Speciation Analysis of Colloidal Silica in Acidic Conditions Ewa Cukrowska, Levi Ochieng, Hlanganani Tutu, Luke Chimuka
School of Chemistry, University of the Witwatersrand, P. Bag X3, 2050 Johannesburg, South Africa [email protected]
Abstract Dissolved silica speciation is of great importance in mining as it interferes in the various metals extraction/removal processes especially in electro-winning. Analytical methods have been devised to identify species in colloidal silica in acidic conditions. Total sil- ica was determined aft er microwave acid digestion by ICP-AES. Monosilicic acid was determined using molybdic method; alpha and beta isomers were quanti ed. Silicon dioxide was determined a er precipitation as potassium uorosilicate by titrimetric method. Various interfering ion species which can co-precipitate with silicon dioxide were investigated. Keywords: colloidal silica, acidic conditions, speciation
Introduction fuse to form interlinking chains. e aggrega- Silicon is the second most common element tion of subunits due to van der Waals forces found in the earths crust and mantle. It is most results in a so then hard gel. abundant in the form of silicates and occurs Most literature on silica analysis focus throughout the mineral realm complexed to on the determination of monomeric form of metals such as magnesium, calcium, iron and silica using colorimetric technique where a aluminium (Ning 2002; Holleman-Wiberg molybdate solution reacts with monomeric 1995). silicic acid to produce silicomolybdic acid Many di erent techniques have been as proposed by Dienert and Wandenbulkcke adapted to extract minerals from ores. e (Govett 1961; Roa 1992). rst step of processes is the milling or crush- Silicic acid is a tetravalent weak acid with ing of ores followed by a dissolution step. A a strong tendency to eliminate water during method of metal extraction is electro-win- condensation reactions with other silicic acid ning, once the sample is completely dissolved. molecules (Ning 2002; Holleman-Wiberg Silica dissolves from the ore samples, 1995). e acid is most stable at a low pH of both under acidic or alkali conditions. Vari- 1 - 3 and least stable at pH 5 - 6. e solu- ous species of silica and its complexes un- bility of silicic acid increases in the presence dergo polymerization to form colloidal sus- of acids such as hydro uoric acid (HF) and pensions. e colloidal silica then interferes reacts with it to form silico ouride anions with electrodes, complexes with metal ions (Govett 1961). Polymerization of silicic acid and metal extraction processes cannot occur results in large polymers that form a colloidal (Ning 2002). solution. e molecules condense to produce ere are two types of colloidal systems a Si-O-Si bond between two molecules. Si- formed by silica. Silica sols are poly-silicic licic acid polymerization is strongly depen- acids in spheroid, amorphous forms. When dant on pH (Ning 2002). Below pH 3.2, the
sodium silicate (Na4SiO4) is acidi ed a sol polymerization reactions result in linear or is formed. Addition of a base causes depro- openly branched polysilicic acids. Above pH tonation and a negative overall charge of the 3.2, the polymerization results in cyclic or molecules that prevent bonding and results cross-linked polymer arrangements. Polym- in a sol. Under very specifi c conditions, a sol erization is rapid in alkaline or neutral envi- can convert itself to a gel (Holleman-Wiberg ronments and in acidic pH it is much slower. 1995; Govett 1961). If the pH is acidic or if a e condensations eventually result in a 3-D
base does not stabilize the silica, the subunits structure with the repeating units of (SiO2)n.
Wolkersdorfer, Ch.; Sartz, L.; Weber, A.; Burgess, J.; Tremblay, G. (Editors) 103
02_Water treatment process BOOK.indb 103 9/3/18 11:58 AM 11th ICARD | IMWA | MWD Conference – “Risk to Opportunity”
Monomeric as well as polymeric silica hydroxide was dissolved in 1 L of deionised species exists in dissolved ore samples. It is water. imperative that their concentration is deter- Preparation of molybdate acidic reagent. A mined. It is therefore not substantial to just mixture of 100 mL of ammonium molybdate know the total concentrations of silica but solution, 500 mL of deionised water and 200 more importantly, its speciation (Holleman- mL of 1.5% (H2SO4) were mixed to give a Wiberg 1995). concentration of 0.0707M MoO4-2, 0.375M e objective of this study was to develop SO4-2 and 0.148M NH+4, with a pH of 1.2. methods for the analysis of various silica spe- cies in colloidal matrix under acidic condi- Alpha Monosilicic acid analysis tions and to apply the developed methods to A mixture of 20 mL of molybdate acidic re- carry out silica speciation in electro-winning agent with no more than 5 mL of sample so- processes. lution was prepared containing about 0.01 to
1.0 mg of SiO2 and the total volume was then Experimental diluted to 25 mL with deionised water. e Determination of silica in this study followed absorption was measured at a wavelength of the following procedure: standard sample 410 nm. Fresh reagent mixtures were made preparation, determination of total silica con- daily (Govett 1961). tent (ICP-AES), monomeric silica determina- Beta Monosilicic acid analysis tion (Molybdate reagent) and SiO2 determi- nation by titrimetric method. Sample solution was diluted to silica concen- tration of between 0.5 - 4 mg/L. All dilutions Preparation of colloidal silica stan- preformed were with 1.5 % H2SO4. 10 mL of dard samples each sample was then into 50 mL asks. 2 mL Analytical grade silica gel was milled to a ne of a 10% w/v solution of ammonium molyb- powder using Fritsch Planetary mono mill date was added. e solutions were mixed GmbH ‘Pulverisette 6’ at a speed of 130 rpm and allowed to react for 20 minutes. 2 mL of for 6 hours, sulphuric acid solution of known 8% w/v solution of tartaric acid was added pH was added and the mixture shaken for 16 to the reagent mixture to prevent ferric (III) hours to prepare colloidal solutions of known ion reduction, followed by 1 mL of reducing concentration (Saccone et al. 2006) . agent (4-amino-3-hydroxy-1-naphthalene- sulfonic acid/sodium metabisul te/sodium Acid digestion and determination of sul te solution). e sample was allowed to total silica by ICP-AES stand for 20 minutes to react before reading Acid digestion was used to convert samples the absorbance on a UV-Vis spectrometer into solution for determination of total sili- ((UV- 1201) at 820 nm). ca (Multiwave 3000 digestion Anton Paar Determination of silica dioxide (SiO ) GmbH). 0.1 g of ground sample masses 2 were weighed into reaction vessels and 5 mL To a 1 mL of sample, 5 mL of 20% calcium chloride and 1 g of sodium uoride was added; HNO3, 5 mL HF and 0.5 mL HCl added. Th e digested samples are then placed in a 100 mL the solution was stirred. Solid potassium chlo- volumetric ask and mixed with boric acid in ride was added until an excess of 2 - 3 g over the ration of 6 mL : 1 mL of hydro uoric acid saturation point was attained. e precipitate [Roa 1992; May and Rowe (1965). Total silica was ltered and washed with a solution con- was determined by spectral analysis using taining 70 g of potassium chloride in 1 L of Spectro ICP-AES Genesis. 1:1 ethanol-water to ensured e ective removal of excess acid. e washed precipitate, lter Determination of soluble silica paper were transferred to a 1 L beaker con- (monosilicic acid complexation). taining 500 mL of boiled deionised water. e Preparation of a stable solution of ammonium solution was titrated with 0.15M sodium hy- molybdate. 100 g of (NH )6Mo O .4H O and droxide and 1% phenolphthalein (Lewis-Russ 4 7 24 2 A, Ranville J and Kashuba AT (1991). 47 g of concentrated (28% NH3) ammonium
104 Wolkersdorfer, Ch.; Sartz, L.; Weber, A.; Burgess, J.; Tremblay, G. (Editors)
02_Water treatment process BOOK.indb 104 9/3/18 11:58 AM 11th ICARD | IMWA | MWD Conference – “Risk to Opportunity”
Results and discussion 1.8 -2.0. ey were le to stand and sampled Total silica at the top, middle and bottom for analysis to assess the stability of the colloidal solutions e concentration of Si determined for the and to get a representative fraction of the col- prepared colloidal solution is shown in table loid samples. e means and percentages of 1. e measurements were done in triplicate monosilicic acids were calculated (table 2) and taken at a wavelength of 251.612 nm. for beta and alpha monosilicic acid values, Percentage of silica in the solution was de- respectively. termined as 99.4% with a good precision of e values obtained are all time depen- 0.15 % RSD. dant as the beta isomer is continually con- Monomeric silicates verting into the stable alpha isomer. e two silicic acids di er due to variation in their Monomeric silicates were quanti ed using ionic charges. UV-Vis spectroscopy. Th e acidic molybdate Beta molybdosilicic acid predominates reagent forms two compounds when react- during maximum colour formation at pH ing with the Monosilicic acid; Molybdenum 1.2 -1.4 and below. Both beta and alpha acids heteropolyacid (MHA) (SiMo O )-4 and 12 40 are formed at pH 1.8 2.0. e absorbance heteropoly anion (SiMo O )-5 complexes. 12 40 of alpha acid increases with time at any pH ese ions are commonly known as the alpha value. e time dependence is minimized (MHA) and beta (anion) forms. e alpha under strongly acidic conditions. erefore form is the most stable while the beta form the results obtained in acidic media are ac- converts spontaneously to the alpha form at curate. pH above 2.5. e MHA complex forms a Above a pH value of 3.6 the alpha form yellow solution that was measured at 410 nm dominates and can be purely alpha. e alpha while the reduced beta species forms a blue molybdosilicic acid absorbance increases with complex that was measured at a wavelength time at any given pH. is occurs due to the of 815 nm. e alpha isomer is very stable conversion of the unstable beta monomer to it. at pH 3.5 - 4.5 while beta formation occurs e precision of the process is indicated more readily at pH 0.8 - 2.5. When the envi- by the RSD values. Generally the higher the ronment is at a low pH the conversion of the pH of the system the lower the RSD value and beta form is very slow (Govett 1961; Maybodi the greater the precision. e rate of conver- and Atashbozorg (2006). If the reduction of sion of the beta isomer to the alpha isomer is beta molybdosilicic acid occurs at a low pH greatly reduced due to the acidic medium and of 1.2 the heteropoly blue anion (beta- (Si- the ROOM temperatures in which the experi- Mo O )-5) results. 12 40 ments were done. ree colloidal solutions were prepared of pH ranges of between 0.8 – 1.0, 1.2 – 1.4 and
Table 1. Concentration of total silica by ICP-AES
Solution # Concentration of silica (ppm)
1 0.994 2 0.993 3 0.996 Mean 0.994 SD 0.002 RSD (%) 0.15
Wolkersdorfer, Ch.; Sartz, L.; Weber, A.; Burgess, J.; Tremblay, G. (Editors) 105
02_Water treatment process BOOK.indb 105 9/3/18 11:58 AM 11th ICARD | IMWA | MWD Conference – “Risk to Opportunity”
Table 2. Beta and alpha monosilicic acid concentration in μg L-1 determined by UV-vis spectroscopy at dif- ferent acidic pH ranges
Sampled region of colloidal solution pH range of the solution analyzed
0.8 - 1.0 1.2 - 1.4 1.8 - 2.0 Beta monosilicic acid Top 47.4 43.8 39 Middle 48.6 49.2 37.8 Bottom 50.4 52.8 38.4 X (mean) 48.8 48.6 38.4 SD 1.5 4.5 0.6 RSD 1.2 3.7 0.5 Total Beta % 87.1 86.8 68.6 Alpha monosilicic acid Top 3.4 5.45 14 Middle 8.0 8.5 11.9 Bottom 6.6 6.8 13.4 X (mean) 6.0 6.8 13.0 SD 2.4 1.6 1.2 RSD 1.9 1.3 0.9 Total Alpha % 10.7 12.2 23.3
Table 3. SiO2 (%) in colloidal solution determined by precipitation and titration method (titre volumes are given in mL)
Sampled region of colloidal solution pH range of the solution analyzed
0.8 - 1.0 1.2 - 1.4 1.8 - 2.0 Top 1.4 1.6 0.4 Middle 1.5 1.8 0.7 Bottom 1.6 1.4 4.1 X (mL) 1.5 1.6 1.7 SD 0.1 0.2 2.1
% SiO2 80.4 85.7 92.9
Silica dioxide (SiO ) 2 for SiO2 by precipitating it out of solution Titrimetric method was used to determine then titrating with NaOH.
the amount of silica dioxide by alkaline fu- Percentage of SiO2 in solution was calcu- sion. e titrimetric values were multiplied lated as follows: % SiO2 = 0.150 mL NaOH/g by a fraction that is dependant on the experi- of sample mental conditions to determine percentage. Hence for the top sample: 1.4 mL e experiment must be done as quick as NaOH titre, (Dilution factor 500) possible as the longer it takes the more errors % SiO2 = (0.015 1.4/1.4) = 105/1.4 = 74%. occur due to co-precipitation by other metal e stability of a system greatly a ects the species Glaso O and Patzauver G (1961). accuracy of the results obtained. For the pH ree colloidal solutions were prepared in region of 1.8 - 2.0 silica stability in solution pH ranges of between 0.8 – 1.0, 1.2 – 1.4 and reduces and the percentage of the compound 1.8 – 2.0. ey were le to stand and sampled increases. erefore, the system exists at a at the top, middle and bottom and analysed meta-stable equilibrium with its monomer.
106 Wolkersdorfer, Ch.; Sartz, L.; Weber, A.; Burgess, J.; Tremblay, G. (Editors)
02_Water treatment process BOOK.indb 106 9/3/18 11:58 AM 11th ICARD | IMWA | MWD Conference – “Risk to Opportunity”
150
130 Ca
110 Al % of SiO2 % of 90 Ti
70
50 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 Conc. of interfering ions (Ca, Al & Ti) Figure 2 Change in SiO2 percentage with increasing interfering ions concentrations
Interferences concentration of 0.2 ppm, % SiO2 approaches Metal ions present in solutions can cause in- a maximum value and stays constant with terferences as they complex with silica diox- continued increase concentration. ide and precipitate out as colloid. e posi- Titanium: Of all the three metal ions in- tively charged ions electro-statically interact vestigated, titanium has the least in uence with the negatively charged surface of the in increasing silica dioxide concentration by colloid. Once the metal co-precipitates with co-precipitation. Initially there is a gradual
silica it is incorporated into the colloidal increase in % SiO2 and a maximum is rapidly structure and becomes di cult to extract reached. % SiO2 stays relatively constant a er during mining. an increase in titanium ion concentration of Silica solutions were spiked with various about 0.15 ppm.
concentrations of interfering ions and then Co-precipitation of SiO2 occurs with cal- analysed. Calcium aluminium and titanium cium, aluminium and titanium in solution. exhibited di erent trends as shown in g. 2. is is due to the solubility constants of these Calcium: Calcium causes an immedi- compounds that are minimal. Calcium ions ate exponential increase in the silica dioxide suppress the co-precipitation of aluminium concentrations right from the start. It is the and titanium. e lowest complexation oc- most reactive interfering ion. It appears that curs with titanium. in any competition for uoride ions, calcium All the graphs show an increase the colloi- is stronger than aluminium and titanium ( g. dal silica percentage to a maximum value and
2). ere is a drastic decrease in % SiO2 a er then decreases gradually as no more metal 0.2 ppm for the calcium curve and this could ions can complex by binding due to satura- be due to reversible attachment. tion of the binding sites. e calcium inter- Aluminium: Initially aluminium causes ference reached a maximum value and then a small increase in percentage silica dioxide decreased which indicates a reversible attach- but exponentially shots up at a concentration ment to the colloidal silica. of about 0.1 ppm. At a slightly higher Al ion
Wolkersdorfer, Ch.; Sartz, L.; Weber, A.; Burgess, J.; Tremblay, G. (Editors) 107
02_Water treatment process BOOK.indb 107 9/3/18 11:58 AM 11th ICARD | IMWA | MWD Conference – “Risk to Opportunity”
Table 4. Mean, standard deviation and relative standard deviation for various species analysed in the Skor- pion leachate samples
Sample pH Alpha Beta Silica ioxide *Total Si (ICP-AES) (mg/L) (mg/L) dioxide T1 Mean SD 0 0 0.26 0.27 RSD 0 0 0.21 1.26 T2 Mean SD 0 0 0.45 0.34 RSD 0 0 0.37 1.96 T3 Mean SD 0 0 0 0.31 RSD 0 0 0 1.77 Percentages (%) given in parenthesis, * Total Si a er acid digestion
Silica speciation in leachate samples Govett GJS (1961) Critical factors in the colo- Acid leachates of Skorpion samples at di er- rimetric determination of silica. Anal. Chim. ent pH values ranging from 0.8 – 2.0 obtained Acta, 25, 69-80 from Anglo Research Lab. were analyzed for Roa CRM Reddi GS and Rao TAS (1992) Acid alpha, beta and total silica, table 4. decomposition procedure for the spectrophoto- metric determination of silica in rocks and min- Conclusions erals at room temperature. Anal. Chim. Acta, Speciation of colloidal silica in acidic condi- 268, 357-359 tions was successfully achieved in this study. Glaso O and Patzauver G (1961) Determination Various interfering ion species that co-pre- of silica in iron ore. Anal. Chim. Acta, 25, 189- cipitate with the silicon dioxide and increase 192. the titrimetric values such as Ca (II), Ti (II) May I and Rowe JJ (1965) Solution of rocks and and Al (III) were assessed. refactory minerals by acids at high tempera- Good percentage recoveries, accuracy tures and pressures. Determination of silica af- and precision were obtained for the methods ter decomposition with hydro uoric acid. Anal. used and they were applied to analyze leach- Chim. Acta, 1965, 33, 648-654. ates from acid leaching of silica at di erent Lewis-Russ A, Ranville J and Kashuba AT (1991) pH values. Di erentiation of colloidal and dissolved silica: analytical separation using spectrophotometry References and inductively coupled plasma atomic emis- Ning RY (2002) Discussion of silica speciation, sion spectrometry. Anal. Chim. Acta, 204, 509- fouling, control and maximum reduction. De- 511. salination 151, 67-73 Maybodi AS and Atashbozorg E (2006) Quanti- Holleman-Wiberg (1995) Inorganic chemistry. tative and qualitative studies of silica in di er- Academic press. Berlin and New York, 822-936. ent rice samples grown in north of Iran using Saccone L, Conley DJ and Sauer D (2006) Method- UV-vis, XRD and IR spectroscopy techniques. ologies for amorphous silica analysis. J. of Geo- Talanta, 70, 756-760. chem. Exploration, 88, 235-238.
108 Wolkersdorfer, Ch.; Sartz, L.; Weber, A.; Burgess, J.; Tremblay, G. (Editors)
02_Water treatment process BOOK.indb 108 9/3/18 11:59 AM