Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The ternary system. Magnesium sulphate-sodium, sulphate-water, and a method for the separation of these… Gale, William Alexander 1923

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

Item Metadata


831-ubc_1923_a7_g2_t3.pdf [ 9.87MB ]
JSON: 831-1.0059556.json
JSON-LD: 831-1.0059556-ld.json
RDF/XML (Pretty): 831-1.0059556-rdf.xml
RDF/JSON: 831-1.0059556-rdf.json
Turtle: 831-1.0059556-turtle.txt
N-Triples: 831-1.0059556-rdf-ntriples.txt
Original Record: 831-1.0059556-source.json
Full Text

Full Text

e Ternary System gnesium Sulphate -Sodium te-Water  THE  TERNARY  SYSTEM  MAGNESIUM SULPHATE - SODIUM SULPHATE - WATER , AND  A  METHOD  FOR  THE  OF T H E S E  SEPARATION  SALTS  *>*  William Alexander  A Thesis  Gale  submitted (or the D^qree. of  MASTER OF APPLIED SCIENCE in (he Department  of  CHEMISTRY . THE UNIVERSITY •/ BRITISH COLUMBIA APRIL,  1923.  TABLE OF CONTENTS. (1). Introduction. (2). Previous Investigations. (3). Review of Problem. (4). Experimental Methods; - Fig. 1, (Apparatus). (5). Solubility Data; - Table I. (6). Isotherms; - Figs. 2 and 3. (Isotherms). - Plate I. (Solid Model). (7). Determination of Solid Phase;  - Table II.  - Figs. 4 and 5. (Trangular Diagrams). (8). Scheme of Separation; - Fig. 6. (Diagram). (9). Separation Tests; - Test No.l - Fig. 7. (Diagram). - Test No. 2. (10). Commercial Application. (11). Summary. -o«  1. THE  TERNARY  SYSTEM  MAGNESIUM SULPHATE - SODIUM SULPHATE - WATER , and A  METHOD of  for the THESE  SEPARATION  SALTS. -O-  INTRODUCTION. The investigation of this system was undertaken primarily in the hope that a commercial method for the separation of magnesium and sodium sulphates by fractional crystallization might be evolved. Such a process would be of considerable value for the utilization of natural deposits of mixed sulphates, such as occur in British Columbia, and at various places throughout the Prairie Provinces. (See "Alkali Deposits of Western Canada", L.H.Cole, Can. Chem. Jour., 1921, page 171.) The British Columbia deposits occur in the dry belt, and are generally in the form of basins containing small circular masses of solid salts separated by mud-rings. Many of the deposits are dry for the greater part of the year. Investigations have shown that the salts found in these basins consist chiefly of mixtures of sodium and magnesium sulphates in various proportions with, generally, small quantities of sodium chloride and possibly carbonate. An examination of the deposits near Clinton, B.C, and at Kruger Mountain on both sides of the international boundry near Oroville, Wash., was made by L.Reinecke ( Can. Chem. Jour., 1919, p. 209) who states that these particular deposits consist of thin surface layers of fairly pure magnesium sulphate underlain by solid bodies of salt composed of mixtures of sodium and magnesium sulphates almost free from other soluble impurities. The deposits at Basque, B.C. have been described by G.C.Crux (Can. Chem. Jour., 1919, p. 179), and are very similar to the Kruger Mountain deposits. They are composed of a chain of lakes having a total area of twenty acres, and are of considerable depth. Impurities, such as chlorides and potassium salts, amount to less than one percent. The Basque Chemical Production Co. Ltd., of Vancouver, B.C., and the Stewart Calvert Co., of Oroville, Wash., have produced large quantities of refined Epsom salt from the surface layers of the Basque, Clinton and Kruger  2. Mountain deposits. However, attempts to recover pure sodium sulphate have apparently proved unsuccessful. It is to be noted, also, that this particular problem does not seem to have presented itself to any extent when the Stassfurt deposits of Germany were under investigation. There the magnesium sulphate occurs combined principally with potassium sulphate in the minerals kainlte and polyhalite, and as kieserite mixed with chlorides in the carnallite and kieserite regions. PREVIOUS INVESTIGATIONS. The solubility relationships of the double salt astrakanite (MgS04,Ha*SO*,4H*0) were studied by R.W.B.Roozeboom (Rev. Trav. Chim., 6, 335-335; and Chem. Abstracts, 1888, p. 1164.) who determined the transition point for MgSO*7Hj.O + Na^SO^lOHiO *• Astrakanite as 22°C, at which temperature 100 moles of water dissolve 2.85 moles of Na.SO* and 4.63 moles of KgSQ*. This corresponds to a solution'containing 14.6$ NaaS0«. and 20.2% MgS04, which is in good agreement with the present work. He also showed that the solubility of astrakanite hardly varied at all with temperature. Extensive investigation was carried out by van't Hoff and his co-workers on the extraction of salts from sea water. Accounts of this work were published by van't Hoff in two books entitled,"Zur Bildung der ozeanischen Salzablagerungen," (Braunschweig Vieweg) in 1905 and 1909, and also in "Uber die Bildungsverhaltnisse der ozeanischen Salzebazerrungen" (Leipzig Verlagsgesellschaft, 1912). The portion of his work having the most bearing on the present problem, was the determination of the solubility relationships of the complex system MgSO^-NazS04- MgCla.-NaCI- KC1 for 25° C and 83* C-. A general outline of these publications is given by J.H.Hildebrand (J.I.E.C.,10, 96, 1918), who attempted to apply the information to a process for the extraction of KC1 and other salts from sea water bittern in California during the latter stages of the war. With regard to the salts of sodium and magnesium, he constructed an equilibrium diagram for the system MgS0« - MgClA- N&*S0«- NaCl at 25* C, from data given by Roozeboom, van't Hoff and Seidell (Dictionary of Solubilities),upon which he based his suggestions. It was proposed to precipitate NaCl and MgSO, Hj.0 (kieserite) by evaporation at higher temperatures, leaving a mother liquor that could be treated by the carnallite process for the recovery of KC1 and the remaining MgSO*. He stated that MgCla apparently had a dehydrating effect, allowing kieserite to be stable in contact with solution at lower temperatures. Thus, by taking advantage of the relatively low solubility of kieserite, the NaCl might be dissolved out of the precipitated solid by a cold solution of MgCl». Such a process would require the presence of sulphates and chlorides  3. in nearly equivalent proportions. He recognized, however, that the formation of astrakanite at ordinary temperatures, and of ldwite (MgS04 ,NaaSO* ,|HZ0) and vanthoffite (MgSQ,3Na*SCD at higher temperatures might possibly render this scheme useless. Prom the work of van't Hoff, it would appear that lowite forms above 43'C and vanthoffite above 46°C, while these two double salts can co-exist at about 60°C. The essential points of the Hildebrand equilibrium diagram were re-determined by W.C.Blasdale (J.I.E.C., 12, 164, 1920) who found that kieserite does not form a part of the system at 25*C, and could not be precipitated even upon inoculation. Previous to this, Blasdale did considerable work on the system Na.SO*- NaCl - K 4 S0 # - KC1, (J.E.I.C, 10, 344 & 347, 1918.) He determined the isothermal relations at various temperatures, and applied this information to suggested processes for the separation of each pair of salts. The only ones apparently capable of giving satisfactory results, were the chlorides. With these the scheme consisted of alternate evaporation with the separation of NaCl, and cooling with the crystallization of KC1. REVIEW Of PROBLEM. In order to put the separation of sodium and magnesium sulphates on an exact basis, it is desirable to have a knowledge of their solubility relationships over the entire range of temperature from the freezing to the boiling points of their solutions. The data given by Hildebrand, Blasdale and Seidell concerning mixtures of these salts, apply only to a very few temperatures. Therefore, the main problem presenting itself was the determination of the solubility relations over a wide range of temperature, together with the identification of the solid phases in equilibrium with the solutions. Owing to the large number of gravimetric determinations necessitated by the solubility investigations, it was not found possible, in the time available, to make examinations of the solid residues at all the points investigated. In a few doubtful cases, experiments were made to determine the solid phase, using Schreinemakers's well-known "residue" method (Z. physik. Chem., 11, 76). However, this procedure was not found very satisfactory for distinguishing astrakanite (MgSOp jNa^SO^^H^O) from lowite (MgSO, ,NaxS0* ,|H»0) since there is no wide difference in their compositions. Difficulties due to supersaturation were overcome, as far as possible, by inoculation with small amounts of paj*t.i._ ally dehydrated mixtures. Any possibilities of the formation of "mixed crystals" or solid solutions of isomorphous salts, were neglected throughout. Thermometers were compared with instruments standardized by the United States Bureau of Standards.  Thcrntomtftr  lemptraTurt  confral  \ G*s  Gas &},»,„;,,.;>>?,</>">}}>>>>*)*}'>?•'•••?!,•,'">,.  • ..:'••.„  FIG. Offcfion Tube  1.  of Thermo sfcf anci  N  Showing  Sftrrer  in  I 7© fdl0W paye  3 .)  Safurafiny place .  4.  The sulphates used were "Baker's Analyzed Chemicals" which were found to contain about 0.00b%> chlorides, but free from other Impurities beyond a trace of Iron. EXPERIMENTAL METHODS. Various mixtures of Na»SO* ,10^0 and KgS0*,7Kt0 were vigorously agitated with small amounts of distilled water In test-tubes Immersed to within a few centimeters of the top In a gas heated thermostat which was kept constant to 0.2*of the desired temperature. For the determination of the Isotherm at 0*C, the bath was filled with melting snow. The type of stirring apparatus used, as shown In Flg.l, was that recommended by Flndlay ("Practical Physical Chemistry" 1920, p. 302.). Two saturating tubes were usually operated concurrently. After agitation for about six hours, the stirrers were removed and the solid allowed to settle, during which time the tubes were stoppered to prevent evaporation. Samples of 1 o.c. of the clear solution were then drawn off by a pipette (previously warmed If necessary) through a cotton wool filter. These samples were weighed In the usual type of weighing bottles, fitted with ground glass stoppers. After dilution the solutions were analyzed as follows. Magnesium was determined by precipitation as MgNH^PO* in the cold, the precipitant being added to the neutral solution which was then made strongly alkaline with NH«0h. The precipitate , after standing several hours, was filtered and washed with NH*0H solution, ignited and weighed as Mg„PAOr. The total sulphate was precipitated and weighed as barium sulphate, a separate sample being used in each case. The sodium and water were then calculated by difference. All determinations were made in duplicate, but only the average values of concordant results are given in the tables. It was not considered necessary to re-determine the solubilities of single salts. Free use was made of information obtainable from Seidell (Dictionary of Solubilities, 2nd.Ed., Van Hostrand,, 1919), Comey and Hahn (Dictionary of Chem. Solubilities, 2nd. Ed., Macmillan Co., 1921), Roozeboom (loc. clt.) and Blasdale (loc. cit.). The original authors of these data have been noted in the following tables. All values have been expressed on the basis of percent by weight, as this is the most useful form, especially where the information is to be applied to a commercial process. SOLUBILITY DATA. Abbreviations used to indicate Solid Phases; Mg-12. . . .MgS0»,12Ht0. Mg-7 . . . .tegSO, ,7H,0, (Epsom Salt).  5,  Mg-6 |fig-l Na-10 Ast. Low. Vant.  . . MgS0^,6Hi0. . . MgSO«.,H*Oj ( K i e s e r i t e ) . . . Na«SOf , 1 0 H X 0 , ( G l a u b e r ' s S a l t ) MgSO^ j N a ^ S O f l , 4 R \ 0 , ( A s t r a k a n i t e ) .  MgSQ* ,NaxSO,. ,%B.afi, (LSwite ). . MgS0„ ,3Naz SO*, (Vanthof fite).  Referencesj Mulder,-"Schlekundige Verhandelungen en Ouderzockingen", v.3, pt. 2, p.52& 123, 1864. Smith, - Landolt & Bornstein's "Tabellen", 4th. Ed., p. 481, 1912. Diacon,-"Jahresberlcht uber die Fortschrltte der Ghemie", 1866, p. 61. Blasdale, - loc. cit. Roozeboom,- loc. cit. TABLE I.  •  Isotherm at 0' C. Exp.No.  foMgSO,  ^Na^SO^  0.0 (Mulder) 62. 9.65 11.25 63. 20.0 64. 20.0 65. (Interpol- td)20.0 (Mulder) 21.2 (Diacon) 19.75  4.58 4.89 4.69 4.63 4.3 0.0 0.0 3.97  %H ,0 95.42 85.46 84.06 75.37 75.7 80.0 78.8 76.28  Solid Phase * Na-10. • • • •  Na-10, Mg-7(?) Mg-12(?) • •  Mg-7. (not stated)  0 In most cases the Solid Phase was assumed from the position on the Isotherm. Isotherm at 10*C. Exp.No.  %MgS0^  (Mulder) 13. 14. 34. (Mulder)  0.0 0.75 19.4 20.9 23.9  foNa^SO^ 8.3 8.3 7.4 7.46 0.0  *H*0  Solid PHASE.  91.7 90.95 73.2 71.64 76.1  Na-10. • • • •  Na-10, Mg-7. Mg-7.  6. Isotherm at 18.7*0 Exp. No. (i). 1. 3. 6. 7. 8. 9. 10. 11. 12. (i).  foMgSO*  0.0  15.05 14.68 13.90 13.50 12.30 12.30 11.80 11.48  1.48 8.92 11.30 15.35 15.70 20.57 20.60 24.4 25.7 26.2  ^H 2 0  %Ne.z S0 #  3.4 0.5 0.0  84.95 83.84 77.18 75.20 72.35 71.90 67.63 67.92 72.2 73.8 73.8  Solid Phase. Na-10.  Na-10, Mg-7 Mg-7. I  o  a  o  • •  (i). . Interpolated from Mulder. Isotherm at 25'C. Exp. No.  #MgS0f  (Mulder)  0.0  16.  0.25 16.6 21.15 19.8 22.6 27.8  (Blasdale) • •  21. 22. (Mulder)  #Na„ SO* 21.8 21.2 17.8 13.0 16.9 10.8  0.0  %KM0  78.2 78.55 65.2 65.85 63.3 66.6 72.2  Solid Phase. Na-10. • *  Na-10, Ast. Ast., Mg-7. (Na-10, Mg-7) Mg-7. • »  Isotherm at 30° £i Exp. No.  0tgSO*  (Mulder)  0.0  18.  6.48 12.3 15.8 23.25 24.5 24.55 29.0  (Roozeboom) »•  23. 26. 25. (Mulder)  ^Na^ SO* 29.0 26.08 23.25 18.6 12.35 9.27 8.75  0.0  ^H z 0 71.0 67.44 64.45 65.6 64.4 66.23 66.7 71.0  Solid Phase. Na-10. •  »  Na-10, Ast Ast. Ast., Mg-7 Mg-7. * 9  • I  Isotherm at 40° C_L Exp. No.  #MgS0„  #Na* SO*.  (Mulder)  0.0 2.54 3.61 11.0  32.8 30.70 30.70 24.7  60. 61. 51.  %EZ0  67.2 66.76 66.99 64.3  Solid Phase. Na^SO*. •  •  • • •  •  e 00 a;  d  p  «  a  <t!  •  •  • l> i bO •  ID  P cd  £ £ in e  0>  Xi P O DO H  <••  (1)  ro CO  » o CO bp  O S3  *  ij  sis.  S3  o^ a R  CO  t  jd •X,  • 35 fc» i -H  K -P  •>P CO CO  T3 N  •H  O  H  •  *  H o  •  •  «<  «  o  09  •  I D ID 0> CD CO tO CV CO  •  cO^tOtDiO^iOtOCO cOCDCQcDcDcocOcOCD  ID ID lOWCOHCOcOCOiOO  ^(O(0d)noo>coo  W W W H H H  ID I D CDOJ-^tOOiCOrHOtO HCVCVJLDo^iDCOrH H H H H C J W W W t O  d  PH  •  u  !*)  s  CMCOtDO-*cot>CO'd LDcDLOlDiDiDt>C^iH  ^-^  o  03  o  w ;rj  51 * o 01  M as S3 S<  o aft,  • o  S3  •  H  Q,  Cj-  o CO  d  P to  >  P • 03 P  -— «=c; eo * <: • o •P CO p  -El  »  CCS CO S3  >  S3 * 6 — cS ai  H  w  P » - en o P 03  <  CD I • c O f  & O id  03 ID ID ID ^ 03!>^tOLOtOlDrHOCT)tOLDiDLn^cD^OCOI>lD COCO£>COlOC\lW>v}<l0^iDLDiD'st,02^tOtOWcOcO cDcOci)(iDcocOcD(£icDcOCDcDCDCDCDcX)cDCDCDcDcD  o  td  o to ID lO iD ID C0 C 0 C ^ O t 0 C 0 O C 0 C 0 O C - t 0 O < H r H t > r H r H C - O O O  d  &•  p  H O O c D t D i n t O H O i C O O O c D c O H O l M O i n c O l O O t O ( O l O W ( M « 0 3 0 J H H H H H H  P  Q d 01  o  H  *  a>  to  cd d  PH  >d  •H rH  0  CO  H  O  *  P  SZ  xl-l  CO > M Cti  d • * >s O • co • * P « • * a • o a •  ai  ft  CD  & o  •  •  id  •  SH  o!  CO  p  o • • ^» • «  LO  coa>r-ia>cDC\}03coo  CO<DCQlOlDlO«st<^CQ cOcOcQCQcDcQCQcOcO  •  «  •  •  •  •  •  « » • • • • • •  •  LD ID CMOOOOcDCOCOcDOO  .3  d  0CV2t>LOcD^HCMCO T3cOCOCOCOCOI>C*-tO  OcDcocOaOO'^iDCD H H iH  •  • ^ lO CD OOOrHCOOiOCDCM  •  O CO  H O O M f l L O ^ H C R i n W M N N W M N H H  •< at  S3  %*  < * •  O  55  m -.<  00 ID LO ID O L n c D ^ C D t ^ - C - r H O t O - ^ ^ ^ ^ C O t O ^ L O r H t D l D OWWCPa>WtO"^CDcOc£»COCO'^C^COOiHrHrHtO H H H H H H H H O J W C M l O t O l O t O t O  o  bO  d  a)  & CO  i ^G0a>c0C0C\}(00>C^rHlDK)c0Ma>^Oa>OC<t p -d ,-i i D i D c D - ^ C O ^ t O ^ ^ C - ^ t - C O ^ - ^ ^ c O C ^ ^ rH H  g  *5  -_-  8.  Exp. No. 29. 30. 37. 31. 35. 38. (Mulder)  foWlgQOq.  JENa»S0« 12.0 9.6 6.5 4.6 3.55 1.92 0.0  23.1 25.9 30.0 33.8 33.8 34.5 35.5  I sotherm a t Exp. No. (Mulder) 89. 103. 102. 73. 74. 104. 105. 107. 108. 161. 122. 123. (Smith)  %MgS0„  #Na k SO*  0.0 4.29 13.9 14.2 14.55 14.7 15.4 16.4 18.1 24.06 33.25 36.3 38.4 38.6  30.4 28.1 22.4 22.25 22.3 22.0 21.4 19.7 17.85 12.6 6.25 4.9 2.8 0.0  %E^0  64.9 64.5 63.5 61.5 62.65 63.58 64.5  Solid  Phase.  Low. •• « 9  Low., Mg-6. Mg-6. .. .»  8 0 ' ll. %B.t0  Solid  Phase.  69.6 67.61 63.7 63.55 63.15 63.3 63.2 63.9 64.05 63.34 60.4 58.8 58.8 61.4  Na z SO*. Na^SO, , V a n t . ( ?) Vant. • 0  V a n t . , Low. Low. . • .. *. . • . • . •  Mg-1. ••  I s o t h e r m a t IOC)°C. E x p . No. (Mulder) 126. 163. 119. 118. 124. 125. (Seidell,?) (Mulder)  £MgS0 4 0.0 13.7 13.75 16.9 23.2 31.1 32.0 40.6 42.5  f o Na A S0*  %Ht0  29.8 20.5 20.6 19.4 12.55 6.5 6.07 0.0 0.0  70.2 65.8 65.65 63.7 64.25 62.4 61.93 59.4 57.5  Solid  Phase.  Na^SO*. Vant. 9  «  V a n t . , L o w . ( ?) Low. .. ••  Mg-1. (Mg-6).  Miscellaneous Data. Ref.  temp.  £MgSOf  %Na,S0*  (1). (2). (3). (2).  -2.9° -3.9 -6.0 1.8  13.9 19.0 19.0 21.1  0.0 0.0 0.0 0.0  Solid Phase. Ice. I c e , Mg-12. ( I c e , Mg-7.) Mg-12, Mg-7.  FIG.2. (to follow pdge  Percenf  &.)  N f l 2 SO,  Fie. 3 . (to  follow  Pcrccnr  Fij.Z.)  NdiSQo  PLATE I. I To follow  Tin  Fig. 3 . J  PHASE RELAI KINSHIPS .* M C . M U N . ^ 1 IN COHTACT VMTH THEIR o  10  19  JO  W  »VATt« SW  PHOTOGRAPH •/ Showing  ^  fhe  Saturation  Sfatfe  Solid  SYSTEM  C*  SOUTIOKS . r*  iw  *•  »#  li^Hinnwf  SOLID MODEL Fields  of  fhe  Phases  McS0 4 -NA z S0 4 -H a O.  9. Ref.  Temp,  #MgSO<  ^Na^SO*  Solid Phase.  (4) (5) (6)  32.383" 48.4 68.0  0.0 (33.0) 37.0  (33.61) 0.0 0.0  Na-10, NaAS0, . Mg-7 , Mg-6. Mg-6 , Mg-1.  (U-  de Coppet; Ann. Chim. phys. (4), 25, 528-532,1872. (2)- Cottrell et al; Sitzber. k. Akad Kiss. Berlin, 1901 y P . 1035, 121 1223; 1922, (3)- Weston; J. Chem. Soc (4)- Richards & Wells; Z i physik. Chem., 43, 471, 1903. Dickenson & Muller; J « A « W • O • , cc? ,1318, 1907. (5)- Carpenter & Jette; J.A.C.S., 45, 578, 1923. (6)- van't Hoff; Sitzber i k. Akad. Wiss , Berlin, 1901, p. 1035. ISOTHERMS. The percentages of NaiSO* were plotted against percentages of MgSO, to give solubility isotherms which are shown in Figs. 2 and 3. A break in a curve indicates a change in the solid phase in equilibrium with the solution at that point. As the change of direction was very slight in some cases, a certain amount of doubt exists as to the transition points, particularly between the double salt portions, and between vanthoffite and Na»S0?. The magnesium sulphate portion of the isotherm at 0*C. is also uncertain since it is doubtful as to which hydrate was present. However, these cases are of little importance from a practical, standpoint, as they do not affect the general shape of the isotherms. Temperatures were then plotted along a third axis at right angles to give a solid model (Plate I) which-shows the saturation areas of the stable solid phases. The extent of the vanthoffite field is at present uncertain, but a more intensive application of Schreinemakers's method, together with the determination of intermediate isotherms would probably fix this area. Dilatometer methods might also yield definite results, by taking advantage of changes in volume on passing from one solid phase to another. Such variations would, no doubt, be fairly distinct, since R.Gorgey (Min. petro. Mitt., 28, 334; & Abstr.A.C.S., 4, 1281, 1910) states that density determinations may readily be used to differenciate between these double salts. DETERMINATION of SOLID PHASE. Schreinemakers's "residue" method is based on the fact that if the composition of the solid with adhering mother liquor is plotted on a triangular diagram, this point must lie on the straight line joining the points representing the compositions of the liquid and solid phases. Thus, when  Schreinemakers\s Method. Isotherm  at  FIG.  4 0 ° C.  4.  Graphic DeferminqT/on of Composition of InUrmediaU Ifo  follow  Solid . pCl%<2  9 J  Sch  Mefhod. Isofherm af  FIG. Dcferminafion {To fdUw  50C.  5. of  Solid Phases.  FiJ. q-.j  10. at least two different solutions are analyzed together with their wet residues, the composition of the pure solid may be found by exterpolatlon. It must be remembered, however, that this method Is only applicable when the solid consists of one phase of constant composition. Mixtures were made up so as to obtain points at intervals along that portion of the isotherm under investigation. Analyses were made of both the saturated solutions and of the solid residues with adherent mother liquor, and the results plotted on triangular diagrams. The straight lines Joining each pair of values were produced to intersection, this point representing the composition of the pure solid. The results obtained from these experiments sre contained In Table II, while the triangular diagrams illustrating the graphical determinations of the solid phases at 40*C. and 50 C. are shown in Pigs. 4 and 5. It is to be regretted that time did not permit the completion of this branch of the work. However, sufficient determinations were made to identify the double salts in their proper order along several of the Isotherms. TABLE II. Solution. Temp. $MgS04 #Na A S0* %UK0  ifiet R e s i d u e . ?bMgSOf p a , S O ,  %hl0  Solid  Phase.  40*C.  13.15 21.60 24,10 19.20  22.80 13.45 10.80 15.00  64.05 64.95 65.10 65.80  19.3 25.2 27.6 23.8  27.2 22.6 18.3 23.07  53,5 52.2 54.1 53.13  50*0  13.3 18.2 21.3 27.8 32.8 30.4  23. 0 16. 0 13. 5 9.75 2.51 6.6  63.7 65.8 65.2 62.45 64.69 63.0  18.8 22.2 25.2 29.5 40.0 31.8  28.2 21.9 21.7 16.3 2.2 16.25  53.0 55.9 53.1 54.2 57.8 51.95  80*C. 13.85 33.25  22.55 6.35  63.60 60.40  17.6 34.0  40.7 15.8  38.3 50.2  100*C. 13.75  20.60  65.65  17.40  42.75  39.85 Vanthofflte.  Astrakanlte  Ast.,m.stable MgS0.6H.O. Ldwite. Vanthofflte. Lttwite.  SCHEME of SEPARATION. A method for the separation of these two sulphates was suggested from a study of their Isothermal relations at  11. lower temperatures. This scheme depends on the fact that belcv 22*C. the two solids, MgSQ, ,7ht0 and Nat;>0# ,10n,0, can coexist in equilibrium only with solutions whose compositions fall along the line AB in Fig. 6, such a system being monovarient. This line represents the boundry between the Epsom and Glauber's Salt areas in the solid model.  Thus on cooling a solution of original composition a n below its saturation temperature, GlauberTs salt will separate out along the line "ab". This may be removed by rapid filtration or centrifuging, while the mother liquor of • composition "b" is then concentrated at higher temperatures to point "cn which would give a saturated solution at say 40* C. On cooling to about 22* C , we would have Epsom salt separating along the line "cd". The mother liquor may then be diluted so as to bring the concentration well within the Glauber'8 salt field before being oooled further for the separation of more sodium sulphate. Similar cycles of operations may be repeated until the solution has been reduced to a very small volume. The process may be commenced from any point on either side of the line AB, and the composition of tbe solution may be brought into the cycle by dilution, concentration, or partial separation of one of the salts. Concentration might also be brought about by the addition of a mixture high in magnesium sulphate. w  12. SEPARATION TESTS. This scheme was put to test on a laboratory scele, and the results were entirely satisfactory. The methods used in controlling the process during operations were fairly simple, and would , on doubt, prove adequate on a lsrger scale with a few slight improvements. Knowing the composition of the original solution, changes in concentration during evaporation and dilution were followed by observing the volume changes. These were plotted on a chart (Fig. 7) showing isotherms for each alternate degree between 0°C. and 22°C.(obtained by interpolation) together with sets of lines along which the salts would separate on cooling. Samples were taken at various points, and were later analyzed. In Test No. 1, a solution having the original composition of 17.1% HgSQ* and 9.2% NaaSO*, was cooled to 2° C. with frequent stirring. The solid separating out was removed on a Buchner filter and sucked as dry as possible, but not washed. The analysis of this solid is shown as Sample A in the following table, while B represents the mother liquor. The "Excess H 4 0" is the difference between the total water and that necessary for the crystallization of the major salt; while the "% Separation" shows the amount of the major salt contained in the total anhydrous material. T e s t No. JU Sample.  %'MgSO^  %Na4 SO*  %iixO  A. ( s a l t )  4.62 18.55  36.10 5.40  59.28 76.05  0.89 10.40 39.40 5.40  B.  (m.l.)  C. ( s a l t ) D.  (m.l.)  45.25 21.70  E. P.  (salt) (m.l.)  3.70 17.60  % Excess HxO 13.58  % Separation, 88.7  53.86 67.90  6.51  98.0  56.90 77.00  7.00  91.5  Solution B was evaporated to two-thirds of its volume (i.e., to a concentration of 27.8% MgSO* and 8.1% Na^SO*), and then cooled to 20*G. The solid C was removed as before, without washing; while D represents the composition of the mother liquor, which was then diluted with one-fourth its volume of water, giving a solution of about 16.5% ffigS0<, and 13.9% Na^O*. This was cooled to 2°C. and the solid E separated as before, while the final mother liquor is shown as F, which is very nearly the same as solution B.  13.  ay  39%  The second test was made on a mixture of salts approximating 25# Glauber's salt and 75% Epaom salt, this being a rough average for many parts of the Basque deposits. 200 grams of the mixed hydrates were dissolved in sufficient water to give a saturated solution at 40*. C. The process was carried through two complete cycles, and the solids separating out were removed,as before, without washing. The results are shown In the following table. Test No. 2. Sample.  #MgS0, *Na,S0#  JtHiO  26.75  7.80  65.45  48.00 22.00  0.74 9.90  51.26 68.10  98.5  Diluted % & cooledr Salt, to 0*C. \U. L.  3.76 20.40  37.60 4.40  58.64 75.20  91.0  Evaporated '/3 & f Salt. f Salt cooled to 20*.1 C.( M.L.  44.40 23.00  2.20 7.62  53.40 69.38  95.3  Original Solution Cooled to 20*C. f  {..  Salt. Liquor.  % Separation.  14. Sample. Diluted '/6 3c f Salt. cooled to 0°C. I M.L.  %MgSO^ %NazSO« %Et0 5.70 19.95  % Separation. 37.80 56.50 87.0 4.80 75.25  The high percentages of MgS0« in the Glauber's salt fractions in Test No.2 are in part due to insufficient dilution previous to their separation. This is shown by the fact that the composition* of the mother liquors are almost identical with the values found for the quadruple point for O ' C , at which both solid salt hydrates can precipitate on cooling. From these preliminary separation tests, it is obvious that the scheme is quite feasible provided the process can be properly controlled. By washing the separated salts with a small amount of cold water, or with a concentrated solution of the respective salt, a 99% separation could easily be obtained, while a single recrystallization would give practically pure products. COMMERCIAL APPLICATION. In applying this scheme on a commercial scale, it is suggested that fresh solution may be added in each cycle so as to keep the volume at any point fairly constant, thus making it a continuous process. If the Glauber's salt in the original mixture is less than about 25%, the operation might even be carried out without evaporation by adding the solid mixture directly to the system in order to concentrate the solution. The accumulation of impurities would necessitate discarding a certain volume of the mother liquor at regular intervals. One of the main difficulties of such a process would be the accurate control of temperature. Artificial cooling would probably not be economical, in which case the scheme could only be operated during the winter months in localities where moderately cold weather is customary, and where fuel is cheap. It is intended to continue the separation tests in order to determine the effect of such impurities as would be met with in practice; e.g., chlorides and potassium salts. It is also hoped that an opportunity may arise for conducting tests on a much larger scale. In conclusion, I wish to express my sincere appreciation to Dr.E.H.Archibald for his advice and encouragement, and also to the "Honorary Advisory Council for Scientific and Industrial Research" for their kindness in granting a Bursary which made the carrying out of this work possible.  15. SUMMARY. (1). A study has been made of the solubility relations of mixtures of magnesium and sodium sulphates at the following temperatures; 0*, 18.7', 25', 30', 40* , 50*, 60* , 60' and 100' C. (2). The solid phaseB were determined In a few doubtful cases, and a solid model was constructed to show the saturation fields of all stable solids In contact with solutions. (3). A scheme for the separation of sodium and magnesium sulphates was evolved, and tests made on a laboratory scale proved quite satisfactory, giving fractions of salt as high as 98.5% pure without washing. (4). Suggestions are offered whereby a continuous process might be developed for the recovery of pure Epsom and Glauber's salts from certain natural deposits of mixed sulphates.  >_o—0--o—  Chemical Laboratory, University of British Columbia, Vancouver, B.C.  


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



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"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items