American Mineralogist, Volume 59, pages 1209-1219, 1974

Synthesis of Barite, Celestite, Anglesite, Witherite, andStrontianite from Aqueous Solutions

Cnenras W. BrouNr

Department of Geology, Uniuersity of Georgia,Athens, Georgia 30602

Abstract

Well-formed single crystals of barite, strontium barite, celestite, and anglesite have beensynthesized in gram quantities from dilute aqueous solutions in borosilicate glass flasks by aslow precipitation procedure over a range of experimental conditions. The largest crystals weregrown in a 0.03 molar HCI solution, near 95"C in a non-uniformly heated flask by combiningdilute metal chloride and sodium sulfate solutions at a rate of less than 5 mllhour. The optimummetal concentrations for growing large crystals were 0.01 molar for barite, O.02 molar forcelestite, and 0.03 molar for anglesite. The optimum Na,SOr solution concentrations were 0.01molar for barites and 0.02 molar for the other sulfates. Maximum dimensions for the thicktabular single crystals were 0.8 mm for barite, 1.2 mm for celestite, and2.3 mm for anglesite,In contrast amorphous or submicron-sized precipitates resulted from the use of more con-centrated solutions, too rapid a solution cornbination rate, the use of plastic ware, the use ofmagnetic stirring, inadequate cleaning of glassware, and uniform heating of the flask.

For the carbonates, witherite and strontianite crystals over I mm long were grown in a flaskdesigned to be completely filled with solution. The crystals were grown by combining 0.02 molarmetal chloride and sodium bicarbonate solutions at a rate near 5 ml/hr. at 80"C in solutionsthat were initially 0.001 molar in HCl. Unlike the sulfate minerals, a uniformly heated flaskresulted in larger carbonate crystals.

The crystals of sulfate and carbonate minerals had the same indices of refraction as naturalminerals. The spacing and intensities of powder diffraction peaks agreed with those publishedin the Powder Diffraction File for the minerals,

fnhoductionMany minerals that exhibit low solubilities in

aqueous solutions are difficult to synthesize in thelaboratory, especially when coarse, crystaUine ma-terial is needed. As an alternative, the experimental-ist may have to use natural materials that may be im-pure because of ionic substitutions or small foreigninclusions. In many such experimental studies thepresence of these impurities and the chemical effectsthat they may cause are neglected. In some chemicalsystems, the effects produced by impurities may benegligible; however, in other systems the effects maybe very large. For example, the presence of strontiumin barite markedly affects the barium concentrationin solutions (Gundlach, Stoppel, and Striibel, lg72).The use of synthetic minerals of known high puritythat are chemically and structurally identical to thenatural minerals would eliminate the need to correctfor coprecipitated impurities or to justify ignoringtheir presence. There are other advantases also. Aradioactive or isotopic tracer could be idded when

the mineral is synthesized as an analytical aid. Also,since synthetic crystals have a well developed ex-ternal morphology, studies of crystal growth, dis-solution, and alteration can be made (Blount andBeckrnan, 1974).

The initial objective of the synthesis studies re-ported here was to develop a reliable procedure formaking gram quantities of pure, coarsely crystallinebarite (BaSOa) for both hydrothermal solubility andphase stability studies. Similarly, procedures werethen developed for synthesizing anglesite (PbSO+),celestite (SrSoa), witherite (BaCOa), and strontia-nite (SrCO3) from acqueous solutions at conditionssimilar to those that are believed to have precipitatedthe natural minerals. Normally, however, precipi-tates from aqueous solutions in the laboratory arevery fine grained, poorly crystalline, and tend to beeasily broken. An extensive survey of the literaturewas therefore undertaken to learn what factors maybe important for the synthesis of these minerals.

Large crystals of synthetic barite, celestite, and

l2to

anglesite have been produced by a number of pro-cedures. Frequently the experimental synthesis con-ditions are far different from those under whichthe natural minerals are believed to have been pro-

duced. For example, Patel and Bhat (1972) grewsingle crystals of SrSO+ and BaSO+ up to 1 mmlong in silica gels at 20-35"C. These crystals, how-ever, contained large amounts of silica, and theirperfection of external form diminished with in-creasing size. Patel and Koshy (1968a) and Pateland Bhat (1971) developed procedures for growingcrystals of SrSO+ and BaSO+ from NaCl melts attemperatures between 700 and 900'C that measuredup to 7 mm long and 4 mm wide.-PbSO4 crlstalsof a poorer quality up to 2 mm long were grownin NazSO+ melts in a similar manner. While the(011) and (102) faces were recognized, photo-graphs show the crystals to be quite irregular witha poorly developed morphology. Moreover, the facesof crystals grown by this procedure showed a den-dritic pattern which was not observed in any of thesamples of natural material. This pattern was be-lieved to have been produced by a rapid precipita-tion of dissolved metal sulfate that took place whenthe flux solidified (Patel and Bhat, l97l). X-raydiffraction measurements and hardness tests gave thesame results for both natural and melt-grown crys-tals. The melt-grown crystals also showed fewerdislocations on etched cleavage surfaces than naturalmaterials (Patel and Koshy, 1968a, 1968b).

A large amount of research work has been under-taken in efforts to improve aqueous precipitationprogedures that are employed for the determinationsof barium and sulfate. The crystals of BaSO+ thatare usually produced by rapid precipitation pro-cedures from aqueous solutions are small, l-20microns (Liteanu and Lingner, 1970a, l97Ob;Lieser, 1969). Crystals produced in this way arealso easily broken yielding much finer material(Nielsen, 1957; Fischer and Rhirlehammer, 1953).These precipitates are therefore not suitable formany types of experimental studies. Precipitates ofSrSO+ and PbSO+ have similar properties. Longdigestion periods in hot aqueous solutions have beenused to recrystallize and coarsen precipitates. Magerand Lieser (1972a,b,c) studied recrystallizationrates and the aging of SrSO+ precipitates and con-cluded that the effects resultant from aging didnot stop even after a very long time. NazSO+ thatwas occluded during the precipitation continued tobe released as recrystallization continued. Aging of

precipitates of Sr, Ba, and Pb sulfates has been usedto prepare crystalline material with an average sizenear 10 microns (Nancollas, 1968). Blount andDickson (1968) synthesized celestite with the sameaverage grain size, but individual grains ranged from1 to 120 microns in length. Recently, Brower(1973) used thermal cycling to grow syntheticbarite crystals ranging from 50 to 500 microns inlength. Barites containing up to 20 mole percent

strontium were also synthesized. Up to 2 monthsof thermal cycling were required in experiments thatyielded 5 to 7 milligrams of solid. A considerableamount of finer material appears to be present inthe photographs in this paPer.

Laitinen (1960) summarized previous studies onthe nucleation, growth, properties, and contamina-tion of BaSO+ and other precipitates used for gravi-metric analyses. While aimed at improving analyticalprocedures, their studies also provided the type ofbasic data needed for synthesizing highly pure sul-fates, as large and dense crystalline grains thatresist breakage and display a crystalline structureidentical to the corresponding natural minerals. Thetwo most important aspects of these prior studiesare the efforts ( 1) to reduce the number of nuclea-tion sites to promote the formation of a coarserprecipitate and (2) to reduce contamination by in-creasing crystallinity.

The number of nucleation sites can be reducedin a number of ways. Nielsen (1957) found thatcareful cleaning of glassware using steam generatedfrom distilled water resulted in nearly a 10 foldreduction in the number of crystals formed. Liteanuand Lingner (I970a,1970b) used a hot alkalineEore solution to accomplish the same purpose.Substances that catalyze the nucleation of the pre-cipitate also must be avoided. For example Brickerand Myers (1969) reported that Teflon catalyzedthe precipitation of BaSO+. Pretreatment of thesolutions used for the precipitation is also important.O'Rourke and Johnson (1955) and Bogen andMoyer (1956) both observed that either filtering oraging the BaClz solutions for 24 ot more hoursreduced the number of crystals l0-fold comparedto freshly prepared solutions. They suggested thataggregates of incompletely dissolved BaCl2 mayhave served as nuclei for BaSOr crystals' The useof dilute solutions is necessary. Liteanu and Lingner(1970a,b) studied the eftects of many differentexperimental conditions on BaSO+ precipitates ob-tained using their rapid precipitation procedure.

CHARLES W. BLOUNT

STNTIIES/S OF SULFATE AND CARBONATE MINERALS l2l1

Depending upon the experimental conditions, theaverage grain size varied between 1.2 and 6 microns.The largest grains were produced in glassware thathad been carefully cleaned and by using dilutesolutions (below 0.001 molar) that had been aged,filtered, or boiled prior to use. Mealor and Town-shend (1966a,b) showed that all solutions containednumerous minute particles (heteronuclei) that pro-vided possible nucleation sites. Attempts to removethese heteronuclei were unsuccessful. Precipitationprocesses from dilute (0.01 molar) solutions werealways heterogeneous. The addition of other dis-solved substances to solutions has brought aboutan increase in the average grain size in rapid pre-cipitations (Fischer and Rhinehammer, 1954). Theaverage grain size of BaSO+ precipitates increasedfrom 14 microns in water to 135 microns in a 0.005molar FeCl3 solution.

Crystal perfection strongly depends upon chem-ical conditions. Nielson (1958) showed that lowconcentrations favored surface-controlled growthrates and the formation of orthorhombic crvstals.At higher reagent concentrations, diffusionally con-trolled growth processes became important and star-like crystals were produced. Gordon (1955) rec-ommended homogeneous precipitation proceduresusing sulfamic acid or similar slow reacting com-pounds as a source for sulfate because a slow rateof precipitation favored the formation of dense crys-tals with a minimum coprecipitation of other ionsfrom solution. Fischer and Rhinehammer (1953)also showed that the perfection of BaSOr crystals(average size 14 microns) was markedly improvedfor rapid precipitations when the solutions wereaged, dilute (0.016 molar), acidic (pH 0.5-1), andhot (98"C). Crystals grown in cool or in near-neutral solutions were larger (up to 60 microns)but tended to be feathery, irregular, friable, andbuoyant.

Espig and Neels (1967) were able to precipitateBaSOa, SrSOa, and PbSO+ crystals with good crys-talline forms at room temperature from rapidlymixed dilute solutions ranging up to 500 times sat-uration concentrations. The crystals produced weresmall: SrSOa to 15 microns, BaSOa to 30 microns,PbSO+ to 50 microns. Crystal forms varied de-pending upon the concentration and composition ofsolutions. They proposed but did not carry outgrowing much larger crystals by repeated additionsof supersaturated solutions to a solution containingsome crystals.

Burkhard (1973) dissolved aqueous precipitatesof barium and strontium sulfates in concentratedsulfuric acid and reprecipitated the solid by slowlyevaporating the acid at 160'C. (Ba,Sr)SOa solidsolution crystals over the whole range were preparedincluding pure barite and celestite. Well-formed in-dividual crystals up to 2 mm long were prepared inthis manner. Intermediate composition crystals weresmaller and less perfect in form with a barrel-shapedform at 55 percent SrSO+.

Experimental

Review of previous studies on the precipitationof BaSO+ indicated a number of experimental con-ditions that could be applied in developing a pro-cedure for producing large and dense single crystalsof barite of high purity. Six of the experimentalconditions that dppeared to be most important arelisted below.

(1) Clean glassware: All glassware was initiallycleaned using a hot chromic-sulfuric acid cleaningsolution. After rinsing thoroughly with distilledwater, the glassware was cleaned using steam fromdistilled water following the procedure describedby Nielsen (1957) or cleaned using a 0.1 molarEore solution as recommended by Liteanu andLingner (1970a).

(2) Solution homogenization: Experimental so-lutions were used only after they had aged forseveral days or were filtered or boiled.

(3) Solution concentrations: Experimental solu-tions were diluted or made at concentrations be-tween 0.001 and 0.03 molar.

(4) pH: Experimental solutions in the flask weremade acidic by the addition of HCI to obtain a con-centration between 0. 1 and 0.6 molar.

(5) Temperature: The precipitations were car-ried out at temperatures between 90 and 98'C.

(6) Precipitation rate: The precipitating solutionat room temperature was added at a rate between2 and 5 ml/hr to the hot solution in the flask sothat the precipitation was carried out very slowlyover a period of many days.

All experimental solutions were prepared usingreagent grade chemicals and distilled water. Mostsolutions were prepared well in advance of the timeof their use. The experimental equipment (Fig. 1a)consists of a gravity feed system for adding theprecipitating solution, an Erlenmeyer precipitationflask, and a standard hot plate with a variable tem-perature control. The gravity feed system is com-

t 2 l 2

posed of a borosilicate glass chromatographic col-umn with a 500 ml reservoir, a precision-type Teflonneedle valve to control rate of solution flow, a dis-posable Pasteur pipet for a drip tip, and severalpieces of Tygon tubing to make connections: TheTeflon needle valve is positioned to the side toprevent air bubbles that accumulate in the tubingabove the drip tip from blocking the valve. Theflasks used ranged in volume from 1 to 6 liters.A perforated watch glass reduces the evaporationrate of the solution. The flask is centrally placed ona short (3 cm high) circular aluminum block whosediameter is about 1./3 that of the bottom of the flask.This block promotes convective mixing in the solu-tion.

obFtc. 1. Experimental apparatus for growing (a) sulfate

minerals and (b) carbonate minerals.

lcgend : A-Chromatographic columnB-Tygon tubing connector I-Variable control hot plate

For a precipitation experiment, the Erlenmeyerflask was partly filled (l/6 to l/2 of its capacity)with one solution, generally the metal chloride. This

solution was acidified, heated to the desired tem-perature, and the precipitating solution, generally

NazSO+, added. The precipitation was allowed toproceed from several days up to several weeks de-pending upon the volume of solutions to be com-bined. The precipitate was separated by filtrationwhile the solution was hot.

Minor difficulty was encountered in maintaininga constant rate of flow for long periods of time with

the gravity flow system. Air tended to collect in the

tubing below the metering valve' This air was ex-pelled periodically to prevent obstruction of the

valve. The valve also required minor adjustment to

keep the flow rate as constant as possible. Variationsin flow rate below 5 ml/hr did not appear to affect

the size and perfection of the precipitate. Additionsat a more rapid rate (e.g', 2O ml/hr) resulted in

a finer grained precipitate. It was not possible to

examine quantitatively the effect of rate of flow

on the size of crystals because of the rather poor

control obtained from the gravity feed system. A

motor-driven infusion pump could be used for such

a study and to keep the flow rate constant.A number of variations in experimental proce-

dures and conditions were examined, but only afew improved the quality and increased the quantity

of solid produced. Many had no noticeable effect

or else markedly degraded the quality of the pre-

cipitate. Eight of these variations are describedbelow.

1. Effect of additional substances: The possibility

that other substances in solution could increasethe size of crystals (Fischer and Rhinehammer'1954) was examined by adding ferric chloride to

a solution. There was no noticeable change in the

size of crystals; however, a reddish precipitate was

observed on some of the crystals. While a singleexperiment is not conclusive, the lack of any changeand the contamination of the crystals made this

approach unattractive. Precipitation of barite froma 2 molar NaCl solution resulted in the formationof finer grained precipitate.

2. Etrect of stirring: Stirring with a Teflon-cov-ered magnetic stirring bar produced an exceedinglyfine grained solid. Possibly the Teflon catalyzedthe precipitation, or the rotating stirring bar ground

the solid. This method of stirring is detrimental.Mechanical stirring of solutions in the precipitating

CHARLES W, BLOUNT

C-Precision valveD-Drip tipE-Thermometer

J-Flask solutionJ'-Precipitating solutionK-Flask closure

M-Outlet tubeN-Silicone oil temperature bath

F-Perforated cover glass L-Inlet tubeG-Erlenmeyer flaskH-Aluminum block

flask was not studied further because of the success-ful use of thermal convection described below.

3. Use of plasticware flask: In one experimenta polycarbonate Erlenmeyer flask was substitutedfor the borosilicate flask. Because of the temperaturelimit of polycaibonate, this flask was heated in awater bath. The BaSO+ precipitate produced wasextremely fine grained, resembling that obtainedduring rapid precipitations. Also a carbonate phasecomposed of long, highly birefringent crystals wasevident. Polycarbonate flasks were not used afterthis result.

4. Use of larger flasks: The initial studies used1 liter Erlenmeyer flasks, which limited the quantityof barite precipitate to about l/2 gram. Use of a6 liter flask produced a better precipitate with theadded advantage of a 2 to 3 gram yield of crystals.

5. Thermal convection: Heating the flask on ahot plate creates thermal convection in the flaskwhereas placing the flask in a silicone oil bathprovided more uniform heating and thus suppressedconvection mixing. Such suppression produced sig-nificantly smaller crystals. In the next experiment,the flask was placed on a small diameter aluminumblock in order to increase thermal gradients andconvection in the solution in the flask. Fewer crys-tals were produced and there was a recognizeableincrease in crystal size and perfection.

6. HCI concentration variation: The concentra-tion of HCI was varied between approximately 0.01and 0.67 molar. Increasing the acid concentrationup to about 0.3 molar seemed to favor the growthof larger crystals. One experiment at 0.67 molarHCI resulted in a finer-grained precipitate.

7. Reversal of solutions: Normally, a NazSO+solution was added to an acidified metal chloridesolution. In a few experiments, the metal chloridesolution was added to an acidified Na2SOa solu-tion. There was essentially no significant differencein the precipitates produced by reversing the solu-tions.

8. Variation in solution concentration: Sulfateminerals were precipitated from metal chloride solu-tions ranging from 0.001 to 0.1 molar using sodiumsulfate solutions ranging from 0.005 to 0.02 molar.Optimum solution concentrations for precipitatinglarge amounts of coarse-grained crystals of barite,celestite, and anglesite could not be sharply defined.Rather these sulfate minerals seem to precipitateover a range of solution concentrations. There is amaximum concentration that may be used for each

l2t3

sulfate mineral. Precipitates from more concen-trated solutions are finer grained and of poorerquality. Maximum (optimal) concentrations arelisted in Table 1.

Considerable efiort was also spent on attemptsto grow barites containing strontium sulfate. In theseexperiments it was desirable to attempt to maintaina constant solution composition in the precipitationflask. This was done by adding a second solution,that had nearly the same composition as the pre-cipitated barite, at the same rate that the sodiumsulfate solution was being added. Here again thegravity feed system presented control difficulties andtwo 50 ml burettes were used to keep track of solu-tion volumes being added.

Since the solubility of BaSOa, increases with in-creasing temperature up to at least 125'C (Temple-ton, 1960; Blount, l97I), the possibility that BaSOacould be recrystallized in HCI solutions along adecreasing thermal gradient was investigated. Ap-proximately 3 grams of very fine-grained analyticalreagent grade BaSOa was placed in one end ofa 1 inch diameter Pyrex glass test tube that wasnecked down to 1/2 inch in the middle. The tubewas filled to 95 percent of capacity with a 0.6 molarHCI solution and sealed. The end of the tubecontaining the BaSOr was placed in a tubular muffiefurnace and heated to 95'C. The cooler end of thetube outside the mufle furnace was near 40oC.After several weeks, small crystals were seen in thecooler portion of the tube. After 6 weeks, the experi-ment was terminated and the solid examined. Therecrystallized solid consisted of well formed singlecrystals of barite that were up to 0.03 mm longcompared to the 0.001 micron size of the originalreagent powder.

For comparison purposes attempts were made to

Tlnrs 1. Experimental Solution Compositions forProducing Large Crystals of Sulfate Minerals*

C o n c e n t r a t i o n } l o l e s / L i t e r

M e l a l S o d i u n

Prec ip iEate HCI Ch lor ide_ i ' * Su l fa te ' ! ' l c rys taL D inens ions

STNTIIES/S OF SULFATE AND CARBONATE MINERALS

B a r i t e( B a s 0 4 )

C e l e s c i t e( S r S 0 / )

Ang les i te

( Pbs04 )

0 . 8 x 0 , 6 5 x 0 . 5 6

1 . 2 x 0 . 9 x 0 . 5

2 . 3 x 1 . 3 x 1 , 0

0 . 3

0 . 1 5 o . o 2 0 . 0 2

0 . 0 2

o T = 9 4 * 5 o C , v o l u m e 2 l i E e r s i n 6 l i l e r f l a s k , F I @ r a E e 5 - 1 0 n l / h o u r

" " t robab ly a max imum concen! ra ! ion fo r g roF ing Large c lys ta ls .

t 2 t4

Fro. 2. Scanning electron microscope photo of barite crystal.Damaged surface shows cleavage. Bar is 0.5 mm long.

produce barite by some of the previously discussedprocedures. Following Espig and Neels (1967), a0.0033 molar BaClz solution was rapidly mixedwith a 0.0033 molar Na2SO+ solution at room tem-perature in a clean flask and allowed to stand for72 hours before filtration. A fine precipitate wasobserved about 1 minute after the solutions weremixed. The resulting precipitate consisted of small1-2 micron crystals. A few plateJike crystals up to15 microns long were observed; however, thesewere so thin (2 microns) that they did not showany apparent birefringence and were very easilybroken into smaller fragments. An attempt was alsomade to grow barite crystals in NaCl melts, follow-ing the procedure described by Patel and Bhat(197I). A mixture of BaSO+ and NaCl was placedin a platinum crucible and heated to 900'C in amuffie furnace. The furnace was cooled in successivesteps to 700"C and then shut off to cool. TheBaSO+ in the crucible had recrystallized to produceshardlike grains with numerous inclusions of NaCl.The platinum crucible also showed considerableetching and pitting.

The precipitation procedure used for the syn-thesis of sulfate minerals was used for precipitatingthe carbonate minerals witherite, BaCO3, and stron-tianite, SrCOg. Dilute (0.02 molar) BaClz and SrCIzchloride solutions that were 0.001 molar in HCIwere heated to 80-85'C on a hot plate. A 0.02

molar NaHCO3 solution was slowly added at arate between 5 and 10 ml/hr until precipitationappeared complete. The precipitate was entirelycrystalline with grains between 0.001 mm and 0.30mm in length. The crystals were generally bladelike with a width between l/4 and I/20 of theirlength. X-ray diftraction and optical parameterswere identical to those obtained for natural witheriteand strontianite. Several variations in the experi-mental procedure were employed in an effort toimprove the quality of the precipitate. Reducingconvection in the flask by placing the flask in ashallow oil bath resulted in some improvement.Further, surface-grown crystals were eliminated byredesigning the flask (Fig. lb) so that the flaskcould be completely filled with solution. Rather largecrystals (Fig. 7) up to about L mm grew at the tipof the inlet tube near the bottom of the flask. Theexcess solution escaped frorn the flask through anexit in the top of the cap piece.

Discussion

The simple precipitation procedure described inthis report has been used many times for the syn-thesis of crystalline barite, celestite, and anglesite.Precipitates of these minerals (Figs. 2, 3, 4, and 5)generally consisted of many individual crystals ofvarious sizes with a distinctive external morphology.Small crystal aggregates were also common. The

Frc. 3. Scanning electron microscope photo of celestite crys-tal. Bar is 0.5 mm lone.

CHARLES W. BLOUNT

success of the procedure is due to establishing con-ditions that allowed slow growth to occur (from0.01 to 0.1 mm/day) on a minimum number ofcrystals. No fine-grained (less than 0.002 mm) oroptically amorphous material was observed in ex-perimental syntheses carried out over a broad rangeof chemical conditions. Very fine-grained precipitateswere obtained when solution concentrations signifi-cantly exceeded those listed in Table 1 (e.9., bariLeor celestite precipitated from 0.1 molar solutions) orwhen other experimental conditions that degradedthe precipitate-e.g., stirring, use of plastic ware,poor cleaning, excessive combination rates-wereused.

The perfection of individual crystals varied some-what. The sulfate crystals grew on the glass surfacein the solution. The crystal faces next to the glasssurface were less well developed. Some crystalsshowed twinning, satellite crystals, and surfaceirregularities due to uneven growth. The larger crys-tals were generally more imperfect in their externalmorphology. The sulfate crystals had very small fluidinclusions that appeared to be filled. The externalmorphology, indices of refraction, and other opticalproperties were the same as for natural minerals.Interplanar spacings and peak intensities agree withvalues published in the X-ray Powder Data File.

Frc. 4. Scanning electron microscope photo (oblique view)of anglesite crystal. Bar is 0.5 mm long.

t2r5

Fro. 5, Photomicrograph of strontium barite crystal (0.1mole fraction Sr) in plain light. Bar is 0.05 mm long.

Variations in average crystal size occurred be-tween successive precipitations of the sulfate min-erals under seemingly identical conditions. The aver-age crystal size increased as the quantity of crystalsdecreased. Crystals produced in poorly cleanedflasks were not only finer grained but occurred instreaks, suggesting that the presence of foreign mat-ter on the glass surface may have aided nucleation.The presence of seed crystals did not prevent thenucleation of new crystals. Also, additional newcrystals appeared in the flask after the appearance ofthe initial group of crystals. La Mer and Dinegar(1951) and Coll ins and Leineweber (1956) showedthat the degree of supersaturation (concentration inexcess of the solubility of the solid) necessary fornucleation was strongly dependent upon the contentof foreign particulate matter (heteronuclei) thatcould serve as nucleation centers. Nucleation onheteronuclei required a supersaturation between 14and 58 times the solubility for BaSOr. Walton (1963)

and Mealor and Townshend (1966b) showed thathomogeneous nucleation of BaSO+ required a super-saturation factor near 1000. Homogeneous nuclea-tion is characterized by the rapid appearance of vastnumbers of new particles rather than nucleation onthe heteronuclei already present in the solution.Homogeneous nucleation is clearly undesirable forthe production of large crystals. Studies by Mealorand Townshend (1966a, b) suggest that true homo-

SYNTHESIS OF SULFATE AND CARBONATE MINERALS

1216

geneous precipitation would be very difficult to ob-tain. The precipitation of sulfate minerals describedin this study must be on heteronuclei. Mealor andTownshend indicated that the number of heteronucleiin aqueous solutions is near 106 particles /cc. Theformation of an estimated 106 to 107 crystals fromliter quantities of solution suggests that some heter-onuclei serve as crystallization centers much betterthan do many others. The very slow precipitationrate prevented the formation of crystals on most ofthe heteronuclei.

Nancollas (1968) showed that crystal growthrates depended upon the crystal surface area andthe degree of supersaturation. When only a relativelysmall number of large crystals are present, the sur-face area is greatly reduced compared to an equiva-lent mass of very small crystals. Variations in addi-tion rates would then cause larger variations ingrowth rates and the degree of supersaturation. Toorapid an addition rate could increase the level ofsupersaturation sufficiently to cause nucleation totake place on additional heteronuclei in spite of thepresence of a solid precipitate. Growth rates of crys-tals would greatly increase under such conditions butmorphological perfection would decrease. The useof persulfate-thiosulfate solutions to generate sulfateion in solutions was not used because a much sloweraddition rate was desired than such a reaction pro-duces. Most persulfate-thiosulfate precipitations werecompleted within 1-2 hours (e.9., Mealor andTownshend, I966a, b) with a significantly smallercrystal size. Slow precipitations were completed onlyafter 3 to 20 days time. The preparation of largecrystals of sulfate requires clean equipment and care-ful control of additional rates. The preparation ofultra-clean solutions with a low heteronuclei contentseems to be unnecessary. If the larger heteronucleiare more favorable nucleation sites for the sulfateminerals, their removal could be detrimental.Smaller heteronuclei probably Vastly outnumberlarger ones. Nucleation on the smaller ones couldresult in the formation of larger numbers of crystals.

There is a considerable difference in maximumcrystal sizes (Table 1). The order of increasing sizeagrees with the order of increasing solubility of theminerals reported by Blount and Dickson (1968),Stri. ibel (1968) and Blount (197I). The solubil ityof these minerals in HCI solutions exceeds that inwater alone and appears to increase with increasingtemperature. Most of the crystals precipitated in thecooler oart of the flask. A roush indication of rela-

tive solubility is provided by the amount of precipi-tating solution that must be added before precipitatedcrystals are observed in the flask. Barite precipitationrequired less NazSO+ solution than celestite whichin turn required less than anglesite. Solubility ofanglesite in HCI solutions was markedly greater inhot than in cold solutions. A considerable amountof fine grained PbSO+ was precipitated from filteredhot residual solutions after these solutions hadcooled to room temperature. Similar precipitateswere not observed in cooled solutions obtained frombarite or celestite synthesis experiments. It appearedthat carrying out the precipitation under conditionsthat enhanced the solubility of the sulfate mineralsaided in the formation of larger crystals.

Most strontium-bearing barite crystals had thesame general morphology as the pure barite or purecelestite crystals. At rather high strontium sulfatecontents (Sr/Ba = 0. 1 or greater), many crystalshad a distorted barrel-shaped appearance (Fig. 5).These crystals showed extinction only along narrowbands that become constricted toward the center ofthe crystal. X-ray difiraction peaks for strontiumbarites were intermediate in position between thoseof barite and celestite. Slight broadening of the 002and 111 difiraction peaks was observed as reportedpreviously by Renault and Brower (1971). Thisbroadening increased with increasing strontium con-tent. Examination of these crystals using the electronmicroprobe suggested that the earliest formed por-tion of the crystal had a higher barium content. Re-sults of spot analyses along the rim (last precipitatedmaterial) indicated that it was uniform to withini5 percent, the precision of the measurements.

A comparison between some solution and solidcompositions for strontium barites (Fig. 6) indi-cates that a very high Sr concentration is requiredin solutions to produce appreciable coprecipitationof Sr in barites. Sr/Ba ratios in the solid vary be-tween 1/30 to 1/80 of the Sr/Ba ratio in the solu-tion. The scatter in the data is due to compositionalvariations in the experimental solutions. The moleratios of Sr/Ba in the solution and in the solid arerelated by the equation,

Sr** . srSooBu**

: n Buso,

in which,t is a distribution coefficient. Starke (1964)showed that r varied with temperature, rate ofgrowth, and solution composition. Hanor (1968)considered that a value of ). : 33 would fit his dis-

CHARLES W. BLOUNT

tribution model for the Sr-Ba sulfate solid solutionseries; however, values as high as 66 would also bepossible. Most of the Sr barites synthesized in thisstudy lie between the lines representing r = 33 andtr : 66. Brower and Renault (I97I) considered thatthe Sr-Ba sulfate solid solution series was ideal inbehavior, so that ,\ is equal to the ratio of thethermodynamic solubility products (KspSrSO+/KspBaSO+) of pure Sr or Ba sulfate. K'*o valueswere estimated at 95oC from celestite and baritesolubility data (Blount and Dickson, 1968; Blount,I97I) . The ratio of the solubility products obtainedfrom these previous solubility studies is 390. A ,1,value this large does not fit the results obtained dur-ing the present study, as shown in Figure 6.

The sulfate crystals produced by rapid precipita-tion or flux fusion were definitely less satisfactory.Possibly much better crystals could be made bythese methods if the proper type of equipment wasavailable and the procedures used by previous work-ers closely followed. The recrystallization of sulfateminerals along a thermal gradient, while possible, isquite slow, and the crystals produced are muchsmaller than those precipitated from solution. This

1217

Frc. 7. Scanning electron microscope photo of witheritecrystals. Bar is 0.1 mm long.

method was not thoroughly investigated becausesynthesis by combining solutions proved to be re-liable, faster, and productive of larger crystals. Baritecrystals prepared by the procedure described byBurkhard (1.973) could be satisfactory for manytypes of studies. However, most crystals contain fluidinclusions that are filled with the solution in whichthey grew. Recrystallization would, in time, causeleakage from these inclusions and alter the composi-tion of experimental solutions used for solubility andphase equilibria studies. The leakage from fluid in-clusions filled with concentrated sulfuric acid wouldhave a significantly greater effect than leakage fromfluid inclusions filled with a 0.01 molar aqueous saltsolution.

Witherite and strontianite crystals grew in a dis-tinctly different manner than the sulfate minerals.Many of the crystals grew on the surface of the solu-tion in contact with the COz rich vapor phase. Alarge amount of the precipitate consisted of thesesmall (1-10 micron) crystals. When the surface-grown crystals were eliminated, most of the precipi-tation took place at the end of the glass tube usedto introduce the NaHCO3 solutions. Very little pre-cipitation took place on the bottom or sides of theflask. Some carbonate precipitated from the overflowsolutions after they stood for several days. Whilesome crystals grew to lengths in excess of 1 mm,

SYN?IIESIS OF SULFATE AND CARBONATE MINERALS

t 8

7

X'= 39O

/

/ \ '=66

//

/

,/./ r . /

./ r .{/a .ru

//

/ X=33

i:t'

l 4

t a

co=oac

++o(D+\+!a.9o

E

oo

oo5 0 . r o r5 0 .2 025 03

Mo le Ro t i o S rSOa /BoSO4 in bo r i t es

FIo. 6. Comparison of Sr/Ba mole ratios in strontium-barites with mole ratios in aqueous solutions.

r2t8 CHARLES W. BLOUNT

several percent of the crystals were less than 2microns. No amorphous material was observed. Itseemed that the carbonate crystals were more easilynucleated than were crystals of sulfate. As for sulfatecrystals, variations in addition rates may result inthe continued nucleation of crystals. While the prod-uct was satisfactory for experimental work on thealteration of sulfate and carbonate minerals (Blountand Beckman, 7974), further improvement in thequality of the carbonate precipitates may be possibleby using solutions more dilute than 0.02 molar andby using a more reliable solution-addition method.

The use of the results of this study to suggestformation conditions for barium and_strontium sul-fate and carbonate minerals in nature is subject toquestion. Some possible restrictions with respect tothe formation of coarse-grained hydrothermal baritesare worth discussing. Laboratory synthesis experi-ments were made using clean equiprnent at stronglyacidic conditions. Natural hydrothermal solutionswere probably much less acidic and the environmentcertainly was not laboratory-clean. The $owth oflarge crystals suggests that nucleation rates were verylow. This also requires low levels of supersaturation.Growth rates then would have been very muchslower than those observed under laboratorv condi-tions.

AcknowledgmentsThis research was supported by the Earth Sciences Sec-

tion, National Science Foundation, NSF Grant GA 1667.The assistance of M. Guill is gratefully acknowledged formuch of the sample processing. Mr. Baskin made micro-probe analysis of the Sr barite crystals.

ReferencesBtouNr, CHenrEs W. (1971) Barite solubility in water at

temperatures to 250"C and pressures to 1000 bars. Geol.Soc. Am. Abstr. Progr.3, 508.

eNr K. L. BncxulN (1974) Barite-witherite and cel-estite-strontianite equilibria in NaoSO,-NaHCOrNaCOrH,O solutions. Trans. Am. Geophys. (Jnion, 55, 456-457.

lNo F. W. DrcxsoN (1968) The'solubility of celes-tite (SrSO.) in H,O from 50' to 250'C and 10O to 1500bars. Geol. Soc. Am. Spec. Pap.101, 19.

BoceN, E. J., rNo H. V. Moyrn (1956) Effecr of aging solu-tions of barium chloride on particle size of barium sulfate.AnaI. C hem. 28, 47 3-47 6.

Bnrcxrn, C. E., euo D. J. Myrns (1969) The effects of ironand thorium ions on the kinetics of precipitation of BaSO,.U. S. Natl. Tech. Inform. Sero. P B Rep. 1969, No.190570, 35 p.

BRowER, Errsn (1973) Synthesis of barite, celestite andbarium-strontium sulfate solid solution crystals. Geochim.C osmochim. Acta, 37, I 55-1 58.

eNo J. Rr,Nlurr (1971\ Solubility and enthalpy of

the barium-strontium sulfate solid solution series. NewMexico Bur. Mines Mineral Resour. Circ. 116,2l p.

Bunrnenn, A. (19q3) Optische und r

SYNTTIES/S OF SULFATE AND CARBONATE MINERALS l2r9

(1958) The kinetics of crystal growth in barium in the barium sulfate-strontium sulfate series. Am. Min-sulfate precipitation. Acto. Chem. Scand. 12, 951-958. eral.56,1481-1485.

O'Rounrn, J. D., eNo R. A. IonNsoN (1955) Kinetics and Srems, R. (1964) Die Strontiumgehalte der Baryt. Frei-mechanism in formation of sligbtly soluble ionic pre- berg.Forsch.Cl50.cipitates. Anal.Chem.27,l699-1704. SrR,ijBEL, GuNrtn (1968) Hydrothermale Lcisungen, Ex-

Perer, A. R., lNo H. L. Bner (1971) Growth of barite perimentelle Untersuchungsergebnisse iiber hydrothermal-group crystals by the flux evaporation method. I. Crystal synthetische Iiisungen bis 600'C and 2000 Bar. GeoLGrowth, ll,166-170. Rundsch.5t,259-273.

AND - (1972) Growth of single crystals of TuurleroN, C. C. (1960) Solubility of barium sulfate inBaSOn and SrSOn from gels. l. Crystal Growth, 12, 288- sodium chloride solutions from 25' to 95'C. J. Chem.29O. Eng. Data,5,5l4-516.

exo J. Kosny (1968a) Growth of barium sulphate WerroN, A. G. (1963) Nucleation and the interfacial ten.single crystals by chemically reacted flux method. J. sion of sparingly soluble salts. Mikrochim. Acta, 1963,Crystal Growth,2, 128-130. 422-430.

AND _ (1968b) Etching of synthetic barite(BaSO.) single crystals. I. AppI. Crystallo,gr. l, 172-175. Manuscript receiued,October 30, 1972; accepted

Rrweur:r, J., rND E. BRowER ( 1971) X-ray line broadening f or publication, May 28, 1974.