- Email: [email protected]
Preparation of radioactively labeled monodisperse silica spheres of colloidal size
Preparation of Radioactively Labeled Monodisperse Silica Spheres of Colloidal S~e H E R O ' A N N F L A C H S B A R T A~D W E R N E R S T O B E R Department of Radiation Biology and Biophysics, School of Medicine and Dentistry, University of Rochester, Rochester, New York 14620
Received March 19, 1969 INTRODUCTION
Laboratory investigations concerning physical, chemical, or physiological properties of suspended particulate matter are greatly facilitated if uniform model substances with well-defined physical properties are available. A recent study (1) described a chemical process for producing silica spheres of predetermined uniform size which are suitable for the generation of monodisperse aerosols. The new technique of growing spherical silica particles in alcoholic suspem sion also lends itself to incorporating radioactive tracers into the bulk of the spheres during their growth. Since such an irreversible label would make the silica particles a versatile tool for many fundamental investigations, a systematic study with a number of radioactive isotopes was conducted. EXPERIMENTAL
METHODS
A simple experimental procedure for attempting the incorporation of radioactive isotopes into growing silica particles was devised. Approximately 1 vc or less of dissolved radioactive material amounting in most cases to not more than a few micrograms of foreign matter was added to 50 ml of various alcoholic solutions of esters of silicic acid which were prepared as described in the preceding study (1). The final yield of silica was approximately 500 rag. In some initial tests, the influence of the added foreign matter on the nucleation and condensation characteristics of the alcoholic solutions was investigated. No significant difference in final particle size was detected between this series and the controls.
A particular recipe for the particle production was then selected for the majority of the subsequent tests. A solution of 10 ml of saturated ammonium hydroxide in 40 ml of pure ethanol was prepared, and 2 ml of tetrapropyl ester of silicic acid and 1 ml of the solution of the radioactive tracer were added. Table I lists the various tracer solutions used in the tests. The solutions turned turbid a few minutes after mixing and the growth of the silica particles continued over a period of about 2 hours, after which time the silicic acid in solution was depleted. Electron micrographs indicated a final particle size close to 8 × 10-~ cm. All samples prepared in this way were filtered (!VIillipore membrane filter GS 0.22 t~), and the activities of the filtrate and the filter cake were measured separately in a well-type crystal counter. The filter cakes were then weighed and resuspended for about 24 hours under various chemical conditions in 100 ml of alcoholic or aqueous solutions. In the latter case, a wide range of p H values was covered b y adding appropriate amounts of hydrochloric acid and ammonium hydroxide, respectively, and different agents in solution were applied in order to remove the exchangeable fraction of the radioactive material from the particles. At the end of the test, the particles were again separated from the liquid phase by filtering the suspension through a membrane filter. The filter cake was weighed and the distribution of activity between liquid and solid phase was measured.
Jou~'nal of Colloid and Interface Science, Vol. 30, No. 4, August 1969
568
P R E P A R A T I O N OF R A D I O A C T I V E L Y L A B E L E D S I L I C A S P H E R E S
569
with the silica particles although noticeable amounts of silica dissolved in the form of silieie acid at increasing pH values. At low Concen- ActivIsotope Compound Solvent tration ity ptI values, however, particularly after add~g/ml) (l~c/ml) ing 1 gm of stable manganese chloride and/or tartaric acid (20 ml of a 20 % acid solution), ~a4Cs CsC1 0.5 N I-IC1 1 1.0 a gradual removal of the activity from the 65Zn ZnC12 1.0 N tICt 3.0 0.3 particles was observed. After 36 hours of re54Mn MnC12 1.0 N HCI 2.5 1.0 124Sb SbC13 6.0 N tIC1 4.5 0.25 suspension at pH = 3.1, the particles lost 59Fe FeCla 1.0 N HC1 2.6 1.0 about 50 % of their radioactivity. However, 5~Cr CrCla 1.0 N HC1 0.025 1.0 at pH = 0.1 only 11% of the ~4manganese raCe CeCla 1.0 N HCI 0.150 1.0 was not exchanged by stable manganese 7Be BeC12 1.0 N HC1 0.030 1.0 chloride. 5sCo CoCI2 1.0 N I:ICI 0.010 1.0 4. 124Antimony. Resuspending the silica 5sCo CoC12 1.0 N HC1 300 1.0 particles in various solutions showed little effect on the activity associated with the particles. No changes were observed in RESULTS alcoholic suspension. There was a small reIsotopes of nine metals of different chemilease of activity in aqueous suspension. cal properties were used for labeling the The highest releases were obtained at silica particles while they were growing. values of pH > 9 and amounted to about Throughout the growth experiments, the 6 %. No additional release was found after activities of the filtrates were essentially at adding stable antimony chloride or tartaric background level and all activity of the acid to the test suspension. Surprisingly, no tracers was associated with the particles on significant dissolution of silica particles octhe flter. There were only small traces of surfed up to pH = 9.7. contamination inside the filter. On resus5. ~gIron. The silica particles were repension the following results were obtained. suspended in a variety of alcoholic and 1. ~a4Cesium. All radioactivity could be aqueous solutions, but the results showed removed from the resuspended silica paralmost no changes of the activity associated ticles by adding 1 gm of stable cesium with the particles. The optimum releases chloride to the suspension. The exchange occurred at high pH values and did not took place within 2 hours at the most. exceed 5%, although 316 mg of silica Filtering the suspension after this time left representing 70% of the original particle very little activity on the filter cakes. 2. 65Zinc. The resuspended particles did mass dissolved at pH = 12. In this case, the released activity was apparently renot release any activity in alcoholic suspenadsorbed on the surface of the remaining sion, but a gradual release was observed in undissolved silica particles. aqueous suspensions. The process was rather 6. 51Chromium. The resuspended silica slow for alkaline suspensions but increased particles released very small, if any, amounts rapidly with decreasing pH values.At pH = 2, the activity of the particle was almost of activity within a period of 36 hours. A instantly released. In general, adding zinc maximum of 3 % was recorded at pH = 1.2. nitrate increased significantly the rate of No significant change was effected by varying the pH values or adding dissolved chrome activity removal from the particles at high pH values. At any given pH value the alton or tartaric acid to the test suspensions. removal could be increased further by adding At high values, there was a significant 20 ml of 0.1 M ethylenediamine tetraacetic dissolution of silica (ca. 60 rag. at pH = 10.2) but no corresponding release of activity acid (EDTA) to the test suspensions. into solution occurred. 3. 54Manganese. The removal of the activ7. 14~Cerium. No release of radioactivity ity from the resuspended particles could not be accomplished in alcoholic suspensions or was observed when the silica particles were in aqueous suspensions at pH > 7. In the resuspended in alcohol. In aqueous suslatter case the activity remained associated pension the release of activity never exTABLE I
])ATA OF TRACER SOLUTIONS
Journal of Colloid and Interface Science, Vol. 30, No. 4, August 1969
570
FLACttSBART AND STOBER
ceeded a few per cent (5% at p H = 8.8). Adding cerium nitrate or E D T A to the suspension did not influence the distribution of activity. Even when a large fraction of the silica particles was dissolved (300 rag at p H = 10.6), there was no significant increase of the activity in the filtrate. The activity released with the dissolved silica was immediately adsorbed on the surface of the remaining particles. I t could be subsequently washed off at low p H values. 8. 7Beryllium. The labeled silica particles were resuspended in alcohol and in aqueous solutions of stable beryllium chloride in combination with tartaric acid at various p H values. No activity was found in the liquid phase of the alcoholic suspension. Over almost the entire p H range the release of activity into the aqueous solution was very small. I t was not influenced by the salts added to the solutions. At p H = 2, the release amounted to 3 %. A loss of 9 % of the activity was recorded at p H = 10. In this case, however, 70 mg of silica representing 14% of the weight of the silica particles were dissolved during the test. The test solution contained tartaric acid. Thus, part of the released activity was prevented from being readsorbed on the remaining silica particles and appeared in the filtrate. 9. 5sCobalt. Two different tracer solutions were used for these tests, a dilute solution of cobalt chloride of high specific activity and a more concentrated cobalt chloride solution with a very low specific activity. The series with the cobalt tracer of high specific activity showed that losses of the activity of the silica particles were almost negligible in alcoholic and aqueous suspension. In the latter case the release was tested over a range of p H values between 1.2 and 11.0. Parallel test series were made in which 20 ml of 0.1 M E D T A a n d / o r 1 gm of stable cobalt chloride were added to the aqueous solutions. A maximum release of 3 % of the activity was encountered in the presence of stable cobalt chloride at p H = 11. The salt had a remarkable influence on the dissolution of silica particles at high p H values. Whereas in its absence about 135 mg of silica were dissolved Journal of Colloid and Interface Science, Vol. 30, ]go. 4, August 1969
at p H = 11, the dissolution rate was significantly reduced in its presence. In this case, only a few milligrams of silica were dissolved over a comparable period of time. The varying amounts of dissolved silica, however, were not significantly reflected by the release of radioactivity. Invariably, only small amounts of radioactive cobalt were found in the filtrate. This indicates that the ammine cobalt complex or the complex with E D T A forming on release of the radioactive cobalt from the dissolving silica is immediately readsorbed on the remaining silica surface. Subsequent successive removal of the adsorbed material at low p H values confirmed this assumption. Resuspending the samples obtained with 300 ~g of cobalt of low specific activity in aqueous solution at low p H values gave different results. At p H = 1.3, for instance, there was a release of 23 % of the activity from the particles which could be increased to 74% b y adding 1 gm of stable cobalt chloride to the solution. At high p H values, the same pattern prevailed as in the tests with small amounts of cobalt of high specific activity. Although 115 mg of silica were dissolved at p H = 11 in the absence of any other salt, no significant amount of activity was released into the filtrate. The ammine cobalt complex was adsorbed on the silica surface. Subsequent washing of the remaining particles at p H = 1,2 greatly reduced the activity associated with the particles. DISCUSSION The experimental results indicate that the radioactive tracers are attached to the silica particles in different ways depending upon the chemical nature of the isotope used. Cesium is representative of all elements which, irrespective of the p H value, remain in the dissociated ionic state. Although adsorbed to the surface of the growing silica particles the cesium ions are never incorporated into the bulk of the silica. They remain mobile on the silica surface, while the silicic acid condenses on the particle and makes it grow. Thus, a simple exchange with an excess of stable cesium ions in the liquid phase will remove all radioactive cesium from the silica.
PREPARATION OF RADIOACTIVELY LABELED SILICA SPHERES A similar behavior was observed with radioactive zinc. The silica particles are grown under high pH values which cause the formation of a zinc ammine complex ion. This ion is adsorbed on the particle surface but it is not incorporated into the framework of the SiO 4tetrahedrons during the growth of the silica particles. Aging of the adsorptive layer during filtration and drying of the filter cake probably transforms the zinc complex to zinc oxide. This would explain the slow exchange rate with zinc nitrate and EDTA at high pH values. The results of pit = 2 indicate quite clearly that zinc is merely adsorbed to the silica particles. The results obtained with manganese don't quite permit such a gross statement. In solution with the growing silica particles the ammonia causes the formation of dissociated manganese hydroxide. The cation of this compound is then adsorbed to the silica surface, where it escapes incorporation as long as it remains in the mobile ionic state. A small fraction of the adsorbate, however, will probably condense to manganese sesquioxide; at high concentrations this would be likely to occur even in the liquid phase. This nonionic oxide fraction becomes immobilized and condenses to the silica surface. Then, the continuing growth finally provides the incorporation into the bulk of the particles, while the major fraction of the manganese remains as hydroxide on the surface. The final surface layer probably turns into oxide during the drying of the filter cake. Thus, after resuspension, complete removal of the surface layer within 36 hours is facilitated only at very low pH values. The small fraction of incorporated oxide, of course, is not affected unless the particle is dissolved. All other radioactive elements utilized in this study at the same level of up to a few micrograms indicated a more uniform mechanism of tracer attachment. In all these tests with small amounts of antimony, beryllium, cerium, chromium , cobalt, and iron, the results suggested that the tracer was incorporated into the bulk of the silica particles. In view of the general chemistry of these metals and the results obtained with zinc
571
and manganese we may expect that resuspending the particles at low pH values would be the most efficient way to remove or exchange any tracer material adsorbed on the silica surface. But neither at low nor at high pH values did the six tracers show any significant release of radioactivity from the particles. Substantial losses could always be related to the dissolution of silica in suspension. Since the silica particles are growing in solutions of ammonia at high pH value, all the six metal ions under consideration form preeipitable hydroxide compounds in the presence of the nucleating silicic acid. Their extremely low concentration is likely to prevent a spontaneous precipitation of the hydroxide, but the silica surfaces undoubtedly adsorb the foreign hydroxide molecules. Once adsorbed, the hydroxides apparently participate in the condensation process of the silicic acid and are linked up with it in the oxide form. The limited solubility for most of the last six tracers did not permit experiments with increased concentrations of foreign matter. However, precipitates of cobalt hydroxide redissolve at high ammonia concentrations. Thus, a test series with an increased amount of cobalt was feasible. The results of the experiments with 300 ~g of cobalt per test show clearly that this concentration, which is of the order of 0.1% of the silica, is not low enough to warrant a complete incorporation. Table II summarizes the results of the various test series. It indicates that antimony, beryllium, cerium, chromium, cobalt, and iron will be incorporated by the growing silica particle if the foreign solute is not in excess of about 1-10 ppm with regard to the final particle weight. Since the silica particles were rather uniform in size as shown in Fig. 1, it could not be decided whether or not the incorporated tracers were homogeneously distributed over the volume of a particle. It appears that the foreign oxides did not serve as condensation nuclei for the particles, because, in this case, they would have influenced the growth characteristics of the silica; this they did not according to the control experiments. Journal of Colloid and Interface ~cience, Vol. 30, No. 4, August 1969
572
F L A C H S B A R T A N D ST(~BER TABLE II ATTACHMENT OF D I F F E R E N T I~ADIOACTIYE 1SOTOPES TO SILICA PARTICLES CONDENSED IN THE PRESENCE OF THE TRACER
Metal ion
Attachment
Tracer concentration (ppm)
n4Cs
~ 1 mg/500 g SiO~ complete
2 X 10-6
65Zn
3.0 rag/500 gm SiO_~ complete
6 X 10-6
54Mn
2.5 mg/500 gm SiO2 complete
5 X 10.6
i24Sb ~gFe 51Cr
4.5 mg/500 gm SiO: complete 2.6 rag/500 gm SiO2 complete 0.5 ~g/500 gm SiO2 complete
9 X 10-6 5 X 10-6 1 X 10-9
~41Ce
150 t~g/500 gm SiO2 complete
3 X 10-7
7Be
30 ~g/500 gin SiO2 complete
6 X 10-8
58Co
10 t~g/500 gm SiO2 complete
2 X 10-s
58Co
300 mg/500 gm SiO2 70% t a k e - u p
6 X 10-~
Maximumdesorption, exchange, or chelation Complete exchange by adding 1 gm of stable CsC1 p i t = 2; E D T A (20%) complete removal p H = 0.1; M n C12 release of 89% p H > 9; release of 6% p~I > 9; release of 5% p H = 1.2; K Cr (SO4)2 release of 3% p H = 8.8; Ce (NO3)3 release of 5% p H = 10; t a r t a r i c acid release of 9% p H = 11; Co C12 release of 3% p H = 1.3; Co C12 release of 74%
Incorporation No No Partial Yes Yes Yes Yes Yes Yes Partial
FIa. 1. E l e c t r o n micrograph of sample of silica particles grown to final size (mean d i a m e t e r 8 X 10-~ era). Journal of Colloid and Interface Science, Vol. 30, No. 4, August 1969
PREPARATION OF RADIOACTIVELY LABELED SILICA SPHERES
The possibility that the tracer was depleted before the particles had grown to their final size cannot be excluded, however. An exact proportionality between the relative amount of silica dissolved at high pH values and the relative activity subsequently recovered from the surface at low pH values could not be established with sufficient certainty. If the activity is completely incorporated in the early stage of the particle growth, i.e., if the foreign oxides are primarily distributed around the core of the particle, then the activity of a sample of particles of a polydispersely grown sol would be representative of the number of particles rather than their mass. The various results obtained for the amount of silica dissolved at high pH values is in keeping with findings reported in the literature. It has been shown (2, 3) that the rate of hydrolysis of silica in aqueous suspension is substantially influenced by the presence of various salts and organic compounds. The concentration of saturated silicic acid increases rapidly (4) from 0.1 gin/liter at pH _~ 8 to 5 gin/liter at pH = 11. However, it may take a few days to several weeks depending upon the additional solutes to establish this equilibrium. Correspondingly, the hydrolysis of resuspended silica particles at high p i t value may appear suppressed, as in the presence of cobalt salts, or moderate, as in the presence
573
of chrom alum, or increased, as in the presence of cerium nitrate. SUMMARY A recently developed technique of growing monodisperse silica spheres of predetermined size in alcoholic solutions facilitates the incorporation of certain radioactive tracers. Spheres growing to a final diameter of 8 X 10-5 cm were used in a study with radioactive isotopes including 7Be, 51Cr, 5~Mn, ~sCo, 59Fe, 65Zn, 124Sb, 134Cs, and 141Ce. The results reveal that those metals which form oxides of low solubility at pH >_ 7 are easily incorporated into the bulk of the silica spheres when present during the growth of the particles in an amount not exceeding 1-10 ppm with respect to the total mass of silica in the system. ACKNOWLEDGMENT This paper is based on work performed under contract with the United States Atomic Energy Commission at the University of Rochester Atomic Energy Project and has been assigned Publication No. UR-49-1069. REFERENCES 1. STOBER,W., FINK, A., AND BOHN, E., J. Colloid and Interface Sei. 26, 62 (1968). 2. FRIEDB]~nG,K., Beitr. Silikose-Forsch., Sonderband, p. 49 (1955). 3. STOBE~,W., Kolloid-Z. 147,131 (1956). 4. ALEXANDER,G. B., HESTON, W. M., ANn ILER, R. K., J. Phys. Chem. 58,453 (1954).
Journal of Colla4dand Interface Science, VoI.30~No. 4, August1969
Received March 19, 1969 INTRODUCTION
Laboratory investigations concerning physical, chemical, or physiological properties of suspended particulate matter are greatly facilitated if uniform model substances with well-defined physical properties are available. A recent study (1) described a chemical process for producing silica spheres of predetermined uniform size which are suitable for the generation of monodisperse aerosols. The new technique of growing spherical silica particles in alcoholic suspem sion also lends itself to incorporating radioactive tracers into the bulk of the spheres during their growth. Since such an irreversible label would make the silica particles a versatile tool for many fundamental investigations, a systematic study with a number of radioactive isotopes was conducted. EXPERIMENTAL
METHODS
A simple experimental procedure for attempting the incorporation of radioactive isotopes into growing silica particles was devised. Approximately 1 vc or less of dissolved radioactive material amounting in most cases to not more than a few micrograms of foreign matter was added to 50 ml of various alcoholic solutions of esters of silicic acid which were prepared as described in the preceding study (1). The final yield of silica was approximately 500 rag. In some initial tests, the influence of the added foreign matter on the nucleation and condensation characteristics of the alcoholic solutions was investigated. No significant difference in final particle size was detected between this series and the controls.
A particular recipe for the particle production was then selected for the majority of the subsequent tests. A solution of 10 ml of saturated ammonium hydroxide in 40 ml of pure ethanol was prepared, and 2 ml of tetrapropyl ester of silicic acid and 1 ml of the solution of the radioactive tracer were added. Table I lists the various tracer solutions used in the tests. The solutions turned turbid a few minutes after mixing and the growth of the silica particles continued over a period of about 2 hours, after which time the silicic acid in solution was depleted. Electron micrographs indicated a final particle size close to 8 × 10-~ cm. All samples prepared in this way were filtered (!VIillipore membrane filter GS 0.22 t~), and the activities of the filtrate and the filter cake were measured separately in a well-type crystal counter. The filter cakes were then weighed and resuspended for about 24 hours under various chemical conditions in 100 ml of alcoholic or aqueous solutions. In the latter case, a wide range of p H values was covered b y adding appropriate amounts of hydrochloric acid and ammonium hydroxide, respectively, and different agents in solution were applied in order to remove the exchangeable fraction of the radioactive material from the particles. At the end of the test, the particles were again separated from the liquid phase by filtering the suspension through a membrane filter. The filter cake was weighed and the distribution of activity between liquid and solid phase was measured.
Jou~'nal of Colloid and Interface Science, Vol. 30, No. 4, August 1969
568
P R E P A R A T I O N OF R A D I O A C T I V E L Y L A B E L E D S I L I C A S P H E R E S
569
with the silica particles although noticeable amounts of silica dissolved in the form of silieie acid at increasing pH values. At low Concen- ActivIsotope Compound Solvent tration ity ptI values, however, particularly after add~g/ml) (l~c/ml) ing 1 gm of stable manganese chloride and/or tartaric acid (20 ml of a 20 % acid solution), ~a4Cs CsC1 0.5 N I-IC1 1 1.0 a gradual removal of the activity from the 65Zn ZnC12 1.0 N tICt 3.0 0.3 particles was observed. After 36 hours of re54Mn MnC12 1.0 N HCI 2.5 1.0 124Sb SbC13 6.0 N tIC1 4.5 0.25 suspension at pH = 3.1, the particles lost 59Fe FeCla 1.0 N HC1 2.6 1.0 about 50 % of their radioactivity. However, 5~Cr CrCla 1.0 N HC1 0.025 1.0 at pH = 0.1 only 11% of the ~4manganese raCe CeCla 1.0 N HCI 0.150 1.0 was not exchanged by stable manganese 7Be BeC12 1.0 N HC1 0.030 1.0 chloride. 5sCo CoCI2 1.0 N I:ICI 0.010 1.0 4. 124Antimony. Resuspending the silica 5sCo CoC12 1.0 N HC1 300 1.0 particles in various solutions showed little effect on the activity associated with the particles. No changes were observed in RESULTS alcoholic suspension. There was a small reIsotopes of nine metals of different chemilease of activity in aqueous suspension. cal properties were used for labeling the The highest releases were obtained at silica particles while they were growing. values of pH > 9 and amounted to about Throughout the growth experiments, the 6 %. No additional release was found after activities of the filtrates were essentially at adding stable antimony chloride or tartaric background level and all activity of the acid to the test suspension. Surprisingly, no tracers was associated with the particles on significant dissolution of silica particles octhe flter. There were only small traces of surfed up to pH = 9.7. contamination inside the filter. On resus5. ~gIron. The silica particles were repension the following results were obtained. suspended in a variety of alcoholic and 1. ~a4Cesium. All radioactivity could be aqueous solutions, but the results showed removed from the resuspended silica paralmost no changes of the activity associated ticles by adding 1 gm of stable cesium with the particles. The optimum releases chloride to the suspension. The exchange occurred at high pH values and did not took place within 2 hours at the most. exceed 5%, although 316 mg of silica Filtering the suspension after this time left representing 70% of the original particle very little activity on the filter cakes. 2. 65Zinc. The resuspended particles did mass dissolved at pH = 12. In this case, the released activity was apparently renot release any activity in alcoholic suspenadsorbed on the surface of the remaining sion, but a gradual release was observed in undissolved silica particles. aqueous suspensions. The process was rather 6. 51Chromium. The resuspended silica slow for alkaline suspensions but increased particles released very small, if any, amounts rapidly with decreasing pH values.At pH = 2, the activity of the particle was almost of activity within a period of 36 hours. A instantly released. In general, adding zinc maximum of 3 % was recorded at pH = 1.2. nitrate increased significantly the rate of No significant change was effected by varying the pH values or adding dissolved chrome activity removal from the particles at high pH values. At any given pH value the alton or tartaric acid to the test suspensions. removal could be increased further by adding At high values, there was a significant 20 ml of 0.1 M ethylenediamine tetraacetic dissolution of silica (ca. 60 rag. at pH = 10.2) but no corresponding release of activity acid (EDTA) to the test suspensions. into solution occurred. 3. 54Manganese. The removal of the activ7. 14~Cerium. No release of radioactivity ity from the resuspended particles could not be accomplished in alcoholic suspensions or was observed when the silica particles were in aqueous suspensions at pH > 7. In the resuspended in alcohol. In aqueous suslatter case the activity remained associated pension the release of activity never exTABLE I
])ATA OF TRACER SOLUTIONS
Journal of Colloid and Interface Science, Vol. 30, No. 4, August 1969
570
FLACttSBART AND STOBER
ceeded a few per cent (5% at p H = 8.8). Adding cerium nitrate or E D T A to the suspension did not influence the distribution of activity. Even when a large fraction of the silica particles was dissolved (300 rag at p H = 10.6), there was no significant increase of the activity in the filtrate. The activity released with the dissolved silica was immediately adsorbed on the surface of the remaining particles. I t could be subsequently washed off at low p H values. 8. 7Beryllium. The labeled silica particles were resuspended in alcohol and in aqueous solutions of stable beryllium chloride in combination with tartaric acid at various p H values. No activity was found in the liquid phase of the alcoholic suspension. Over almost the entire p H range the release of activity into the aqueous solution was very small. I t was not influenced by the salts added to the solutions. At p H = 2, the release amounted to 3 %. A loss of 9 % of the activity was recorded at p H = 10. In this case, however, 70 mg of silica representing 14% of the weight of the silica particles were dissolved during the test. The test solution contained tartaric acid. Thus, part of the released activity was prevented from being readsorbed on the remaining silica particles and appeared in the filtrate. 9. 5sCobalt. Two different tracer solutions were used for these tests, a dilute solution of cobalt chloride of high specific activity and a more concentrated cobalt chloride solution with a very low specific activity. The series with the cobalt tracer of high specific activity showed that losses of the activity of the silica particles were almost negligible in alcoholic and aqueous suspension. In the latter case the release was tested over a range of p H values between 1.2 and 11.0. Parallel test series were made in which 20 ml of 0.1 M E D T A a n d / o r 1 gm of stable cobalt chloride were added to the aqueous solutions. A maximum release of 3 % of the activity was encountered in the presence of stable cobalt chloride at p H = 11. The salt had a remarkable influence on the dissolution of silica particles at high p H values. Whereas in its absence about 135 mg of silica were dissolved Journal of Colloid and Interface Science, Vol. 30, ]go. 4, August 1969
at p H = 11, the dissolution rate was significantly reduced in its presence. In this case, only a few milligrams of silica were dissolved over a comparable period of time. The varying amounts of dissolved silica, however, were not significantly reflected by the release of radioactivity. Invariably, only small amounts of radioactive cobalt were found in the filtrate. This indicates that the ammine cobalt complex or the complex with E D T A forming on release of the radioactive cobalt from the dissolving silica is immediately readsorbed on the remaining silica surface. Subsequent successive removal of the adsorbed material at low p H values confirmed this assumption. Resuspending the samples obtained with 300 ~g of cobalt of low specific activity in aqueous solution at low p H values gave different results. At p H = 1.3, for instance, there was a release of 23 % of the activity from the particles which could be increased to 74% b y adding 1 gm of stable cobalt chloride to the solution. At high p H values, the same pattern prevailed as in the tests with small amounts of cobalt of high specific activity. Although 115 mg of silica were dissolved at p H = 11 in the absence of any other salt, no significant amount of activity was released into the filtrate. The ammine cobalt complex was adsorbed on the silica surface. Subsequent washing of the remaining particles at p H = 1,2 greatly reduced the activity associated with the particles. DISCUSSION The experimental results indicate that the radioactive tracers are attached to the silica particles in different ways depending upon the chemical nature of the isotope used. Cesium is representative of all elements which, irrespective of the p H value, remain in the dissociated ionic state. Although adsorbed to the surface of the growing silica particles the cesium ions are never incorporated into the bulk of the silica. They remain mobile on the silica surface, while the silicic acid condenses on the particle and makes it grow. Thus, a simple exchange with an excess of stable cesium ions in the liquid phase will remove all radioactive cesium from the silica.
PREPARATION OF RADIOACTIVELY LABELED SILICA SPHERES A similar behavior was observed with radioactive zinc. The silica particles are grown under high pH values which cause the formation of a zinc ammine complex ion. This ion is adsorbed on the particle surface but it is not incorporated into the framework of the SiO 4tetrahedrons during the growth of the silica particles. Aging of the adsorptive layer during filtration and drying of the filter cake probably transforms the zinc complex to zinc oxide. This would explain the slow exchange rate with zinc nitrate and EDTA at high pH values. The results of pit = 2 indicate quite clearly that zinc is merely adsorbed to the silica particles. The results obtained with manganese don't quite permit such a gross statement. In solution with the growing silica particles the ammonia causes the formation of dissociated manganese hydroxide. The cation of this compound is then adsorbed to the silica surface, where it escapes incorporation as long as it remains in the mobile ionic state. A small fraction of the adsorbate, however, will probably condense to manganese sesquioxide; at high concentrations this would be likely to occur even in the liquid phase. This nonionic oxide fraction becomes immobilized and condenses to the silica surface. Then, the continuing growth finally provides the incorporation into the bulk of the particles, while the major fraction of the manganese remains as hydroxide on the surface. The final surface layer probably turns into oxide during the drying of the filter cake. Thus, after resuspension, complete removal of the surface layer within 36 hours is facilitated only at very low pH values. The small fraction of incorporated oxide, of course, is not affected unless the particle is dissolved. All other radioactive elements utilized in this study at the same level of up to a few micrograms indicated a more uniform mechanism of tracer attachment. In all these tests with small amounts of antimony, beryllium, cerium, chromium , cobalt, and iron, the results suggested that the tracer was incorporated into the bulk of the silica particles. In view of the general chemistry of these metals and the results obtained with zinc
571
and manganese we may expect that resuspending the particles at low pH values would be the most efficient way to remove or exchange any tracer material adsorbed on the silica surface. But neither at low nor at high pH values did the six tracers show any significant release of radioactivity from the particles. Substantial losses could always be related to the dissolution of silica in suspension. Since the silica particles are growing in solutions of ammonia at high pH value, all the six metal ions under consideration form preeipitable hydroxide compounds in the presence of the nucleating silicic acid. Their extremely low concentration is likely to prevent a spontaneous precipitation of the hydroxide, but the silica surfaces undoubtedly adsorb the foreign hydroxide molecules. Once adsorbed, the hydroxides apparently participate in the condensation process of the silicic acid and are linked up with it in the oxide form. The limited solubility for most of the last six tracers did not permit experiments with increased concentrations of foreign matter. However, precipitates of cobalt hydroxide redissolve at high ammonia concentrations. Thus, a test series with an increased amount of cobalt was feasible. The results of the experiments with 300 ~g of cobalt per test show clearly that this concentration, which is of the order of 0.1% of the silica, is not low enough to warrant a complete incorporation. Table II summarizes the results of the various test series. It indicates that antimony, beryllium, cerium, chromium, cobalt, and iron will be incorporated by the growing silica particle if the foreign solute is not in excess of about 1-10 ppm with regard to the final particle weight. Since the silica particles were rather uniform in size as shown in Fig. 1, it could not be decided whether or not the incorporated tracers were homogeneously distributed over the volume of a particle. It appears that the foreign oxides did not serve as condensation nuclei for the particles, because, in this case, they would have influenced the growth characteristics of the silica; this they did not according to the control experiments. Journal of Colloid and Interface ~cience, Vol. 30, No. 4, August 1969
572
F L A C H S B A R T A N D ST(~BER TABLE II ATTACHMENT OF D I F F E R E N T I~ADIOACTIYE 1SOTOPES TO SILICA PARTICLES CONDENSED IN THE PRESENCE OF THE TRACER
Metal ion
Attachment
Tracer concentration (ppm)
n4Cs
~ 1 mg/500 g SiO~ complete
2 X 10-6
65Zn
3.0 rag/500 gm SiO_~ complete
6 X 10-6
54Mn
2.5 mg/500 gm SiO2 complete
5 X 10.6
i24Sb ~gFe 51Cr
4.5 mg/500 gm SiO: complete 2.6 rag/500 gm SiO2 complete 0.5 ~g/500 gm SiO2 complete
9 X 10-6 5 X 10-6 1 X 10-9
~41Ce
150 t~g/500 gm SiO2 complete
3 X 10-7
7Be
30 ~g/500 gin SiO2 complete
6 X 10-8
58Co
10 t~g/500 gm SiO2 complete
2 X 10-s
58Co
300 mg/500 gm SiO2 70% t a k e - u p
6 X 10-~
Maximumdesorption, exchange, or chelation Complete exchange by adding 1 gm of stable CsC1 p i t = 2; E D T A (20%) complete removal p H = 0.1; M n C12 release of 89% p H > 9; release of 6% p~I > 9; release of 5% p H = 1.2; K Cr (SO4)2 release of 3% p H = 8.8; Ce (NO3)3 release of 5% p H = 10; t a r t a r i c acid release of 9% p H = 11; Co C12 release of 3% p H = 1.3; Co C12 release of 74%
Incorporation No No Partial Yes Yes Yes Yes Yes Yes Partial
FIa. 1. E l e c t r o n micrograph of sample of silica particles grown to final size (mean d i a m e t e r 8 X 10-~ era). Journal of Colloid and Interface Science, Vol. 30, No. 4, August 1969
PREPARATION OF RADIOACTIVELY LABELED SILICA SPHERES
The possibility that the tracer was depleted before the particles had grown to their final size cannot be excluded, however. An exact proportionality between the relative amount of silica dissolved at high pH values and the relative activity subsequently recovered from the surface at low pH values could not be established with sufficient certainty. If the activity is completely incorporated in the early stage of the particle growth, i.e., if the foreign oxides are primarily distributed around the core of the particle, then the activity of a sample of particles of a polydispersely grown sol would be representative of the number of particles rather than their mass. The various results obtained for the amount of silica dissolved at high pH values is in keeping with findings reported in the literature. It has been shown (2, 3) that the rate of hydrolysis of silica in aqueous suspension is substantially influenced by the presence of various salts and organic compounds. The concentration of saturated silicic acid increases rapidly (4) from 0.1 gin/liter at pH _~ 8 to 5 gin/liter at pH = 11. However, it may take a few days to several weeks depending upon the additional solutes to establish this equilibrium. Correspondingly, the hydrolysis of resuspended silica particles at high p i t value may appear suppressed, as in the presence of cobalt salts, or moderate, as in the presence
573
of chrom alum, or increased, as in the presence of cerium nitrate. SUMMARY A recently developed technique of growing monodisperse silica spheres of predetermined size in alcoholic solutions facilitates the incorporation of certain radioactive tracers. Spheres growing to a final diameter of 8 X 10-5 cm were used in a study with radioactive isotopes including 7Be, 51Cr, 5~Mn, ~sCo, 59Fe, 65Zn, 124Sb, 134Cs, and 141Ce. The results reveal that those metals which form oxides of low solubility at pH >_ 7 are easily incorporated into the bulk of the silica spheres when present during the growth of the particles in an amount not exceeding 1-10 ppm with respect to the total mass of silica in the system. ACKNOWLEDGMENT This paper is based on work performed under contract with the United States Atomic Energy Commission at the University of Rochester Atomic Energy Project and has been assigned Publication No. UR-49-1069. REFERENCES 1. STOBER,W., FINK, A., AND BOHN, E., J. Colloid and Interface Sei. 26, 62 (1968). 2. FRIEDB]~nG,K., Beitr. Silikose-Forsch., Sonderband, p. 49 (1955). 3. STOBE~,W., Kolloid-Z. 147,131 (1956). 4. ALEXANDER,G. B., HESTON, W. M., ANn ILER, R. K., J. Phys. Chem. 58,453 (1954).
Journal of Colla4dand Interface Science, VoI.30~No. 4, August1969