UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Synthetic testing for flotation McLachlan, Charles Gordon 1925

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_1925_a7_m2_s9.pdf [ 72.15MB ]
Metadata
JSON: 831-1.0099564.json
JSON-LD: 831-1.0099564-ld.json
RDF/XML (Pretty): 831-1.0099564-rdf.xml
RDF/JSON: 831-1.0099564-rdf.json
Turtle: 831-1.0099564-turtle.txt
N-Triples: 831-1.0099564-rdf-ntriples.txt
Original Record: 831-1.0099564-source.json
Full Text
831-1.0099564-fulltext.txt
Citation
831-1.0099564.ris

Full Text

U.B.C. L l&HA RY  *^?£»VaatKmr^:^"- •*••••  SYNTHETIC TESTING FOR FLOTATION. by  Charles Gordon McLachlan  1 0  T.»  r4  j4t*JL A A+fn+Xim. t~ .  ./ *  LE3-37 » f^r* i—  SYNTHETIC TESTING FOR FLOTATION, by Charles Gordon MeLachlau  A Thesis submitted for the Degree of MASTER OF APPLIED SCIENCE in the Department of MINING AND METALLURGY  The University of British Columbia April 1925  i -"synthetic Testing for Flotation". Table of Contents. Prefatory Note -  pag  e jy  INTRODUCTION -  --  - page 1  PRELIMINARY WORK Consideration of the Importance of Selective Adsorption -- -..pag Development and Description of Surface Tension Instrument pag  e9 e 1 1  Method of Operating Instrument -  pag  e 1 8  Surface Tension Experiments with Pure Liquids and Solutions- pag  e 24  Surface Tension Experiments with Emulsions---page 25 Investigation of Methods for Estimating Adsorption from Emulsions pag  e 22  ADSORPTION TEST S WITH OLEIC ACID pag  e 35  ADSORPTION TESTS WITH ANILINE pag  e 3?  ADSORPTION TESTS WITH Cu S04 pag  e 43  ADSORPTION TEST S WITH LIME -  pag  THE RELATION BETWEEN THE CRYSTAL STRUCTURE OF MINERALS AND THEIR FLOTABILITY-- CONCLUSION _  pag  BIBLIOGRAPHY -  -  page 51 e 56  SUMMARY OF RESULTS AND CONCLUSIONS pag ACKNOWLEDGMENTS -  e 49  - - pag page^6  e 58 e 60 l  "Synthetic Testing for Flotation" i  i  Table of Contents (con»t)  PLATE I — Photograp h of Improved Surface Tension Instrument T  o face page 13  PLATE II—Photograph of Complete Set up of Surface Tension Instrument Opposit PLATE III—Detailed Drawing of Surface Tension Instrument M  e page 14  w  1  3  PLATE IV — Type s of Dishes for use witk Surface Tension Instrument - - - To face page 20 PLATE 7 — Grap h showing Decrease in Surface Tension of Aqueous Solutions with Increasing Concentration of Acetic Aeid "  "  PLATE 71—Photograph of a Mechanical Shaker " PLATE 711—Graph showing Amount of Oleic Aeid Adsorbed by Various Minerals -  M  " "  2 5 3 3  -" «  "  3 7  PLATE VIII-Graph showing Amount of Aniline Adsorbed by Various Minerals---" "  "  4 0  PLATE 1 2— Graph Showing Amount of Copper Adsorbed from Solution by Various Minerals '  • «  "  4 4  PLATE I — Grap h Showing Amount of Calcium Adsorbed from Solution by Various Minerals «  "  "  5 0  iii "Synthetic Testin g for Flotation"  Table of Contents (eon»t )  Table 1 Table 2  Langmuir's Contact Angle Results - -pag Surface Tension Results for Liquids pag  e 2 e 24  Table 3  Surface Tension Results for Petrolatum Emulsions pag  e 26  Table 4 Table 5 Table 6 Table 7  Table 8  Surface Tension Results for Oleic Acid Emulsions pag Results showing Accuracy of Method for Recovering-Oleic Acid pag Results obtained using Aiiiline as a "Collector" pag Amount of Copper contained in the various Minerals used as Adsorbents pag -Physical Properties of Floatable and Hon floatable Minerals t  Pig. 1 - An Ideal Surface Tension Curve for Estimating Adsorption - -  e 28 e 34 e 41  e 43 o face page 51 - - - page 10  Pig. 2 - Space Lattice Diagram of the Oalcite Group Opposite page Pig. 3 - Space Lattice Diagram of ChalFig. 4 Pig. 5 Pig. 6 -  copyrite —  "  Space Lattice Diagram of Pyrite " Space Lattice Diagram of Sphalerite " Space Lattice Diagram of Molybdenite"  52 52 53 53 '53 54  Pig. 7 Space Lattiee Diagram of Cuprite " Fig. 8 - Space Lattice Diagram of Cupric Oxide » Fig» 9 — Space Lattice Diagram of Zincite  54  Fig. 10 — Spac e Lattice Diagram of Arsenic Trioxide- - - -  55  55  1111  PREFATORY NOTE  All the experiments described in the following pages were either designed to bring out some particular point in connection with the flotation process or were intended to establish certain methods which might be used to do so« This has resulted in quite a number of interesting points having been brought out which do not seem to have a direet bearing on the present research, and for this reason they have received only very cursory treatment. Two examples of such points will be found in the section dealing with surface tension in which only a very brief discussion is given as to why, the surface tension instrument, which has been developed, seemed to give consistently high results, and why surface tension measurements made by the direct method on certain types of emulsions give discordant values* A continuous method of numbering the references in the text has been adopted in order; to "avoid confusion when quoting for the second time one of several papers by the same author. I f a certain paper has been referred to more than once the reference number given refers back to the point at which the reference was first cited and the number of the page indicates the particular portion of this reference on which the new point cited will be found, e«g» lee eit p. 22 Ref. Ho. 8 means that Ref. No. 8 gives the title of the article referred to and the point in question will be found on page 22 of that article.  SYNTHETIC TESTING FOR FLOTATION. Introduction.  Synthesis as defined by the New Oxford Dictionary "is the putting together of parts or elements so as to make up a complex whole". The process of flotation is most certainly "a complex whole" and the object of the present research has been to try to settle definitely certain points connected with it. No attempt will be made here to give a historical sketch of the development of flotation. Thi s has been done admirably by T.A. Rickard in his book "Concentration 1 by Flotation1*. Neither will any attempt be made to summarise the large number of articles dealing with the process which have appeared in the technical journals and periodicals during the last eight or nine years. Fro m these articles three papers have been selected which seem of special interest and preliminary discussion will be limited to them.  1. "Concentratio n by Flotation", Chapt. I, John Wiley & Sons (1921) also Min. & Sci. .tress, March 17th db  2.  The first of these three is due to Langmuir2 who in 1919 presented a paper before the Faraday Society on "The Mechanism of the Surface Phenomena of Flotation". In his paper he showed that monomolecular films of very great stability could be produced on solids with oleic acid, but tfeat pure paraffin oil, though spreading readily on a clean dry surface, was easily displaced by water. H e also showed that the contact angle made by a large drop of water on a solid was greater if the water contained oleic acid. Thi s is illustrated in Table I reproduced from his work. Table I. Solid  Clean H 2 0 h e height of drop in m.m.  Mica Quartz Glass Platinum Caleite Sphalerite Galena  .9 mm -  2.1 2.9 5.1 5.6 5.7  Water Saturated with Oleic Acid  9 s Contact angle 18° 4  *°.  6  ^n 70° 820 86°  h  e  .9 mm 1.2 .. 1.5 ,  24° 51° 420 72° 2.45 •• 82° 2.75 3.0 92° 5.55 " 106°  2. Trans . Farada y Society Vol. 25, Part III p. 64-74. Abstracted Mining & Scientific Press. Vol. 121:(1920) p. 915.  3.  The equation which Langmuir used to obtain the contact angle from the height of the drop was  h - a \J 2 sin (£ 6) where a - 2  V i  n which  Y r the surfaee tension of the liquid forming the drop d - the density of the liquid. In the above calculations he took V for clean water as 72.8 dynes and for water saturated with oleic acid as 42.8 dynes. A statement made by Langmuir at the conclusion of his article seems of considerable interest. I t is this, "The results indicate that the selective action by which substanees like galena are separated from qaiartz and oalcite is dependent upon the oontact angle formed by the oiled surfaces rather than by any selective tendency for the oil to be taken up by some minerals more than others". The foregoing conclusion together with the results contained in the table will be referred to again later. The second of these three papers appeared in 1922 when Taggart and "*audin^ published the results of a number of experiments. Thei r work will be found in the *  J. Trans . A.I.M.S. Vol. 68 (1923J p. 479.  4.  Transactions of the American Institute of Mining and Metallurgy under the title of "Surface Tension and Absorption Phenomena in Flotation". Thei r investigations were based on the use of an instrument for measuring surface tension which had been devised by Fahrenwald.4 Thi  s  instrument they found gave satisfactory results on solutions but would not give consistent readings when employed for measuring the surface tension of oil-in-water emulsions which contained insoluble oils known to be of value in flotation. The y accordingly filtered their emulsions after carrying out adsorption testa and before measuring the surface tension of the solutions. Unfortun ately they failed to realise that when they filtered their emulsions they permitted a second adsorption to take place, the which resulted in practically allAsolutions they tested giving the same result. Apar t from this error their other results are extremely interesting. Particularl y those which deal with maximum foaming^ and the distribution of organic reagents.  4. Minin g & Scientific press. Vol . 123. (1921) p. 227. J>. Quit e recently a paper has been published by 0. Batseh in "Kolloid Chem. Beihefte" Vol. 20 (Oct. 1?24) p. 1-49 entitled "Uber Schaumsysteme". Thi s paper bears a very marked similarity to that of Taggart & Qaudin and seems to be merely a repetition of their work.  The third paper to which introductory reference is made is that by A.W. Fahrenwald. I  t embodies the re-  sults of a three years' research and is probably by far the most important contribution which has yet been made to the study of the process. Thi s paper which is entitled "Surfaee Reactions in Flotation" was presented before the American Institute ©f Mining and Metallurgy in February 1924. I t was the present writer*s privilege, due to the kindness of Mr. Fahrenwald, to see a M.S. copy of this paper several weeks before its preliminary publication and the basis of the present research is a continuation of those phases of his work which seemed to indicate the greatest possibilities of successful treatment. I t would be difficult to completely summarise here Fahrenwald's paper which is divided into four parts. Th e headings of these are (1) "Surface Reactions with Special Reference to the Air-Liquid Interface". (,ii) "Surfaee Reactions with Special Reference to the Liquid - Air Interface", (iii) "Surface Reactions with Special Reference to the Electrical Charge on Mineral Particles", (iv) "Surface Reactions with Special Reference to the Adsorption of Flotation Oils by Mineral Separation."x Perhap s the most striking conclusions which can be drawn from his paper are. First , That *»  6. Trans . Am. Inst. Min. & Met. (Vol. 70 (1924) p. 647. x. Note . As many references will be made to Fahrenwald's paper in the following pages it is suggested that if the reader is not already acquainted with this work, a preliminary survey of it will prove of considerable assistance.  6. the surface tension of the solutions in which flotation is conducted is not very much lower than that of ordinary water; second, that organic eompounds with polar or active groups play an important part in flotation; and third, that Fahrenwald's results seemed to suggest that selective adsorption of oils and re-agents is directly connected with the flotability of minerals. Thi s last result, it will be seen, is at direct variance with that of Langmuir. The surface tension values obtained by Fahrenwald and Taggart & Gsudin are in very satisfactory agreement. Surface tension measurements made at the static surface of a solution are not of course a true indication of the forces existing in a bubble film, when sufficient time has elapsed for a contaminant which lowers the surface tension to concentrate in it. Tha t such concentration does exist has been proved theoretically by Gibbs in the well known equationx that bears his name and has been shown experimentally by both Taggart & Gaudin? and Rahn. x Gibb»s Equation Z s -  c x d  y  X • excess of eontaminent in the surface layer* c z molar concentration of the eontaminent in the solution. R m the gas constant - 83.15 X 10° ergs. T = Temperature in degrees absolute. y - the surface tension of the solution dy » is determined from a surface tension concentration "dc curv e for the solution under investigation. 7. LOG . cit. p. 509. Se e Reference No. 3« 8. "Di e Entstellung der Butter" Proc. World's Dairy Congress 1923 Vol. 2. p. 1008 Mote. Thi s last reference was brought to my notice by frof. Golding of the Department of Dairying, University of B.C.  7.  It should however be remembered that in flotation the bubbles are formed under dynamic conditions and that at the time they are formed the concentration of the solution is uniform throughout. Consequentl y the surface tension of the air-liquid interface will be the same as that which exists momentarily at the bubble-liquid interface. Obser vation of what takes place in any small testing (flotation) cell will show that these bubbles attaeh themselves to floatable minerals almost as they are formed. Consequentl y though subsequent diminution of the force of surface tension undoubtedly takes place in the bubble film, flotation of the mineral commences at once. Fo r this reason the values given by JPahrenwald for solutions may be taken as representing the surface tension of newly formed bubbles in those solutions. The conditions for the stability of bubble films are explained by R.B. Elder? as follows. " A oontaminent capable of changing the surface tension must be present in order to give the film a variable tension. I f the contaminent lowers the surface tension, it is adsorbed at the surface, and is present there in greater amounts than in the interior of the liquid. Stretchin g such a film exposes the liquid containing less of the contaminent, and thereby raises the surface tension and re-enforees the film where rupture is threatened. A  re-concentration  9. "Interfaeia l Tension Measurements and some Applications to Flotation". Pamphle t Ho. 1 (1?21) p. 23. Universit y of Idaho.  %  8.  of the contaminent then takes place, calling for another adjustment of the tension; the net result is constant motion and a dynamic rather than a static equilibrium". So far then, the probable surface tension of newly formed bubbles and the conditions under which they will be stable have been dealt with . Th e reason why particular minerals under certain conditions attach themselves or become attached, to these bubbles is not so clear. I n faet though many theories have been advanced to explain this phenomenon not one at present exists which can be said to be of any great practical value. Almos t the entire process has been built up on empirical results, and the development of flotation to the stage where it is now used to treat over seventy million tons of ore annually ha  s taken place with practically nothing def-  inite being known with regard to the theory which underlies it.  10. Thi s figure is taken from Alexander^ "Colloid Chemistry" p. 82. 2nd Ed (1924) Van Nostrand Coy. This writer does not quote his authority and a search made for statistics has failed to produce confirmation of his statement, but a rough estimate of the total tonnage treated annually by mills known to be using flotation indicates that the total of 70 million tons may be considered as a very fair one.  9. PRELIMINARY WORK Consideration of the Importanoe of Selective Adsorption  It has already been pointed out that the results of Langmuir and those of Fahrenwald as to the relative adsorption of oils by minerals do not agree. Definite information on this point seemed to be of such primary importance that it was chosen as the basis for the first set of investigations. This necessitated a mtlthod being selected for measuring adsorption* Taggart & Gaudin 11 and Fahrenwald hav  e estimated  adsorption by determining the surface tension of a solution before and after treatment with an adsorbent* Their method has already been referred to but the following explanation of its application may help to illustrate it more fully*  11. Loc . cit. Kef. No. 3* 12. Lo« . cit. Hef. Nos. 4 & 6*  10.  LL*i Let the curve DCBA, see Pig. 1, represent the change in surfaee tension of water contaminated with different amounts of a reagent (oil or salt) which we will call x. The surfaee tension of several solutions containing various concentrations of x are measured and these values for surfaee tension plotted against those for concentration as shown* Suppose that the point B on the curve represents the surface tension of our original solution then B» will represent the concentration of x. I f then this solution be agitated with a mineral y, and the surface tension of the solution after agitation is found to he C,then the change in the concentration of x will he B^-C,which valu e maybe regarded as the amount of x adsorbed by y. This method of estimating adsorption was accordingly selected for the present work because the principle on which it was based had already been tested by the investigators mentioned*  11. As a result of this decision the various ways of measuring surface tension were examined. Development and -Description of Surface Tension Instrument Any surface tension method which is employed for estimating adsorption should embody rapidity of manipulation with a high order of relative accuracy. Th e direct method is the only one which fulfills these requirements, the other methods such as the capillary tube method, the drop weight method and the ripple method all failing to satisfy one or other of them. •* The direct method depends upon measuring the force which is required to detach a solid from the surface of the liquid Whose surface tension is being tested. As already stated Fahrenwald has devised an instrument for this purpose and gives details for its manipulation in the references cited.14 H e uses a straight pieee of silver foil held with the thin edge vertical for measuring the pull of the liquid. Thi s piece of silver foil is about 6.4 c.m. long and 1.5 c m . deep. I t has a rectangle 6 c.m. by 1 c.m. cut from 12. Fo r a review of 20 methods of measuring Surface Tension see - Ferguson, Allen, "Capillary Constants and their Measurement," Scienc e Progress Jan. 1915* p.428. 14. Jnl . Am. Optieal & Rev. of Sei. Instruments, Sept. 1922. See also Kef. Ho. 4.  12.  its lower side so that two prongs, 2 m.m. wide and 1 c m , deep, project down into the liquid when the 6 c m. edg e is brought in contact with it. Thes e two prongs serve to support the liquid film, which forms as the metal is withdrawn due to the lowering of the dish which contains the liquid, and the pull which results is transmitted by a silk thread that passes over the circumference of a thin cork wheel. Thi s wheel is pivoted at its centre and is fitted with a pointer which normally projects vertically downwards. Th e pull exerted on the wheel causes the pointer to move from its vertical position, and this movement can be traced on a scale placed immediately behind it. Suc h movement is, however, resisted by the pointer itself, whose moment of weight acts as a force opposing that exerted by the liquid film formed along the edge of the silver foil. Fahrenwal d calls this piece of silver foil a "knife edge." Th e force which is required to detach it from the liquid is measured by the angular displacement of the pointer. Du Nouy ^ has devised a similar instrument. H e measures the force required to detach a platinum ring from the surface of a liquid. Thi s force is transmitted to a lever arm attached to a torsion wire balance. This * lever arm also serves as a pointer to measure deflection.  1^, Jour , of Physiology Vol. 1 No. 5 (May 20th. 1?19).  To face pa;?e 13 '  Pla$e I 'hoto graph of Improved -Direct Lleasuring Surface Tension I n s t r u m e n t .  135.  The instrumen t illustrate d i n Plat e I was devise d for th e present wor k an d i s A  combinatio n o f thos e o f  Fahrenwald an d D u Bouy. D u Houy' s torsio n wir e ha s bee n substituted fo r F a h r e n w a l d ^ pivo t an d counte r weight , and hi s leve r replaced b y ffahrenwald's w h e e l . Th e pointer s which ar e used i n eac h o f thei r tw o instrument s t o indi cate th e scale , have bee n eliminate d an d thei r plac e take n b y a smal l mirror attache d t o th e axi s o f the w h e e l . Plate I I show s th e complet e se t u p o f th e apparatu s an d the manner i n which t he light i s directe d o n to th e mirro r and thenc e reflecte d t o th e scale . Befor e proceedin g t o deal with th e metho d o f operatin g th e instrumen t a bsie f description wil l be give n o f its construction , whic h augments th e detaile d dimension s foun d i n Plate III . Th e base i s o f iro n an d w as cas t i n on e piece . Whe n cas t i t w a s se t up s o tha t bot h shoulder s eoul d b e machine d a t the sam e time , and groove s cu t i n the m fo r th e slide s which hol d th e end s o f the torsio n w i r e . Th e w h e el itsel f was turne d fro m a piece o f aluminium , an d befor e removin g it from th e lath e a scratc h w as mad e o n it s circumferen tial surfac e a s a guid e fo r th e sil k threa d use d t o trans mit th e pul l o f the knif e edge . I t i s very importan t t o note tha t a groov e i s no t required , a s thi s migh t interfer e with the threa d maintainin g a tru e tangentia l positio n to t he whee l whil e a measurement i s bein g taken . Befor e  To face pa^e 13•  -;  Plate  II.  -hoto^raph of >iet-U|?pf -direct l.Ieasuring Surface Tension I n s t r u m e n t .  14.  cutting out the hollow portions of the wheel a hole was drilled in its oentre for a spindle.wfeeei.  This spindle  had one side flattened to which the mirror could be stuck with shellac.  This flattened side was purposely  made at a slight angle to the two small set screws, used for clamping the spindle to the torsion wire, so that the eccentric weight of these screws would counter balance the eccentric weight of the mirror.  The wheel was  fixed to the spindle by two hexagonal nuts and the details of this construction are shown on the left hand side of the Plate III. of steel piano wire.  The torsion wire used was a piece  The photographs show that the stand  immediately beneath the wheel is covered.  This was done  in order to provide a soft surface on which the wheel might rest during adjustment. used for this purpose.  A piece of sheepskin was  The hole bored in the base of  the stand serves for a pin by means of which the instrument is firmly attached to the table while still permitting sufficient movement to enable the light from the mirror to be directed onto the scale. The measurements for the stand have been given in inches and those for the wheel and its accessories in centimetres. The knife edge was made of platinum foil but gold may be substituted if desired, the thickness used (.01 cm.; being Just heavy enough to prevent bending taking place in the  15.  -•• blade.  o r  it s prongs. A platinum wire forms the  loop by means of which it is hung on the hook attached to the end of the silk thread. No dimensions have been given for the small balancing pans which can be seen hanging on either side of the wheel. Thes e are made of cork, but any light material can be substituted. The shoulders of the stand were purposely designed with sufficient length so that two more holes could be made for the clamping screws if desired. Thi s provides increased adjustment over and above that which is given by the slots cut in the slides. A s the deflection of a torsion wire varies directly as its length and inversely as the fourth power of its diameter, practically any increment of deflection on the scale can be obtained provided the elastic limit of the wire selected is not exceeded. In constructing such an instrument as the one just described the inertia effect of the wheel and scale pans should not be overlooked, as it has to be withstood by the delicate liquid film. I n this respect an improvement in the present instraraent might be made by reducing** the thickness of the wheel to 2 millimetres and also cutting down the width of the snokes. Th e inertia effect of the spindle and central adjustments is almost negligible.  16.  The chief advantages v/aich this instrument appears to h ve over that devised by Fahrenwald are: jj'irst: Th e jfriction of his pivot is eliminated and consequently the scale v/ill alv/ays have - constant Kero. Second: riis device to facilitate easy reading of the ;~cale with increased deflection reduces the apparent sensitiveness of his instrument. I n the present case if the measuring scale is placed vertically, as illustrated, the increments of turning will be multiplied by a tangential function of the an«?le turned from the horizontal. Thi s causes toe increment of scale deflection to be greater at the point of rupture than at aero. 'xhe particular improvement,which th e present instrument has over that of Du Houy, is that the moment of torque on the wire is constant. A  special correction, has  to be applied in the case of his instrument in order to take care of the apparent shortening of the lever arm of his machine due to its moment varying as the cosine of the angle which the lever makes with the horizontal. This correction has been dealt with by ICLopsteg1 , but as far as is knovm Ms result s have not yet been published. The substitution of the small mirror for the pointers used in the other tv/o machines appears also to be a very decided improvement. 16. Scienc e Vol. 60 (1924) p. 319.  17.  The light which is directed on to the mirror is furnished by a j?0 watt Edison Mill type lamp.x I n the present instance a piece of J? cm. diameter drawn glass tubing filled v/ith boiled distilled water was used as a condenser. A  silk thread drawn across a slit  cut in the tin box, which served as a lantern, was focussed by means of a movable lens until it was reflected on the scale as a hair line. A considerable amount of time was spent before the light could be brought into satisfactory adjustment. Lamp s focussing a hair line are supplied for use with galvanometers present ease and might easily be adaoted for use ih'the A > Par ~ D Ct W6CT1  ticularly if t;he distanceA-£# the scale :d.ncl the m:'rror is rediided. A s illustrated the scale is about four feet from the mirror, and the hair line at that distance can be focussed until the entire w;; dth of its image does not exceed one millimetre on the scale. Th e instrument is so nicely balanced, that if the wheel is given a slight oscillating motion it does not again come to rest for four or five minutes. I t was constructed entirel y within the University. x. Note: A drawn wire filament lamp is useless as the condenser makes the filaments show up on the scale as a series of red lines. Thi s difficulty is practically eliminated when a lamp fitted v/ith a circular mill type filament is used.  18.  Method_-a.f~i5ff-eT,grting "Instrument. The scale was calibrated by placing known weights on the weighing pan and making a series of scale readings which could be plotted graphically using weights as ordinates and scale readings as abscissa. The formula which connects the strength of the pull with the surface tension in dynes is given by the ecpation: Surface tension = W  . g "2.1  in which W  =  th e force of the pull measured in grams weight  g = the gravity constant. 1 • the length of the knife edge in centimetres. If a knife edge of definite length is employed a second graph ca n be drawn ,  usin g surface tensions in  dynes as abscissa and again weights as ordinates. The reason that dynes cannot be read direct from one graph is due to the fact that a correction has to be made to take care of the pull exerted by the two prongs. This necessitates tv/o readings being taken, the first when the whole blade is in contact with the liquid, the second when only the two prongs are touching it. Th e difference betv/een the weight corresponding to the first reading,and  1?.  and that corresponding to the second is the pull exerted by the film along the length of the knife edge between the prongs, and as such can be converted into dynes. If a circular instead of a perpendicular scale were used dyne* could be read off directly from one graph. Considerable time was spent in trying to eliminate the pull of the prongs. Coverin g them with paraffin wax and also with graphite was tried unsuccessfully in both cases. Thei r pull could be reduced but not eliminated. 17 1  9  8 1  Taggart & G-audin ' , Fahrenwald an  d Du Nouy  7  all  give methods for cleaning the knife blades or ring used. All these were tried besides several others, but the only really satisfactory method found was that of immersing the blade in sulphuric acid which was jus t on the point of fuming. I f the lieu id being tested were of a very oily nature the blade was removed from a first bath of acid after a period of ten minutes, and placed in a second for a similar period of time. The n the acid was washed off under the tap and the stirface tension measurement madeas soon as possible. Distille d water may be used for this last washing, but it quickly becomes contaminated, and if tap water is sufficiently 17« Loc . cit. p. 47? Sef. No. 3. 18. loc . cit. Hef. no. 4. I?. Loc . cit. Hef. no. 15,  )i  20  pure it seems to give the more satisfactory results. The reason for using the two oaths of sulphuric acid is due to the first rapidly becoming; dirty if much oil adheres to the knife blade. J t therefore s erve s to remove the major portion of the contaminent while the second completes the process. T o cite only one foT cleaning of the several methods triedA, heating the blade to a red heat die? not <?ive nearly such satisfactory results as that just described. To o much emphasis cannot be laid on the importance of thoroughly cleaning the knife blade before a reading is made. Th e instrument is much more delicate than that used by the other investigators whose methods of cleaning were tried , and differences which would pass unnoticed on their scales show up in the present case as differences of 2 millimetres or more. Converted int o dyne s this variation is ecu ivalent to .14 dynes. I t may be noted that one of Du Kouy's methods for cleaning his ring was Yery simila r to the present one, barring the fact that he does not emphasize that the sulphuric acid used must be practically boiling. Either one of the two types of glass dishes shown in Plate IV may be used to hold the liquid whose sur- i> face tansion is to be measured. Th e dish which is marked Typ e 1 was designed s o that surface tension measurements could be made on an expanding surface.  C.J.*  Such a surface it was thought v/ould closely represent that existing in a "bubble film at the instant of its formation, and shoul d therefore give results v/hich v/ould parallel flotation actual conditions more tglosely than measurements made on a static surface . Typ e 2 indicates a more useful size of dish than that shown in Type l,as th e knife blade is less likely to be attracted to the side of the dish v/hile a reading is being made. An y dish v/hich is used must be clean. Cleanin g may be effected by first washing it with alcohol and then with several changes of water. The liquid - the surface tension of v/hich is to be measured - is poured into whichever type of dish is selected. Thi s dish is then placed on the adjustable top of the stand shown in the bottom left hand corner of Plate II. Th e top of the stand i s raised until the knife edge engages in the surface of the liquid. Th e stop cock under the dish is turned so that the liquid may drip out slowly into the beaker below, and the top of the table is then lowered until the light on the scale shows that a deflection has taken place which is about five sixths of that v/hich will ultimately be produced. The n the top is fixed and the remainder of the scale deflection is brought about by the lowering of %» Mote : Surface tension measurements made on containing completely dissolved substances that no variation in surface tension could tected between the results obtained by the dynamic method.  solutions showed be destatic or  22.  the surface which takes piaoe as the liquid drips out of the glass dish. Th e scale reading will usually be found to remain stationary for a short time after fixing the top of the stand. The n when the cross hair starts to climb it will continue to do so until it is a distance above the final reading of the surface tension by an amount which corresponds to a scale reading of from one to three dynes. Pro m this maximum point the cross hair will gradually commence to slip back until it comes to rest for an instant before the film finally snaps. Thi s final point is the one to be recorded. Sometime s the break occurs before the reading has reached its fin,al value. Shoul d this happen, it usually takes place when the first reading is being made with a freshly cleaned blade. (N0TI3 : I n this case the pull of the liquid on the whole blade seems to reach such a maximum value that when the film ultimately has to withstand this excess force it seems unable to do it, and rupture takes -place at a point on the scale somewhere between the maximum and what should have been the final reading.) Wit h a little practice such breaks will be easily recognised and in cases such as the above it seems good practice to take another reading immediately without re-cleaning the blade. I n fact, according to the solution being measured, two and sometimes three consecutive readings may be made without re-cleaning.  23.  No rule can be laid down on this point, save that no result for surface tension,which is subsequently recorded,was considered as satisfactory unless two out of three readings checked exactly on the scale and the third was not, more than ..j?  m.m. removed from  the other two. Thi s difference of .p m.m . in the present case represents a difference of about .0 7 dynes. j| source of error which must not be overlooked is that caused by changing temperature. Th e surface tension of water falls about .17 dynes for each rise of 1° 0. Th e temperature at which solutions are tested should therefore be that of the mean temperature of the laboratory in which the instrument is set up, as this will considerably facilitate rapid work. The correction for the prongs is usually made without any difficulty an d the reading is taken at the point when one or both of them break free from the surface of the liquid. This , it should be noted, is not a true correction, as the pull of the film is measured while both prongs are projecting into the liaiid surface. On the other hand it is practically impossible to measure exactly the pull of the prongs when they are in this position, and consequently the point of rupture is taken as this is definite and no difficulty is found in repeatins; the reading.  24.  Surface Tensio n Experiments with Pure Licuids and Solutions.  The error introduce d wit h th e prong correctio n mos t probably account s i n a large part fo r the fac t tha t surface tension measurements made with th e instrument ar e slightly high.- Thi s is shown rathe r clearl y "oy the results given i n Table 2. Table 2 . Substance  Distilled Jaie z Acetone Benzol Ether  Temperature  Results Obtaine d with Instrumen t  20° G 17.3° C 22.3° 0 20° G  73 • 3 dynes 23.O 1 30.6 " 17.3 " •  Results Ootaine d oj Othe r gjnvestigators. 72.8 23.3 29.4 16.8  Martin s Jaege r Canto r Brunne r  The substances teste d i n the present cas e were not specially purified, but were o f th e ordinary 0.}? . quality. A s hov/ever the aosoiut e valu e o f the surfac e tensio n of som e o f these liquids is still in dispute th e genera l high trend o f t.oe present result s rathe r than thei r actual value is of importance. It has hov/ever cUre-cl-dy been state d tha t the instrument was intende d t o be used fo r estimating adsorption. Consequentl y a slight inaccurac y i n the absolute values o f surface tension obtaine d would b e immateria l provided that the erro r remained constant . 20. These , results ar e take n from : "Handbook o f Chemistry and Pnysics" p. 43. lOthlMition (I^ZAj O h em. Rubber publishing Co.  0  K  Ui  h  O ifl Z Ul  51  a  >•  'UJ Z  49  SI  0.1 M  Ul  !3  7C  :  Mil  of  •  :  .  :  '  .  '  •  •  :•  '  -  •  u:  51  •  •  •  PLATEte •  •  •  •  •  —  •  —  IN  CURVES SHOWIN G DECREAS E  ;i  !' •'•  •  ;  .  •  •  •  •  .  l o f 85  PERCENTAGE O F W A T E R  I N ' T A B. ANN .  fa  SI N PERCEN  T 0/  ^~~~^x  . D E C O N S T A N T S " VOL. 4 P A R T H .  E FRO M DAT A GIVE N FO R ZO'C.  10/ I CON C ENTHATION O F A C E T I C A C I D I  * CURV  *  5TR U M ENT . l> « PLATE S i,vm.)  INCREASING CONCENTRATIO N OFACETI C ACI D  WITH  SURFACE TENSIO N O F AQUEOUS SOLUTION S  ~ - \ INTERNAT  .  "•  7st  S»2  o  For this reason it w- £ felt that insie: d of spending further time investigating why these vslues were high it was more important to test the relative accuracy of the instrument. This was done by measuring the surface tension of aoueous solutions which contained known percentages of acetic aoid. Thes e results ha", e been plotted and appear in Plate T# I t v/ill be seen that nearly all the values obtained lie directly on a smooth curve and that this ourve is in excellent agreement with the second one plotted from data given in "Tables Anrmelles Internationales de Constants".21 Thi s fact indica'ed that the instrument might at least be considered as giving excellent relative results. Th e next tests to be made were those on emulsions.  Surface Tension Experiments with Emulsions The first emulsions tested were those of petrolatum in water.  The method of procedure w- s as follows:-  The surface tension of a sample of distilled water was measured, and the remainder of the water used to make up various emulsions which contained petrolatum in t various concentrations. A s these emulsions quickly separated into tv/o phases surface tension measurements v/ere made on them as qluickly as possiole. 21. Tae*. Ann. Internat. de Constants Vol. 4, Part I,  »  2b.  It was fouiid that the value of surface tension obtained fo r these emulsions was the same as thr~t which had been found for the water providing the measurements were made before a complete lens of oil had formed under the "knife blade". Thi s result was in- substantial agreement with that obtained by Taggart £•. J-audi n for emulsions of nujol in water. Thes e experimenters record nujol a value of surface tension for their.A emulsions which is only .2 dynes lower than the value which they obtained for the sample of water, and state that such variation is within the limits of their experimental 21a error.  the  The result ofA present tests is given in Table 3 r  but no concentrations of petrolatum are specified as the final point at which constant readings can be made varies with the type of dish used for holding the emulsion and the speed with -which the measurement is ms.de. Table 3 Temp.  liquid Die t i l l e d IVater D i s t i l l e d 7/ater Petrol;-turn  The n e x t t e s t s o l e i c a c i d in 81a L o o  3 u r f s. c e T e n s i o n  20°G  73.p  dynes  20CG •  7-3 -3 dynes  c a r r i e d o u t were w i t h e m u l s i o n s  water.  . p i t . p . 49 9 Ref .  Wo. 3 .  of  Oleic acid v/us used in some of tne early tests carried out in the laboratory- of the Minerals Separa22 tion Company in London bu  t it has not had much  commercial application most probably on account of its high cost. I t is an excellent example of an insoluble oil which contains a polar "roup and has the formula G17 H53 GOOH. Th e COOH group being the polar portion of the compound. Th e oil used for the following tests is that ordinarily supplied by the "Merit11 Company and no further purification of it was undertaken. Th e value of the surface tension obtained for the oil itself was ;>4.0 dynes. A set of emulsions was prepared in which the concentration' of the oil was 1 [ in 150,000, 1 in 120,000, 1 in 90,000 and 1 in 60,oo0. As oleic acid is practically insoluble in water even the most dilute emulsion contained a considerable portion of the oil added in an undissolved state. Th e Type 1 dish, Plate IV, was used t o hold these emulsions, but the results which were obtained from the tests were very discordant. The y indicated, however, that surface tension measurements made when the surface of the liquid was expanding vrere several dynes higher than when no expansion took place. the n the 22 Lo c cit p. 17. 3ef . Ho. 1.  eo«  stop cock Of the dish w e closed another variation was noticed*  If the film were ruptured by lowering the  platform on which the dish w- s placed, and the measurements repeated without re-oleaning the blade, the surface tension value obtained was found to drop until it became practically constant.  Shis was the case even  for the most dilute emulsions. The foregoing results have been summarised in Tables*^. Table 4. '-*  Ligaid  Temp. Dynamic Test  3urface Tension Static Test  If  Oleio aoid. Emulsions of Oleio Aoid la Water  >4*0 dynes  j&4*0 dynes  18° C * 18° 0  Increasing with rapidity of expansion of Surface* rax. (approx) 4^.0 dynes  diminishing with repetition* Mia* (approx) pt>*> dynes  • * ' • • " * .  The above maximum and minimum values are of interest as Fahrenwald ' gives 37 dynes as the surface tension of oleic acid saturated with water and 4b*2 dynes ». for water saturated with oleio aoid*  tansmuir2  gives  42.8 dynes for this latter value* 23* 24*  Loo. cit. p. 693 Hef lo. 6. Loo* oit. Hef. IS* 2* abstracted on present text*  pa-e of  29.  The similarity which exist s between the high and low results obtained i n the present case and. those cited above is very marked. Apparentl y repeated measurements resulted in an oil film forming along the under side of the knife edge,(IIote. Th e start ox the formation of such a film could be plainly Scen) and the minimum value recorded for surface tension must have been that of this oil film, and would be expected to correspond to the surface tension of oleic rcid saturated with water. Henc e the similarity between the present minimum value of 36.5 dynes and that of 37 dyne give n by i!'ahrenv;ald for oleic acid saturated with v/ater. Th e maximum value would naturally be expected to approach that of v/ater saturated with oleic acid as the more quickly the measurement was made the less chance the oil would have of spreading along the knife edge and influencing the determination. lang muir's value for the surface tensio n of v/ater saturated with oleic acid was 42.8 dynes,and the maximum recorded in the present case for the emulsion s was 43.O dynes. Consequentl y it might be expected that, for emulsions of oleic acid in v/ater,all values of surface tension between that of oleic acid saturated with water and water saturated with oleic acid might be obtained,according to the particular momentary  30.  condition of the surface on which the measurement was made. Surface Tension measurements made on emulsions of G-eneral Naval Stores oils Eos, j > and 22 also gave discordfiHt results, presumably for the same reason. Similar results to those just recorded for oleic acid emulsions have been found by the following investigators but no explanations are suggested to account for the variations observed. Tag5art & Oaudin2^ as already pointed ou t (p.4 ) , stated that they could not obtain consistent values for surface tension when working with emulsions. Bu Uouy^, after testing the surface tension of various colloidal solutions .states that their surface tension decreases rapidly as a function of the time. 27 Fahrenwald ',i n some experiments made with aqueous solutions containing oleic acid ,obtained discordant results which he marks with an interrogation. Ivlilner foun d that the surface tensio n of solutions of sodium oleate in water decreased o n standing until it reached a definite value, but this definite » value could not be duplicated and sometimes differed as much as 5° o . 2^. 26. 27. 28.  Loc. Phil. loc. Phil.  c i t . p . 514 Hef. Mag. S e r i e s 6. c i t . p . i>Si Hef. Mag. S e r i e s 6,  #0. Vol. No. Vol.  3. 48 (1924) p . 264-77. 6. 4;>, p . 6 6 3 .  31.  The present results together with those of these other investigators indicated that adsorption from emulsions could not he determined directly by a change in their surface tension^?. Thi s meant that some other method for measuring; adsorption .from emulsions would have to be found. The surface tension method could of course still be applied to true solutions but most of the best "collectors"are compounds which are only slightly soluble in water. . These compounds are used solely for au-gmenting the attraction which exists between the &iTbubbles and floatable minerals and not to augment frothing. Stud y of their action should therefore •  .  •  •  •  I  produce valuable information as to whether selective adsorption actually takes place or not. 29. -0. Bat^ch in a recent paper entitled "iieitrag zur Theorie des Sehaiunschwimmerf ahrens". Koll . Chem. .Beihefte Vol. 20 (Oct. l?24j p. 50-77 measured the surface tension of oleic acid emulsions by the drop weight method after adding to them a known amount of KOH more than sufficient to convert all the acid to potassium oleate and so render it soluble in water. liilne r (Ref. Ifo. 28)found that the surface tension of a similar solution was not constant and Batsch does not submit any data to show that his dynamic method give s consistent results.  Investigation of Kgthods for Estimating Adsorption from gmulsions. The proportion of oil to water used in flotation practice varies from 1 in 2,700 to 1 in 16,000. A method had to "be found which v/ould permit determination of fractions of these amounts. Th e methods ]£= considered in the present case follow:ffirst: A conductivity method - dependent upon the change in conductivity of an emulsion with a change in concentration of the oil. oecond: A n optical method based either on the principle of refraction or that of turpidity or opacity. Third: A  method of direct analysis. /  After considerable preliminary investigation the last of thece me hods was selected as likely to give the best results. Reference s to the other two methods will be found in the attached bibliography but only the second seems to offer possibilities o f successful application. The method developed for the present work depends upon th e fact that oleic acid can be titrated with Na OH in a solution of alcohol and water using phenolphtalein as an indicator. Th e procedure for collecting the emulsified oil so that this might be  To face page 33'  Plate VI Photograph of Mechanical Shaker. The size may he judged from the large driving wheel which is 8" in diameter.  33.  done was as follows:2? c.c . of neutral CC.l^ v/ere introriuced int o a 500 c m . suctio n flas k containin g 20 0 c.c. o f an emulsion o f oleic aci d i n water. Th e flask was corked an d evacuate d afte r which it was placed i n the mechanical shake r show n i n Plate V I and shake n for fifteen minutes. A t the end o f this time th e contents of th e flask were poured int o a separatin g funnel an d left fo r twelve hours. ih;rin ^ this time the emiilsic n separated int o two phases and the GC1 , settled ou t car r in? ? with i t the dissolved olei c  Abortion  20c.c A of this CCi 4 solutio n was then taken and 40 c. c o f neutral alcoho l were added t o it, after which the acid was titrate d wit h a n IT solu 2u~0~~ tion o f i!a OH. i'h e weight of oleic aci d containe d in the origina l am eoue solutio n could then be cal culated. The reason tha t only 2 0 c.c. of the origina l 2i? c.c. o f CCl, solution were take n was because absolutely complet e separation o f the emulsio n int o two phases took several days an d preliminary determinations made as described indicate d tha t satisfac tory relative results coul d be obtained without waiting for this to take place. Inspectio n o f Table j> illustrates thi s point.  } * •  Table 3« '•.it of Oleic ^cid Taken 11.73 mf? 17.33 '• 22.60 " 23.30 "  V,t of Oleic Acid Pound 11.30 16.15 21.20 21.80  rag " » «  Amount Lost .45 ms 1.2 " 1.4 " 1.30 »  The results show that all the acid added is not recovered but rhat the loss is a straight line function due doubtless to a distribution equilibrium having to be reached between oleic acid and v/ater. Furthe r experiments made to determine whether this lose varied with the temperature showed that for temperature variations of lese than 5° 0 the error could be neglected.  » •  ADS03PTI0K TEST S 171TH OLEIC ACID.  Adsorption test s on emulsions of oleic acid in water msjrng galena, pyrite, and granite as adsorbents were' carried out as follows:200 e.c. of boiled water containing a known weight of oleic acid were placed i n the suction flask. 1. 5 g.m. of galena screened to -100 +1.50 mesh v/ere added and the flask evacuated. Th e whole was shaken for thirty minutes in the mechanical shake r after which the contents of the flask were allowed t o settle. Afte r ten minutes a known quantity (usuall y about l80-l?0 c.c.J o f the superasaat liquid was decanted from the first flask into a second one. T o this second flask 2.5 c.c. of COI4 were added and the determination of oleic acid made by the method described. Th e temperature throughout being maintained between 16 - 21°C ' .  Blan k tests v/ere also  run with no oil present to determine what correction,if any,v/as required for the effect of the ore itself. Sim ilar sets of determinations v/ere made using pyrite and granite, a weight of screened or e being taken v/hich had the same surface area as the galena. Thes e tests were . 30. • Owing to the conditions under v/hich the work was done a closer range of temperature than this v/as very difficult to maintain. Langmui r however in his paper-on the "Constitution of Solid and liquids'1 Part I. Jour. Am. Chem. Soc. Vol. 38 ii (1?16J p. 224? states that,"there are good reasons for believing that the intensity of the surface field of force is substantially," independent of the temperature."  *  36.  repeated using ore which had" "been screened to -1J30 +200 mesh and the weights of ore taken were calculated so that each sample would present the same surface area es that already used for the tests with -100 +1.50 mesh galena. Thes e calculations for comparative surface area were made from data given in one of the Tyler -Screen Company's catalogues31 . Th e screens used in the present work were not specially standardised bu t the same set of screens was used throughout the tests and any error resulting from their use would therefore be constant and would not affect the comparative value of the results. The results of these adsorption tests have been plotted on a practical ba.se and will be found in plate VII. _ (See next page.) 31. Th e Profitable Use of Testing Sieves. Catalogue Uo. 48 p. 37. Tyle r Coy. Cleveland u-d. J... x Ilote . A more accurate method than that of screening, for finding the siae of the particles is to apply a formula based o n dtoke's law of settling. Thi s formula is: 6 rrnrv= 4 rrr-^ (it-djg in which  J i) - the viscosity of medium in which settling takes place - usually water r = th e radius of the particle v = the rate of settling v r the density of the particle which settles d z th e density of the medium in which settlin? takes place. This formula was net used in the present ec-so to find tie si£e of particles ss the error introduced by takin? the mean values rive n in the Tyler Coy's catal: \ae was ' less than the other experimental errors involved.  To face page 3f«  LO  Ul  5  o  <  m Z3  • <  o rr < > •"—'  Ul  U  u  o  h  Ul  — -< ul ' 7. Ul -J  7-  >• < < <n ; 1  o  <  > o  - </ > ZD  u  LJ  1=i <  __i  a  ?5VJd(1S 1V&3III  U J O 1 0 0 J JSvfiS ? 5 1 5 OJAOM/ S ">I 0 J O «ONH0 d .  37  Inspection of Plate Vli will show that apart yt fro m the two low points which fall below the curve drawn for galena and which are enclosed in a small rectangle the other results obtained are in very satisfactory agreement considering th e experimental difficulties involved. Thes e two low points are rather interesting, as they illustrate the difference produced by not grounding the contents of the bottle during agitation in comparison with the other pair - indicated by the arrow - which were grounded. Groundin g was effected in the above case by passing a copper wire through the rubber cork of the suction bottle. Thi s particular result was only noticed when -150 +200 galena was used, and the results for granite and pyrite showed no such difference. Thi s difference is undoubtedly due to the fact that the emulsified oil and the mineral are both negatively change d with respect to water and as the relative proportion of the change possessed by a small particle increases as the size of the particle decreases, no such difference was manifested "ay the larger particles. Moreover the charge on the smaller particles of pyrite and granite did not affect the adsorption values which these minerals gave* :  - Jthis is only further proof that the  size of the charge on a particle depends upon the composition of the particle. I n practice nearly all flotation cells . are ->  grounde d - usually quite unintention-  ally. Keepin g this fact in mind th e following conclusions can be drawn from the graph.  38.  First* Olei c aoid is adsorbed more strongly by galena and pyrite than by granite. aecond. Adsorptio n of oleic acid is a surface phenomenon and almost entirely independent of the size of the particle. Third. Ther e is a definite adsorption of oleic acid by granite. The above results are largely confirmed by those 32 Of Batseh. Despite the fact that the curves shown in Plate VII may at first sight indicate selective adsorption yet on account of the extreme thickness of the adsorbed oil film,if reduced to molecular dimensions , the results with olsio acid v/ere considered as inconclusive and it was decided t o carry out another set of tests using if possible a compound with a different polar group.  L 32. Loo . cit. p. 68. Ref . no. 29. 33« Langmui r gives 11.2 x 10~s om as the length of a molecule of oleic acid - Kin. & Sc. PreoS. Vol. 121 (1920; p. 913  59.  ADSORPTION TESTiO AITK AKITJ11E.  -Aniline (C^ H HH 2 ) which contains the polar IIBA group was selected as a suitable compound on which to carry out further adsorption tests,. The reasons for selecting it were these - first a quantitative analytical method v/as available for its determination-^ and second- though aniline itself does not seem to have been used in practice to promote flotation the polar NHg group which it contains is found in a large number of the coal tar products that are constituents of the tar oils which are employed commercially as "collectors". The method of carrying out the adsorption tests with aniline was as follows:A standard aniline solution was made up • containing one gram of jfj?eshly distilled aniline per litre. Variou s amounts of this solution were then diluted with water to 200 c.c. 1?  g m of -100 +150  galena or a corresponding weight of equal surface area was added and the whole corked and shake n in a suction bottle for half an hour, as had been done in the case of oleic acid. Thi s time however the bottle was not evacuated bu t all the tests were run grounded. A t the end of the half hour's shaking the solution v/as decanted and set aside for testing. Th e results of these tests. 34, Jour . Ind. Eng. Ghem. Vol. ? (1917) P« 953.  <  < n  a o <n DC Q  CD  Id to  QT 3  o  Ul  O  3  a  a £11 z  MJJ O f  !»  9 -0  *5*io' .  1  . [.. _ j  5 -l  -_—--"""X  GR  ..  . ;  •I  •j  •  •  -T-.  -  -  •  •  :  IN POUND S PE R TO N O F WATE R  — -  :. . - ; . . .: .; .  '••  •t'  •  i  „,.jg .  j-  •  -  !  -  -  :  ••  -•  •  •  "™" """" " • "  4 -r..;... - :  •  •  ;  - :  i• •  3*  *?  igL  rC  ft  -,-  .:  - 1 L. . . "~ : - ' - ;- - - ----4 r °-—! PY F I T E .,. - - -- ; j . _ :— rs ! E NA— 7i . - - - - GA N DU F {Hoi 2z : .... „ ....:.. mm ...... ~r HT --- r ;' ; ' •. 'i-n . ».* .:. z . H . . . '. . ! i : r ... H . . . . .__ _ _ _ \ ; H 1 —i 3> -:-f . -. ! ...; .• 4 fH Ur: ! b •h I ' 1 1 •i  _. .|.. ._,._. . .  " i  =pi"rar. j1  ADSORBED BY VARIOUS MINERAL S ±  ^PH S H 0 ^ NGAMOU kTS OF /  - ^ : T -  CONCE NTRATON O F ANILIN E I N WATE R AFTE R AGITATIO N WIT H MINERAL ,  5* «£_._.._, L . ;  PU^TE m j  j;  40  are shown in Plate ViH . The conclusions which can be drawn from these results for aniline are as follows: First. Th e amount of aniline adsorbed by granite and galen a is identical. Second. Pyrit e shows a. slight preferential adsorption of aniline. Third. Adsorptio n is independent of the sise of the particle and dependent onl y on the surface area. If there v/ere anything in the adsorption theory of flotation then supposing experimental tests v/ere run in an ordinary laboratory-testing-machine using an ore containing galena, pyrite and granite , and aniline v/ere added as a collecting oil either the granite and galena should remain unfloated o r they should both float together. I n order to test this point a synthetic ore was made u;o and a series ox determinations made on portions of it v;hich had been screened to -100 +1.? 0 mesh. The results of these tests are given in Table o which follows:-  41.  £aole 6. O i l s ~ddea  P e r c e n t a g e of l e a d Hecovered  Blank - u s i n g f r o t h i n g o i l , e . u i v . to .05 l b s r e r t o n of ore Using e ^ u i v . t o «0p l b s of f r o t h i n g o i l + . 2 7 I D S of A n i l i n e ;:er t o n Using e o u i v . t o .Op l b s of f r o t h i n g o i l + .-*o l o 8 of A n i l i n e r e r ton lining e q u l v . t o .05 l b s of f r o t h i n o i l + . b 8 l b s of Aniline t e r t o n  P e r c e n t a g e of I r o n Recovered  15.2.: 23.0JC  5*4jL  27»0jl  9-Sj.  These results show conclusively that the percentage Of galena and pyrite( represented in the concentrate by lead and iron) have both increased at the expense of the gangue,and consequently the statement onn be Bade that •elective flotation may take place when selective adsorption of oils does not* Since performing this experiment some of the re* •alts of ea investigation carried out by Sshafer*'* have been given is Chemical Abstracts and the author is credIted with the statement that selective adsorption and selective flotation are not parallel phenomena* Shis statement is farther substantiated by results given ia the works of Batsebr0 and Wilson & Barsard*7 which show respectively that oleic and stearic acids * • are adsorbed by non-floatable oxides. M , :heE. Abet -act* Vol. 19 Ho. J (1*25) p. 457 36. loc. clt. Table 9 p. 72. lef. Is* 29 37* Jour. Ind. Sng. Chen. Vol. 14 (1922) p. 68-  42.  As a result of the tests made with aniline it was felt that further time spent on adsorption tests with oils would not lead to anything of practical value, as selective adsorption and differential flotation v/ere clearly not necessarily parallel phenomena* The next question considered v/as what was the funotion of added electrolytes?  43.  "ADSOHPXIOU" TESTS WITH Cu S01  Ihe fir8t electrolyte with which adsorption tests were carried out was Cu SO4, 5H2 0 (Blue stone) Copper sulphate was selected as it is a very common addition a -ent used to promote the flotation of sphalerite, furthermore the iodide method for determining copper is very accurate and presented a readily applicable method of analysis. The minerals to be used for these tests were assayed for copper and the results of the^e assays are given in Table 7. Table 7  Mineral Quarts Sphalerite 3-alena Pyri e jpyrrhotite  .percent Cu. Hil Nil flil Trace • 4>  These results showed that any error which might possibly be introduced by the copper contained i n these minerals going into solution and affecting the tests could he neglected. The method of carrying out these tests was identical with that used in the case of aniline baring the fact that the bottles were not grounded, and a standard CuSO. solution v/as used in place of a standard solutio n  To face page 4 4 .  44.  of aniline. Th e analytical method used for determining the copper, which remained in solution after agitation with the mineral, was as follows:- A  measured quan-  tity of the Cu SO4 solution - usually about 190 c.c. was decanted and either evaporated direct to about jj c.c. or first filtered to remove suspended particles and the filter paper thoroughly washed with hot water. Test s made to see whether this filtering of the solutions resulted in any of the copper being lost showed conclusively that such was not the case. Afte r evaporation* just enough NE4 OH was added to make the solution alkaline. Th e whole was then covered v/ith a watch glass and boiled after which it was cooled and made acid with 4 c'.c. of acetic acid - The copper was then determined by titration. ** The results obtai ed for these tests are given in Plate I X • Inspectio n of this graph will show that though hardly any copper was adsorbed from solution by the quartz an appreciable amount was taken up by each of the sulphides. Furthe r study of these sulphideadsor tion curves will show that there is a very marked similarity between them.and leads to the conclusion '* x Note. I f after evaporation it was thought that any of the copper h d been deposited on the surf ce of any finely powdered mineral which w,< s present in the evaporating beaker a few drops of HNO^ were added so as to bring everything into solution and in this cabe 2 c.c. of IIH4F solution were added before titration. xx ITote. In order to eliminate the personal error as far as possible these titrations were verj kindl y carried out for me as "unknowns" bj LIr . ,/• Bishop, Assistant in Metallurgy and formerly Head Chemist of the Dranby Hining Co»  «>•  that some action other"than that of selective adsorption must take place when Ju S04 is used a s an addition a^ent to promote th e flotation of sphalerite in a.complex ore. In this connection it was noticed while the tests were bein* carried ou t that all solutions decanted from the sulphides cleared afte r standing twelve hours with the exception of those of sphalerite which had originally contained less than 2$ m.g. of Cu S04 , 5 HaO x and that of galen a which had contained less than 10 mg. As the "adsorption" results showed that definite amounts of co-per had been removed in all these.cases the explanation which suggested itsel f w s that a chemical reaction had taken place and consequently the precipitating power of the copper ion had notoee-nable to manifest itself until a definite concentration of copper had been present in solution. Suc h being the ca- e it was argued that the Cu SO. solution which remained after sphalerite had been agitated in  A  , should also contain  in solution an amount of sino equivalent to that of the oopper which had been "adsorbed". Tw o of these solutionswere therefore tested for zinc as follows. Afte r being twice filtered their bulk was reduced by evaporation to 100 oo( and the dissolved copperd precipitated with BL3 in the pretence of HC1• Th e solutions were x« •2V.J-D* . of CuSG^, 5 HoO on the graph represent* tSjblK^. p>> tan M waief»»t l the ^Teeent ttsts.  4b  then heated and H2S passed in again to make certain that all the copper had been removed after which they were made just alkaline with NH4 OH ana then neutral with then acetic acid. H2 0 wssApasse& into the warm solutions and a heavy whitish yellow precipitate was seen to form almost immediately. Thi s was suggestive of 'i inc. whe n complete precipitation had resulted the solutions v/ere filtered and the ureoipates from them, re-dissolved by pouring an HO solutio n through the filter papers. These View solutions "'--r e then tested for zinc oy adding to them a few drops of JQ Fe OH/ and in each case the typical whitish blue precipitate of zinc resulted. The foregoin g tests showed that,In the case o f i  sphalerite., the a d d i t i o n  of CuSO^ solution h a d resulted in in which  a definite chemical reaction taking place,A the surface of the Zn a had bYesu-mably been filmed, with OugS — a much more floatable mineral. Though similiar confirmatory tests v/ere not carried out in the c; se of galena it is believed that a similiar chemical action also takes place. Thi s is borne out by results obtained in practice, for addition of Gu SO4 to the zinc circuit at th e Sullivan Mill, Simberley .B.C., ' results in the galena which  is nat floate  d in the  lead circuit floating in'the first two cells of the zinc circuit.  47.  Batsch^k a s the result of some tests carried out with Cu SO4 and sphalerite also arrives at the conclusion that a chemical reaction takes place though he does not submit any confirmatory tests to prove his point "but instead quotes an article in the Mining Eng. WOrld *' . I t has not been possible to look up this reference* Somewhat similiar results were obtained by Watanabe wh  o describes some tests on replacemen t  which he carried out and shows that Cu9 3 is deposited from a  CUSO4 solution and gradually replaces ZnS.  The curve tfor ohalcopyrite is also of interest, as in the oases of low concentration more copper was found in solution than was actually added, showing clearly that some of the copper had dissolved. Thi s result indicates v/hy H2S04 may be used to promote the flotation of Zns i n certain oases, as ores containing a certain percentage of readily soluble copper will themselves, in combination with H2SO4, furnishj^he CUSO4  necessar  y "t ° promote the flotation of the ZnS*  The results for the tests with Cu SO4 may be summarised as follows*  28. lo c cit. p. 70. Ref. No. 29. 39. Min . Eng. Wld. Vol. ?1, IJo. 24, p. 1071* 40. Ec . (tool. Vol. 19 (1924 ; p. 497.  48.  First: Sphalerite , and to a lee er extent galena j enter into a chemical reaction vit h Cu SO4 consequently the beneficial affect of this electrolyte does not depend, upon selectiv e adsorption but upon chemical reaction. Second. Chalcopyrit e has been found to dissolve appreciably in a very dilute Cu oQ solutio n and points 4-  to' t h e conclusion t h a t many a d d i t i o n a vents are added by t h e ores themselves. I'he next t e s t s to be undertaken were those v/ith lime.  4?.  AiDSOHPTIO? gaiSgS i/Il'H IIME.  Calcium hydroxide is added in practice to retard the flotation of pyrite in the presence of chalcopyrite. The amount used varying fro m 1 to 6 lbs per ton of ore treated. Thesetests were run in a similar way to those carried out with Cu 50,,but the method employed had to be modified t o a certain extent to reduce the possibility of the Ca (OH;2 bein g precipitated fro m solution as Ca GO/,. Consequently all water used in the tests was boiled just previous to the experiments being made, and after the mineral had been introduced into the suction flask, the flask V/PS evacuated and sealed off. Th e time of agitation w s also reda.cer' to twenty minutes so as to reducd the possibility of air leaking back into the bottle while it was being shaken. Afte r agitation the solution in the flask was pou r d out . at once and filtered. Th e filter paper being washed thoroughly with a hot solution of NH4 Gl .  Th e lime  in the filtrate was then determined by the gravimetric method used by [iilleb.rand. *  41. "Th e Analysis of Silicate and Carbonate Rocks'' n. 140-141 Bulleti n Ho. 700 Wash. d-ov. Print. Office.  To faoe p a g e .50.  j  ;  4'  CALCULATED THICKNESS,I N CM- ,  ' "  •  OF; CALCIU M L A Y E R  4Z2 i ;•• 4.1:3 -4-3 ;-[oi _;J<Li  ,.  -r-bj 1 ! <_3 ..; I ! |<  •>  "  •  '  •  "  .  Li-  'I  N O  !-  • .  ".  .  'j  .  -4— i  •  f  ! 4" ; -i •'  •  ;:.!:„  T I |.  _'  ! ^  i O  — i - !  •IS  ! •*  ;,..  -'±-\.-cc  :!  >  -  i'  !  <  •  ••  11  !  \  . :.. I < AQtl  |cp  ;^c  v  ..J.... j .  :  4  ;i \  | ®  .  !  .  •• •  .  • ; • •  • :  .. :  —  ;.j  :  51 -  : | gpp::: , .;...  .  k t, : i  -! F r|-=  \ '  -.  :  r  • . y  •  •  •  •  . j. •  •  -  , : r -  \  • -•  •  ••S.!^:1! s! z i  "" "T ^  ill , •  U., I - 1  ^  \\: :V_ : .YV JL L \\H- h . 1 - ;  :-;v.;.. M V . !. • -i-1 | , ' li • i ; "i  z•  i  ^i  \ .  -  o «  • .  '  j .. .  J  z  •  -  '  _5 .. _ .  ,,.^......  I  .  "• '  iva3Ni w J O XOO J savno s a j j - i - j ^ -44-1 \Gaassdosav. wnmv o J O e ONOOd i _ , - ' : : • i I •  •••  " M! LOt .jans  " : ' i i' ;  • ; "  i  :•  :  |  L'._ !  .: L::L._ .  4\|#r  :• . > .....  . . I . I  :  j  JO  ....i_ ^'...:j' J  .:::4-.;...i,-:i.:.J.-.-.  ;  "• • > •  :  \.  • -V \  ..- i • . ;  -  Zft i ' l - :.....:.  .  4; 4 V  •  ..;.. A  .! ' .':  .  i  ::  .  - i  .. \ . •;._;._i,L J ** S •  4  ;  I.!...  .v  f"-;  Ml  Z:  1UJ '  \\  V  h i  —fea-  .•  1  ...  ;  - - k .<  :  ;  ....: : ' \  '•  4f: \ _____ i Ji „  •  1  - •:••• ,  1.\ ,  H CC <  -' est  \  ••  2 ...;  •  \  - . ; ' . _ : . .i  > -_ <__ r3 Or a  ibJ  •  !  :.-i..i _  L; I..:..:.. : ;.. L.-..:.. .!._. . ! : ... ....  !  •  - j: - - - •  2  IfcO  :  i •: "  ' — -j -  ;  i  :1  „.ii,,44._.: :u _  a  SI  1: i  :•  p--p-.r"~--T —  t  -ftsi  :  '  . L.J...QJ . -N  4<S  4"  '  •  -4-o  --•r  '7 1  l-za  .4.  "i  i;  -i <i 0  •i  •:r  DC Ld  ;  •i  14i  -  11 *o•  ; • • ; - '•  , j...; , i  -; :  .]  ; !  . ....^—L........i-...! —  :  Itl  iL:4L.i..:..  52.  A search was consequently made to disoover what data existed on the structure of the sulphides, oxides, and carbonates* Typical space lattioe diagrams of the unit crystals of these minerals have been reproduced from available works * and are shown in' figures 2 -10. Figure 2 represents the unit of structure of the calcite group and the difference between this structure and that shown in Fig* 3 for chalcopyrite is easily apparent* figure s 4,5 an d 6 show respectively the crystal units of pyrite, sphalerite and molybdenite* In the case of these last two minerals it will be seen that the crystal surface consists of both metal and sulphur atoms, and but for the fact that sulphur takes the place of oxygen these minerals are very similar in structure to the oxides of copper, sine and arsenic shown in Figures 7 -10» (fchbosiTe . Jiaojea 5M5Jt; If however the compounds formed by oxygen and sulphur are considered a difference will at once become apparent* Whe n sulphur combines with oxygen it assumes the role of a metal. Consequentl y if this reaction only is involved the surface of a sulphide mineral may • be considered as a metallic surface. Such being the case we have an immediate explanation for the phenomena of flotation, it is this :44* Unles s otherwise indicated these figures have been reproduced from "The Structure of Crystals'1* Wyckoff (1?24) Th e Chemical Catalog Company*  Au Ag Gu Ag 2 s M0S2 pbS ZHS  Gold Silver C coper Argenbite Molybdenite Galena Sphalerite Chalcocite Chalcopyrite Pyrite Pyrrhotite  Sl02  <i  3.3 ,3.8 6.3 2.7 4.3 6.0 7.0  SnO FeCO-PbCOz Ca CO.. BaSO^ KAlSlzOq  Zincite Siderite Cerus^it'e Calcite Barite Orthoclase Quartz  rt • /\ S  3.2 6.0  F e 2Ox Gu20  4.2 3.0 4.6  V7  1?.3 10.3 8.8 7.3 4.7 7-3 4.0  Specific Gravity  Hematite Cuprite  CuFeS2 FeS 2 FenS8  c^2s  Formula  Mineral  Hexagonal Hexagonal Orthorhorabic Hexagonal Orthorhorabic Monoclinic Hexagonal  Hexagonal Isometric  Fracture Metallic Metallic Metallic Metallic Metallic Metallic Adamantine Metallic Metallic Metallic Metallic  Lustre  _  Metallic Interrupted A dama n tin e or Sub Metallic Subconchoidal Sub Adamantine Perfect Vitreous Perfect Uneven Adamant ine Distinct Conchoidal Vitreous Highly Perfect Gonchoidal Uneven Vitreous Perfect Gonchoidal Vitreous Perfect Indistinct Subconchoidal Vitreous Subconchoidal Gonchoidal  Hone Hackly Hone Hackly None Hackly Hone Subconchoidal Basal Cubic Even Dodecahedral Uneven Indistinct Cfanchoidal Indistinct Uneven Indistinct Gonchoidal Uneven  Non-Floatable Minerals.  Isometric Isometric Isometric Isometric Hexagonal Isometric Isometric Orthorhorabic Tetragonal Isometric Hexagonal  GrystalSystem Cleavage  Floatable. Minerals  Table 8  H  CD  P= 0<3  •~d  O CD  P  o  31.  THE RELATION iET.'/BBM THE GH-iTSTAL STRUCTURE OF MINERALS AND THEIR PLOATABILITY.  The results obtained in the preceding experiments have indicated that flotation is purely a surface phenomenon and cannot be attributed to selective adsorption^ itac^mts. Taggart in his "Manual of Flotation Processes"4* has stated that floatable minerals are those with a "metallic, ad antin e or resinous lustre"* I K to the present time this is the most satisfactory rule which has been given to answer the question as to whether a mineral will float or not* Table 8  o  n the opposite page lists some of the  - /  physical characteristic^ of typical sulphides and 43 oxides etc. Inspection of this table shows that all the properties given are shared by both floatable and non floatable minerals and suggests that Taggartfs rule should be amended to state that, "If a mineral has metallic or adamantine lustre it may be floatable"• Unless one were to doubt the validity of the foregoing experiments the only similar surface characteristic which might still exist between floatable and nonfloatable minerals is that of atomic structure* 42. " A M nual of Flotation Proces: es" p* 1 (1921) John Wiley & Son* 43• I>ana» 8 Textbook of Mineralogy, 3rd Edition (1922) John Wiley db Son*  To f a c e pa<?e £ 2 .  FIQ. 13.— Tba  M U n « ( It o  p n a t o f tb » I h w h i U> « uni t o f .1riu-tur # o f group, Hftt,.  Pig. 2* Th e arrangement of the atoms in the unit •£.structure of members of the oaloite group (HCO3J (Am. Jour. Soienoe Vol. 50 p. 3^7)  M  i  Mmmm ^i  i  *  .  3 * * C l L*TTiC i V CHAkCP#V*iT I  Pig. 3« Th e arrangement of the atoms in the unit crystal of chaloopyrite (CuFeS2) Larg e black circles represent copper atoms, ringed circles' iron atoms and small black circles sulohur atoms. (.Jour. Am. Chera. Soc< V. 39 (1?17) ii P« 2518,  j?2.  A search was consequent^ made to discover what data existe d o n the structure of the sul -hides, oxides, and carbonates. Typical space lattice diagrams of the unit crystals of these minerals h; ve oeen reproduced fro m available works «,n  d are shown in /imres 2 -10 .  Fir-vire 2 represents the unit of structure of the calcite grou p and the difference between this structure and that shown in .Fig. j> for chalcopyrite is easily apparent, figure s 4, j? and 6 shov/ respectively the crystal units of pyrite, sphalerite and molybdenite. In the case of these last two minerals it will be seen that the crystal surface consists of both metal and sulphur atoms, and but for the fact that sulphur takes the place of oxygen these minerals are very similar in structure to the oxides of copper, zinc and arsenic shown in Figures 7 -10. (0$>pos>)tz jadcje s 5h+55) If however the compounds formed "ay oxygen and sulphur are considered a difference will at once become apparent. ./he n sulphur combines with oxygen it assumes the role of a metal. Consequentl y if this reaction only is involved the sur:.";.-ce of a sulphide mineral may •* be considered as a metallic surface. Such being the case we have an immediate explanation for the phenomena of flotation, it is this :44. Unles s otherwise indicated these figures have been reproduced from "The structure of Crystals". Wyckoff (1924) Th e Chemical Catalog Company.  To faee page 52*  Fig- 4» Th e unit cub© of pyrite (FeS 2 ). Tke iron atoms are represented by the ringed circles, the atoms of sulphur by the blaek ones* a0 » :U}& x 10-8 0. m ,  Fig. 5. Th e unit cube of sphalerite (ZnS) Either the zine of the sulphur atoms can be represented by the ringed elroles. a0»5.4QxlQ-0'e,Hu  Fig, 4* Th e unit crystal of molybdenite (MoSg) ^h e black circles represent molybdenium atoms A= 1X1Q-°C«H U (Jour. Am. Chem. 30c» Vol, 43 Vl?23) p. 1466.)  53.  "Flotation is the result of a quasi-chemical-reaction. The energy available for this reaction is measured by the affinity which exists between metals and the gases in the air. ,/he n the metallic surface is immersed in water this energy may be augmented by the addition to the water of certain' types of organic compounds. The foregoing statement may be expressed more particularly as follows:All metals tend to-form compounds particularly with oxygen. j=. t ordinary temperatures this reaction proceeds very slowly indeed. Consequentl y the metal holds the gas to its surface and the force of this attraction may be measured by the energy which would be used up if the reaction we're'to take place.. This energy therefore represents a definite amount of potential energy which can he measured as chemical *affinity. A special consideration of flotation upon a basis of chemical affinity is being undertaken in a subsequent paper but an idea of the forces involved may be Judged from the fact that the affinity of one gram molecule of silver for oxygen at 20° C is equivalent to about 2.0 x lofl ergs. The fallowing facts are however very strong arguments in.favor of the chemical affinity hypothesis for flotation.  r  fo face page J52,  i n ig. 7« The u n i t cube of c u p r i t e Cu 2 0. Th e copper atoms a r e r e p r e s e n t e d by r i n g e d c i r c l e s and o::ygen atoms by b l a c k c i r c l e s . a. 0 =~4.32 x 10~8  F i g . 8 . ' The u n i t p o s s i b l e arrangement of c u p r i c oxide GuO. The r i n g e d c i r c l e s r e p r e s e n t , J e i t h e r the copper or oicygen a t o m s . . a-0= 3»74xlu~ c b Q = 4.67 x 10~ u c.m. c 0 - 4.,67 x l O ' ^ c . m .  .54.  First. Flotatior i i s almost entirel y limite d t o th e metals an d sulphide s an d ha s bee n show n t o depen d upo n surface conditions . Investigation ' of th e physica l properties o f floatabl e an d no n floatabl e mineral s ha s failed t o revea l an y propert y which i s alon e particiila r to eithe r class . O n th e othe r han d examinatio n o f th e space lattic e diagram s o f th e uni t crystal s o f thes e minerals hav e indicate d that,considere d fro m th e stand - \ point o f chemica l affinity^ a definit e variatio n exist s j between mineral s whic h ar e floatabl e an d thos e v/hic h are,' not. Second. Thi s theory suggest s a satisfactor y ex planation fo r bot h fil m an d frot h flotation . Third* , It ha s bee n state d tha t onl y organi c compound s with pola r group s ar e o f valu e i n flotation . I f thes e groups ar e considere d i n connectio n wit h th e affinit y hyoothesis i t will b e see n tha t i f chemica l combinatio n were t o tak e plac e betwee n th e minera l an d th e organi c compound th e pola r #rou p woul d b e th e on e with which th e mineral woul d react;&sthes e organi c compound s als o hav e a tendenc y t o concentrat e a t th e bubbl e liqui d interfac e t h e y a r e therefor e abl e t o augmen t th e attractio n whic h exists betwee n th e minera l an d th e ga s bubble. Owin g t o the chemica l compositio n o f most o f th e polar group s which for m par t o f thes e organi c reagent s the y woul d aa^ ;  To f a c e page, J? 2.  iJPig- . 9 . The u n i t c r y s t a l of zinc oxide (znO). The zinc atoms a r e r e p r e s e n t e d by t h e r i n g e d c i r c l e t h e atoms of oxygen by f u l l c i r c l e s . afc = 2.7x 10S< 4 . 4 x 10"°c.m. t h e p a r a m e t e r T i s about j? •o;; -  S**  ^ W 5 3 ^ s\  1^\y^\  J  Aw, .fl  m.  L  <^  \yQzf  ^fSC >^" i ^  1  ^ N-^*'  C*  P«q. 172.—The unit cubo ol lao ntomic nrranjEcnent in cubic Atfi,. For llie reeon.ftraciioD. of ihi« ifomic Rroupintr llin symbols 0 and * JO A »h<n»ld be. replaced by IJ*nJC;lhe«ril<M jf ihc nyni^.U art] (•> l » tiikoa coincident wild the cMi(ci*of the imall cnl»~*, r.anp- l/l;yt r m ] . - . o'l-r.—-nt ill'- arsenic staira; the aroull circk* >r*  Fie; 1 0 . The u n i t cube of the atomic arrangement Of i58t20j Data c o n c e r n i n g t h i s c r y s t a l a r e given i n t h e photographed t y p e . a 0 - 11.0b x 10~°c.rn. Y - + . 21  71 of Abe  expected to have an affinity for only metals and  sulphides, but for other solids as well, and this attraction should show up in the organic compound being adsorbed from solution b^ such solids. Tha t this is the case has been shown in the experiments conducted with aniline and was found to be so by other investigators, whose results are given in the references cited (Hef. Nos.  fourth* I f the supposition is made that oxygen is the active gas on which flotation depends the following points can be explained. (a) Sphalerit e is the least floatable sulphide. The space lattice diagram for this mineral (Fig. 4) indicates that either zinc or sulphur atoms are exposed on its crystal surface. No w both zin« and sulphur are very readily oxidised, and therefore it would seem probable that in this particular case partial combination with oxygen actually takes place and the chemical affinity of the surface for oxygen is reduced. (b) Mineral s ground in air are less floatable J 45 \ than minerals ground under water. y (c) A n emperical rule that increasing the oxygen content of the water improved flotation was used with marked success in applying the De Bavay process for. 45. Writer' s Graduation Thesis, -0.^. University, 1?24.  $6.  treating the ore from the Broken Hill Mine in Australia.x (d) Goa l is floatable because carbon has a definite chemical affinity for both oxygen and nitrogen.  Conclusion.AS far as is known this is the first definite suggestion which has been made to explain flotation on the 'basis of chemical affinity or to apply existing data on crystal structure to substantiate such a hypothesis. A certain similarity exists however between this hypothesis and that presented by L&ngmuir to account for adsorption^. Bu t it might be pointed ou t Langmuir*s work on adsorption was published tv/o years before his own paper on fldtation4"' , and that in this last paper he makes no suggestion whatever that the primary force underlying flotation is a result of the affinity which exists between floatable minerals and the gases in the air. I n fact the weakness of Langmuir*s flotation experiments lay rhdt air i n t h e l i q u i d d r o n s he u s e d d i s p l a c e d • a t t h e A  x  ~  surface  A  of the mineral while in practice when a mineral is floated air displaces liquid. x. note. I am indebted to Mr. H.W. Gepp General Manager of the Australasian ^inc Corporation for this piece of information. 46. "Th e Constitution and fundamental Properties of Solids and Liquids." part s I & II Jour. Am. Ohem. doc. Vol. 38 ii (1916) p. 2221 and Vol. 59 (1917) p. 1B4?. 47. Lo c cit. Hef. no, 2.  .57.  S.S. Dean1'0 in an article published in 1?21 undoubtedly had a somewhat similar idea to that just advanced when he states that flotation is a result of the selective adsorption of gases, but a second article ' by him in 1922 indicates that his work has only lead him to further contact angle experiments similar to those already carried out by Langmuir and that he has failed to discover the reason why this attraction exists. The summary of the results of the present work follows.  481 Kin . Soi. Press. Vol. 122 (1?21) p. 2?1. 4?. Min . Sci. Press. Vol. 124 (1922) p. 410.  58. SUMTA5Y OF RSSUITS AND CONCLUSIONS. 1. A n improved type of direct measuring surface tension instrument has been devised and the method of using it described. 2. Surfac e tension measurements made on emulsions cannot be. employed to estimate adsorption and a reason is advanced to show why this is the case. 3. Adsorptio n results obtained from surface tansi on measurements of solutions may give erroneous values due to a reaction taking place which has been overlooked. 4. Selectiv e adsorption of oils by minerals is not the primary controlling factor in flotation though cases will be found in which selective adsorption of these reagents is a parallel phenomenon to that of selective flotation. 5. Th e forces involved in flotation are surface forces and consequently the relative magnitude of these forces increases as the size of the mineral particle decreases. 6. Th e electrostatic charge normally possessed by a mineral particle is a very small force in comparison with these surface forces and plays little part in flotation. 7. Coppe r sulphate promotes the flotation of  s p h a l e r i t e by filming i t with Cu 2 S. Co 5/  59.  8. Chemica l reactions which take place in concentrated solutions also take place in dilute solutions and produce shell or film changes. 9. Certai n sulphides have been shown to dissolve under the conditions existing in flotation. Th e dissolved compounds formed should he considered as addition agents* 10. A n example of the application of Conclusion "9" Is found in the suggestion that the beneficial action resulting from the addition of H2 SO4 in the treatment of certain zinc ores is due to the copper associated with these ores going into solution as CuSO^. 11. Test s with lime indicate that its value in flotation does not depend upon selective adsorption. 12. Th e crystal structure of floatable and nonfloatable minerals has been considered and points to the conclusion that flotation takes place as the result of chemical affinity*  60.  Acknowledgments.  In conclusion I should like to express grateful appreciation; to Dr. T.C. Hebb, Mea d of the Department of Physics,for his advice and helpful criticisms made in connection with the design of the surface tension & B 4 instrument described in this paper; to Drs. W.F. Seyer and M.J. Marshall,of the Department of Chemistry, for their suggestions as to possible methods for determining adsorption; to,Dr. D.7. Do Image,o f the Canadian Geological Survey, -fr. A.E. Hennings, Department of Physics and Mr. W. Smitheringale, post graduate student in Geology, for their advice as to the applicability of my idea for using crystal structure data in connection with the problem of flotation; to the members of the faculty of Mining and Metallurgy for the assistance which they have at all times been ready to render; and to the Advisory Board of the Bational Research Council of Canada for their award of a Research Bursary in connection with the present work.  61* BIBLIOGRAPHY Articles and Book Dealing with the Theory of Flotation. H.R. Adam, "A Resume of Literature on the Theory of Flotation with Critical Notes". Min . & Scientific Press, Vol. 121 (1920) p. 765. (Contains 28 references.) R.J. Anderson, "The Flotation of Minerals". Trans. A.I.M.E. Vol. 53 (1916 ) p. 327. (Contains a very complete Bibliography up to date of publication (1916) on Colloids and Colloid Chemistry.) W#D. Bancroft* "Ore Flotation". Met . & Chem. Eng. Vol. 14 (1916) p. 251. 0. Batsoh. "Ube r Sohaumsysteme". (Foamsytems) Kolloid Chem. Beihefte Vol. 20 (1924) p. 1-49. (Thi s paper is practically a reprint of that quoted as No* 22 in this list*) n  "Beitra g zur Theorie des Schaumschwimmerfahrens". (A contribution to the Theory of Froth Flotation) Kolloid. Chem. Beihefte Vol. 20 (1924) p. 50-77* (Contain s the results of adsorption test s carried out with oleio acid*)  W.H. Coghill & C O. Anderson. "The Nolecular Physics of Ore Flotation." Jour . Phys. Chem. Vol. 22* (1918) p. zyi. (Deal s with surface tension and contact angles.) f w  "  " "Certai n Interfacia l Tension Equilibria, Important in Flotation." U.S. Bureau of Mines Tec. Paper. No. 262 (1923) (Thi s paper deals mainly with film flotation.)  H.P. Corliss & C.L. Perkins. "Theor y of Ore Flotation." Min. & Scientific Press. Vol. 114 (1917) P* 803. (A general investigation of theoretical aspects.) R.S* Dean, "The theory of Flotation." Min . & Scientifi c Press. Vol. 122 (1921) p. 291.  62.  10.- R.S. Dean & R.A. White, "Adsorption of Gases during Froth Flotation." Min . & Scientific Press Vol. 124 (1922) p. 410. 11.- O.T. Durrell, "Universal Flotation Theory", Met. & Ckem. Eng. Vol. 14 (1916) p. 251. 12.- E. Edser, "Molecular Attraction and the Physical Properties of Liquids." Fourt h Report on Colloid Chemistry, Issued by The British Association for the -advancement of Science (1922) ( A mathematical consideration of adhesional forces.) 13.- " "  "Th e Concentration of Ores by Flotation." Fourth Report on Colloid Chemistry, Issued by the British Association for the Advancement of Science (1922) (Thi s paper, perhaps purposely, gives little definite information.)  14«- A.W. Fahrenwald, "Surface Energy and Adsorption in Flotation." Min . & Scientific Press, Vol. 123 • (1921) p. 227 . 15.- "  1  "Surfac e Reactions in Flotation". Trans . A^m. Inst, of Min. & Met. Vol. 70 (1924) p. 647 . (A summary of this paper is given on page £ of the present text.;  16.- W.D. Harkins, "The Fundamental Principles of Flotation and Adhesion". Hat . Acad, of Science Vol. 3 (1919) P- 569. 17.- I. Langmuir, "The Mechanism of the Surface Phenomena of •flotation". Trans . Faraday Soc. Vol. 25, Part III, p. 64-74. Abstracted . Min. & Scientific Press, Vol. 121 (1920) p. 913. ( A summary of this paper is given on page 2 of the present text*) 18.- F.G. Moses, "Surface Energy in Flotation". Eng . Min. Jour. Vol. 111. (1921) p. 7. 1 9 . - T.A. Rickard, "Concentration by F l o t a t i o n " , McGraw-Hill Co. (1921) (This book i s a compilation of a r t i c l e s which have appeared i n the Min. & S c i . Press up to 1921) 2 0 . - W. Schaf»r t "Adsorption and F l o t a t i o n Capacity of Various Minerals". Met. £££ Ee#. Vol. 21 (1924) p . 401. (Gives r e s u l t s to disprove the hypothes i s advanced bylraube & WIshizawa - see p r e s e n t l i s t No. 23.)  63 •  21.- H.I . Sulraan, "A Contribution to flotation", Min. & .Scientific Press, Vol. 120 (1920) p.p. 14, 47, 122. A criticism of this paper appears under the heading of "A Contribution to the Study of Flotation", Met. & Chem. Eng. Vol. 22, (1920) p. 397. 22«- A.F . Taggart & Gaudin, "Surface Tension and Adsorption Phenomena in Flotation". Trans . Am. Inst. Min. Eng. Vol. 69 (1923) p. 479 ( A summary of this paper is given on page 3 o f the present text*) 23#- J . Tiaube & E. Nishizawa, "Adsorption and Cohesion. A Contribution to the Problem of Flotation". Koll. Chem. Zeitschrift Vol. 32 (1923) p. 383. (A consideration of flotation from the standpoint of selective adsorption.) 24.- P . Vageler, "The flotation Process from the Standpoint of Colloid Chemistry," Met . und Erz. Vol. 17, (1920J p. 113.  Articles and Books on Flotation Practice. 23.- C.C . Freeman, "Froth Flotation at Broken Hillf>. Proo . Aus. Inst, of Min. & Met. No. 36 (1919) p. 89.x 26,- T.J . Hoover, "Concentrating Ores by Flotation". Pub . by the Mining Mag. london (1914) 27.- G.E . lock, "Milling and Flotation", Eng. & Min. J.P. Vol. 119 (1923) P. 109. 28.- W.T . McDonald, "Selective Flotation at Nacozari." Eng. & Min. J.P. Vol. 118 (1924) p. 443. 29•- R.D . Nevett, "Some Controlling Factors in Flotation". Proc. Aus. Inst. Min. & Met. No. 37 (1920) p. 3 3 .* 30.- Th e Staff of the Consolidated Mining & Smelting Co, "The Development of the Sullivan Mine". Can. Inst, of Min. & Met. Monthly Bull. sto. 146, (June 1^24) I am indebted to the Secretary of the Australasian Institute of Mining and Metallurgy for fqrwarding me copies of these papers*  64. 31«- A.P . Taggart, "A Manual of Flotation Processes". Pub. by Wiley (1922) 32.- L.V . Materhouse, "Flotation Practice at Mount Lyell". Proc. Aus. Inst, of Min. & Met. No. 38. (1920) 105.x 33•- Se e also articles at end of No. 19. Surface Tension and Polar Groups. 34«- P.L . Du. Nouy, "A New Apparatus for Measuring Surface Tension". Jour , of Phys. Vol. 1, No. 3 (191?) 35,. H M n "Surfac e Tension of Colloidal Solutions". Phil. Mag. S.6 Vol. 48 (1924) p. 264. 36*- R.B . Elder, "Interfacial tension Measurements and Some Applications to Flotation". Universit y of Idaho Publications. Pamphle t No. 1 (1921) 37.- »  "  "Th e Measurement of Surface Tension". Jour. Phys. Ghem. Vol. 26 (1922) p. 538.  38.- Ferguso n & Allan, "Capillary Constants and Their Measurements", Science Progress, Jan. 1915, p. 428. 39.- W.D . Harkins, "Jour. Am. Chem. Soc. Vol. 39 (1917) p. 39 & 541. (Thes e are the first of a series of articles which appeared in the above journal from 1917 $0 1922. The y deal with surface tension, polar groups and the orientation of molecules.; 40.- Hartridg e & Peters, "Interfacial Tension and HydrogenIon Concentration". Proc . Hoy. Soc. Vol. 101,A (1922) p. 348. 41.- I . Langmuir, "Surface Tension Phenomena", Met. and Chem. Eng. Vol. 15 (1916) p. 468. 42.- "  "  "Th e Constitution and Fundamental Properties of Solids and Liquids", Part II. Jour . Am. Chem. Soc. Vol. 39 (1917) P. 1849.  x# I am indebted to the Secretary of the Australasian Institute of Mining and Metallurgy for forwarding me a copy of this paper.  65. S. Sugddn, "Th e Influence of the Orientation of Surface Molecules on the Surface Tension of Pure liquids". Jour, Chem. Soc. Vol. 125, (1?24) p. 1167. See also Nos. 6; 7; 12; 13; 14; 15; 19; 21. Optical Methods of Analysis. L.H. Adams, "Th e Use of the Interferometer for the Analysis of Solutions". Jour . Am. Chem. Soc. Vol. 57 (1?15; p. 1181• 0. Wolff, "Uber die Messung von Adsorptionsvorganden mit Hilfe des Interferometers." (Measuremen t of Adsorption with the aid of the Interferometer). Koll . Zeit . Vol.. 52, (1923) p. 17. Woodman, ^ookin, & Neath, "Th e Nephelometric~.Determination of Small Amounts of Essential Oils." Jour, Ind. Eng. Chem. Vol . 8 (1916) p. 128. Conductivity Determinations. J.W. McBain & M. Taylor, "Zur Eenntnis der Eonstitution von Siefenlosungen". (Th e Composition of Soap Solutions) Zeit. Phys. Chem. Vol. 76 (1?11) p. 17?. H.C. Jones, Publication s of the Carnegie Inst, of Washington, Nos. 65, 80, 170, 210, 230. The Leeds Northrup Co. Philadelphia, "Th e Measurement of Conductivity of Electrolytes". Catalogu e No. 48 (191?) Reference to conductivity in practical flotation, see No. 28. Addition Agents. H.E. Frederick, "A Basis for Selecting flotation Agents", Eng. Min. J.P. ¥ol. 118 (1924) p. 11. (Th e dlassification given seems unnecessarily muddled,)  66.  M.H. Thomberry & H.-T. Mann. "Th e Effect of Addition Agents in .flotation". Universit y of Missouri. Tec. Bull. Vol. 4, No. 2. (1?17) T. Varley, "The Consumption of Addition Agents Used in Flotation". Can . Chem. & Met. Jour. Vol. J ? (1921; p. 51. Tfre X Ray Analysis of Crystals. W.H. & W.i*. Bragg, "X Rays and Crystal Structure", G. Bell & Sons, London (1?15) W.P. Davey, "Th e Study of Crystal Structure and Its Applications". (Jen . Sleet. Review. Vol, 27 (1?24) p. 742. (Thi s is the first of a series of articles^ R.W. Wickoff, "Th e structure of Crystals". Th e d Chemica l Catalog Company. (1?24) (Thi s book contains a bibliography complete up to the end of 1923.)  1  ( )  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0099564/manifest

Comment

Related Items