Citation: Hodder, A.P.W.; Naish, T.R.; Lowe, D.J.. 1996. Towards a? under~tanding of the~?dynamic and kinetic con~rols on t h~ formation of clay minerals from volcaruc glass under vanous envrronmental_ conditions_. Pp. 1-11 (Chapter 1) m Pandalru, s. G. (Ed) Recent Research Developments in Chemical Geology. Research Stgnpost, Tnvandrum.
Receut Res.llfvel. iu Chemical Geol.. 1 (1996)
Towards an understanding of thermodynamic and kinetic controls on the formation of clay minerals from volcanic glass under various environmental conditions A.P.W. Hodder, T.R. Naish and D.J. Lowe Department of Earth Sciences, University of Waiklto, Private Bag 3105, Hmnilton, New Zealand
Thermodynamic stability diagrams based on a behind these processes. Ruxton (6) recognised that the topological procedure constrained by thermodynamic formation of clay minerals from glass was essentially a data for idealised smectites and imogolite indicate that first-order process: that the proportion of clay minerals imogolite is the favoured alteration product of rhyolitic (Pc) was related to the initial proportion of glass (P,:) volcanic glass in both the soil and nearshore marine by a relationship of the form. environments. However, smectites are observed in PJP,,c =exp (- ktt) (1) . young nearshore marine sediments, suggesting a rapid where kt is the rate constant for the process and tis the . alteration. The rate of formation of clay minerals has time for weathering. been previously demonstrated to be a function of their equilibrium constant (K) and the reaction activity On a rather shorter time-scale, the development quotient (Q) - effectively the divergence from of hydration rinds on obsidian artefacts (e.g., 7; 8; 9) equilibrium. Extending this treatment to include the and studies on the dissolution of glass in the laboratory process of glass dissolution (8) as a prelude to clay suggested that this was a diffusion controlled process (10; 11), for which a typical expression for the change ($), mineral formation the rate of clay mineral in concentration (C) with time (t) is of the form: formation is demonstrated to be a function of the parameter: C= CNI/2 (2) where C0 is the concentration initially and kp is the so· called "parabolic rate constant". For young tephra k.t{l-Q&'Ke)- k'+(l-Q41~) deposits, this diffusion process may be an important Where k+ and k' + are the rate constants of dissolution part of the total weatherin$ process. A model coupling for glass and clay respectively. The values of Q91Ke the first-order and diffusional process represented by are such that in general, the rate determining step is equations (1) and (2) was developed (12) and then determined by Qq.JK~ applied to glasses weathering to allophane (Si-rich allophane with AI:Si ratio of 1:1 as in halloysite), Based on this approach, the rate of formation of imogolite (Al-rich [proto-imogolite] allophane with smectite is fast in silica saturated interstitial waters in AI:Si ratio of 2: 1), or halloysite in the soil-forming the nearshore marine environment, perhaps a thousand environment (13). In the marine environment times faster than the formation of imogolite or bentonite has long been recognised as an alteration halloysite in soil waters. product of volcanic glass (e.g., 14). Most studies of bentonites imply that because their formation has been INTRODUCTION from volcanic ash or tuff deposits in deep-sea sediments The weathering of rhyolitic volcanic glass, of at least Pleistoce.ne age (e.g., 15; 16; 17; 18; 19), .chiefly from tephric materials, to form clay minerals factors such as time, burial and temperature are has been extensively studied (e.g. I; 2; 3; 4; 5). Clay important to bentonite formation. "Bentonite" has been mineralogists and pedologists have been principally used quite generally to describe the alteration ~roduc ts, concerned about the structure and composition of the v:hich include smectite clay minerals and'zeohtes (20). derived clay minerals, and tllefr effects on soils or More recently late Holocene estuarine sediments paleosols, but there has been comparatively little adjacent to a volcanic terrain (Firth of Thames, New consideration of the thermodynamics and kinetics Zealand) have been recognised as containing large
~ : 2 A.P.W. Hodder rta/.
~nicarus (wtll) :-: 0 ., , \ \ \
\ ,' \ ,
, '
100 0 100 0
Waltoa River
~ Land above 250m Sample locations •
FIG I. A. Composition of sediments in rivers entering the Firth of Thames, New Zealand, and surficial sediments of the firth. B. Compositional logs for a representative core in the Firth of Thames (21). (Types of allophane are not distinguished in the core logs).
amounts of smectite (21 ). The mineralogy of the formed in the nearshore marine environment, but streams draining the region and entering the Firth does imogolite (or allophane or halloysite) in the soil not include smectite, but does include abundant forming environment? By the use of thermodynamic volcanic glass (Fig. 1). Naish (21) considered it likely stability diagrams and the recognition that relative that the smectite was forming from the volcanic glass mineral solubilities may be proxies for rate constants under early diagenetic conditions in the Firth, and for mineral dissolution or precipitation, this paper Hodder et al. (22) proposed a two-stage model for the shows that high silica concentrations in solutions of kinetics of smectite formation from detrital volcanic high ionic strength are required to precipitate smectite, glass analogous to that previously developed for but that the solution needs to have only a high silica imogolite formation from glass in tepha-c!erived content to ensure formation of imogolite. · paleosols. THERMODYNAMIC STABILITY DIAGRAMS The question remains: what are the The representation of thermodynamic stability environmental conditions that c;nable smectite to be for clay minerals formed as a result of 'weathering' (in Fonnation of clays from volcanic alass 3 the soil-forming environment) or 'diagenesis' (in the 1.0....----- nearshore marine environment) of glass is made more difficult by the absence of free energy data for the glass IDM>ile and for allophane. To some extent this difficulty can 0.11 lb~~At:Malcime G Glau be reduced by a topological approach (23). For this Gi Gibblil• H HaDO)'Site approach, only the compositions of the minerals are I lmogoftto required to be known. For two minerals to be in rS 0.6 Sm Na-bel Ito amec1it equilibrium, the compositions on an appropriate plot ~ are joined (see Fig. 2). Such tic-lines may not ~ 0.4 intersect, and the tic-Hnes arc at right angles to the line ~ defining equilibrium between the minerals on the thermodynamic stability diagram. Such compositional diagrams and their derived stability diagrams are shown for the Na20-Si02·Al:z03-H20 and CaO-Si02·AI203- H20 systems on Figs 2 and 3 , respectively. Data for · 6 8 10 the free energies of the clay minerals enable these Si02/AI20 3 diagrams to be better constrained. For this purpose the Gibbs standard free energy of formation of Si-rich allophane is taken as that of halloysite, the Gibbs standard free energy of formation for smectite clay minerals in taken as that for idealised N and Ca beidellites {23}, and the value for imogolite is as calculated by Percival (24). 6 i The limiting positions of the stability lines are shown (broken lines on lower part of Fig. 2) as defining glass in equilibrium with imogolite and halloysite only, I.. or with smectite and halloysite only. An intermediate 0 position (solid line) allows equilibrium between glass and all the clay minerals. 0 Also shown on these diagrams are the compositions of seawater and the range of compositions of soil water from topsoils formed from rhyolitic tephra (volcanic ash) (25). These soil-water compositions lie mainly within the calculated stability ....s -•.o -3.s -3.o field for imogolite but some are in the halloysite log aH•SIO• .._ stability field. This is in accord with the observations (26) of low soil solution silicon is present in soils where proto-imogolite allophane is predominant, but halloysite predominates in soils of relatively high FIG. 2 A. Composition diagram for the Na20-siQ:z solution silicon. The values for seawater are close to AI203-H20 system. Joining lines indicate the zeolite-imogolite boundary, well distant from the 'allowed' reactions; the slopes of the ~base smectite stability field. If smectite is to be formed from boundaries are at right angles to these JOi:Js. rhyolitic glass in the marine environment, either the This serves to constrain the topology of the solution chemistry has to change during early phase boundaries involving glass. The diagenesis or the process is kinetically controlled, or composition diagram assumes indealised both. compositions of the clay minerals and the composition of glass is taken as that typical of New Zealand rhyolites. A decrease in the (log ONa +pH) or log (ac12t + 2pH) values could reasonably occur with the reduction B. Stability diagram: loga.Na+ + pH vs of pH, reasonably expected during diagenesis by the logGH4Si04· Hatched region is range of soil reduction of seawater sulfate and in response to the water compositions from Percival (25) bacterially-assisted reduction of iron and manganese. The stability diagram is constructed on the In the case of (log ac.2+ + 2pH), this parameter could assumption that aluminium is conserved in the also be decreased by the precipitation of diagenetic system; positions of the stability lines further calcite, although there was no evidence of this in the constrained by the use of standard Gibbs free Firth of Thames cores. The increase of dissolved silica ener~y data where available, giving results required to attain the smectite field may be achieved consJGtenfwith Percival (36). :
A.P.W. Hoddcrttal.
during diagenesis (28). An alternative is to consider that although in seawater the formation of imogolite COMPOSITTON M An011Nia may be thermodynamically favoured, the activation 0.8 l laui'TIOI'IIile energy for its formation in interstitial waters may be G Glass much greater than the activation energy for smectite. 1 lmogolte Under these conditions smectite might be formed at a A AlopNine rather greater rate than the formation of imogolite. a 0.6 H Haloysite Sm Ca beidelile The cores in the Firth of Thames have penetrated the i (smectite) estuarine sediments to a depth of about 5.5 m with a 0 Gl Gibbsite radiocarbon age of ca. 5000 BP at this level. The j 0.4 down-core Increasing trend in smectite abundance suggests not only that smectite is forming under these conditions- at the expense of glass (see Fig. I)-but 0.2 also that it is forming quickly.
2 4 6 8 10 Si02/Al20 3 100~------,~------~ Par&nt material
14 ~} Rh)ooiJic eluvium
: } Rhyolitic lephta 12
~ 10 + ~ iii ~ >::s 8
~ !0 E 0 6 .. ~.0 -3.75 -3.5 -3.25 -3.0 -2.75 '!5 100 I ~ 4 I . I ! 110 • I HALt..afSITE 'S. ~ 4.5 4.0 3.5 3.0 2.5 '000 log aH,SO, ~ ·~ ~ ~ Mod. wee fo drain«! ~ FIG. 3 A. Composition diagram of the CaO-Si02- A120J-H20 system. The procedure to achieve QU-----~--~~~~~~~~----~ the stability diagram from this composition -4. -3.75 -3.5 -3.25 -2.75 diagram is as for Fig. 2. I OQ all•SIO. B. Stability diagram: logac.2+ + 2pH vs logtzH4Si04· Hatched region is range of soil Gibbsir•J1 Jmogoit ~ !AiaSi03COHl.l J Hailoysile lJ water compositions from Percival (25). ~So lubil ity ol QUBr1Z SolubiUiy of amorPh,ous siDca '""'! The stability diagram is constructed on the assumption that aluminium is conserved in the system; positions of the stability Jines further FIG. 4 Halloysitic (A) and 'allophanic' (B) clay constrained by the use of standard Gibbs free mineralogical compositions of soils derived energy data where available, giving results from rhyolitic volcanic materials in northern consistent with Percival (36). New Zealand, related to dissolved silica (26). Formation of days llom volcani~ glass 5
KINETIC CONTROL ON CLAY MINERAL mineral 9, ne has values in the 0-1 range and k+ is FORMATION given by: There have been few studies on the kinetics of ~ = ~ exp (·E+IRT) (5) clay mineral formation. However, since as is shown assuming the rate constant k+ has an Arrhenian later, it is possible to infer the rate of precipitation from temperature deJ?endence, E+ is an activation energy, A+ dissolution kinetics, it is appropriate to comment briefly a pre-exponential factor, R the gas constant, and T the on the mechaisms of mineral dissolution. temperature. If the dissolution rates do not depend on the concentrations of other components in solution, The solubility of selected silicates and inorganic then (30) as the solution approaches equilibrium compounds seems associated with their mechanism for equation (4) can be recast as: dissolution: more soluble compounds dissolvedtby !!5. Ae Ae Qem transport- or diffusion - controlled mechanisms, those dt = V Ui9 k+ (aH+)n9 • v Ui 9 Kern k+ less soluble by surface reaction control (29). Diffusion controlled processes typically have activation energies (aH+)nB (6) of about 20 kJ mol -1. Surface reaction control where m can be any real number, not necessarily an mechanisms may from Fig 5 be inferred to have higher integer, and Qe is the reaction activity quotient, and Ke activation energy of dissolution, but these are, nevertheless, generally Jess than that typical of bond is the equilibrium constant for dissolution. The second energies. For the clay minerals considered here the term in equation {6) represents the rate law governing values of the equilibrium constant for dissolution, log precipitation. K, is typically S -1 0, implying surface reaction control. The essence, then, of equation ( 6) is that the overall rate of a dissolution process can be expressed in For the dissolution of a particular mineral, terms of thermodynamic parameters associated with expressed as: that dissolution. More sisnific andy for the present case, the rate of precipitation of a mineral (9) + n H+-+ silica± alumina+ cation:; mineral - the second term of equation (6) - is also (i). (3) related to thermodynamic parameters associated with its The rate of dissolution, dependent particularly on pH dissolution. and temperature, can be expressed (30): de· Ae If a two-step reaction sequence is considered: yl>i9 4 (aH+)n 9 (4) Tt-= mineral ( 9) + vH+ ~ silica+ alumina+ cations (i) where Ae is the surface area of mineral e, V is the ~mineral(~)+ pH+ (1) volume of solution in contact with the mineral e . \li9 is the stoichiometric content of chemical species i in then the corresponding expression for ( 6) is: 160 i ':- 140 =!~Yo ... (.,.)"['- ~Hl 1 120 i! 100 (8] r eo -l~v;;k;('>.)'f~)) ] so j l! ~0 From equation (8) the rate of transformation 2(1 from glass (9) to a clay mineral (~ ) depends on the equilibrium constants for the dissolution of glass and of 20 3J 40 :ill 60 log K clay minerals and their respective reaction activity 0 Surrilc:e reaclon eont1ol coefficients. These coefficients can be calculated from the concentrations of the appropriate chemical species in soil and interstitial waters that host the reactions (see [] Transport tt•. ciffuskwll conlrol Table 1)
As a first step to determining whether the FIG. 5 Relationship between activation energy for dissolution of silica-bearing glass or the formation of dissolution of minerals and their equilibrium the clay mineral is rate determining, if it is assumed constants for dissolution (30) Also shown is that: the inferred mechanism for dissolution (29) Ae=At=A;v9=p41=1 ;v:=p= l (9) 6 A.P.W. HoddcrctaL
TABLE 1: Clay mineral dissolution data
MINERAL (pH-range) Reaction
log~ logO.
IMOGOLITE (basic) AI2Si03(0H)4 + SH20-+ 2AI(OH)4- + H4Si04 + 2H+ log K1 = • 36.04 log Qt =log 8H4Si04 + 21og8AI(aq) -2pH
(neutral) Al2Si03(0H)4 + H20 + 2H+-+ 2AI(OH)2+ + J:4Si<>4 log KI =- 9.38 log Ql = 2 log aAl(aq) + log8H4Si04 + 2pH
(acid) AI2Si203(0H)4 + 6H+-+ 2AJ3+ + H4Si HALLOYSITE (basic) AhSi20s(OH}4 + 7H20-+ 2Al(OH)4- + 2H4Si04 + 2H+ logKH = - 38.6 log QH =2 log 8A)(aq) + 2log8H4Si04 • 2pH {neutral) A)zSi20s(OH)4 + 3H20 + 2H+-+ 2AI(OH)2+ + 2}4Si<>4 logKH=· 11.94 log QH= 2log 8At(aq) + 2los8H4Si04 + 2pH (acid) AI2Si2 0s(0~)4 + 6H+ -+ 2AJ3+ + 2}4Si<>4 + HzO log KH =+ 7.54 log QH= 21og 8Al(aq) + 21og8H4Si04 +6pH Na - SMECTITE (oasic) Na o.33AI 2.33Si 3.67 0J0(0H)2 +12H20-+0.33Na+ + 2.33 AI(OH)4· + 3.67 H4Si04 + 2H+ log Ks=- 47.5 log Qs =0.33" log8 Na+ + 2.33 log8 Al(aq)-+:_ 3.67 log8 H4Si04- 2pH Fonnatlon of clays &om volemic glass 7 Table l.contd. (neutral) Na o.33Al2.33Si 3.67 010(0H)2 + 7.33H20 + 2.66H+~0.33Na+ + 3.67 H4Si04 + 2.33 Al(OH)2+ 8 8 log Ks= - 16.05 log Q5 = 0.33 loga Na+ + 2.33 log AI(aq) + 3.67 log H4Si04- 2.66pH (acid) Na 0.33AI 2.33Si 3.67 0J0(0H)2 + 7 .33H+ + 2.68H20~0 . 33Na+ + 3.67 H~i04 + 2.33 AJ3+ 8 8 log Ks= - 8.48 log Q5 = _0.33 toga Na+ + 2.33 log A!(aq) + 3.67 log H4Si04 +7.33pH SILICA GLASS log Ke ., :- 1.454 then equation (8) reduces to: In most interstita1 and soil waters ( see Tables 2-4 ), a reasonable approximation is dc;/dt =(~a H+ )(4 (1- SelKe) (1- Q~nc.> = 1 (13) - k'+ (1- eq,fKt)] (10) whence equation (12) becomes: Although the activation energy for diffussion in glass is generally cited as 20 kJ mol -1, activation dCj/dt=(~a H +) energies for process involving silica loss from glass are [ 2.2•10·7 + 6.7 • JQ-IS ( Qa/K(I -1)) (14) inferred to be higher (see Fig 5,- 60 kJ moJ-1). Data obtained under hydrothermal conditions (31) yielded an For the formation of the clay mineral kaolinite to have a activation energy of 54 kl mol-l. Assuming an comparable effect on the overall rate of its formation as Arrhenian temperature dependence of rate constant, the dissolution of glass, the reaction activity coefficient these same data give a rate constant (k+) of 2.2 • 1O· 7 s · must be such that 1 when extrapolated to 25°C. For the dissolution of Jog (QI()/~) > 7.5 (15) kaolinite a rate constant of 1.24 • 10·12 mol m·:.!s~l at a Thus, an indicator of the likelihood of the formation of pH of 3 and at 80"C is recorded (32). From the a given clay mineral being rate determining can be empirical relationship of Fig 5: estimated from the parameter log (Q'I~). or from the E.,.: -1.1251og K + 68 (E+ in kJ mol ·I) (II) parameter R(l where the activation energy for the dissolution of kaolinite in R¢ = 2.3 • to-7(1- OelKe) calculated as 83 kJ mol·l. Assuming an Arrhenian - 6.7 • to-Is (1- Q.p/~) (16) temperature dependence for this reaction, the rate This is done in the next section for the formation constant at 2soc is determined as of imogolite, halloysite, and smectite under estuarine and -inferred shallow marine diagenetic conditions and 6.7 • JQ-7 mol m·2s-l, whence equation (10) reduces to also in soil waters associated with rhyolitic tephra. dq/dt =(~a H +)[ 2.2 •1Q-7 (1- O&'I(e) The chemical reaction and the equilibrium constants (log K(l) for the dissolution of the clay . - 6.7 • 10-IS ( 1- Q~K+) ) (12) minerals imogolite, halloysite, and smectite, together 8 A.P.W. Hodda'tlal. equal proportions, and if interstitia) waters are taken to range from seawater to a solution wherein the solution is saturated with respect to amorphous silica, values of the parameters Qf~Ke for glass Q~+ for clay minerals and the denied parameter ~ for the clay minerals can be calculated. The results of these calculations arc shown in Table 2 and plotted in Fig. 6. Although the plot Fig. 6B shows more dramatically that smectite is formed rapidly than imogolite once the concentration of silica starts to increase (i.e., with diagenesis), the trend is also apparent on Fig. 6A, where there is no particular assumption about the relative rate constants for '[ ] dissolution (viz, k+, k+' ). '[ ·--·--·~ -· .,-.... .·--·-·=---·--~ Similar calculations for the formation of ·' smectite, halloysite, and imogolite in a free - draining soil formed from rhyolitic tephras show that the rate of fomtation of imogolite far exceeds that of smccti!e and is slightly higher than that for halloysite (Table 3). ,". Comparison of the rate proxies (R~) between Tables 2 and 3 indicate that the rate of formation of smectite during early diagenesis may be orders of magnitude l" higher than the rate of formation of imogolite in soil j :: water. For a soil with impeded drainage the rates of I •• formation of halloysite and imogolite seem similar, but .. in the lower honzons there seems little difference ,f .." between these rates and those inferred for the " dissolution of glass (Table 4). There seems a .I " sympathetic relationship between the relative rate of J •• formation of halloysite and imogolite and the actual proportions of each of halloysite and allophane in soils (Fig.7). This suggests that the formation of the clay minerals in these soils is more a consequence of kinetics than thermodynamic stability. The inferred similar rate to that of glass dissolution may also suggest that the clay - mineral forming processes in soil are diffusionally controlled rather than surface reaction FIG. 6 Inferred precipitation rate of imogolite and controlled. smectite in estuarine and interstitial waters in marine sediments. MECHANISTIC IMPLICATIONS A. the parameter log (Qiji/K~) vs mixing extent. As noted by Lasaga (30) and shown on Fig. 5, B. the parameter R41 vs mixing ex!ent. activation energies of the dissolution of clay minerals are generally higher than those typical for transport in solution (- 20 kJ mol ·1), but less than that expected with expressions for their reaction activity coefficients for the breaks of bonds ( 160-400 kJ mot · l). The (logQ41) are given in Table I. In these expressions and cqulibrium constants for halloysite and imogolite calculations it is assumed that the dominant aluminium dissolution under acid conditions - as in the C - horizon of the poorly drained tephra derived soil - are rather species in basic solution is Al(OH)4-, in near neutral higher (7.54 and 10.08, respectively). On th~ basis of solution Al(OH)2+, and in acid solution AJ3+. Also Fig. 5, these processes could be diffussionally shown on Table 1 are the corresponding parameters controlled with activation energies of 60 and 57 kJ mol· (log Kg, log Qe) for the dissolution of silica glass. I respectively. To use this approach to predict the likely rate of Jambon (33) derived a relationship between the clay mineral formation the concentration of dissolved activation energy for diffusion (E*) in obsidian and the species in estuarine waters, interstitial waters in marine radius (r) and charge (z) of the diffusing ions, viz. sediments, and soil waters are required. If estuarine waters arc assumed to be a mixture of average river E" = 4.184 [ 8 + I 28( r· 1.34) 2 + 33 r.21( r + 1.34)] (17) water (23) and seawater, with all components mixed in In this el\pression r is the radius of the ion •. in Formation oC clays from volcanic glass 9 _ .., -: ~- ·· ... SOLl hor1ZGn (I, .... ~ " 04 -•...... I" wifti~ i ..II All I/A ., .. lrJl 28tfl 21 r12 l C• .. 1.0i ~, ...: •n"" -=~ -illl ...... •• , ...,.'·" ....'"' ... ..••1 .... D• pr11 ••lo• awtra.:e (eM) ...... •• ...... IZ•.u 3M ,.,.,. ,..., ::-~ ....."'"" ... 1~ l ::~ ...... ,., ....'"' ... . ;c.• lt-22 I ""-38 I..,... WAT£Roww:TEN11'1Q Cll GLASS lfiO.ac. ·2.lll '·J•\ ·- .. ., ·US ..... ·lAS ~~· ...... us I.ID 1.35 uo 5~0 1.25 4.115 4.10 - pH I 1.511 l .lf I . II 101 Ul ~II 1-J.M --2.il -3.13 ...... •2.97 .JI.1o4 ...12 ...II .Q ... log- u.c J.t.t 1U , .J...... &II ~I AI"'> ..., f:·-~"" ... "'~' ...... , -3.~ ·'.31 I ·3.31 .... I •3.17 -3.11 -:1.00 "'I logattqj()& ..... OGO ·2.21 .Z.0 7 •2.01 ·1.29 •2.U •2.23 ..... TABLE 5: Calculated activation energies for the GlASS loCI .C: 'I nearshore marine environment. On the basis that the rate of conversion of glass to clay minerals is a function of the solubility of the clay mineral, smectite is expected to be formed under mildly diagenetic conditions, and formed more rapidly than imogolite in soil. The derived activation energies for formation of imogolite from glass in soils are appropriate for a diffusion controlled reaction, and appear consistent with the diffusion of the tetrahedrally co-ordinated species Al[ivJ(OH)2(H2Q)+. In the marine environment, however the mechanism for all reactions appear to be surface reaction control. ACKNOWLEDGEMENTS This paper was first presented at the ;::[ . ~ International Inter INQUA Conference and Workshop I -GG~ ', on Tephrachronology, Loess and Paleopedology held ... b '~ !!:-oao~---· in Hamilton, New Zealand in February 1994. The & l ::.~ authors are grateful for the assistance of M.A. Velbel, H.J. Percival,• and K.L. Nagy in the preparation'of later ~ =:) I I '•,,, I I I I • o o • 0 I • • a 2 4 s a ~ ~ " ~ d a a M • • versions of this paper. The diagrams were/repared by IJ, •aiJophane' in soil Frank Bailey and the typescript prepare by Elaine Norton, Sydney Wright and Pam Fannin. FIG. 7 Relative rates of formation of halloysite to imogolite as a function of (A) the percentage REFERENCES of halloysite and '(B) the percentage of allophane in soils derived from rhyolitic I. Wada, K. (1987). Minerals formed and mineral tephra. Closed circles are horizons in a well - formation from volcanic ash by weathering. drained profile, open circleS are horizons in a Chemical Geology, 60, 17-28. poorly drained profile. Lines join similar 2. Mizota, C. and van Reeuwijk, L.P. (1989). Clay horizons. Arrows show expected trends. mineralogy and chemistry of soils fonned in volcanic. materials in diverse climatic regions. International Soil Reference atzd Jnfonnation Centre (V{ageningen) Soil Monograph, 2. (285 to105 kJ moJ·l) equation (17) can be amended so 3. Parfitt, R.L. .and Kimble, J.M. (1989). Conditions that for hydrated media it becomes: for formation of allophane in soils. Soil E* = 4.184 [13 + 20.6 (r- 1.34)2 + 7.5 z2/(r + 1.34)] Science Society of AmerictJ Joumtil, 53, 971- (18) 977. . 4. Dahlgren, R., Shoji, S. and Nanzyo, M. (1993). The· most likely species to be involved in any Mineralogical characteristics of volcanic ash diffisional process leading to the formation of soils. Developments in Soil. Scimce, 21, 101- halloysite or imogolite involve AI or Si. From the 104. calculated values of E* (equation 18), in Table 5 the 5. Lowe, D.J. and Percival, H.J. (1993). Clay mineralogy of tephras and associated paleosols most probable species is AI(OH)2+ having an activation and soils and hydrothermal deposits, North energy for diffusion close to that calculated on the basis Island {New Zealand]. Guide Book for New of the equilibrium constaP.t for clay mineral dissolution. Zealand Pre-Conference Field Trip FJ, lOth To be consistent with the re~ognition (35) that the International Clay Conference, Adelaide. "AI (iv) in ... allophane corresponds to the solution form 6. Ruxton, B.P. (1968). Rates of weathering of of AI from which the allopbane precipitates ", the Quaternary volcanic ash in north-eastern Papua. diffusing species must be AI[JV](OHh(H20h+. Transactions, 9th lntemational Congress ofSoil Science, 4, 367-376. CONCLUSION 7. Friedman, I., Smith, R.L. and Long, W.D. (1966). lmogolite is the kinetically and Hydration of natural glass and the formation thermodynamically favoured weathering product from of perlite. Geological Society of America rhyolitic volcanic glass in the soil-forming Bulletin, 77, 323-328. environment. However, on thermodynamic grounds 8. Fri edman, I. and Long, W.D. (1976). Hydration imogolite would also appear to be the favoured rate of obsidian. Science, 191,347-352. alteration product of rhyolitic glass deposited in' the 9. Hughes, R.E. (ed.) (1989). Currellt Directions in F01111lltion of ~lays from volcani~ glass 11 California Obsidian Studies. Berkley 24. Percival, H.J. (1985). Soil solutions, minerals, Department of Anthropology, University of and equilibria. New Zealand Soil Bureau California. B.. rkley Report, 69. 10. White, A.F. and Claassen, H.C. (1980). Kinetic 25. Percival, H.J. {1986). Data In: Joe, E.N. {comp.), model for the short-term dissolution of a Soil water characterisation studies of six soils in rhyolitic glass. Chemical Geology, :ZS, 91-109. the Waikato distrh:t, New Zealand. New I L White, A.F. (1983). Surface chemistry and Zealand Soil Bureau SWAMP Data Sheets dissolution kinetics of glassy rocks at 25°C. 1984, 1-6, New Zealand Department of Geochimica et Cosmochimica Acta, 47, 805- Scientific and Industrial Research, Lower Hutt. 815. 26. Singleton, P.L., McLeod, M. and Percival, H.J. 12. Hodder, A.P.W. (1985). Effects of composition (1989). Allophane and halloysite content and on hydration kinetics in volcanic glasses: soil solution silicon in soils fro~ rhyolitic implications for dating. Journal of Colloid and volcanic material, New Zealand. Austrolian Interface Science, 103, 4549. Joumal ofSoil Research, 21,67-77. 13. Hodder, A.P.W., Green, B.E. and Lowe, D.J. 27. Burdige, OJ. (1994). The biogeochemistry of (1990). A two-stage model for the formation manganese and iron reduction in marine of clay minerals from tephra-derived volcanic sediments. Earth Scitnce Reviews, 35, 249- glass. Clay Minerals, 25, 313-327. 284. (4. Ross, C.S. and Shannon, E.V. (1926). Minerals of 28. Sholkovitz, E. { 1973). Interstitial water chemistry bentonite and related clays and their physical of the Santa Barbara Basin Sediments. properties. Journal of the Amtrican Ceramics Geocl:imica et Cosmochimica Acta, 37, 2043- Society, 9, 77-96. 2073. • IS. Peterson, M.N.A. and Griffin, J.J. (1964). 29. Berner, R.A. {1980). Early Diagenesis : a Volcanism and clay minerals in the Thtorttical Approach. Princeton University southeastern Pacific. Joumal of Marine Presl,Princeton. Research, 22, 287-312. 30. Lasaga, A. C. {1984). Chemical kinetics of water 16. Hein, J.R. and Scholl, D.W. {1978). Diagenesis rock interactions. Journal of Geophysical and distribution of late Cenozoic volcanic Research, 89, 40094025. sediment in the South Bering Sea. Geological 31. Hawkins, D.B. (1981). Kinetics of glass Society ofAmer ica Bulletin, 89, 197-210. dissolution and zeolite formation und.:r 17. Hein, J.R., Scholl, D.W. and Miller, J. (1978). hydrothermal conditions. Clays and Clay Episodes of Aleutian Ridge explosive Minerals, 29,331-339. volcanism. Science,199, 137-141. 32. Nagy, K.L., Blum, A.E. and Lasaga, A.C. (1991). 18. Gardner, J.V., Nelson, C.S. and Baker, P.A. Dissolution and precipitation kinetics of (1986). Distribution and character of pale kaolinite at SO"C and pH3: the dependence on green laminae in sediment from Lord Howe solution saturation state. American Journal of Rise: a probable late Neogene and Quaternary Science, 291, 649 • 686. tephrostratigraphic record. Initial Reports of 33. Jambon, A. (1982). Tracer diffusion in granitic the Deep Sea DrillinR Project, 90. 1145-1159. melts: experimental results for Na, K, Rb, Cs, I 9. Charnley, H (1990). Clay Sedimentology. Ca, Sr, Ba, Ce, Eu to 1300°#C and a model of Springer, Berlin. 623pp. calculation. Joumal of Geophysical Research, 20. Millot, G. {1 970). Geology of Clays. Springer, 87, 10797-10810. Berlin, 425pp. 34. Wittaker, E.J.W. and Muntus, R. (1970). Ionic 21. Naish, T.R., Nelson, C.S. and Hodder, A.P.W. radii for use in geochenistry. Geochimica et ( 1993). Evolution of Holocene sedimentary Cosmochimica Acta, 34, 945-956 bentonite in a shallow marine embayment, .Firth 35. Childs, C.W., Parfitt, R.L. and Newman, R.H. of Thames, New Zealand. Marine Geology,. (1990). Structural studies of Silica Springs 109,267-278. atlophane. Clay Minerals, 25, 329-341. 22. Hodder, A.P.W., Naish, T.R. and Nelson, C.S. 36. Percival, H.J. {1995). Relative stabilities of {1993). A two-stage model for the formation selected clay minerals in soils based on a of smectite from detrital volcanic glass under critical selection of solubility constants. shallow marine conditions. Marine Geology, Proceedings I Oth International Clay 109, 279-285. . Conference, (Adelaide). CSIRO Publications, 23. Drever, J.I. (1982). The Geochemistry ofNatural Melbourne (in press) Waters. Prentice-Hall, Englewood Cliffs. ~ :