THE FLOTATION OF APATITE AND DOLOMITE IN ORTHOPHOSPHATE SOLUTION by DAVID LAWRENCE JOHNSTON B.A.Sc., The University of British Columbia, 1963 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF APPLIED SCIENCE in the Department of MINERAL ENGINEERING We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA March, 1969 In p r e s e n t i n g t h i s t h e s i s i n p a r t i a l f u l f i l m e n t o f t h e r e q u i r e m e n t s f o r an a d v a n c e d d e g r e e a t t h e U n i v e r s i t y o f B r i t i s h C o l u m b i a , I a g r e e t h a t t h e L i b r a r y s h a l l make i t f r e e l y a v a i l a b l e f o r r e f e r e n c e and S t u d y . I f u r t h e r a g r e e t h a t p e r m i s s i o n f o r e x t e n s i v e c o p y i n g o f t h i s t h e s i s f o r s c h o l a r l y p u r p o s e s may be g r a n t e d by t h e Head o f my D e p a r t m e n t o r by h i s r e p r e s e n t a t i v e s . I t i s u n d e r s t o o d t h a t c o p y i n g o r p u b l i c a t i o n o f t h i s t h e s i s f o r f i n a n c i a l g a i n s h a l l n o t be a l l o w e d w i t h o u t my w r i t t e n p e r m i s s i o n . D e p a r t m e n t o f Mineral Engineering The U n i v e r s i t y o f B r i t i s h C o l u m b i a V a n c o u v e r 8, Canada D a t e A p r i l 9, 1969 TEMPORARY RESTRICTION OF USE This thesis contains information of a confidential nature and access to the thesis shall be only with the written permission of the Head of the Department of Mineral Engineering. The period of restricted access to this thesis shall end on Ap r i l 30, 1971. A^opy«of^the=theslsJjwaJ.Jl«at==that0 time'=be="forwarded=to=the=Nati'ona4-=Library. Apr i l 16, 1969. i ABSTRACT A study has been conducted on the loss of phosphate ions and excessive oleic acid consumption encountered i n the selective flotation of dolomite from apatite. An attempt has been made to evaluate the mechanism by which orthophosphate ions depress apatite flotation. -2 - 2 Replacement of SO^ ions on gypsum by HFO^ ions i s found to occur rapidly i n solution. The presence of CaHPO^ * 211^0 on gypsum surfaces i s shown using infrared reflectance spectroscopy. CaHPO^^I^O i s isomorphous with gypsum and has identical lattice parameters. Experiments show that the addition of sulfate ion to the system suppresses the reaction of orthophosphate ions with gypsum by the common ion effect. The proposed reversible reaction i n the system i s : C a S 0 , - 2 H o 0 + HPO"2-*—r CaHPO - 2 H 0 + SO - 2 4 2 4 4 2 4 A proposed mechanism by which orthophosphate ions select-ively depress apatite flotation i s shown to f i t a l l experimental observations. Orthophosphate ions are known to be potential determ-ining for calcite, dolomite and apatite. It i s observed that i n the presence of orthophosphate ions calcite and dolomite recovery i s higher at pH 6.0 than at pH 8.5. Apatite and gypsum exhibit opposite behavior. It i s proposed that the adsorption of strongly hydrogen -1 -2 bonding R3PO4, H 2 P0^ and HPO^ in the electrical double layer results in collector species being slow to penetrate and adsorb on the minerals. Acid attack on calcite and dolomite results in disruption of the hydrogen bonded layer and allows rapid c o l l e c t o r adsorption. Two equally important effects of adsorbed phosphates are to decrease c o l l e c t o r adsorption due to the large negative zeta potential generated and to impede f r u i t f u l particle-bubble interactions. B r i t t l e froths encountered i n f l o t a t i o n at pH 6.0 are related to condensation of surface films formed by o l e i c acid -sodium oleate complexes. The fro t h s t a b i l i z i n g effects of hydro-phobic dolomite p a r t i c l e s i s noted. ACKNOWLEDGEMENTS The continued advice and help of Dr. J. Leja, the author'8 research director, i s gratefully acknowledged. An expression of thanks i s due to Dr. H. E. Hirsell and R. Tipman who sacrificed some of their time to the discussion of ideas and explanation of experimental techniques. Appreciation i s expressed to Cominco Limited whose financial help made this work possible. TABLE OF CONTENTS ABSTRACT i ACKNOWLEDGEMENTS i i i LIST OF TABLES iv LIST OF FIGURES v 1 INTRODUCTION 1 1-A The Flotation of Sedimentary Phosphate Ores . . . . 1 1-A-l Mineralogy 1 1-A-2 Phosphate Ore Flotation 2 1-A-3 The Selective Flotation of Dolomite from Apatite 4 1-B Chemistry of The Flotation System 5 1-B-l Solution Composition . . . . 5 1-B-2 Solubilities and Dissociation Constants . . . 7 l-B-3 Chemistry of Solid-Solution Interfaces . . . . 9 1-B-4 Chemical Reactions 11 1-C Previous Work 12 1-C-l Chemical Studies 12 l-C-2 Infrared Studies 14 l-C-3 Electrokinetic Studies 15 1-D Scope of Present Investigation 17 1- E Experimental Methods. 18 2 ANALYTICAL METHODS 19 2- A Surface Analysis of Minerals by Infrared Spectro-scopy 19 2-B Reference Spectra 20 2-C Absorption Spectra 22 2- D Internal Reflection Spectroscopy . . 22 3 EXPERIMENTAL 24 3- A Equipment 24 3-A-l Flotation 24 3-A-2 Frothing 25 3-A-3 Infrared 25 3-A-4 Zeta Potentials 26 3-A-5 Surface Tensions 26 3-B Materials 27 3-B-l Minerals 27 3-B-2 Organic Chemicals 29 3-B-3 Inorganic Chemicals . . . . . . . . . . . . . 29 3- C Procedures 30 3-C-l Flotation Tests 30 3-C-2 Frothing Tests 31 3-C-3 Oleic Acid Adsorption Tests 31 3-C-4 Orthophosphate Reactions with Solids . . . . 32 3-C-5 Sulfate Ion - Orthophosphate Ion - Gypsum Reaction 32 3-C-6 Zeta Potentials 33 3- C-7 Surface Tension Measurements . . . . . . . . 33 4 RESULTS 34 4- A Flotation Tests 34 4- A - l Apatite and Dolomite Flotation Tests . . . . 34 4-A-2 The Effects of Particle Size on Dolomite Flotation 34 4-A-3 The Effects of C0 2 on Dolomite Flotation . . 37 4-B Frothing Tests 37 4-C Oleic Acid Adsorption Tests 37 4-D Orthophosphate Ion Reactions with Solids in the System 44 4-D-l Apatite and Dolomite 44 4-D-2 Gypsum . . . . . 45 4-D-3 Anhydrous Dicalcium Phosphate 46 4- D-4 Sulfate Ion - Orthophosphate Ions -Gypsum reactions . . . . . 46 4-E Zeta Potential Measurements 48 4- F Surface Tension Measurements 48 5 INTERPRETATION OF RESULTS . . . 49 5- A Flotation and Frothing 49 5- A- l The Flotation of Dolomite and Apatite . . . 49 5-A-2 Frothing D i f f i c u l t i e s Encountered i n Flotation 51 5-B Loss of Orthophosphate 55 5-C Adsorption Studies 56 5-C-l Oleic Acid Adsorption and Correlation with Zeta Potentials 56 5-D Depressing of Apatite by Orthophosphate Ions . . 57 5-E Orthophosphate Requirements as a Function of Flotation Variables 59 6 CONCLUSIONS 60 7 FURTHER WORK 61 8 BIBLIOGRAPHY 63 9 APPENDICES 66 I Zeta Potential Results of Others 66 II Reference Spectra 67 III Zeta Potentials - Numerical Results 71 iv LIST OF TABLES Table 1 Unavoidable Ions 6 Table 2 Flotation Reagent Ions 6 Table 3 Solubility Products for Minerals 7 Table 4 Probable Solution-Solid Reactions 12 Table 5 Main Infrared Absorptions of Materials Encountered . 21 Table 6 Chemical Analyses of Minerals 28 Table 7 Infrared Absorptions of Oleates 38 Table 8 Impurity Absorptions in the Apatite Spectrum . . . . 40 Table 9 Fixing of Orthophosphate Ions by Gypsum 47 Table 10 Surface Tensions Values for Solutions Containing Orthophosphates and Oleic Acid 49 V LIST OF FIGURES Figure I Microphotographs of phosphate ores . 3 Figure 2 Orthophosphoric Acid Dissociation Products . . . . 8 Figure 3 Suggested Dissociation Curve for Oleic Acid . . . 9 Figure 4 The Electrical Double Layer 10 Figure 5 Total Internal Reflection of Radiation 23 (a) Reflection versus Angle of Incidence (b) Penetration versus Angle of Incidence Figure 6 Total Internal Reflections i n a Truncated Prism. . 24 Figure 7 Frothing Apparatus . . . . 25 Figure 8 Plate Holder for Infrared Analysis 26 Figure 9 Flotation of Apatite and Dolomite 35 Figure 10 Dolomite Flotation at pH 6.0 36 Figure 11 Dolomite Flotation at pH 8.5 36 Figure 12 Spectrum of Oleic Acid Adsorbed on Apatite i n Sulfate Solution 39 Figure 13 Spectrum of Oleic Acid Adsorbed on Apatite i n Phosphate Solution 41 Figure 14 Spectrum of Oleic Acid adsorbed on Dolomite . . . 42 Figure 15p Photographs of Froths 52 Figure 15 Spectra Showing Conversion of Oleate Salt to Oleic Acid by pH Shift 43 Figure 16 Spectra of Apatite and Dolomite Treated in Phosphate Solution . . . . . . 44 Figure 17 Spectra of Gypsum Samples Treated i n Phosphate and Sulfate Solutions 45 v i Figure 18 Spectrum of Anhydrous Dicalcium Phosphate Treated in Phosphate Solution . . . . . 46 Figure 19 Zeta Potential versus pH for Apatite and Dolomite . 50 Figure 20 Zeta Potential versus Orthophosphate Ion Concentration . . . . . 50 Figure 21 Zeta Potential versus Oleic Acid Concentration . . 50 1 1 INTRODUCTION 1-A The Flotation of Sedimentary Phosphate Ores 1-A-1 Mineralogy Sedimentary phosphate ores are formed on the margins of continental shelves under well defined physical and chemical con-ditions. The common mode of formation results i n considerable mineralogical similarity among phosphate ores*. The dominant minerals are carbonate-fluorapatite, s i l i c a and calcite. Dolomite, clays, hydro-carbons, gypsum and ferro-magnesium sil i c a t e s occur in varying ratios. Chemical weathering of phosphate ores results in the leaching of calcite and gypsum leaving residual material enriched in phosphate content; Such weathered ores are sandy and contain clay minerals. Beneficiation may consist of washing and drying to 2 derive a marketable product . Unweathered or "hard" ores contain apatite particles i n a siliceous or carbonate matrix. The texture i s often complicated by fine post-deposition veining with s i l i c a , calcite or dolomite. These ores must be finely ground and treated by flotation to derive a marketable phosphate concentrate. Common to a l l ores are the roughly spherical "oolites" of carbonate-fluorapatite which are formed by the phosphatization of chemically precipited calcium carbonate pisolites. Many oolites are nucleated around s i l i c a grains and are impure i n themselves. Phosphorite oolites contain 35% to 40% P2°5* Stoichiometric 2 fluorapatite would contain 42% ?2°5* Figures 1(a) and 1(b) are microphotographs of unweathered phosphorite ores showing oolites. 1-A-2 Phosphate Ore Flotation The beneficiation of "hard" phosphate ores requires the separation of apatite from s i l i c a or apatite from dolomite and calcite. Apatite can be floated e f f i c i e n t l y from s i l i c a t e s using anionic collectors. The separation of calcite or dolomite from 3 apatite by flotation i s only possible under certain conditions . 2 It i s common in North Africa to calcine and water leach ores to remove calcite and dolomite from apatite . These processes are relatively expensive. Flotation of apatite from siliceous sedimentary ores is carried out in a water-bearing pulp at pH levels between 8.0 and 9*0. Collector solutions may contain crude fatty acids or petroleum sulphonates and fuel o i l . Sodium s i l i c a t e , sodium hydroxide and sulfuric acid are used as modifiers. Stable froth for flotation i s provided by the crude fatty acid or petroleum sulfonate collectors. Several stages of flotation cleaning may be necessary to obtain the desired concentrate grade. A minimum of 30% P2O5 content i s often specified for concentrates to be used in the pro-duction of phosphoric acid by the wet process. It i s normal for dolomite and calcite contained in the ores to be concentrated with apatite. Oleic acid, the main constituent of industrial fatty acids, i s recognized as a universal 4 5 6 collector for apatite, dolomite and calcite. ' ' Figure 1(a) Mlcrophotograph of Typical Sedimentary Phosphate Ore Showing Oolites X35 Figure 1(b) Microphotograph of Sedimentary Phosphate Ore Showing S i l i c a Veins X35 p a r t i a l l y polarized l i g h t 4 The presence of carbonate minerals, particularly dolomite, i s highly undesirable in concentrates produced for the wet phosphoric acid process^. The carbonate neutralization reaction consumes sulfuric acid and causes reactors to overheat. Carbon dioxide gas generated contributes to severe foaming problems in reactors. Magnesium phosphate precipitation w i l l hinder acid settling and w i l l clog f i l t r a t i o n equipment. The greatest expense may be caused by a decreased plant capacity and low acid grade. A selective process for floating calcite and dolomite from apatite i s of economic interest to phosphate rock producers and f e r t i l i z e r manufacturers alike. 1-A-3 The Selective Flotation of Dolomite from Apatite The observation that soluble phosphate in flotation pulps using oleic acid inhibits the flotation of apatite was f i r s t published g i n a paper by Borisov in 1956 . Others have since applied this 9 10 9 effect i n processes which are patented ' . The Cominco process , most familiar' to the author, i s highly successful technically. However, in common with other modifications*® economic considerations have hindered i t s application. Unexplained losses of phosphate ion and unexpectedly large collector requirements appear to be the 11 1 2 greatest problems ' . 12 The Cominco process consists of the flotation re-cleaning of fatty acid flotation concentrates i n a solution containing 2 to 5 gm/liter ?205* Orthophosphoric acid or f e r t i l i z e r grade ammonium phosphate i s used to supply the required orthophosphate ions and to control the pH between 4.5 and 6.0. An industrial fatty 5 a c i d containing 60% o l e i c a c i d i s used as a c o l l e c t o r - f r o t h e r . An in t e n s i v e conditioning step at 65% s o l i d s i n the phosphate s o l u t i o n i s found necessary to deactivate a p a t i t e . F l o t a t i o n of the carbonate minerals i n the re-cleaning step i s rapid. Clean-up of the carbonate f l o a t i s hindered to some degree by a b r i t t l e f r o t h condition. It was noticed that as f l o t a t i o n proceeded there was a marked tendency f o r the f r o t h layer to thin. 12 Laboratory work done by Cominco personnel indicated that up to 4 pounds of f a t t y a c i d and 5 to 7 pounds of ammonium phosphate were required per ton of re-cleaning feed. A s i g n i f i c a n t reduction i n these reagent requirements would make the r e f l o t a t i o n process more a t t r a c t i v e economically. A fundamental study of the phosphate f l o t a t i o n system with p a r t i c u l a r reference to i n v e s t i g a t i n g the high reagent usage i n the Cominco r e f l o t a t i o n process forms the basis f o r t h i s thesis. 1-B Chemistry of the F l o t a t i o n System 1-B-l Solution Composition The aqueous s o l u t i o n present i n the f l o t a t i o n system has a much higher i o n i c strength than natural water. Numerous un-avoidable ions are contributed to the system through d i s s o l u t i o n of minerals present. These minerals and ions are shown i n Table 1. Table 1 Unavoidable Ions Mineral Formula Apatite Calcite Dolomite Gypsum Miscellaneous ions Ca 5(P0 4) 3 (F,C03) CaCO. CaMg(C03)2 CaSO *2H O 4 2 Ions i n Solution Ca*2, P0"\ H2PG^, HPO"2, F" 1, CO"2, 4 3 HCO"1 Ca* 2, of 2, HCO^1 Ca*2, Mg*2, CO"2, HCO" Ca*2, SO"2 4 C l " 1 , Na + 1, K*1, H*1, OH" 1, SiO" 2 Flotation modifiers and chemicals introduced to the solution are soluble and contribute the species shown in Table 2 Table 2 Flotation Reagent Ions Flotation Reagent Orthophosphoric acid Oleic acid Formula H 3P0 4 C 9 H 1 8 S C 8 H 1 5 C O O H Ions in Solution H 2H>; 1, HPO; 2, PO; 3 H* 1 ( C18 H33°2 )" 1' 7 l-B-2 Solubilities and Dissociation Constants Solubility product data for a l l the minerals are not available. Table 3 shows some values found in the literature. 13 Apatite has been shown to dissolve in an irregular way . Table 3 Solubility Products for Minerals at 25°C Mineral Solubility Product Ref. -25 Apatite (variable) Ksp (est.) 1 x 10 13 Calcite 2.9 x l ( f 9 to 4.8 x l ( f 9 14 Dolomite 3.7 x 10* 1 1 14 Gypsum 2.4 x IO - 5 14 Brushite 3 to 5 x 10 - 7 14 Solution content of ions present at pH 6.0 during flotation recleaning using the Cominco process i s postulated to follow the decreasing order: H2PO4 1, HPO^2, Ca* 2, SO^2, H*1, Mg*2 HCO"1, CjgH^O" 1, and CO*2. Orthophosphoric acid dissociates i n a well known manner. The three dissociation constants are: KI = 1.1 x IO - 2 K2 = 7.5 x IO - 8 K3 « 4.8 x 10" 1 3 Figure 2 i s a graphical representation of molar quantities -1 -2 -3 of H3PO4, H2PO4 , HPO4 and PO4 present in solution at pH values between 3 and 11 when a 0.1 molar H3PO4 solution i s neutralized with NaOH. Between pH 4.5 and 6.0 H2PO4 1 i s the most abundant phosphate species in solution. I I I I I I I I 1 I 2 3 4 5 6 7 8 9 10 11 PH Figure 2 Orthophosphoric Acid Dissociation Oleic acid i s known to be soluble i n water to a s l i g h t degree. The l i t e r a t u r e contains few references to the s o l u b i l i t y 15 - i l i m i t of t h i s compound. One reference c i t e s a value of 0.6 x 10~ moles/liter at 25°C. The observed s o l u b i l i t y i s dependent upon the degree of o l e i c acid dissociation and th i s i n turn i s governed by the pH. The dissociation curve resembles that shown i n Figure 3^. Solubilities of sodium oleate and calcium oleate quoted in the literature are 100 gm/liter and 0.4 gm/liter at 25°C respectively l-B-3 Chemistry of Solid-Solution Interfaces Collector adsorption and flotation depend upon the surfac properties of minerals. The identity of cations and anions in the mineral solution interface and their mode of adsorption are of utmost importance to flotation. It is recognized that reaction between collector ion and the mineral surfaces to produce surface hydrophobicity can be modified by the addition of certain ions to -1 -2 -2 *V2 solution. Examples of modifying ions are CN , S , SO, , Cu , -2 -1 SiG\j and MnO^ . The specific action of these ions i s not always predictable since i t is governed by numerous parameters of each system. -2 -1 42 Ions such as S , CN and Cu react chemically on mineral 4, 6 surfaces forming compounds . One characteristic of their behavior is that a small addition of ions can produce gross effects. Colloidal compounds such as starches and lignin salts form physically adhering layers around mineral particles and prevent collector adsorption. This effect of starches is used in potash flotation to "blind" fine clay particles and prevent adsorption of amine collectors. Modifiers may suppress exchange reaction between minerals -2 and collector ions. Sodium s i l i c a t e and S are known to have such properties. -1 *3 -3 Highly polarizing species such as F , Al and PO^ i n the electrical double layer act through changes in chemical and electrokinetic potentials on collector adsorption. Figure 4 is a schematic diagram of the electrical double layer around a mineral particle. Solid •i G s y I a v u o Stern layer Shear plane Solution Potential determining ions Zeta potential Distance Figure 4 Structure of the Electrical Double Layer The electrical double layer theory is well described else-4 where and will not be explained in detail. It is important to note that in order to change the zeta potential sign of a mineral, oppositely charged ions must be adsorbed into the inner fixed portion of the Stern layer in excess of the previously fixed ions. Such ions are termed "potential determining". The adjacent portion of the Stern layer contains hydrated specifically adsorbed ions or molecules. The Gouy portion of the double layer is a diffuse layer of cations and anions* The zeta potential is measured between a shear plane in the Gouy layer and the solution bulk. The value is determined by the degree of shear and therefore the position of the shear plane. The formation of collector - mineral "complexes" enabling flotation of minerals involves the formation of bonds between the collector ion and mineral cation . It is due to the common calcium atoms of apatite, calcite and dolomite that the selective flotation of one mineral from the other requires the use of a specific depressant. l-B-4 Chemical Reactions Flotation systems are seldom in a state of chemical equi-librium. Cations and anions are supplied to solution and precipitates are continuously being formed. Indications of such reactions taking place are shown by changes in solution content of ionic species. The phosphate salt solution in which the flotation of dolomite from apatite is effected is a typical example. A rise in pH and a decrease in phosphate assay of solutions take place gradually as flotation proceeds. The probable reactions involving solids in the system are shown in Table 4. Table 4 Probable Reactions CaCG3 + 2H* Ca*2 + CG2 • H 2© 1 Ca Mg(C03)2 • 4H + I^*Ca* 2 • Mg*2 f 2C02 * 2^0 2 Ca 5(P0 4) 3(P) 1_ 2 x(C0 3) x + 2xH* L^5Ca* 2 • 3P0"3+ (l-2x)F" 1 • xC02 +3fl 20 3 CaS04 • 21^0^ Ca*2 • SO"2 * 2H20 4 3Ca*2 t 2P0 4 3^Ca 3(P0 4) 2 5 Ca*2 • HP042 + 21*20 ,^CaHP04 • 2H20 6 Ca*2 • 2H2P04V^Ca(H2P04)2 7 3Mg*2 t 2P0 4 3 + xH 20 T^Mg 3(P0 4) 2 • x^O 8 Mg*2 • HP042 * xH20^MgHP04 • xH20 . 9 Ca*2 • 2(C 1 8H 3 30" 1 4 2 ^ 0 ^ C a ^ H ^ O ^ . 2^0 10 Mineral - Collector ^ - J * Mineral: Co Hector 11 The sequence of the four most dominant reactions in the flotation system at pH 6.0 is postulated to be 1, 6, 2, 4; the others follow. 1-C Previous Work 1-C-l Chemical Studies g Borisov (1956) noted that phosphoric acid, used as a pH regulator in dolomite-apatite flotation with oleic acid, decreased the flotability of apatite. Since 1956, Montecatini, Cominco Ltd., 13 and the Israel Mining Industries Institute for Research and Develop-ment have sponsored work on the use of soluble phosphate in flotation processes. The only published information on these efforts are the 10 11 Montecatini patent and the paper by Mitzmager et a l of Israel Mining Industries Institute. The Montecatini patent and a patent applied for by Cominco Ltd. do not discuss the mechanism by which phosphate ions depress phosphate minerals. The paper published by Mitzmager et a l includes two hypotheses on the action of soluble phosphate i n creating a flotation differential between calcite and apatite. Mitzmager contends that: (a) Solution acidity promotes flotation of calcite by generating CO2 bubbles on the particle surfaces. (b) The phosphate ions react with gypsum included on the apatite surfaces resulting i n the formation of a layer of dicalcium phosphate. This layer was assumed to blind the surface or "close the surface" to collector species. Mitzmager reported an almost complete loss of phosphate ion from solution during the flotation of a fine reject ore fraction. The loss of phosphate ion was attributed to the reactions 1, 4 and 12. CaC03 * 2 H 4 1 ^ —*Ca* 2 • H20 • 0>2 1 C a S 0 4 « 2 H 2 0 v Ca* 2 f 2H20 4 NaH2P04 • Ca 4 2^—^ C a P 0 4 4 Na*1 * H* 12 A portion of the same study i n which radioactive tracers were used demonstrated that calcium ions and phosphate ions were adsorbed by gypsum. The formation of CaHP04 on the gypsum surface was postulated. 14 18 Petrova et a l described the use of sodium hexametaphosphate to depress apatite flotation in treating a scheelite ore. An addition of 0.02 lbs,/ton of ore was made. 19 Chernyi summarized the work of Borisov and others in a study. He stated that H** i s adsorbed on apatite and through selective *2 dissolution of ?2®5 t n e s u r ^ a c e i s Ca *" rich. He theorized that H^ PO'* i s adsorbed in the Stern layer and displaces collector ions. f 2 The effect postulated for dolomite was a selective dissolution of Ca . Adsorbed collector ions in the Stern layer were postulated to bond with the Mg*2 rich surface. H** was said to be consumed in hydrating the dolomite surface. H2SO4 was stated to activate apatite flotation. The use of CaHPO^ (38.4 Kg/ton) with a small amount of inorganic acid was recommended as depressant. l-C-2 Infrar Stu ies 16 An infrared study by Knubovets and Maslennikov indicated that i n a 1.5 gm/liter H^ PO^ solution oleic acid occurs i n a molecular form adsorbed on phosphate minerals. The conclusion drawn was that the presence of excess phosphate ions saturated the free valence of the phosphate mineral surface and hindered chemisorption of the anionic collecting species. It was also noted that under these conditions calcium oleate s t i l l formed to some extent on the dolomite surfaces. A flotation differential involving depression of phosphate mineral would therefore be generated in the presence of soluble phosphate. L. I. Stremovskii stated i n a later paper 1 7, "In our opinion the selective action mechanism of the mineral acid consists 15 of conversion of the collector from the ionic state into molecular form which i s responsible for the selective adsorption of the collector toward dolomite." Evidence given by Peck 2 0 and Knubovets 1 6 that under acid conditions oleic acid occurs mostly in molecular form on both calcite and apatite invalidates such an argument. 21 Chapman and Thirlwell reported an extensive investi-gation u t i l i z i n g infrared spectra of orthophosphates. Their comparison of Raman and infrared spectra of I^PO^1 and H-jPO^ indicated hydrogen bonding was sufficiently strong to alter the appearance of OH absorption peaks at 2900 - 3000 cm"1 i n water. In addition the threefold axial symmetry in the tetrahedral H3PO4 is disturbed. It was concluded that the hydrogen bonding between water molecules and phosphate ions i s considerably stronger than that between water 21 molecules. This conclusion agreed with that of Simon who observed that water molecules and orthophosphoric acid molecules are bonded into a macrostrueture. l-C-3 Electrokinetic Studies of Adsorption Reactions 7, 8 Borisov and his co-workers have reported detailed investigations into the effects of ion adsorption on the zeta potential of minerals. The observed changes in zeta potential have been correlated to changes in flotation properties. The absolute changes in zeta potential have been termed "electrokinetic indices". Thus the preferential adsorption of a cation would result i n a positive index and the adsorption of an anion a negative index. The positive electrokinetic index of apatite i s shown to become increasingly positive with increasing H,P0, concentration i n solution. The values for dolomite become increasingly negative. The values given are for levels between 0 and 1.4 gm/liter. The pH range covered would be from 8.0 to 1.8. An opposite trend was shown by gradually neutralizing a mineral suspension con-taining 1.4 gm/liter H3PO4 by adding magnesium carbonate. Borisov also showed that for a fixed concentration of sodium oleate (150 mg/liter) the electrokinetic index of dolomite became increasingly negative with increasing H3PO4 addition. The index for apatite in the same experiments became less negative and approached zero at an orthophosphoric acid addition of 1.4 gm/liter. Flotation tests on sedimentary apatite and dolomite con-firmed that a flotation differential was indeed generated by H3PO4. On the basis of the measurements made, Borisov concluded that in orthophosphoric acid solution hydrogen ions rather than oleate ions adsorbed on apatite. Flotation of apatite was thus impossible while flotation of dolomite proceeded. Borisov assumed that dolomite does not adsorb hydrogen ions due to reaction 13 taking place. 2H*1 • CO" 2^—A H20 * C0 2 13 Borisov attributed the negative increase in electrokinetic index of dolomite to a gradual penetration of orthophosphate ions into the dolomite surface in the absence of oleate ions. The value of Borisov's data i n interpreting the selective depression of apatite by phosphate ions i n the Cominco process was 17 minimized by several factors: (1) Borisov used low concentrations of phosphate ion. (2) The system pH was varied with changes in phosphate ion level. (3) The pH levels used were low. (4) Only one level of sodium oleate addition was used in the presence of soluble phosphate. (5) Isoelectric points were not given. (6) Only index values and not signed zeta potentials were used. 22 Somasundaran investigated the effect of soaking time on the zeta potential of apatite. The effects of OH**1, • 2 - l Ca , F and orthophosphate ions on the zeta potential of apatite were also discussed. The f i n a l isoelectric point of apatite was found to be 7.0 from pH change observations and 6.0 by streaming potential measurements. Hydrogen, hydroxyl and orthophosphate ions were found to be strongly potential determining. Appendix I shows zeta potential curves taken from 4 Borisov*s early work and Somasundaran. 1-D Scope of Present Investigation The summary of previous work indicates a lack of agreement between authors on the mechanism by which orthophosphoric acid depresses apatite in oleate flotation. Extensive losses of phosphate ion from flotation solutions have been attributed to precipitation by calcium ions supplied through calcite and gypsum dissolution. Other possible reactions have been inadequately investigated. In addition, no efforts have been successful in circumventing these losses. The b r i t t l e froths encountered in the oleic acid flotation of calcite and dolomite from apatite have not been mentioned in the literature. It i s conceivable that oleic acid consumption could be decreased through a better understanding of solution properties. Therefore a study was undertaken to evaluate the adsorption of oleic acid on apatite and dolomite in the presence and i n the absence of orthophosphate ions. Simultaneously, possible reactions between these ions and calcite, dolomite and apatite were also looked at. B r i t t l e froths encountered in the practical flotation system were analyzed. 1-E Experimental Methods The flotation characteristics of pure apatite and pure dolomite were studied in solutions containing oleic acid in the presence and absence of orthophosphate ions. The results were evaluated using weights of dried products. The adsorption of oleic acid on apatite and dolomite i n the presence and absence of orthophosphate ions was studied using two infrared techniques. The results from this work were correlated with zeta potential determinations made by the electrophoretic mobility technique. The products of reaction of orthophosphate ions with apatite, dolomite and gypsum were investigated by infrared and 19 chemical methods. Surface tensions were determined by the drop volume method to correlate with solution frothing. Frothing effectiveness was evaluated visually. 2 ANALYTICAL METHODS 2-A Surface Analysis of Minerals by Infrared Spectroscopy Infrared spectroscopy i s a sensitive method of identi-fying compounds. It is only recently that this powerful method has been applied to the elucidation of problems involving natural minerals. The advent of internal reflection techniques has made possible the determination of surface spectra for highly opaque surfaces 2 3. The spectrum of a mineral i s a graphical representation of the proportion of energy absorbed from an infrared beam at any given frequency. A l l minerals have a definite spectrum depending only upon internal bonding. The presence of energy absorptions at anomalous frequencies in the spectrum of a mineral w i l l there-fore indicate the presence of an impurity i n the mineral or on i t s surface. This idea has been successfully applied in studying the adsorption of surface active compounds on minerals and metals by 24 25 15 14 Poling and Leja , Scowen and Leja * A. S. Peck , R. G. Knubovec 29 and M. J. D. Low . The lower limit of detection for infrared spectroscopy -8 26 i s near 10" moles of absorbing material. Kirkland mentions a sample of 10 micrograms being used to obtain absorption curves. Simple calculations based on a molecular cross-sectional area of 25A*2 would indicate that a compact vertical monolayer of molecules 2 over 15 cm would be required for detection of a surfactant. Alternatively, i f a mineral powder with an adsorbed surfactant were scanned for surfactant detection the mineral particles would have to be in the order of one micron size based on a cubic mineral' nominal surface area. The quantity of mineral used for analysis i s in the order of 0.75 mg. Reflection techniques have an advantage over transmission techniques i n that the ratio of adsorbate to adsorbent traversed by the infrared radiation can be readily controlled for the former methods. 2-B Reference Spectra Reference spectra of minerals, organic compounds and inorganic compounds were required for the infrared study of oleic acid adsorption and orthophosphate ion reactions. The spectra of calcite, calcium oleate, oleic acid and sodium oleate were found 20 in the work of Feck on infrared studies of oleic acid adsorption on fluorite, barite and calcite. The spectra of dolomite, calcite, gypsum and dicalcium phosphate were found in a publication by 27 Hunt, Wisherd and Bonham . The author checked the spectrum of each material against those available and contributed the spectra of fluorapatite, hydroxyapatite, a mixed fluor-hydroxyapatite, dicalcium phosphate dihydrate, tricalcium phosphate and primary calcium phosphate monohydrate. The spectra of materials used in this study are found 21 in Appendix II. Table 5 l i s t s the main absorption peaks of minerals and inorganic compounds encountered in this study. Table 5 Main Infrared Absorptions of Minerals and Inorganic Compounds In Wavenumbers V - very S, M, W - strong, moderate, weak S, B - sharp, broad Fluorapatite 1090 1050 960 600 570 455 Ca 5(PG 4) 3F SS VSB MS VSS VSS W Hydroxyapa t i te 3400 1450 1410 1090 1030 955 867 Ca 5(P0 4) 3GH MB MS MS S VSB MS WS Dolomite 1400 940 862 710 CaMg(C03)2 SB WB VSS SS Gypsum 3530 3395 1680 1617 1110 665 598 CaS04* 2H20 MB MB MS MS VSS MS MS Brushite 3535 3482 3275 3150 1650 1212 1132 CaHP04» 2H20 MS MS WB WB SS MS VSS 1058 984 872 782 655 578 528 VSS SS MS MB WB MS SS CaHP04 3400 2800 2300 1630 1400 1350 1125 WVB WVB WVB WVB WB WB SS 1060 990 890 562-78 520 VSS SS SB SS MS Ca(H2P04)2.H20 3455 3000 2270 1650 1235 1150 1088 WS MVB WB WB SS MB VSS 860-80 665 568 500 MB WB WS SS Ca 3(P0 4) 2 1095 1030 960 860 605 565 455 SS VSS WS WB MS SS WB Ca(OH)2 3621 3400 1420-50 872 400 600 VSS WB VSB MS VS VB 603 562 455 SS SS MS 22 2-C Absorption Spectra Absorption spectra were obtained by the KBr pellet method. This method i s widely known and i s well explained in a short a r t i c l e 26 by Kirkland . Only details specific to this study w i l l be given here. The KBr suspensions pressed into pellets contained 0.5% of minus 10 micron mineral. Pellets for determining the spectra of calcium oleate and sodium oleate contained 0.1% of the respective compounds. Inorganic salt spectra were obtained from KBr pellets containing 0.3% to 0.5% of the salts. Pressing of the 400 mg pellets was done in a 16 mm evacuated die. One gram KBr sample suspensions were finely ground in a fused alumina mortar. Ground suspension was transferred to the die and the die was evacuated for a minimum of two minutes. 2 A pressing load of 45 tons/in was then applied for one minute. Vacuum and load were released and the pellet recovered. 2-D Internal Reflection Spectroscopy 23 Harrick i s the author of a well known paper on the application of total internal reflection to surface studies. Detailed explanation of the physics of the method w i l l not be given here. Two important features of total internal reflection as applied to surface studies are that energy losses in total re-flection are much less than in metallic reflection and the depth of penetration into the surface being studied i s controllable. Figures 5(a) and 5(b) ill u s t r a t e these points. 23 Total Internal Reflection after Harrick B O •H 4J O V at 100 i 80 _ 60 40 20 ± / / ± J L -Total reflection -Metallic reflection 0 20 40 60 80 90 6 degrees Figure 5(a) Reflection Coefficient versus Angle of Incidence CO a o u u •iH s 2.0 1.0 0 ± Prism Air 0 20 40 60 80 90 0 degrees Figure 5(b) Penetration Depth versus Angle of Incidence Figure 6 illustrates the path of infrared radiation traversing an infrared-transparent truncated prism. The total internal reflection of the radiation and the penetration of the radiation through the prism surfaces are shown. In an actual prism the two large surfaces of the prism are entirely bathed in reflected radiation. Radiation penetrates interface ^ during r e f l e c t i o n Figure 6 Total Internal Reflection in a Prism 3 EXPERIMENTAL 3-A Equipment 3-A-l Flotation Flotation tests were done using a Denver Model D-l flotation machine. A 250 gram capacity glass bowl was used, impeller speed was maintained at 1500 rpm. Low pressure a i r was supplied to the c e l l . Continuous pH monitoring was effected with a suspended combination electrode feeding a Beckman Zeromatic pH meter. 3-A-2 Frothing Frothing tests were conducted in a 200 ml pyrex vessel with a f r i t t e d glass bottom. A schematic diagram i s shown in Figure 7. o n I K r Froth layer Clamp Solution Valve Air supply Bleed valve Figure 7 Frothing Apparatus 3-A-3 Infrared A modified model 521 Perkin-Elmer spectrophotometer was used for obtaining infrared spectra'. Pellets in the sample beam were held in a stock sample holder. The reference beam passed through a variable attenuator or through a reference pellet. Figure 8 illustrates the method used for holding polished mineral plates for surface analysis. An even pressure to assure good contact between the prism and mineral specimen was required. In pressure application, an aluminum-coated glass 26 slide was used to minimize energy losses from the prism. The truncated infrared-transparent prism was KRS-5 a thallium bromide iodide salt. The dimensions of the largest prism surface were 20 mm by 50 mm. 3-A-4 Zeta Potentials Zeta potentials were obtained by conversion of electro-phoretic mobility measurements made using a Riddick Zetameter. A 10 cm lucite capillary c e l l with platinum cathode and molybdenum anode was used for a l l measurements. A check on the polarity of the potentials obtained was made using a glass c e l l with platinum electrodes. Allan head screws Hold down plate •Mirror Prism Mineral plate Holder body Mounting pin holes Figure 8 Plate Sample Holder 3-A-5 Surface Tensions Surface tension determinations were made by the drop 27 volume method. A glass micrometer syringe was employed. 3-B Infrared Spectra of Materials Infrared spectra of materials used i n the study are given i n Appendix II. 3-B-l Minerals Apatite from three sources was used. The main sample was crystalline Durango apatite from Mexico. Flotation was done on a green apatite obtained from the Industrial Minerals Division of the Mines Branch in Ottawa. A portion of a cow's tooth was analyzed to obtain a reference spectrum of hydroxyapatite. A sample of crystalline gypsum (selenite) from Utah was purchased through Eckert Minerals i n Denver. The gypsum was perfectly transparent. Dolomite was obtained by the author from a recrystallized section of the Kootenay limestone formation near Crawford,Bay, B. C. The dolomite was similar to a coarsely polycrystalline white marble. Scattered grains of s i l i c a t e impurity were noted. Table 6 contains chemical analyses of minerals used i n the study. X-ray diffraction patterns of the minerals were obtained to confirm their identities. Mineral plates and powders were used i n the study. The plates and powders were prepared from the same specimens. Plates were cut using a diamond saw with care being taken to maintain parallelism of the main faces. One face of each plate was polished to a mirror fi n i s h by standard grinding and lapping techniques. Table 6 Mineral Analyses Mineral %CaO %MgO » 2 2 s %F %InSol. %S Durango Apatite 54. 7 38.2 2.8 0.1 0 Ottawa Apatite 52.5 36.2 3.6 1.7 .09 Theoretical Fluorapatite 55.5 42.1 3.8 0 Kansas Calcite 50.0* 0.3 Theoretical Calcite 56.1 Kootenay Dolomite 30.0 20.4 4.4 Theoretical Dolomite 30.3 21.9 Utah Gypsum 32.0 Theoretical Gypsum 32.6 *low 29 Final polishing was done using only water and alumina g r i t . Polyethylene gloves were worn under s i l k gloves to avoid contact of specimens with the fingers. Mineral powder was obtained by grinding specimens in a ceramic mortar and pestle. Final grinding was done in a hardened-steel Bueller vibrating m i l l . The powders were reduced to minus 10 microns as indicated by Bahco analysis. Spectrographs analyses indicated the dolomite powder to be a pure dolomite with 2% s i l i c o n content. The apatite powder was found to contain 1*7% combined rare earth elements as the major impurity. 3-B-2 Organic Chemicals Oleic acid used in adsorption tests was lot G-l-A - 90% pure acid obtained from the Hormel Institute at the University of Minnesota. Sodium oleate and calcium oleate were prepared by standard chemical means. Calcium oleate dihydrate precipitated from solution was dehydrated at 105°C. The dehydrated salt was leached with cold methanol to remove co-precipitated oleic acid. 3-B-3 Inorganic Chemicals Inorganic chemicals were C. P. grade from commercial sources. Dicalcium phosphate dihydrate was prepared by pre-cipitation from solution. A confirmation of the precipitate identity was obtained by x-ray diffraction. Sodium hydroxide used was analyzed by spectrographs means and was found to be pure. o 30 3-C Procedures 3-C-l Flotation Tests Four series of flotation tests were completed. The general procedures and flotation techniques were standard. A 200 gram sample of mineral was floated using U. S. P. oleic acid as collector-frother. Atmospheric air was used as flotation gas in a l l but one flotation test. Control of pH was effected using 1% solutions of NaOH or l^SO^. For convenience of collector addition oleic acid was converted to sodium oleate solution containing 5 gm/liter oleic acid equivalent. Phosphate ion was supplied i n flotation solutions by neutralization of H3PO4 with NaOH. A head solution containing 20 gm/liter H3PO4 equivalent neutralized to pH 5.0 was used as one half of the flotation solution volume. D i s t i l l e d water and NaOH solution for f i n a l pH adjustment f i l l e d the c e l l to the operating level of 1100 ml. The flotation solution contained 7.2 gm/liter P 2O 5. Conditioning times were standardized at 5 minutes for i n i t i a l wetting of minerals before collector addition and 2 minutes between collector addition and flotation. Flotation was carried out unti l the froth became barren. From three to six consecutive floats were pulled i n each test. The effects of sulfate ion and orthophosphate ions on the flotation of apatite and dolomite were determined at pH 5.5 to 6.0. The influence of particle size on the collector consumption 31 of dolomite was determined at pH 6.0 and pH 8.5 i n the presence of (1) sulfate ions and (2) orthophosphate ions. A single test was done to determine i f CO^ depressed dolomite flotation. 3-C-2 Frothing Tests Frothing tests were carried out by visual observation of the volume and s t a b i l i t y of froth obtained in bubbling compressed air through solutions. Figure 7 is a schematic diagram of the simple frothing apparatus. Frothing tests were done in the absence and presence of oleic acid, sodium lauryl sulfate, dolomite and apatite. A l l tests were at pH 5.5 i n 7.2 gm/liter ?2°5 solution. 3-C-3 Oleic Acid Adsorption Tests Six series of oleic acid adsorption tests were done. The procedure followed was to expose the mineral specimens to the desired solutions for 30 minutes, remove the specimen, rinse, dry in vacuo and mount for infrared scanning. Solutions were prepared with double d i s t i l l e d water using the high purity reagents listed under materials. Glassware was degreased with chromic acid cleaning solution. Specimens were handled with degreased tongs and spatulas. Drying was carried out i n a vacuum dessicator over s i l i c a gel. Mineral plate specimens were removed from solution only after the solution surface was overflowed to remove any impurities. Powder specimens were fi l t e r e d using a millipore f i l t e r with 0.3^1 cellulose acetate media. Mineral plates (20 mm x 50 mm x 2 mm) were immersed in 50 ml solutions. Powder samples weighing 50 mg were suspended i n 20 ml solutions. 32 Infrared spectra of plates and pellets were obtained as outlined i n sections 2C and 2D. Differential spectra were obtained for powder samples. Interpretation of the spectra was by qualitative comparison with the reference spectra i n Appendix II. The adsorption of oleic acid onto dolomite and apatite from 200 mg/liter solutions in the pH interval 5.5 to 9.5 was investigated. The effects of 7.2 gm/liter P2O5 as orthophosphate on adsorption were studied. 3-C-4 Orthophosphate Reactions with Solids Specimens of apatite, dolomite, gypsum and CaHPO^ were exposed to solutions containing 7.2 gm/liter F 20^ i n t n e ^ o r m °^ orthophosphate ions. Each material was subsequently scanned to determine i f changes had occured in the infrared spectrum. Both powders and plates were used. Solution volumes and sample sizes were identical to those used i n adsorption tests outlined in 3-C-3. Exposure times of 30 minutes to 65 hours were employed for apatite and dolomite. Gypsum was exposed for 20 minutes, 30 minutes and 2 hours. One gypsum plate was exposed to a solution containing 10 gm/liter Na 2S0 4 and 7.2 gm/liter p 2°5* T h e e x P o s u r e times for CaHPO^ were from 2 hours to 92 hours. 3-C-5 Sulphate Ion - Orthophosphate Ions - Gypsum Reaction Series 1 Four gram samples of C. P. gypsum powder were contacted for 3 hours with 120 ml orthophosphoric acid solutions containing from 1.2 to 8.2 gm/liter P2°5* Solution pH ranged between 1.5 and 2.0. The solutions were analyzed for phosphorus and sulfur. Series 2 Samples of Utah gypsum, 5 gm each, were ground to minus 200 mesh and suspended in 500 ml solutions containing from 0 to 4 gms of sodium sulfate and 7.2 gm/liter P 20 s at pH 5.5. The suspensions were maintained at 25°C and stirred for 30 minutes. The solids were then f i l t e r e d out, washed and dried. Solutions and gypsum residues were both analyzed for phosphorus by standard methods. The solutions were analyzed for total sulfur. 3-C-6 Zeta Potentials The effects of pH, oleic acid and orthophosphate ions on the zeta potentials of apatite and dolomite were determined. A 2 gm sample of mineral powder was suspended in 100 ml solutions. Solution pH was adjusted with NaOH or HC1. Measurements were obtained after a five minute time interval. Oleic acid was added to solutions after mineral additions. A pH check was run on a l l samples after determinations had been made. The effects of oleic acid from 0 to 215 mg/liter, pH from 5.0 to 10.0 and orthophosphate from 0 to 7.2 gm/liter P 20 s were investigated. 3-C-7 Surface Tension Measurements Surface tensions were determined for d i s t i l l e d water and solutions containing orthophosphate ions and oleic acid. Ortho-phosphate content was constant at 7.2 gm/liter P ?0 S. Oleic acid 34 levels of 50 and 100 mg/liter were employed. Solution pH's were 6.0 and 8.0. 4 RESULTS 4-A Flotation Tests A-A-l Apatite and Dolomite Flotation Tests Ottawa apatite and Kootenay dolomite floated well at pH 6.0 i n the presence of sulfate ion. Dolomite floated in the presence of orthophosphate at pH 6.0 but apatite did not. Complete depression of apatite by orthophosphate ions was observed up to pH 8.2. Figures 9(a) and 9(b) are graphical presentations of test results. 4-A-2 The Effects of Particle Size on Dolomite Flotation Figures 10 and 11 are graphical illustrations of the flotation recovery of dolomite versus particle size. Collector addition in dolomite flotation was minimum per unit of recovery between 100 and 150 mesh sizes in the presence of only sulfate ion at both pH 6.0 and 8.5. In the presence of 7.2 gm/liter ^2^5 at pH 6.0 collector addition was again minimum at particle sizes between 100 and 150 mesh. At pH 8.5 a strong depressing effect was encountered and collector addition per unti of recovery de-creased with decreasing particle size. Froths formed at pH 6.0 were observed to collapse when a total collector addition between 50 and 60 mg/liter had been reached. This collapse was pronounced in the presence of fi^SO^. A total addition of 1.7 gm of H2SO4 was made during flotation at pH 6.0 Oleic Acid Added mg/liter Figure 9(a) Flotation of Apatite and Dolomite at pH 8.5 Figure 9(a) (b) Legend 0 100 200 300 400 500 600 700 Oleic Acid Added mg/liter Figure 9(b) Flotation of Apatite and Dolomite at pH 6.0 36 -65 + 100 -100 • 150 -150 • 200 -200 • 270 -270 • 400 Mesh size Figure 10 Dolomite Flotation o x LEGEND H 2S0 4 H 3P0 4 Oleic Acid Total 100 80 J u v > o o e> cn 4.0 7. 7 2.10 0.1 ) ,S 4-E Zeta Potential Measurements Numerical results are presented in Appendix III. Results are shown graphically in Figures 19, 20 and 21. The zeta potential curve of apatite i n water shows a minimum near pH 6. 7 indicating that the isoelectric point i s near this value. An anomalous rise in the positive value above the minimum for apatite was noted. The curve obtained conforms with 7 22 Borisov's results . Somasundaran obtained a more definitive curve indicating the isoelectric point to be at pH 6.0. Analysis of Figure 20 reveals that in 0. 72 gm/liter P2O5 orthophosphate solution the zeta potentials of apatite and dolomite are 27 and 26 mi l l i v o l t s lower than in water at the same pH. An increase in ?2°5 t o 7*2 gm/liter decreases the zeta potential another 6 mil l i v o l t s for apatite and 11 mi l l i v o l t s for dolomite. Orthophosphate ions are seen to be strongly potential determining for apatite and dolomite between pH 6.0 and pH 9.0. Figure 21 reveals that oleic acid i s specifically adsorbing on dolomite but not on apatite in 7.2 gm/liter P2O5 solution at pH 6.0. Oleate ion i s potential determining for apatite at pH 6.0 i n water. 4-F Surface Tension Measurements Oleic acid at 50 mg/liter and 100 mg/liter lowered the surface tension of solutions containing orthophosphate ions. The lowering effect would be compatible with normal frother behavior. Surface tensions of 51 and 38 dynes/cm were found. Table 10 contains the measured surface tensions. 49 Table 10 Surface Tension Values P2O5 Oleic Acid gm/liter mg/liter pH dynes/cm 0 0 natural 71.5 7.2 0 6.0 70.5 7.2 0 8.0 56.0 7.2 100 8.0 23.8 7.2 100 6.0 38.0 7.2 50 6.0 51.3 Drops were observed to be slow to reach equilibrium i n the presence of orthophosphate. 5 INTERPRETATION OF RESULTS 5-A Flotation and Frothing 5-A-1 The Flotation of Dolomite and Apatite Apatite was shown to be strongly depressed during flotation with oleic acid between pH 5.5 and 8.0 i n the presence of ortho-phosphate ions. The apatite specimens were analyzed for sulfur and were found to be gypsum-free. It can be concluded that contrary to Mitzmager* s observations 1 1 gypsum plays no essential role in the depression of apatite flotation by orthophosphate ions. Dolomite was readily floated with oleic acid at pH 6.0 and at pH 8.5 in the presence and absence of orthophosphate ions. Flotation efficiency of dolomite was reduced by low pH contrary to 11 observations made by Mitzmager et a l for calcite flotation. The 50 Zeta Potentials of Apatite and Dolomite > s 4 J a ,2 CS u 0) [S3 + 40 _ + 20 0 --20 -•40 - - " X * • Dolomite in water — Apatite i n water Apatite i n 7*2 gm/liter P 20 5 1 1 1 1 1 1 r 5.0 6.0 7.0 8.0 9.0 10.0 11.0 pH Figure 19 Zeta Potentials versus pH 4 J c u Q 01 + 40 -i •20 0 -•20 -• 40 pH 6.0 Figure 20 Zeta Potentials versus P2O5 in Solution > s r-l t» •H U c tt) u o cu 0) + 40 -, +20 pH 6.0 _ „ X-Dolomite in water -40 Apatite in ?2°5 Solution Apatite in water Dolomite in P2O5 Solution 40 I 80 = X -~ ~ - —d - _ I 1 1 1 120 160 200 240 Oleic Acid mg/liter Figure 21 Zeta Potentials versus Oleic Acid Concentration in Water and i n 7.2 gm/liter phosphate solution slower response of dolomite than calcite to acid attack could account for the discrepancy. Orthophosphate ion depressed the flotation of dolomite to some extent at pH 6.0 and markedly at pH 8.5. The addition of oleic acid per unit of recovery in the flotation of dolomite i s seen to be lowest at a particle size near 100 mesh (150 microns). Particle size alone would not account for the high oleic acid consumption noted i n the industrial process. 5-A-2 Frothing Dif f i c u l t i e s Encountered i n Flotation Frothing during the flotation of dolomite was poor at pH 6.0. The formation of undissociated molecule-ion complexes of oleic acid and calcium oleate could be responsible, since a condensed film at the air-solution interface could then be formed. Frothing tests indicated that with only oleic acid i n orthophosphate solution at pH 6.0 a stable froth does not form. The role of mineral particles in stabilizing the froth during the separation of dolomite from apatite would account for the observed thinning of froth at the end of flotation in the re-cleaning step. Efforts to obtain a maximum cleanup lead to increasing oleic acid additions in order to obtain a stable froth. Instead, the selection of an efficient and compatible frothing compound for the system should be attempted and could result in much lower oleic acid consumptions. The observation that orthophosphate solutions containing oleic acid do not froth at pH 6.0 would indicate a lack of suitable molecular species at the air-solution interface. Figure 15p illustrates frothing obtained in the flotation of dolomite at pH 6.0 and pH 8.0. 52 Figure 15p(b) Frothing During Flotation of Dolomite With Oleic Acid at pH 8.0 53 The same froth conditions were obtained i n solutions containing H 3 P 0 4 " N a 0 H » a » d only H2 S 04* T n e froth s t a b i l i t y i s therefore a function of pH, and of ionic composition of solution (Ca* 2 vs Na* 1). Sodium oleate is formed at basic pH* s i n the system and i t can be shown that the ionized form i s a strong frothing agent. Above pH 10.0 maximum froth i s obtained but no flotation of calcium minerals occurs. Oleic acid i n the undissociated form adsorbs at the air-solution interface and acts as a frother. At pH 6.0 the system, therefore, contains two potential frothing agents; unionized oleic acid and oleate ions. At pH 6.0 oleic acid and oleate ions would be expected to form molecular aggregates giving a more con-densed film than either species alone. Air bubbles would not be stabilized due to the interfacial film being too r i g i d and prone to rupture with bubble vibrations. It was noted that a solution containing 50 mg/liter oleic acid at pH 6.0 was a hazy blue indi-cating a colloidal suspension. The same solution at pH 8.0 was clear. Surface tension measurements of oleic acid solutions at pH 6.0 indicated values of 38 dynes/cm for 100 mg/liter oleic acid in the presence of 7.2 gm/liter ?2°5' I t : w a s o b s e r v e d that drops were extremely slow to equilibrate during measurements by the drop volume method. The surface tension would slowly decrease over several minutes and f i n a l l y a drop would f a l l . Gaudin^ discusses the problem of slow formation of sodium stearate layers on newly created air-solution interface. Up to 10 seconds were required for a measurable change in surface tension to occur. In one minute the surface tension was lowered from 71.5 dynes/cm to 58 dynes/cm. In one hour the value obtained was 36 dynes/cm. The diffusion of sodium stearate to the inter-face was obviously slow. Observations indicate that oleic acid exhibits similar behavior. A contributing factor to unstable froths would be the short time which a bubble-solution interface exists in the pulp. The slow migration of oleic acid to the interface would be explained by the existence of molecular aggregates and their insoluble nature. 55 5-B Loss of Orthophosphate Of the major mineral components present in the phosphate flotation system, gypsum i s the only mineral to remove significant quantities of orthophosphate ions from solution. The presence of 8 gm/liter Na 2S0 4 i s seen to suppress the reaction between gypsum and orthophosphate ions. The reaction taking place would be: -2 -2 CaSO '2H„0 • HPO — * CaHPO »2H 0 * SO 13 4 2 4 ^ 4 2 4 Infrared spectra showed that the compound CaHPO^^H^O forms directly on the gypsum surface. It i s known that dicalcium phosphate dihydrate and gypsum are isomorphous and have identical lattice parameters. The direct crystallization of CaHPO^'ZI^O on the gypsum surface has been 28 confirmed by Cecconi . Excess calcium ions i n a solution would also be pre-cipitated by reaction 6 Ca* 2 • HPO~2 4 2H20 ^ C aHPO^I^O 6 Gypsum particles could act as nuclei for the crystallization of CaHP04» 2H20. Calculated and actual sulfate ion concentrations required to suppress the fixation of soluble phosphate by gypsum agreed within experimental limits. The calculated value i s 9.2 gm/liter _2 SO^ the actual quantity used in a successful test was 6.3 _2 gm/liter SO^ . The common ion effect was therefore concluded to be the mechanism involved i n suppressing the reaction. 5-C Adsorption Studies 5-C-l Oleic Acid Adsorption and Correlation with Zeta Potentials The infrared study of oleic acid adsorption on apatite has been correlated qualitatively with the results of flotation tests and zeta potential determinations. Apatite recovery was high in water over the pH range 6.0 to 8.5. Oleic acid was shown to adsorb well on apatite i n this pH range with a maximum calcium oleate formation between pH 6.5 and 7.5. A "dip" i n the zeta potential versus pH curve exists for apatite at pH 6.7 i n water and correlates with the isoelectric point for apatite as determined by Somasundaran. The isoelectric point i s considered to coincide 4 with maximum collector adsorption for most minerals . It i s interesting that the oleic acid flotation of phosphate ores i s best carried out at a pH between 8 and 10. Ionic species i n an ore pulp could shift the pH of maximum collector adsorption upward and in-crease flotation recovery at higher pH's. Equally probable i s that with more sodium oleate and oleate ions as frother species at pH 8.5 than at pH 6.0 ai r bubbles are rapidly stabilized and a more rapid attachment of hydrophobic particles to bubbles occurs. Figures 20 and 21 ill u s t r a t e that oleate ion i s not strongly potential determining for dolomite i n water at pH 6.0 i n contrast to the effect on apatite. It would appear that H*1 i s potential determining for dolomite under these conditions and ^18 H33°2^ ^ i s P ° t e n t i a l determining for apatite. The opposite situation holds for apatite i n the presence of orthophosphate ions. This of course correlates well with observed flotation behavior. 57 Zeta potential measurements indicate an anomalous rise i n potential of apatite above the "dip" corresponding to the isoelectric point. The presence of a potential determining multivalent cation in the Stern layer would be required to explain this behavior. Divalent calcium ions or another cationic complex could be responsible as similar curves have been obtained by others for fluorite, apatite, 4 dolomite and calcite . Spectrographs analyses have revealed a combined content of Y, La, Nd and Ce of 1.7% i n the apatite sample used. These elements form multivalent cations which would be strongly potential determining i f concentrated i n the double layer during mineral dissolution. The Hardy - Schulze rule predicts that the concentration required to effect the same change i n zeta potential 1 2 3 decreases for multivalent ions in the ratio 10 : 10 : 10 where the indices represent the valence of the ion. 5-D Depressing of Apatite by Orthophosphate Ions The specific depressing effect on apatite flotation of orthophosphate i s believed to arise from a combination of electro-kinetic effects and physical interference of adsorbed orthophosphate species on collector adsorption and flotation. It has been shown that less oleic acid i s adsorbed on apatite i n the presence of orthophosphates. The large negative -1 -2 zeta potential induced by specific adsorption of H2PO4 , HPO^ and -3 P0^ on apatite could account for this. Under the same conditions adsorption of oleic acid proceeds on dolomite and calcite with less hindrance. Figure 20 shows the zeta potential of dolomite to be very similar to that of apatite i n orthophosphate solution at pH 6.0. 58 The neutralization of H* through acid attack on dolomite could be increasing the effective pH near the particle surfaces. Oleic acid would therefore be more readily adsorbed on dolomite. A contributing factor to the action of orthophosphates could be due to the formation of a strongly hydrogen-bonded layer of orthophosphates around the particles. This layer would be negatively charged and hydrophilic. The effects on flotation would be twofold. Collector adsorption would be slowed and the layer would interfere with particle-bubble interaction. The depressing action of Zn(0H>2 on sphalerite has been attributed to 4 a similar effect by Livshitz and Idelson . It has been observed that flotation of depressed apatite w i l l occur after a fifteen minute conditioning time. Kinetic factors i n the depressing action are therefore indicated. Tests have shown that the calcium minerals apatite and brushite which are not prone to acid attack are strongly depressed at pH 6.0. The minerals calcite and dolomite which are prone to acid attack are not depressed at pH 6.0. At pH 8.0 a l l the minerals are depressed. It would therefore appear that mineral structure i s not a determining effect i n orthophosphate depressing action. Structure does affect the flotation properties of the minerals i n the usual way as adsorption site density i s an important property to flotation. Calculations of calcium (or magnesium) spacings in the lattices of apatite, brushite, calcite and dolomite did not indicate differences which could cause preferential depression due to steric factors alone. 59 Important facts which resulted i n the explanations given are: (1) The absence of solid reaction products on apatite and dolomite, exposed to orthophosphate solutions. (2) The zeta potential correlations showing ortho-phosphate ions to be potential determining for both minerals. (3) The observation that orthophosphate ions w i l l not reverse the adsorption of oleate by apatite but only prevent i t . (4) The strong hydrogen bonding exhibited by dibasic 21 and tribasic orthophosphates which would increase with de-creasing pH. (5) The observation that with decreased pH lower total Q orthophosphate ion i s required for apatite depression . (6) The greater depressing effect of orthophosphate ions on dolomite at pH 8.5 than at pH 6.0. (7) The known solubilization of dolomite through acid attack in the system with C0 2 evolution at pH 6.0. (8) The observed depressing of gypsum flotation i n orthophosphate solution 1 1. Under the given conditions gypsum is converted to brushite CaHPO..2H-0. 4 2 5-E Orthophosphate Requirements as a Function of Flotation Variables A corollary of the theory presented above i s that the level of orthophosphate ions necessary for apatite depression w i l l be lowest at the lowest pH tolerable i n a flotation system. This pH w i l l be determined by the mineral present which i s most reactive 60 to acid attack. A mineral mixture containing only dolomite could be floated at a much lower pH than one containing calcite with a similar loss of orthophosphate ions. Balancing the HPO^ present _2 i n solution with SO^ could enable lower pH levels to be used with equivalent or lowered orthophosphate loss. The introduction of a suitable frother and high speed flotation with minimum solids-solution contact time could give additional economies. 6 CONCLUSIONS Gypsum has been found to react with orthophosphate ions to form surface layers of CaHPO^^H^O. This reaction can be suppressed through the common ion effect by the presence of 6.3 gm/liter sulfate ion. The pH of maximum calcium oleate formation (through adsorption) on apatite coincides with the isoelectric point of pH 6.7. The depressing effects of orthophosphate ions on apatite -2 and dolomite are due to the adsorption of hydrogen bonding HPO^ H2PO4* and EjPO^ i n the Stern layer of the electrical double layer of the minerals. The hydrophilic barrier formed slows oleic acid and oleate ion penetration and interferes with f r u i t f u l particle-bubble interaction. The large negative zeta potential reduces collector adsorption. The specific action of orthophosphate ions i n depressing apatite more than dolomite i s due to selective solubility of dolomite. 61 The evolution of CO^ p a r t i a l l y removes the hydrogen-bonded ortho-phosphate barrier to allow rapid o l e i c acid adsorption. The neutralization reaction produces a higher l o c a l pH which results i n increased f l o t a t i o n of dolomite. The concentration of orthophosphate ions necessary to depress apatite i n the f l o t a t i o n of ores could be decreased by lowering the pH of the system and supplying sulfate ion to reduce -2 HPO^ ion loss through brushite formation. The consumption of f a t t y c o l l e c t o r i n the f l o t a t i o n of ores i n orthophosphate solution could be decreased by supplying a compatible frother which possesses strong frothing characteristics under a c i d i c conditions. The unstable froths observed may be due to either of two factors: f i l m condensation caused by o l e i c acid-oleate ion complex formation and low mobility of the frothing species. Suggested optimum conditions for the f l o t a t i o n of dolomite from apatite would include: 5 gm/liter P2°5 a n d 10 gm/liter -2 SO^ i n solution; f l o t a t i o n between pH 4.0 and 5.0; the inclusion of a strongly frothing compound i n the solution; and minimum contact time between solution and solids. 7 FURTHER WORK The application of the findings of this thesis to p r a c t i c a l f l o t a t i o n systems has yet to be undertaken. Effects of orthophosphate ion hydrogen-bonding on c o l l e c t o r 62 adsorption and particle-bubble interaction could be assessed by various methods. An NMR study using hydrogen isotopes could reveal the concentrations of ions at the s o l i d - s o l u t i o n interface. The d i f f u s i o n rates of o l e i c acid through a layer of orthophosphates could reveal possible k i n e t i c effects i n c o l l e c t o r adsorption. A determination of the k i n e t i c s of particle-bubble interactions would reveal any interferences due to the proposed hydrogen-bonded layer. Obtaining the s o l u b i l i t y versus pH and dissociation curve for o l e i c acid would be a challenging problem. The a v a i l a -b i l i t y of these data could ease interpretations of f l o t a t i o n studies. The effects of mineral impurities on zeta potential could be assessed and result i n explanation of the anomalous shape of the zeta potential curves obtained for more soluble minerals. 63 BIBLIOGRAPHY 1. Sauchelli, V., "The Origin and Processing of Phosphate Rock With Particular Reference to Beneficiation", The F e r t i l i z e r Society, No. 70, July (1962). 2. The British Sulfur Corporation Ltd., "A World Survey of Phosphate Deposits" 2nd Ed., Section II p. 51, The British Sulfur Corporation Ltd., London 1964. 3. Miche, R., "Differential Flotation of Calcite and Lime Phosphate", i n the Phosphate Ores of North Africa. Translation by Lauzon, P., March (1954). 4. Klassen, V. I., and Mokrousov, V. A., "An Introduction to the Theory of Flotation", translated by Leja, J., and Poling, G. W., pps. 247 - 259, 329, 391, 30 - 35, 427, 200 - 202, 304 - 346, 353, 359, 366 - 370, 376, Butterworths, London (1963). 5. Sutherland, K. L., and Wark, I. W., "Principles of Flotation", pps. 314 - 319, 335 - 338, Melbourne, Australian Institute of Mining and Metallurgy (Inc.), (1955). 6. Gaudin, A. M., "Flotation" pps. 465, 471 - 475, 185 - 187, 189, 230 - 231, 256 - 258, 335 - 336 2nd Edition, McGraw-Hill, 1957. 7. Borisov, V. M., "Prerequisite for Enrichment of D i f f i c u l t -to-Enrich Phosphorite Ore of Kara Tau Deposits", State Institute of Mines and Chemical Raw Materials (1956). 8. Borisov, V. M., "The Effect of Surface Electrokinetic Properties of Minerals on Their Flotation A b i l i t i e s " , Doklady Akad. Nauk SSSR, 1954, Volume XCIX, No. 3. 9. Cominco Patent Pending 10. Montecatini Inc., A Process for Enriching Phosphate Minerals. British Patent 859,155 July 3 1959; after Italian Patent July 8 (1958). 11. Mitzmager, A., Mizrahi, J., and Fischer, E., "Flotation of Calcite from Phosphate Slimes: Effect of Soluble A l k a l i Phosphate Salts", Institution of Mining and Metallurgy, Section C, September 10 (1966). 12. Hirsch, H. E., "Separation of Calcite and Dolomite from Douglas Creek Phosphate Rock", February 26, 1966 Cominco Internal Report - Confidential 64 13. Rootare, HiHare M., Deitz, Victor R., and Carpenter, Frank 6., "Solubility Product Phenomena i n Hydroxyapatite - Water Systems", Journal of Colloid Science 17, 179 - 206 (1962). 14. Chemical Rubber Publishing Co. "Handbook of Chemistry and Physics" 35th Edition. 15. Leja, J., Course Notes Mineral Engineering 471. 16. Knubovec, R. G., and Maslennikov, B. M., "Infrared Spectroscopy Studies of the Adsorption of Flotation Reagents on Mineral Surfaces", Translation by V. Ositis. 17. Stremovskii, L. I., The Flotation of Calcium Minerals i n Acid Media, Reference misplaced. 18. Petrova, I. V., Poboikova, E. G. and Tumanova, V. G., Dobychai Obogashch Rud. Tsevtn. Met., Nouchn. -Tekhn. Sb 1964(4) 42 (Russ.) CA Vol. 62 7426e. 19. Chernyi, L. M., Mechanism of The Action of Some Inorganic Acids i n the Flotation of Phosphate-Dolomite Ores in Acid Media, Khim Prom., No. 5 (1963) 341 - 4. 20. Peck, A. S., "Infrared Studies of Oleic Acid and Sodium Oleate Adsorption on Fluorite, Barite and Calcite", U.S. Bureau of Mines, Report of Investigations, 6202, (1963). 21. Chapman, A. C., and Thirlwell, L. E., "Spectra of Phosphorus Compounds - I, The Infrared Spectra of Orthophosphates" Structural Inorganic Chemistry, 3rd. Ed. 22. Somasundaran, P. "Zeta Potential of Apatite i n Aqueous Solutions and Its Change During Equilibration", Journal of Colloid and Interface Science, Vol. 27, No. 4, August(1968). 23. Harrick, N. J., "Total Internal Reflection and i t s Appli-cation to Surface Studies", New York Academy of Science, Volume 101 A r t i c l e 3, pp. 928 - 959, January 23, (1963). 24. Poling, G. W., Leja, J. and L i t t l e , L. H. "Xanthate Adsorption Studies Using Infrared Spectroscopy, II, Evaporated Lead Sulfide, Galena and Metallic Lead Substrates", Bull. Inst. Min. and Met., London, (1963). 25. Scowen, R. V. and Leja, J., "Spectrophotometric Studies on Surfactants II. Infrared Study of Adsorption from Solutions of Single and Mixed Surfantants on Copper Substrates", Canadian Journal of Chemistry, Volume 45,2829 (1967). Kirkland, J. J., "Quantitative Application of Potassium Bromide Disc Technique i n Infrared Spectroscopy", Analytical Chemistry, Volume 27, (1955). Hunt, John M., Wisherd, Mary P. and Bonham, Lawrence C., "Infrared Absorption Spectra of Minerals and Other Inorganic Compounds", Analytical Chemistry, Volume 22, page 1478, December (1950). Cecconi, S. and Martelli, M., Phosphatation of Soil II, Reactions of Calcium Sulfate With Phosphate Ions Agrochimica 8(3) 232 - 48 (1964) Ita l . Low, M. J. D., Inoue, H., Infrared Emission Spectra of Fatty Acids on Steel Surfaces. Canadian Journal of Chemistry, Vol. 43, p. 2047 - 2051, 1965. 66 APPENDIX I Appendix 1(a) Figure 1 Figure 2 20 50 100 200 500 1000 Conditioning Time, Hours Zeta Potentials by streaming after Somasundaran Zeta potential of apatite as a function of pH and time in 10~2 M KN03 1 i n i t i a l value 2 two weeks equilibration Change of zeta potential of apatite with time in 10"2 M KN03 for various i n i t i a l pH values. pH I n i t i a l Final 1 11.0 7.5 2 7.5 7.5 3 3.0 6.5 4 2.0 5.3 Appendix > \ • i \ 1 — v 2 -A-,0^ Figure 1 2 4 6 8 10 12 PH 1(b) Zeta Potentials of Calcite and Apatite after Borisov (1955) Influence of medium pH on the electrokinetic potentials of minerals. 1 Calcite 2 Apatite 3 Fluorite 67 APPENDIX I I 2000 1600 1200 800 400 Wavenumbers Appendix 11(a) Infrared Spectra of Apatites Transmission Spectra 68 o 2000 1600 1200 800 400 Wavenumbers Appendix 11(b) Infrared Spectra of Mineral Powder Samples Transmission Spectra Appendix 11(c) Infrared Spectra of C. P. Calcium Orthophosphate powders - Transmission Spectra 70 Wavenumbers Appendix 11(d) Infrared Spectrum of Oleic Acid Sodium Oleate and Calcium Oleate Transmission Spectra 71 APPENDIX III Zeta Potentials - Numerical Results Apatite Solution Composition pH Zeta Potential 1. NaOH - HC1 3.4 +19 5.4 +18 6.7 0 7.2 + 3.4 9.9 +10.8 11.1 +11.3 2. NaOH - HC1 pH 6.0 Oleic Acid mg/liter 26.9 +10.2 53.7 -16.0 80.6 -23.5 107.4 -29.0 214.8 -29.0 3. NaOH pH 6.0 H 3 P 0 / i gm/liter 1 -14.8 2 -15.0 5 -18.4 10 -21.0 4. NaOH - H 3P0 4 10 gm/liter £H 6.0 -21.0 7.0 -13.5 8.0 -11.3 9.0 - 9.4 5. NaOH - H 3P0 4 10 gm/liter Oleic Acid mg/liter pH 6.0 26.9 -13.2 53.7 -20.0 80.6 -18.0 107.4 -18.8 214.8 -13.0 72 Dolomite Solution Composition pH Zeta Potential 1. NaOH - HC1 5.4 + 19.0 6.0 + 18.6 7.0 *17.0 8.0 +23.0 10.0 +24.8 2. NaOH - HC1 pH 6.0 Oleic Acid mg/liter 26.9 53.8 107.6 215.2 •22.6 +25.0 + 17.0 + 19.0 3. NaOH pH 6.0 HoPO, gm/liter 0 +18.6 1 - 7.5 2 - 9.5 5 -14.8 10 -19.3 4. NaOH - H 3P0 4 10 gm/liter Oleic Acid mg/liter pH 6.0 0 -19.3 26.9 -29.6 53.8 -29.1 215.2 -34.0