Experimental Precipitation/Genesis Sepiolite At Earth-Surface Conditions | CHEM 247A, Lab Reports of Organic Chemistry

Download Experimental Precipitation/Genesis Sepiolite At Earth-Surface Conditions | CHEM 247A and more Lab Reports Organic Chemistry in PDF only on Docsity! THE AMERICAN MINERALOGIST, VOL. 53, SEPTEMBER_OCTOBER, 1968 EXPERIMENTAL PRECIPITATION AND GENESIS OF SEPIOLITE AT EARTH-SURFACE CONDITIONS Ror.ewo Wor.resr, Institut de Chimie Industrell,e, Unilersit/ Libre d.e Bruxel,les, Brurell,es, Belgium FnBo T. MacrrNzro, Department of Geology, North- weslern Unittersit^v. and. Evanston. Illinois 60201 AND OwBN P. Bnrcxrn, Department of Geology, The Johns H o phins U n iver s'ity, B altimore, M aryland, 2 2 I I 8 . ABsrRAcr Reaction ol aqueous silica with sea water produces a hydrated magnesium silicate similar to sepiolite in structure and composition. Our investigations indicate that sepiolite is the only non-aluminous, cation-bearing silicate that can be precipitated directly from sea water. Experiments with magnesium-free sea water and with magnesium chloride solu- tions shor,r' that cations other than Mg2+ are not involved in the precipitation of this compound. The equilibrium constant for the reaction: 2Mg2+t3SiOz.n*(n-|2)Hzo : MgrSLO8(H2O)"+4H+ has been determined experimentally at 25'C and one atmosphere total pressure and the AG"y of dehydrated sepioiite calculated to be - 903.3 + 0.5 kcal/mole. Stability relations in the system MgO-SiOrHsO at 25oC and one atmosphere total pressure are examined and models for the genesis of sepiolite in the marine environment, the saline lake environment and by chemical weathering of ultramafic rocks are discussed. INrnonucrroN The genesis of clay minerals at earth-surface conditions has aroused the interest of investigators for many years. In particular, much atten- tion has been paid to the origin of the magnesium-bearing clays, sepiolite, attapulgite, palygorskite, and magnesium montmorillonite. Recently the discovery of sepiolite and palygorskite in deep-sea sediments (Bon- atti and Joensuu, 1968; Hathaway and Sachs, 1965), the excellent work of Millot (1962a, b) on the occurrence and origin of the magnesium- bearing clays. in sedimentary rocks and that of Siffert (1962) and Siffert and Wey (1962) on the laboratory synthesis of these phases have pro- vided us with much insight into the conditions of formation of these minerals. The stability of sepiolite and some other magnesium silicates at 25oC and 1 atmosphere pressure has been discussed by Hostetler (1960); however, little is known about the conditions necessary for equilibrium between these phases and natural waters. Because of our continued interest in silicate mineral-sea water inter- actions (see references) as mechanisms for controlling the activities of chemical species in sea water, we decided to study the reactions that occur during the addition of sodium metasil icate to sea water. This addi- tion results in the precipitation of a hydrated magnesium silicate similar I 645 1646 WOLLAST-, MACKI],NZIE AND BRICKL,R in composition and structure to sepiolite. In this paper, we define the conditions necessary for precipitation of sepiolite from sea water, and determine the equilibrium constant for the precipitation reaction and the free energy of formation oI sepiolite. Also, we discuss the experimental results with regard to the formation of sepiolite in sedimentary deposits and to the chemistry of sea water. PnBvrous Wonr Recently, Hathaway and Sachs (1965) reviewed the studies pertaining to the occurrence and origin of sepiolite; we will present only a brief summary of the more recent work. Authigenic sepiolite, as well as other magnesium-bearing clay min- erals, is commonly found in Cretaceous and younger sedimentary de- posits of marine and Iacustrine origin (Bartholom6,1966). Sepiolite has been found in modern deep-sea sediments associated with clinoptilolite, serpentine, chert, and coccolith ooze (Hathaway and Sachs, 1965). In- terestingly, palygorskite has also been discovered in the deep-sea asso- ciated with the same minerals (Bonatti and Joensuu, 1968). Mil lot (1962a, b) reviewed thoroughly the literature concerned with the oc- currence and genesis of sepiolite and other magnesium-bearing silicates in sedimentary rocks, particularly those of the African and French Tertiary basins. He concluded that these minerals form during or shortly after deposition of the associated sediments by reactions involving dis- solved silica and magnesium. Hathaway and Sachs concluded that sepiolite obtained from samples dredged from the Mid-Altantic Ridge formed by the reaction of dissolved Mg with silica derived from devitri- fication of silicic volcanic ash. Bonatti and Joensuu reported that paly- gorskite in a deep-sea sediment core collected from the Barracuda Escarp- ment in the western Atlantic formed by the action of hydrothermally- derived, Mg-rich solutions on montmorillonitic clay minerals. Siffert and Wey (1962) and Siffert (1962) synthesized sepiolite at room temperature by adding NaOH to solutions saturated with amor- phous silica and containing various concentrations of MgCl2. Unfor- tunately, these authors do not report sufficient chemical data to deter- mine the equilibrium conditions between sepiolite and aqueous solution. However, they do report that sepiolite precipitation obtains when the initial pH of their solutions is greater than 10. At these pH values, bru- cite (Mg(OH)) precipitates rapidly, and the formation of sepiolite may be promoted initially by precipitation of brucite that later reacted with dissolved silica. Cole and Hueber (1956) reported the occurrence of a hydrated mag- nesium sil icate of composition 4MgO.SiOr.4H2O formed by the action PRT,CIPITATION OF SEPIOLITE 1649 rnission Laiie camera and Guinier focussing camera patterns of the ran- domly oriented precipitates were obtained on a standard Phillips-North American x-ray machine using Cu Ka radiation. Aluminum metal was used as an internal standard. Table 2 gives the X-ray powder data for the precipitates compared with data for synthetic and natural sepiolites. Comparison of the X-ray data indicates that the structure of our syn- thetic compound is similar to that of synthetic and natural sepiolite. Ifowever, the high background in the region of the 12 A and 7.5 A peaks prevented their detection. To confirm the chemical and x-ray data for Tngle 2. X-R.q,y-Po'nonn Dera lor Nerunlr, nno SvNrnnrrc Soptolrtns Natural Sepiolite Synthetic Sepiolite Siffert Epprecht (re62) (1947) This work 110 130 200,040 150 331 261 370 081 421 0. 10.0, 510 ML,28l 530 ll2, 37 l, lgl 2 . t 0 .0 r32 202,042 t .11 .0 ,222 ,461 062,312,2.10.1 620, 57O,332 4. 53 4 . 1s J . t l 3 .32 3 .20 060 131 330 260 24r 080 2 .49 2 .43 2 .36 2 . 2 4 2 .08 12.05 (100)' 7.47 ( 10) 6.7s ( s) s .01 ( 7 ) 4.498 ( 2s) 4.306 ( 40) 4.o22 ( 7) 3.7s0 ( 30) 3.s33 ( 12) 3.366 ( 30) 3.1e6 ( 3s) I 3.0s0 ( 12) 2.e32 ( 4) 2.82s ( 7) 2 .77 r ( 4 ) 2.69r ( 20) 2.617 ( s0) 2 . s86 (NR) 2.s60 ( ss) 2.47e ( s) 2.Me ( 2s) 2.406 ( 1s) 2.263 ( 30) 2.206( 3) 2 . 5 2 2 . 5 6 2 .46 2 .32 2 . 2 2 2 . l l t 2 . 1 3 / . J O 6 . 6 8 5 . U / 4. 503 4.312 4.03 3 . 7 3 5 3 .509 3.344 3. 184 2.684 2.6 t9 2 .592 2 . 5 6 r 2 . 5 1 2 2 . M 3 2.26r 1 2 . 3 t - o 4 . 9 4 . 5 4 . 3 3 . 7 5 3 . 4 9 2 . 9 8 2 . 6 7 Mid-Atlantir Lrtile uotton- KloSe cod. utah. usA ,-- , " , " - , , (Ha thaway asv & .bradlev. 1O(( \ - ' ano Sacns' 196s) * Relative intensities. NR:not resolved. 1650 WOLLAST, MACRDNZIE AND BRICKER the precipitate we also pertormed infrarecl and electron rnicroscopic investigations. Infrared analysis. lnfrared absorption spectra of the precipitate and of natural hydrated magnesium silicates were obtained from KBr disks with a double-beam spectrophotometer. The spectra obtained are com- pared in Figure 1. The absorption peaks of the precipitate are similar to those of sepiolite in agreement with the results of the chemical and x-rav data. It is important to note the absence of the sharp peak (3720 cm -r) of brucite in the spectrum of our precipitate. Brucite was not precipitated in our experiments because the titrations were carred out at pH values of about 8.0. At higher pH values, brucite and calcium carbonate precipi- tate from sea water. The broad bands at 3400 cm-l and 1630 cm-r in the precipitate spectrum indicate that the solid contains a large amount of non-structural water. Electron microscopy. An electron micrograph of the sepiolite precipi- tate is shown in Figure 2. The precipitate consists of packets of needles approximately 0.2-0.3 pn in length, and is similar in appearance to the synthetic sepiolite pictured in an electron micrograph by Siffert (1962). Composition of the aqueous phase in equilibrium wi,th sepiolite General experiments. Table 3 shorvs the composition of sea water be- fore and after addition of SiOz in the form of solutions of sodium meta- sil icate. Before addition of sil ica, the sea water contained about 0.03 ppm SiOz and 58.8 mMoles of Mg2+, and had a pH of 8.2 to 8.3. The addi- tion of SiOz resulted in the precipitation of sepiolite, and SiOz and Mgz+ were removed from the sea water solutions. The pH of the solutions de- creased slightly during reaction. The reactions were carried out until the pH and concentrations of SiOz and Mgz+ remained constant. At this point it was assumed that equilibrium between the precipitated solid and solution had been achieved, and the final pH, SiOz and Mg'+ coo- centrations were determined (Table 3). Specific experiments. Several additional experiments were carried out to clarify the sepiolite-sea water reaction. Magnesium and calcium were removed from sea water by addition of NaOH. Enough calcium was then added to the sea water in the form of Ca(OH)z to restore the cal- cium concentration to its original value, and the pH adjusted to 8.2 with hydrochloric acid. The resulting Mg-free sea water was titrated with a sodium metasil icate solution, adjusted to a pH of 8.2 by addition of hydrochloric acid, until 80 ppm of dissolved silica were added to the sea water. No solid precipitated from the sea water immediately, or upon standing for 30 days. The SiOr concentration of the solution after PRECIPITATION OF SDPIOLIT'E 1651 Fro. 1. Infrared absorption spectrum of the product precipitated from sea water by addition of sodium metasilicate compared with the spectra of some natural magnesium silicates and brucite. The snectra were obtained from KBr disks with a double-beam spectrophotometer. 30 days was 80 ppm. We conclude from this experiment that a mag- nesium silicate is the only non-aluminous, cation-bearing silicate that wil l precipitate from sea water. To confirm this conclusion and to det.ermine whether other ions are ANTIGORITE 1654 lo WOLLAST, MACKENZIE AND BRICKER ppm oq pH t 2 3 ml NoOH odded I'rc.3. Titrationcurveforadditionof0.l0MNaOHtoa0.lMMgClssolutioncontain- ing 140 ppm SiO2.n initially. 1, after one hour. 2, after 26 hours. 3, after 15 days. Dashed curve shows ppm SiOs. after 15 days. the SiOr was consumed. From Figure 3, it can be seen that the rate of reaction depends on the concentration of OH- ion. Several other NaOH titrations of MgCIz-SiOz solutions of various concentrations were carried out. Ifowever, the mechanism of precipita- tion of hydrated magnesium silicates by addition of NaOH to MgCl2-SiO2 solutions is complicated, and the initial solids are commonly richer in Mg than the sepiolite precipitate obtained from sea water. These initial Mg- rich precipitates may be metastable. Equrrrnnrulr CoNsraNr Car,currlrroN AND AGor SeprorrrB The chemical and mineralogical data show that the precipitate is sepiolite containing SiOz and MgO in the ratio of 3 to 2. Because we know the composition of the solid and the composition of the sea water in PRECIPITATION OF SEPIOLITE 1655 cquil iblium with thc soli<I, we are ahlc to calcrrlale i ln erlrri l ibrium con- stant for the sepiolite precipitation reaction at.2.5o(l and l atmosphere total pressr"rre. The general reaction lnay be written rMgz+ * ySio2.o + (n * r)H2O : (MgO)"(SiOr)u(H2O)" t 2rH+, (l) and in terms of a solid with a SiOr/MgO raLio ol3f 2 2Mgz+ + 3SiO2.q + (" * Z)HrO: MgzSi'Os(HrO)" + 4H+. (2) The equil ibrium constant for this reaction is 4 aH+ - : K l z r r ( 3 ) aMsr* aoSiO 2uq assuming both the solid and H2O have an activity of one. The values of the equilibrium constant for reaction (2) as calculated from the data in Table 3 and an activity coefficient for Mg2+yy*r+ of 0.36 (Garrels and Thomp- son, 1962) are given in Table 3. The mean value ol Ket is 10-18 78. The free energy of formation of sepiolite containing no water (AGor ""piorit") -uy be calculated from the relationship AGor"u"lioo : AGor products - AGor reactantsr (4) where AGo.un"rioo f 25oC and 1 atmosphere total pressure is AGo. - - 1 .3641o9K . (5") The AGo" for the sepiolite precipitation reaction is AGo" : - 1 .364 ( -18.78) :25.62 kcal . (5b) From equations (2) and (4), neglecting the free energy contribution of HsO, except that necessary to balance the reaction, we may write AG"" : AGor ""piot i t" * 44Got"* - 24Go 1**'+ - 34Gor sio2aq - 24Gor nro, and from 5b and free energy values of Rossini, et aI. (1952) 2 5 ' 6 2 : A G o r s e p i o r t e + 4 ( 0 ) - 2 ( - 1 0 8 . 9 9 ) - 3 ( - 1 9 9 . 2 ) - 2 ( - 5 6 . 6 9 ) AGol " "o ;o1 ; * : - 903.3 * 0 .5 kca l /mo le . This value is for dehydrated sepiolite; the amount of HzO in the sepiolite structure is variable and thus the free energy of formation of natural sepiolites varies somewhat. However, we may calculate AGoi ,uoio1i1u for the theoretical composition of the structure of sepiolite proposed by 1656 WOLLAST, MACKENZIL, AND BRICKER Brauner and Preisinger (1956) to provide an estimate of the AGol of na- tural sepiolites. The theoretical composition is 8HzO'Mgs(HzO)r(OH)a SirrOao. According to the general reaction (1), the reaction for precipita- tion of this solid from sea water is written 8Mg'9+ f 1 2 SiO 2^q+ 22H2O : 8HzO . M gs(HrO) 4 (OH) nSirrOr0+ 1 6H+. The equilibrium constant is l 6 &H+ K I : : x i r , : 1 0 - 7 5 1 2 , 8 1 2 oMe2+agiozoq and the free energy of formation for the theoretical sepiolite is AGof sepiol r tc : AG"o * 8AGor r , re*, * 124Go1s16roo - 164Go6+ : 4(25 '62) + 8(-108.99) composition of I 22AGom,o + r2(-ree.z) + 22(-56.69) : - 4 ,107.0 kcal /mole. DrscussroN The results of the experiments and calculations reported above provide us with the information necessary to draw some conclusions concerning the genesis of sepiolite at earth-surface conditions. Figure 4 shows an activity diagram for the system MgO-SiOz-HzO at 25"C, unit activity of HzO and one atmosphere total pressure. The diagram was constructed according to the techniques given in Garrels and Christ (1966) using the AGor sepiolite calculated in this paper, the AGor u.,cite (Hostetler, 1963), and a AGor of amorphous sil ica consistent with a solubil ity of 115 ppm dissolved sil ica (Morey, Fournier, and Rowe, 1964). The diagram il lus- trates well the influence of lH on the sepiolite-sea water equil ibrium. Less sil ica is required to precipitate sepiolite from moderately alkaline sea waters than from moderately acid waters. If si l ica is added to sea water and the pH maintained at a relatively low value, sepiolite does not form, even metastably; instead the sea water becomes saturated with re- spect to amorphous silica. We also have plotted in Figure 4 the compositions of some natural waters that are of interest to the problem of sepiolite genesis as functions of the concentrations of Mgz+, H+, and SiOr. Surface sea waters and most fresh waters (average chemical compositions of streams and of the Great Lakes are given as examples) are undersaturated with respect to sepiolite; however, some saline waters are saturated or oversaturated. PRECIPITATION OF SEPIOLITE the area representing marine interstitial water compositions. Some of these waters are saturated with respect to sepiolite. However, cold bot- tom waters are not saturated; their sil ica concentrations are too low. In cold marine interstitial waters containing moderately high dissolved silica concentrations due to the dissolution of diatoms or the devitrification of volcanic glass, sepiolite may precipitate directly from aqueous solution by the reaction presented previously. The solubility of sepiolite probably decreases with decreasing temperaturel thus, the increased thermody- namic stability of sepiolite with decreasing temperature could promote the formation of sepiolite in the cold, interstitial waters of deep-sea sediments. However, if aluminum is available in reactive aluminous phases in the sediment a Mg-aluminosilicate (chlorite-like clay minerals) will form instead of sepiolite. Experimental and theoretical work (e.g. Sil l6n, 1961; Holland, 1965; Mackenzie, et al., 1965, 1966, t967) suggests that cation-aluminosilicates instead of nonaluminous silicates are stable in the marine environment under most conditions. The scarcity of sepio- Iite in modern and ancient marine deposits and the relative abundance of chlorite in these same types of deposits substantiate this conclusion. The mechanism proposed above for the formation of sepiolite in marine sediments should not be overgeneralized. Sepiolite commonly is found in Iimestones and calcite marls of presumed marine origin, particularly from the Cretaceous and Tertiary periods. Some of these carbonate sediments probably formed in small, shallow-water, marine basins suffi.ciently isolated from communication with open-ocean water so that the waters in these basins became highly saline. In this case, if silica were not removed from solution by organisms, sepiolite could precipitate from the con- centrated sea water. Shallow marine waters often have higher pH values during the day because of the photosynthetic activity of marine plants; such diurnal increases in pH could initiate the plecipitation of sepiolite. Alternatively, some of these carbonate sediments were probably de- posited in normal salinity sea waters, and the sepiolite in them may have formed as described above for deep-sea sediments. Interestingly, the common association of sepiolite with carbonate sediments that are low in aluminum, but not with shales that are high in aluminum and commonly contain chlorite, suggests that sepiolite forms in the carbonate-depositing environment because of the lack of aluminum. Genesis of sepiolite by the chemical weothering of ultramaf,c rocfts. Sepiolite is found associated with serpentine (e.g. Cailldre and Henin, 1949), and is believed to be an alteration product of mafic minerals. Chemical analyses of several ground waters derived from the chemical weathering of ultra- mafic rocks containing primarily olivine, pyroxene, and serpentine are 1660 WOLLAST, MACKENZIE AND BRICKER shown in Figure 4. The ultrabasic waters with pH values greater than 11 are oversaturated with respect to sepiolite; the slightly alkaline mag- nesium bicarbonate waters are only slightly undersaturated. The origin of these waters is discussed in detail by Barnes, et oI. (1967). The chemical weathering of ultramafic rocks may lead to the formation of brucite and magnesium carbonates, as well as a spectrum of magnesium silicates. The specific weathering reactions and the compositional changes involving mafic minerals and weathering solutions are not well known. However, it is obvious from Figure 4 thaL sepiolite could form by reactions involving mafic minerals, such as olivine, and weathering solutions. These reactions could conceivably result in the production of ground waters saturated with respect to sepiolite. Suuuenv The behavior of magnesium in the geochemical cycle has been of in- creasing interest to geologists, particularly with respect to sea water- sediment equilibria. We investigated the conditions under which the magnesium silicate sepiolite precipitates from sea water, determined the equilibrium constant for the reaction at 25oC and one atmosphere total pressure and calculated a free energy of formation for sepiolite. Addi- tional experimental work in magnesium-free sea water and in magnesium chloride solutions indicate that: (1) sepiolite is the only nonaluminous, cation-bearing silicate that will precipitate directly from sea water as the dissolved silica concentration is increased, (2) other ions present in sea water (aside from Mg2+, SiO2nq, and OH-) are not necessary for the formation of sepiolite, and 3) depression of the pH of sea water prevents the precipitation of sepiolite and allows the concentration of silica (in the absence of other regulatory mechanisms) to increase to saturation with respect to amorphous silica. The genesis of sepiolite under some common environmental conditions and its significance as an environmental indicator have been discussed. Sepiolite occurs in saline andf or alkaline Iakes, as a weathering product of ultramafic rocks, and in the marine environment. The most important factors in the genesis of sepiolite are: (1) absence of aqueous aluminum species and of reactive aluminous solid phases, and activities of Mg2+, OH-, and SiOz consistent with the equilibrium constant for sepiolite (e.g. moderately high silica concentrations and Iow a,yqrr+f afi+; moderately low silica concentrations and high a.y""+fafi*). In the presence of al- uminum, magnesium will be incorporated into aluminosilicate structures such as chlorite instead of precipitating as sepiolite. If there is biologic (e.g. diatoms) or mineralogic (e.g. clay minerals) control that keeps silica concentration low, sepiolite will not precipitate. If the pH of the en- PRECI PITATION OF SIiPTLITE 1661 v i ronment i s n ra i r r ta ined in the s l igh t l l ' ac id ic to ac id ic lange, sep io l i te wil l not precipitate and si l i t :a ( in the absence of anothel regtt latory mechanism) wil l increase to saturation with respect to amorphous si l ica. AcKNoWLEDGEImNTS We wish to thank R. M. Garrels of Northwestern University for his helpful suggestions and critical review of this manuscript. This work was supported by NSF Grants GP-4140, GA-828 and the Petroleum Research Fund of the American Chemical Society. Contribution no. 439, Bermuda Biological Station for I{esearch, St. George's lVest, Bermuda I{EIERENcEs BARNES, I., V. C. LeMencnn, Jn., elro G. Hrunnrnrne (1967) Geochemical evidence of present-day serpentinization . Sc ienc e, 156, (3776), p 830-832. B.l,nrnor,ou6, P. (1966) Sur I'abondance de la clolomite et de la sepiolite dans les series sedimentaires. Chmt.. GeoL,l, p. 33-48. BoNArrr, E., .rNo O. JoENSUU (1968) Palygorskite from Atlantic deep sea sediments. Amer. MineraL,53, 975-983. Bn,ruNnx, K., e,No A. Pr.ersruoun (1956) Struktur und Entstehung des Sepioliths. Tscher- mah's Mineral,. Petrogr. Mitt., q p. 120-140. Bntnorrv, G. W. (1959) X-ray and electron diffraction data for sepiolite. Amer. Minerd., .14, 495-500. Cer,r,rinn, S. (1936) Contribution d 1'6tude des min6raux des serpentines. Bul'|. Soc. Franc. Mi,nerd. ,59, p. 163 326. S. HnNrx (1949) Occurrence of sepiolite in the Lizard serpentines. Notwre,163, p. 962. Cor.o, W. -F , aNl H V. Hursen (1956) Hydrated magnesium silicates and aluminates formed synthetically and by action of sea water on conctete. Sil'icates Intl.,3l, p. 75-85. Errnocur, W. (1947) Versuche zur Synthese von Serpentin-Mineralien. Schweiz. Minera.l'. Petrogr. Milt., l, p. l-2O. Gmnrr-s, R. M., .rNr C. L. Cnms:r (1966) Solulions, Minerals, ond. F,quil'ibria Harper and Row, New York,450 p. -- AND F. T. M,c,crnNzrn (1967) Origin of the chemical compositions of some springs and lakes. 1z R. F. Gould, (ed.), Equilibriwm Concepts in Nalural Water Systems, Ailronees i,n Chemi.stry Series, No.67, Amer. Chem. Soc., Washington, D. C., p.222- 242. - AND M. E. TrroupsoN (1962) A chemical model for sea water at 25"C and one atmosphere total pressure. Amer. f. Sci.,260, p. 57-66. Hl:urlwlv, J. C., lNn P. L. S,lcns (1965) Sepiolite and clinoptilolite from the mid-Atlantic Ridge. Amer. Mineral , 50, p 852-867. Hor,r.lNo, H. D. (1965) The history of ocean water and its efiect on the chemistry of the atmosphere. Proc. Nat. Acad. Sci. U. S., 53, (6), p. 1173-1183. Hosrnrr,rn, P. B. (1960) Low temperature relal.i.ons i.n the systan: MgO-SiOz-COz-HzO. Ph.D. Diss , Harvard University. - (1963) The stability and surface energy of brucite in water at 25"C. Amer. J. Sci., 261, p. 238-258. LtvrNcstont, D. A., (1963) Chemical composition of rivers and lakes. In, (ed), M. Fleis- cher, Data of Geochemistry, U.S. GeoI. Suru. ProJ. Pap.,44O,p. l-64. MacrtNzrr, F. T., ,r.Nr R. M. Gannr'r,s (1965) Silicates: Reactivity with sea water. Science, 150, (3692), p. 57-58.