294 MINERALOGICAL NOTES

Rnnnnnwcns Bowlrer, J. F., axn W. LoorrNc (1969) The PensaukenFormation-A Pleisto- cene Fluvial Deposition in New Jersey. 1z S. Subitsky (ed.), GeologA ol Se- Iected, Areas in New JerseE and E,astem Penrngluania. Rutgers University Press, New Brunswick, N. J. Cmozz4 A. V. (1960) Microscopic Sedimentarg Petrography. John Wiley and Sons, New York. Dnnn, W. A., R. A. flowm, .cNn J. ZussMeN (1962) Roclt-Forming M'i'nerals, VoI.6. John Wiley and Sons.New York. Dn-Wnrssn, G. (1964) Bauxite lateritique et bauxit karstique. In M. Karsulin (ed.), Symposium sur les Bauuites, Zagreb. Gannnon, L. R. (1970) A chemical model for the origin of gibbsite from . Amer. Minerar. 55, 1880-1889. Knr,r,nn, W. D. (1958) Argillation and direct bauxitization in terms of concen- tration of hydrogen and metal cations at surface of hydrolizing silicates, Bzll. Amer. Ass.P etrol. Geol. 42. 222-245. LolrrNc, W. (1961) Gibbsite vermiforms in the Pensauken formation of New Jersey.Amer. MinenaL 46,394401. Rocnns, A. F., eNn P. F. Krnn (1933) Thin Section Mineralogy. McGraw-Ilill, New York.

American Mineralogist Vol. 57, pp. 294-300 (7972)

CONDITIONSFOR DIRECT FORMATIONOF GIBBSITE FROM K-FELDSPAR-FURTHERDISCUSSION

Lnowlnr Rosnnr GennNon,Department of GeologE,[Jntuersitg of South Carolina.Columbia. South CaroLina

Assrnacr

Theoretical analysis indicates that direct alteration of K-feldspar to gibbsite in the zone of weathering probably occurs only under conditions of good soil drainage and only if the pH is above 4.3. The analysis also indicates that gibb- site should not form if the soil water has an initial dissolved silica, concentra- tion above 10-'o moles/liter. The calculations further show that, in the pH range 2 to 8, K- should not form as an initial precipitate in the hydrolysis of K- feldspar but must always be preceded by the precipitation of some kaolinite and/or gibbsite. This result may partly explain the rather infrequent occurrence of K-mica as a direct alteration product of K-feldspar. MINERALOGICAL NOTES 295

INrnolucttow

The occurrence of delicate gibbsite vermiforms and pseudomorphs after K-feldspar in the Pensauken Formation of New Jersey indicates that under suitable conditions it is possible for K-feldspar to alter

or K-mica rather than gibbsite. An alternative model

concentration of dissolved silica is greater than 10-n'6moles/liter. An attempt will now be made to show that the same conditions also apply to the direct formation of gibbsite from K-feldspar. The calculations that follow will employ the thermodynamic data and equilibrium con- stants shown in Tables 1 and 2 of Gardner (1970).

pH LrurreuoNs

As for the pH limitations let us consider the congruent dissolution of K-feldspar in solutions which have no initial dissolved potassium, m6 MINERALOGICAL NOTES

TABLE I

Calculatedl Activities of Speeies and Appanent SolubiJ-ity

Proilucts of Intersection of K-Feldspar Congruent Dissolution

Line andLGibbsite Stabitity Boundany.

pA!+3 pK+ PTI DHr'Si0" PK.. PKsp Bk-P.. Ka6Iinite K-mica r\-rerqspar

8r0 16.I 7.60 7.L2 82.34 - 2.7+ l_3.02

7.0 13.l 8.60 8.12 84. 4rt 2.35 18.06 6.0 10.1 8.95 8.47 85.I+ 4.66 20.46 5.0 7.L 6.80 6.32 80.84 - 2.9+ 12.86 4.5 5.6 5.48 5 .00 78.20 - 7.72 8.08

4.0 4.I 4. 06 3 .58 75.36 -12.90 2.90

3.0 I.l t.ro 0. 62 69.42 -23.76 -7.9t+

3 .04 I.I 1.IO 2.60 73.40 -r7.80 - 2.00

a' Recalculated, assuming saturation with respect to arnor:phous

silica, pHqSi04 = 2.60. aluminum,or silica. Thus for the congruentdissolution of K-feldspar (KALSiBOs)under theseconditions we can write:

)dissolvedSiO2:3 X >dissolvedAl: 3 X concentrationofK*. (1) Let us next assumethat at any pH the first stability boundary inter- sectedby the congruentdissolution line is that of gibbsite.As indicated above,the calculationsof Helgeson,et aI. (1969) show that for dis- tilled water with an initial pH of 7 this is actually the case.Assuming that activities of aqueousspecies equal their concentrationswe can calculatethe activities of aqueousspecies at the intersectionof the congruent dissolution line and the gibbsite stability boundary at any MINERALOGICAL NOTES

TABLE 2

Calculated Activities of Species and.Appanent Solubility

Productb at Intersection of K-FeIdspar Congruent Dissolution

Line ancl Kaolinite Stability Boundany.

pH pAf,+3 pK+ PHr'Si0r' PKSp pKsp pKsp Gibbsite K-mica K-feldspar

8.0 14.80 6.30 5.82 32.80 -rr.54 6.56

7.O II .3I 6.81 6.33 32.31 -L0.27 15.II

6:0 8.I4 6.99 6.51 32.I+ - 9.06 10.66

5.0 6. 2I s.9I q lr ? 33.2I - 9.L7 8.41

4.5 5.37 s.25 33.87 - 9.33 6. 93

4.0 4.57 4. 53 t+.05 34.57 - 9.61 5.25

3.0 3.06 3.06 36.06 -10.02 I.8 6

2,0 I.56 1.56 1.08 37.56 -r0.5 2 -1. 64 2.04 0.05 0 .05 2. 60 36.05 -I2.00 -0.r0

a' Recalculated assuming saturation wi.th respect to amorphous

silica, pHOSiOO= 2.60. pH, given the solubility product of gibbsite (K*o gibbsite = lFsa:l). For exampleat pH 5 , rAr.,l:41#$&: $#: 1o-, (z) and by Equation1,

lK.l: Ier: lAl.'l(r + K,[oH-]+ K,K,[oH-]n) o :10-'.'(1 + 10n'ox lO-n + 10n'ox 10"'ux l0-tu (B)

: 10-6'8 298 MINERALOGICAL NOTES where K1 an Kz are the equilibrium constants given in Table 2 in Gardner (1970) for the reactions Al(oH)'. -- AI'* + oH- (4) and Al(oH).- --+Al(oH)'* + 3 OH- (5) Thus by Equation 1, - [H*SiO4]:3 X >Al:3 X 10-u8 10-6'32 (6) The solubility product given by Gardner (1970) for kaolinite is - K,, kaolinite : [A1.3]'[HnSion]'[oH-]' 10-77'3 (7) At the intersectionof the gibbsitestability boundary and the congru- ent dissolutionline for K-feldspar at pH 5 the apparent solubility product for kaolinite is given by,

K", kaolinite apparent - : [Al*']'lHnsion]'[oH-]' l0-8o'8410-77'3 (8) Sinceat pH 5 the apparent solubility product for kaolinite is much smallerthan the actual solubility product this indicatesthat at pH 5 the congruentdissolution line of K-feldspar intersectsthe gibbsite stability boundary,as assumed,before reaching the kaolinite stability field. Thus at pH 5 the hydrolysis of K-feldspar is behaving in a mannersimilar to its behaviorat pH 7; lhat is, gibbsiteappearc as the initial precipitate and may later dissolve.The results for the above seriesof calculationsat other pH values are summarizedin Table l. From the data of Table t it can be seenthat below a pH of about 4.3 the apparentsolubility product of kaolinite exceedsthe actual solu- bility product. This meansthat below pH 4.3 the congruentdissolu- tion line of K-feldspar must intersectthe kaolinite stability boundary beforereaching the stability boundaryof gibbsite.Thus we shouldnot expectdirect formation of gibbsitebelow a pH of about 4.3. AIso shown in Table 1 are the calculatedapparent solubility pro- ducts for K-mica and K-feldspar.Based on free energiesof formation of -892.817 kcal and -1330.1 kcal for K-feldspar and K-mica (Robie and Waldbaum, 1968) respectively,the actual solubility productsare

K", K-mica: I4ilt#r#dqd - 10,82 (e) MINERALOGICAL NOTES and

(10)

It should be noted in Table 1 that at pH 3 the apparent solubility productsof both K-mica and K-feldspar also exceedtheir actual solu- bility products.In order to clarify the meaning of these results the apparentsolubility productsfor gibbsite,K-mica, and K-feldspar have beencalculated using the procedureemployed in Table 1 but with the assumptionthat the congruentdissolution line for K-feldspar first intersects the kaolinite stability boundary rather than the gibbsite stability boundary.The resultsof the calculations,shown in Table 2, confirmthe gibbsite-kaolinitetransition of pH 4.3 arrived at in Table 1. Also, as can be seenfrom Table 2, the apparentsolubility product for K-mica doesnot exceedits actual solubility product in the pH range2 to 8. This suggeststhat K-mica forms as an alterationproduct only when the soil solution closelyapproaches final equilibrium with K-feldspar and after precipitation of some kaolinite. The fact that K-mica doesnot appearas an initial precipitatein this pH rangemay explain its relatively rare occurrenceas a direct alteration product of K-feldspar.

Ernncr oF INrrIAr, Drssor,vppSrr,rce stability relationships in the K-feldspar system depend upon four independentvariables: pH, pK., E dissolvedsilica, and I dissolvedAl. Thus a four dimensionaldiagram would be requiredto showthese relationships. Regardless of how one may chooseto repre- sent this system,the stability boundariesfor quartz, kaolinite, and gibbsitewill be independentof pK-. Thereforethe intersectionof the gibbsite and kaolinite stability fields, which occurs at [HrSiO+] = 10+'6moles/liter (Figure3 in Gardner,1970), will alsobe independent of pK-. Thereforeany solution approachingequilibrium with K-feld- spar with an initial silica concentrationgreater than 10-a'6moles/liter will not be able to intersect the gibbsite stability boundary without fi.rstencountering either the kaolinite, K-mica, or K-feldspar stability boundary.Thus direct gibbsite formation should not be expectedat soil depthsbelow which dissolvedsilica exceeds10-a'6 moles/liter. Also the occurrenceof authigenicgibbsite in soil that containsquartz or other primary silicatesis probably an indication that the kinetics of dissolution of the primary silicattjs is very slow comparedto the kinetics of feldspardissolution and the rate of soil water drainage. 300 MINENALOGICAL NOTES

Suulreny er.roCoNcr,usroNs The resultsof this theoreticalanalysis indicate that direct gibbsite formation probably occursunder conditionsof high soil permeability and good soil drainage.Also gibbsite formation probably does not occur where the soil pH is less than about 4.3 or where the initial concentrationof dissolvedsilica exceeds10-n.6 moles/liter. The analysis further indicatesthat alteration of K-feldspar to K-mica probably doesnot occurwithout antecedantproduction of kaolinite. In general, the analysisdemonstrates that the formation of clay mineralsin the zoneof weatheringinvolves the complexinterplay of parent material, pH, rates of soil drainage,kinetics of dissolutionof primary silicates and precipitationof clays,initial compositionof the soil solution,and biologic factors. There is probably no simple universal sequenceof alteration products in the zone of weathering,the exact sequenceof formation and relative proportionsof clays in a particular situation beingdetermined by the abovefactors.

RnrnnpNcns Gennxnn, L. R. (1970) A chernical model for the origin of gibbsite from kaolinite. Arner. M'inerar. 55, 1880-1889. Ifnr,cnsox, I1. C., R. M. Gennnr,s,exl F. T. MecnnNzrn (1969) Evaluation of irreversible reactions in geochemical processesinvolving and aqueous solutions-*Il. Applications. Geocbi.m.C osmochim. Acta 33, 45H81. Lotoruc, W. (1961) Gibbsite vermiforms in the Pensauken tr'ormation of New J ersey. Am,er. M'ineral. 46, 3g+-40I. Ronro, R. A., er.rn D. R. Wer,osAulr (1968) Thermodynamic properties of min- erals and related substancesat, 2gE.lb.K (2S.0"C) and one atmosphere (1.018 bars) pressure and at higher temperaturas. tl. S. Geol. Suru. BuIl, tzlg.

American Mineralogist Vol. 57, pp. 300 316 (1972)

STABILITYOF BIOTITE: A DISCUSSION Rosunr F. Mupr,r,en,National Aeronautics and SpaceAdmin;istration, GoddardSpace Fkght Center,Greenbelt, Margland 20771

Assrnect

Distribution data for magnesium and iron from a variety of ferro-magnesium silicates and coexisting biotites indicate that the mixing properties of these constituents in biotite and the other silicates do not deviate greatly from the ideal solution (activity coefficients within a factor or two of unity). This result is in conflict with some inferences which have been drawn from exferimental data on the phlogopite-annite system. Examination of these experimental data