R I 1916 T Nor REMOVE FRCl1 r~·ReS~:r~~ n~:nter . ~ E. 315 Montgomery Ave. Spokane, WA 99207 Bureau of Mines Report of Investigations/1988 LI BRARY

Na2Cr04 From Domestic Chromite Concentrates by an Alkali-Fusion Method

By Gary L. Hundley, R. E. Mussier, R. A. Holmes, and R. S. Olsen

UNITED STATES DEPARTMENT OF THE INTERIOR ..

Report of Investigations 9167

Na2Cr04 From Domestic Chromite Concentrates by an Alkali-Fusion Method

By Gary L. Hundley, R. E. Mussier, R. A. Holmes, and R. S. Olsen

UNITED STATES DEPARTMENT OF THE INTERIOR Donald Paul Hodel, Secretary

BUREAU OF MINES T S Ary, Director Library of Congress Cataloging in Publication Data:

Na2Cr04 from domestic chromite concentrates.

(Bureau of Mines report of investigations : 9161) Bibliography: p. 12. Supt. of Does. no.: I 28.23:9167. 1. Chl'omium--Metallurgy. 2. Sodium chromate. 3. Chromite. 4. Sodium hydroxide. 5. Fused salts. 6. Leaching. 1. Hundley, Gary L. II. Title: Alkali-fusion method. III. Series: Report of investigations (United States. Bureau of Mines) ; 9167.

TN23.U43 [TN799.C5] 62 s [669'.734] 88-600002 gas

CONTENTS

Abstract ...... •...•..••...... •..•...... •..... 1 Introduction ••••••••••••••••••••••••••••• · ...... 2 Raw materials ..••••.••••.•.••.•.•...•••.. 4 Countercurrent leaching •••••••••••••••••• ...... 5 Equipment and procedures ••••••••••••••• 5 Results and discussion ••••••••••••••••••• 6 Solution purification •••••••••••••••••••• 7 Crystallization •••••••••••••••••••••••••• ·...... 9 Equipment and procedures ••••••••••••••••• 9 Results and discussion •••••••••••••••• 10 Synthetic leach solutions ••••••••• · ...... , .. . 10 Actual leach solution ••••••••••••• 11 Summary and conclusions •••••••••••••••••••• . , ...... 11 Ref erences •...... ••.•...•...••...•.•.•.• 12

ILLUSTRATIONS

1. Chemical processing of chromite •••••••• 3 2. Countercurrent leach flowsheet ••••••••• ...... 5

3. Vacuum evaporative crystallizer •••••••• • •••••••••••••••••••••••••• CI •• 0 ••••• 9

TABLES

1. Head analyses of chromite concentrates ••••••••••••••••••••••• ...... 4 2. Results of two-s tep, three-stage countercurrent leach tests •• ...... " 6 3. Results of solution purification tes t s ...... ~ . • •••• (It 8 4. Results of vacuum crystallization tests ...... , ..... 10 UNIT OF MEASURE ABBREVIATIONS USED IN THIS REPORT

°c degree Celsius kg kilogram cm 3/g cubic centimeter per gram L liter g gram min minute gil gram per liter mL milliliter h hour mL/min milliliter per minute in inch pct percent in Hg inch of mercury ppm part per million (atmospheric pressure) » ZE£2L!!

Na2Cr04 FROM DOMESTIC CHROMITE CONCENTRATES BY AN ALKALI"FUSION METHOD

By Gary L Hundley,1 R. E. Mussler,2 R. A. Holmes,3 and R. S. Olsen4

ABSTRACT

The Bureau of Mines has devised a procedure to recover chromium chemi­ cals from concentrates derived from low-grade domestic chromites. These domestic chromites contain silicon and aluminum impurities at levels that are too high to permit processing by present industrial processes. The Bureau procedure consists of reacting chromite with molten NaOH under oxidizing conditions to form sodium chromate (Na2Cr04)' The reac­ tion product is leached with methanol to recover the majority of the unreacted NaOH, then with water to extract the Na2Cr04 and the remainder of the NaOH. The Na2Cr04 product is recovered by evaporative crystal­ lization from the aqueous solution. This report presents laboratory results of studies to determine the optimum leaching conditions for the reaction products of several domestic chromite concentrates. Results of solution purification and crystallization studies are also presented. The concentrates are from a variety of sources including the Stillwater Complex in Montana, Red Bluff Bay and the Kenai Peninsula in Alaska, and a Ni-Co laterite from southern Oregon. Sodium chromate crystals were produced that contained less than 0.01 pct NaOH.

Chemical engineer. 2Research chemist. 3Chemist. 4Supervisory chemical engineer. Albany Research Center, Bureau of Mines, Albany, OR. 2

INTRODUCTION

Chromium is a commodity that is essen­ Na2Cr04 in a methanol solution containing tial to the Nation's metallurgical, chem­ 10 pct NaOH is only 0.013 pct (4). The ical, and refractory industries. The methanol solution is then evaporated to United States has no domestic production recover the NaOH, which is recycled to or economic reserves of chromite, the the fusion reactor. only commercial ore of chromium, and must The residue from the methanol leach is rely on imports to meet national needs. water leached to recover the remainder of The chemical industry consumes approxi­ the NaOH and the Na2Cr04. Silicon and mately 25 pct of the chromite for the aluminum compounds are removed from the production of pigments, chromic acid for aqueous solution by adding a soluble electro plating, and other chemicals used silica compound to adjust the Si-Al ratio for leather tanning, wood preservatives, so the compound, NaAlSi04, will form and catalysts, and corrosion inhibitors precipitate from solution. The final (1 ).5 Na2Cr04 product is recovered from the -Commercial processes presently used to aqueous solution by evaporative crystal­ chemically treat chromite concentrates lization. The mother liquor exiting the include an oxidizing roast of the chro­ crystallizer is evaporated, and the mite with Na2C03 and CaO in a rotary kiln solids are recycled to the fusion reac­ at temperatures around 1,100 0 to tor. The Na2Cr04 product from this pro­ 1,150 0 C. The amounts of reagents and a cedure is a basic industrial chemical and diluent are controlled so that the reac­ can be used to produce the other common tion mixture remains as a solid phase in Cr compounds in commercial use. the kiln (2). Concentrates produced from This report describes experimental domestic ~hromite deposits contain too results for the leaching, solution puri­ much silicon to be processed by this fication, and crystallization steps of method. The silicon forms molten, sticky the procedure. A previous report (5) reaction products with the Na2C03, which describes the experimental results of the cause balls or rings of material to form fusion step for a variety of concen­ in the kiln, hindering its operation. trates. The general equation for the The aluminum content in domestic re­ fusion reaction of the chromite with the sources also is high, resulting in an NaOH is excess consumption of reagents (3). A simplified flowsheet for the Bureau of Mines procedure to recover chromium chemicals from domestic chromite concen­ + 2Na2Cr04 + 1/2 Fe203 + 2H20. (1) trates is shown in figure 1. Briefly, the procedure consists of reacting the Previous work by Chandra (~) showed chromite at 550 0 to 650 0 C with an excess that an excess of NaOH must be used to of molten NaOH under oxidizing conditions maintain a fluid reaction mixture. A to produce Na2Cr04" The product from the NaOH-to-chromite weight ratio of ~4:1 fusion reaction is solidified, crushed, (22.4:1 mole ratio) typically is neces­ and leached with methanol in a counter­ sary to maintain fluidity and to obtain current procedure. The methanol leach good conversion of the chromium in the removes the majority of the excess NaOH chromite to Na2Cr04' Reaction products while only removing a trace of the build up on the surface of the chromite Na2Cr04" This separation can be accom­ particles preventing further reaction so plished because of the large difference the mixture must be kept well agitated to in solubility of the two compounds in remove these surface products. A methanol. For example, the solubility of stirred-pot-type reactor was used in the present work. Studies also are being numbers in parentheses re­ conducted in Japan on a similar procedure fer to items in the list of references at using sodium nitrate as the oxidizing the end of this report. medium rather than air as used in the s

3

Makeup NaOH

Oxidation and fusion Air-~ 550°-650° C

Solidified fusion product Recycle methanol

Methanol leach

Solid-liquid Solution Recycle separation vaporation 1--NoOH --til.... ~_....--~::.J

Water Recycle NaOH

Condenser Soluble silica

Solution Purification ; ~~ ~dr~: fa"~ d I mllllllll~1II1II1II I

Residue to AI and Si disposal compounds

FIGURE 1.-Chemical processing of chromite.

Bureau method (L-~). The results of the marginal high-iron concentrates, the Bureau study indicated that chromium con­ chromium conversions were in the 90- to versions in the fusion reactor as high as 94-pct range, and with submarginal 98.9 pct can be obtained for high­ chromites the chromium conversion was in chromium, high-grade concentrates. For the 68- to 81-pct range. 4

RAW MATERIALS

The chromite concentrates tested in A concentrate from southern Oregon was this study and presented in a previous derived from a Ni-Co laterite leach resi­ Bureau report (2) were obtained from a due. This material was the residue variety of sources in Alaska, California, remaining after processing in the Bu­ Montana, and Oregon. They were catego­ reau's roast-leach procedure for recover­ rized in one of the following groups: ing nickel and cobalt. Beneficiation of this residue resulted in a high-iron 1. High chromium (metallurgical-grade) chromite concentrate containing 41.5 pct chromite that contained a m1n1mum of Cr203 and 1.7 pct Si02 with a Cr-Fe ratio 46 pct Cr203 with a Cr-Fe ratio >2.0:1. of 1.8 (.!.!). A concentrate from the Mouat Mine in 2. High-iron (chemical-grade) chromite the Stillwater complex in Montana also that contaIned 40 to 46 pct Cr203 with a was tested. This was a marginal, high­ Cr-Fe ratio of 1.5 to 2.0:1. iron material having a Cr-Fe ratio of 1.5 and a high silicon content of 8.5 pct 3. Marginal chromite that met either Si02' the grade or Cr-Fe ratio requirements for The impurity content of the chromites one of the classifications above and very is generally of two types: either gangue nearly met the other. components associated with the chromite grains or lattice impurities in the These classifications were obtained from chromite mineral itself. The silicon Dahlin (9-10). The four concentrates impurity is generally in the form of tested in-Che leaching and solution silicate minerals such as olivine (Fe-Mg purification steps are listed in table 1 silicate) and serpentine (magnesium with their origin, composition, Cr-Fe silicate). Silicon is not present in ratio, and quality classification. Ten significant quantities in the form of additional samples were tested, with silica minerals such as quartz. results reported previously (2)' The chromite mineral is a spinel struc­ Test results for two concentrates from ture theoretically represented by the Alaska are reported in this paper. One formula FeO'Cr203' Magnesium can substi­ sample, from Red Mountain, was a high­ tute for the Fe 2+, and A13+ and Fe 3+ can chromium concentrate containing 56.4 pct substitute for the Cr 3+ in the crystal Cr203 and 1.4 pct Si02 with a Cr-Fe ratio lattice, giving the formula (Mg,Fe)O of 2.8. The other sample was a high-iron '(A1,Fe,Cr)203' In addition to the concentrate from Red Bluff Bay containing lattice impurities, additional iron in 41.7 pct Cr203 and 8.0 pct Si02 with a the form of magnetite (Fe304) is commonly Cr-Fe ratio of 2.0. associated with the chromite grains.

TABLE 1. - Head analyses of chromite concentrates, percent

Sample Cr203 Fe MgO A1203 Si02 Cr-Fe Quality ratio Alaska: Red Mountain (Kenai Peninsula) ••• 56.4 13.7 15.2 8.8 1.4 2.8 High-Cr. Red Bluff Bay (Baronof Is land) ••• 41.7 14.3 16.9 9.2 8.0 2.0 High-Fe. Southern Oregon: Eight Dollar Mountain laterite ••••••••••••••••• 41.5 16.1 12.4 22.3 1.7 1.8 Do. Montana: Mouat (Stillwater complex) .•••••....•••••••••••••••• 36.1 16.7 16.3 15.0 8.5 1.5 Marginal. 5

COUNTERCURRENT LEACHING

EQUIPMENT AND PROCEDURES conditions (12). Each leach stage was performed in a 250-mL sample bottle using The solidified fusion product from the a small, three-bladed propeller for first step of the process (fig. 1) was mixing. The solid-liquid separation crushed and ground to minus 20 mesh. All step, after leaching, was conducted by grinding and screening operations were centrifuging the sample in bottles. The performed in a dry box to prevent the liquid was decanted and added to the next absorption of moisture by the material. appropriate stage. The solids in their Single-stage batch leach tests were con­ same bottle were advanced to the next ducted on this material to determine the leaching stage. In this manner, the appropriate conditions to use in counter­ solids remained in the same bottle for current leach tests. These showed that all three leach stages and did not have the solid-to-liquid ratio and the amount to be removed, minimizing handling of water in the methanol were important losses. factors. In particular, the water in the Fusion products from the four concen­ methanol solvent had to be limited to trates reported in the Raw Materials sec­ 5 pct or less to prevent solubilization tion were leached in three countercurrent of NazCr04. stages with methanol, then the residue Countercurrent leach tests were per­ from the methanol leach was leached with formed in a stepwise manner using the water in three countercurrent stages Shanks system to simulate steady-state (fig. 2) in the equipment previously

NaOH-riCh:j methanol ~MethanOI in 1- 2 3 Fusion .1 1- .1 Solids to product in step 2 Step I, methanol leach

t Water In Final residue Step 2, aqueous lea ch to disposal

Solids from step 1 f Solids from step I r-- Water in Na 2 Cr04-rich I 2 3 4 aq ueous soluton - r---- Final residue to disposal I t Step 2, split-feed aqueous leach

FIGURE 2.--Countercurrent leach flowsheet. 6

described. All the leach tests except and trace amounts «0.4 pct) of NaOH were one were conducted with 30 pct solids in detected in all residues from the water the leach slurry; the other test was leach except for the Red Bluff residue. conducted at 15 pct solids. The 30-pct­ The soluble chromium remaining in the solids value was chosen in order to leach residue from all four materials produce methanol leach solutions that represents inadequate washing of the were nearly saturated with NaOH, while residue. This material could be readily the 15-pct-solids value was chosen to removed by the addition of one or two determine if a greater NaOH extraction more aqueous leach stages. Additional could be obtained with a more dilute aqueous stages were not used in the test solution. This information is necessary program because the number of cyles to to select the procedures that will mini­ reach equilibrium was increased greatly mize the energy required to evaporate the and also required the use of much more methanol so that the NaOH can be recy­ material. Five complete leach cycles cled. Leach time was 30 min in the were used for both the methanol and methanol leach step and 10 min in the aqueous leach steps to attain equilibrium water leach. A particle size of minus 20 conditions. plus 32 mesh was used in most of the The NaOH concentration in the methanol tests in the methanol leach. The parti­ solution ranged from 146 gIL for the Red cles resulting from the methanol leach Mountain fusion product to 224 gIL for was used in the water leach, and these the Mouat material. A saturated NaOH particles were quite fine because the solution in methanol contains approxi­ original particles were broken down as mately 240 gIL NaOH. The Na2Cr04 concen­ the methanol removed the NaOH. A tration in the aqueous solution was in methanol solvent with a concentration of the range of 249 to 280 gIL, and the NaOH 95 wt pct methanol and 5 pct water was was 35 to 80 gIL. An aqueous saturated used in all tests. solution at 25° C would contain approxi­ mately 530 gIL Na2Cr04 and 150 gIL NaOH. RESULTS AND DISCUSSION The chromium concentration in the aque­ ous solution was relatively dilute com­ The results indicate that three-stage pared with the mother liquor concentra­ countercurrent methanol leaching removed tion that would be used in a crystallizer 84 to 94 pct of the unreacted NaOH. The in the product recovery step of the higher extraction was obtained from process. This requires that a large fusion products containing a higher ratio volume of water be evaporated in the of NaOH to chromite. Testing showed that crystallizer, with its corresponding high only 1 pct more NaOH would be removed energy consumption. The solution concen­ with a fourth stage. As shown in table tration cannot be increased much by 2, only a trace of the chromium was found increasing the percent of solids in the in the methanol solution. The aqueous leach slurry because the mixture would solution contained the remainder of the contain too many solids to mix effective­ NaOH and 90 to 99 pct of the soluble ly. To produce a more concentrated solu­ chromium. The remainder of the chromium tion, a split-feed arrangement (shown in

TABLE 2. - Results of two-step, three-stage countercurrent leach tests, percent NaOH-chromite Solids Extraction d- Residue Concentrate wt ratio in Methanol Wate NaOH Cr in fusion leach NaOH Cr NaOH Cr Red Mountain •••••••••••• 2: 1 30 88.8 Tr 11.1 98.7 0.1 1.2 Red Bluff Bay ••••••••••• 4:1 30 87.9 Tr 7.7 96.5 4.4 3.5 Eight Dollar Mountain ••• 4: 1 30 84.0 Tr 15.6 96.8 .4 3.2 Mouat ••••••••••••••••••• 6:1 30 94.3 Tr 5.5 90.0 .2 10.0 6:1 15 92.7 Tr 7.2 98.6 .1 .4 Tr Trace. 7 figure 2) was tested. In this procedure, leaching sequence was repeated for seven fresh feed equivalent to 30 pct solids cycles to produce the final equilibrium was fed into each of the first two leach conditions. The separation of solids stages. The residues from each of these from liquids in this procedure was very stages was combined in stage 3. Enough difficult, especially in the first stages soluble material was removed from the with the concentrated solutions. The solids in each of the first two stages so solutions were very viscous, and the that when the solid residue from each of particle size was small, making for a these stages was combined, the resulting difficult separation. A more reasonable slurry contained about 30 pct solids. procedure would be to operate the first Four stages were used in this test. This two leach stages at 20 to 25 pct solids was not enough because the final residue to allow an easier solid-liquid separa­ contained an appreciable amount of solu­ tion while still producing solutions more ble chromium and NaOH. The final leach concentrated than those obtained with a solution from this procedure contained straight countercurrent leach at 30 pct 449 giL Na2Cr04 and 185 giL NaOH. The solids. At least two additional stages final residue contained 12.1 pct of the would also be required to insure that all soluble chromium fed to the leach circuit the soluble chromium and NaOH were and 14.1 pct of the soluble NaOH. The removed from the residue.

SOLUTION PURIFICATION

No significant amounts of impurities the fusion reaction generally resulted in were found in the methanol solution. an increased silicon extraction. Aluminum and silicon were found in the The total amount of silicon found in 10- to 50-ppm range. the aqueous solution after countercurrent The major impurities that were solubi­ leaching ranged from 1.6 to 13 pct of the lized in the aqueous solution by the total silicon content in the chromite. fusion reaction were silicon and alumi­ The aluminum content in the solutions num. A minor amount of ferrous iron ranged from 32 to 71 pct of the total (Fe 2+) also was soluble, but it oxidized amount in the chromite. on exposure to air and precipitated from Dilute solutions such as those obtained solution. Magnesium was a major impurity from a single-stage leach can be purified in the chromite concentrates but did not of silicon and aluminum compounds qy become soluble to any extent. Solution sparging CO 2 into the solution to reduce concentrations were typically 1 ppm Mg or the pH so that silicon and aluminum less. The aluminum extraction tended to compounds precipitate. This procedure is follow the same trend as the chromium not effective with the stronger solutions extraction. This would be expected produced by countercurrent leaching. because the aluminum substitutes for Even after 90 pct of the unreacted NaOH chromium in the chromite lattice. As the in the methanol solution was removed, the chromium was reacted, the aluminum also aqueous solution still contained too much would be exposed and react with the NaOH to reduce the pH by sparging. So NaOH. much Na2C03 was produced that the entire The silicon extraction appeared to be solution solidified as the Na2C03 sol­ more random, although it generally ubility was exceeded. decreased with time after increasing for A procedure patented by Holtz (ll) was the first hour of the reaction. Reac­ used to purify solutions resulting from tions at 550 0 C often resulted in greater countercurrent leaching tests. In this extractions than those at 650 0 C. These procedure, a soluble silica compound was lower extractions at the longer reaction added to the solution so that the ratio times and higher temperature may be due of silicon to aluminum was adjusted to to the formation of high molecular form the compound NaA1Si04 , which then weight, insoluble metal silicates. precipitated from solution. The silica Increasing the NaOH-to-chromite ratio in compound was added in the form of 8

waterglass (Na2Si03) or Ludox6 suspen­ used. The waterglass with a crystalliza­ sion of colloidal silica. The solution tion seed appeared to result in a better was initially heated to 70° to 90° C for aluminum extraction, and the precipitate 30 min to 2 h after the silica compound was easier to filter than when using the was added, then the solution was boiled Ludox colloidal silica. This colloidal for 30 min to 1 h. In some cases, a silica alone (i.e., without a crystalli­ crystallization seed in the form of soda­ zation seed) resulted in a gelatinous lite (NaAlSi04) was also added to promote precipitate that was very difficult to the formation of a more crystalline filter. Boiling for a longer period of product. Other variables investigated time also resulted in improved aluminum were (1) solution temperature when the extractions and improved filterability. soluble silica compound was added, The amount of crystallization seed was (2) temperature of initial heating step, based on the aluminum content of the (3) time at initial heating step, and solution and was calculated to be approx­ (4) time at boiling temperature. These imately 10 pct of the expected weight of tests are summarized in table 3. the precipitate. When the silica compound was added to The precipitate obtained in these stud­ the solutions, a thick gel initially ies was identified by X-ray diffraction formed, particularly in the case of the as 1.0INa20·A1203·1.68Si02·1.73H20. The Ludox colloidal silica. The gel eventu­ chromium content of the precipitate ally broke down, but the best results ranged from 0.24 to 1.70 pct, with most were obtained when the silica was added of the values in the 0.7 to 0.9 pct slowly and a crystallization seed was range. Final aluminum content in the solutions was in the 0.3 to 0.9 giL 6 Refe renee to specific products does range, with the silicon content at 0.1 to not imply endorsement by the Bureau of 0.6 giL. Mines.

TABLE 3. - Results of solution purification tests

Solution Addition Crystal- Preheat Preheat Boiling Al extrac- Si extrac- and soluble temp 1 lizat ion temp, °c time, h time, h tion, pct tion, 2 pct silica type seed LATERITE Ludox •••••••• Hot ••••• Yes ••••• 70-80 1 0.5 100 98.8 Do ••••••••• • • • do ••• No •••••• 70-90 .5 .5 99.8 97.8 Waterglass •.• • •• do ••• No •••••• 70-80 1 .5 94.5 99.9 Ludox •••••••• Cold •••• Yes ••••• 70-80 1 .5 92.9 97.3 Do .•••••••• Hot ••••• Yes ••••• 70-80 1 .5 95.1 96.6 Waterglass ••• Cold •••• Yes ••••• 70-80 1 .5 96.0 97.0 Do ••••••••• Hot ••••• yes •••.. 70-80 1 .5 96.5 96.8 Ludox •••••••• · . . do ... yes ••••• 80-90 2 .5 96.7 96.2 Waterglass ••• • • • do ••• Yes ••••• 80-90 2 .5 94.9 97.0 RED MOUNTAIN Ludox •••••••• Cold •••• No •••••• 70-80 1 0.5 43.3 98.0 Do ••••••••• · . . do ... yes •••.• 70-80 1 .5 59.6 98.3 Waterglass ••• • • • do ••• No •••••• 70-80 1 .5 84.4 95.8 Do ••••••• ~ • · . . do ... Yes ••••• 70-80 1 .5 87.8 99.1 Ludox •••••••• Hot ••••• yes ••••• 80-90 1 1 97.2 99.2 Waterglass ••• • •• do ••• yes ••••• 80-90 1 1 97.5 99.2 lCold--treated solution at ambient temperature. Hot--treated solution at preheat temperature. 2Si extraction includes amount in original solution plus amount added to precipi­ tate aluminum. 9

CRYSTALLIZATION

EQUIPMENT AND PROCEDURES In the general operation of the crys­ tallizer, approximately 2 L of mother A semi continuous vacuum evaporative liquor of the desired composition was crystallizer (fig. 3) was constructed of added to the crystallizer. Synthetic standard taper-joint glassware. The solutions were used for the mother liquor crystallizer body was constructed from a and feed solution rather than solutions 3-L round-bottom flask that was heated produced from actual chromite concen­ with an electric heating mantle. A trates during most of the testing. Syn­ ground-glass shaft sealed with a mineral thetic solutions were used because of the oil gland was used to stir the contents lack of personnel and equipment to pro­ of the reactor. A vaCUum was obtained in duce large quantities of actual leach the system with a water aspirator, and solution. The solution was then heated the desired vacuum level was maintained to the desired temperature, and the with a pressure-relief valve. The con­ vacuum was adjusted to induce boiling at densate was collected in 500-mL flasks this temperature. The temperatures gen­ that were kept in an ice bath to prevent erally used were 65° C or 80° to 85° C. excessive loss of condensate due to the Limited testing was done at 65° C because vacuum in the system. Fresh feed was the necessary vacuum could not be mai.n­ added to the system at a known rate with tained to induce boiling with the more a piston-type metering pump. Product concentrated solutions. After crystals crystals and mother liquor samples were began to form, the feed was started. The obtained by means of a dip-tube placed in feed composition was kept constant at the crystallizer slurry. This tube was 40 giL NaOH and 260 giL Na2Cr04 for all sealed to a hot-water-jacketed filter, the testing. A fixed feed rate of 10 mLI and samples were obtained by releasing min was used i.n all tests. Every 30 min the system vacuum and applying a vacuum samples of the mother liquor and crystals to the filter flask. Crystals were col­ were taken and the condensate flask was lected on the filter and a mother liquor changed and the volume recorded. The sample was obtained from the filter weight of crystals and volume of mother flask. liquor taken were determined by a

Product dip tube

vacuum-release I valve ~;~~mCO\ctIOO

Condensate '-t..I;;;====;;;u' H eat e d receiver ~~::::!J filter vacuum-control/ valve Electric 3-L crystallization heating body mantle Ice bath Metering pump/ 4eed reservoir 'Filter flask FIGURE 3.--Vacuum evaporative crystallizer. 10

material balance so that the concentra­ and the remainder was returned to the tion of NaOH and Na2Cr04 in mother liquor crystallizer. The crystals were dried in and the volume of mother liquor in the a drying oven and then kep't in air-tight crystallizer remained unchanged. The sample bottles while awaiting chemical rate of heat input to the crystallizer analysis and a screen analysis. also was adjusted slightly to maintain steady-state conditions. As samples of RESULTS AND DISCUSSION mother liquor were obtained during the run, they were titrated for NaOH and Synthetic Leach Solutions Na2Cr04 concentration to insure that steady-state conditions were maintained. The NaOH concentration in the mother The product crystals collected on the liquo~ was varied from 422 giL to filter were washed with methanol to 1,058 giL. The corresponding saturated remove residual mother liquor adhering to concentration of Na2Cr04 ranged from 330 the crystals. The volume of methanol was to 93 giL at these NaOH levels. As shown normally 2 cm 3 /g of crystals. The in table 4, the average crystal size methanol was applied in three portions ranged from 28 to 100 mesh, and the NaOH and mixed with the crystals with each content in the crystals was generally application. The resulting slurry was less than 0.01 pet. The particle size then vacuum filtered. Approximately 75 g shown is the average of all samples taken of crystals was obtained with each and represents the size for 50 pet of the sample. Removal of only 25 to 30 mL of weight. Higher levels of NaOH in the mother liquor normally was required to crystals are d~e to inadequate washing of maintain steady-state conditions in the the crystals and not to NaOH cocrystal­ crystallizer, but to obtain the necessary lized in the Na2Cr04 matrix. As the weight of crystals, approximately 250 mL mother liquor became more concentrated, of mother liquor had to be processed it became more difficult to wash the through the filter. A sample of mother crystals. The maximum practical limit liquor was taken from the filter flask, for the mother liquor composition was

TABLE 4. - Results of vacuum crystallization tests on synthetic solutions

Av mother liquor Av crystal size, NaOH content in composi tion J5./L Vacuum, in Hg Temp, °c mesh crystals, pct NaOH Na2Cr04 422 330 27 74 28 <0.01 436 336 26 78 60 <.01 535 288 22 100 28 <.01 547 241 26 88 42 .01 564 223 26 84 (I ) <.01 604 205 27 84 42 <.01 654 192 27 85 48 <.01 655 221 27 88 60 <.01 676 188 27 88 42 <.01 680 194 27 80 60 <.01 685 202 27 87 65 <.01 686 192 27 85 42 .01 742 170 27 90 65 .05 775 178 27 93 48 .13 802 176 27 95 48 .01 886 191 27 105 100 .02

1,058 93 27 110 I (2) i NAp NAp Not applicable. lSaturated Na2Cr04 aqueous solution used for wash; crystals fused together. 2Crysta1s too fine to wash. 11

around 800 giL NaOH. As this level was and time left in the testing program. exceeded, the solution became very The residue from the methanol leach was viscous, and the particle size became given a single aqueous leach at 35 pet very small, making washing difficult. solids to produce solution for crystalli­ The very small crystals also are not zation. The solution c.oncentration using desired. 35 pct solids is approximately equal to An attempt was made to correlate that which would be obtained in a coun­ crystal size with operating conditions, tercurrent leach at 30 pct solids. This but this was not accomplished because of solution, containing 46 giL NaOR, 212 giL the nature of the equipment and its semi­ Na2Cr04, 17.6 giL AI, and 0.24 giL Si, continuous method of operation. The then was treated with waterglass to crystal size distribution was not consis­ remove the aluminum impurity. The solu­ tent from batch to batch. The crystal tion, after treatment, contained 0.11 giL size generally began as a small size that Al and 0.69 giL Si. Because of the grew larger and was eventually removed as extended period of time the solut ion was product. Then a new batch of fine exposed to air during the fi oper­ crystals was generated that in turn grew ation after the aqueous leach and during in size. A scale apparatus with the impurity removal procedure, the solu­ some sort of crystal classification tion absorbed CO 2 from the air to produce system, such as an elutriation leg, would 15 to 16 giL Na2C03' Attempts to remove give more meaningful results on the this impurity by precipitation with lime crystal size distribution. High NaOH met with limited success. Fifteen liters concentrations generally resulted in more of a final solution containing 11 giL sm~ll crystals than did the lower NaOH Na2COS, 49 giL NaOH, and 253 giL Na2Cr04 concentrations. It was demonstrated on were available for crystallization this small-scale equipment that NaOH-free studies. crystals of moderate particle size can be A synthetic mother liquor with the com­ produced from these solutions by vacuum position of 700 giL NaOR and 175 giL evaporative crystallization. More Na2Cr04 was used in the crystallization refined operating conditions will have to tests. The crystallizer was operated at be evaluated in larger scale, continuous a vacuum of 27 in Hg, resulting in a equipment that more closely simulates boiling temperature of 90° C. A total of commercial-scale equipment. 11.6 L of feed solution was processed in the crjstallizer, producing 2.84 kg of Actual Leach Solution crystals. This represented a recovery of ------96.6 pct of the Na2Cr04 fed to the Four crystallization tests were per­ system. The average size of the crystals formed with a solution produced .from was 42 mesh. The crystals contained less leaching a chromite fusion product rather than 0.01 pct AI, Si, and NaOH, but the than using synthetic solutions. Twenty­ Na2C03 content averaged 3.7 pct. In any four kilograms of a fusion product pro­ future tests, care must be taken to mini­ duced from Eight Dollar Mountain laterite mize the time the aqueous solutions are was leached with methanol in three sepa­ exposed to air to avoid the problem of rate stages at 30 pct solids to remove carbonate contamination. The Na2C03 is the bulk of the unreacted NaOH. A the least soluble component and will countercurrent leach could not be per­ contaminate the crystals if present. formed because of insuffi.cient material SUMMARY AND CONCLUSIONS Laboratory-scale testing showed that with fused NaOH under oxidizing con­ low-grade domestic chromite concentrates ditions to form Na2Cr04, which is then that are not suitable for chemical pro­ recovered by leaching and crystalliza­ cessing by present commercial methods can tion. Product crystals of Na2Cr04 were be treated successfully by a procedure obtained that contained less than devised by the Bureau of Mines. The 0.01 pct NaOH and whose average particle procedure involves reacting the chromites size ranged from 28 to 65 mesh. 12

REFERENCES

1. Papp, J. F. Chromium. Ch. in Min­ 8. Kashiwase, K., M. Mita, T. Kon, and eral Facts and Problems, 1985 Edition. T. Okabe. Methanol Leaching of Reaction BuMines B 675, 1985, pp. 139-156. Product of Chromite With Molten Sodium 2. Copson, R. L. Production of Chro­ Salts. Nippon Kagaku Kaishi, v. 9, 1975, mium Chemicals. Ch. in Chromium, Chemis­ pp. 1491-1495. try of Chromium and Its Compounds, ed. by 9. Dahlin, D. C., L. L. Brown, and M. J. Udy. Rei,nhold, v. 1, 1956, J. J. Kinney. Podiform Chromite Occur~ pp. 262-282. rences in the Caribou Mountain and Lower 3. Hartford, W. H., and R. L. Cops on. Kanuti River Areas, Central Alaska. Part Chromium Compounds. Ch. in Kirk-Othmer II; Beneficiation. BuMines IC 8916, Encyclopedia of Chemical Technology, ed. 1983, 15 pp. by A. Standen. Wiley, v. 5, 2d ed., 10. Dahlin, D. C., D. E. Kirby, and 1964, pp. 473-516. L. L. Brown. Chromite Deposits Along the 4. Kashiwase, K., G. Sato, E. Narita, Border Ranges Fault, Southern Alaska. and T. Okabe. Solubility of Sodium (In Two Parts). 2. Mineralogy and Chromate in Sodium Hydroxide Solution and Results of Beneficiation Tests. BuMines in Methanol Solution. Nippon Kagaku IC 8991, 1985, 37 pp. Kaishi, v. 7, 1974, pp. 1224-1229. 11. Kirby, D. E., D. R. George, and 5. Hundley, G. L., D. N. Nilsen, and C. B. Daellenbach. Chromium Recovery R. E. Siemens. Extraction of Chromium From Nickel-Cobalt Laterite and Laterite From Domestic Chromites by Alkali Fusion. Leach Residue. BuMines RI 8676, 1982, ~uMines RI 8977, 1985, 14 pp. 22 pp. 6. Chandra, D., C. B. Magee, and 12. Treybal, R. E. Leaching. Ch. in L. Leffler. Extraction of Chromium From Mass-Transfer Operations. McGraw-Hill, Low-Grade Chromium-Bearing Ores (grant 1955, pp. 590-591. G0284009, Denver Res. Inst.). BuMines 13. Holtz, J. W. Process for Removal OFR 151-82~ 1982, 164 pp.; NTIS PB 83- of Alumina From Aqueous Metal Chromate 106781. Solutions. U.S. Pat. 4,173,618, Nov. 6, 7. Okabe, T., and K. Kashiwase. Pro­ 1979. cess for Production of Alkali Metal Chromates. U.S. Pat. 3,859,412, Jan. 7, f975.

U.S. GOVERNMENT PRINTING OFFICE: 19B8 - 547,000/80,029 INT.-8U.OF MI NES,PGH. ,PA. 28682