A STUDY OF THE EFFECT OF VARIOUS ELECTROLYTES ON ELECTRICAL ENDOSMOSIS JOHN ARTHUR ANDERSON THESIS FOR THE DEGREE OF BACHELOR OF SCIENCE IN CHEMISTRY COLLEGE OF LIBERAL ARTS AND SCIENCES UNIVERSITY OF ILLINOIS 1922 ■ /&2Z An 2.3 UNIVERSITY OF ILLINOIS June 1, lp^2. THIS IS TO CERTIFY THAT THE THESIS PREPARED UNDER MY SUPERVISION BY JOHN ARTHUR ANDERS Oil ent ITLED A STUDY 0F TIIE effects of various ELECTROLYTES ON ELECTRICAL END OSMOSIS . IS APPROVED BY ME AS FULFILLING THIS PART OF THE REQUIREMENTS FOR THE DEGREE OF BACHELOR OF SCIENCI3__ # __ Instructor in Charge Approved : _U_ _ rC _Ur_ _ rLA ULr. — Gc# . >'*- '-1 HEAD OF DEPARTMENT OF CHEHISTEY ■5 — V? Digitized by the Internet Archive in 2016 https://archive.org/details/studyofeffectofvOOande Acknowledgment* The author of this thesis wishes here to express his appreciation to Dr. Emmett K. Carver for continued interest and direction during the study of this problem. . . Contents Page 1. Introduction 1 2. Adsorption 1 3. Adsorption and the Electrical Migration of Colloids. 5 4. Historical 8 5. Experimental A. Apparatus 11 B# Experimental Part • 12 C. Conclusions 15 6. Bibliography 19 DIAGRAM OE APPARATUS 12 GRAPHS 17 18 1 . 1. Introduction. The purpose of this study wrs to rresent the fcctors upon which electrical endosmesis depends , especially adsorption, and to check up experimentally the effect of ions of differ- ent valence. 2. Adsorption. A great many solids and liquids exhibit the property of drawing in tightly to their surfaces other solids and liquids and also gases. This phenomena is known as adsorption. There is only a slight difference between adsorption r nd absorption, and that is in the point of view taken by the user. For in- stance , porous charcoal will take up many times its own vol- ume of some gases. If we merely think of the gases as being drav/n into the mass of the charcoal, we say that they are bei g b sorbed, bait if we think of the gases as being drawn to the pores of the charcoal, and concentrated at the surfaces by some attractive force which the charcoal surface possesses, we call the phenomena adsorption, because the word absorption is too general a term. Examples of adsorption are the removal of gases from high vacuum apparatus, by charcoal, the removal of colored impur- ities from syrups by filtration through bone black, and the adsorption of iodine from sea water by certain sea weeds to which we are indebted for our iodine supply. The last mentioned example illustrates one of the pecul- iarities connected with ads orpt ion, namely : that some sub- stances may "be adsorbed in much greater proportion than some others are. Iodine is present in sea water in very small amounts, and cannot be detected by ordinary qualitative tests, but the sea weed removes considerable amounts of it by adsorp tion, while the plants seem to have little power to take up some of the salts which are present in greater concentration. In the same way, hardly any two substances are adsorbed by a 1 given adsorbent in equivalent amounts. In short , adsorption is specific for any substance , and depends on a number of factors Ions may be adsorbed in the same manner as molecular substances. It is the adsorption of ions from an added salt which is used to cause the precipitation of the troublesome colloidal arsenious sulfide solution which often forms when arsenic is precipitated with hydrogen sulfide in anal sis. The ions of the electrolyte added are adsorbed, and neutral- ize the charges on the colloidal particles of arsenious sulfide. The particles can then get together in masses which are large enough to settle rapidly. The original charge on the particles is supposed to cone from adsorption of ions from an excess of hydrogen sulfide. Ion adsorption is also selective, i. e. , no two ions are necessarily adsorbed in equivalent quantities. For instance, not all salts will cause precipitation of the colloidal sulfide mentioned above , because there is not enough ion adsorbed from them to neutralize the charge ontthe particles of the sol. Some salts will cause partial prec ip it at ion, but ' , ' ’ ■ * - , 5 * none work so well as ammonium nitrate. The ions show an order of adsorption 2 which is almost the same for any t— adsorbent phase 0 . Consia.ai.acle significance is attached to this order of adsorbtion of ions and salts, because of two striking facts. Fir*o, ^Le amount of adsorption varies directly as the valence of tne icn 4 . This has been called Schulze 1 s rule of valence, ilieie are some exceptions to this rule* - , but it is a very use- approximation. Second, the order of adsorption of an ion is generally the order in which the salt decreases the surface tension of the liquid from which it is adsorbed 0 . Some results ox Eaaglund’s for sodium chloride and potassium chloride indicate exceptions to this rule , but perhaps he is wrong, since Taylor 8 gi\ss data for sodium chloride which contradicts E§aglund's. Freundlich* s theory of adsorption is based on this generality, that adsorption from solution varies as the power of the solute to decrease surface tension. Surface energy always tends to decrease, and one way in which it can decrease is by a decrease in the surface tension. Therefore, if the tension between two Phases decreases with increasing adsorption, then adsorption will precede until the tendency of the surface energy to decrease is balanced by forces which tend to promote distrib- ution of the substance adsorbed. These forces may come from several sources. Bancroft says, "An equilibrium will be reached when the change in surface tension is balanced by the difference 4 . of osmotic pressure between the surface film and the mass of the solution"*'. In the case of adsorbed ions, the forces opposing adsorption might also include the repelling force of the charges for each other as they are concentrated by adsorpt- ion" 1 'h Hatschei gives another possible source of charges which might oiDpose adsorption. It is based or ji. theory that whenever two electrical conductors are brought together, there is a re-distribution of electrons bet¥/een the two substances which leaves one charged positively and the other negatively. These two layers of electricity might have an influence in opposing adsorption on account of the tendency of the electrons in each layer to spread outward. Whatever the sources of these forces opposing adsorption may be does not affect the validity of the theory that adsorption consists of the establishment of such an equilibrium. The theory helps one to form a concrete picture of the mechanism of adsorption. 12 lagergren has a different theory of adsorption, based on the formation of layers in compression. Surface layers are thought to be in a state of compression. To this idea lagergren has applied the he Chatelier principle, which states that an increase in pressure displaces any equilibrium in which the volume decreases, and he has arrived at the conclusion that adsorption would be greater, the greater the increase in the density of the surface layer as adsorption goes on. He also concludes that if a substance decreased the density of a solution it would be negatively adsorbed. Sodium chloride proves to be an exception, although various other salts support the theory. 5. 3. Adsorption and the Electrical Migration of Colloids, The particles in most colloidal solutions migrate when in an electric field. Gold, silver, and platinum solutions and suspended particles of shellac, clay, cotton wool, starch, etc. move to the anode in water. Methylene blue, methyl violet, and many hydroxide sols of metals move to the cathode. There- fore, the particles in the former case must he charged negatively and those in the latter case, positively. This phenomenon is known as cataphoresis. The source of the charge which causes the migration is generally considered to he the ions in the dispersing medium (water in the above cases). Air bubbles migrate in distilled water*'", hut do not in turpentine which does not ionize 14 . As to the mechanism by which the ions charge the particles, there are two main theories: the one based on the diffusion of ions and the other on adsorption. The former assumes that some ions in the dispersing medium are able to diffuse into the colloidal phase more readily than others and charge the particles. The experimental basis for this theory is the fact that hydrogen ions are known to be able to penetrate aluminium hydroxide films, whereas hydroxyl ions are not. This theory, however, does not fit all cases. Yfhen the disperse phase is an air-bubble, it would not be consistent to say , in view of our ideas of ions, that the bubble becomes charged by diffusion of ions into the air. , - -> , 6 The theory which does seem to fit all cases is the adsorption theory. According to this theory, ions are drawn to the surface of the particles in order to decrease the inter-facial tension. This charges the particles and leaves the dispersing medium oppositely charged. When the system is placed in an electric field, the two phases have to move in opposite directions. ITernst has added to this simple adsorption theory with his theory of solution pressure of metals. He believes that metals give off ions which leaves the metals charged. The ions, thern- 11 selves, may be partially readsorbed by the metal. If the charged solid material is in the form of an immovable diaphragm or capillary tube, the liquid will be the only movable phase present. This case is known as endosmosis. All the characteristics of cataphoresis apply equally well to endosmosis. The flow nearly always takes place when ions are 16 present in the liquid used. According to Perrin a number of alcohols, nitrobenzene , and most salt solutions give a flow, while non-ionizing substances such as ether, chloroform, benzene, carbon disulphide, petroleum, and oil of turpentine give no flow. Quinke, on the other hand, found that both turpentine and carbon disulphide as well as acetone do give a flow. Perhaps, this can be explained on the basis of impurities in the materials used by Quinke since some of his results conflict with Perrin’s and with Me. Taggart’s experiments with air-bubbles in turpentine. Perrin, however, did find that acetone did give a flow -1 . If we neglect these exceptions, endosmotic flow can be credited to T , V ion adsorption in the same manner as cataphoretic migration, i.e. that some ions are adsorbed by the solid, which becomes 17 charged with a layer of electricity. The opposite ions left in the liquid form a second layer farther from the solid phase so that when the system is placed in an electrical field this outer layer of ions moves and carries water through the capillaries with it. If the charges depend on the adsorption of ions, any change in the nature of the ions available for adsorption should have an effect on the osmotic flow. In general, the addition of ca.tions of higher valence decreases the flo w to the highest extent if movement is toward the cathode. In this case, the anions seem to have no effect. On the other hand, the addition of anions of higher valence has the greatest effect in decreasing the flow if movement is toward the anode. The effects of the addition of H-ions and QH-ions are exceptions to this rule. Both produce effects far greater than would be expected from a consideration of their valence. The experimental part of this study shows some of these effects. , . * . • : It , . ' * . . . 8 4. Historical, he In 1808, while A was investigating the passage of an electric current through a thick suspension of clay. Reuss noticed that the level of the liquid rose in one of the electrode tubes and fell in the other, and that the liquid in the latter became cloudy with fine particles of clay. The other tube remained clear. This was the first recorded experiment on either cataphoresis or endosmosis. Reuss’ work interested a great many other physicists 18 who sought for an explanation of the phenomenon. Porett thought it was something analogous to osmosis, and gave the name "endosmosis” to the passage of a liquid through a diaphrgm under the influence of a drop in potential. At this time it was noticed that electrolytes added to the liquid either increased or decreased the flow of liquid during endosmosis. 19 In 1852, Wiedeman was able to state very clearly the most important generalizations regarding endosmosis. These are as follows: 1. The amount of liquid carried through is proportional to the current. It is independent of the size of the diaphragm for a given diaphragm material. ■ . * 9 2. The hydrostatic pressure developed is proportional to I* (I— Current in Amperes) 3. For a given diaphragm material the hydro- static pressure developed is proportional to the Current (I) and independent of the dimensions of the diaphragm, 20 Several years later, Quincke formulated. laws for the converse of endosmosis where liquid is forced through a diaphragm and an E.M.F. is produced. The statements were the converse of those given for endosmosis. In order to explain endosmosis, Quinke and Helmholz developed the theory of a double layer of electricity in the interface between the solid and liquid phases, i.e., one kind of electricity stuck to the solid, while the other kind stayed in the liquid part of the interface. With a such a double layer given, it is easy to show that in an electrical field, the liquid and the solid will tend to move and go in opposite directions. When the solid is held in place as in a diaphragm, the liquid alone will move. Neither Quinke nor Helmholz could explain why a double layer was formed between liquid and solid, but Perrin came forward with a very plausible explanation. He said that the smaller and more mobile ions crowded to the surface and gave the interface its peculiar properties. Freundlich and Bancroft have enlarged this idea to the adsorption theory. 10 Some attempts have "been made lately to prove that the flow of the liquid in endosmosis is due to a hydration of the 21 ions . If this were so, it would sometimes he necessary for common ions to carry as much as 370 molecules of water in order 22 to account for the flow" . E.W. Washburn has shown that the hydration of Ha, li, K, H, and Mg is less than three molecules of water per ion. It seems, therefore, that the flow through the diaphragm is caused by the layer of moving ions which pushes water through the capillaries. . • ■ • ' ' VC-. •; 11 5, Experimental. jU Apparatus . The apparatus used in this study was similar to that of 04 Briggs , except that the form was modified in order to measure the pressure of endosmosis instead of the flow of liquid. The tubes from which the gauge was made were of Pyrex glass, 3 mm. in inside diameter. This size was chosen because larger tubes required too long a time for filling and smaller ones could not be obtained of sufficient uniformity for use. Pieces of platinum, one cm. square, were used for electrodes. They were placed in small vertical tubes sealed to the horizontal part of the apparatus so that the gases formed at the electrodes might escape from the main body of the liquid and thus be prevented from dissolving in the liquid between the electrodes and changing the conductivity. The substance used for diaphragms was packed into sections of Pyrex tubing 3 cm. long and 2 cm. in diameter, and was held in place by plugs of shredded, quantitative filter paper, about 2 mm. thick. These sections were then inserted into the apparatus in which two Gooch discs prevented the cotton plugs from washing out. The convenience with which diaphragms can be made and inserted is one of the advantages of this type of apparatus. ♦ I ' } . ■ i,*ct fgj 12 . Apparatus for Measuring Endosmotic Pressures. 1. Gauge tubes 2. Pinch Clamp 3. Gooch Pise 4. Rubber Stopper 5. Piaphragm 6. Platinum electrodes IS. The substances which are suitable for diaphragms must be very insoluble. Barium sulphate, for instance, is soluble enough to make it useless. The most satisfactory substances for use are those which, in addition to being very slightly soluble, are composed of hard, smooth grains. With material of this nature , diaphragms can be made which are uniform; whereas, if a soft, fibrous substance such as asbestos is used, the diaphragms are not uniform, because the asbestos can never be packed twice in the same condition. Substances which show stenolysis, or the phenomenon of precipitating a solute in an electric field are to be avoided. 25 Holmes gives a list of materials which produce stenolysis and those which do not. B. Experimental Part. The following tables give the endosmotic pressures obtained with conductivity water and with solutions of electro- lytes. fl) Carborundum Diaphragm and CedfO^lg. Molal Concentration. Pressure in cm. on Cathode Chamber. 00000 39 00048 13.2 00144 2.7 00288 - 3.4 r'Tfn . * * 14. (2) Carborundum Diaphragm and Ba(UOg)g. Molal Concentration Pressure in cm. on Cathode Chamber. .000 39 • 008 24.2 .020 6.5 .028 2.8 (3) Carborundum Diaphragm and NaCtfOs). .000 39 .008 30.5 .016 15.1 .020 13.9 .040 12.3 (4) Asbestos Diaphragm and h 2 so 4 . .000 40.3 • 004 11.5 .008 3.7 •*0195 2.5 (5) Asbestos and HC1. .000 40.3 .002 19.4 .006 4.7 15 Substance (6) Asbestos Diaphragms with Organic Electrolytes Molal Cone. Water Aniline 95% Ethyl Alcohol Pressure in cm. on Cathode Chamber. 40.3 2.4 2.5 C. Conclusions . The curves on Page 17 for Ce(N0g) s v Ba(U0g) 2 , and UaltfOg show very clearly the effect of the valence of the valence of the cation in lowering the endosmotic flow to the cathode. The cerium ion has a far greater effect than either the barium or the sodium causing even a reversal of the flow at a concentr- ation of 0.002 Molal. The barium and the sodium would probably never reverse this flow. According to the theory of the adsorpt ion of ions f the effect may be ascribed to an adsorption of cations by the diaphragm, the cerium ions suffering the greatest adsorption because of their tri-valent condition while, in like manner, the sodium ions are the least absorbed because of their moni-valent character. This furnishes a check on the general valence rule proposed by Schulze. The acid curves on Page 18 show that hydrochloric acid has approximately the same effect on endosmose as has sulphuric. This agrees with the general rule that when endosmose is taking place in the direction of the cathode, anions added have little 16 effect, but cations have a very noticeable effect. Here, the effect of the sulphate ion and chlorine ion is so small in comparison with the effect of the hydrogen ion that the two acids appear to have the same influence. , . ' 19 6. Bibliography. (1) ( 2 ) (3) (4) ( 5 ) ( 6 ) (7) ( 8 ) (9) ( 10 ) ( 11 ) ( 12 ) ( 12 ) (14) (15) (16) (17) (18) Bancroft; "Applied Colloid Chemistry.” p.214-215, (1921) ” It 11 11 IT || |f Freundlich and Hathansohn; Eolloid Z.,2£, 258, Schulze; Z. Prakt. Chem. 25, 43,(1882) & 27, 320,(1884 Wo. Ostwald; Kolloid Z., 26, 69-81. Freundlich; "Eapillarchemie", p.354, (1909) G. von Georgievies; Monatsh., 33, 45-62. Patrick; Z. Physik. Chem., 86, 545-563. Haaglund; Z. Chem. Ind. Kolloide, _7* 21-22. Taylor; "Chemistry of Colloids", p.252, (1920) Bancroft; "Applied Colloid Chemistry", p.130, (1921) Hatschek and Willows; "Surface Tension and Surface Energy", pp. 59-60, (1915) w w if i» it n ti pp. 65-66. Taylor; "Chemistry of Colloids’; pp. 250-252, (1920) Hatschek and Willows; " S. T. and S. E." pp. 73-74. Taylor; "Chemistry of Colloids", p.90. w ii ti it it n Perrin; J. Chim. Phys., 2, 601, (1904) iamb ; Phil. Mag., 25, 52, (1888) Thomson’s Annals of Philosophy, 8 , 74, (1816) 20 (19) Pogg. Ann. » 87, 321, (1852) Tt II 99, 177, (1856) (20) it n 107 , 10, (1659) II II 110, 38, (1860) (21) Am# Chem. J# , 40, 319, (1908) (22) J# PhyS. Chem., 21, £30, (1917) (23) J# AM. Chem. Soc., 36, 322, (1904) (24) Trans. Am. Electrochem. Soc., 31, 260, (1917 (25) J. Am. Chem. Soe., 36,781, (1914)