Amcricat Misesalookt

Vol. 57, pp. rg23-1631 0972\

CATTONEXCHANGE SELECTTVITY IN MICA AND

VERMICULITE

D. L. Dor,cersn,t
M.L. Jecrsox,ANDJ. K. Svens,
Departmento'f SoiIScience,Uniuersityof Wisconsin,
M'adison, Wi saonsin59706

ABsrBAcr
Layer charge density differences between potassium saturated, dried micas
and vermiculite resulted in a cation exchange selectivity (cES) for potassium
in biotite and mwcovite (1.7 and 4.0, expressedas equivalent ratios,KX/cavJ()
which were 4 and 10 times, respectively, the CES in vermiculite (0.4). Interlayer
wedge sites created by the potassium saturation and drying treatment (which
made the cation exchange capacity nearly the same for the three materials,
8 to 10 meq,/100 g) increased the CES both of the micas and of vermiculite
by a factor of two, relative to that in the undried vermiculite, biotite, and
muscovite (0.2,0.9, and 1.7, respectively). In addition to effects on CES of
charge of the layers making up the crystal wedges and external (001) cleavage
faces, the obliqueness of the hydroryls in dioctahedral layers contributed con-
siderably to the range in cES for the potassiun relative to calcium on the
extemal sites of vermiculite, biotite, and muscovite (0.0, l.B, and B.Z, respec-
tively).

INrnouuctroN
The selectiveexchangeof lyotropic seriescations by layer silicate
minerals is controlled by the mineral structural configurations and
layer charge densities in addition to the well-known effects of the
cation propertiessuchas valence,size,and degreeof hydration. struc-
tural properbiesof layer silicate minerals which give rise to difrerences
in cation exchangeselectivity (cES) appearto arise from the presence
of interlayer wedgesand the charge density of the layers making up
the wedges(Dolcater et at.,lg68). Suchwedgesoccur (Jackson,1g68;
1968) at the boundary betweenopened (weathered) and unopened
(unweathered)portions of micaceousvermiculite formed by weather-
ing.
This paper showsthat the chargedensity of layers on the (001)
cleavages,in addition to the presenceof interlayer wedgesand the
chargedensity of layers making up thg wedgespresentin micaceous
vermiculite, biotite, and muscovite, affects the selectiveadsorption of
potassiumrelativeto calcium.

lPresent address: Amoco Production

Company, P.O, Box d01, Tulsa, Okla-
homa 74102.
1824 DOLCATER, JACKSON.AND SYERS

Expnnrrupr.trer,
Vermiculite (size fraction (5rrm, biotitic, from Transvaal, South Africa),
biotite, and muscovite (both of size fraQtion 242pm, from Ward's Natural
science Establishment, Rochester, N.Y.) were separated by sedimentation fol-
lowing treatment of a water suspension of large flakes in a Waring blender.
Each fine material was divided into two portions, one of which was sodium
saturated and left undried, while the other was potassium saturated and dried
at 110"c. samples of both portions were then washed with a solution 0.005 N
with respect to both KCI and cacL. The cES was determined as the equivalent
ratio of exchangeable potassium to calcium (KX/Caat2X, in which x represents
one charge of the cation exchanger anion). The potassium and calcium selec-
tivity values were determined and calculated according to the procedure given
by Dolcater e, al. (1968).

Rnsur,rs
The potassium saturated and dried vermiculite, biotite, and mus-
covite gave K/Ca equivalentratios (Fig. 1) of 0'4, L-7, and'4'0, re-
spectively,which were double those for the sodium saturated and
undried materials (0.2, 0.9, and 1.7' respectively).These ratios were

I80
E CEC
N CES
Elt
I40 x F\S

(g
o (J
o
^><
a tnn
v;
l+l
L)
lrl
c) 60

20
0 K, K, K,
Na!ndried *, N a l n d r i e *d, N a ! n d r ie d* ,
u n d r ie d d r i e d undrei d dri ed undried dried
Vermiculite Biotite M u s c o vi t e

ftc. 1. CEC and.K/Ca CES for sodium saturated (undried), potassium satu-
rated (undried), and potassium saturated then dried vermiculite, biotite, and
mrucovite.
 C,ES. IN MICA AND VERMICULITE 7825

inversely related to the cation exchangecapacity (CEC) of each

material (Fig. 1). This relationshipsuggeststha:bthe interplanar sites
of vermiculite are more selectivefor calcium relative to potassium
than the e>ichange sites of the micas. The doub ing of the selectivity
values as a result of potassium fixaiion was attributed (Dolcater
et aI.,1968) to an increasein the proportion of interlayer wedgesites
to interplanar sites. After potassium saturation and drying, which
reducedthe CEC of the three materialsto essentiallythe samevalue
(8 to 10 mey'100 g, Table 1), muscoviteand biotite had CES values
10 and 4 times as great, respectively,as that, of vermiculite. These
higher selectivity values for the micas comparedto vermiculite point
to the importanceof the layer chargedensity of not only the layers
making up the wedges,but also that of layers exposedon external
cleavagefaces,in the selectiveadsorptionof small, unhydrated ca-
tions such as potassium.The three-fold differencein CES between
the dioctahedraland trioctahedral micas, which have nearly equal
charge densities,is attributable to the orientation of the hydroxyls
(Serratosaand Bradley, 1958) below the ditrigonal holesin the silica
sheets.

DrscussroN
The relative contributionof the externalplanar surfacechargeden-
sity effectsand the interlayer wedgeeffectswere evaluatedby con-
siderationof the followingcases:
Case 1. The value of a-b (Table l) was assumedto be an estimate
of the quantity of interlayer wedgesites,which were oc-
cupiedby potassium.
Case 2. A value of 3.6 mey'100 g (van Olphen, 1966) was taken
as an estim,ateof the quantity of external planar surface
. sites.
Case 3. A value of 5.0 meq/100 g (Alexiadesand Jackson,1965)
was taken as an estimateof the quantity of externalplanar
surfacesites.
Exchangeable potassiumwas assignedfirst to the wedgesiteswhereas
exchangeable calcium was assignedflrst to the external planar sites.
The excessof either ion was secondarilyassignedto the alternativesite.
Many other possibleassumptionsbesidescasesl, 2, and 3 were con-
sidered,but had to be rejectedon the basis of the experimentaldata
in Figure 1 and in Table 1.
Calculationsmade for each of thesethree casesshow the distribu-
tion of exchangeable potassiumand calcium on the external planar
1826 DOLCATER,
JACKSON,AND SYERS

surfacesand in the interlayer wedges (Table 2). In the first case,

calcium in the equilibration solution apparently causeda small amount
of openingof wedgesin the potassiumsaturated materials, thereby
exposingadditional cation exchangesites (a-b in Table 1). If these
sites are largely filled with potassium,and considerableevidencesug-
gests they are selectivefor potassium (Jackson, 19631Boll et al.,
1963; Rich and Black, 1964; Scott and Bates, 1967; Dolcaler et aI.,
1968),the selectivityvalue of the wedgesiteswould approachinfinity.
The selectivity values for the surfacesites of biotite and muscovite
are then 3 and 9 times as great, respectively,as the overall value for
vermiculite (Table 2). Thus, the surfacesof the micas show a high
degreeof selectivity for potassium,which can be explainedsatisfac-
torily only as a result of their higher external surface chargedensities
than that of vermiculite (Fig. 2).
The K selectivity difference between biotite and muscovite is ex-
plained primarily by the oblique orientation of the hydroxyl groups
in muscovite(Serratosaand Bradley, 1958),giving a greaterattraction
for K by muscovite.In addition, decreasein layer chargethrough such
ionic reactions as iron valence changeor substitution of other cations
andfor hydrogen (Jackson, 1963; 1968) produces the local areas
shownwith no net negativecharge (Fig. 2).1 The remainingareasof
high chargeare relatively selectivefor potassiumand such areashave
a frequency: muscovite ) biotite ) vermiculite.
In case2, the interpretations are similar, exceptthat the selectivity
of the surface sites of the micas is somewhatlower becausemore
than 50 percent of the CEC is attributable to wedgesites for each of
the three materials (Table 2). Such a high percentageof wedgesites
would be unlikely (especiallyin micas)I however,the exchangesites
on the (001) cleavageof biotite and muscovite still show a high
selectivity for potassium (Table 2) relative to those of vermiculite,
an efect attributed to the higher charge density of the mica layers.
The indicatedpotassiumselectivityof the siteson the (001) cleavages
rThe charge density of biotite (FiS. 2) has been taken as being lower than
that of muscovite because of the lower K/EC (column b, Table 1). The layer
charge densities (rn-') and (/EC give specific surface areas of 25 and t2 mn/g
for these 24.2 pm fractions of muscovite and biotite, respectively. Taking any
higher charge density for biotite would further lower the calculated specific
surface of bioiite, an unreasonable premise. Lowering the biotite layer charge by
a 10 percent deficiency in K+ (through compensating trivalent octahedral ions)
and 5 percent by surface weathering reactions is assumed to be a plausible basis
for the 3.0 meq,/lffm'? layer charge represented in Figure 2. The specific surface
for vermiculite (2.4 meq/lfm2 and 6.9 meq,/lffS) is 29 me/g, suggesting cleavage
into much thinner particles in (5 pm fraction employed.
 C"ES.IN MICA AND VERMICULITE L827

TABLE 1. CATION EXC}IANGE CAPACITY OF VERMICI]LITE, BIOTITE,

AND MUSCOVITE AND SITES EIGOSED BY Ca PRESEM IN THE MIXED

K + Ca EOUILIBRATION SOLUTION

Cation exchange capacity Portion of

cEC on
cEc K/ECa
Material (Kx+Ca!X) (K,AHUoAo) sites exposed
bY ca
K, dry *ig*"
(a) (b) (a-b)

meq,/100 g

Vemiculite q.5 6.9 2.6

Biotite 7 .7 6.5 t.2

Muscovite 9 .7 9. 1 0.6

'Washed with 0.5 .lf KOAc and 0.01 .l'I KCI to clean:
weighed and dried in 0Ol .lf KCI; K displaced in 0.5 N
NH.OAc. K/EC (as distinct from KEC measured with-
out drying) indicates that the K+aturated sample was
dried at 110"C.

TABI,E 2. EXCHANGEABLE K AND Ca ON INTBRI,AYER WEDGE AND EXTERNA]- PLANAR

SUR.FACE SITES OF VERMICU],ITE, B]OTITE. AND MUSCOVITE IN THE MIXED K + CA

EQUILIBRATTON SOUTION

Vemiculite Biotite Muscovite

Exchantre
sitg KX Catl. Yj{/Catj. KX Carix'. r<x/Cayx Io( CaLX l<x/CaLX
-
Location _ _
(meq,/loo g) (neq,/100 g) (neq,/100 g)

Overall 2.8 6.7 0.41 4.9 2.8 t.7 7.a 1.9 4.0
Case 1. Distribution of KX and CaqX assuning that a-b (Table 1)
is an estjmate oF the quanrity of riedge sites:
Wedge 2.6 0 -- 1.2 O -- 0.64 0 --
Surface 0.L8 6.7 0.027 3.7 2.a 1.3 7.1 I.9 3.7
Case 2. Distribution of KX and Ca2X assuming that 3.6 neq,/loo g
is an estimate of the quantity of external pLanar surface sites:
lJedge 2.a 3.1 0.89 4.L 0 -- 6.1 O --
Surface 0 3.6 O O.79 2.8 0.28 1.7 I.9 0.86
Case 3. Distribution of (X and Ca2X assuning that 5.0 meq,/100 g
is an estimate of the quantity of external surface sites:
Wedge 2.a L.7 1.6 ?.7 0 -- 4-7 0 --

Surface 0 5.0 O 2.2 2.8 0-78 3.1 1-9 1.6

VERMICULITE

MOLAR RATIO

K X / C oX 2 ! O . O 6 7

E Q U I V A L E N TR A T I O ' cEs
RX/Cohlx.0.055

Oxt Oco t+

BIOTITE

MOLAR RATIO

x\ / CoX2z l.g

E O U I V A L E N T R A T I O' C E S

xX/ Cot/zx.1.4

O*. O c o* *

MU S C OTVEI

MOLAR RATIO

KX/CoX2=7.9

E O U I V A L E N T R A T I O= C E S

Kxlcor/2x.3s
O** Ocot*
 C.E.S.IN MICA AND VERMICULITE 1829

of micas is higher still in case3 than it was in case2, becausethe

percentageof the cEC attributable to the wedgesitesis smaller (Ta-
ble 2). If the values of 96 percent external surfacesites and 4 per-
cent wedgesites (Bolt et at., L963)are used,the selectivityof the sites
on the (001) cleavagesof the micas becomeseven higher.
Comcr,usroNs
The conclusionfrom this researchis that, in addition to cation
selectivity of the wedgesites,sites on (0Ol) cleavagesof micas have
an appreciablygreaterselectivitythan thoseof vermiculite for potas-
sium as compared.to calcium. The best explanation for this high
selectivity of mica relative to that of vermiculite appearsto be the
differencebetweenthe layer charge density of mica and vermiculite.
The higher potassium selectivity of the muscovite exchangesites
relative to thoseof biotite can be explainedby the smaller ditrigonal
holesin the tetrahedrallayer o,fmuscovite(Radoslovichand Norrish,
1962) and the oblique orientation of the hydroxyl g"roupsin the bottom
of the holes of muscovite (Serratosaand Bradley, 1958; Bassett,
1960), both of which result in a strongerpolar bonding (Shainberg
and Kemper, 1966) of potassiumto muscovitethan to biotite' The
strongeradsorptionof cesiumthan sodiumon the surfacecleavagesof
muscovitemacroflakeswas also explained (Mokma et aL.,1970) by
this reasoning.Any tendencyfor the development(on weathering)of
oblique orientationof the hydroxyls, favoring K selectivity (Barshad
and Kishk, 1970),in this biotite and this biotitic vermiculite is ap-
parently more than offset by the extensivelowering of the charge in
the layers.

Frc. 2. Distribution of Caa and K* ions on planar cleavage surfaces on which

each shaded side of a structural hexagon (oxygen at each apex) carries a net
CEC charge of -l/12. Layer stmctural ion reactions are assumed to have elimi-
nated the surface charge in indicated (not shaded) areas of vermiculite (2.4
meq,/lO"m'; 170 meq,/100g) and biotite (3.0 meq/lffm'; 210 meq/100 g) as com-
pared to the full theoretical charge of muscovite (3.5 meq,/lfm'?; 250 meq/100 d.
The K equivaLent to the wedge sites (a-b in Table l) has been subtracted from
KX before the varying selectivity ratios of the cleavage surfaces were derived
(approximately case 1),
1830 DOLCATE&, JACKSON,AND SYERE

When either vermiculite or mica is consideredindividually, the

interlayer wedgesites have a greater selectiveadsorptionfor potas-
sium relative to calcium than do the external planar surface sites.
When comparingthesetwo materials,however,the greaterselectivity
of the micasis attributable to the higher chargedensity of the layers
on the (001) cleavagesurfacesof the micas. Consequently,the con-
tribution of the layer charge density of minerals to the selective
adsorption of cations should be given consideration,especially in
studies concerningthe selectiveadsorption of cations in fertilizers,
radioactivedebris and wastes,and various other pollutants.

Acrnowr,nocmnwrs
This resea,rchwas mrpported in part by the College of Agricultural and Life
Sciences, University of Wisconsin, under project 1336, and in part by the Di-
vision of Biology and Medicine, Environmental SciencesBranch, United States
Atomic Energy Commission Contract AT(11-1)-1515-Jackson (paper COO-
l5l5-24). Comments on the manuscripts by Dr. W. F. Bradley and Mr. K. Sridhar
are warmly appreciated.

R"nnnnnr.rcns
Ar,nxrams, C. A., er.ro M, L. Jecrsow (1965) Quantitative determination of
vermiculite in soils. SoiI $ci. Soc.Amer. Proc.29,522-527.
Bensnen, f., .lwo tr'. M. Krsnr (1970) Factors affecting potassium fixation and
cation exchange capacities of soil vermiculite clays. CI,aAsCIaA Mi,neral. tB,
rn-87.
Bessnrt, W. A. (1960) Role of hydroxyl orientation in mica alleration. GeoI.
Soc.Amer. BuIl,. lt,449-4.E5.
Borr, G. 8., M. E. Suurvnn,exn A. Karrpsonsr (1963) A study of the equilibria
between three categories of potassium in an illitic soil. Soi,I 9ci,. Soc. Amer.
Proc.27,m4=W.
Dor,catnn.D. L., E. G. Lorsn, J. K. Srmns,aNo M. L. JacrsoN (1g68) Cation
exchange selectivity of some clay-sized minerals and soil materials. Soil
Sci,. Soc. Amer, Proc. 92, 795-798.
Je.crsox, M. L. (1963) Interlayering of expansible layer silica.tesin soils by chemi-
cal weathering. Clags Clag Mineral. tt,2s.46.
- (1968) Weathering of primary and secondary minerals in soils. Int. Soc.
Sor.lScc'.Trans.,9th.Congr. ( Adetnide),4,AjL-292.
Morue, D. L., J. K. Srnns, aNn M. L. JecrsoN (1g70) Cation exchangecapacity
and weathering of muscovite macroflakes. SoiI Sci. Soc. Amer. Proc. 34, 146-
151.
Reoostovrcn, E. W., eNl K. Nonnisu (1962) The cell dimensions and symmetry
of layer lattice silicates: f. Some structural concepts. Amer. Mineral 4T,
599-616.
Rrcu, C. I., arvn W. R. Br,ecr (1964) Potassium exchange as affected by cation
size, pI[, and mineral structure. 9o,i,lSci,.92,884-890.
Scom, A. D., arn T. E. Beros (1g67) Changes in exchangeable potassium ob-
served on drying soils after treatment with organic compounds: IL Reversion.
SoiI Sci,.Soc.Amer. Proc.3t,481-485.
 C.Eg.IN MICA AND VERMIAULITE 18:}1

Sunnamse,J. M,, ar.roW. tr'. Bneolnv (1gd8) Determination of the orientation of

o'E bond axes in layer silicates by infrared absorption. J. phys. chem. 62,
u64-1167.
SnArNanno,f., eNo W. D. Kerrrpnn (1g66) Electrostatic forces between clay and
cations as inferred from electrical conductivity. ctnys clay Mineral. L4, tlz-
t32.
Vew Or,psulv, E. (1966) Collapse of potassium montmorillonite clays upon heat-
ing-"potassium fixation." Clays CIay Mi,neral.l4, Bg}40b.

Manwcri,pt recei,ued,Februarg S, lglg; atcepted;

lor p,ublicatinn June #), 1g/2.