ullmann - potassium compounds

� 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Article No : a22_039

Potassium Compounds

HEINZ SCHULTZ, Kali und Salz AG, Kassel, Federal Republic of Germany

G€uNTER BAUER (RETIRED), Kali und Salz AG, Kassel, Federal Republic of Germany

ERICH SCHACHL (RETIRED), Kali und Salz AG, Kassel, Federal Republic of Germany

FRITZ HAGEDORN, Kali und Salz AG, Kassel, Federal Republic of Germany

PETER SCHMITTINGER, H€uls Aktiengesellschaft, Werk L€ulsdorf, Niederkassel, Federal

Republic of Germany

1. Introduction. . . . . . . . . . . . . . . . . . . . . . . . 640

1.1. Occurrence . . . . . . . . . . . . . . . . . . . . . . . . 640

1.2. History. . . . . . . . . . . . . . . . . . . . . . . . . . . . 640

2. Potash Salt Deposits. . . . . . . . . . . . . . . . . . 641

2.1. Minerals. . . . . . . . . . . . . . . . . . . . . . . . . . . 641

2.2. Geology of Potash Deposits . . . . . . . . . . . . 641

3. Mining of Potash Salts . . . . . . . . . . . . . . . . 647

3.1. Shaft Mining . . . . . . . . . . . . . . . . . . . . . . . 647

3.2. Extraction, Conveying, and Haulage . . . . . 647

3.3. Solution Mining . . . . . . . . . . . . . . . . . . . . . 648

4. Treatment of Potash Ores . . . . . . . . . . . . . 649

4.1. Intergrowth and Degree of Liberation . . . 649

4.2. Grinding . . . . . . . . . . . . . . . . . . . . . . . . . . 650

5. Potassium Chloride . . . . . . . . . . . . . . . . . . 652

5.1. Properties . . . . . . . . . . . . . . . . . . . . . . . . . 652

5.2. Production byCrystallization fromSolution 652

5.2.1. Phase Theory . . . . . . . . . . . . . . . . . . . . . . . 652

5.2.2. Hot Leaching Process . . . . . . . . . . . . . . . . . 655

5.2.3. Processing of Carnallite . . . . . . . . . . . . . . . . 658

5.2.4. Equipment. . . . . . . . . . . . . . . . . . . . . . . . . . 662

5.3. Flotation . . . . . . . . . . . . . . . . . . . . . . . . . . 665

5.3.1. Potash Ores Suitable for Flotation . . . . . . . . 666

5.3.2. Carrier Solutions . . . . . . . . . . . . . . . . . . . . . 666

5.3.3. Flotation Agents . . . . . . . . . . . . . . . . . . . . . 666

5.3.4. Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . 667

5.3.5. Flotation Equipment . . . . . . . . . . . . . . . . . . 668

5.3.6. Processes . . . . . . . . . . . . . . . . . . . . . . . . . . 669

5.4. Electrostatic Separation . . . . . . . . . . . . . . . 671

5.4.1. Theoretical Basis . . . . . . . . . . . . . . . . . . . . . 673

5.4.2. Equipment and Processes . . . . . . . . . . . . . . . 673

5.5. Heavy-Media Separation . . . . . . . . . . . . . . 675

5.6. Debrining and Drying . . . . . . . . . . . . . . . . 676

5.7. Process Measurement and Control . . . . . . 677

5.8. Waste Disposal and Environmental Aspects 678

5.9. Granulation . . . . . . . . . . . . . . . . . . . . . . . 680

5.10. Quality Specifications. . . . . . . . . . . . . . . . . 682

5.11. Toxicology and Occupational Health . . . . . 683

5.12. Economic Aspects and Uses . . . . . . . . . . . . 683

6. Potassium Sulfate . . . . . . . . . . . . . . . . . . . 685

6.1. Properties . . . . . . . . . . . . . . . . . . . . . . . . . 685

6.2. Raw Materials . . . . . . . . . . . . . . . . . . . . . . 685

6.3. Production . . . . . . . . . . . . . . . . . . . . . . . . . 686

6.3.1. From Potassium Chloride and Sulfuric Acid

(Mannheim Process) . . . . . . . . . . . . . . . . . . 686

6.3.2. From Potassium Chloride and Magnesium

Sulfate . . . . . . . . . . . . . . . . . . . . . . . . . . . . 687

6.3.3. From Potassium Chloride and Langbeinite . . 689

6.3.4. From Potassium Chloride and Kainite . . . . . 689

6.3.5. From Potassium Chloride and Sodium Sulfate 689

6.3.6. From Potassium Chloride and Calcium Sulfate 690

6.3.7. From Alunite . . . . . . . . . . . . . . . . . . . . . . . 690

6.3.8. From Natural Brines and Bitterns . . . . . . . . . 691

6.4. Granulation . . . . . . . . . . . . . . . . . . . . . . . . 691


6.6. Toxicology and Occupational Health . . . . . 691


7. Potash–Magnesia . . . . . . . . . . . . . . . . . . . . 692

8. Production of Potassium Salts from Other

Raw Materials . . . . . . . . . . . . . . . . . . . . . . 692

8.1. The Dead Sea. . . . . . . . . . . . . . . . . . . . . . . 693

8.2. The Great Salt Lake . . . . . . . . . . . . . . . . . 693

8.3. Searles Lake . . . . . . . . . . . . . . . . . . . . . . . 694

8.4. Other Sources . . . . . . . . . . . . . . . . . . . . . . 694

9. Storage and Transportation . . . . . . . . . . . . 694

10. Analysis of Potassium Compounds. . . . . . . 695

11. Potassium Hydroxide . . . . . . . . . . . . . . . . 695

11.1. Properties . . . . . . . . . . . . . . . . . . . . . . . . . 695

11.2. Production . . . . . . . . . . . . . . . . . . . . . . . . . 695



12. Potassium Carbonate . . . . . . . . . . . . . . . . 696

12.1. Properties . . . . . . . . . . . . . . . . . . . . . . . . . 697

12.2. Production . . . . . . . . . . . . . . . . . . . . . . . . . 697

12.3. Quality Specifications and Analysis . . . . . . 699

12.4. Storage and Transportation . . . . . . . . . . . . 700


13. Potassium Hydrogencarbonate . . . . . . . . . 701

13.1. Properties and Production . . . . . . . . . . . . . 701

References . . . . . . . . . . . . . . . . . . . . . . . . . 701

DOI: 10.1002/14356007.a22_039

1. Introduction

1.1. Occurrence

Potassium occurs in nature only in the form of itscompounds. It is one of the ten most commonelements in the earth’s crust. Several widelydistributed silicate minerals contain potassium,in particular, the feldspars (! Silicates) andmicas (! Mica). The weathering of theseminer-als produces soluble potassium compounds,which are present in seawater and occur in ex-tensive salt deposits. Potassium is important inthe metabolism of plants and animals, and istherefore found in ash from plant materials andin the bodies of animals. Potassium compoundsare obtained almost entirely by the mining of saltdeposits.

1.2. History

Before the discovery and exploitation of potassi-um salt deposits, the production of potassiumcompounds consisted almost entirely of potash(K2CO3) obtained from natural sources such aswood ash, residues from distilleries, Bengal salt-peter, wool grease, and mother liquors from seasalt production. Quantities were small and wereused only for the production of soap, glass, andexplosives.

In 1840, J. VON LIEBIG laid the foundations ofthe theory of mineral fertilizers in his paper ‘‘TheApplication of Organic Chemistry to Agricultureand Physiology’’ (! Fertilizers, 1. General,Chap. 2.). This spread the knowledge that potas-sium was one of the most important plant nutri-tion elements. In 1851, some deepmineworkingsin Stassfurt struckminerals containing potassiumandmagnesium, although these could not be usedas fertilizers. In 1861, ADOLPH FRANK started thefirst plant using the process that he developed forproducing from carnallite a potassium salt thatcould be employed as a fertilizer. This soon led tothe development of other processes and to theestablishment of many potash mines and works.Many attempts were made up to the end ofWorldWar I to hinder the building of new factories inGermany, which had a world monopoly in theproduction of potash fertilizers, and attemptswere made to ration production and supply, andto regulate prices. In spite of this, 69 factories

were in existence in 1910, and 198 in 1918.WhenWorld War I ended, Alsace was returned toFrance, and the potash works that were builtthere shortly after 1900 became French property,so that Germany lost her monopoly in potash.

The German national assembly of 1919 en-acted the potash regulations and the so-calledclosure order, which reduced the number ofoperating factories to 29 by the year 1938. Thisalso caused production to be concentrated in afew large potash companies [17, 18].

After the end of World War II, ca. 60% ofGerman production capacity went to the area thatwas later to become the German DemocraticRepublic, and became the VEB Kombinat Kali.After German unification in 1990, the newlyformed Mitteldeutsche Kali AG took over theseoperations. In the Federal Republic of Germany,several groups of works were formed, leading in1970 to the formation of the company Kali undSalz AG following a series of amalgamations.This company owns seven potash works, alllocated in the former Federal Republic ofGermany [19, 20].

Most of the French potash works in Alsacewere nationalized after World War I. DuringWorld War II, production was greatly increased.Several of the deposits have now become ex-hausted, andmany of theworks have been closed.Production has been concentrated in two largeworks, although these are likely to be closeddown soon after the end of this century [21].

Potash production in Spain began in 1926withthe start up of a factory in Catalonia. Otherfactories were established later both there andin Navarra.

The kainite deposits in Sicily have beenworked since 1959–1960; the salts are used forpotassium sulfate production [22].

The Soviet Union began potash production in1931 at a plant in the Northern Urals. In 1939,plants that had been operating in Eastern Polandsince 1920 were taken over by the Soviet Union.In 1963, the first plant was started to exploit avery extensive deposit in White Russia. The CIStoday has several large operations in the UralsandWhite Russia, giving it the largest capacity ofall potash-producing countries.

In the United States, potash production beganduring World War I because the United Stateseconomy could no longer buy German potashfertilizers. Potassium salts were obtained from

640 Potassium Compounds Vol. 29

Searles Lake in California and in northernNebraska. Most of these operations ceased pro-duction in the 1920s. After the discovery ofpotash deposits in the area of Carlsbad, NewMexico, a large number of potash works werefounded since 1931 [23]. Since the foundation ofthe Canadian potash industry, which has theUnited States as its main outlet, the number ofworks in theCarlsbad region and their productionrate have continually declined. New potashworks were started in southern Utah at Moab in1964 and at the Great Salt Lake in 1970 [24].

The most important potash deposit in NorthAmerica was discovered during World War IIin Saskatchewan, Canada. After initial pro-blems due to its great depth and the presenceof water-bearing overlying rock, which wasdifficult to deal with, several potash works werefounded in the early 1960s, and today Canada issecond only to the CIS as a potash producer.Two more potash plants were established in the1980s in New Brunswick on the east coast ofCanada.

In 1986, a potash plant began production atSergipe in Brazil.

In 1974, a potash mine was opened in York-shire, United Kingdom.

Potash production in Palestine began on thenorth bank of theDead Sea in 1931. Potash plantshave now been producing since 1952 at thesouthern end on the Israeli side and since 1982on the Jordanian side.

Potash ores are treated today by three basicprocesses: leaching–crystallization, flotation,and electrostatic treatment. Gravity separationis of minor importance because of the smalldensity differences between salt minerals.

The oldest process is leaching–ocrystalliza-tion. In this process, salt solutions were origi-nally cooled in open vessels. Vacuum cooling ofthe solutions in crystallizers was introduced in1918 in the United States, so that much lessenergy was required and cooling times werereduced from days or weeks to minutes. Flota-tion was introduced in 1935 in the United States.This proved so efficient, especially for the treat-ment of sylvinite ores, that it is now the mainpotash treatment process worldwide. The elec-trostatic process was first used on a large scale inthe German potash industry in 1974. It is nowwidely used in Germany for treating complexhard salts.

2. Potash Salt Deposits

2.1. Minerals

Potash salt deposits were formed by the evapo-ration of seawater [25]. Their composition isoften affected by secondary changes in the pri-mary mineral deposits. More than 40 salt miner-als are now known, which contain some or all ofthe small number of cations Naþ, Kþ, Mg2þ, andCa2þ; the anions Cl� and SO2�

4 ; and occasionallyFe2þ and BO3�

3 , as well. Most of these are listedin Table 1 [26].

The more important salt minerals are halite,anhydrite, sylvite, carnallite, kieserite, polyha-lite, langbeinite, and kainite. Gypsum occurs atthe edges of salt deposits and in the overlyingstrata. Bischofite, tachhydrite, glauberite, thenar-dite, glaserite, and leonite occur additionally insome deposits [4, 27, 28].

Other minerals, not described in detail here,are useful in elucidating difficult geologicalquestions with regard to the origin of salt depos-its. In special geochemical investigations, smallamounts (ppm) of Rbþ and Csþ in place of Kþ;Sr2þ replacing Ca2þ; Mn2þ replacing Fe2þ; Br�

replacing Cl�; etc., are important [29, 30]. Theindividual minerals can be identified microscop-ically (grains or thin sections) and by X-rayanalysis.

Potassium salt deposits always consist of acombination of several minerals (Table 2). TheGerman term Hartsalz (hard salt) refers to thegreater hardness of sulfate-containing potashminerals in potash deposits.

2.2. Geology of Potash Deposits

Potash deposits occur worldwide in almost allgeological systems. The most important depositswere formed in the Devonian, Carboniferous,Permian, Cretaceous, and Tertiary periods [4,25, 27, 31–37]. All major potash deposits are ofmarine origin. Bodies of seawater became iso-lated from the open ocean when bars formedunder the water surface, and under arid climaticconditions, the seawater became concentrated,finally depositing the dissolved salts. The impor-tant feature is that exchange between normalseawater and concentrated salt solution generallydoes not occur. During sedimentation, the less

Vol. 29 Potassium Compounds 641

Table 1. Principal salt minerals

Mineral CAS registry no. Formula Crystal system/

crystal class

Refractive indices na : nb :

ng or nopt. : nC ;

optical activity

Density,

g/cm3

Hardness

(Mohs)

Anhydrite [14798-04-0] CaSO4 rhombic 1.570 : 1 .575 : 1.614 2.96 3.8

D2h-mmm opt. þAscharite [13768-64-4] Mg2[B2O5] �H2O monoclinic 1.575 : 1.646 : 1.650 2.70 3

C2h-2/m opt. �Astrakhanite [15083-77-9] Na2Mg[SO4]2 � 4 H2O monoclinic 1.483 : 1.486 : 1.487 2.23 3

C2h-2/m opt. �Bischofite [13778-96-6] MgCl2 � 6 H2O monoclinic 1.495 : 1.507 : 1.528 1.60 1.5

C2h-2/m opt. þBoracite [1303-91-9] Mg3[Cl/B7O13] rhombic 1.662 : 1.667 : 1.673 2.95 7

(stassfurtite:

lumpy form

of boracite)

C2v-mm 2 opt. þ

Carnallite [1318-27-0] KMgCl3 � 6 H2O rhombic 1.467 : 1.475 : 1.495 1.60 2.7

D2h-mmm opt. þD’Ansite [12381-13-4] Na21Mg[Cl3(SO4)10] cubic 1.489 2.65

Td-43 m

Epsomite [14457-55-7] MgSO4 � 7 H2O rhombic 1.432 : 1.455 : 1.461 1.68 2.5

D2-222 opt. �Glaserite [13932-19-9] K3Na[SO4]2 trigonal 1.491 : 1.498 2.70 2.7

D3d-3 m opt. þGlauberite [13767-89-0] CaNa2[SO4]2 monoclinic 1.515 : 1.532 : 1.536 2.85 3

C2h-2/m opt. �Gypsum [13397-24-5] CaSO4 � 2 H2O monoclinic 1.521 : 1.523 : 1.530 2.32 2

C2h-2/m opt. þHalite [14762-51-7] NaCl cubic 1.5443 2.168 2.5

(rock salt) Oh-m3m

Kainite [1318-75-2] (KMg[ClSO4])4 � 11 H2O monoclinic 1.494 : 1.506 : 1.516 2.13 3

C2h-2/m opt. �Kieserite [14567-64-7] MgSO4 �H2O monoclinic 1.518 : 1.531 : 1.583 2.57 3.7

C2h-2/m opt. þKoenenite [12252-18-5] [Mg7Al4(OH)22] trigonal 1.55 : 1 .58 2.15 1

[Na4(CaMg)2Cl12] D3d-3 m opt. þLangbeinite [14977-37-8] K2Mg2[SO4]3 cubic 1.534 2.83 4.2

T-23

Leonite [15226-80-9] K2Mg[SO4]2 � 4 H2O monoclinic 1.479 : 1.483 : 1.488 2.20 2.7

C2h-2/m opt. þL€oweite [16633-52-6] Na12Mg7[SO4]13 � 15 H2O trigonal 1.495 : 1.478 2.34 2.5–3

C31-3 opt. �Mirabilite [14567-58-9] Na2SO4 � 10 H2O monoclinic 1.394 : 1.396 : 1.398 1.49 1.7

(Glauber’s salt) C2h-2/m opt. �Polyhalite [15278-29-2] K2MgCa2[SO4]4 � 2 H2O triclinic 1.548 : 1.562 : 1.567 2.78 3–3.6

C1-1 opt. �Rinneite [15976-45-1] K3Na[FeCl6] trigonal 1.588 : 1.589 2.35 3

D3d-3m opt. þSchoenite [15491-86-8] K2Mg[SO4]2 � 6 H2O monoclinic 1.461 : 1.463 : 1.476 2.03 2.6

C2h-2/m opt. þSylvite [14336-88-0] KCl cubic 1.4903 1.99 2

Oh-m3m

Syngenite [13780-13-7] K2Ca[SO4]2 �H2O monclinic 1.501 : 1.517 : 1.518 2.58 2.5

C2h-2/m opt. �Tachhydrite [12194-70-6] CaMg2Cl6 � 12 H2O trigonal 1.520 : 1.512 1.67 2

D3d-3m opt. �Thenardite [13759-07-4] Na2SO4 rhombic 1.471 : 1.477 : 1.484 2.67 2.7

D2h-mmm opt. þVanthoffite [15557-33-2] Na6Mg[SO4]4 monoclinic 1.485 : 1.4876 : 1.489 2.69 3.6

C2h-2/m opt. �


soluble salts were deposited first, and the mostsoluble salts last. Inmost salt-forming sea basins,this process was repeated many times, resultingin cyclical salt formation [36, 38].

A complete salt deposition cycle begins withbasic carbonates (limestone, dolomite, andsometimes magnesite), followed by sulfates(gypsum and anhydrite), rock salt, and finallypotassium and magnesium salts (Table 1). In-termediate layers of clay sediments are the

result of repeated influxes of fresh water, whichcan lead to partial redissolution of the saltdeposits. Also, eolian (airborne) transportationinto the basin can occur. A recent example of asalt deposition basin is provided by the Kara-Bugas, a lagoon on the eastern side of theCaspian Sea.

In the course of the earth’s history, the depos-ited salts underwentmany changes, often leadingto the formation of a modified mineral constitu-tion. The original formations were affected byrelatively low-temperature thermal and hydrau-lic metamorphoses, and sometimes by volcanicaction, producing the mineral compositions thatexist today (Table 2) [39, 40].

Rock movements have changed the originalhorizontal stratification in many salt deposits.Saltmigration, folding, and upwardmovement toform diapirs have led to tilting of the potashlayer, sometimes to an acute angle; to thinning;or to local accumulation. As a result of move-ments, salt deposits often came into contact withgroundwater and were partially or completelyredissolved.

Table 3 lists the most important potash saltdeposits, showing the main potashminerals pres-ent, the geological systems, and the geographicallocations. Table 4 shows the distribution of pot-ash deposits, with estimates of the reserves [4, 37,41].

Table 2. Marine salt rocks: mineral constituents*

Salt rock Main components Secondary components

Rock salt Na A, Po, Ki, La, clay

minerals, etc.

Anhydrite A Na, Dol, Mag, gypsum,

Sy, C, clay minerals, borates

Carnallitite C, Na Ki, Sy, A

Sylvinite Sy, Na C, Ki, A

Hard salt

kieseritic Na, Ki, Sy C, A, La, Po, borates

langbeinitic Na, La Ki, A, Sy, C

anhydritic Na, Sy, A Ki, C

Kainitite Na, Kai Ki, Sy, C

Bischofitite Bi, C and/or Ta Na, Ki, Po, A, borates

Tachhydritite Ta, Bi C, Na

Claystone Clay minerals,

quartz, mica, A,

Dol, Mag

Na, Sy, C, Bi, Ta, La,

coenenite, rinneite

*Abbreviations: A ¼ anhydrite; Bi ¼ bischofite; C ¼ carnallite;

Dol ¼ dolomite; Kai ¼ kainite; Ki ¼ kieserite;

La ¼ langbeinite; Mag ¼ magnesite; Na ¼ halite; Po ¼polyhalite; Sy ¼ sylvite; Ta ¼ tachhydrite.

Table 3. Important potash deposits, mineral compositiona, geologic ageb

Country Geological age

Rect Plei Tert Cret Jura Perm Carb Dev Sil Camb

Australia B

Brazil SC

Canada S SC

CIS KS LKSC SC S S

Congo SC

Germany HSC

Ethiopia SCK

France S

United Kingdom S

Israel B

Italy K

Jordan B

Netherlands C

Spain S

Thailand CS

United States B SL S SC S

aS ¼ sylvinite; C ¼ carnallite; H ¼ hard salt; L ¼ langbeinite; K ¼ kainite; B ¼ brine.bRect ¼ recent; Plei ¼ Pleistocene; Tert ¼ Tertiary; Cret ¼ Cretaceous; Jura ¼ Jurassic; Perm ¼ Permian; Carb ¼ Carboniferous;

Dev ¼ Devonian; Sil ¼ Silurian; Camb ¼ Cambrian.


The concept of a reserve presupposes thepossibility of economical extraction, which de-pends on the presence of a useful potash contenttogetherwith usable quantities of othermaterials.Other important factors are the type of deposit, auniform and usable seam thickness at a depth thatis neither too great nor too low (water problems),economic workability, the possibility of solutionextraction, and above all, proximity to consumersand profitability. Losses incurred during the ex-traction process (10–20%) must be deductedfrom the total size of the reserve.

The two largest known potash deposits in theworld, which are in Saskatchewan (Canada) andWhite Russia, are of Devonian origin. The Perm-ian deposits (Germany, United States, and CIS)were for a long period the classical salt depositsand were the most important potash reserves, butthese lost their economic importance afterWorldWar II.

However, the known potential of extractablepotash deposits is so large that theworld supply isguaranteed for many hundreds of years.

Cambrian. In the central part of theAngara-Lena Trough, ca. 800 km north-northeast ofIrkutsk (CIS) in the vast eastern Siberian basin,an evaporite seriesmore than 2 km thick containsa potash deposit. A sylvinite series up to 28 m

thick is situated at a depth of 600–900 m in theinterior of this Nepa potash basin. At the edges itmerges into carnallitite, with over- and underly-ing carnallite–halite. The potash content of thebest sylvinite seam, up to 13 m thick, is 19–30%K2O, with an extremely low MgCl2 content [4,27, 37].

Silurian. In the center of theMichigan basinis a sylvinite zone having a total thickness of28 mwithin a 900-m-thick evaporite series of theSalina group covering an area of 34 000 km2. Asolution-mining pilot plant (see Section 3.3) is inoperation [4, 37].

Devonian. Canada. The western Canadianbasin between the RockyMountains and HudsonBay accomodates eight Devonian salt-bearingunits, of which the Prairie Evaporite is the mostimportant both in extent and in economic impor-tance, having four potash seams, partly sylviniticand partly carnallitic. The upper limit of thePrairie deposit is inclined to the south, beginningat a depth of 600 m in the northeast and ending at>3000 m in the southwest. Of the total area (ca.200 000 km2) of the potash seams, ca.50 000 km2 is extractable bymining. Other areascan only be exploited by solution mining. Theclay-bearing sylvinites, each 3–5 m thick, con-tain 25–30% K2O, but no potassium magnesiumsulfate. At present, ten plants are in operation [4,27, 37].

United States. The potash deposit describedabove, which is of great importance for Canada,stretches across the border toMontana and NorthDakota. The potash-bearing area covers morethan 30 000 km2, but because it is at a depth of>1000 m and up to 3500 m, extraction by con-ventional mining is impossible [4, 37].

CIS. The potash deposits of Soligorsk andStarobinsk lie within the Pripyat marshes,25 000 km2 in area, ca. 120 km south of Minskand 800 km from the Baltic harbor of Ventspils.Potash extraction began in 1963, and outputreached 4.2� 106 t of K2O by 1980. The UpperSalt is 3200 m thick and contains ca. 60 potashzones in the upper half. These can be groupedinto four workable potash-bearing formations.Each of these consists of interbedded layers ofsalt clays, rock salt, sylvinite seams, and some

Table 4. Minable potash reserves in units of 106 t K2O

Canada, Saskatchewan 4500–6000

(conventional mining only)

New Brunswick 60–80

United States 100–150

Brazil 10–40

Chile and Peru 30–50

Congo ca. 20

Germany 400–800

United Kingdom 30–50

Italy 10–20

France ca. 20

Spain 20–30

CIS 2000–3000

Dead Sea (Israel and Jordan) 100–200

China 10–100

Thailand up to 160

Laos up to 20

World* 7500–10 000

* If reserves only extractable by solution mining are included

(particularly those in Saskatchewan, but also including those in

the CIS), the figure for the total minable reserves increases by a

factor of 4 or 5. Other deposits of potash salts are either

unimportant or of minor local importance compared with the

above figures. These exist in Australia, Ethiopia, Iran, Libya,

Morocco, Poland, and Tunisia.


carnallitite. Only stage II (thickness: 1.8–4.4 m,K2O: 17.7%, insolubles: 5%) and the lower partof stage III (thickness: up to 2.8 m, K2O: 13.4–16.4%, insolubles: 9%) are extracted. Potash ismined at depths of 350–950 m, where levelseams with only slight deformation exist [4,28, 37].

Carboniferous. United States. The potashdeposit of the Paradox formation at Moab, Utah,is extracted by Cane Creek Mine. The valuablesylvinite deposit is at present extracted by solu-tion mining. The extent of the salt deposit is ca.25 000 km2, ofwhich 15 000 km2 is potash bear-ing. It contains about ten potash seams, of whichfour are workable, containing sylvinite (25%K2O) and some carnallitite. The salt has beenformed by tectonic action into anticlines, some-times diapiric, stretching fromnorthwest to south-east [4, 27, 37]. Because the deposit is undulating,conventional mining is complicated and madehazardous by the presence of oil and gas.

Canada. In the eastern provinces of Canadaon the Atlantic coast (i.e., New Brunswick, NovaScotia, andNewfoundland) are a number of smallsedimentary basins with evaporites of the LowerCarboniferous Windsor group. Gypsum, anhy-drite, and slightly to strongly folded rock salt arepresent, along with several mainly tectonicallybounded small structures containing potash salts.In New Brunswick a high-quality sylvinite seamoccurs that contains 25–30% K2O up to 40 mthick. At present, potash is extracted by PCA(Potash Corporation of America) 7 km north ofSussex at a depth of 460 –760 m, and by PMC(Potacan Mining Corporation) 25 km south-southwest of Sussex at a depth of 600–1000 m[4, 27, 37].

Permian. The Permian period was one ofimmense salt accumulation, which took place inthree vast evaporite basins: (1) the Central Euro-pean Basin, (2) the East European Basin, and (3)the American Midcontinental Basin, as well asnumerous smaller sedimentation areas in Europe,Asia, North and South America, and the Arctic[27].

1. The Central European Basin extends fromYorkshire to Central Poland and Lithuania,and from the River Main in northern Germany

to thenorthernpart of theNorth Sea [42]. In theZechstein, seven sedimentation periods aredistinguished, of which the three lowest–theWerra, Stassfurt, and Leine series – are ofeconomic importance for potash extraction.Germany. In the Werra and Fulda areas, theHessen and Thuringia potash seams of theWerra series are mined (hard salt and carnal-litite in level deposits at a depth of 400–1000 m with a thickness of 2–5 m, K2O: 9–12%). The Stassfurt potash seam of the Stass-furt series is mined in the Harz–Unstrut–Saalearea (hard salt and carnallitite at a depth of500–1000 m and a thickness of 5 m, K2O:20%). The potash seams Ronnenberg andRiedel of the Leine series are mined in theHanover area in salt diapirs (sylvinite in in-clined deposits, depth: 350–1500 m, thick-ness: 2 –40 m, K2O: 12–30%). Hard salts ofthe Stassfurt series are at present extracted toonly a minor extent. Finally, potash is minedon the Massif of Calv€orde near Zielitz (depth:600–900 m, Ronnenberg sylvinite inclined at<18–25�, thickness: up to 10 m, K2O: 14–20%) [4, 37].United Kingdom. In Yorkshire, a level depositof carnallitic sylvinite is extracted, which cor-relates with the German Riedel seam bothpetrographically and stratigraphically (aver-age thickness: 5 m, K2O: ca. 27%, depth:1100 m) [4].The Netherlands. In northeast Holland, thecarnallitic potash salts of the Leine series havebeen drilled at various places. The plans are toextract carnallitite by solution mining [4].Po-land. The Zechstein series in Poland correlateswith that in Germany. The carnallitic forma-tions of the Stassfurt seam have been drilled,and the Leine deposits also contain somecarnallitite and sylvinite.

2. The Eastern European Basin extends from theBarents Sea to the Caspian Sea and covers anarea of 1.5� 106 km2.CIS. Potash depositsare known in several parts of the westernfoothills of the Urals from the Pechora De-pression to the Orenburg District, and in theregion of the Caspian Depression. The mostimportant reserve is in the area of Solikamskand Berezniki on the upper Kama, 200 kmnorth of the town of Perm. It has been minedsince the late 1920s. Potash seams occur in theupper part of a salt series>400 m thick that is


part of the Kungurian. These seams are syl-vinitic below and sylvinitic–carnallitic above,interlayered with rock salt. The lower section,consisting of four sylvinite seams of totalaverage thickness 20 m, with a K2O contentof 15–20%, is mined at a depth of 300–500 min predominantly level deposits [4, 27, 37].

3. The Midcontinental Basin of North Americastretches from North Dakota to Texas but isbroken up into numerous smaller basins. Hor-izons of rock salt are present in almost thewhole of the Permian system. The depositionof potash in large quantities apparently tookplace only during a small part of the UpperPermian period in a subregion of theDelawarebasin of New Mexico and Texas [27].United States. In Carlsbad, New Mexico, syl-vinite and langbeinite are mined in the 50 –140-m-thickMcNutt potash zone of the Saladoformation in five out of a total of eleven potashseams up to 3 m thick in a level deposit at adepth of 250–575 m. The sylvinite contains12–30% K2O, and the langbeinite 7–9% K2O.Massive and disseminated polyhalite, carnall-ite, kieserite, kainite, leonite, glauberite, andthenardite are also present [4, 27, 37].

4. Brazil. In the Amazon basin at Nova Olinda,150 km south of Manaus, drilling of a leveldeposit of Lower Permian sylvinite at a depthof 1000–1150 m has been carried out. Thethickness is up to 4.5 m, and the K2O content17%. Extractable reserves amount to ca.35� 106 t of K2O [4, 37].

Jurassic. Turkmenistan/Uzbekistan. Up tosix seams of sylvinite can be found at a depthof 200–1200 m in the Gaurdak formation in theareas of Gaurdak–Tyubegatan, Kugitang, andOkuzbulak; and carnallite may be found at Kar-lyuk. The seams of sylvinite, up to 6 m thick,contain 12.5–25%K2O, sometimes with as muchas 12% insoluble material. A deposit at Karshiin Uzbekistan is to be exploited, and solutionmining on a pilot scale has been started atKarlyuk [4, 37].

Cretaceous. Along the east coast of Braziland the west coast of Africa from Gabon toAngola, Lower Cretaceous potash-bearing eva-porites occur. Their similar formations indicatethe existence of a common sedimentation areabefore the two southern continents began to drift

apart during the tectonic movements of the Cre-taceous period [4, 37].

Brazil, Sergipe. The Taquari-Vassouras de-posit of the Muribeca formation has been minedsince 1985. It has a basal layer of tachhydrite,above which is a thin layer of carnallitite and twoseams of sylvinite, 10 m and 8 m thick, contain-ing 15% K2O, with a rock salt bed 4–10 m thickbetween the two seams. The reserves of sylviniteamount to 525� 106 t, with 16 000� 106 t ofcarnallitite [4, 37].

Congo. Salt deposits have been formed inten sedimentation cycles, of which the secondhighest contains four sylvinite seams. The othercycles contain largemasses of carnallitite, and, asin Sergipe, tachhydrite.Mining of the sylvinite inthe third layer (depth of 370 m, 3–16% K2O)began in 1969 but ceased in 1976 when the minebecame flooded in the space of two days [4, 37].

Thailand. The lower part of the Maha-Sarakham formation below the Khorat Plateaunortheast of Bangkok consists of evaporiteswith a potash sequence up to 90 m overallthickness at depths of 90–530 m. The K2Ocontent of the carnallitite, which can be 15 mthick, is usually ca. 10%, but can occasionallyreach 14%. Sylvinite, with a maximum thick-ness of 3.7 m and a K2O content of 18%, hasbeen found in only a few of the 48 explorationboreholes, with tachhydrite in some locations.Test mining has begun in the concession areas,which cover 300 km2 [4, 37].

Tertiary. France.The LowerOligocene de-posit in Alsace contains two sylvinite seams in amarl–rock salt series. The upper layer has athickness of up to 2 m and contains 19–25%K2O; the lower, up to 5.5-m-thick layer, with15–23% K2O, also contains 15% insolubles(clay, anhydrite, and dolomite).Mining is carriedout at comparatively high rock temperatures at adepth of 500–1000 m in flat or slightly inclinedseams that have been disturbed by faults [4, 37].

Spain. Deposits are located in two areas ofthe Ebro basin. In Catalonia and Navarra, potashsalts lie above rock salt. These deposits are up to10 m thick in Catalonia and up to 15 m inNavarra. Above this occurs an interbedded


deposit of rock salt, carnallitite, marl, and anhy-drite. Only the sylvinite seams A and B aremined. These are up to 4 m in total thickness ata depth of 1020 m, some deposits being level andsome inclined. The crude salt contains 12.5–14%K2O [4, 37].

Italy. The Miocene evaporites starts with ananhydrite–polyhalite–halite series followed byrock salt with up to six kainite interbeds and aroof of kieserite–sylvinite. Since 1976, only thekainite seams, which are up to 32 m thick, andcontain 12–14% K2O, have been mined at ca.800-m depth in seams inclined at 25–65� [4, 37].

CIS. In the eastern Galician foredeep ofCarpathia, potash salts have been known since1804 (mainly langbeinite and kainite, with a littlesylvinite) and have been mined since 1864 atStebnik and Kalusch. The individual lenticulardeposits occur in five horizons of 1–5 m thick-ness, the kainite containing 10–12%K2O and thesylvinite 8–19%. The deposit is inclined at 45�[4].

Pleistocene. In the Danakil depression,Ethiopia, is a salt basin of ca. 40� 100 km,100 m below the surface. Boreholes have re-vealed a potash zone containing sylvinite, carnal-litite, and kainite at a depth of ca. 740 m [4].

Salt Lakes and Subterranean Brines.Israel and Jordan. Potassium chloride, magne-sium chloride, rock salt, soda, chlorine, andbromine are obtained from the Dead Sea [4, 37].

United States. Potassium and sodium sulfateare extracted from the Great Salt Lake [4], [37].Potassium and magnesium chloride are obtainedfrom subterranean brines from the Great SaltLake Desert [4]. Potassium chloride and sulfate,sodium borate, and boric acid are obtained fromsubterranean brines in Searles Lake [4, 37].

Australia. Langbeinite is obtained from sub-terranean brines at Lake McLeod.

China. Tsaerhan Lake, a dry lake, ca.1100 km northeast of Lhasa, the largest salt lakein the Chaidamu basin, appears to be the mostimportant potential reserve of potassium chloridein China [4, 37].

3. Mining of Potash Salts

3.1. Shaft Mining

Mineral salts are very soluble, and therefore anyflow of water into the mine from the overlyingstrata, which are normallywater bearing, must beprevented. This causes difficulties when sinkinga shaft, and severe accidents have been caused byinfluxes of water. The freezing technique, espe-cially deep freezing, is comparatively safe and isused to sink most mineshafts. The shaft itself isprotected from water-bearing rocks by the use oftubbings, which are segments bolted together toform rings and are usually made from cast iron orsteel-reinforced concrete. The shafts generallyhave a diameter of 5–7 m, and the depth canexceed 1000 m.

3.2. Extraction, Conveying, andHaulage

In most potash mines, the salt is mined fromsubhorizontal deposits. Generally, rooms arecreated by removing the salt, and pillars are leftbetween these to prevent the cover rocks fromcollapsing. This enables an extraction rate of25 –60% to be achieved. For cost reasons themined-out rooms are not backfilled. In somemines, the total ore is extracted, which causessubstantial subsidence of the overlying strata(Alsace).

In steeply dipping deposits (e.g., in the saltdomes of northern Germany), roof mining wasoriginally carried out. This was later replaced byfloormining and then by funnelmining, which isnow being used increasingly in numerous var-iations [43]. Entry drifts are driven one abovethe other at intervals of 15–20 m, and the re-maining potash salt is mined by sublevel stop-ing. Material loosened by explosives falls viathe lowest funnel-shaped region into the mainhaulage level underneath. The mined-out room,100–250 m in height, is usually backfilled withsalt waste after mining. Funnel mining is muchsafer and cheaper than roof mining, becauseboth ore stoping and backfilling take place undergravity, and the mined room need not be reen-tered. During drilling of the blast holes, theminers are protected by horizontal pillarsbetween the sublevels.


Drilling and blasting operations are carriedout with the help of trackless vehicles. Largeholes with a diameter of ca. 40 cm and a length of7 m are drilled bymobile drilling equipment, andanother mobile drilling rig is used to drill blastholes around the larger hole in a predeterminedpattern. The explosive (generally ammoniumnitrate with addition of oil) is brought to theworkplace in tanks carried by diesel vehicles andis blown into the shot holes by compressed air.

In predominantly horizontal sylvinite depos-its, the most frequent method of extraction is bycutting with heavy machinery (180 t per unit) athigh cutting rates. These machines produce ma-terial suitable for transport by conveyor belt,enabling continuous extraction in 30–60-msections.

Extraction by borers with two to four rotors ismainly used to produce long pillars in a room andpillar system (Canada), the length of the cham-bers being as much as 1000 m. In longwallmining, however, which is usually carried outas a caving operation, two or three drum shearersare used (e.g., in Alsace and the CIS). Thesemachines enable daily outputs between 1500 and4000 t to be achieved.

The recovered material is transported bytrackless diesel or cable-fed electric front-endloaders with a capacity of 15 t. Transport alongunderground roads is increasingly by conveyorbelts but also by electric or diesel trainswith 30-t-capacity wagons or by dumper trucks with ahopper capacity of 40 t. For transportation byconveyor belt, the material must first be brokeninto suitably sized pieces.

The network of roadways for conveying,traveling, and ventilation usually extends >100km in large potash works. A radio communica-tion system is generally used.

Since the introduction of very heavy machin-ery and diesel-powered vehicles, extra attentionmust be given to ventilation. Powerful fans sup-ply fresh air at up to 30 000 m3/min.

In hoisting shafts, the skips have a capacity ofup to 25 t. These operate automatically andsupply large intermediate storage bins in thefilling station, so that the continuously operatingmanufacturing plants can be supplied with ore ata steady rate. Average daily throughput can beca. 30 000 t of ore.

Improvements in mining methods and theintroduction of new techniques have enabled the

output per worker underground to be increasedconsiderably. In Germany during 1965–1974 theoutput of potash ore increased by 20%, despite a>50% reduction in the number of employees.

3.3. Solution Mining

Solution mining is an alternative to conventionalmining for the extraction of potash ore. Theadvantages of this method are that the highexpense of sinking a shaft is eliminated andreserves can be exploited where conventionalmining is impossible (e.g., at great depth). Also,this method can be used where existing mineworkings are available but conventional miningmethods are no longer feasible, even thoughextensive reserves may still exist.

Since 1964, Kalium Chemicals in Saskatch-ewan has operated a plant in which brine ex-tracted from a potash deposit at a depth of ca.1500 m is used to produce very pure potassiumchloride. The process is based on a series ofpatents [44], but details have not yet been pub-lished. Water or an unsaturated solution of po-tassium chloride is passed through a system ofboreholes into the potash seam, which is 20 –25 m thick; potassium chloride and sodium chlo-ride are dissolved. The almost saturated solutionis pumped to the surface and fed to the productionplant. Rock salt above the potash seam is pro-tected from dissolution by an oil or air cushion(see Fig. 1). The brine produced passes throughmultiple-effect evaporators in which sodium

Figure 1. Solution mininga) Overlying rock; b) Oil cushion; c) Partly unsaturated saltsolution; d) Saturated salt solution; e) Cavity produced bydissolution; f) Potash deposit; g) Sodium chloride layer


chloride crystallizes. Potassium chloride is thencrystallized in a series of vacuum coolers [45].

In Utah (United States), where it was neces-sary to terminate a conventional mining opera-tion due to severe geological and technologicalproblems, operation was resumed in 1972 byusing solution mining. Shafts and undergroundcavities were flooded, providing a route for thebrine formed when water was fed in to dissolvethe salt. The brine was passed into surface pondswhere solar evaporation caused a mixture ofpotassium chloride and sodium chloride to crys-tallize, which was treated in a flotation plant toproduce 60% K2O potassium chloride [46].

In Canada a conventional mining operationwas also converted to the solution mining meth-od after penetration of water led to completeflooding of the mine and forced operations tocease. Water is passed via boreholes into theflooded mine and is converted into a concentrat-ed brine, which is withdrawn and cooled in apond during the very cold Canadian winter. Thepotassium chloride that crystallizes is recoveredand processed to give a salable product [47].

In the former German Democratic Republic,extensive research into the solution mining ofcarnallite or potassium chloride from carnalliticdeposits has been carried out. An experimentalplant with a KCl capacity of ca. 50 000 t/a wasoperated for a long period (see Section 5.2.3)[48].

4. Treatment of Potash Ores

4.1. Intergrowth and Degree ofLiberation [49, 50]

The salt minerals in potash ores are intergrown tovarying extents. Before the minerals can beseparated and the useful components recovered,the oremust be sufficiently reduced in size so thatindividual components are accessible to the pro-cessing method to be used. In the hot leachingprocess, sylvite is extracted, and therefore it mustfirst be liberated (i.e., it must not be occludedinside other minerals). To achieve this, it issufficient to break down the ore to a particle sizeof 4–5 mm or less.

For the mechanical treatment processes (i.e.,flotation, electrostatic treatment, and gravity sep-aration), liberation of the minerals must be com-

plete (i.e., individual grains must consist as muchas possible of pureminerals). The extent towhichthe minerals in the potash ore are intergrown canvary greatly from deposit to deposit (see Fig. 2),which means that the crude salt must be size-reduced to varying degrees before furtherprocessing.

The degree of intergrowth of individualminerals can be determined by examination ofa thin section. A photomicrograph shows thesizes and shapes of individual minerals in rela-tion to each other [51]. The disadvantage of thismethod is that a thin section gives only a two-dimensional view of a relatively small region ofthe salt mineral, and a very small sample of thesubstance is examined. The three-dimensionalarrangement of minerals present and their distri-bution are not observed. For these reasons, andalso because of the high cost of preparing thinsections, this method is now of only minor im-portance for the industrial processing of potashores.

Figure 2. Thin sections of sylvinite oresA) Coarsely intergrown (potash works in Lanigan, Canada);B) Finely intergrown (Kaliwerk Sigmundshall, Germany)


More usually, the degree of liberation in size-reduced samples is determined. The degree ofliberation of amineralmeans the percentage ratioof fully liberated mineral particles to the totalcontent of the mineral in the sample.

Degree of liberation:

%Free mineral

%Freeþ%Intergrown mineral�100

The degree of liberation L depends not only666on the grain size achieved by the grinding oper-ation but also on the type of grinding. It can begiven for particular ranges of grain size so thatthe way in which it depends on grain size can bedetermined, or it can be expressed as an averagefor the total sample (the integral degree of liber-ation �L). The arithmetic mean of the degree ofliberation of each range of grain sizes is obtainedfrom the following formula:

�L ¼Pni¼1

Lipiai

100a

where pi is the mass fraction of the ith size rangein percent; ai the percentage of useful mineral inthe ith size range; and a the percentage of usefulmineral in the total sample. Apart from thefraction of useful mineral a, which is determinedby chemical analysis, the liberated, noninter-grown fraction of useful mineral in each grainsize range must be determined. Two methods fordoing this are possible:

1. Visual estimation (by counting under themicroscope) of the proportion of intergrownparticles

2. Heavy-medium separation of the free or al-most free grains

The first method is easy to apply to saltminerals because the intergrowth effects arereadily recognized owing to the transparency ofthe grains. For coarse-grained materials such asthose usually found in ground products fromcoarsely intergrown potash ores, this processcannot be used.

The separation of free or nearly free mineralgrains from a size fraction is carried out by usingheavy liquids of appropriate density, such astetrabromoethane–toluene mixtures. A float –sink separation is carried out to determine the

fraction of free mineral grains. This methodcannot be used for salt particles with a grain size<0.5 mm because of the agglomeration of finegrains. In this case, the method of counting underthe microscope must be used, or the degrees ofliberation of the coarser size ranges must beextrapolated.

The results are expressed in liberation curves,which give the degree of liberation as a functionof grain size (Fig. 3).

4.2. Grinding [52, 53]

Potash salts are easily size-reduced. Therefore,fines may be formed, which can cause problemsin later stages of processing. Great care must betaken in selection of the equipment for variousstages of grinding.

The maximum grain size to which the potashore is ground depends on the processing methodused and the degree of intergrowth of the ore. Forthe hot leaching process, an upper grain size limitof 4–5 mm is adequate. For mechanical proces-sing, the ore must be ground to a degree ofliberation> 75%. For German sylvinite ores andhard salts, this is achieved by grinding to amaximum grain size of 0.8–1.0 mm. For themuch coarser sylvinite ores of New Mexico, amaximumgrain size limit of 2.4 mm is sufficient.The sylvinite ores of Saskatchewan are evenmore coarsely intergrown, so that size reductionto < 9 mm would give adequate liberation.However, such large crystals cannot be treated

Figure 3. Liberation curves of sylvinite oresa) Coarsely intergrown sylvinite ore (from New Brunswick,Canada); b) Finely intergrown sylvinite ore (from a northGerman salt deposit)


by conventional flotation, and the material istherefore normally ground to < 2.3 mm. Onelarge Canadian manufacturer produces a coarsecrystalline product by grinding the potash ore to< 9 mm, removing grains < 1.7 mm, and treat-ing this fine material by conventional flotation.The remaining fraction (1.7–9 mm) is treated ina heavy-medium separation plant (see Section5.5), giving a product with 60% K2O (95% KCl)and a size distribution of the granular commercialgrade.

The preliminary size reduction of potash ore iscarried out by underground mobile crushers,usually in the vicinity of a mining operation.When the ore is mined by continuous miningthis gives sufficient size reduction for it to godirectly to the haulage line. Further size reduc-tion to the grain size required for processing iscarried out in two stages on the surface. An initialsize reduction with impact or hammer mills (!Size Reduction) is always carried out to produce4–12-mm particles, depending on the raw mate-rial and processing method to be used. A coarsegrinding plant usually includes two grinding andscreening stages (see Fig. 4).

The final fine grinding stage is carried outeither by wet grinding in rod mills or by dry

grinding with rollers or impact crushers (Figs. 5and 6). Wet grinding with rod mills in a recircu-lating system with classification is standard formost flotation plants. Classification is by spiralclassifiers, vibrating screens, curved screens, orcyclones. Wet screening produces only a smallamount of fines and has the further advantage ofproviding a scrubbing effect that facilitates theremoval of clay from clay-bearing ores; a neces-sary step before the flotation process.

The production of a fine product by dry grind-ing is rarely used in flotation plants. It is indis-pensable as a preparation for electrostatic pro-cessing, which is not compatible either with thechanges to the mineral surfaces caused by aque-ous solutions or with excess moisture. The grind-ing operation must be carried out carefully togive a product low in fines. Roller mills andimpact mills can be used. Roller mills have thedisadvantage that throughputs are relatively low

Figure 4. Two-stage grinding systema) Conveyor belt; b) Grid; c) First screening stage; d) Firstgrinding stage; e) Second screening stage; f) Second grindingstage; g) Bucket elevator

Figure 5. Wet grinding and screening of coarsely inter-grown potash ore for flotation [5]Reprinted with permission of the Society for Mining, Metal-lurgy, and Exploration, Inc.


and maintenance costs are high. Although thiswas the preferred method in the early days ofelectrostatic processing owing to its gentle grind-ing action, it has now been largely replaced byimpact grinding, which is also used for sizereduction of the middle product from electro-static separation [54].

5. Potassium Chloride

Potassium chloride [7447-40-7], KCl, mineralname sylvite [14336-88-0], forms colorless non-hygroscopic crystals. It occurs in many salt de-posits (see Section 2.2) mixed with halite andother salt minerals. Natural sylvite is usuallyopalescent or milky white, as are crystals ob-tained from an aqueous solution. Sylvite is oftencolored red by hematite. With magnesium chlo-ride it forms the double salt carnallite [1318-27-0], KCl �MgCl2 � 6 H2O, which is also common-ly found in salt deposits. Potassium chloride isproduced in large quantities from mined potash

ores and from salt-containing surface waters.More than 90% of the potassium chloride pro-duced is used in single- or multi-nutrient fertili-zers, either directly or after conversion to potas-sium sulfate (! Fertilizers, 2. Types, Chap. 1.).The remainder has various industrial uses and isthe raw material for the manufacture of potassi-um and its compounds.

5.1. Properties

Potassium chloride crystallizes in the cubic sys-tem, usually as actual cubes. Some physicalproperties described are as follows:

Relative molecular mass 74.55

Melting point 771 �CCrystal system and type cubic 05hRefractive index n20D 1.4903

Density 1.987 g/cm3

Specific heat cp 693.7 J kg�1 K�1

Heat of fusion 337.7 kJ/kg

Enthalpy of formation DH 0 � 436.7 kJ/mol

Entropy S 0 82.55 J mol�1 K�1

Dielectric constant (at 106 Hz) 4.68

Thermal coefficient of expansion (15–25 �C) 33.7� 10�6 K�1

Solubilities in water at various temperaturesappear in Table 5, and the phase diagram of thesystem KCl–H2O is shown in Figure 7.

In the system KCl–H2O, the only solid phasesformed are KCl and ice. The cryohydric point(ice þ KCl) is � 10.7 �C (29.7 mol K2Cl2/1000 mol H2O). The boiling point of the saturat-ed solution at 1.013 bar is 108.6 �C (71.6 molK2Cl2/1000 mol H2O).

5.2. Production by Crystallizationfrom Solution

5.2.1. Phase Theory

The salt deposits were formed by the evaporationof seawater, which contains the principal ions

Table 5. Solubility of potassium chloride in water (g/100 g) [16]

Temperature, �C 0 10 20 30 40 50

Solubility 28.1 31.2 34.2 37.2 40.2 43.1

Temperature, �C 60 70 80 90 100

Solubility 45.9 48.6 51.3 53.8 56.2

Figure 6. Dry grinding and screening of coarsely inter-grown potash ore for flotation [5]Reprinted with permission of the Society for Mining, Metal-lurgy, and Exploration, Inc.


Naþ, Kþ, Mg2þ, Ca2þ, Cl�, and SO2�4 . With

water, these ions constitute a six-componentsystem. The concentration of Ca2þ in the mostinteresting region of the system is negligiblysmall, so that the system reduces to quinary. Itis nevertheless very complicated with, 23 differ-ent salts being formed between 0 and 100 �C,depending on the temperature and the ratio ofconcentrations of the components.

Understanding how the salt deposits wereformed and how they behave in dissolutionprocesses requires knowledge of the solutionequilibria.

The theoretical foundations for this were laidby VAN’T HOFF et al., who between 1896 and 1906investigated the formation of oceanic salt depos-its [55]. Many investigators have continued thiswork up to the present.

J. D’ANS critically evaluated all publisheddata up to 1933, expressinghis results in graphicalform [56]. In the same book, he described experi-mental methods for determining solution equilib-ria and gave recommendations for the graphicalrepresentation of experimental results [57].

Much of the equilibrium data published up to1967 are given in [58, 59].

In the years following World War II, in theKaliforschungs-Institut (Potash Research Insti-tute) of Hanover, AUTENRIETH carried out com-prehensive and detailed research into the stableand metastable equilibria of most relevance topotash production (especially from hard salt),giving the results in a form suitable for practicalapplication [60–66].

The intensive investigations carried out intothe quinary system make it the most thoroughlyinvestigated system with more than four compo-nents. However, only parts of the system that areofmost relevance to potash production have beenthoroughly investigated. An obstacle in the ap-plication of equilibrium data to practical pro-blems is that such a complex system is verydifficult to represent in a two-dimensional dia-gram. However, by fixing parameters, workingwith projections on a plane, and using diagramsshowing lines of constant parameters, even non-experts can work with them.

The best-known region is that in which thesolutions are saturated with sodium chloride.This is also the most important region for potas-sium chloride manufacture, both by the hotleaching process and by flotation, because inboth cases a solution saturated with sodiumchloride is used. The most important part of theso-called NaCl saturation space at 25 �C isshown in Figure 8 as a three-dimensional view.

Each point in the interior of this space corre-sponds to a solution saturated with NaCl,in which the concentrations of MgCl2, K2Cl2,and MgSO4 are given by the distance of thepoint from the axes. For practical reasons, the

Figure 7. Solubility curves for potassium chloride in waterE ¼ cryohydric point : Ice–potassium chloride solution

Figure 8. Three-dimensional view of the quinary system(saturated with NaCl) at 25 �C with 0–65 mol MgCl2/1000 mol H2O, showing stable and metastable regions


concentration figures are given in moles per1000 mol of H2O, and the concentrations of KCland NaCl are given in double moles. If one ormore of the salt concentrations increase, and if thesaturation concentration of another salt is ex-ceeded, this salt separates out. Its identity dependson the concentration ratio in the solution. Suchtwo-salt solutions (saturated with two salts) lie ontwo-salt surfaces, which form the boundaries ofthe saturation space. Of the twelve other saltswhose saturation spaces form the boundary of theNaCl saturation space, seven can be seen onthe figure because their two-salt surfaces lie inthe concentration range of Figure 8.

The KCl–NaCl two-salt surface on the frontside of Figure 8 is of special significance for thepotash industry, because it allows solutions to bemade up that are saturated with both salts at25 �C. Such solutions are important in the treat-ment of sylvinitic potash ores.

In many instances, a salt does not crystallizespontaneously when its concentration exceedsthat indicated on a two-salt surface. Instead,highly supersaturated solutions are formed,which can remain stable for hours or days, de-pending on temperature and composition. Thesesupersaturated solutions are termed metastable.They become stable saturated solutions by crys-tallization of a salt. In Figure 8, continuations ofthe stable schoenite–NaCl two-salt surface to theright and left into the unstable region are shownas broken lines. In potash manufacture, stablesolution equilibria are seldom attained. This isparticularly true for hard salt processing in whichNaCl-saturated solutions with high MgSO4 con-tent often lead to the undesired crystallization ofdouble sulfates such as schoenite, leonite, lang-beinite, and glaserite. Here, the rates of dissolu-tion, nucleation, and crystallization of these saltsas a function of temperature and composition ofthe solutions are especially important [64–66].

For practical application of equilibrium data,the boundary surface of the NaCl saturationspace is projected, for example, in the directionof the K2Cl2 axis in the MgSO4–MgCl2 plane. InFigure 9, the two-salt surface NaCl–KCl for thestable and metastable 25 �C isotherm is shown.The indicatedK2Cl2 andNa2Cl2 lines of constantconcentration enable the complete compositionto be read off for each of the solutions shownhere. By using this diagram, the dissolution andcrystallization processes possible in this part of

the quinary system can be described quantita-tively. To control the crystallization of potassiumchloride from 90 �C solutions in an industrialplant, for example, the most important boundarysurfaces of the 90 �C isotherm of the system(Fig. 10) are additionally required.

If any of the components of the quinary sys-tem are present in such small quantities that theyhave a negligible effect on the process, themathematical treatment can be simplified bydealing only with the remaining subsystem. Thefollowing subsystems are of importance:

Figure 9. 25 �C isotherms of the quinary system saturatedwith NaClStable (—–) andmetastable (–––) surfaces saturatedwithKCland NaCl. Concentrations of K2Cl2 (—–) and Na2Cl2 (–––)are indicated by lines of equal concentration

Figure 10. 90 �C isotherms of the quinary system saturatedwith NaClStable (—–) andmetastable (–––) surfaces saturatedwithKCland NaCl. Concentrations of K2Cl2 (—–) and Na2Cl2 (–––)are indicated by lines of equal concentration


1. Naþ, Kþ, Mg2þ, Cl�, and H2O with NaClsaturation (see Fig. 17) for the conversionof carnallite into potassium chloride andbischofite

2. Kþ, Mg2þ, Cl�, and H2O (see Fig. 16) for thedecomposition of carnallite

3. Naþ, Kþ, Cl�, and H2O [57] for the selectivedissolution of NaCl (e.g., from crystallineproduct obtained from hard salt in the hotleaching process)

4. Kþ, Mg2þ, Cl�, SO2�4 , and H2O [61] for the

production of potassium sulfate and potash–magnesia

5. Naþ, Mgþ, Cl�, SO2�4 , and H2O [57] for the

production of thenardite or Glauber’s salt

5.2.2. Hot Leaching Process

The hot leaching process is the oldest industrialprocess used to produce potassium chloride frompotash ore. It was first used in 1860 in Stassfurtand since then has been developed further inGermany, where it is still the dominant process.It is especially suitable for treating very finelyintergrown ores or ores that contain other saltminerals or insoluble minerals in addition to thesylvite and halite. It enables a high-purity productwith a uniform grain size to be produced. Inmanyplants, especially in Canada, where flotation isthe main production process, small hot leachingplants are also operated, in which the productfines (<0.2 mm) are recrystallized, or potassiumchloride is extracted and crystallized from theflotation residues or thickened clay slurries.These procedures give a considerable improve-ment in total yield and result in a very pure,completely water-soluble product.

Two different processes are used, dependingon the composition of the ore. In the sylvinite hotleaching process, the other salts present in addi-tion toKCl andNaCl play only aminor role in theprocess solutions. In hard salt leaching, processsolutions contain appreciable amounts of MgCl2and MgSO4. In the case of carnallite-containinghard salts, preliminary carnallite decompositionmust be carried out (see Section 5.2.3.) if theamount of carnallite present exceeds a criticallevel.

The different solubility properties of sodiumchloride and potassium chloride are shown inFigure 11. The solubility of potassium chloride islower in hard salt solutions than in sylvinite

solutions. The difference between the potassiumchloride contents of saturated solutions at lowand high temperatures is less for solutions of hardsalt than for solutions of sylvinite, so that theamount of potassium chloride that can be crys-tallized from a given amount of solution is smal-ler, which has a marked effect on the energyrequirement. Furthermore, an important differ-ence between the two solution types is that thesolubility of sodium chloride in sylvinite solu-tions decreases with increasing temperature,whereas it increases in hard salt solutions. Thisis apparent in Figure 11 B, which shows thebehavior of process solutions in a carnallitic hardsalt plant with a magnesium chloride content ofca. 240 g/L. This dependence of the solubility ofsodium chloride on temperature and magnesiumsalt content explains why the sodium chloridecontents of crystallized products differ, depend-ing on whether they came from the treatment ofsylvinite or hard salt.

The Process (Fig. 12). The potash ore,ground to a fineness of <4–5 mm, is stirred ina continuous dissolverwith leaching brine heatedto just below its boiling point. The leaching brineis themother liquor from the crystallization stageof a previous cycle of the process. The quantity ofleaching brine required is determined by theamount of potassium chloride in the ore. Thepotassium chloride should be extracted from theore as completely as possible, and the resultingproduct solution should be as nearly saturated as

Figure 11. Solubility curves for KCl and NaCl (schematic)A) Sylvinite leaching (with nonevaporative cooling); B) Hardsalt leaching


possible. The residue consists of two fractions ofdifferent particle size. The coarse fraction isremoved from the dissolver and debrined. Thefine fraction (fine residue or slime) is removedfrom the dissolver along with the crude solution,which is clarified with the aid of clarifyingagents. The slime that separates is filtered off,and the filtrate from the coarse and fine residues isrecycled to the recirculating brine. The residuesare washed with water or plant brines low in

potassium chloride to remove the adhering crudesolution, which has a high potassium chloridecontent. The residues are then disposed of bydumping (see Section 5.8).

The hot, clarified, crude solution is cooled byevaporation in vacuum equipment. Potassiumchloride and sodium chloride crystallize as thewater is removed. The sodium chloride contentof the crystals formed can be controlled bycomplete or partial replacement of the evaporat-ed water during crystallization. The crystalsformed are separated from the mother liquor andprocessed further. The mother liquor is heatedand recycled to the dissolver as leaching brine.

The leaching process is usually carried out intwo stages in a main dissolver and a secondarydissolver. The ground ore is first added to themain dissolver where it is mixed with the al-ready partially saturated solution from the sec-ondary dissolver. This causes the solution to bealmost completely saturated, and it is then re-moved from the leaching equipment. The partlyextracted ore is next fed to the secondary dis-solver where it comes in contact with freshleaching brine, and the potassium chloride thatwas not extracted in the main dissolver is takenup by the solution, which is then fed to the maindissolver. The leaching process in the maindissolver can be cocurrent or countercurrent(Fig. 13).

Figure 12. Overall schematic of a hot leaching process

Figure 13. Schematic arrangements of a hot leaching apparatusA) Cocurrent flow; B) Countercurrent flow


The residence time in the leaching equipmentis insufficient to give complete KCl–NaCl equi-librium. The crude solution always contains lesspotassium chloride and more sodium chloridethan corresponds to equilibrium. Another reasonfor the potassium chloride concentration to bemaintained below saturation is that completerecovery of potassium chloride from the ore ispossible only if the hot solution at the end of theleaching process still has a small capacity fordissolving potassium chloride. Also, the crudesolution cools by 1–2 �C as it travels from themain dissolver to the overflow from the followinghot clarification stage. If the solution leaving themain dissolver were completely saturated inpotassium chloride, the potassium chloridewould begin to crystallize at this point, leadingto losses in the slime.

For the crystallization process the potassiumchloride content should be maintained as near tosaturation as possible in the leaching equipment,so that the concentration difference between themother liquor and the hot crude solution is asgreat as possible. The greater this difference, thesmaller is the amount of solution used, and thelower is the energy consumption.

Also, a high concentration of potassium chlo-ride in the solution leads to a high potassiumchloride content in the crystalline product ob-tained on cooling. If the solution is unsaturatedwith respect to potassium chloride, a correspond-ing amount of sodium chloride in excess of itssaturation concentration is taken up by the solu-tion to compensate for the missing potassiumchloride. This means that, on cooling, a quantityof sodium chloride crystallizes, causing a de-crease in the potassium chloride content of theproduct even in sylvinitic solutions in which thesolubility of sodium chloride increases with de-creasing temperature.

Cooling the crude solution by evaporationincreases the concentration of salts present belowsaturation, and the saltswithwhich the solution issaturated crystallize. Unwanted crystallizationof sodium chloride can be prevented by addingwater to the vacuum cooling equipment, espe-cially in the case of solutions from sylvinite oreleaching. If the required potassium chloride con-tent in the crystalline product cannot be obtainedby addingwater to the solution,which is often thecase when hard salt is being processed, the sameresult can be obtained by treating the product

with cold water. A product containing 60% K2O(95%KCl) is usually required. By this cold-watertreatment technique, it is even possible to obtain aproduct with 62% K2O (98% KCl). The spentsolution is recycled to the process.

For KCl of analytical or pharmaceutical qual-ity, potassium chloride produced by the hotleaching process must be purified by single ormultiple recrystallization.

Processing of Hard Salt. Unlike the sylvi-nitic potash ores, whose principal constituentsare sodium and potassium chloride, hard saltscontain not only the alkali chlorides but also largeamounts of kieserite, usually with varyingamounts of carnallite, langbeinite, and anhydrite.Therefore, process brines produced by the leach-ing of hard salt are characterized by high contentsof magnesium chloride and magnesium sulfate,which make potassium chloride production moredifficult.

Magnesium sulfate comes mainly from kie-serite, which is very soluble (Fig. 14) but has aslow rate of dissolution that becomes even slowerif large amounts of dissolved MgCl2 are present.The amount of magnesium sulfate that dissolvesdepends on the grain size of the ore fed to thedissolving equipment, the kieserite content of theore, the time for which the crude salt–solutionmixture is stirred, the temperature, and the mag-nesium chloride content of the brine.

Magnesium chloride results from reaction ofdissolved magnesium sulfate with potassiumchloride to give sulfate-containing double salts,and also from any carnallite (KCl �MgCl2 � 6H2O) present in the hard salt. In general, the

Figure 14. Metastable solubility of kieserite in the quinarysystem saturated with KCl and NaCl between 75 and 110 �C,and solubilities of langbeinite and glaserite at 90 and 100 �C


MgCl2 and MgSO4 contents are kept constant inthe circulating brines, so that the rate at whichmagnesium is taken up from the ore is balancedby the rate at which it leaves the system in theresidues and products. If the ore contains largeamounts of carnallite, part of the circulatingbrine must be removed continuously from thesystem to prevent the MgCl2 level becoming toohigh.

Because of the large amounts of magnesiumsalts in process brines, the hard salt leachingprocess has three important disadvantages com-pared with the sylvinite process:

1. In solutions of hard salt, the range of solubi-lities is considerably less than in solutions ofsylvinite (Fig. 11). To extract a given amountof potassium chloride, much more liquor, andhence more energy, is required comparedwith sylvinite.

2. With solutions of hard salt, crystallization ofdouble sulfates such as schoenite, leonite, andlangbeinite often occurs. These salts can ap-pear in the residue or the product, which leadseither to potassium losses in the residue or to alower K2O content of the product. Also, dou-ble salts can crystallize in pipelines, vessels,and pumps, interferingwith the process and inextreme cases bringing it to a standstill. Theparticular double salt that crystallizes dependson the MgCl2 content of the solutions, thetemperature, and other operating conditions(Figs. 9 and 10). The extent of supersaturationwith double sulfates must therefore be con-trolled to prevent uncontrolled crystallizationof double salts and consequent introduction ofimpurities into the product or disturbance ofthe process. To achieve this, the circulatingbrine, or part of it, is fed to a reactor inwhich itis agitated intensively in the presence of nu-clei (20–40 wt%) of the double salt to beremoved until the brine is no longer supersat-urated with respect to it [67].

3. Another disadvantage of the hard salt leach-ing process is that, with higher magnesiumsalt contents, the temperature dependence ofthe solubility of NaCl becomes unfavorable(Fig. 11), that is, when the potassium chlorideis crystallized by cooling, considerableamounts of sodium chloride can also crystal-lize (Fig. 15). In practice, equilibrium is notcompletely reached when the KCl–NaCl is

dissolved, and cooling of the solution oftenoccurs by water removal, so that the K2Ocontent of the crystals formed is usually onlya little more than 40%. Since the usual potas-sium chloride for fertilizers (excluding spe-cial products) has a minimum potash contentof 60% K2O, this primary product was for-merly crystallized in a second leaching plant.Alternatively, a product can be made with60% K2O directly, if the crystallization ofsodium chloride during vacuum cooling isprevented by addition of sufficient water be-fore or during crystallization of the solution toensure that only potassium chloride crystal-lizes [67]. The K2O content in the product of aprimary crystallization can be increased to therequired level by treating the product withwater or a plant solution unsaturated withrespect to sodium chloride. Both the excesswater added during vacuum cooling and thewater used for treating the product must beremoved from the recirculating system of theplant. Both cause a loss of yield whose extentdepends on operating conditions. To avoidthis, excess water must be evaporated fromprocess liquors.

5.2.3. Processing of Carnallite

Carnallite, KCl �MgCl2 � 6 H2O, is the mostabundant potassium mineral in salt deposits andoccurs widely in mixtures with halite or withhalite and kieserite in the form of carnallitite ore

Figure 15. K2O content of the crystalline product obtainedby cooling an equilibrium solution saturated with NaCl andKCl from 90 to 25 �Cwithout evaporation while maintainingsaturation


(see Section 2.2). In the early days of the potashindustry in Germany, it was the preferred startingmaterial for the production of potassium chlo-ride. Today, sylvinitic ores and hard salts areused almost exclusively, because the extractionand processing of carnallite ore are considerablymore difficult and expensive for the followingreasons:

1. Carnallitite ore has unfavorable mechanicalproperties that make mining more difficult.

2. The K2O content of pure carnallite is ca. 17%,compared with ca. 63% for pure sylvite.

3. Whereas the separation and purification ofsylvite from sylvinite ore can usually be car-ried out by flotation, which does not involve aphase change, the extraction of potassiumchloride from carnallitite ore necessitatesdissolution or decomposition of the carnallite,and a high energy consumption for decompo-sition or purification of the decompositionproduct, depending on the process.

4. The treatment of carnallite generates largequantities of concentrated magnesium chlo-ride solution, which must be disposed of.

For these reasons, mined carnallitite ore istoday seldom used as a raw material. However,carnallite is often a major component in mixedsalts of the hard salt type, and hence influencesthe choice of processing method.

Large quantities of a carnallite–halite mixtureobtained by solar evaporation of water from theDead Sea are used for the production of potassi-um chloride in Israel and Jordan.

Theoretical Basis. The theory of the pro-duction of potassium chloride from carnallite andcarnallite-containing mixed salts is based on theK2Cl2–MgCl2–H2O system shown in Figure 16,which is valid between� 3 and 117 �C. In VAN’T

HOFF coordinates (moles of salt per 1000 molH2O) the points representing the composition ofwater ðPH2OÞ, bischofite (Pbischofite), and carnall-ite (Pcarnallite) are indicated. All possiblemixturesof water and carnallite are shown on the straightline between PH2O and Pcarnallite, the molar ratioK2Cl2 : MgCl2 here being always 1 : 2, as incarnallite. The curve fromL4 via R and E to pointL0 represents an arbitrary isotherm, and indicatescompositions in which solutions are in equilibri-um with the corresponding solid phase. Point L0

gives the solubility of KCl in water, and thesolutions L1 are in equilibrium with KCl asthe solid phase. Solution E is in equilibrium withthe solid phases KCl and carnallite, and the solu-tions L2 are in equilibrium with carnallite only.Solution R is in equilibrium with bischofite andcarnallite. The solutions L3 are in equilibriumwith bischofite, and point L4 indicates the solu-bility of bischofite in water.

If carnallite is dissolved in water, the compo-sition of the solution follows the straight linePH2O�Pcarnallite;, which intersects curve L1 atpoint D. Here, the solution is saturated with KCl,and further addition of carnallite results in disso-lution of MgCl2 and crystallization of KCl untilpoint E is reached. This incongruent solubility isthe basis for the simple method of processingcarnallite (i.e., cold decomposition by motherliquor).

Carnallitic potash ores or crystallized pro-ducts from solar evaporation always contain somuch halite that solutions produced during pro-cessing are saturated with sodium chloride. Fig-ure 17 is a section from the quaternary systemK2Cl2–NaCl–MgCl2–H2O saturated with NaCl,showing the 25 �C and 105 �C isotherms. Thekieserite content in the ore in the region of thesolutions E has in practice a negligible effect onthe composition of the solutions. Some analysesof equilibrium solutions of the quaternary systemare given in Table 6.

Figure 16. The system K2Cl2–MgCl2–H2O (not to scale)represented by using van’t Hoff coordinates


Processing Methods. Many processes forthe treatment of carnallite have been describedin the literature and used [10], but only a few areimportant today. By far the most important iscold decomposition by mother liquor. The com-plete dissolution process is still used occ-asionally.

Cold decomposition is carried out at ambienttemperature (e.g., 25 �C). Carnallitite ore ismixed and agitated with water or with a solutionof low MgCl2 content such that the compositionof the mixture corresponds to point B in Fig-ure 17. This causes the crystallization of anamount of potassium chloride corresponding tothe line B–E25, with formation of a solution of

composition E25. The very fine potassium chlo-ride produced still contains fine, salted-out sodi-um chloride, undissolved halite, and sometimeskieserite and clay, depending on the compositionof the carnallitite ore used. Approximately 85%of the potassium chloride contained in carnallitecan be obtained as a crystalline product by thisprocess. The yield can be increased by evapora-tive concentration of the decomposition motherliquor E25 or Q25 (analyses are given in Table 7).During evaporation, sodium chloride and some-times magnesium sulfate crystallize and must beremoved. The synthetic carnallite that crystal-lizes when the concentrated solution is cooled isfed to the cold decomposition process.

Potassium chloride from the decomposition,which consists of very finely divided particlesand is rather impure, must be purified. Purifica-tion by flotation is difficult owing to the finenessof the decomposition product (see, however,Section 5.3.6). For this reason, purification isnearly always carried out by the hot leachingprocess, which yields a very pure, completelywater-soluble, coarse-grained product.

The mother liquor E25 or Q25 from the de-composition process can take up a certainamount of sodium chloride (see Table 7). If thecarnallitite used in the cold decomposition con-tains only small amounts of halite and if water isused for the decomposition, potassium chloridecan be produced that requires no further purifi-cation, because sodium chloride has dissolved inthe decomposition liquor. In Israel, a large pro-portion of the carnallite–halite mixture recov-ered from theDead Sea by solar evaporation is so

Figure 17. Quaternary system (K2Cl2–MgCl2–Na2Cl2–H2O) saturated with NaClSee text for explanation

Table 6. Composition of stable saturated solutions in Figure 17*

Point Temper-

ature, �CDensity,

g/cm3

Concentration,

mol/1000 mol H2O

K2Cl2 MgCl2 MgSO4 Na2Cl2

E25 25 1.275 5.8 70.8 4.4

Q25 25 1.291 5.8 68.0 5.2 4.2

E105 105 1.328 13.2 93.7 4.4

Q105 105 1.325 13.7 92.9 1.0 4.3

*Solutions Q52 and Q0

15 are saturated with MgSO4 and NaCl, and

correspond to solutions E52 and E0

15, which are saturated with

NaCl.

Table 7. Typical grain-size distribution in potash fertilizers (cumula-

tive percentage retained)

Mesh Tyler Cumulative subsieve fraction

width, mm mesh Granular Coarse Standard

þ 3.36 þ 6 2–12

þ 2.38 þ 8 30–45 2–20

þ 1.68 þ 10 75–90 25–50

þ 1.19 þ 14 95–98 70–90 5–15

þ 0.84 þ 20 99 90–98 20–45

þ 0.60 þ 28 99 50–75

þ 0.42 þ 35 70–90

þ 0.30 þ 48 85–95

þ 0.21 þ 65 96–98

þ 0.15 þ100 97–99


completely freed from halite by fractional crys-tallization and hydraulic classification that agrade of potassium chloride with > 60% K2Ocan be produced directly by cold decomposition.This so-called cold crystallization process iscarried out so that the crystals produced have

the same grain size as those from the crystalliza-tion plant of the hot leaching process and nofurther increase in grain size is required. Aschematic diagram showing various processesused in the treatment of carnallite from the DeadSea is given in Figure 18.

Figure 18. Flow diagram for potassium chloride production by the Dead Sea Works Ltd. (Israel) [4]Reproduced from [4] with permission


In the electrostatic treatment of carnallitichard salts (see Section 5.4.2), concentrates canbe produced that contain various amounts ofhalite together with sylvite and carnallite. Theseconcentrates are reacted with suitable processbrines or with water to bring about cold decom-position of the carnallite. The K2O content of theproduct, and hence its potential use, depend onthe halite content of the concentrate.

The complete dissolution process is used in anexperimental plant in central Germany in whichcarnallite is extracted from a salt deposit bysolution mining (see Section 3.3) [48].

The crude salt is dissolved by a solution inwhich the magnesium chloride content is ad-justed so that at the chosen temperature of ca.80 �C, no decomposition of carnallite takesplace, and hence no crystallization of potassi-um chloride, and the carnallite is completelydissolved. The solution, which is almost satu-rated with carnallite, is pumped out, evaporat-ed, and cooled, causing crystallization of thecarnallite, which is then treated by the colddecomposition process to produce potassiumchloride.

For mixed salts containing less than ca. 15%carnallite, carnallite decomposition is generallydispersed with, and the ore is treated directly bythe hot leaching process. The magnesium chlo-ride that enters the process in the carnallitemust be removed from the brine circuit inappropriate process brines. If the carnallitecontent of the ore is high, a carnallite decom-position stage is carried out before the hotleaching process.

5.2.4. Equipment

Leaching. The choice of leaching equip-ment depends on the properties of the materialto be leached and the throughput required. Thetechniques used in plants in which the potash iscompletely leached (e.g., in many German andFrench plants) differ fundamentally from those inplants (e.g., in Canada) where only the finematerials from a screening operation, cyclonefines, and slimes are recrystallized or extracted.

Potash ores are leached at rates of up to1000 t/h salt, producing up to 2000 m3/h solu-tion. Screw dissolvers, up to 14 m long and 3 min diameter, are widely used (Fig. 19). Highmaterial transfer is achieved by fitting partitionsin the upper part of the dissolver at intervals of 2–3 m. These are immersed in the suspension ofcrystals and force the solution to flow perpendic-ular to themain flowdirection. At the outlet of thedissolver, the solids are scooped from the solid–liquid mixture by an elevator system with perfo-rated buckets from which liquid drains duringconveying; the residual water content is ca. 15–20%. Alternatively, bucket wheels can be used.

Elevators or bucket wheels remove only thecoarse residue (ca. 75% of the total residue) fromthe screw dissolver. The amount of potassiumchloride-containing solution adhering to thecoarse residue is usually lowered to 2–4% bycentrifugation.

The fine residue (ca. 25% of the total residue),which consists of very fine salt and insolublecomponents of the ore (mostly anhydrite andclay), flows with the raw solution from the screw

Figure 19. Leaching system with residue debrininga), b) Screw leachers with draining elevators; c) Vibratory screen centrifuge; d) Hot clarifier; e) Rotary filter


dissolver into the hot clarifier, where flocculationagents are added and sedimentation occurs.Sealed, insulated circular clarifiers are used. Thefine residue is removed from the hot clarifier as asuspension with a solids content of ca. 50% anddebrined on rotary filters to a residual watercontent of 10–16%. To avoid losses of potassiumchloride, the solution adhering to the residuemust be removed. For small quantities of residue,this is done by washing on the filter with water ora process brine having a low potassium chloridecontent. For large amounts of residue, two- orthree-stage water washing is used; the filter cakeis slurried with the water, and the resultingsuspension is filtered. This treatment is repeatedonce or twice.

In North America, where the flotation processpredominates and only small quantities of ore aretreated by the hot leaching process, a series ofagitated vessels is used. The leaching processtakes place in either cocurrent or countercurrentflow. Between the stages of countercurrent leach-ing, solid–liquid separation is carried out withhydrocyclones or hydroclassifiers [68]. The lay-out of a leaching plant including crystallization isshown in Figure 20.

Crystallization [69, 70]. The hot solutionfrom the clarifier of the leaching plant is almostsaturated with potassium chloride and is cooledby expansion evaporation in vacuum equipmentto cause crystallization. The vapors are con-densed in surface or barometric condensers, withprocess brines used for higher temperatures andcooling water for lower temperatures. Theamount of heat that can be recovered dependson the number of stages in the cooling system. Inthe past, vacuum cooling plants were constructedwith up to 24 stages, to give maximum possibleheat recovery. However, the number of stages ina modern plant is usually between four and eight.

The high cooling rate of a vacuum plantresults in a high degree of supersaturation thatproduces very fine crystals, unless the design ofthe plant and the crystallizing conditions areoptimized to give a coarse product. Several suit-able crystallizers are now available (! Crystal-lization and Precipitation, Chap. 5.).

If it is not important to have coarse crystals(e.g., if they are to be reprocessed), stirred crys-tallizers or conical-based evaporators are used.The former consist of horizontal cylindricalvessels divided into chambers by vertical walls.

Figure 20. Leaching–crystallization process for production of potassium chloride from potash ore [8]Reproduced from [8] with permissiona) Crusher; b) Screen; c) Leach tanks; d) Thickener; e) Centrifuge; f) Heater; g) Steam ejector; h) Barometric condenser;i) Vacuum cooler–crystallizer; j) Dryer


The solution or suspension flows from chamberto chamber through openings in the walls as it iscooled. These crystallizers are provided withmechanical stirrers or air agitation (Fig. 21).

The conical-based evaporator consists of avertical cylinder with a lower conical section.The solution is sprayed into the top, and much ofthe water evaporates. A certain liquid level ismaintained in the evaporator, into which liquidfrom the nozzle falls as a supersaturated solution.

Several evaporators are arranged in series oneabove the other so that the suspension flowsthrough each stage in the direction of decreasingpressurewithout the need for intermediate pumps(Fig. 22). The advantage of this arrangement liesin its very simple and economical construction,with no moving parts. However, the product isvery finely divided.

The aim of the crystallization process is usu-ally a dust-free product with optimum grain size.

Figure 21. Eight-stage horizontal crystallizera) Surface condensers; b) Barometric condenser; c) Hot well; d) Cyclone; e) Centrifuge; f) Filtrate tank; g) Slurry holding tank.Reprinted with permission fromR. M. McKercher et al., Proc. 1st Int. Potash Technol. Conf.,Oct. 3–5, 1983, Pergamon Press

Figure 22. Conical-based evaporator–crystallizer plant with eight evaporation stagesa) Surface condensers; b) Barometric condensers; c) Barometric collection vessel for condensate; d) Barometric collectionvessel for salts and brines; e) Barometric collection vessel for cooling water; f) Pumps; g1–g8) Evaporators


Two main requirements must be met to achievethis: the solution must have only a slight degreeof supersaturation, and this should be removed bycrystallization on seed crystals that are alreadypresent. The slight degree of supersaturation isachieved by providing internal circulation withineach crystallization stage, in which the cooledsolution is mixed with the added hot solution inan exactly controlled ratio. The continuous pres-ence of seed crystals is ensured by keeping thesolids content in each crystallizer at ca. 30 wt%.

These conditions are fulfilled by two types ofcrystallizer: the fluidized-bed (Oslo) type and thedraft-tube agitated type.

In the Oslo crystallizer (Fig. 23), the solutionis supersaturated in a separate evaporator andflows into a crystal slurry vessel below, where thesupersaturation is removed by crystallization onKCl crystals already present. To reduce the de-gree of supersaturation in the evaporator, a largeamount of clarified solution is pumped from thecrystal slurry vessel into the evaporator, togetherwith the incoming hot feed solution. Clarifiedsolution for the next stage is removed from thevessel at another takeoff point. A crystal suspen-sion is also removed from the vessel at a ratecorresponding to the rate of crystallization, and isfed either to the next stage as seed crystals or tothe debrining stage. Crystallizers of this type areconstructed by companies such as Lurgi, Swen-son, and Struther-Wells.

In draft-tube agitated crystallizers, a state ofsupersaturation is created and removed in thesame crystallizer. This type is manufactured bySwenson (DTB¼draft tube baffle), Standard-Messo, and Kali und Salz. In the Swenson andStandard-Messo (! Crystallization and Precip-itation, Section 5.1.) crystallizers, the entire sus-pension is brought into continuous contact withthe evaporator surface by a propeller stirrer thatprovides internal circulation; immediately aftersupersaturation is achieved, a large quantity ofsalt is made available to remove the state ofsupersaturation. The coarser crystals collect pref-erentially in the lower part of the crystallizer,where they are removed as a suspension and fedto the next stage or to the debrining stage. The hotsolution is passed through a clarifying zone and isfed to the next stage.

All the processes described so far are operatedcocurrently (i.e., the crystallizing salt is trans-portedwith the cooling solution fromone stage tothe next). In contrast, the Kali und Salz process(Fig. 24) operates countercurrently [71]. Thesalt produced in each crystallizer is passedthrough a classifier at the bottom of the crystal-lizer as soon as a predetermined minimum crys-tal size has been reached; then, unlike the othersystems, it is passed to the next hottest stage.The classifier is fitted with equipment that en-ables the upward flow rate, and hence crystalconcentration and product crystal size, to becontrolled independently of the throughput rateof the solution. Mixing of the cooler solutionwith the hotter solution in a crystallizer is pre-vented by transferring the salt with the help ofthe solution from the stage into which it ispumped.

5.3. Flotation

Since the early 1900s, ores of many differentkinds have been processed by froth flotation. Ageneral description of this process is given in! Flotation.

Investigations into the flotation of potash oresbegan in the early 1930s in the United States andthe Soviet Union [72, 73]. The first full-scaleplant began production in 1935 inCarlsbad (NewMexico, United States) [74].

The first reports of investigations into theflotation of potassium salts in Germany appearedFigure 23. Oslo crystallizer


in 1939 [75]; the first full-scale plant beganoperating in 1953 [76, 77].

At present, most of the worldwide productionof potash fertilizers is by flotation. This is themost widely used process in Canada, the UnitedStates, the CIS, and France. Small hot leachingplants are often attached to the flotation plantsfor treating slimes and intermediate products.The flotation process is not widely used inGermany, where a combination of electrostatictreatment with the hot leaching processpredominates.

5.3.1. Potash Ores Suitable for Flotation

Various types of crude potash salts can be treatedby flotation:

Sylvinite Ores are mixtures of sylvite(KCl) and halite (NaCl) in varying ratios. Theyrepresent the majority of potash ores treated.The sylvinites of Canada, the United States, andthe CIS also contain up to 8% clay components[78].

Hard Salts contain kieserite (MgSO4 �H2O), as well as sylvite and halite, and some-times also anhydrite (CaSO4).

Mixed Salts consist of amixture of sylviniteore or hard salt with carnallite (KCl �MgCl2� 6 H2O).

Polymineral Salts contain not only sylvite,halite, and kieserite, but also langbeinite (K2SO4

2 MgSO4), kainite (4 KCl � 4 MgSO4 �11 H2O), polyhalite (K2SO4 �MgSO4 � 2 CaSO4

2 H2O), and clay.

5.3.2. Carrier Solutions

In the flotation of water-soluble salts the carrierliquids are salt solutions that are saturated withthe components of the raw material.

Thus, sylvinite ore flotation is carried out in aKCl–NaCl solution. For the flotation of hard salt,the brine also contains various amounts of mag-nesium sulfate and magnesium chloride.

5.3.3. Flotation Agents

The collectors are the true flotation agents, whichselectively coat the surface of the component tobe floated. For the flotation of sylvite, straight-chain primary aliphatic amines are used in theform of their hydrochlorides or acetates [79].

Figure 24. Kali und Salz countercurrent crystallization processa) Crystallizers; b) Bells; c) Draft tubes; d) Liquor overflow; e) Stirrers; f) Condenser; g) Thickener; h) Centrifuge; i) Dryer;j) Combustion chamber


Mixtures of amines of various chain lengths,which largely eliminate the effects of pulp tem-perature variation, are extremely useful. A typi-cal flotation amine has, for example, the follow-ing composition: 5%C14 – NH3Cl, 30%C16 –NH3Cl, 65% C18–NH3Cl, iodine number: 4.

Foamers contribute to the dispersion oflong-chain amines and to the stabilization andhomogeneous distribution of amine micelles[80]. The following substances are preferred:aliphatic alcohols with chain lengths >C4, ter-pene alcohols, alkylpolyglycol ethers, and meth-yl isobutyl carbinol, which is used mainly inCanada and the United States.

Extenders for sylvite flotation are nonpolarmaterials, especially oils of various types. Theyare probably incorporated in the micelles andincrease their hydrophobic properties. Extendersare especially effective in the flotation of coarseparticles [81].

Clay Depressants are used in salt flotation toblock clay components, which would otherwisebind large amounts of flotation agent. Clay con-tents of 1.5–2% can be controlled by treatmentwith these depressants. If larger amounts arepresent, additional steps must be taken (clayflotation or classification). Clay depressants in-clude guar and starch products, carboxymethylcellulose, and polyacrylamide [78].

5.3.4. Theory

The combinations of reagents used for sylviteflotation have been found by empiricalinvestigation.

Theoretical studies of the separation of potas-sium and sodium chloride by flotation have led tovarious interpretations.

The theories developed up to 1961 are thor-oughly discussed in [82], which also describesexperimental results that have contributed great-ly to the understanding of salt flotation.

The most important theories are reviewedbelow:

According to the exchange theory [83–85],hydrophobic properties can be imparted to amineral if the polar group of a collector can beincorporated into the crystal lattice in place of an

ion. In the flotation of sylvinite ore, for example,exchange between the Kþ ion (radius: 0.135 nm)and the NHþ

3 group of the amine (radius:0.143 nm) is assumed to occur, whereas the Naþ

ion (radius: 0.095 nm) is too small (Fig. 25).This does not account for the fact that kieserite(MgSO4 �H2O), for example, is readily floated byamines in water although the Mg2þ ion has aradius of only 0.065 nm.

The structure theory [86] postulates that theinteranion distance in the crystal lattice of the saltmatches the intermolecular distance in the latticeof the amine hydrochloride within � 20%. Insylvite, this condition for oriented growth of thecollector on the crystal is fulfilled (Fig. 26),whereas the interionic distance of 0.398 nm inthe halite lattice is too small. However, langbein-ite has good flotation properties, although thesesteric requirements are not met [86].

Figure 25. Incorporation of a polar group of the collectorinto the crystal lattice [83]Reprinted with permission of the Society for Mining, Metal-lurgy, and Exploration, Inc.

Figure 26. Lattice analogy between the amine and the salt[86] (distances in nm)


According to the hydration theory [87], flota-tion of a mineral is impossible if it has a positiveheat of solution, whereas salts with negativeheats of solution can be flotated. If heat isevolved when a salt dissolves, more energy isproduced by hydration of the ions than is requiredfor breakdown of the lattice. This suggests thatsuch salts are strongly bonded to water mole-cules, preventing adhesion of the collector. Acomment on this theory is given in [88].

Another hydration theory is based on newinformation on the hydration of ions in dilutesolutions [89]. This differentiates between posi-tive local hydration (in which the ion reduces themobility of the neighboringwater molecules) andnegative local hydration (water molecules closeto the ions are more mobile than in pure water).

According to [90], in the case of positive localhydration (e.g., with Naþ ions), the Cl� ions areshielded by water molecules, so that the bondbetween a cationic amine and sodium chloridecan only be weak. In the case of negative localhydration (e.g., with Kþ ions), formation of abond with the collector is unhindered. Applica-tion of these theories, developed for dilute elec-trolyte solutions, to the hydration state of saltsurfaces is difficult, but should be qualitativelyvalid [90].

The G zone theory (G ¼ Grenzfl€achen ¼ in-terfaces) [91] applies both to the flotation ofsylvinite with amines and to the flotation ofkieserite with an anionic surfactant (the sodiumsalt of a highly sulfated fatty acid). In both cases,adsorption of the collector proceeds by the samemechanism.

Amines and the anionic surfactant, in a satu-rated solution of salts that can be floated withthese reagents, are present either as micelles or intrue solution. In the usual carrier brines, floccu-lation (amines) or formation of very small dro-plets occurs (anionic surfactant).

Based on experimental results, zones are as-sumed to be formed on the surfaces of crystalssuspended in a mixed-salt solution (carrier liq-uid) in which ions from the lattice are present inthe form of a saturated solution.

When collectors are introduced into the sus-pension, they migrate to the zones where theirsolubility is greatest due to their thermodynamictendency to dissolve (i.e., to the boundary zonesof those particles that they cause to float). Thebonding of the collectors in these zones is re-

inforced if their solubility in the surroundingcarrier brine is low. The success of flotationdepends on the strength of this bond (Fig. 27).

Froth cannot be produced in the usual carrierliquids either by collectors or by foaming agents.The collision of an air bubble with a crystalcoated with the collector material causes thelatter to spread over the surface of the bubble,stabilizing it and causing the crystal to float.

Large-scale operations of sylvite and kieseriteflotation has thus been placed on a good unifiedtheoretical foundation.

5.3.5. Flotation Equipment

The flotation equipment used in the potash in-dustry resembles that used for flotation of otherores (! Flotation, Section 5.3.).

Mechanical flotation cells operate with a ro-tor–stator system that causes both thorough mix-ing of the pulp and thorough distribution of theair, which can be drawn in by suction or injectedfrom a compressed air network.

In recent decades, high-capacity flotation hasbeen introduced almost everywhere. In this pro-cess, several stirrersoperate inasingle trough[92].

Agitair high-capacity flotation equipment isused in Germany, and the preconcentrate is puri-fied in individual cells of the Denver type. Wem-co high-capacity machines are used by theFrench potash industry, and Mechanobr high-capacity cells by the former Eastern-bloccountries.

Figure 27. Flotation of potassium chloride with octylaminein saturated solutions of KCl and KCl þ NaCl (Feed materi-al: 20% KCl, 80% NaCl; 0.1–0.315 mm)a) Carrier liquid saturated in KCl and NaCl (low solubility ofoctylamine); b) Carrier liquid saturated only in KCl (highsolubility of octylamine)


Modified Mechanobr M 7 fluidized-bed units(Fig. 28), manufactured by Mescerjakov, arestandard equipment in the CIS [93, 94]. Typicalfeatures include the presence of a grid and circu-lation of the suspension via a box on the frontwall of the cell, connected to the agitator by acirculation pipe.

Pneumatic flotation equipment operates with-out stirrers. In the free-jet flotation processdeveloped inGermany [95], the conditioned pulpis pumped through an aerator fitted with a porousgas distributor or annular nozzle, either outside orinside the separation vessel, and flows as a free jetinto the separation vessel. Air bubbles with theadhering minerals rise to the top and form a froththat flows over the edge of the vessel into achannel (Fig. 29). The pneumatic flotation pro-cess has been tested successfully for a specialcase of sylvite flotation. It is used industrially forthe recovery of kieserite from leaching residues.

In the CIS, a pneumatic flotation cell has beendeveloped for the treatment of coarse sylviniteores with a maximum grain size of 3 –5 mm [96,97] (Fig. 30). The conditioned pulp is fed fromabove onto a layer of froth immediately above theperforated-tube aeration system. The hydropho-bic sylvine crystals are retained in the froth, sothat mechanical damage of the bubble–crystalcombination is kept to a minimum. The flotationrate is very high.

5.3.6. Processes

Sylvinite Ore Flotation. Pure sylvinite ofGerman origin can be floated without serious

problems. The components of the ore are rela-tively strongly intergrown [50] and must bebroken down by grinding until each grain con-sists as far as possible of only one component.

The ore is first ground to a particle size of< 4 mm, screened to remove fines, and thenslurried with carrier brine and ground to a grainsize of< 0.8–1.0 mm in a circulating system thatincludes rod mills, spiral classifiers, or wetscreens (Fig. 31).

The solids content of the flotation pulp isadjusted to 30–40 wt%, and it is then mixed withflotation agents (ca. 40 g of oil, 20 g of foamingagent, and 40–80 g of collector per tonne ofcrude salt). As much of the desired product aspossible is then extracted by the three-stage high-capacity rougher flotation. The rougher tails areclassified in hydrocyclones, debrined (the coarsefraction in centrifuges, and the fine fraction inrotary filters), and dumped. Since the filtrate stillcontains finely divided salt, it is clarified in athickener and recycled for slurrying the ore.

The material floated in the rougher flotation isthen concentrated in a cleaner flotation process,usually in three stages. This does not produce amarketable product (with > 60% K2O). Theconcentrate is therefore separated from the brineand washed with water in salt washingequipment.

Figure 28. Mechanical flotation cell with fluidized beda) Flotation cell; b) Stirrer; c) Stator; d) Tube of the stirrersystem; e) Grid; f) Circulation box; g) Circulation pipe;h) Shaft

Figure 29. Principle of pneumatic free-jet flotationReproduced from [95] with permission


This washing process dissolves more halitethan sylvite, producing a potash salt that meetsquality requirements. The washing liquor is re-moved by centrifuges, and the product is driedand conveyed to silos.

The cleaner flotation process also producessome intermediate-quality product with a fairlyhighK2O content, whose coarse particles containmost of the residual intergrownmaterial from theore. They are therefore recovered by hydrocy-clones and fed to the screening–grinding stagefor size reduction. The energy consumption forthe entire process is ca. 10 kW � h/t ore.

The treatment of sylvinite ores from Canada,the United States, and the CIS is more difficultowing to the clay content of up to 8%. The claycomponents disintegrate in the brine, formingvery fine slimes that absorb large quantities offlotation agent. The flotation process is thereforesometimes preceded by a desliming stage(Fig. 32). The clay is first detached from thesurface of the salt by vigorous agitation in scrub-bers, so that it can be floated in a separateoperation. The clay is usually removed by mul-tistage classification [98]. The fine material isthen thickened and washed with water to reducelosses of K2O. Up to 1.5–2% of residual clay inthe crude salt can be handled by adding claydepressants (see Section 5.3.3).

An advantage of clay-bearing sylvinite oresis that they are usually sufficiently liberated atgrain sizes of < 3–5 mm. Sometimes, coarsecrystals are floated separately from fine crystals

Figure 30. Pneumatic coarse-grain flotation cell (Gogorch-improect type) [96]a) Feeding device; b) Spray nozzle; c) Aeration system(perforated rubber tubes); d) Device for removing rejectmaterial

Figure 31. Flow diagram of the flotation of potash oresa) Fine screen; b)Wet grinding; c) Classification; d)Rougher flotation (high volume); e) Cleaner flotation (three stage); f)Waterwashing for concentrate and debrining; g) Drying; h) Classification of intermediate-quality product; i) Residue (tails)collection; j) Residue debrining; k) Cyclone; l) Clarification of brine for recirculation


(! Flotation, Section 5.4.4.! Flotation, Sec-tion 5.4.5.).

Hard-Salt Flotation is similar to the sylvi-nite flotation process (Fig. 31). However, theflotation product also contains kieserite, whichcannot be removed by washing because it dis-solves only slowly in water.

Sylvite, which is intergrown with kieserite,must therefore be liberated as fully as possible,which leads to increased grinding costs. Becauseof the increased production of fines, more flota-tion agent is needed. Also, in warm summermonths, potassium-containing double salts (leo-nite, schoenite) can separate from the circulatingbrine. Since these salts end up in the residue, theyield is reduced.

Mixed-Salt Flotation. Carnallite (KCl �MgCl2 � 6 H2O) in the mixed salt is usually de-composed by a brine with low MgCl2 content.The potassium chloride produced has a grain sizeof < 0.04 mm and is accompanied by sylviniteore or hard salt components.

In the 1980s, this complex salt mixture wassuccessfully treated by flotation alone for the firsttime [99, 100].

Potassium chloride produced by the decom-position is first floated with a fairly small amountof amine in pneumatic or stirred flotation cells;then the sylvite component of the sylvinite ore or

hard salt is floated with a further measuredamount of amine, and a discardable residue isobtained (Fig. 33). The flotation froth is thentreated by flotation in cleaner cells to give aconcentrate with > 55% K2O. A final washingprocess gives a potash fertilizer salt containing> 60% K2O.

Polymineral Salts are rarely treated by flo-tation. Potash ore from Stebnik (CIS) containsnot only sylvite, halite, and clay materials, butalso large amounts of kainite, langbeinite, andpolyhalite [101]. Clay depressants are added first,followed by flotation of halite with a mixture ofC7–C9 fatty acids. Sodium hydroxide solution isused to adjust pH [101].

Schoenite can be floated from a salt mixtureby using coconut oil acids as the collector [102].

The flotation of anhydrite to produce a saltmixture of kieserite, langbeinite, and polyhaliteis recommended in [103].

A combination of cold leaching and flotationfor the treatment of salt mixtures containinglangbeinite and polyhalite is described in [104].

5.4. Electrostatic Separation

Electrostatic separation depends on the direc-tional movement of electrically charged bodiesor particles in an electric field. The processes

Figure 32. Schematic of flotation of potash oresa) Screen; b) Crushers; c) Scrubber; d) Classifier; e) Thickener; f) Centrifuge; g) Brine tank; h) Dryer; i) Cleaner flotation cells;j) Rougher flotation cells; k) ConditionersReproduced from [98] with permission


used differ according to the various methodsof producing or separating the charges andaccording to the separation equipment used(! Electrostatic Separation). The separation ofmixtures of salt minerals (e.g., in the treatmentof crude potash ores) invariably involves theselective exchange of electric charges that takesplace on contact between the various mineralparticles, followed by separation in free-fallseparators. Although roll separators are widelyused for treating other minerals, they have notbeen used for salts, mainly because of their lowthroughput.

The first investigations into the industrial-scale separation of potassium chloride andsodium chloride were carried out after WorldWar II by the International Minerals & ChemicalCorp. in Carlsbad, New Mexico [105]. The ore,which contains alkali-metal chlorides andvarious amounts of clay and sulfate minerals, wasground, heated to ca. 500 �C, cooled to ca.110 �C, and separated in a free-fall separator

with a field strength of 2–6 kV/cm. Resultswere not encouraging and the method wasabandoned.

Research in the Kaliforschungs-Institut inHannover in 1956 led to an industrial break-through. The addition of organic and inorganicreagents much improved the electrical chargeex-change between the mineral components,and the separating temperature could be reducedto < 100 �C [106, 107]. The potash works ofNeuhof-Ellers, a subsidiary of Kali und Salz,was the first industrial plant to use this process toproduce kieserite in 1974. In the followingyears, plants for the electrostatic production ofkieserite and potash concentrates, and for thedry removal of residues, were installed in threeother factories, with capacities up to ca. 1000 t/h[108]. Investigations aimed at the introductionof this process for processing sylvinite ores fromthe potash deposits in Saskatchewan, Canada,have been carried out by PCS Mining, Saska-toon [109].

Figure 33. Flotation of fine salt (< 0.1 mm) from carnallite decomposition, showingK2O content of each size fraction [100]Yields are based on amounts of K2O in each fraction


5.4.1. Theoretical Basis

The basis of the process is the mutual selectiveexchange of electrical charge between the saltcomponents, which occurs on contact. The di-rection, selectivity, and intensity of the chargeexchange can be influenced by a large number ofreagents (conditioning agents) [110]. In additionto this chemical conditioning, treatment with airof specific relative humidity is necessary. This isusually controlled by means of the air tempera-ture and should be between 5 and 25% [110],[111]. By the appropriate choice of conditioningagent and relative humidity, the charging prop-erties of the individual mineral components in apotash ore can be controlled so that the desiredcomponents are recovered. In this way, an ore ofcomplex composition can be completely sepa-rated into its components [112].

The mechanism of charging by contact de-pends on the transfer of electrons between thetouching mineral surfaces, which must havesuitable surface properties (i.e., surface energiesappropriate for the exchange of charge). Theseenergies are influenced (or created) by chemicalconditioning agents and partial water vapor pres-sure [113–115].

5.4.2. Equipment and Processes

Before the separation process, the potash oremust be size-reduced to give complete liberationof its components. Fine particles (< 0.1 mm)behave nonselectively in an electrostatic field,so the grinding process must be carried out ascarefully as possible (e.g., by impact grinding).The conditioning agents (20–100 g/t ore) areadded to the ore in a mixer or introduced in thevapor state into the fluidized-bed dryer that heatsthe salt to the separation temperature (25–80 �C),whereby the relative humidity is adjusted to asuitable value for selective charging of the saltparticles [116–118]. Alternatively, a rotary dryercan be used. Depending on the conditioningagent used and the relative humidity and temper-ature, sylvite generally becomes positivelycharged, whereas halite and kieserite can bepositive or negative. Accompanying mineralssuch as langbeinite, carnallite, or kainite can beseparated individually or together with othermineral components [119–122].

Separation of the mixture of charged mineralsis carried out in free-fall separators with anelectrical field strength of 4–5 kV/cm. For adistance between the electrodes of 25 cm, theapplied voltage is 100–125 kV. Formerly, verti-cal belt separators (Fig. 34) were common, butthey are now used only for special applications.The electrodes consist of rotating rubber beltswith conducting coatings. Brushes on the sideopposite the electric field remove salt fines,which otherwise settle on the electrode surface,forming a coating that weakens the electric fieldand hinders separation. The separator used ex-clusively today is the tube free-fall separator(Fig. 35), which was developed by the potashindustry [123]. It consists of two opposed rows ofsteel tubes, ca. 2 m long. The tubes rotate on theiraxes, and salt fines are removed by brushes on theside remote from the falling salt. The maximumworking length of separators of this type is ca.10 m. The charged salt mixture leaving the flu-idized bed is fed to the top of the separator andfalls through the electric field between the elec-trodes. This causes the particles to move

Figure 34. Vertical belt separatora) Rotary brush; b) Rubber belt; c) Stationary brush; d) Flaps


sideways, the direction depending on the sign ofthe charge. At the bottom of the separator, adjust-able flaps enable a cathodic fraction, an anodicfraction, and a middle fraction to be recovered.

These fractions are then treated further or re-moved from the process as product and waste (ormiddlings).

The particle sizes for complete liberationshould not exceed 1.5 –2 mm for sylvite andhalite, or 1.2 mm for kieserite, because themovement of particles in the electric field of theseparator is determined by both the horizontalelectrical force and the vertical effect of gravity.Since the electrical force depends on the chargeon the particle, which in turn is a function of thesurface area, the surface–volume ratio is a veryimportant factor in determining the extent ofsideways movement in the electric field. For finematerials, the effect of the electrical force is thegreater; the converse is true for coarse particles.

Modern tube free-fall separators have athroughput of 20–30 t h�1 m�1. Their energyconsumption is very low because for thisthroughput the current is only ca. 2 mA.

Generally, even for a single-stage separationstep two free-fall separators are arranged oneabove the other so that the concentrate andwaste material from the upper separator bypassthe lower one, while the middlings flow directlyto the lower separator where further separa-tion occurs. This is shown schematically inFigure 36. In most cases, multistage separationor treatment is necessary, in which the concen-trate produced in the first stage is purified orconcentrated in another single stage or in mul-tistage separation (Fig. 37).

Electrostatic separation has thus far been usedindustrially on a large scale only for the treatment

Figure 35. Tubular free-fall separatorA) Cross section; B) Plan viewa) Salt feed; b) Upper bearing;c) Motor; d) Tubes; e) Brushes; f) Lower bearing; g) Flaps;h) Receiver for products

Figure 36. Schematic of single-stage electrostatic separationa) Mixer; b) Fluidized-bed preheater; c) Elevator; d) Free-fall separator; e) Middle-fraction grinding


of hard salts with the principal components syl-vite, halite, kieserite, and various percentages ofcarnallite. Three basicmethods of separation intothe individual components are

1. Separation of sylvite in the first stage [105,111, 124], and separation of kieserite andhalite in the second stage

2. Separation of kieserite in the first stage [125,126], and separation of sylvite and halite inthe second stage

3. Separation of halite in the first stage[127–129], and separation of kieserite andsylvite in the second stage

In all cases, carnallite appears in the sylvitefraction.

The first stage of method 2 is used to producekieserite on a large scale, andmethod 3 is used inseveral variations for the production of a dryhalite residue and of kieserite and potash con-centrates [108]. So far, these plants have beenoperated almost entirely in association withleaching and flotation plants because electro-static treatment alone gives unsatisfactoryyields.

5.5. Heavy-Media Separation

Heavy-media separation is not used widely inthe potash industry because of the generallyunfavorable extent of intergrowth and the smalldensity difference between the main compo-nents sylvite (r ¼ 1.99) and halite (r ¼ 2.17).However, the process can be used for verycoarsely intergrown high-quality sylvinite oresas mined in Saskatchewan. International Miner-als & Chemical Corporation (IMCC) operates aplant of this type in Esterhazy, Saskatchewan[130]. The heavy medium is a suspension ofmagnetite in a saturated salt solution whosedensity is adjusted to 2.10. Small salt crystals(< 1.7 mm) are very difficult to separate frommagnetite; therefore, only materials with a par-ticle size of 1.7–8 mm are treated by this pro-cess. Separation is carried out in hydrocyclonesin two stages. A process scheme is given inFigure 38.

IMCC also operates a heavy-media separationplant in Carlsbad, NewMexico, where langbein-ite (r ¼ 2.83) is recovered from a potash ore inwhich sylvite and halite are the other maincomponents.

Figure 37. Flow diagram of electrostatic separation plant


5.6. Debrining and Drying

The products and waste from all the treatmentprocesses except for the dry electrostatic processare obtained as suspensions with various solidscontents and must be debrined. The two mainaims of debrining potassium chloride product are(1) to achieve as low a moisture content aspossible to minimize drying costs and (2) toremove as much of the adhering brine as possibleto maximize product purity. As the brine adher-ing to the residues contains potassium chloride,its recovery minimizes yield losses.

Suspensions must often be thickened in circu-lar thickeners or hydrocyclones before debrining.

The choice of equipment is determined main-ly by the particle size of thematerial to be treated.The extent to which adhering brine can be re-moved by washing with water is also important.

Pan filters are used where debrining of a fine-grained product is combined with water washingto increase theK2O content andwhen the product

is to be stored intermediately. Residual moisturecontent in this case is 12–14%.

Drum filters (! Filtration, 2. Equipment,Chap. 10.) are generally used for debrining fineresidues or when washing of the filter cake isnecessary; they give a residual moisture contentof 9–11%. Alternatively, belt filters (! Filtra-tion, 2. Equipment, Chap. 8.) are used becausethey have a high capacity and allow the filter caketo be washed with recovery of the washingliquids.

Themost commonly used debrining apparatusconsists of centrifuges of various designs (!Centrifuges, Filtering) [131]. In potash works inCanada, the United States, Jordan, and the CIS,screen-bowl and solid-bowl centrifuges of 1400-mm diameter are in general use. Screen-bowlcentrifuges have throughputs of 60–110 t/h andachieve a residual moisture content of 3–8%.They are used to treat both products and residues.Solid-bowl centrifuges mostly are used for fineresidues, and achieve throughputs of 70–120 t/h

Figure 38. Heavy-media separation of potash ore [130]Reprinted from [130] with permission of John Wiley & Sons, Inc.


and a residual moisture content of 6–8%. InEuropean and some Canadian factories, large-diameter (> 900 mm) pusher centrifuges areused, with throughputs of 40–50 t/h and a resid-ual moisture content of 3–6%. For coarse pro-ducts and residues, vibratory screen and screwscreen centrifuges are generally used. These havea diameter of 900–1200 mm and a throughput of35–70 t/h, and give a residualmoisture content of2–4%.

The products are usually dried in drum dryers(! Drying of Solid Materials, Section 2.1.2.)with diameters up to 3 m and lengths of 20 m[132]. They are heatedwith oil or gas in cocurrentflow at throughputs up to 120 t/h. The dryers arefitted with internals to promote heat exchangeand prevent caking of the salt. Exhaust gases arededusted, first by cyclones and then by electro-static filters, wet scrubbers, or fabric filters. Themain reasons for the widespread use of drumdryers are their ruggedness, safe operation, lackof sensitivity to throughput variations, and adapt-ability to differing grain sizes and moisture con-tent. Figure 39 gives a schematic diagram of adrying plant of this type.

Since the early 1960s, fluidized-bed dryershave been used to an increasing extent (! Dry-ing of Solid Materials). At first, they were usedmainly for coarse products (> 0.5 mm) with lowinitial moisture content. Later, improvements influidized-bed technology enabled products withgrain size down to 0.1 mm and initial moisturecontent up to 8% to be dried. Their most impor-tant advantages compared with drum dryers are

improved heat and mass transfer, more efficientuse of energy, and much smaller floor spacerequirement. Cyclones and bag or electrostaticfilters are used for dedusting the exhaust gasesfrom a fluidized-bed dryer. A typical plant isshown in Figure 40.

5.7. ProcessMeasurement andControl

The methods normally used in the chemicalindustry for measuring and controlling processparameters are also used in potash plants. Specialmethods include the analytical determination ofpotassium by means of its natural radioactivity[133], the use of flame photometry for so-calledratio analysis, and mineral analysis by infraredspectrometry.

The natural mixture of potassium isotopesincludes ca. 0.012% of the radioactive isotope40K, which is both a b and a g emitter. Both typesof radiation aremeasured to determine potassiumcontent. The b emissions are usually measured inthe laboratory,whereas on-linemeasurement of gemissions is widely used for process control.Owing to the great penetrating power of g rays,the reliability of measurement is strongly depen-dent on the geometry of the measuring equip-ment. Equipment for the radiometric determina-tion of potassium in bulk products is shown inFigure 41 [134].

The on-line method of ratio analysis [135,136] of KCl–NaCl mixtures is based on measur-ing the potassium and sodium contents of a

Figure 39. Drying plant for potassium chloridea) Combustion chamber; b) Rotary dryer; c) Deduster cyclone; d) Electrostatic filter; e) Screen; f) Grinding equipment


sample with a double-beam flame photometer,and calculating the ratio from the measured data.Since this ratio is independent of the weight anddilution of the sample wet materials with varyingwater content can be analyzed without weighingthe sample. Ratio analysis is used for determin-ing residual NaCl content in flotation concen-trates that have been water washed. The NaClcontent is used to determine the amount ofwashing water required, and enables losses ofK2O due to excessive use of washing water to beavoided. The method has also been used forcontrolling the K2O content in products fromleaching–crystallization plants.

Infrared spectroscopy is suitable for continu-ous determination of carnallite or kieserite incrude potash salts. Reflection photometry can beused to measure water of crystallization, andhence the content of these minerals. In somecircumstances, the carnallite content can be de-termined even when kieserite is present [136].

5.8. Waste Disposal andEnvironmental Aspects [137, 138]

The main environmental problem of thepotash industry is disposal of process waste. The

Figure 40. Fluidized-bed dryer with dedusting equipmenta) Fan; b) Combustion chamber; c) Fluidized-bed dryer; d) Deduster cyclone; e) Suction bag filter; f) Exhaust fan; g) Fan forcooling air

Figure 41. Equipment for radiometric determination ofpotassium in solid materialsa) Inlet for material to be analyzed; b) Container for materialto be analyzed; c) Bypass; d) Detector; e) Time-controlledvalve


total world production of potash ore is ca.250� 106 t/a, whose processing necessitates thedisposal of ca. 200� 106 t of waste withoutdamage to the environment [137]. In theCanadian province of Saskatchewan alone,300� 106 t of solid waste has been generatedduring the last 30 years, covering an area of ca.35 km2 with solid and liquid materials [139].

The composition of the waste depends on thetype of ore treated. Waste from the treatment ofsylvinite consists mainly of halite. Waste mate-rials from hard salt treatment are halite andkieserite, and from carnallitite ore processing,halite and magnesium chloride, which is alwaysproduced in the form of an aqueous solution. Saltsolutions that must be disposed of are also gen-erated during the production of potassium sulfatefrom potassium chloride and magnesium sulfate,and during the recovery of kieserite from resi-dues from the treatment of hard salt by dissolvingthe halite (! Magnesium Compounds).

Four methods for disposing of waste aredumping, backfilling, pumping into the ground,and discharge into natural water systems.

The disposal of waste by dumping is by far themost important method. Salt solutions that runoff the dumped materials must be demonstrated

not to harm the environment when they areabsorbed into the ground. Salt solutions canoriginate from the brine adhering to wet residuesor from the carrier liquid for transporting solidwaste to the dump, or they can form whenatmospheric precipitation dissolves the salt fromwaste material. If the ground underneath thedump is not impermeable, it must be sealed bylayers of clay or plastic sheeting, or a combina-tion of the two. The salt-containing runoff wateris collected in ditches at the edges of the dump,and as much as possible is returned to the re-circulating brine system in the plant. Excess brineis disposed of along with other liquid waste. InGermany, solid waste is formed into steep coni-cal heaps after drying or debrining as fully aspossible. This reduces the amount of salt-containing runoff water formed by atmosphericprecipitation and also minimizes the ground arearequired. In most Canadian installations, filtra-tion residues (tails) are slurried with brines andpumped as a suspension into a large lagoonsurrounded by dykes. Flat deposits are formedover a very large area. The brine that runs off ispumped back to the plant and reused for slurryingsolid waste (Fig. 42) [140, 141]. If the salt solu-tion enters groundwater-bearing layers despite

Figure 42. Conventional potash waste handling in Saskatchewan [139]Reproduced with permission


sealing, boreholes are sunk, and salt-bearinggroundwater is pumped back into the lagoon.Attempts to cover dumped waste material toprevent the formation of salt-containing waterby the leaching effect of rainfall have in the longrun proved unsuccessful [139].

Under certain geological conditions, and ifmining methods are suitable, solid residues canbe transported underground for backfilling. Sincethe bulk density of the residue ismuch lower thanthat of the potash ore, often only a part of theresidue can be accommodated by the space leftafter extraction of the ore. Backfilling is the mainmethod of waste disposal in North German saltworks where the salt beds are steeply inclined, aswell as the potash works of New Brunswick inCanada. In most potash works, where the potashis mined from level deposits, backfilling is notpossible for technical and economic reasons.

Pumping salt solutions back into the ground ispossible if certain geological requirements aremet. The formation used for this purpose mustpossess sufficient porosity and permeability, andmust have no contact with formations that couldprovide a water supply or contain salt deposits.Salt solutions are generally pumped under pres-sure through lined boreholes into the porousformation.

Since 1926, in the Werra potash region ofGermany, large quantities of brine have beenpumped into a porous dolomite layer 20–25 mthick. Injectionwells have an absorption capacityup to 1000 m3/h at head pressures up to 11 bar[142]. Since most of this waste brine comes fromkieserite production, the amount produced hasdecreased drastically with the introduction ofelectrostatic ore treatment [143]. In Saskatche-wan, excess brine is pumped into deep forma-tions of dolomite or sandstone. The capacity ofthesewells is up to ca. 200 m3/h at head pressuresup to 60 bar. Kalium Chemicals, Saskatchewan,has been disposing of waste brine since 1979 incaverns produced by solution mining.

The possibility of disposing waste brine inrivers and lakes depends very much on the loca-tion. Sea disposal, as practiced in the UnitedKingdom and by one of the potash works in NewBrunswick, presents few major problems if theoutfall is sufficiently remote from the coast. Thepotash works on the Dead Sea can dispose ofwaste brine in the Dead Sea itself without harm-ful consequences. The Potash Company of

America in Saskatchewan discharges its solidand dissolved wastes into the southern end ofthe salt-containing Lake Patience, which is iso-lated from the rest of the lake by a dam.

For most potash works, waste brine cannot bedisposed in natural salt water. If undergrounddisposal is impossible, disposal into flowingnatural water is the only alternative. In all coun-tries, this is subject to ever-stricter regulation.One particular problem, for which a solution inthe foreseeable future is being sought, is the highsalt content in the Werra River due to brine fromthe potash works in the eastern part of the Werraregion [143].

The only significant atmospheric pollutioncaused by potash works is salt dust emitted bythe drying plant and from the handling of the oreand products during production and supply. Dustremoval from waste gases from dryers is dis-cussed in Section 5.6. The dusts produced duringproduction and conveying are usually removedby air extraction, and trapped in fabric filters orwet scrubbers. During the drying of productscontaining magnesium chloride brines, hydroly-sis of the magnesium chloride can lead to theemission of hydrogen chloride, which can beremoved from the gas by wet scrubbing or ab-sorption in calcium hydroxide in combinationwith awoven filter [144]. This procedure can alsogreatly reduce the level of sulfur dioxide if it ispresent in the exhaust gas.

5.9. Granulation [145–149]

Potassium chloride can be produced in a widerange of crystal sizes, depending on the compo-sition of the potash ore, its degree of intergrowth,and the process used. Different particle-size dis-tributions are needed for various applications. Tomeet these requirements, the potash industryoffers products with standardized size distribu-tions (see Section 5.10) obtained by screening togive the various fractions. The resultant distribu-tion of the product among the various standard-ized grades does not always correspond tomarketrequirements. Demand for products with a parti-cle size of ca. 1–5 mm (coarse and granulargrades) exceeds their normal production rate ina potash works. Most potash works must there-fore increase the proportion of coarse product bygranulating part of the primary product.


The main reason for the high demand forgranulated potassium chloride is the techniqueof applying fertilizers that was developed mainlyin North America and Western Europe in thepostwar years. This technique, which is now themostwidely used, requires coarse particleswith arather narrow size range. This is needed both forsingle nutrient fertilizers and for bulk-blendedmaterials, which are widely used in NorthAmerica. Also, granulated potassium chloridehas a lower tendency to cake or form dust thanthe fine product.

Two methods of granulation are commonlyused in the fertilizer industry: agglomeration ofmolten or wet material in rotating drums ordishes, or compaction of dry material in rollpresses. The latter is the method most often usedfor potash fertilizers.

The starting material for compaction usuallyconsists of a mixture of fine material from theproduction process with recycled material fromthe grinding and screening system of the com-paction plant. Although dusts from the plant canalso be compacted, a high proportion of fines(< 0.1 mm) must be avoided because they causeproblems in the compaction equipment. Aminesfrom the flotation process on the surfaces of theparticles can also interfere with compaction and

must be destroyed or rendered inactive by heat orchemical treatment.

Roll presses used in the potash industry usu-ally have 60–125-cm-long rollswith diameters of60–100 cm (! Size Enlargement). The feedmaterial, which is usually at 100 –120 �C, isgenerally predensified by force feeders that feedit into the nip between rolls, where it is deaeratedand compressed, with plastic deformation ofeach particle, to produce a dense sheet of mate-rial. The compression force is 40–50 kN/cm ofroller length. The sheet of material is fed to aprebreaker that size-reduces it for ease of trans-portation. It then goes to a grinding–sievingplant, which produces either a granular productor granular and coarse products. Undersize ma-terial is recycled to the compaction press (!Fertilizers, 4. Granulation, Section 3.5.). A typi-cal flow diagram for a compaction–granulationplant is given in Figure 43.

The compaction product consists of irregu-larly shaped angular particles. Handling causesabrasion of the edges and corners to form un-wanted dust. The granular material is thereforeusually treated with water after screening andthen dried in a fluidized-bed dryer at 180 –200 �C. This smooths the corners and edges, andgives a dense surface to the granules. Another

Figure 43. Flow sheet of a modern potash compaction–granulation plant [149]a) Double screens; b) Hammer mills; c) Compactors; d) Screens; e) Fresh feed vessel; f) Recycled materials vesselReproduced from [149] with permission


method of dust reduction is to treat the granuleswith liquids that bind the dust formed by abrasion(see Section 5.10).

5.10. Quality Specifications

The main application of potassium chloride is inpotash fertilizers, either as a single-nutrient fer-tilizer or as the potash component in mixed orcomplex fertilizers. In English-speaking areasand on the international market it is often calledmuriate of potash (MOP).

Single-nutrient fertilizers, which were for-merly themost common type, have been replacedlargely by complex or mixed fertilizers. Also, thefinely divided fertilizers formerly used have in-creasingly been replaced by granulated products.Single potash fertilizers often contain the addi-tional nutrient magnesium sulfate.

Potassium chloride must have differing grainstructure and nutrient content depending on itsintended use, e.g., for the production of granu-lated NPK or PK fertilizers, for application assuspension or liquid fertilizers, for bulk blendingwith other components (! Fertilizers, 2. Types,Chap. 1.), or as a single-nutrient fertilizer. Thepotash industry has developed internationallyaccepted quality standards.

The three main grades differ in grain size:

Standard 0.2–0.8 mm

Coarse 0.8–2.0 mm

Granular 1.2–3.5 mm

Other products, such as those designated soluble,special standard, or special fine, can contain ahigh proportion of grains <0.2 mm. Grain-sizedistributions of the principal grades are listed inTable 7. Products fromvariousmanufacturers candiverge considerably from these specifications.

The granular and coarse products are used forbulk blending with other granulated fertilizers oras single-nutrient fertilizers. The standard andother fine materials are used in the manufactureof granulated multinutrient fertilizers, and forsuspension and liquid fertilizers.

The potassium chloride content, which forfertilizers is usually given as % K2O (100% KCl63% K2O), is generally at least 60% K2O (95%

KCl) for this application. The main impuritiesinclude sodium chloride, or magnesium sulfate,and sometimes anhydrite or clay minerals, de-pending on the raw materials and productionmethod. Fertilizer-quality potassium chlorideproduced by flotation is often colored red tored-brown by hematite inclusions or salt clay.

For single-nutrient fertilization, especially inEuropean agriculture, potash fertilizers with alow K2O content are often used. These generallycontain a guaranteed level of water-soluble mag-nesium salts. In Germany, for example, twogranulated potash fertilizers with a guaranteedMgOcontent arewidely used: 40erKornkaliwithMgO (40%K2O, 6%MgO) andMagnesia-Kainitcoarse (11% K2O, 5% MgO, 24% Na2O).

About 5% of the potassium chloride producedworldwide is used as an industrial chemical,mainly for the production of potassium hydrox-ide (see Chap. 11) and chlorine by chloralkalielectrolysis. Material of this quality is producedby recrystallizing low-purity potassium chloride.A distinction is made between industrial-gradeand chemical-grade material, depending on theimpurity content and area of use. Typical analy-ses of these products are

Industrial grade: KCl: 99.0–99.5%

NaCl: 0.8– 0.3%

Chemical grade: KCl: 99.8–99.9%

NaCl: 0.05–0.03%

The remaining impurities consist mainly ofbromide and alkaline-earth sulfates, dependingon the raw material and production process.

Dust content and free-flowing properties areimportant quality criteria, and depend on chemi-cal composition and grain size. With granulatedmaterials, dust formation by abrasion occurs dueto handling during manufacture and transporteven if the materials are specially treated duringproduction (see Section 5.9). To bind the dust, thegranulated material is treated with conditioningagents such as polyglycols, mineral oil products,vegetable oils and waxes, or mixtures of these.The addition rates are usually 0.3 –5.0 kg/t[150], [151]. Various methods have been devel-oped to assess or measure the abrasion resistanceand dust content of the granulated product. Abra-sion resistance is tested by subject-ing samples ofthe granules to a process of screening or rolling in


a rotating drum with steel spheres or rods, andweighing the fine material formed after a giventime [152]. Dust content is determined by airblowing a sample under precisely defined con-ditions and weighing the dust collected [150].

Fine potassium chloride (standard quality orfiner) tends to cake on storage or transport overlong distances. Anticaking agents are thereforegenerally added, usually mixtures of aliphaticamines, or sometimes higher fatty acids. Saltproducts that do not contain alkaline-earth sul-fates are treated with potassium hexacyanofer-rate(II). Addition rates for organic anticakingagents are 50–300 g/t, but potassium hexacya-noferrate(II) is effective even at 15–25 g/t [150,151]. Anticaking agents can affect the wettabilityof potassium chloride, which in turn can affect itsgranulation properties when complex fertilizersare being produced. The anticaking treatmentmust therefore suit the user’s requirements ex-actly [153].

5.11. Toxicology and OccupationalHealth

No toxic hazards are associated with the normalhandling of potassium chloride. According toUSP XVII of 1970, the usual therapeutic dose(e.g., for treating potassium deficiency) takenorally would be 1–10 g/d. The LD50 (oral, guineapig) is 2500 mg/kg.

Protective measures for storage and handlingand personal protection such as breathing appa-ratus or gloves are unnecessary.

5.12. Economic Aspects and Uses

A review of world potash production since 1900is given in Table 8. The figures show the totaloutput from the potash industry including potas-sium sulfate and potassium products for indus-trial use [4, 154, 155].

Table 8. World potash production (in 1000 t K2O) by country

Country Year

1900 1910 1920 1930 1940 1950 1960 1970 1980 1985 1986 1987 1988 1989

Germany 322 905 924 1381 1746

West

Germany 906 1978 2306 2737 2583 2161 2201 2290 2186

former East

Germany 1200 1598 2419 3422 3465 3485 3510 3510 3200

France 194 506 517 896 1580 1765 1894 1745 1629 1539 1502 1195

Spain 25 88 159 265 521 658 659 702 740 766 742

Italy 25 152 102 143 109 117 126 154

United

Kingdom 306 337 391 429 460 463

Middle and

Western

Europe 322 905 1118 1912 2351 3161 5446 7163 9119 8932 8477 8536 8654 7940

United

States 44 56 344 1168 2394 2467 2240 1245 1171 1218 1461 1580

Canada 3173 7300 6637 6698 7267 8327 7360

North

America 44 56 344 1168 2394 5640 9540 7882 7869 8485 9788 8940

Poland 1 70

former Soviet

Union 221 312 1084 4087 8064 10 367 10 228 10 888 11 300 10 231

Eastern

Europe 1 70 221 312 1084 4087 8064 10 367 10 228 10 888 11 300 10 231

Israel 45 83 546 790 1163 1240 1253 1244 1273

Jordan 545 662 722 805 792

Dead Sea 45 83 546 790 1708 1902 1975 2049 2065

Brazil 11 37 48 109

Congo 123

China 26 20 40 24 40 54 56

Others 149 20 40 35 77 102 165

Total 322 905 1163 2038 2961 4641 9007 17 585 27 533 28 929 28 511 29 961 31 893 29 341


Germany was the sole producer up to WorldWar I, and was later joined by France and theUnited States, and then Spain, the Soviet Union,and Poland. After World War II, the leadingproducers were the central and western Europeancountries and the United States. In the 1960s, theSoviet potash industry grew strongly, and theSoviet Union became the leading producer. Alsoin the 1960s, the first potash works in Saskatch-ewan was started up. In a few years, several largepotash works were in operation there, andCanadabecame the second largest producer after theSovietUnion.Thecapital investment in theSovietUnion and Canada and the rapidly increasing useof fertilizers in agriculture in the 1960s and 1970sled to a steep increase inworld potash production.Since 1980, the average annual increase in worldpotash production has been only 0.7%.

Almost two-thirds of world potash productionis exported. All the potash-producing countriesare exporters except for Brazil and China.Canada is by far the largest exporter. Worldexports of potash in 1989 (in 103 t K2O) wereas follows [154]:

Canada 6 808

France 517

former East Germany 2 251

West Germany 1 339

Israel 1 007

Italy 64

Jordan 755

Spain 354

United Kingdom 251

United States 482

former USSR 3 407

Total 17 235

The economic situation, particularly in devel-oped countries, greatly influences the extent andregional distribution of exports. Both the quanti-ty exported and its distribution among consumersare greatly affected by the state of their agricul-ture, especially in developed regions, and by thedemand for or availability of convertible curren-cy in the exporting or importing country. Trans-port costs for potash fertilizers have a consider-able bearing on total cost to the consumer, andlogistical considerations also influence the direc-tion and size of exports or imports. Finally,fluctuations in the rate of exchange of currenciesbetween the countries concerned are veryimportant.

Because of the conditions described above,certain special regional relationships developed.The agricultural requirements of the former CO-MECON countries were satisfied by the Sovietand East German potash industries only. In west-ern Europe, the market was supplied almostentirely bywestern European producers. InNorthAmerica, Canadian and United States producerswere in a dominating position. Canada had agood export market in Asia, as did Europe andJordan. The Latin Americanmarket was suppliedmainly by Canadian producers, together withformer East Germany and the former SovietUnion. The political and economic changes inthe former Eastern Bloc, the unification ofGermany, and the collapse of the dollar have allled to changes in the supply–demand relation-ships described above, although the distancebetween the producer and the consumer is stillof overriding importance. The future develop-ment of potash exports will be influenced greatlyby imports into China and Brazil. Both countrieshave onlyminimal production and are compelledto import fertilizers on a large scale. Problemsassociatedwith their internal economies andwithforeign exchange have thus far limited imports[156, 157].

The capacities of the potash producers invarious countries for 1990–1991 (in 103 tK2O) were [155]:

Middle and Western Europe 8 700

Germany 5 700

France 1 500

Spain 750

Italy 250

United Kingdom 500

North America 13 028

United States 1 838

Canada 11 190

former Soviet Union 12 880

Western Asia 2 220

Israel 1 380

Jordan 840

Brazil 150

China 50

Others 200

Total 37 218

Estimated world demand for potash fertilizersin the business year 1990–1991 by regions (in1000 t K2O) was as follows:


Europe 7 520

Eastern Europe 2 120

Western Europe 5 400

former Soviet Union 5 600

America 7 320

North America 5 100

Central America 350

South America 1 870

Asia 4 705

Western Asia 155

Southern Asia 1 360

Eastern Asia 3 190

Africa 522

Oceania 260

World 25 927

In using these figures, it should be borne inmind that the capacities quoted by individualproducers are generally too high and that onlyca. 95% of the total potash production is used inthe form of fertilizers. Hence, the total consump-tion of products of the potash industry is ca.1.5� 106 t of K2O higher than the figure givenabove. Nevertheless, considerable overcapacityexists worldwide, as can be seen by comparingthe two tables.

For the use of potassium compounds in ferti-lizers, see ! Fertilizers, 2. Types.

Industrial-grade and chemical-grade potassi-um chloride are used mainly for the electrolyticproduction of potassium hydroxide. Other im-portant uses include the production of drillingfluids for the oil industry, aluminum smelting,metal plating, production of various potassiumcompounds, and applications in the food andpharmaceutical industries [158].

6. Potassium Sulfate [159]

Potassium sulfate [7778-80-5], K2SO4, mineralname arcanite, forms colorless, nonhygroscopiccrystals. It occasionally occurs in nature in thepure state in salt deposits (e.g., in Germany, theUnited States, and the CIS) but is more widelyfound in the form of mineral double salts incombination with sulfates of calcium, magne-sium, and sodium (see Table 1). Potassium sul-fate is, after potassium chloride, the most impor-tant potassium-containing fertilizer, being usedmainly for special crops. Potassium sulfate con-stitutes ca. 5% of the world demand for potashfertilizers.

6.1. Properties

Potassium sulfate forms orthorhombic crystals,which transform to the trigonal modification at583 �C. Some properties of potassium sulfate arelisted below:

Mr 174.25

fp 1069 �CCrystal system and type orthorhombic D16

2h

Phase change at 583 �CCrystal system and type at > 583 �C trigonal D3

3d

Refractive indices n20D 1.4933; 1.4946;

1.4973

Density 2.662 g/cm3

Specific heat capacity cp 752.9 J kg�1 K�1

Heat of fusion 197.4 kJ/kg1

Heat of transformation (orthorhombic/trigonal)48.5 kJ/kg1

Enthalpy of formation DH 0 � 1438 kJ/mol

Entropy S 0 175.6 J mol�1 K�1

Dielectric constant (at 4� 108 Hz) 6.3

Thermal coefficient of expansion (cubic) 130� 10�6 K�1

Apart from the naturally occurring double-saltmineralsmentioned above, potassium sulfate alsoforms double salts and mixed crystals with am-monium sulfate and the sulfates of beryllium,magnesium, calcium, strontium, barium, andlead. It is reduced to potassium sulfide or potassi-um polysulfides by reducing agents such as hy-drogen and carbonmonoxide at high temperature.

The solubility of potassium sulfate in water islisted in Table 9. The cryohydric point is � 1.51C (7.1 gK2SO4/100 gH2O), and the boiling pointof the saturated solution is 101.4 �C (24.3 gK2SO4/100 g H2O). The solid phases formed inthe systemK2SO4–H2OareK2SO4,K2SO4 �H2O,and ice. In aqueous ammoniacal solution, thesolubility decreases rapidly with increasing am-monia concentration [160]. Potassium sulfate isvirtually insoluble in industrial organic solvents.

6.2. Raw Materials

Potassium sulfate is produced from single ormixed minerals or brines, or by the reaction of

Table 9. Solubility of potassium sulfate in water (g/100 g) [15]

Temperature, �C 0 10 20 30 40 50

Solubility 7.35 9.24 11.1 12.9 14.8 16.6

Temperature, �C 60 70 80 90 100

Solubility 18.4 20.0 21.5 22.8 24.0


potassium chloride with sulfuric acid or sulfates[161]. The economically important minerals in-clude the deposits of hard salt in Germany,langbeinite in New Mexico (United States), andkainite in Sicily. In the United States, sulfate-containing crystalline products from the evapo-ration of water from the Great Salt Lake (Utah)and Searles Lake (California) are also used forthe production of potassium sulfate.

Potassium chloride is usually converted topotassium sulfate by reaction with sulfuric acid,but SO2–air mixtures can also be used. Reactionswith sodium sulfate and gypsum are also ofinterest. Small amounts of potassium sulfate areobtained during the production of alumina fromalunite.

6.3. Production

The choice of production method and the loca-tion of a potassium sulfate plant depend onhaving a plentiful economic supply of the startingmaterials and being able to utilize or dispose ofthe byproducts or waste. Most production plantsare located on salt deposits from which at leastsome of the rawmaterials can be obtained. Plantsin which potassium chloride is reacted withsulfuric acidwith liberation of hydrogen chlorideare usually located in regions with a demand forhydrochloric acid, (e.g., for a acidification ofpetroleumboreholes), or they operate in conjunc-tion with a chemical works having a process thatuses hydrogen chloride. In such cases, the ex-ploitability of hydrogen chloride determines thecapacity of the potassium sulfate plant.

6.3.1. From Potassium Chloride andSulfuric Acid (Mannheim Process)

The reaction of sulfuric acid with potassiumchloride takes place in two stages:

KClþH2SO4!KHSO4þHCl

KClþKHSO4!K2SO4þHCl

The first reaction step is exothermic and pro-ceeds at relatively low temperature. The secondis endothermic and must be carried out at highertemperature. The relationship between total re-

action time and temperature is shown in Figure44. In practice, the process is operated at 600 –700 �C. To minimize the chloride content of theproduct, a small excess of sulfuric acid is used,which is later neutralized with calcium carbonateor potassium carbonate, depending on the purityrequirements for the product.

The reaction is usually carried out in so-calledMannheim furnaces (Fig. 45) [162].

The furnace has a closed dish-shaped cham-ber, with diameter up to 6 m, heated externallyby an oil or gas burner. Potassium chloride andsulfuric acid are fed into the chamber in therequired ratio at an overhead central point. Themixture reacts with evolution of heat and ismixed by a slowly moving stirrer fitted withstirring arms with scrapers (rabbles), which pro-pels themixture from the center of the chamber tothe periphery. Potassium sulfate leaves the reac-tion chamber at this point and is neutralized andcooled. It normally contains 50–52% K2O and1.5–2% chloride. Hydrogen chloride gas formedis absorbed in water to form hydrochloric acid orused in gaseous form.

The Mannheim process is the most widelyused method of producing potassium sulfate dueto its simplicity, high yield, and themanyways inwhich the byproduct can be utilized. Hydrogenchloride is used to produce dicalcium phosphate,vinyl chloride, or calcium chloride if it cannot besold as hydrochloric acid.

Disadvantages of the process include highenergy consumption, severe corrosion, and highcapital cost. In the United States, reductions incorrosion and energy consumption are achievedby using the Cannon process, in which the reac-tion is carried out in a directly fired fluidized bed.

Figure 44. Temperature dependence of the reactionbetween potassium chloride and sulfuric acidReproduced from [162] with permission


Another variation is the Hargreaves process,which is also used in the United States. Bri-quetted potassium chloride is heated in reactionchambers in a stream of sulfur dioxide from thecombustion of sulfur, excess air, andwater vapor.The yield and the degree of conversion are bothca. 95%.

The Mannheim and Hargreaves processes arealso used to produce sodium sulfate from sodiumchloride and sulfuric acid. Mannheim furnacescan be used to produce potassium and sodiumsulfates alternately. Research into the reaction ofpotassium chloride with sulfuric acid in a liquid–liquid extraction process did not result in theconstruction of a production plant [163, 164].

6.3.2. From Potassium Chloride andMagnesium Sulfate [165, 166]

In a process used mainly in Germany, the sulfaterequired is provided by kieserite,MgSO4 �H2O, acomponent of German hard salt deposits. Meth-ods of extraction are described in! MagnesiumCompounds, Chap. 5. The reaction can be repre-sented by the following overall equation:

2 KClþMgSO4!K2SO4þMgCl2

Kieserite reacts very slowly and must beground finely before reaction. Alternatively, itcan first be recrystallized to give epsomite,MgSO4 � 7 H2O.

The basis of the process is explained in Fig-ure 46. The fundamental relationships for the

single-stage process of KUBIERSCHKY and the two-stage process of KOELICHEN and PRZIBYLLA areshown as broken lines on the isotherm diagram.

For the single-stage process, the most favor-able mixing ratio of the starting materials is givenby point C. In the presence of sufficient water, thismixture reacts to form potassium sulfate and asulfate mother liquor (point m). This solution hasthe highestmagnesiumchloride content attainableby direct reaction, which determines the yield.The magnesium content of solution m reaches amaximum at 25 �C, and the process is thereforecarried out at this tempera-ture. The single-stageprocess achieves a theoretical potassium yield ofonly 46.1% and sulfate yield of 67.5%.

For this reason, the two-stage process is nowused exclusively. In this process, the startingmaterials are first mixed in the presence of adefinite quantity of water corresponding topoint S to form schoenite, K2SO4 �MgSO4 � 6H2O. So-called potash–magnesia liquor,which has a high magnesium chloride content(point P), is also formed. The schoenite isreacted with additional potassium chloride(point D) to form potassium sulfate and sulfatemother liquor:

2 KClþ2MgSO4þxH2O!K2SO4�MgSO4�6 H2O

þMgCl2 ðaq:Þ

2 KClþK2SO4�MgSO4�6 H2OþxH2O

!2 K2SO4þMgCl2 ðaq:Þ

This process gives a theoretical potassiumyield of 68% and sulfate yield of 83.7%.

Figure 45. Schematic diagram of a Mannheim furnaceReproduced from [162] with permission


The flow diagram of the process is shownschematically in Figure 47.

In the first stage, called the potash–magnesiastage, schoenite or leonite, K2SO4 �MgSO4 �4 H2O, is produced by stirring solid epsomite orfinely ground kieserite with potassium chloridein sulfate mother liquor recycled from the secondstage. The suspension produced is filtered onrotary filters; the potash–magnesia brine, whichcontains 180–200 g/L magnesium chloride, isremoved; and the solid crystalline product, also

known as potash–magnesia, is fed to the nextstage, sometimes after being washed with sulfatemother liquor, where it is stirred with potassiumchloride solution at ca. 70 �C.The temperature ofthe mixture is 35–40 �C, and solid potassiumsulfate is formed. This is thicknened, debrined bycentrifuges, and dried in drum or fluidized-beddryers.

If the sulfate reaction is carried out in aclassifying crystallizer at a high solids content[167], a very pure, coarsely crystalline product is

Figure 46. Isothermals of the system K2–Mg–Cl2–SO2�4 –H2O at 25 �C according to J€ANECKE

Figure 47. Flow diagram of the two-stage production of potassium sulfate from potassium chloride and magnesium sulfate


obtained with K2O content of 53% and chloridecontent of < 0.5%.

In the industrial process, sodium chloride isalways present. If the molar ratio of Na2 : K2 inthe sulfate mother liquor exceeds 2 : 5, glaserite,3 K2SO4 �Na2SO4, is formed instead of potassi-um sulfate. To prevent this, the potassium chlo-ride used must be of adequate purity.

The yield is determined by losses to the wastebrine. The higher the magnesium chloride con-tent of the potash–magnesia brine is, the lower isthe potassium content, and therefore the potassi-um loss. Excess potassium-rich sulfate motherliquor must be recovered.

6.3.3. From Potassium Chloride andLangbeinite [168]

Large deposits of langbeinite, K2SO4 �2 MgSO4, can be found in New Mexico (UnitedStates). Langbeinite can be converted to potassi-um sulfate according to the following overallequation:

K2SO4�2MgSO4þ4 KCl!3 K2SO4þ2MgCl2

The potash ore also contains halite and vary-ing amounts of sylvite, fromwhich langbeinite isseparated by gravity separation, flotation, anddissolution of halite, giving various crystal sizes.The coarser langbeinite fraction is sold as pot-ash–magnesium fertilizer, and the finer fractionis reacted with potassium chloride to producepotassium sulfate. A flowdiagram for the processis given in Figure 48.

The potassium sulfate formed is granulatedand marketed in three different grain sizes: gran-ular (0.8–3.4 mm), standard (0.2–1.6 mm), andspecial standard. The latter has a high content ofgrains < 0.2 mm.

6.3.4. From Potassium Chloride andKainite [161]

In Sicily, kainite, KCl �MgSO4 � 2.75 H2O, isobtained from a potash ore by flotation. It is thenconverted into schoenite at ca. 25 �C by stirringwith mother liquor containing the sulfates ofpotassium and magnesium from the later stagesof the process. Schoenite is filtered off and

decomposed with water at ca. 48 �C. This causesmagnesium sulfate and part of the potassiumsulfate to dissolve and most of the potassiumsulfate to crystallize. The crystals are filtered anddried. The sulfate mother liquor is recycled to thekainite–schoenite conversion stage. The motherliquor produced there, which still contains ca.30% of the potassium used, is treated with gyp-sum, CaSO4 � 2 H2O, causing sparingly solublesyngenite [13780-13-7], K2SO4 �CaSO4 �H2O,to precipitate. Syngenite is decomposed withwater at ca. 50 �C, which dissolves potassiumsulfate and reprecipitates gypsum. The potassi-um sulfate solution is recycled to the schoenitedecomposition stage, and gypsum is reused toprecipitate syngenite. A simplified flow diagramof the process is given in Figure 49.

6.3.5. From Potassium Chloride andSodium Sulfate [164]

The production of potassium sulfate from potas-sium chloride and sodium sulfate takes place intwo stages, with glaserite, Na2SO4 � 3 K2SO4, asan intermediate, according to the following equa-tions:

4 Na2SO4þ6 KCl!Na2SO4�3 K2SO4þ6 NaCl

Na2SO4�3 K2SO4þ2 KCl!4 K2SO4þ2 NaCl

Figure 48. Flow diagram of the production of potassiumsulfate from langbeinite


Potassium chloride and sodium sulfate arereacted at 20–50 �C in water and recycled pro-cess brines to form glaserite, which is filtered andthen reacted with more potassium chloride andwater to form potassium sulfate. Because themother liquor from the glaserite stage has a highpotassium and sulfate content, the maximumpotassiumyield is 73%, and themaximum sulfateyield is 78%. The yield can be increased consid-erably by cooling the mother liquor to producemore crystals and by including a final evapora-tion stage.

A production plant in the CIS uses the glaser-ite process [166], and an experimental plant isoperating in Canada [169].

Alternatively, sodium sulfate solution can beused to charge an anion exchanger with sulfate,which reacts with potassium chloride solution togive a high yield of potassium sulfate [170, 171].This process is used atQuill Lake, Saskatchewan,Canada, which has a high sodium sulfate content.

6.3.6. From Potassium Chloride andCalcium Sulfate

Processes based on gypsum, CaSO4 � 2 H2O,have often been proposed, because it is so readilyavailable. Two processes have been tested inexperimental plants:

1. Reaction with potassium chloride in stronglyammoniacal solution

2. Reaction with anion exchangers

Potassium chloride reacts with gypsum inwater to give syngenite. If the reaction is carriedout in a concentrated solution of ammonia at lowtemperature, potassium sulfate with a very lowsyngenite content is obtained [163, 172, 173].

CaSO4�2 H2Oþ2 KCl!K2SO4þCaCl2 ðaq:Þ

By carrying out the reaction in two ormore stages, high concentrations of calciumchloride in the waste brine can be produced, andhence very high yields. The ammonia requiredfor the reaction medium must be recovered bydistillation. This process has attracted someinterest [174].

The anion-exchange process is carried out intwo stages [175]. First, the ion-exchange resin(R) is treated with a suspension of gypsum,charging it with sulfate. The charged resin isthen treated with a concentrated solution ofpotassium chloride, to replace the dissolved chlo-ride by sulfate. Potassium sulfate crystallizesfrom the solution, sometimes after the additionof solid potassium chloride. The process takesplace according to the following equations:

R2Cl2þCaSO4!R2SO4þCaCl2

R2SO4þ2 KCl!K2SO4þR2Cl2

6.3.7. From Alunite [176]

Alunite [12588-67-9], K2SO4 �Al2(SO4)3 � 4 Al(OH)3, occurs in several extensive deposits. On

Figure 49. Flow diagram of the production of potassium sulfate from kainite


being heated to 800–1000 �C, it decomposeswith liberation of sulfur trioxide to form a mix-ture of alumina and potassium sulfate. The lattercan be extracted from the mixture. A plant in theCIS uses this process.

6.3.8. From Natural Brines and Bitterns

In theUnited States, large quantities of potassiumsulfate are produced from the brines of the GreatSalt Lake and smaller amounts from the brines ofSearles Lake (see Section 8.3). Extensive inves-tigations and project studies have been carriedout in various countries into the extraction ofpotassium sulfate from the mother liquors (bit-terns) produced when salt is extracted fromseawater, and from concentrated brines in Tuni-sia and the Atacama desert of Chile [164].

6.4. Granulation

The demand for granulated potassium sulfate hasincreased greatly for the same reasons that thedemand for potassium chloride has increased.The method of production with compaction rollsis widely used (see Section 5.9), although potas-sium sulfate does not behave under pressure likepotassium chloride, which undergoes plastic de-formation with merging of the grain boundaries.However, the granulated product is much moredense and solid if the potassium sulfate is wettedbefore compaction by adding up to 2% water, orif steam is introduced through force feederslocated above the compaction rolls. Also, thepressure used must be considerably higher thanthat for potassium chloride [177, 178]. A com-paction plant for potassium sulfate is otherwisesimilar to that for potassium chloride, althoughthe throughput is much lower.


The important quality specifications for agricul-tural-grade potassium sulfate, known in theEnglish-speaking world and on the internationalmarket as sulfate of potash (abbreviated to SOP),are the K2O and chloride content. All producersguarantee a minimum K2O content of 50%(92.5% K2SO4), typical values being 50.5 –

51.0%. A completely water-soluble potassiumsulfate with K2O content of > 52% is suppliedfor the production of liquid fertilizers.

The maximum permissible chloride content is3%, according to EC Guidelines, but the usualcommercial upper limit is 2.5%. The chloridecontent depends very much on the productionmethod, and some producers offer potassiumsulfate with chloride content < 0.5%.

The most common impurities, apart fromalkali chlorides, are the sulfates of calcium andmagnesium.

The demand for granulated potassium sulfateas a single-nutrient fertilizer and for bulk blend-ing has greatly increased. The crystal size andsize distribution of the products granular, coarse,and standard are the same as those for potassiumchloride (see Section 5.10).

Potassium sulfate does not cake as readily aspotassium chloride and is therefore not treatedwith anticaking agents. Dust-reducing agents areadded at the rate of 5 kg/t to the fine and granu-lated products; the same additives are used as forpotassium chloride (see Section 5.10).

Industrial-grade potassium sulfate with aK2SO4 content of 99.6–99.9%, purified by re-crystallization, is supplied as a raw material forthe production of other potassium compoundsand as an auxiliarymaterial or reagent for variousbranches of industry. Specifications with respectto chemical purity and crystal size distributionare suited to the individual user’s requirementsand can sometimes be very detailed.

6.6. Toxicology and OccupationalHealth

No health hazard is associated with potassiumsulfate if it is handled in accordance with regula-tions. According to Swiss law relating to toxicsubstances, it is a Class 4 material (LD50: 500–5000 mg/kg). Otherwise, the information givenin Section 5.11 for potassium chloride applies.


Potassium sulfate accounts for ca. 5% of worldproduction by the potash industry, expressed inK2O units. It is produced in 11 countries. Of thetotal, ca. two-thirds comes from Belgium and


Germany. Other European producers are Italy,Spain, Finland, and Sweden. Potassium sulfate isalso produced by the CIS. Other productionplants are in the United States, Japan, SouthKorea, Taiwan, the Philippines, and China. Pres-ent world production and consumption both ex-ceed 1.5� 106 t K2O [179, 180].

A list of the most important exporting coun-tries (exports for 1985–1986 in 103 t K2O) isgiven below:

Western Europe 586

Belgium 256

Finland 18

West Germany 227

Italy 54

Spain 31

former East Germany 61

United States 74

Asia/Middle East 16

Israel 3

South Korea 8

Taiwan 5

Total 737

The consumption (1985–1986 in 103 t K2O)amounted to:

Western Europe 558

Eastern Europe 106

Africa 121

North America 115

Latin America 51

Asia/Middle East 271

Oceania 6

Total 1228

Potassium sulfate is up to twice as expensive aspotassium chloride based on K2O content due tothe costs of the rawmaterial (potassium chloride)and processing. It is therefore only used as apotash fertilizer for applications where it per-forms much better than potassium chloride.

The sulfur content of potassium sulfate is anadvantage where there is a deficiency of sulfur inthe soil. Also, it has only a slight oversaltingeffect on soil in arid or semiarid areas. It is usefulfor fertilizing crops that are sensitive to chlorideor whose quality is improved by chloride-freefertilization. This applies particularly to tobacco,vegetables, potatoes, vines, and citrus or variousother fruit [179, 181].

Potassium sulfate is used in industry for theproduction of other potassium compounds, ac-

celerators for rapid-setting cements, syntheticrubbers, desensitizers for explosives, lubricants,powdered fire extinguishers, dyes, explosives,and pharmaceuticals.

7. Potash–Magnesia

Crops sensitive to chloride can be fertilized withpotassium sulfate alone or by fertilizers contain-ing the sulfates of potassium and magnesium.These consist either of a mixture of the twosulfates or of the double sulfates schoenite, leo-nite, or langbeinite in dehydrated form.

Formerly, large quantities of amixture of fine-grained potassium sulfate and kieserite weremarketed under the name potash–magnesia(K2O: 27–30%, MgO: 9–12%). This has sincebeen almost completely replaced by a granulatedpotash–magnesia fertilizer, which is produced byfirst mixing potassium sulfate and kieserite at95 �C with hot synthetic langbeinite. The hotmixture is granulated in a drum granulator, andthe product is quickly cooled to < 60 �C. Itcontains 29–30% K2O and 10% MgO.

Langbeinite from the potash deposits in Carls-bad, NewMexico (United States), can be used inthe pure state directly as a potash–magnesiafertilizer. Crude langbeinite also contains haliteand varying amounts of sylvite, and the methodof purification utilizes the low rate of dissolutionof langbeinite in water. If the sylvite content islow, the potash ore is first ground to< 6 mm andthen simply washed with water to dissolve thehalite. The langbeinite that remains is dried andscreened to obtain the commercial size gradings.For higher sylvite content, an additional stage isrequired to separate the components by a combi-nation of gravity separation and flotation. Thelangbeinite produced has a K2O content of 20–22% and an MgO content of 18 –19%, and ismarketed as Sulpomag or K-Mag.

8. Production of Potassium Salts fromOther Raw Materials

Potassium salts occur not only in salt deposits,but also in solution in many types of inland lakesand in seawater. Where the concentration of thesalt solution is high enough and the climatic and


topographical conditions are suitable, potassiumchloride or potassium sulfate can be produced bysolar evaporation. Typical analyses of some ofthemore important sources are listed in Table 10.

8.1. The Dead Sea

Potash production by solar evaporation began atthe northern end of the Dead Sea in 1931. Later,production switched to the southern end but wasinterrupted by the war of 1947–1948. In 1952,potash production was resumed in Sodom by theIsraeli State Dead Sea Works. The capacity wasbuilt up in several stages and now amounts to> 2� 106 t of potassium chloride in all the usualcommercial grades [182, 183].

Brine from the lake is concentrated in evapo-ration ponds with a total area of ca. 90 km2, fromwhich the crystallized salts are recovered. Mostof the dissolved sodium chloride separates outfirst in the primary evaporation ponds. A mixtureof carnallite and sodium chloride then crystal-lizes in the main production ponds. This is re-moved as a suspension by suction dredgers andpumped to the works where the crystals arefiltered off and treated by the carnallite colddecomposition process (see Section 5.2.3). Mostof the resulting NaCl–KClmixture is then treatedin a hot leaching plant (see Section 5.2.2), wherepotassium chloride with> 60%K2O is producedin crystallizers. Part of the crop of crystals fromthe main production ponds is treated by a coldcrystallization process developed by the DeadSea Works. This decomposition process directlyyields a product containing > 60% K2O.

A potash works was started by the ArabPotash Co. in 1982 on the southeast bank of theDead Sea near Safi, Jordan. Evaporation andcarnallite production are carried out in a pond

system with a total area of 100 km2. The processresembles that of the Dead Sea Works. Magne-sium chloride is removed from carnallite in acold decomposition process, and the salt mixtureformed is then treated in a hot leaching plant togive potassium chloride [182, 184].

8.2. The Great Salt Lake [185]

The Great Salt Lake is the result of the shrinkageby evaporation of the former Lake Bonnevilleand lies in the eastern part of the basin. It has ahigh salt content and is the reason for the exis-tence of several plants that produce sodiumchloride and, since 1968, potassium salts.

At the western end of the Great Salt Lake, inthe Great Salt Lake Desert near Wendover, arethe Bonneville Salt Flats. Here, under a salt crust,are porous sediments containing brines, whichare regenerated by water from atmospheric pre-cipitation. Potassium chloride has been producedfrom these brines since 1937. They are collectedby a systemof ditches and evaporated in ponds. Amixture of potassium and sodium chloride crys-tallizes, from which potassium chloride is ob-tained by flotation [186].

Unlike the Wendover brines, the Great SaltLake brines contain considerable amounts ofsulfate (see Table 10). The process used inWend-over cannot therefore be used here. The oppor-tunity exists of obtaining substantial quantities ofvaluable potassium sulfate rather than potassiumchloride, which has been carried out since 1970in the Great Salt Lake Minerals & ChemicalsCorporation (GSLM & CC) plant at Ogden onthe east bank. Sodiumchloride is first crystallizedin the 56-km2 pond system until the solution issaturated in potassium salts. Further solar evap-oration then takes place in the main productionponds, producing a mixture containing varyingproportions of kainite, carnallite, and schoenite,with small amounts of sodium chloride [187].This mixture is converted into schoenite in theplant by treatment with recycled process brine.The sodium chloride that is not dissolved by thisreaction must be removed before further treat-ment. This is carried out by flotation of a sidestream [102]. Schoenite is then decomposed bywater, which produces very pure potassium sul-fate. The brine from this decomposition stage hasa high potassium content and is recycled to the

Table 10. Composition of some natural brines compared with seawa-

ter (in wt%)

Dead Sea Wendover brine Great Salt Lake Seawater

Kþ 0.6 0.6 0.7 0.04

Naþ 2.9 9.4 7.6 1.08

Mg2þ 3.4 0.4 1.1 0.13

Ca2þ 1.3 0.016 0.04

Cl� 17.0 16.0 14.1 1.94

SO2�4 0.04 0.2 2.0 0.27

Br� 0.5 0.01 0.006

H2O 74.3 73.3 74.5 96.5


first stage of the process. The brine produced bythe reaction at this stage is recycled to the evapo-ration pond. The sulfate content of the main cropof crystals is higher than the potassium content,so that further potassium sulfate can be producedby addition of potassium chloride from an exter-nal source [188].

The level of the Great Salt Lake increased somuch during the 1970s and 1980s that the pondsystem of the Great Salt Lake Minerals & Che-micals Corporation (GSLM & CC) overflowedduring 1984, and production had to stop [189].The lake level then fell, the pond system wasreinstated, and production started again in 1989.

8.3. Searles Lake [190]

In Trona, on the northwest bank of the almost drySearles Lake, brines are obtained that contain notonly sodium and potassium chlo-rides, but alsoconsiderable quantities of sulfate, carbonate, andborate ions. Recycled process brines are firstadded, and evaporation produces sodium chlo-ride and the double salt burkeite [12179-88-3],Na2CO3 � 3 Na2SO4. Potassium chloride is ob-tained by vacuum cooling of the potassium- andborate-containingmother liquor. Part of the chlo-ride is reacted with part of the burkeite to formglaserite, Na2SO4 � 3 K2SO4, an intermediatestage in potassium sulfate production. Anotherreaction used for potassium sulfate production isthat of potassium borate in the end brines withsulfuric acid, to form potassium sulfate and boricacid:

K2B10O16�8 H2Oþ6 H2OþH2SO4!K2SO4þ10 H3BO3

8.4. Other Sources

In addition to the sources mentioned under po-tassium sulfate, other salt lakes exist whosepotassium content would appear to offer thepossibility of extracting potassium salts. In1969 at Lake McLeod in Western Australia, aworks produced langbeinite, K2SO4 � 2 MgSO4,for a short period of time but ceased operationsfor unknown reasons. In China, in the QinghaiProvince, potassium chloride production hasbeen carried out for several years at Tsarhan

Lake by using solar evaporation. Large increasesin production are planned [191].

Seawater has a low potassium content (Table10) so that economical extraction of potassiumsalts is not possible. The production of sea salt bysolar evaporation in salt gardens yields residualbrines with increased potassium content. In someplaces (e.g., India), small quantities of low-per-centage potassium salts are produced from thesebitterns. However, economical production ofpotash fertilizers of marketable quality is notpossible from the amounts of mother liquoravailable from even the largest sea salt producers.

9. Storage and Transportation(! Fertilizers, 5. Analysis, Transport,Application, Section 2.1.)

The demand for potash fertilizers fluctuatesgreatly throughout the year, but because potashplants need to produce at as steady a rate aspossible, large storage capacities are needed toaccommodate periods of low demand. Therefore,potash plants usually have high-capacity productstorage facilities. Also, in seaports, where ferti-lizers are loaded onto ships, the largest potashcompanies or their subsidiaries have large stor-age capacities. In both cases, long storage shedsare used, usually with walls sloping to match theangle of repose of the potash salt. The sheds areusually filled by means of conveyor belts locatedunder the shed roof. They are emptied either bybucket loaders or scrapers that move the salt intoa channel under the floor of the silo or onto aconveyor belt at the side, which carries it viasloping bands or elevators to the loading plant.More recently, especially where there is a short-age of land, round silos have been used, oftenarranged in rows. These too are filled from aboveby conveyor belts and are emptied through open-ings at ground level. The majority of potashfertilizer is transported in bulk in self-dischar-ging wagons with a capacity up to 100 t, by rail,truck, or inlandwaterway. Transport from potashplants remote from a seaport or the main con-suming area is usually by special trains that runon a fixed timetable between the potash plant andthe seaport or intermediate storage facility. Forexample, the transport of potash fertilizers fromSaskatchewan to a cargo-handling plant in


Vancouver uses continuous-loop train tracks thatenable 10 000 t to be delivered in 102 wagonswith a capacity of 98 t each [192].

A small proportion of potash salts is suppliedin sacks, usually containing 50 kg. The sacks arefilled either by the supplier or at the loading plantat the seaport by automatic sack-fillingmachines.The sacks are usually paletized.

10. Analysis of PotassiumCompounds

Potassium is usually determined gravimetrically.In the United States, it is precipitated as thehexachloroplatinate [193]. Precipitation as thetetraphenylborate is anotherwidely usedmethod,being the standard ISOmethod for fertilizers, andcan be either a gravimetric or a volumetricprocedure (ISO 5318 and 5310) [194]. Precipita-tion as the perchlorate or tartrate is seldom used.Flame photometry is used for both laboratory andprocess control analysis. X-ray fluorescence canbe used for the analysis of solids or brines.

A review of methods recommended for po-tassium determination in fertilizers is given in! Fertilizers, 1. General.

11. Potassium Hydroxide [195–198]

11.1. Properties

Pure, solid potassium hydroxide [ 1310–58–3 ],KOH, caustic potash, Mr 56.11, r 2.044 g/cm3,mp 410 �C, bp 1327 �C, heat of fusion 7.5 kJ/mol, is a hard, white substance. It is deliquescentand absorbswater vapor and carbon dioxide fromthe air. Potassium hydroxide dissolves readily inalcohols and water (heat of solution 53.51 kJ/mol). The solubility of KOH (g KOH/100 gH2O) in water is shown below:

Temperature, �C 0 10 20 30 50 100

Solubility 97 103 112 126 140 178

The mono-, di-, and tetrahydrates are formedwith water. Aqueous potassium hydroxide is acolorless, strongly basic, soapy, caustic liquid,whose density depends on the concentration:

Concentration, wt% 10 20 30 40 50

Density, g/cm3 1.092 1.188 1.291 1.395 1.514

Technical caustic potash (90–92% KOH)melts at ca. 250 �C; the heat of fusion is ca.6.7 kJ/mol.

11.2. Production

Today, potassium hydroxide is manufacturedalmost exclusively by potassium chloride elec-trolysis. The diaphragm, mercury, and mem-brane processes (! Chlorine) are all suitable forthe production of potassium hydroxide, but themercury process is preferred because it yields achemically pure 50% potassium hydroxidesolution.

In the diaphragm process, a KCl-containing,8–10% potassium hydroxide solution is initiallyformed, whose salt content can be reduced to ca.1.0–1.5% KCl by evaporation to a 50% liquor.Further purification is complicated, and the qual-ity of liquor from mercury cells cannot beachieved.

In the mercury process a very pure KCl brinemust be utilized, because even traces (ppb range)of heavy metals such as chromium, tungsten,molybdenum, and vanadium, as well as smallamounts (ppm range) of calcium or magnesium,lead to strong evolution of hydrogen at the amal-gam cathode. The very pure potassium hydroxidesolution running off the decomposers is cooled,freed from small amounts of mercury in pre-coated filters, and in some cases sent immediatelyto the consumer as a 45–50% liquor in drums,tank cars, or barges.

Since about 1985, new cell rooms for themanufacture of potassium hydroxide solutionhave used the membrane process. At present,the cell liquor has a low chloride content (10–50 ppm); the KOH concentration is 32%. Beforedispatch, it is concentrated to 45–50% byevaporation.

Nonelectrochemical processes have been pro-posed for the manufacture of chlorine and potas-sium hydroxide from KCl by thermal decompo-sition of potassium nitrite in the presence ofFe2O3 [199].

This method involves reacting KCl with NO2

to obtain Cl2 and potassium nitrite, reacting the


KNO2 with iron(III) oxide and oxygen to givepotassium ferrate (K2Fe2O4), and reacting theferrate with water to produce KOH. Anothermethod consists of reacting an aqueous solutionof KCl with NO2 and O2 to give Cl2 and KNO3,which is reacted with water in the presence ofFe2O3 to produce KOH.

Largely water-free, ca. 90–95% potassiumhydroxide (caustic potash) is obtained by evapo-rating potassium hydroxide solution. The resid-ual content of 5–10% H2O is bound as amonohydrate.

Suitable evaporation processes are single- ormultistage falling-film evaporators [200], Bad-ger single-tube evaporators, or boilers connectedin cascade. Heating is carried out with steam orby means of heat-transfer agents (salt melts,Dow-therm). Flash evaporators are used as thefinal stage in large-capacity plants [201].

To counter the strong corrosiveness of thepotassium hydroxide solution and retain the pu-rity of the caustic potash, the equipment is madelargely from high-purity nickel (LC 99.2) or issilverplated. The equipment is often protected bypolarization.

For dispatch, caustic potash comes on themarket poured directly into drums or packed inpolyethylene bags after cooling; in blocks,molded pieces, flakes, prills, and as a powder.Potassium hydroxide is classified as a corrosivematerial:

UN no. 1814 (for aqueous solution)

UN no. 1813 (for dry material)

GGVS/GGVE Class 8

RID/ADR Class 8

Handling is described in [202].


Potassium hydroxide solution is supplied in purequality [total alkalinity 49.7–50.3%,KOH48.8%(min.), NaOH 0.5% (max.), CO2�

3 0.1% (max.)]or in technical quality [total alkalinity 49.7–50.3%, NaOH 1.0% (max.), CO2�

3 0.3% (max.)].The contents of Cl�, SO2�

4 , Fe2þ, and Ca2þ are< 30 ppm. Solid caustic potash produced fromamalgam liquor has a total alkalinity (calculatedas KOH) of 89–92%, NaOH 1.5% (max.), CO2�

30.5% (max.), Cl� 0.01% (max.). The values for

SO2�4 , Fe2þ, and Ni2þ are < 50 ppm. Caustic

potash from diaphragm electrolysis has a Cl�

content of 2.5–3.0% and higher content of heavymetals.

For analysis, see 12.3.


Pure-quality potassium hydroxide is used as araw material for the chemical and pharmaceuti-cal industry, in dye synthesis, for photography asa developer alkali, and as an electrolyte in bat-teries and in the electrolysis of water. Technical-quality KOH is used as a raw material in thedetergent and soap industry; as a startingmaterialfor inorganic and organic potassium compoundsand salts (e.g., potassium carbonate, phosphates,silicate, permanganate, cyanide); for the manu-facture of cosmetics, glass, and textiles; fordesulfurizing crude oil; as a drying agent; andas an absorbent for carbon dioxide and nitrogenoxides from gases.

World production is estimated at ca. 700–800� 103 t/a. Main producers are the UnitedStates [203], Germany, Japan, and France. Otherimportant producer countries are Belgium, theUnited Kingdom, Italy, Spain, South Korea,India, Israel, former Yugoslavia, former Czecho-slovakia, Sweden, and Romania.

12. Potassium Carbonate [195–198]

Potassium carbonate was produced in antiquityand used for many purposes. In the Old Testa-ment, potash is mentioned in Jeremiah (writtenin the 7th century B.C.). ARISTOTLE describesthe extraction of wood ash with water; theRomans manufactured soap from fat and pot-ash. LAVOISIER identified potash as potassiumcarbonate.

The production of potash from wood ash forthe manufacture of glass and soap was a flourish-ing industry in the Middle Ages in areas having aplentiful supply of wood such as Russia, and alsoin Scotland. Since 1860, potash salts have re-placed wood as a raw material for the manufac-ture of potassium carbonate.

In Anglo-American usage, the term potashtoday includes potassium carbonate as well asall potassium salts, such as KCl, K2SO4, and


K2SO4 �MgSO4 � x H2O, that are used as ferti-lizers; the potassium content is given as K2O.

Potassium carbonate occurs in small amountsin a fewAfrican lakes (e.g., Lake Chad and in thevicinity of Lake Victoria), as well as in the DeadSea.

12.1. Properties

Anhydrous potassium carbonate [584-08-7],K2CO3, Mr 138.21, r 2.428 g/cm3, mp 891 �C,is a white, hygroscopic, powdery material thatdeliquesces in moist air. It is readily soluble inwater with the formation of an alkaline solution.The solubility of K2CO3 (g K2CO3/100 g H2O)in water is given below:

Temperature, �C 0 10 20 30 40 50

Solubility 105.5 108.0 110.5 113.7 116.9 121.2

Temperature, �C 60 70 80 90 100

Solubility 126.8 133.1 139.8 147.5 155.7

On addition of acid, potassium carbonatereacts with the evolution of carbon dioxide:

K2CO3þH2SO4!K2SO4þCO2þH2O

K2CO3 forms several hydrates, of whichK2CO3 � 1.5 H2O is the stable phase in contactwith the saturated solution from 0 �C to ca.110 �C. This hydrate (Mr 165.24, r 2.155 g/cm3)crystallizes in glassy, virtually dust-free crystals.It is also hygroscopic and deliquesces in moistair. It is completely dehydrated at 130–160 �C.

12.2. Production

From Caustic Potash and Carbon Di-oxide. Themost important process for the man-ufacture of potassium carbonate begins withelectrolytically produced potassium hydroxidesolution. The almost chemically pure solutionobtained by the mercury process (see 11.2) isreacted with carbon dioxide or CO2-containingoff-gases (flue gas, lime kiln gas).

2 KOHþCO2!K2CO3þH2O

Solid potassium carbonate is then obtained bycrystallization (under vacuum and with cooling)from liquors or in the fluidized-bed process.

In the continuous crystallization process(Fig. 50), the filtered, fresh carbonate solution ismixed with mother liquor and concentrated inseveral preliminary evaporators connected in se-ries until the hydrate K2CO3 � 1.5 H2O finallyprecipitates in the crystallizer after cooling undervacuum [204]. The mother liquor is separatedfrom the crystal suspension in hydrocyclones andcentrifuges, filtered, and fed back to the process.The crystals are dried at ca. 110–120 �C in rotarykilns orfluidized-beddryers andpacked for sale aspotashhydrate, or theyare calcinedat200–350 �Cto give 98–100%K2CO3. Impurities such as soda,sulfate, silicic acid, and iron that concentrate in themother liquors can be partially removed [205] byremoving a partial stream of the mother liquor,which is either used for brine purification in theelectrolysis process or sold as a low-grade potas-sium carbonate solution, or by crystallizing thedouble-salt NaKCO3 at elevated temperature in aseparate crystallization and drying process.

The resulting potassium carbonate is verypure and meets the requirements of USP, BP,DAB, and JP if the process is operated in appro-priate manner.

Starting from potassium carbonate solution,prills can be produced in a combined reactor, inwhich spray drying and fluidized-bed granulationtake place simultaneously [206].

In the fluidized-bed process, aqueous potassi-umhydroxide solution is sprayed into a fluidized-bed reactor from above and exposed to a coun-tercurrent of CO2-containing hot gas (Fig. 51)[207, 208]. Carbonization and calcination takeplace in the same reactor. Hard, spherical potas-sium carbonate prills are formed having a highpacking density. The prills are discharged andsieved. The coarse grains are ground and returnedto the reactor together with the very fine grains,where they act as crystallization seeds. The sal-able, dust-free, medium grains are cooled andpacked. Because no mother liquor is formed, thequality of the potassium carbonate depends onthat of the raw materials. Compared to the crys-tallization process the chlo-ride, soda, and sulfatecontents are usually higher, but the investmentand production costs are lower.

Amine Process. In the Mines de Potassed’Alsace process, potassium chloride is reac-ted under pressure in autoclaves with carbondioxide in precarbonated isopropylamine


solution. Potassium hydrogencarbonate precipi-tates and is filtered off, carefully purified of amineby intensive washing, and dried. It can be con-verted to potassium carbonate by calcination. Freeamine, containing carbon dioxide, is recovered

from the mother liquor by distillation and re-cycled. The chloride, predominantly present inthe mother liquor as amine chlorohydrate, isreacted with hydrated lime to give free amine andan aqueous solution of calcium chloride [209].

Figure 51. Production of potassium carbonate by the fluidized-bed processa) Burner; b) Gas cooler; c) Fluidized-bed reactor; d) Cooler; e) Elevator; f) Screen; g) Mill; h) Silo; i) Cyclone; j) Exhaust gasscrubber

Figure 50. Preparation of potassium carbonate with continuous crystallizationa) Carbonization; b) Crude liquor filter; c) Fresh liquor tank; d)Mixed liquor tank; e1), e2) Preliminary evaporation; f) Vacuum/cooling crystallization (Chemietechnik Messo system); g) Preheater; h) Vapor condenser; i) Vacuum pump; j) Hydrocyclone;k) Centrifuge; l) Centrifuge liquor tank; m) Filter for mother liquor; n) Mother liquor tank; o) Drying or calcining rotary kiln;p) Cooling device for calcined K2CO3; q) Storage for hydrated potash; r) Storage for calcined potash


The use of triethylamine [210], hexamethy-lenimine [211], or piperidine [212] is also pat-ented. All the processes have the disadvantagethat calcium chloride liquor is obtained, whichcan be utilized today only to a small extentand therefore represents an environmentalpollutant.

NephelineDecompositionProcess [213]. Inthe CIS, considerable amounts of potassium car-bonate are formed as a byproduct in the nephelinedecomposition process for aluminum hydroxideproduction. Themineral nepheline is decomposedwith limestone by sintering at 1300 �C:

Alumina, portland cement, soda, and potashare obtained from the product in a complexprocess. The sinter product is leached with anNa2CO3–NaOH solution. After filtration, a filtercake is obtained that is processed to give portlandcement and an aluminate solution containingsilicic acid. After precipitation of the silicic acidas alkaline aluminum silicate the purified alumi-nate solution is reacted with carbon dioxide:

2ðNa;KÞAlO2þCO2þ3 H2O!2 AlðOHÞ3þðNa;KÞ2CO3

The aluminum oxide hydrate is filtered off,and the carbonate solution is concentrated byfractional crystallization in a three-stage processand separated into sodium carbonate andK2CO3 � 1.5 H2O.

Feldspar (KAlSi3O8) and leucite (KAlSi2O6)can also be decomposed analogously and used foralumina, cement, and potassium carbonate man-ufacture [214].

Themagnesia process (Engel–Precht process)is of limited interest:

3ðMgCO3�3 H2OÞþ2 KClþCO2

!2ðMgCO3�KHCO3�4 H2OÞþMgCl2

In hot water the double salt (MgCO3 �KHCO3

� 4 H2O) decomposes under pressure into mag-nesium carbonate and dissolved potassiumcarbonate.

Other Processes. Le Blanc process:

K2SO4þCaCO3þ2 C!CaSþK2CO3þ2 CO2

Formate process:

K2SO4þCaðOHÞ2þ2 CO!2 HCOOKþCaSO4

K2SO4þCaðOHÞ2þ2 CO!2 HCOOKþCaSO4

HCOOKþKOHþ1=2 O2!K2CO3þH2O

‘‘Piesteritz’’ process:

K2SO4þ2 CaCN2þ2 H2O!2 KHCN2þCaSO4þCaðOHÞ2

2 KHCN2þ5 H2O!K2CO3þ4 NH3þCO2

These processes are uneconomical today be-cause of high energy consumption and poorproduct quality, and are no longer used.

In the decomposition of chromium ores withpotassium hydroxide solution, a chromate-con-taining potassium carbonate is obtained as by-product. The production of potassium perman-ganate yields considerable amounts of potassiumcarbonate solution [215].

Organic raw materials, such as sunflowerstalks, molasses, and suint, are used to a smallextent for potash manufacture. They are ashed,leached with water, and processed to potashby fractional crystallization and calcination[216].

Ion-Exchange Process [217, 218]. An aci-dic ion exchanger loaded with ammonium ions ischarged with KCl solution, Kþ being absorbedand an ammonium chloride solution running off.The ion exchanger is then eluted with an excessof ammonium carbonate solution (regenerationof the exchanger). The eluate, a K2CO3–(NH4)2CO3 solution, is separated by thermalcleavage to give ammonia and carbon dioxide.The ammonium chloride is reacted with magne-sium hydroxide to give magnesium chloride andammonia, which is recycled.

12.3. Quality Specifications andAnalysis

Depending on the intended use, potassiumcarbonate is offered in varying commercialforms and degrees of purity: as granules, aspowder, and as potassium carbonate hydrate


(K2CO3 � 1.5 H2O). The material manufacturedfrom mercury-process potassium hydroxidesolution is of high purity, particularly withrespect to chloride content. In the amine pro-cess, the chloride content is higher and thesodium carbonate content lower, while nephe-line decomposition gives high sodium carbon-ate contents and relatively high sulfate contents(Table 11).

Analysis. The total alkalinity includesK2CO3 þ KOH þ Na2CO3; it is determinedwith 0.5 N H2SO4 by potentiometric titrationor with a methyl orange indicator (change tobrown-red). Sodium is determined by flamephotometry. The chloride content is determinedby turbidity measurement after addition ofAgNO3. The sulfate content is determined byion chromatography or gravimetrically afterprecipitation as barium sulfate. The metalcontent is determined by atomic absorptionspectroscopy or photometrically by complexformation (Fe2þ as sulfosalicylate, Si4þ as themolybdato complex, Cu2þ as pyrrolidinodithio-carbamate, and Ni2þ as the diacetylglyoximecomplex). Test methods for photographic-grade potassium carbonate, anhydrous aredescribed in ISO 3623–1976 (E).

12.4. Storage and Transportation

Potassium carbonate is stored in bunkers; theventilation air must be dry because of the hygro-scopicity of the product.

Silo vehicles and bulk containers are used fordispatch to bulk customers. Smaller amounts arepacked in polyethylene valve sacks of 25–50 kg.The material is not hazardous; for pharmaceuti-

cal use it is classified as GRAS (generally recog-nized as safe) by FDA [219].


The glass industry is the most important consum-er of K2CO3. Large amounts are also required forpotassium silicate manufacture.

Potassium carbonate is used for many organicsyntheses. Numerous inorganic and organic po-tassium salts are manufactured from potassiumcarbonate (potassium phosphate, bromide, io-dide, dichromate, cyanide, and ferrocyanide); inaddition it is a starting material for drying, neu-tralization, and condensation agents. As a regen-erable absorbent for carbon dioxide, hydrogensulfide, and sulfur oxides, it is attaining impor-tance in environmental protection. Potassiumcarbonate is used as a fertilizer for acidic soil.

Other users are the electrical industry, the dyeindustry, the printing trade, the textile industry,the leather goods industry, and the ceramic in-dustry. Soft soap manufacture has lost its earlierimportance as a customer. Potash solutions areused as fire retardants and as cooling brines(freezing point � 36 �C at 576 g/L¼40.5 wt%K2CO3).

The food industry uses potassium carbonate asa leavening agent in baked goods, as a debit-terizing agent for cocoa beans, and as an additivefor drying raisins. Potash in DAB quality isfrequently used in the pharmaceutical industryas a raw material and auxiliary.

The most important producer countries forpotassium carbonate are the CIS, France,Germany, the United States, and Japan. Otherproducers are Israel, Spain, India, South Korea,Belgium, Italy, former Yugoslavia, and China.

Table 11. Analyses of calcined potassium carbonate of varying origin (data in %, remainder H2O)

From potassium hydroxide solution

Mercury process Diaphragm

process

Amine

process

Nepheline

decomposition

Crystallization Fluidized bed

K2CO3 98.0–99.8 98.5–99.5 97–99 99 97.5–98.5

Na2CO3 0.1–0.5 0.1–0.5 0.5–1.0 0.01 0.1–1.0

Cl 0.001–0.002 0.004–0.013 0.2–1.0 0.19 0.01–0.03

SO4 0.003–0.005 0.005–0.013 0.005–0.010 0.25–0.60

Si þ Ca 0.005 0.006 0.010 0.009

Fe 0.0003–0.0005 0.0002–0.0006 < 0.0005 < 0.0024 0.0007–0.0021


13. Potassium Hydrogencarbonate[195–197]

13.1. Properties and Production

Potassium hydrogencarbonate [298-14-6],KHCO3, Mr 100.12, r 2.17 g/cm3, is a white,crystalline powder that is sparingly soluble inwater and insoluble in alcohol. When heatedabove 120 �C, it decomposes into potassiumcarbonate, water, and carbon dioxide.

It is manufactured industrially by passingcarbon dioxide into concentrated potassium car-bonate solutions or exposing these to a counter-current of purified, cold flue gas in trickle towers(overcarbonization). Because of its low watersolubility (22.4 g of KHCO3 in 100 mL ofH2O at 20 �C) it precipitates in crystalline form,is separated by centrifugation, and dried at ca.110 �C. In some potassium carbonate productionprocesses, potassium hydrogencarbonate is ob-tained as a precursor (see 12.2).

The total alkalinity of the industrial material,calculated as KHCO3, is at least 98– of high-purity potassium carbonate and other pure potas-sium salts. Producing countries are the UnitedStates, Germany, and France.

References

General References1 Fertilizer Manual, International Fertilizer Development

Center, Muscle Shoals, Ala., 1979, pp. 225–247.

2 Kirk-Othmer, 3rd ed., 18, 920–950.

3 Winnacker-K€uchler, 4th ed., 2, 268–333.

4 The British Sulphur Corporation, World Survey of Pot-

ash Resources, London 1985.

5 V. A. Zandon in N. L. Weiss (ed.): SME Mineral

Processing Handbook, Society of Mining Engineers,

New York 1985, section 22.

6 R. M. McKercher (ed.): ‘‘Potash Technology,’’ 1st In-

ternational Potash Technology Conference, Saskatche-

wan, Canada, Oct. 3–5 1983, Pergamon Press, Toronto

1983.

7 W.H. Eatock: ‘‘PotashRefining in Saskatchewan,’’Min.

Eng. (Littleton, Colo.) 34, (1982) no. 9, 1350 –1353.

8 M. P. Kurtanjek: ‘‘Mining and Beneficiating Potash, Re-

cent Developments and Trends outside North America,’’

Phosphorus Potassium 128 (1983) Nov./Dec., 26–31.

9 Gmelin, System no. 22.

10 F. Serowy: Verarbeitungsmethoden der Kalirohsalze,

W. Knapp, Halle/Saale 1952.

11 H. Schubert: Aufbereitung fester mineralischer Roh-

stoffe, VEB Deutscher Verlag f€ur Grundstoffindustrie,

Leipzige, vol. I, 1968; vol. II, 1986; vol. III, 1986.

12 D. Fulda et al.: Kali, das bunte, bittere Salz, VEB

Deutscher Verlag f€ur Grundstoffindustrie, Leipzig 1990.

13 A. Heinz, R. v. d. Osten (eds.): ABC Kali und Steinsalz,

VEB Deutscher Verlag f€ur Grundstoffindustrie, Leipzig1982.

14 I. Barin: Thermochemical Data of Pure Substances,

VCH, Weinheim, Germany 1989.

15 M. Broul, J. N�yvlt, O. S€ohnel: Solubility in Inorganic

Two-component Systems, Elsevier, Amsterdam 1981.

16 Ullmann, 4th ed., 13, 447–496.

Specific References17 R. Slotta: ‘‘Die Kali- und Steinsalzindustrie,’’ Tech-

nische Denkm€aler in der Bundesrepublik Deutschland,

vol. 3, Deutsches Bergbaumuseum, Bochum 1980.

18 D. Hoffmann: Elf Jahrzehnte Deutscher Kalisalzberg-

bau, Verlag Gl€uckauf, Essen 1972.

19 Kaliverein e.V.:DieKaliindustrie in derBundesrepublik

Deutschland, 6th ed., 1988.

20 PhosphorusPotassium 166 (1990)March/April, 17 –19.

21 R. Weissenberger: Chronique des mines de potasse

d’Alsace, Ziegler, Bergholtz 1985.

22 Phosphorus Potassium 168 (1990) July/Aug., 12–13.

23 J. W. Turrentine: Potash in North America, Reinhold

Publ., New York 1943.

24 Phosphorus Potassium 68 (1973) Nov./Dec., 39–43.

25 O. Braitsch: Entstehung und Stoffbestand der Salzla-

gerst€atten, Springer Verlag, Berlin 1962.

26 Ullmann, 4th ed., 13, 450.

27 M. A. Zharkov: Paleozoic Salt Bearing Formations of

the World, Springer Verlag, Berlin 1984.

28 E. A. Vysotsky, V. Z. Kislik: ‘‘Epochs of Bischofite

Deposition in Geological History,’’ Int. Geol. Rev. 29

(1987) no. 2, 134–139.

29 R. K€uhn, Kali Steinsalz 1 (1955) no. 9, 3–16.

30 R. K€uhn, Geol. Jahrb. 90 (1972) 127–220.

31 H. Borchert: Ozeane Salzlagerst€atten, Borntr€ager, Ber-

lin 1959.

32 F. Lotze: ‘‘Steinsalz und Kalisalze,’’ Die wichtigsten

Lagerst€atten der ‘‘Nicht-Erze’’, vol. III, part 1,

Borntr€ager, Berlin 1938.

33 F. Lotze: Steinsalz und Kalisalze, 2nd ed., part 1,

Borntr€ager, Berlin 1957.

34 F. Lotze: Die Salzlagerst€atten in Zeit und Raum, Ar-

beitsgem. Forsch. Landes Nordrhein-Westfalen,

no. 195, Westdeutscher Verlag, K€oln-Opladen 1969.

35 G. Richter-Bernburg: ‘‘Salzlagerst€atten,’’ in Bentz-Mar-

tini (ed.): Lehrbuch der Angewandten Geologie, vol. 2,

part 1, Enke Verlag, Stuttgart 1968, pp. 916 –1061.

36 G. Richter-Bernburg, Z. Dtsch. Geol. Ges. 105 (1953)

593–645.

37 M. Brongersma, Mar. Geol. 11 (1972) 123–144.

38 C. Kippenberger et al.: Untersuchungen €uber Angebot

und Nachfrage mineralischer Rohstoffe, XX Kali, BGR

Hannover-DIW Berlin, Schweizerbart, Stuttgart 1986.


39 J. D’Ans, Naturwissenschaften 34 (1947) 295–301.

40 J. D’Ans, R. K€uhn, Kali Steinsalz 3 (1960) 69–84.

41 H. Mayrhofer: ‘‘World Reserves of Mineable Potash

Salts Based on Structural Analysis,’’ Proceedings of the

6th International Symposium on Salt, vol. 1, The Salt

Institute, Alexandria, USA, 1983.

42 P. A. Ziegler, Geological Atlas of Western and Central

Europe, Elsevier, Amsterdam 1982.

43 E. Messer, Kali Steinsalz 5 (1970) 244–251.

44 Pittsburgh Plate Glass Co., US 3 058 729, 1962 (J. B.

Dahms, B. P. Edmonds); CA 627 308, 1963 (J. B.

Dahms); US 4 007 964, 1977 (E. L. Goldsmith); US

3 262 741, 1966 (B. P. Edmonds, J. B. Dahms); US

3 433 530, 1969 (J. B. Dahms, B. P. Edmonds); US

4 329 287, 1980 (E. L. Goldsmith).


46 D. Jackson, Eng. Min. J. 174 (1973) no. 7, 59–68.


48 G. Duchrow, I. Fitz, N. Gr€uschow, Phosphorus Potassi-

um 167 (1990) May/June, 26–32.

49 H. Schubert: Kali, das bunte, bittere Salz, 1 VEB

Deutscher Verlag f€ur Grundstoffindustrie, Leipzig

1990 72–84.

50 R. K€uhn, Kali Steinsalz 5 (1970) 307–317.

51 R. K€uhn,Erzmetall 8 (1955) Suppl. B 93–B 107, B 115.

52 H. Autenrieth, O. Braun, W. Otto: Winnacker-K€uchler,

4th ed., 2, 281–283.

53 V. A. Zandon in N. L. Weiss (ed.): SME Mineral

Processing Handbook, Society of Mining Engineers,

New York 1985, 22–4–22–5.

54 J. G€otte, Kali Steinsalz 10 (1990) 261–264.

55 J. H. van’t Hoff: Untersuchungen €uber die Bildungs-

verh€altnisse der ozeanischen Salzablagerungen, Akad.

Verlagsges., Leipzig 1912.

56 H. D’Ans: Die L€osungsgleichgewichte der Systeme der

Salze ozeanischer Salzablagerungen, Verlagsges. f€urAckerbau, Berlin 1933.

57 J. D’Ans, Z. Elektrochem. 56 (1952) 497–505.

58 Gmelin, System no. 22.

59 A. B. Sdanowsky, E. I. Lyakhowskaya, R. E. Schley-

mowitch: Handbuch der L€oslichkeit der Salzsysteme,

Gaskhimisdat, Leningrad, vol. I 1953; vol. II, 1954;

vol. III, 1961; vol. IV, 1964 (Russ.).

60 H. Autenrieth, Kali Steinsalz 1 (1953) no. 2, 3–17.



63 H. Autenrieth, Kali Steinsalz 2 (1958) no. 6, 181 –

200.


65 H. Autenrieth, Rev. chim. min�er. 7 (1970) 217–229.


67 G. Peuschel, Kali Steinsalz 9 (1986) no. 9, 296–303.

68 W. P. Wilson, A. G. McKee: Proceedings of the 4th

Symposium on Salt, vol. I, The Northern Ohio Geologi-

cal Society, Cleveland, Ohio, 1974, 517–525.

69 W. P. Wilson: Proceedings of the 3rd Symposium on

Salt, vol. II, The Northern Ohio Geological Society,

Cleveland, Ohio, 1969, 20–29.

70 J. H. Wolf: R. M. McKercher (ed.): ‘‘Potash Technolo-

gy,’’ 1st International Potash Technology Conference,

Saskatchewan, Canada, Oct. 3–5, 1983, Pergamon

Press, Toronto 1983, 711–716.

71 H. Domning, Kali Steinsalz 7 (1977) no. 4, 155–160.

72 W. H. Coghill, J. R. De Vaney, J. B. Clemmer, S. R. B.

Cooke, Report of Investigations of the US-Bureau of

Mines, Report no. 3271, 1935.

73 A. S. Kusin, Kalii 6 (1937) 17–27.

74 R. A. Pierce, L. D. Anderson, Eng. Min. J. 143 (1942)

38–41.

75 Kreller,Kali, Verw. SalzeErd€ol 33 (1939) 35–37, 46–47,

53–57.

76 E. R€usberg, Chem.-Ing.-Tech. 27 (1955) 1–4.

77 O. Karsten, in W. Gr€under: Erzaufbereitungsanlagen in

Westdeutschland, Springer Verlag, Berlin 1955, 343–

345.

78 V. A. Arsentiev, J. Leja, CIM Bull. 3 (1977) 154–158.

79 Du Pont, US 2 088 325, 1937 (J. E. Kirby).

80 C. M. Aleksandrovic, Freiberg. Forschungsh. A 544

(1975) 73–81.

81 H. Schubert, Aufbereit. Tech. 7 (1967) 365–368.

82 A. Singewald, Chem.-Ing.-Tech. 33 (1961) 376–393,

558–572, 676–688.

83 D.W. u. M. C. Fuerstenau,Min. Eng. (Littleton, Colo.) 8

(1956) 302–307.

84 A. F. Taggart, Elements of Ore Dressing, John Wiley &

Sons, New York 1951.

85 A. M. Goudin, Testimony in Transcript of Evidence,

Civil Action no. 1829, District of New Mexico,

p. 1255.

86 R. Bachmann,Erzmetall 8 (1955) Suppl. B 109 –B 116.

87 J. Rogers, J. H. Schulman: Second International Con-

gress of Surface Activity, vol. II, Reprints, Butterworths,

London 1957, 330–338.

88 A. Singewald, Erzmetall 12 (1959) 121–135.

89 O. J. Somojlov: Struktur von w€aßrigen Elektrolytl€osun-

gen, B. G. Teubner, Leipzig 1961.

90 H. Schubert, Aufbereit. Tech. 6 (1966) 305–313.

91 F. Hagedorn, Kali Steinsalz 10 (1991) 315–328.

92 D. Uhlig, Neue Bergbautech. 5 (1975) 145–155.

93 H. K€ohler et al., Neue Bergbautech. 16 (1986) 45–50.

94 N. F. Mescerjakov, Y. W. Rjabov, V. N. Kuznetzov,

Freiberg. Forschungsh. A 594 (1978) 33–54.

95 A. Bahr, K. Legner, H. L€udke, F. W. Mehrhoff, Aufber-

eit. Tech. 1 (1957) 1–9.

96 N. F.Mescerjakov:Flotacionnyemaschiny, Isdatel’stwo

Nedra, Moskau 1972.

97 H. Schubert, Neue Bergbautech. 4 (1974) 223–228.

98 Phosphorus Potassium 145 (1986) 29–33.

99 Kali & Salz, DE 3 435 124, 1987 (F. Hagedorn, G.

Peuschel, A. Singewald).


101 H. K€ohler, W. Kramer, Neue Bergbautech. 11 (1981)

362–366.

102 R. B. Tippin, Chem. Eng. (N.Y.) 184 (1977) no. 15, part

1, 73–75.

103 VEB Kali, DD 220 237, (L. Herrmann et al.).


104 S. Mildner, R. Ecke, DD 35 637, 1965.

105 I.M.LeBaron,W.C.Knopf,Min. Eng. (Littleton, Colo.)

10 (1958) 1081–1083.

106 H. Autenrieth, Kali Steinsalz 5 (1969) 171–177.

107 Kali-Forschungsanstalt, DE 1 056 551, 1957 (H.

Autenrieth).

108 G. Fricke, Kali Steinsalz 9 (1986) 287–295.

109 D. Larmour in: R. M. McKercher (ed.): ‘‘Potash Tech-

nology,’’ 1st International Potash Technology Confer-

ence, Saskatchewan, Canada, Oct. 3–5, 1983, Pergamon

Press, Toronto 1983 597–602.

110 R. Bock, Chem.-Ing.-Tech. 53 (1981) 916–924.

111 Kali & Salz, DE 1 249 783, 1966 (A. Singewald, G.

Fricke).

112 A. Singewald, U. Neitzel: R. M. McKercher (ed.):

‘‘Potash Technology,’’ 1st International Potash Tech-

nology Conference, Saskatchewan, Canada, Oct. 3–5,

1983 Pergamon Press, Toronto 1983, 589–595.

113 L. Ernst, Kali Steinsalz 9 (1986) 275–286.

114 L. Ernst, Ber. Bunsenges. Phys. Chem. 93 (1989) 857–

863.

115 L. Ernst, Ber. Bunsenges. Phys. Chem. 94 (1990) 1435–

1439.

116 Kali & Salz, DE 3 603 166, 1986 (G. Fricke, I. Giesler,

R. Diekmann).

117 Kali & Salz, DE 3 603 165, 1986 (H. Balzer, H. Bur-

ghardt, F. Maikranz).

118 Kali&Salz,DE 3 603 167, 1986 (U.Neitzel, G. Fricke).

119 Kali & Salz, DE 1 077 611, 1959 (H. Autenrieth, G.

Peuschel).

120 Kali & Salz, DE 1 142 802, 1961 (H. Autenrieth, G.

Peuschel, G. Weichart).


Fricke, D. Jung).


Fricke, D. Jung).

123 Kali & Salz, DE 1 154 052, 1960 (H. Autenrieth, H.

Dust).

124 Kali & Salz, DE 1 076 593, 1957 (H. Autenrieth).


Fricke).

126 Kali & Salz, DE 1 667 814, 1968 (G. Fricke, A.

Singewald).

127 Kali & Salz, DE 1 283 772, 1967 (H. Autenrieth, H.

Wirries).


Fricke).


Fricke).

130 W. B. Dancy: Kirk-Othmer, 3rd ed., 18, 931–933.

131 T. E. Burus, B. J. Clarke, W. B. Coome, A. H. New-

combe: R. M. McKercher (ed.): ‘‘Potash Technology,’’

1st International Potash Technology Conference, Sas-

katchewan, Canada, Oct. 3–5, 1983, Pergamon Press,

Toronto 1983 541–546.

132 R.Diekmann: Lecture held at Int. PotashTechnol. Conf.,

2nd, Hamburg, May 26–29, 1991.

133 E. Weps, Kali Steinsalz 8 (1981) 177–183.

134 Kali&Salz,DE3 434 190, 1984 (O. Pfoh,C. Radick,H.

Thenert).


136 T. Fleischer, Kali Steinsalz 9 (1986) 304–313.

137 H. J. Scharf: ‘‘Environmental Aspects of K-Fertilizers in

Production,Handling andApplication,’’Development of

K-Fertilizer Recommendations, International Potash In-

stitute, Worblaufen-Bern 1990, 395–402.

138 PhosphorusPotassium 148 (1987)March/April, 30 –35.

139 M. D. Haug, K. W. Reid: Lecture held at Int. Potash

Technol. Conf., 2nd, Hamburg, May 26–29, 1991.

140 J. E. Tallin, D. E. Pufahl: R. M. McKercher (ed.):

‘‘Potash Technology,’’ 1st International Potash Tech-

nology Conference, Saskatchewan, Canada, Oct. 3–5,

1983, Pergamon Press, Toronto 1983 755–760.

141 K. W. Reid, G. A. Maki: Lecture held at Int. Potash

Technol. Conf., 2nd., Hamburg, May 26–29, 1991.

142 H. E. Schroth, Phosphorus Potassium 67 (1973) Sept./

Oct., 38.

143 A. Singewald, Die Weser 57 (1983) no. 516, 1–8.

144 N. Kn€opfel: Lecture held at Int. Potash Technol. Conf.,

2nd., Hamburg, May 26–29, 1991.

145 H. Stahl, Aufbereit. Tech. 20 (1980) 525–533.

146 W. B. Pietsch: R. M. McKercher (ed.): ‘‘Potash Tech-



Press, Toronto 1983 , 661–669 .

147 L. Medemblik: R. M. McKercher (ed.): ‘‘Potash Tech-



Press, Toronto 1983, 653–659.

148 A. S. Middleton, D. A. Cormode, J. E. Scotten: R. M.

McKercher (ed.): ‘‘Potash Technology,’’ 1st Interna-

tional Potash Technology Conference, Saskatchewan,

Canada, Oct. 3–5, 1983, Pergamon Press, Toronto 1983,

647–651.

149 Phosphorus Potassium 173 (1991) May/June 28–36.

150 K. Kahle, G. Leib: Lecture held at Int. Potash Technol.

Conf., 2nd., Hamburg, May 26–29, 1991.

151 L. I. Skvirski, A. A. Chityakov, Z. L. Kozel: Lecture

held at Int. Potash Technol. Conf., 2nd., Hamburg, May

26–29, 1991.

152 H. Rieschel, K. Zech,Phosphorus Potassium 115 (1981)

Sept./Oct., 33–39.

153 H. Rug, K. Kahle, Phosphorus Potassium 170 (1990)

Nov./Dec., 23–27.

154 International Fertilizer Industry Association (IFA): Pot-

ash Statistics 1989, Paris.

155 Prognose-Arbeitsgruppe Weltbank/FAO/Unido, 1991,

unpublished.

156 C. Childers, Phosphorus Potassium 169 (1990) Sept./

Oct., 16–20.

157 O. Walterspiel, Kali Steinsalz 10 (1989) 168–174.

158 Phosphorus Potassium 165 (1990) Jan./Febr., 18 –19.

159 Gmelin, System no. 22, Suppl. vol., pp. 1280–1338.

160 J. N€ather, H. H. Emons, Bergakademie 21 (1969) 310–

313.



162 Phosphorus Potassium 122 (1982) Nov./Dec., 36–39.

163 N. P. Finkelstein, S. H. Garnett, L. Kogan: R. M.

McKercher (ed.): ‘‘PotashTechnology,’’ 1st International

Potash Technology Conference, Saskatchewan, Canada,

Oct. 3–5, 1983, PergamonPress, Toronto1983,571 –576.

164 U. Neitzel, Kali Steinsalz 9 (1986) 257–261.

165 H. Autenrieth, O. Braun, W. Otto: Winnacker-K€uchler,4th ed., 2, 320–322.

166 H. Scherzberg, G. D€oring: Lecture held at Int. Potash

Technol. Conf., 2nd., Hamburg, May 26–29, 1991.

167 Kali & Salz, DE 3 418 147, 1984 (E. Menche, H. G.

Diehl, H. Eberle).

168 W. B. Dancy: Kirk-Othmer, 3rd ed., 18, 945.

169 D. K. Storer: R. M. McKercher (ed.): ‘‘Potash Technol-

ogy,’’ 1st International Potash Technology Conference,

Saskatchewan, Canada, Oct. 3–5, 1983, Pergamon

Press, Toronto, 1983, 577–582.

170 R. Phinney, EP 0 199 104, 1986.

171 Kali & Salz, DE 3 607 641, 1986 (S. Vajna, G.

Peuschel).

172 Soci�et�e d’Etudes Chimiques pour L’Industrie et

l’Agriculture (SECPIA), DE 956 304, 1954 (J. Lafont).

173 J. A. Fernandez Lozano, A. Wint, Chem. Eng. J. (Lau-

sanne) 23 (1982) 53–61.

174 Phosphorus Potassium 167 (1990) May/June, 11.

175 Superfos A/S, Vedbaek, DK, DE 3 331 416, 1983 (K. C.

B. Knudsen); US 4 504 458, 1983 (K. C. B. Knudsen).

176 Chem. Eng. (N.Y.) 81 (1974) 98–99.

177 Kali & Salz, DE 2 810 640, 1978 (N. Kn€opfel, F.

Wartenpfuhl, A. Hollstein).

178 A. Hollstein, Kali Steinsalz 7 (1979) 498–500.

179 Phosphorus Potassium 141 (1986) Jan./Feb., 17 – 21.

180 Phosphorus Potassium 151 (1987) Sept./Oct., 16–21.

181 G. Kemmler, Kali Steinsalz 9 (1985) 167–169.


ash Resources, London 1985 pp. 62–64.

183 Phosphorus Potassium 24 (1966) 40–44.

184 A. M. Amarin, K. Manasrah, Proc. IFA-NFC Joint

Middle East-South Asia Fertilizer Conference, Lahore,

Pakistan Dec. 3–6, 1988.

185 Kirk-Othmer, 2nd ed., Suppl. vol., 438–467.


ash Resources, London 1985. pp. 38–39.

187 U. Neitzel, Kali Steinsalz 5 (1971) 327–334.

188 P. Behrens, Industrial Processing of Great Salt Lake

Brines, Utah Geological and Mineral Survey Bulletin

116, 1980, 223–228.

189 Phosphorus Potassium 132 (1984) July/Aug., 6.


ash Resources, London 1985, p. 42–43.


ash Resources, London 1985, p. 70.

192 Phosphorus Potassium 173 (1991) May/June, 26–27.

193 W. Horwitz (ed.): Official Methods of Analysis of the

AOAC, 11th ed., Association of Official Analytical

Chemists (AOAC), Washington D.C. 1970.

194 VerbandDeutscherLandwirtschaftlicherUntersuchungs-

und Forschungsanstalten (LUFA): ‘‘Die Untersuchung

von D€ungemitteln,’’ Methodenbuch, vol. II, Verlag J.

Neumann-Neudamm, Melsungen 1972, method 4.1.

General References195 Kirk-Othmer, 3rd ed., 18, 936–939.

196 J. Ford: ‘‘Caustic potash,’’ Encycl. Chem. Process. Des.

7 (1978) 22–34.

197 Ullmann, 4th ed., 13, 489–496.

198 H€uls, Handbook KOH-, K2CO3-, KHCO3-Products,

Marl 1992.

Specific References199 N. Takeuchi, Soda to Enso 39 (1988) no. 461, 277 –290.

200 Bertrams, Concentration Plants for NaOH-, KOH-,

Na2S- and CaCl2-liquors, Muttenz, Switzerland, 1979.

201 Sulzer-Escher-Wyss, US 4 927 494, 1990 (R. Winkler

et al.).

202 Oxy-Occidental Chem. Corp., Caustic Potash Hand-

book, Irving 1987.

203 Oil Paint Drugs, Chemical Marketing Reporter, 28th

May, 1990.

204 Messo Chemietechnik, Brochure, Mass Crystallization,

Duisburg, Germany, 1990.

205 Mannesmann, DE 3 816 061, 1989 (R. Schmitz).

206 VEB Kombinat Kali, DD 255 328 A, 1986 (K. Will, G.

Elberling).

207 Bertrams, Fluid Bed Processes, Muttenz, Switzerland,

1979.

208 Diamond Shamrock Corp., Company brochure, Cleve-

land, Ohio, 1969.

209 Inf. Chim. 99 (1971) Aug./Sept., 125.

210 Kali-Chemie, DT 1 220 401, 1962 (P. Schmid).

211 J. N. Shokin et al., Khim. Promst. (Moscow) 9 (1978)

685.

212 FMC Corp., BE 616 193, 1962 (A. B. Gency, M. J.

McCarthy).

213 D. M. Ginzburg, A. A. Tripolskii, Tr. Gos. Nauchno-

Issled. Proektn. Inst. Osnon. Khim. 30 (1973) 26.

214 IMC Corp., US 3 073 443, 1960 (R. E. Snow).

215 A. Schmidt, Angewandte Elektrochemie, Verlag Che-

mie, Weinheim, Germany 1976, p. 183.

216 Lemar Developments, AU 563 487, 1987 (B. W. Levy).

217 Chem. Age Int., 29th Sept., 1972.

218 DynamitNobel,DT1 812 769, 1968 (D.Labriola et al.).

219 FDA, Fed. Regist. 48 (1983) no. 224, 52 440–3.

Further Reading

M. B. Freilich, R. L. Petersen: Potassium Compounds, ‘‘Kirk

Othmer Encyclopedia of Chemical Technology’’, 5th

edition, John Wiley & Sons, Hoboken, NJ, online DOI:

10.1002/0471238961.1615200106180509.a01.pub2.

R. B. King (ed.): Encyclopedia of Inorganic Chemistry, 2nd

ed., Wiley, Chichester 2005.

G. Kreysa, M. Sch€utze (eds.): Corrosion Handbook, 2nd ed.,DECHEMA/Wiley-VCH, Weinheim 2007.


ullmann - potassium compounds

Documents