dissolved in 1 ml 50% HF, water was added to make 10 ml solution, and it was further diluted as needed used for the ICAP.

Determination of oxalate and hydrogen peroxide by titration with potassium permanganate

A stock solution of 0.1 M was prepared by dissolving 15.8 g in a one-liter volumetric flask and adding water to make one liter of solution. This solution was diluted hundred fold to obtain 0.001M potassium permanganate. Both solutions were used for the titration. Oxalic acid was used to standardize these solutions. The titration was performed by warming 1 ml of simulant and 10 ml 2M sulfuric acid in a 250 ml Ehrlenmayer flask on a hot plate. Potassium permanganate solution was added from a 50-ml burette. During the titration the temperature of the solution was maintained at 85°C and the end point was indicated by the pink color lasting longer than 30 seconds.

Synthesis of TAM5

We have attempted the synthesis of TAM5 five times and the first three failed. The first batch produced an amorphous product with XRD pattern without any peaks. The Kd of TAM5-1 was close to zero. The XRD pattern of TAM5-2 was fair, but the Kd was very low. TAM5-3 was amorphous with very low Kd. Since we could not get hydrous niobium oxide (38% water) from the original vender and due to the difficulty in locating a reliable new source, we prepared hydrous niobium oxide from NbCl₅ by ‘dissolving’ it in water and neutralizing it with sodium hydroxide. The precipitated niobium hydroxide was washed with water until chloride-free and the wet cake was used for the synthesis of TAM5-1, TAM5-2 and TAM5-5. Later we received a ‘filter cake’ of hydrous niobium oxide (50% water) from a source and used it for the synthesis of TAM5-3 and TAM5-5. Even though first three batches failed, the analysis of data from those batches helped to correct the recipe and to produce TAM5-4 and TAM5-5 with good Cs ion exchange performance.

Nuclear Waste Simulants

We have prepared seven simulants by adding carbonate, oxalate and hydrogen peroxide in to our standard nuclear simulant (5.1 M NaNO₃, 0.6 M NaOH, 100 ppm Cs). The compositions of seven simulants are listed in Simulants #1 -#8. Potassium permanganate titration was used to estimate oxalates and hydrogen peroxide in the simulant before and after ion exchange experiments. Initial hydrogen peroxide concentration in simulant #2 and #3 was 0.0025 M. Simulant #7 and Simulant #8 had hydrogen peroxide concentration of 1M and 0.1M. Hydrogen peroxide in the simulant decomposed continuously and the presence of CST did not have any significant effect on the rate of decomposition of H ₂O₂ to H₂O and O₂. In 9 days H ₂O₂ concentrations dropped from 1M to about 0.1M. Simulant #4 contained 0.02M oxalic acid, which did not dissolve completely. Oxalic acid was difficult to dissolve in the standard simulant how much of it dissolved was not estimated. It was just assumed that the standard simulant was saturated with oxalate. The ion exchange experiment was conducted by shaking 100mg of each CST with 10 ml of simulant for periods of 1-day, 2-days and 5-days. The exception is in the case of experiments with simulant #2 where 1-day, 4-days and 6-day periods were used due to equilibration on a weekend. Hydrogen peroxide at 0.0025 M levels in simulant#2 did not have any effect Kd. The Kd value dropped drastically in experiments with simulant #7 with 1 M H₂O₂. An intermediate effect was observed in experiment using simulant #8 with 0.1 M H₂O₂. The results of the Ion exchange experiments are listed in Tables 1a –8a.

Our first CST with TAM5, or sitinakite structure without niobium, had very high Kd values of over 100,000 in neutral and acidic nuclear waste simulant but it dropped below 100 in basic simulant. The CST was synthesized with the substitution of 25% of Ti atoms with Nb atoms and it showed improved the Kd values in basic simulants to above 1000 while lowering the acidic and neutral Kd value to about 10,000. We could not monitor the leaching of Si, Ti and Nb from the CST since at that time we had only an AA spectromer for the elemental analysis and the leaching issue was not raised. Now we have an ICAP and can monitor low levels of Si, Ti and Nb. All the simulants were analyzed for the leaching of CST during the equilibration experiments. Tables 1b –8b and Figures 1 –8 shows the analysis results. The CST has three different phases. The elemental analysis of the first six simulants does not show any significant amount of Ti loss indicating that the CST phase with sitinakite structure is very stable. A small amount of the silicate and niobate phases are leached as indicated by the presence of some Si and Nb species in the simulant. Almost all the CST broke down in simulant #7 with 1M H₂O₂ and about half the CST broke down in simulant #8, over a 5-day period. The effect of anions such as Carbonates, oxalates, nitrates on Kd values were studied and the results were illustrated in Figures 1-2, 1-4 and 2-5. The effect of replacing some of the nitrate with carbonate is shown in Figure 1-2. It appears that carbonate tends to increase the Kd values.

We have synthesized two batches crystalline silicotitantes (TAM5) with good ion exchange performance. The leaching of Si and Nb in the Cs ion exchange experiments using simulants #1 through #6 could be attributed to the silicate and niobate phases. The absence of Ti in the supernatent indicates that CST is not disintegrating. Hydrogen peroxide can attack and brake up the CST as indicated by a high level of Ti leached in ion exchange experiments using simulants #7 & #8, which contained 1M and 0.1 M H₂O₂ respectively. A low level of H₂O₂ (0.0025 M) in the simulants does not have any effect on leaching of Si, Ti and Nb or on the Kd values.

The leaching mechanism from CST must be studied. Using a CST low in silicate and niobate phases may minimize the leaching of CST. A thermal treatment of CST has a good potential of improving the product.

References

1. Anthony, R. G.; Philip, C. V.; and Dosch, R. G. Waste Manage. 1993, 13,503.
2. Method of Using Novel Silicotitanates. R. G. Anthony, R. G. Dosch and C. V. Philip. US Patent Number 6,100,378. August 29, 2000.
3. Krumhansl, J, Private communication from Sandia National Laboratory, 09/02/93
4. Sokolova, E. V.; Ratsvetaeva, R. K.; Andrianov, V. I.; Egorov-Tismenko, Yu. K; Men'shikov, Yu. P. Dolkl. Akad. Nauk sssr 1989. 307,114.

Table 1. Simulant #1 (Basic salt solution, with 0.16 M Na₂CO₃)

Table 1a. Average Cesium Distribution Coefficients (Kd) and standard
distributions for Simulant #1 (0.16M CO₃^2-)

Table 1a. Illustrates that after two days there is very little change in the distribution coefficients for DG141 and TAM5-#5. If one compares the distribution coefficients for TAM5-4, TAM 5-5, and DG 141 the differences are well within the standard deviations. These three batches were prepared using the same procedures. The UOP samples have lower values for the distribution coefficients, which is probably due to synthesis at lower temperatures than used for our samples.

Table 1b. Leaching of Si, Ti, and Nb from CST as a result of shaking with simulant #1 (0.16M CO₃^2-)

Table 1b illustrates the weight percent of Si, Ti and Nb that dissolves in the solution. This percent is based on gram of Si, Ti or Nb per gram of CST charged to the ion exchange experiment times 100.

Table 2. Simulant #2 (Basic salt solution with 0.0025 M H₂O₂, 0.16M Na₂CO₃)

Even though this simulant shows to have had hydrogen peroxide in the charge, as a result of titration with potassium permanganate we do not believe any significant amount of peroxide was present.

Table 2a. Average Cesium Distribution Coefficients (Kd) and standard distributions
for Simulant #2. (0.0025M H₂O₂, 0.16M Na₂CO₃)

Table 2b. Leaching of Si, Ti, and Nb from CST as a result of shaking with
simulant #2 (0.0025M H₂O₂, 0.16M Na₂CO₃)

Table 3. Simulant #3 (Basic salt solution with 0.48 M Na₂CO₃)

The effect of 0.48 M sodium carbonate in the presence of trace amounts of H₂O₂ on distribution coefficients and leaching was studied using Simulant #3.

Table 3a. Average Cesium Distribution Coefficients (Kd) and standard
distributions for Simulant #3 (0.0025M H₂O₂, 0.48 M Na₂CO₃₎

Table 3b. Leaching of Si, Ti, and Nb from CST as a result of shaking with
simulant #3 (0.0025M H₂O₂, 0.48 M Na₂CO₃₎

Table 4. Simulant #4 (Basic salt solution with 0.02 M oxalate, 0.16M Na₂CO₃)

note: Oxalic acid (H₂C₂O₄) did not completely dissolve. The effect of using Oxalic Acid on Distribution Coefficients and Leaching was studied using Simulant # 4.

Table 4a. Average Cesium Distribution Coefficients (Kd) and standard
distributions for Simulant #4 (0.02M H₂C₂O₄, 0.16M Na₂CO₃)

Table 4b. Leaching of Si, Ti, and Nb from CST as a result of shaking
with simulant#4 (0.02M H₂C₂O₄, 0.16M Na₂CO₃)

Table 5. Simulant #5 (Basic salt solution with 0.0025M H₂O₂)

Attempted for 1 M H₂O₂

Table 5a. Average Cesium Distribution Coefficient (Kd) and standard
distributions for Simulant #5 (0.0025M H₂O₂)

Table 5b. Leaching of Si, Ti and Nb from CST as a result of
shaking with Simulant #5 (0.0025M H₂O₂)

Table 6. Simulant #6 (Basic salt solution with 0.0016M oxalate)

Table 6a. Average Cesium Distribution Coefficients (Kd) and standard
distributions for Simulant #6 (0.0016 oxalate)

Table 6b. Leaching of Si, Ti and Nb from CST as a result of shaking
with Simulant #6 (0.0016 oxalate)

Table 7. Simulant #7 (Basic salt solution with 1M H₂O₂)

Table 7b. Leaching of Si, Ti and Nb from CST as a result of
shaking with Simulant #7 (1M H₂O₂)

Table 8. Simulant #8 (Basic salt solution with 0.1M H₂O₂)

Table 8b. Leaching of Si, Ti and Nb from CST as a result of
shaking with Simulant #8 (0.1M H₂O₂)

Figure i. Structure of sodium silicate (TAM5) showing the parallel channels. In the cube,
corners with circles represent Ti and corners with out circles represent O.
Si is between a pair of cubes and Na is between two a
pair of Si, which are not shown.

Figure ii. A section of TAM5 structure showing the Ti₄ clusters linked with tetrahedral silicates.
The hole in center represents the cross-section of the channel

Figure 1. Leaching of Si, Ti, and Nb from CST as a result of shaking with simulant #1 (Basic salt solution, with 0.16M Na₂CO₃)

Figure 2. Leaching of Si, Ti, and Nb from CST as a result of shaking with simulant #2
(Basic salt solution with 0.0025 M H₂O₂, 0.16M Na₂CO₃)

Figure 3. Leaching of Si, Ti, and Nb from CST as a result of shaking with simulant #3
(Basic salt solution with 0.0025 M H₂O₂, 0.48M Na₂CO₃)

Figure 4. Leaching of Si, Ti, and Nb from CST as a result of shaking with simulant #4
(Basic salt solution with 0.02 M oxalate, 0.16M Na₂CO₃)

Figure 5. Leaching of Si, Ti, and Nb from CST as a result of shaking with simulant #5
(Basic salt solution with 0.0025M H₂0₂)

Figure 6. Leaching of Si, Ti, and Nb from CST as a result of shaking with simulant #6
(Basic salt solution with 0.0016 oxalate)

Figure 7. Leaching of Si, Ti, and Nb from CST as a result of shaking with simulant #7
(Basic salt solution with 1M H₂O₂)

Figure 8. Leaching of Si, Ti, and Nb from CST as a result of shaking with simulant #8
(Basic salt solution with 0.1M H₂O₂)

Figure 1-3. CO3^2- ion or NO₃^- effect (same Na⁺ ion), See Simulant #1 and #3 for composition,
Cs Kd Values are listed in Tables 1a and 3a.

Figure 1-4. C₂O₂^2- ion ^- effect (same Na⁺ ion), See Simulant #1 and #4 for composition,
Cs Kd Values are listed in Tables 1a and 4a.

Figure 2-5. Na⁺ ion (5.7M vs 5.4M) or CO₃^2- ion (0.16M vs 0.0M) effect, See Simulant #2 and #5 for
composition, Cs Kd Values are listed in Tables 2a and 5a.