American Mineralogist, Volume 68, pages 818-826, 1983

Kenyaite--synthesis and properties

Kleus BeNere eun Gpnneno Lacely

Institut filr anorganische Chemie der UniversittitKiel, OlshausenstraSe 40160, 23 Kiel, Germany

Abstract

The hydrous sodium silicate kenyaite can be synthesized from aqueous suspensionscontaining SiO2 and NaOH with SiO2A.{aOH ratios ranging from 5 to 20 and HzO/NaOHratios from 50 to 500 at 100-150'C. This phase can also be prepared from concentrated ordiluted water glass solutions above 120'C. At 100'C magadiite generally precipitates as thefirst reaction product and then alters to kenyaite. The stable end product is quartz. Thisleads to the suggestion of the transformation: magadiite --+ kenyaite --+ quartz in agreementwith field observations of Eugster (1969).

Formation of kenyaite at 100'C requires several months. The reaction times are muchdecreased at higher temperatures but under these conditions quartz forms rapidly.Synthetic kenyaites form spherical aggregates of well-developed plates.

Kenyaite gives intracrystalline reactions like other layer silicates. The intracrystallinereactivity reveals some differences between the different samples which are not detected byother methods. The reactivity mainly depends on the time of synthesis and is lessinfluenced by the composition ofthe starting mixture. Through exchange ofprotons for theinterlayer sodium ions, kenyaite is transformed into a crystalline silicic acid.

Introduction

The four naturally-occurring hydrous alkali silicatespresently known were discovered during the last fifteenyears at the alkali lakes in Africa: magadiite, kenyaite(Eugster, 1967, Eugster et al. 1967) and makatite (Shep-pard et al. 1970) at Lake Magadi, Kenya, and kanemite atKanem, Lake Chad (Johan and Maglione, 1972). Thechemical compositions of the natural samples were re-ported as NaHSi2O5 . 3H2O (kanemite), Na2SiaOe. 5H2O (makatite), Na2SilaO2e .9H2O (magadiite) andNa2Si22O45 . l0H2O (kenyaite).

Synthetic magadiite forms at 100-120'C from suspen-sions typically containing I mole NaOH, 2 to 6 molesSiO2 and 10 to 60 moles H2O (Lagaly et al., 1973).Kanemite and makatite can also be easily prepared (Ben-eke and Lagaly, 1977a), but synthesis of kenyaite fromalkaline suspensions of SiO2 at lfi)'C seemed to bedifficult, supporting Eugster's suggestion that kenyaite isan alteration product of magadiite (Eugster, 1969). How-ever, when the runs were held at 100"C for longer than 90days (up to two years) kenyaite was formed.

The X-ray diffraction powder pattern and the Na2O/SiO2/H2O ratios of natural and synthetic kenyaites canvary considerably. It will be shown that the variabilityresults from intracrystalline changes brought about bycation exchange reactions and one-dimensional intracrvs-talline swelling.

Materials and methods

The synthetic kenyaites were prepared from suspen-sions of silica (silica gel: "Kieselsiiure, geflllt", Merck,Germany, or "Aerosil 200", Degussa AG., Germany) inaqueous NaOH or Na2CO3 solutions. In a typical run(sample 27Y, Table 1) 50 e silica gel (: 42 e SiO) weredispersed in 250 ml water containing 1.6 g NaOH (: l7 .5moles SiO2 and 354 moles H2O per mole NaOH). Thesuspension was allowed to stand at 100'C with occasionalshaking. After 3 months a small amount of kenyaite wasformed. Two months later a large amount of precipitateformed which was separated by filtration and then air-dried. Washing with water was usually avoided to preventthe alkali ions from being leached out of the structure.

Most of the runs were performed at 100'C in lfi), 250 or500-ml Teflon vessels with nearly gas-tight screw-on-typecaps. Reactions at 125'C were made in 500-ml steelvessels with Teflon packing rings. For the 150'C experi-ments we used a 750-ml agitated autoclave (Forschungs-institut Berghof) with complete Teflon lining.

A natural kenyaite from the Magadi area (alluvialchannel, Lake Magadi) provided by the Magadi SodaComp., Kenya, was studied for comparison.

The kenyaites were stored under diferent conditions(under mother liquid, washed, wet, air-dried, freeze-died, in vacuo and over P4O1s, heated to 60-1200'C).The X-ray powder patterns were recorded in Debye-

0003-004x/83/0708-08 I 8$02.00 81E

BENEKE AND LAGALY: KENYAITE 819

Scherrer-cameras (114.59 mm diameter, CuKa-radia-tion). The amount of impurities in the samples (magadiiteor quartz) was roughly estimated from the intensities ofthe X-ray reflections in the range 40 : 5l-57" . Calibrationwas accomplished by comparing the powder patterns tothose made from artificial mixtures of synthetic kenyaite,synthetic magadiite and quartz.

The composition of the precipitates of the diferent runswas obtained in the usual way (SiOz by gravimetric orcolorimetric analysis, sodium by atomic absorption, wa-ter from weight loss at ll00'C). It was, however, verydifficult to obtain the exact composition of the kenyaiteitself. The precipitates absorbed mother liquid whichcould not be removed without leaching sodium ions out ofthe kenyaite structure. The procedure successfully usedfor kanemite (Beneke and Lagaly, 1977a) only led to anapproximate composition SiO2/Na2O : 19-22. Manyprecipitates apparently contained amorphous silica orquartz as impurities (c/. Fig. 2b,f,g) and could not be usedfor the evaluation of the chemical composition of ken-yaite. More reliable values were obtained for the potassi-um form of kenyaite (see below). This material can easilybe prepared from suspensions of SiO2 in aqueous KOH at100-125"C (Lagaly, 1980). The precipitate shows noamorphous materials or quartz in scanning electron mi-crographs. For analysis it was transformed into thecorresponding crystalline silicic acid (see below andfootnote I, p. 823). The acid was equilibrated withpotassium hydroxide solutions of different concentra-tions. On the basis of the dissolved amounts of acid theSiO2/K2O ratio was determined to be 19.8-20.4.

Intracrystalline reactions of the kenyaites (cation ex-change and intercalation) with different types of organicmolecules were studied using X-ray powder techniques(Lagaly et aI.1975 a,b, Beneke and Lagaly, 1977a).Piorto these experiments the synthetic kenyaites were care-fully washed with small amounts of water to remove mostof the adhering mother liquid and air-dried or freeze-dried.

The H-form of kenyaite was prepared by addition oflarge excess of 0.1 m hydrochloric acid. The sampleswere allowed to stand some days (typically one week) in arefrigerator, then were filtered off and air-dried.

Results

Synthesis

Kenyaite forms from suspensions containing high SiOriNaOH ratios (>8, Tabla l). Kenyaite can be detected inthe X-ray powder patterns of the precipitates from solu-tions held at lfi)"C for 3 to 5 months.

The amount and crystallinity reach optimal valuesbetween 8 and 24 months. Longer run times transformkenyaite into quartz, the stable end product. Above 100"Cthe rate of kenyaite formation is much increased but alsoincreased is the rate of quartz formation. Preparation ofkenyaite free of quartz requires special conditions. Forexample, run 37L (Table l) yielded a relatively well-crystallized kenyaite without qtartz.

The first crystalline product formed in the 100'C-experiments is usually magadiite (Table l). This phase

Table 1 Synthesis of kenyaite from SiO2, NaOH, H2O-suspensions or water glass solutions at 100-150"C

( m o l e s )srdt-E-----r568 i n te rmed ia te

uonvers l0n tou p u r e " * * + * k e n y a i t e

c o n c e n f , r a f ,nuctea t l0n

of kenya i tedura t ion( months )

d i i t e) 1 7

t 7 r2 E M

l 9 N3 7 L2 8 D

z B E

1 0 0 0 c1 0 0 0 c1 0 0 0 c

1 0 0 0 c1 0 0 0 cl 2 q o a

1 0 0 0 c

1 0 0 0 c1 0 0 0 c

4 1 U 1 5 0 o C4 1 T 1 5 0 o C

41 . o 150ocq 1 R 1 5 o o c4 1 S 1 5 0 o CL l1 K 1500C4 1 P 1 5 0 o C

4 0 Y r z S o c4 0 x 1 z 5 o cqo r \ t r 25oc

q 2 2 21 0 9 37 7 1 0 0

18 35\2 2 2 2 2

I 25010 250

72 41 7

! )

1 01 0

2 4

q2 \

1 71 a

I

834

41 0

395

3t0 . 1

< 3

34

0 . 10 . 1

It about

) 7 2 h

o .25 K+e (<5 tr )o ,25 K+Q (<5t r )

K+Q (20 t r )KK

K + M g ( 1 5 t )R+Q G5%)KK

KK + Q ( < 5 t )

anorphous

K , Q ( t r a c e s )K+Q (<5 t r )K+Q (<5# )

r + Q ( 5 - 1 o t )K + Q ( 5 - 1 0 t )K + Q ( 5 t )

1 67 6r 61 6

L 6r 67 67 67 6

7 51777384lr 4

2502502 q n

250250

ztl h7 2 } ]9 6 h0 . 20 . 5

444

WGSWGS}JGs

( 18wttrsio^ )( 24wtfsio: )(3owt%sioi) 0 . 5

r . )f t l

K = k e n y a i t e , M g = m a g a d i i - t e , Q : q u a r t z , w e i g h t p e r c e n t s o f Q o r M g

water g lass so lu t ion

no in te rmed ia te magad i i te observed

X- ray d iagram showing re f lec t ions o f kenva i te on lv . Presence o f anorDhous s i l i ca no t exc luded

820 BENEKE AND LAGALY: KENYAITE

s -otlo. s,orti.or" ff* "

Fig. l. Syntheses from SiO2-NaOH-H2O suspensions at 100"Cor 150'C. O kenyaite (100"C), O kenyaite (150'C), x magadiite(100'C), I "octosilicate"; (100'C), D unidentified phase (150'C),stippled area: makatite and "octosilicate" field (100'C), dashedline is the speculated stability field boundary between magadiiteand kenyaite.

appears in I to 5 months depending on the SiO2, NaOHand H2O content and slowly transforms into kenyaite.

In the lfi)'C runs with high SiO2A{aOH ratios (>16)kenyaite can be directly formed and magadiite was notobserved as an intermediate. The corresponding compo-sition lie on the "outside" of the speculated stability fieldof kenyaite in Figure I (above the dashed curve).

In the 100"C runs with lower SiO2/NaOH ratios (<10)and compositions lying in the speculated stability field ofmagadiite, magadiite was formed first and then trans-formed into quartz without going through kenyaite as anintermediate (run32F2, Table 2).

When both SiO2/NaOH and H2OA.{aOH ratios are low(stippled area in Fig. l) makatite and "octosilicate" areformed. "Octosilicate" is a hvdrated sodium silicate

synthesized and described by Iler Q964). As yet, thisphase has not been found to occur naturally.

Surprisingly, kenyaite was also formed by heatingsodium water glass solutions to < l25oC, even though thecomposition of these solutions (concentrated solution:SiO2/H2OA.{aOH=2:2:|) lies in the makatite-"octosili-cate" field and far away from the kenyaite field (Fig. l).An amorphous precipitate appears after some days (forinstance, 10 days in runs 40X, Y, W) and slowly trans-forms into poorly crystallized kenyaite which can only bedetected by the intense 19.8A reflection. The crystallinityincreases within 3 months but quartz is simultaneouslyformed. It is likely, that the mechanisms for the formationof kenyaite from water glass and from alkaline SiO2-suspensions are diferent.

The alkali ion appears to influence the type of thesilicate formed. Sodium ions favor the formation ofmagadiite which is slowly altered to kenyaite at higherSiO2A.[aOH ratios (>8). In runs utilizing KOH instead ofNaOH the potassium form of kenyaite was first precip-itated rather than the potassium form of magadiite. In run36,4, NaOH was replaced by a 1: I mixture of NaOH andKOH, and a mixture of sodium magadiite and potassiumkenyaite was obtained. In this respect Li+, Cs+, Sl* andBa2* ions behave like potassium ions.

In Table 3 the X-ray powder patterns of some synthetickenyaites from diferent runs are listed. The variations inthe position and intensity of the reflections are evident.These changes are caused by the intracrystalline changes,in particular by _changes of the layer distances. Thereflection at3.34A is from quartz (3.344, I =- 100), andthe intensity of the reflection at 1.82-1.83A may beslightly increased by the superimposition of a quartzreflection (1.818A, | = l7).

The SiO2/Na2O ratio of the air-dried precipitatesranged from 16 to 28. The water content varied betweenl0 to 30 moles H2O/mole Na2O (the SiO2/NazO ratio fornatural kenyaite is 21.7 (Eugster, 1967)). Due to thedfficulties in analyzing the precipitates this ratio couldnot be determined with high reliability.

r

2

E

t'"

E

Table 2. Reaction sequences of some kenyaite synthesis experiments

Run ,2 F2 Run 27 Z

5 / 3 8 / 7 ( r o o o c ) 9 / 2 2 2 / 7 ( r o o o c ) 7 0 / 9 3 / 1 ( ! o o o c ) r 1 / 7 o o / 7 ( 1 o o " c ) ! 6 / 1 0 0 / r ( 1 5 o o c )

t ime produc ts( d a y s )

i 5 M o

t ine produc ts( days )

6 0 a h ^ ' ^ h

1oq K+Me( ! o t r ): .36 K+Me( 4oE)167 K+Ms( l o t )276 K+Ms( l o t )

107 K+YIc(2o%)j \ 9 K+Ms ( 20 t )

t ime products(days )2\ Ms

776 Ms

I a 7 . h ^ F ^ h

255 t raceso f K

1 8 4 K

t ime produc ts( d a y s )

15 ano"ph.5 \ amorph.

706 amorph.

753 K+I4a(75%)) ) A K + M o / 2 o 4 \261. K+Me( 1o t r )2 9 3 K + M c k s % )32\ Ka 7 ) x

t ime produc ts( d a y s )

1 amorph,J sone K4 x , . 5 % e6 x ' < 5 1 Q

1 4 K , < 5 t Q2 0 0 M g2 4 0 M g100 Ms

J90 a570 a

no lar ra t ios S i02 /HZOINaOH

Mg: magad i i te , K : kenya i te , Q: quar tz , we igh t percent o f Mg or Q

BENEKE AND LAGALY: KENYAITE

Table 3a. X-ray diffraction powder data from synthetic kenyaites and the derived crystalline silicic acids

82r

Natura l kenya i tez) t ) t h e t i c k e n v a i t e s J ) C r v s t a l l i n e s i l i c i c a c i d#run 25 M run 27 Z run 40 Y f rom run 25 M' / f ron run 40 Yr l

! I-----------TrDT-_----------rF'I------.-"rfiTT__--_-rF'I-____--TrrTT-------------"r$_1 9 . 6 8 1 o o 1 9 . 8 0 1 o o 1 9 . 8 2 1 o o 1 9 . 7 0 1 0 0 1 8 . 0 1

5 0 9 . 9 3 3 0 9 . 8 8 3 0t 2 5 . 7 4t 5 \ . 9 72 8 \ . 6 9

1 0 0

1 0 1 . 9 5

9 . O 2a ) t

4 . 2 5

- ) . oo

< 1 0r o ) ) 4 . 8 9 < 1 0

F \ 2 07 0 ' t \ , 2 \ 4 0

F \ 1 A7 0 " 3 . 7 9 t ;

t . \ 3 8 01 0 0

3 . 7 9 8 0

9 . 8 8 t 05 . 0 4 . 7 04 . 9 4 1 0\ , 6 7 2 0\ . 2 6 J o c r1 . 9 3 2 0 - ' ,3 . 6 \ 2 0

1 9 . 8 6 1 o o9 .94 r . l 05 . 0 4 2 04 , 9 8 1 04 . 6 8 2 0

l . 4 l ( \ 8 01 . 3 4 - ' ,3 . 7 9 6 0

9 . 8 1 4 07 . 2 2 2 0

4 . 5 0 2 0

2 0 3 ' 6 \

\ . 2 6t . g 1 3 o ) ) 4 . 0 1 3 03 . 6 5 l l o 3 . 6 2 t o

1 0 1 . 5 1 1 03 . \ 1 e t 8 03 . 1 4 - ' , 7 03 . 7 9 7 0

2 285\ 5

7 \

3 . \ 3 8 0

1 . 2 0 5 0

3 . 4 4 1 0 0

2 . 9 3 1 02 . 8 3 1 0

3 , 2 1 2 0 . r 1 . 2 02 . 9 1 7 0 ' ' , 2 . 9 3

1 2 2 . 8 3 1 0 2 . 8 1 1 0 2 . 8 3 1 0 2 . 7 7_ 3 1 . 8 8 2 0 1 . 8 1 5 0 1 . 8 2 5 0 1 . 8 1 4 0 7 . 8 3 5 0 1 . 8 1

2 . 9 3 2 02 . 8 31 . 8 8 2 0

2 . 9 3 1 0 2 . 9 3 3 0

* ' D e b y e - S c h e r r e r c a n e r a , d i a . = 1 1 q . 5 9 m m , C u 6 O ( - n o t a t i o n

2 ) f t o r E u g s t e r ( 1 9 6 7 ) a n a t M c A t e e E ! a 1 , . , 1 9 6 8 ( n o t a c o n p l e t e l i s t )J ) a i . r - a r i e a , 4 ) w e t , 5 ) t " " y u " o a d , 6 ) q u a r b z ?

Magadiite-keny aite transformatio n

In most experiments with SiO2A.{aOH <16 magadiiteappears as precursor of kenyaite; the final end product isquartz. A transformation magadiite -+ kenyaite -+ quartzis evident. This same sequence was postulated by Eugster(1967) and documented by his subsequent field work atthe magadi area (Eugster, 1969). However, the transfor-mations as expressed by the equations (2) and (a) (Eue-ster, 1969) cannot proceed simply by leaching the sodiumions out of the magadiite structure. Exchange of thesodium ions leads to H-magadiite ("crystalline silicahydrate", see footnote 1,p.E23, Eugster, 1967; Brindley,1969; Lagaly et al., 1975b). Silhydrite found in Trinitycounty, California (Gude and Sheppard, 1972) is in factthe natural form of H-magadiite (Beneke and Lagaly,t977b).

The silicate layers of magadiite are about I lA thick andthose of kenyaite about 17.7A. A solid-state transforma-tion requires a thickening of the magadiite layers by

condensation of silicic acid which has to penetratethrough the interlayer spaces to the reacting sites. Fromour present knowledge about the intracrystalline reactiv-ity of magadiite this process is not very likely. Analternative explanation would be that sodium ions are lostin every second interlayer space followed by condensa-tion of the two adjoining silicate layers into pairs. Brind-ley and Chang (1974) suggested a similar mechanism forthe dehydration of a chlorite. The silicate layers caneasily condense with each other (Lagaly et al. , 1979) , butthe loss of sodium ions in every second interlayer spacecannot be easily achieved or requires a high degree ofcooperativity.

The natural process in the field as an open systemcannot be specified at present. The 100-150'C experi-ments gave no evidence for the occurrence of theseproposed mechanisms.

During the synthesis runs small amounts of the suspen-sions were X-rayed in Debye-Scherrer cameras. In thefirst stages of the runs only reflections of magadiite were

Table 3b. X-ray diffraction powder data from synthetic potassium silicate (K2Si2qO4r . H2O), its sodium exchanged form(Na2Si26Oa1 . HzO), and the derived crystalline silicic acids

K ^ s i ^ ^ o i , . . x x ^ o 1 )- < - z r a ! - z -N a r S i . r O O U r ' x H r O ' / f r o m K 2 S i 2 O O 4 1 . x H 2 O f r o n N a 2 S i . 2 O O ! 1 ' x H 2 O

run 25H run JAs f rom run 14S run 25H2) f rom run J4s1)r -aGT @ T- l'@ E3"G)-

1 o o ! 9 . 8 64 0 9 . 7 2

t o 4 . 8 6

1 0 3 . 8 72 0 1 , 6 69 0 3 . \ t4 0 3 . 2 O2 0 2 . 9 35 0 1 . 8 4

1 0 0 2 0 . 73 0 9 . 8 8

< r o 7 . 2 23 0 4 . 9 1

. t o l ) 3 . 8 31 0 3 . 6 39 0 3 . \ 36 0 t . z o2 0 2 . 9 35 0 1 . 8 3

1 o o 1 9 . 8 63 0 9 . 8 3

2 0 , \ 4 . 8 97 o i < 4 . ? 6l o " 3 . 8 3

1 o o 1 7 . 8 55 0 8 . 8 9) u 7 \ / . z o7 0 ' / 6 . 7 3? o 5 . 9 73 0 4 . 4 47 0 7 \ 3 . 6 75 0 " 1 . 3 9

4 )4 0 1 - 8 l l

1 0 0 1 8 . 0 1t o 8 , 9 1

4 )

4 0 \ . 5 c

l 0 3 . 2 84 )

? o 1 . 8 i

8 0 t , \ 35 0 3 . 2 O2 0 2 . 9 12 0 1 . 8 i

1 ) a i r - d r i e a , 2 ) f r e e z e - d r i e d , J ) v e r y b r o a d , 4)

" "u . " .1 b road re f lec t ions w i th I < 10

822 BENEKE AND LAGALY: KENYAITE

observed. The intensity of the reflections, in particularthe basal reflection at d : 15.6A, then decreased whilereflections of kenyaite, in particular the 19.84-reflection,appeared and increased in intensity. During the runs thereflections of magadiite and kenyaite changed in intensitybut the position and half-width did not change. If linebroadening was observed it did not exceed that expectedfor the decreasing particle size. Thus, a phase intermedi-ate between magadiite and kenyaite or an interstratifiedstructure was not detected. The transformation, there-fore, does not proceed as a topochemical reaction. Mostlikely, it proceeds through an amorphous phase. In run37I (Table 2), for example, the first precipitate wasmagadiite which then became X-ray amorphous after I 16days and then transformed into kenyaite.

Morphology and particle size

Synthetic kenyaites are often obtained in more or lessspherical nodules of plate-like crystals (Fig. 2) which

resemble synthetic magadiite aggregates (c/. Fig. 10,Lagaly, 1979). Sometimes the nodules are embedded inmatrix of amorphous silica (Figs. 2a, b). The samplesfrom water glass solutions at 125'C consist of openaggregates of well-developed plates (Fig. 2d) and resem-ble the natural samples (Fig. 2e). These different types ofsample aggregation suggest that the formation of kenyaitein water glass solutions and in SiO2-suspensions followsdifferent reaction mechanisms.

Quartz formed by decomposition of kenyaite appearsas small spheres (Figs. 2f-h) which are often muchsmaller than I g,m in diameter and often possess apockmarked surface (Fig. 2h).

Intracry s t alline r e ac tio ns

The unit cell dimensions of kenyaite often change whenthe crystals are subjected to changing physical condi-tions, are placed in various liquids, or are treated withacids or solutions of organic cations. Most pronounced is

Fig. 2. Scanning electron micrographs of kenyaites (parentheses: run, temperature, reaction time and SiO2/H2O/NaOH in moles):(a) run 28 D (100'C, 2years, 16:444:1), closely packed nodules ofkenyaite plates; (b) run 37 L (125"C,3 months, 16:138:l), sphericalaggregates of the plates and amorphous material; (c) run 37 S (100'C, l0 months, l6:75: l), loosely packed plates; (d) run 37 U (125"C,3.5 months, water-glass) less regular developed plates; (e) natural kenyaite (Lake Magadi, Kenya); (f) run 27 Z (100'C? 2 years,9:222:l\, alteration of kenyaite into small quartz particles; (g) run 28 E (100'C, 2 years,22:222:l), well developed kenyaite plates andspherical quartz particles; (h) run 28 E, surface morphology ofthe quartz particles.

the shift of the intense X-ray reflections at 19.8A and9.9A to higher d-values (Table 4). The changes resemblethose observed when magadiite is similarly treated andare typical for reactive hydrated layer structures (Eugster1967, Brindley 1969,Lagaly et al. 1973,1975 a,b). Thus,the reflections at 19.8A and 9.94 are considered to bebasal reflections; probably the first and second orderreflections.

Some of the basal spacing changes listed in Table 4should be explained. The decrease by 2A from 19.7Arc17.7A suggests a lo^ss of a monolayer of water. Thestability of the 17 .7 A-spacing up to the transformationinto quartz (700'C) suggests that this spacing correspondsto the van-der-Waals-thickness of the single kenyaitelayer. Thus, the silicate layers in air dried kenyaite appearto be separated by a monolayer of water molecules.

Long-chain organic cations (cationic surfactants) re-place interlayer sodium ions by ion exchange. The larged-spacings (24-444, Table 4) indicate that th; cations areeggregated in bimolecular layers as previously describedfor magadiite (Lagaly et al. 1973,1975 a,b) and kanemite(Beneke and Lagaly, 1977a). In general, the reaction ofprimary alkylammonium ions with alkali layer silicatesrequires special experimental conditions (Lagaly et al.,1973;Kalt et al., 1979). This is also true for kenyaite, butunlike other alkali silicates kenyaite easily reacts withphenyl- and benzylammonium cations.

In the presence of glycol, glycerol, N-methyl forma-mide, urea, etc., the spacings are increased (Fig. 3)because the organic molecules displace interlayer watermolecules and occupy more interlayer space. The basalspacings often depend on the method of preparation ofthe parent compound (kenyaite) and particularly on thereaction times. The lattice spacings tend to decrease withincreasing reaction times and reflect the presence ofunexpanded interlayer spaces.

Table 4. Basal spacings "::#lH*

kenyaites under differenr

BENEKE AND LAG,4LY: KENYAITE 823

\\lHts \*

**'{ -*-\

t-\ "-t"€o

a+-

1 1 4 1 1

-h37 10 3t

t-*----------___

s 4 4 q 2 pI4 7d lU 8()d 35m l(!n llm t3m zLn

$oqc 1250c 1000c

Fig. 3. Intracrystalline reactivity of kenyaites (prepared at 100,125 and 150"C and at diferent reaction times) with respect tovarious liquids (O ephedrine, A ethylene glycol, n glycerol, V :N-methyl formamide, I : dimethyl formamide).

H-kenyaite

If kenyaite is stored in an excess of diluted acids (forinstance 0.1 m hydrochloric acid) all interlayer sodiumions are exchanged by protons. The product is a crystal-line silicic acid.t Titrating kenyaite suspensions with 0.1m HCI shows that the exchange starts at pH 9-10 butmainly proceeds at pH about 7 (Fig. 4). However, theexchange is not quantitative at the "equivalent points"even though the hydrochloric acid was added in rates notfaster than 1.3 mVh. At the "equivalent points" 40 to 70percent ofthe sodium ions were exchanged. Some ofthesodium ions are evidently fixed in the interlayer sites.Usually, the basal spacing of the kenyaites in suspensionremains nearly unchanged (19.6-19.8A) (Table 4) duringthe H*/Na* exchange. The spacing then collapses afterair drying to 17.6-18.0A. One monolayer of water is

'Crystalline silicic acids are obtained by exchanging theinterlayer cations of alkali layer silicates and some other layersilicates with protons (Schwarz and Menner, 1924; Pabst, 1958;Wey and Kalt, 1967; Lagaly and Matouschek, 1980; Frondel,1979; review: Lagaly,1979). Since these compounds are ratheracidic (Werner et al., 1980) "crystalline silicic acid" or "crystal-line phyllosilicic acid" (Liebau, 1964) is a useful nomenclature.Nevertheless, some authors use the notation "crystalline silicahydrate" ("SH-phase, Eugster, 1967) or, more simply, "leachedgillespite" (Pabst, 1958) or "leached magadiite." At presentmore than 15 diferent silicic acids are known. Since a systematicnomenclature is lacking, the acids, at least those derived fromminerals, are called "H-magadiite, H-apophyllite, H-kenyaite,etc . " .

Eugster (1967) reported 6SiO2 ' H2O for his "SH-phase"which is the crystalline acid from magadiite. The crystallinesilicic acid reported here and prepared from synthetic kenyaitehas the approximate composition SiO2 . H2O in air-dried stateand SiO2 . 0.2H2O after freeze-drying. Note that the watercontent changes with the relative humidity.

!E

Treatment b a s a l

k e n y a i t e , a i r d r i e d( d l f f e r e n t s a n p l e s )

u n d e r o . 1 n a c i d( H C 1 , H 2 S O x , H C l O t r , H N O ? e t c . )t h e n a r ? - d r i e d

k e n y a i t e , 1 2 0 o C ( v a c u u n < t O - Z t o " " )k e n y a i t e , 2 0 0 o Ck e n y a i t e , 4 0 0 o Ck e n y a i t e , 6 0 0 o Ck e n y a i t e , 7 0 0 o C

k e n y a i t e + 0 . 1 n a q u e o u ss o l u t i o n o f

d e c y l a n m o n i u n c h l o r i d ed o d e c y l a n n o n i u m c h t o r i d et r i m e t h y l d e c y l a m o n l u m c h l o r i d et r r m e t h y l d o d e c y ) . a m o n i u m b r o m i d ed i n e t h y l d i d e c y l a m o n i u n c h l o r i d ed i m e t h y l d i d o d e c y l a m o n i u n c h l o r i d et r i n e t h y l p h e n y l a m o n i u m c h l o r i d et r i m e t h y l b e n z y l a m o n i u n c h l o r i d et r i e t h y l b e n z y l a n n o n i u n q b t o r i d e

1 9 . 7 - 7 9 . 9

1 9 . 6 - 7 9 . 8

1 7 . 6 - 1 8 . 0

7 7 . 8

1 1 0 . 64 4 . 43 9 . 2\ 2 . 53 9 . 24 3 . 62 \ . 62 7 . 2

_ e x c e p t . l o n s s e e F i g u r e 5

824 BENEKE AND LAGALY: KENYAITE

- ml Olm Hct

Fig. 4. Titration data ofkenyaite suspension (lgll(X) ml water)with 0.1 m HCI: rate of addition of HCI < 1.3 mVh. Percentsodium ions exchanged by protons at the "equivalent points"are indicated. Na2O/SiO2/H2O in 2ED: l:28:17.4; in 37Ll :17.7:29.5; in 25M: l :25.5:10.9; in 41T l :15.8:23.

apparently desorbed from each interlayer space. Rehy-dration of this H-kenyaite in water does not increase thespacings above 18.6A. Evidently, only some of the col-lapsed interlayers are rehydrated. Dehydration at highertemperatures (24 hours at 2fi)"C) decreases the propor-tion of the expandable interlay-ers and the maximumspacing after rehydration is 18.0A. In repeated dehydra-tionirehydration cycles the maximum spacing is furtherdecreased and finally the lattice expansion is completelyblocked. A single dehydration step at 300"C (24 hours)has the same effect.

Some synthetic kenyaites, in particular from runs at125 and 150"C (for example, runs 40Y, 4lP, 4lT) givecrystalline silicic acids with basal spacings Iess than19.6A, even in the wet state (Fig. 5). During the H"lNa*exchange some interlayer spaces apparently lose thewater monolayers leading to an interstratification of hy-drated (d : 19.8A) and dehydrated interlayers (d :17.64). Treatlnent of H-kenyaite with sodium hydroxidesolutions reconstitutes the kenyaite.

The surface acidity of H-kenyaite in aqueous disper-sion (-3 < pK, < +1.5) resembles that of other highlycondensed crystalline silicic acids (Werner er a/., 1980). Itis independent of the kind of the acid (HCl, H2SO4,HNO3, acetic acid, erc.) used for the H*/Na*-exchange.

H-kenyaite like the other crystalline silicic acids (Laga-ly,1979) intercalates several organic compounds (Fig. 5).Most reactive are aliphatic amines. Some H-kenyaites donot expand to the maximum value and reflect the loweredreactivity of the parent kenyaites.

Comparison with other layer silicates

Kenyaite appears to be related to a potassium silicatewith approximate composition K2Si26Oa1 ' xH2O (x - ll)(Lagaly, 1980). The X-ray powder patterns (Table 3) arevery similar if conditions are chosen so that the basalspacings are equal. The relations between kenyaite andKzSizoOrr . xH2O are best revealed by their intracrystal-line reactions. The crystalline silicic acid prepared from

the potassium silicate reacts with NaOH and gives asodium silicate which strongly resembles kenyaite. Thus,kenyaite appears to be the sodium form NazSizoOal ' xH2Oof the potassium silicate K2Si2qOa1 ' xH2O.

One is not surprised that intracrystalline reactions anddehydration/hydration experiments reveal some difer-ences between kenyaite and the sodium and potassiumsilicate as well as among the corresponding acids (Table5). When placed in contact with water the basal spacing ofthe air-dried potassium silicate expands to nearly 224(corresponds to a bilayer of water molecules between thesilicate layers) whereas that of the sodium silicate andsynthetic kenyaite remains the same (19.SA) and corre-sponds to a monolayer of water molecules between thesilicate layers. In the air-dried state or after dehydrationat higher temperatures the spacings of the three com-pounds become identical.

Surprisingly, the potassium ions promote a higherhydrated interlayer state. In the titration experimentsreported in Figure 4 the sodium ions are not quantitative-ly exchanged by protons. This reveals a fixation ofsodium ions in the interlayer spaces. K2Si26Oa1 ' xH2Oquantitatively exchanges the potassium ions when used inthe same procedure. This behavior contrasts with thepotassium fixation by higher-charged 2: I clay mineralssuch as vermiculites and micas. It should be noted thatpotassium fixation not only results from the reducedhydration tendency of the potassium ions but also fromspecific interactions with the silicate surface. One might,therefore, speculate that the sodium fixation and therestriction to a single water monolayer in the Na2Si2sOal ' xH2O and kenyaite structures give evidence ofspecific interactions with the silicate layer which over-come the increased hydration tendency of the sodiumlons.

22

d--2(C

9 le

Eo

I d X L d

t5o.c

Fig. 5. Basal spacings of H-kenyaites under differentconditions and in contact with various liquids. Lower third offigure: O kenyaites in contact with HCl, I H-kenyaites, air-dried, O H-kenyaites, rehydrated; upper two thirds of figure:H-kenyaites under dimethyl sulphoxide (O), ethylene glycol (A),N-methyl formamide (V), dimethyl formamide (I), ephedrine(o).

80d 35m l3.n 2am not

r250c tocc

b=\>al 4t 4lS I P

--\o

a

q}

dA

I

. d *

I\

c

1? L0L Y

H!

a\+--+

25 28M D

D

BENEKE AND LAGALY: KENYAITE

Table 5. Basal spacings of kenyaite, sodium and potassium silicate, and the corresponding crystalline silicic acids under variousconditions

E25

wef,B a s a l s p a c i n g s ( A )

a i r -d r ied rehydra ted KOH,NaOH25"C (under water ) ( )1n)

heated to heated to2 0 0 o C 2 0 0 o C , t h e n

rehvdre fedSample and run No,

na tura l kenya i te

s y n t h . k e n y a i t e * ( 2 5 M ) 1 9 . 8s o d i u m s i l i c a t e 1 9 . 9 - 2 0 . \p o t a s s i u n s i l i c a t e ( 2 5 H ) 2 7 . 8

s i l . i c i c a c i d s f r o n :( 1 ) n a t u r a l k e n y a i t e 1 8 , 0( 2 ) s y n t h e t i c k e n y a i t e ( 2 5 M ) 1 9 . 7( J ) s o d i u n s i l i c a t e ( l 4 S ) 2 0 . 7

79 .7

1 9 . 87 9 . 919 .9

. 1 a

1 8 . 0

1 9 . 8

1 9 . 81 9 . 827 .6

7 7 . 91 8 . 51 8 , 0

NaOH:7 9 , 92 0 , 720 .3

Kofl:

1 0 A

7 7 . 7

7 7 . 7

7 7 . 67 7 , 7a a a

19 .6

1 8 , 0 . *79 .6 * t2 0 . 4

7 8 . 21 a a

t a s s i u m s i l i c a t e

prepared f ron the

c v c l e s a t 2 q 0 o c :

p o t a s s i u n s i l i c a t e

1 7 . 7 S

7 9 . 5( l q s ) , scheme ( r ) , af te r dehydra t ion , / rehydra t ion

1

The acid form ofthe potassium silicate has spacings of21.3-2l3A which decrease to 19.0-l9.ZA UV air-drying.Rehydration reconstitutes the 22A-form. The acid formsof synthetic kenyaite and the sodium silicate have spac-ings of 19.5-20.84, 17.7-18.0A after air-drying and<18.64 after rehydration. Thus, the transformation of thepotassium silicate into the sodium silicate and vice versaproceeds through two acid forms which are very similar

Na*-(scheme I)

Since the spacings of both acid tbrms become identicalafter dehydration (>120"C), the thickness of the silicatelayers appears to be the same in all compounds. Theinterlayer cations probably influence orientation and ar-rangement of the [Sio4]-tetrahedrons. Scheme I impliesthat the silicate layers retain some flexibility which allowsan adaptation to the interlayer cations.

Conclusion

The reported studies make evident that the chemicalcomposition and the X-ray powder patterns of syntheticand natural kenyaites can be sensitively changed: (l) byvariations in the amount of interlayer water, if the physi-cal conditions (for example, relative humidity, tempera-ture) change, (2) by partial or complete exchange of theinterlayer sodium ions by protons if kenyaite comes intocontact with water. The changes are typical for intracrys-talline reactive minerals. Other minerals showing thisbehavior are magadiite (Lagaly et al., 1975a), kanemite(Beneke and Lagaly, 1977a), krautite (Beneke and Laga-ly, l9El), uranium micas (Weiss and Hofmann, 1952\,uvanite (Hilke er al., 1973) and clay minerals.

Sodium interlayer cations are exchanged by protonseven in alkaline solutions (pH =9, Fig. a). Thus, whenkenyaite is brought into contact with water, the waterleaches sodium ions out of the structure. In the firststages of exchange (exchange rutio <20Vo) the basalspacings are only slightly changed. Eugster (1967, 1969)and Eugster et al. (1967) report a SiO2A.{a2O ratio of 22for kenyaite. The ratio in synthetic kenyaites is about 20and appears to be confirmed by the composition of theanalogous potassium silicate K2Si2eOa1 . xH2O. A higherratio for natural kenyaite obviously can result from partialleaching by percolating water.

The variability of the powder pattern complicates thedetection of kenyaite in nature. The changes reported inthis study may aid in a proper identification of kenyaiteeven in mixtures with other sodium silicates.

Acknowledgments

Thanks are extended to Dr. Christian Samtleben, Mr. W.Reimann, and Mrs. U. Schuldt, of the Institute of Geology of theUniversity of Kiel for taking the scanning electron micrographs.We are grateful to the "Fonds der Chemischen Industrie" forfinancial support.

References

Beneke, K. and Lagaly, G. (1977a) Kanemite-innercrystallinereactivity and relations to other sodium silicates. AmericanMineralogist, 62, 763-:77 l.

Beneke K. and Lagaly, G. (1977b) Pecularities of intercalationreactions of the natural silicic acid silhydrite. Clay Minerals,12. 363:365.

Beneke, K. and Lagaly, G. (l9El) Krautite, MnHAsOa .HzO-an intracrystalline reactive mineral. American Mineralogist,66,432435.

Brindley, G. W. (l%9) Unit cell of magadiite in air, in vacuo andunder other conditions. American Mineralogist, 54, 1583-1591.

Brindley, G. W. and Chang, Tsien-Show (1974) Development oflong basal spacings in chlorites by thermal treatment. Ameri-can Mineralogist, 59, 152-158.

Eugster, H. P. (1967) Hydrous sodium silicates from Lake

but not identical:

potassiumsilicate

1lK* l l H*

I tacid form-I

K--acid form-Il

1 lH* ll Na*

Isodium silicateand kenyaite

826

Magadi, Kenya: Precursors of bedded chert. Science, 157,1177-1 180.

Eugster, H. P., Jones, B. F., and Sheppard, R. A. (1967) Newhydrous sodium silicates from Kenya, Oregon and California;possible precursors of chert. Geological Society of America,Program Annual Meeting, 1967 , New Orleans, Louisiana, 60.

Eugster, H. P. (1969) Inorganic bedded cherts from the MagadiArea, Kenya. Contributions to Mineralogy and Petrology, 22,l -31 .

Frondel, C. (1979) Crystalline silica hydrates from leachedsilicates. American Mineralogist &, 799-804.

Gude, A. J. and Sheppard, R. A. (1972) Silhydrite,3SiOz.HzO, anew mineral from Trinity county, California. American Miner-alogist, 57, 1053-1065.

Hilke, K. J., Grosse-Brauckmann, U., Lagaly, G., and Weiss,A. (1971) Zur Chemie von Uranylvanadaten mit Trimetavana-dationen und Schichtstruktur. Zeitschrift fiir Naturforschung,28b,239-245.

Iler, R. K. (1964) Ion exchange properties of a crystallinehydrated silica. Journal of Colloid and Interface Science, 19,648-6s7.

Johan, Z. and Maglione, G. F. (1972) La kanemite, nouveausilicate de sodium hydrat6 de n6oformation. Bulletin de laSoci€td francaise de Min€ralogie et de Cristallographie, 95,37r-3E2.

Kalt, A., Perati, B. and Wey, R. (1979) Intercalation compoundsof KHSi2O5 and H2Si2O5 with alkylammonium ions and alkyla-mines. Proceedings International Clay Conference, Oxford1978, Elsevier, Amsterdam. 1979. 509-516.

Lagaly, G. (1979) Crystalline silicic acids and their interfacereactions. Advances in Colloid Interface Science, I I , 105-148.

Lagaly, G. (1980) Kristalline Kieselsiiuren, ihr Kaliumsalz, Ver-fahren zu ihrer Herstellung und ihre Verwendung. GermanPatent, P 2742912 (August, 23,1977).

Lagaly, G. and Matouschek, R. (1980) Die kristallinen Kiesel-siiuren aus Apophyllit, Carletonit und Gillespit. Neues Jahr-buch fiir Mineralogie, Abhandlungen 138, 8l-93.

Lagaly, G., Beneke, K., and Weiss, A. (1973) Organic complex-

BENEKE AND LAGALY: KENYAITE

es of synthetic magadiite. Proceedings International ClayConference. Madrid 1972. Division de Ciencias C.S.LC.,Madrid 1973,663-671.

Laf:aly, G., Beneke, K. and Weiss, A. (1975a) Magadiite and H-magadiite: I. Sodium magadiite and some of its derivatives.American Mineralogist, 60, 642-649.

Lagaly, G., Beneke, K., and Weiss, A. (1975b) Magadiite and H-magadiite: II. H-magadiite and its intercalation compounds.American Mineralogist, 60, 650-658.

Lagaly, G., Beneke, K. und Kammermeier, H. (1979) NeueModifikationen des Siliciumdioxids. Zeitschrift f0r Naturfor-schung, 34b,666-674.

Liebau, F. (1964) Uber Kristallstrukturen zweier Phyllokiesel-siiuren, H2Si2Os. Zeitschrift fiir Kristallographie, 120, 427-449.

McAtee, J. L., House, R., and Eugster, H. P. (1968) Magadiitefrom Trinity County, California. American Mineralogist 53,2Mr-2069.

Pabst, A. (1958) The structure of leached Gillespite, a sheetsilicate, American Mineralogist, 43, 970-980.

Sheppard, R. A., Gude, A. J., and Hay, R. L. (1970) Makatite, anew hydrous sodium silicate mineral from Lake Magadi,Kenya. American Mineralogist, 55, 358-566.

Schwarz, R., and Menner, E. (1924) Zur Kenntnis der Kiesel-s?iuren (I). Chemische Berichte 57, 1477-1481.

Weiss, A. and Hofmann, U. (1952) Reaktionen im Innern desSchichtgitters von Uranglimmern. Zeitschrift ftir Naturfors-chung, 7b, 362-363.

Werner, H.-J., Beneke, K. und Lagaly, G. (1980) Die Aciditatkristalliner KieselsAuren. Zeitschrift fiir anorganische undallgemeine Chemie, 470, l18-130.

Wey, R., and Kalt, A. (1967) Synthdse d'une silice hydrat6ecristallis6e. Comptes Rendues d'Academie de Sciences fran-cais, Paris, 265, 1437-1440.

Manuscript received, April 23, 1982;accepted for publication, December 8, 1982'