mineralogy and surface charge properties of two acid ... papers/pert vol. 9 (2) aug. 1986... ·...

Pertanika 9(2), 167-176 (1986)

Mineralogy and Surface Charge Properties oftwo Acid Sulfate Soils from Peninsular Malaysia

J. SHAMSHUDDIN, S. PARAMANANTHAN and NIK MOKHTARDepartment of Soil Science,

Faculty of Agriculture,Universiti Pertanian Malaysia,

43400 Serdang, Selangor, Malaysia.

Key words: Mineral; jarosite; variable charges; buffering.

ABSTRAK

Dua siri tanah asid sulfat yang diambil dari ladang kelapa sawit di Perak, SemenanjungMalaysia telah dikaji. Tanah-tanah ini dicirikan oleh pH yang rendah dan kehadiran motel kuning,yang dikenalpasti sebagai jarosit dan/atau natrojarosit. Mineral lain yang turut ada ialah kaolinit,mika, mika-smektit dan smektit. Cos negatif di atas permukaan tanah didapati bertambah denganbererti dengan kenaikan pH. Kenaikan pH itu tidak bergantung kepada kehadiran oksida dan/atauhidroksida kerana bahan-bahan tersebut terlalu sedikit di dalam tanah. Di atas kenaikan pH itu,adalah disyorkan bahawa KPK tanah asid sulfat sepatutnya ditentukan di pH tanah untuk meng-qambarkan keadaan yang sebenarnya. Disamping itujuga, tanah tanah itu menampan dengan kuatdipH 3-5. Penampanan di pH itu bersangkutan dengan kehadiran aluminium yang tinggi.

ABSTRACT

Two acid sulfate soils sampled from an oil palm estate in Perak, Peninsular Malaysia, werestudied. The soils were characterized by a low pH and the presence of yellowish mottles, which wereidentified as jarosite and/or natrojarosite. Other minerals present were kaolinite, mica, mica-smectiteand smectite. The negative charges on the soil surface were found to increase significantly with anincrease in pH. This increase was not related to the presence of oxides and/or hydroxides as thesematerials were present in very small amounts in the soils. On account of the increase in the negativecharges with an increase in pH, it is suggested that CEC of acid sulfate soils be determined at soil pHto give a better reflection of the soil properties. Further, it was found that the soils were highlybuffered at pH 3-5. The strong buffering at this pH range was related to the presence of highamounts of aluminium in the soil.

INTRODUCTION

In Malaysia, acid sulfate soils usually under-go an amelioration programme before they areused to grow oil palm, coconut and paddy (rice).These soils are characterized in the field by thepresence of yellowish mottles, with a hue of 2.5Yor yellower and a chroma of 6 or more. The pHof most of the acid sulfate soil samples below 50cm depth, is reported to be around 3 — 4 (Bloom-field and Coulter, 1973). Van Breemen (1982)

and Carson et al. (1982) attribute the soil buffer-ing at that pH range to the action of jarosite,Fe(OH) 3and kaolinite.

Acid sulfate soils are widespread, occupyingabout 110 000 ha, along the west coast of Penin-sular Malaysia (Kanapathy, 1973). Undernatural conditions, the soils are occupied bymangrove (Rhizosphere macronata) and/ornipah palm (I\/ipa frutescens). On draining,pyrite in the potential acid sulfate soils is

J. SHAMSHUDDIN, S. PARAMANANTHAN AND NIK MOKHTAR

oxidized and extreme acidity is subsequentlyproduced. The high acidity attacks the phyllosi-licates in the soils particularly the 2 : 1 minerals;1 : 1 minerals like kaolinite are more resistant toacid attack (Van Breemen, 1973).

Bloomfield and Powlson (1977) estimatethat about 200 tonnes of CaCO s are needed toneutralize the potential acidity in 1 ha of soil,containing 1% pyrite-S to a depth of 100 cm.Considering that some acid sulfate soils inMalaysia are known to contain 5% or morepyrite-S (Bloomfield et aL, 1968), it is obviouslyimpractical to reclaim the soil by liming alone.

In this experiment, some of the physico-chemical processes taking place in some commonacid sulfate soils are explained in terms of theirmineralogical and surface charge properties.The data available from this study should beuseful for the reclamation of acid sulfate soils forcrop production.

MATERIALS AND METHODS

Two soil series sampled from an oil palmestate, in Teluk Intan, Perak, PeninsularMalaysia, were selected for the study. These soilsare characterized in the field by the presence ofyellowish mottles at 50-100 cm depth. Theybelong to the Tongkang and Jawa Series and areclassified as Sulfic Tropaquepts (Parama -nanthan, 1983). The soils, which were sampledaccording to horizons, were air-dried, groundand sieved ( < 2 mm). Routine chemical analysiswas carried out,on the samples and the results ofthe analysis are given in Table 1.

The clay fraction, from the texturalanalysis, was kept for mineralogical studies.Minerals in the clay fraction were identified byX-ray diffraction (XRD) and their amounts weresubsequently estimated by ThermogravimetricAnalysis (TGA) or by Differential ScanningCalorimetry (DSC). Complementary identification of minerals was also carried out by Scanning

Electron Microscope (SEM) and transmissionelectron microscope (TEM). The yellowishmaterials scraped from the ped faces and/orvoids were subjected to X-ray diffractionanalysis.

Titration curves were prepared by titrating5 g of soil, equilibrated overnight in 50 ml INKC1, by IN KOH or 0.1N HCI, using an auto-titrator. The bases needed to raise the soil pH to5.5 and from 5.5 to 9.0 were estimated from thecurves. The negative and positive charges on thesoil surfaces at pH 3,4 ,5 and 6 were determinedby the method of Gillman and Uehara (1980).Fe3+ titration curve was prepared by titrating 50ml 0.01N FeCl^with 0.1N KOH while the Al ' +

titration curve was prepared by titrating 2 mgAls * in 50 ml IN KC1 by 0. IN KOH.

RESULTS AND DISCUSSION

Mineralogy

The X-ray diffractogram of the yellowishmottles is shown in Fig. I; clear peaks at 6.0A,5.1A, 4.25A, 3.35A, 3.11A and 3.08A arepresent. XRD peaks at 6.0A, 5.1 A, 3.1 lA ami3.08A indicate the presence of natrojarosite(NaFe^ (SOJ., (OH)fi) and/or jarosite (KFe,(SO ()v (OH)vj in the sample (De Coninck,1978). The 4.25A and 3.35 A XRD peaks are thepeaks of quartz, which are present together withthe yellowish mottles.

Fig. 1: X-ray diffractogram of untreated yellowishmottles scraped from ped faces and voids.

168 PERTANIKA VOL. 9 NO. 2, 1986

MINERALOGY AND SURFACE CHARGE PROPERTIES OF TWO ACID SULFATE SOILS

Q-

a ou 2a. o< e

bo

i?

© oo in oo

w e n > ) w — — i n n© © © © © © © ©

© © © ©

- d -

doo

~ coTj- 00

© © © ©

!N f 00-H in w-^ o —«

© ©CO

to ©

40

.

©©

oo ©

<o in

3

t o

!>• 00© ^

OO ^J O1

—' © ©

— CO O roo (N ©

© ' © © ' ©

o

o(21

©

©

101

i©

©

©

+

©

00

1©

in©

©001

00

CO©

120

i©

d

o

+

a, « si< ffl PQ

a »i «< CO PQ

£

PERTANIKA VOL. 9 NO. 2, 1986 169


Under the scanning electron microscope(Plate 1), jarosite and natrojarosite appear ascubes. These cubes are usually found in voids orold root channels and occur as clustered crystals.A similar kind of occurrence has been reportedin Canada (Mermut et aL, 1985). It is notpossible to distinguish the two forms of jarositesjust by their morphology. From the point of viewof morphology and crystal system, it has beensuggested that jarosite (natrojarosite) is a pseudo-morph of pyrite (cubic). This view has beenrejected by Mermut et al. (1985) who foundnatrojarosite to be smaller than pyrite. Theysuggested that natrojarosite was formed byrecrystallization from the solution. In nature,jarosite (natrojarosite) is believed to have beenformed by the oxidation of pyrite in the presenceof K + and Na +. The oxidation of pyrite (VanBreemen, 1982) is given by the equation: —

FeS2+ 15/40 2- Fe(OH)3+

In the presence of K +, jarosite (KFe5 ( ^(OH) ) is formed, while natrojarosite (NaFe3

(SO ) (OH) ^ is formed if Na + is present. Someof the potassium and sodium needed for thereaction can be supplied by soil mineralsreleased from weathering (acid attack) undervery acid conditions.

Fig. 2 is a XRD diffractogram of deferrified(dithionite) clay fraction from the Ap horizon ofTongkang Series. This diffractogram (Mg-saturated) shows the presence of mica (lOA, 5A,3.35 A), kaolinite (7.2 A, 3.5 7 A), gibbsite(4.85A) and quartz (4.25A). Some halloysites arealso present in the sample as shown by the trans-mission electron microscope (Plate 2). UnderTEM, halloysite is normally tubular, whilekaolinite is hexagonal in shape. Goethite isabsent as indicated by the very low amount offree iron oxide present in the soil (Table 1).

When the Mg-saturated sample was saturat-ed with glycol, the 15 A* peaks completelyexpanded to 17.5A (Fig. 2). The expansion ofthe 15A peaks to 17.5 A points to the presence ofsmectite. No mica mixed layer is present in soilsof this horizon, but mica mixed layers were

Plate 1: SEM micrographs (A and B) of octahedralshaped jarosite crystals in C horizon offawa Series; jarosite crystals are clusteredin voids.

Fig- 2: X-ray diffractogram of the treated claysample from the Ap horizon of TongkangSeries.



Plate 2: TEM micrographs showing kaolinite(hexagonal) and halloysite (tubular) fromthe soils ofjawa (A) and TongkangSeries (B).

present in the C horizon (not given). This isindicated by the presence of a ISA peak in theMg-saturated sample. The ISA peak completelyexpanded to 14.5A on glycolation, pointing tothe presence of mica-smectite. There is only aslight mineralogical difference between the Aphorizon and the C horizon.

After the sample was saturated with K \part of the 14 — 15A peaks collapsed and when itwas heated to 550°C, the whole peaks collapsed

to 10 A. These and the absence of a 16 A peak inthe glycolated sample show that the soils do notcontain chlorite and/or chlorite mixed layers.

The clays ofjawa Series gave a more or lesssimilar XRD pattern, pointing to a similar kindof mineralogy. This study indicates that some ofthe mica in the Ap horizon have been completelytransformed into smectite. Whereas, in the Chorizon, under anaerobic conditions, the mica ispartly converted to mica-smectite and partly tosmectite.

Fig. 3 gives the TGA thermograms of thestudied soils. The weight loss at 291 °C and513°C show the presence of gibbsite and kaoli-nite respectively. The amount of kaolinite wasestimated from the thermogram and is given inTable 2. The amount of gibbsite was too low forestimation by the TGA. The DSC, which is moresensitive than TGA, was therefore used toestimate the gibbsite; the estimated amount ofgibbsite in the clay fraction was small (< 1 %).

By X-ray diffraction and thermal analysis,it was found that the dominant minerals in theclay fraction of the studied soils were kaolinite,smectite, mica and mica-smectite. Knowing thepercentage of kaolinite and considering that the

400 MO'C

Fig. 3: TGA thermograms of clay samples fromJawa and Tongkang Series.



sand and silt tractions are entirely made ofquartz, it is possible then to estimate the com-bined percentage of mica, mica-smectite andsmectite (together). The data in Table 2, showsthat the combined percentage of mica, mica-smectite and smectite exceeds that of kaolinite.Mica, mica-smectite and smectite are usuallyreferred to as permanent charge minerals(Uehara and Gillman, 1980; 1981), meaningthat the charges on these minerals do not change

24

20

16

12

I6 4

Ap

significantly with the change in pH. Theirproperties are considered different from theoxides in Oxisols which tend to change with thechange in pH (Tessens and Shamshuddin, 1982;1983).

Surface Charge

The active part of the soil, that is the clayfraction, was composed almost entirely of phyllo-

Ap

Tbngkang

pH pH

Fig. 4: The change of negative charges on the soil surfaces with increasing pH.

TABLE 2Mineralogical composition of soils of Jawa and Tongkang Series (per 100 g soil)

in the Ap and C horizons

Series Horizon Silt sand Kaolinite Smectite micamica-smectite

Jawa

Tongkang

Ap

C

Ap

C

45.2

31.5

47.9

39.6

26.1

30.8

24.6

27.9

27.7

37.7

27.5

32.5

172 PERtANIKA VOL. 9 NO. 2, 1986


silicates, namely kaolinite, mica, mica-smectiteand smectite. Oxides and hydroxides, which arevariable charge minerals, were present in verysmall amounts. In spite of the dominance of thepermanent charge minerals, the negativecharges on the soil surface were found to increasewith pH. This is illustrated clearly by the soils ofTongkang and Jawa series (Fig. 4).

The negative charge at pH 6 is aboutdouble (Fig. 4) that of pH 3 (Table 3). This isnot in agreement with the current thinking onthe charge properties of phyllosilicate (Ueharaand Gillman, 1980). The increase in negativecharge with increasing pH is not due to oxides,as in the case of oxisols (Tessens and Sham-shuddin, 1982; 1983). It is therefore assumedthat the charges on the mica, mica-smectite andsmectite are not completely permanent. Some ofthe charges on these minerals are variable,possibly those located at the broken edges of theminerals. This explanation is consistent with thework of JHendershot and Lavkulich (1983), whoshowed that when the pH was increased from 3

to 7, the CEC of mica was increased by twicewhile that of kaolinite increased by four times.

The CEC (meq/100 g soil) of acid sulfatesoils is usually measured at pH 7. For the soilsstudied, the CEC was also determined at pH 7.The apparent CEC (meq/100 g clay) value at pH7 for the soils studied is more than 30 meq/100 gclay (Table 1). The apparent CEC at the soil pH(around pH 3) could be much lower. So basesaturation calculated on the basis of CEC at pH7 underestimates the true value. Fertilizerrecommendations based on this base saturationtherefore leaves much to be desired. For acidsulfate soils, it may be better to determine theCEC by unbuffered NH 4 C1, which is the CECclose to the soil pH.

Following the low oxides/hydroxidescontent, the positive charges in the soils are verylow, with less than 1 meq/100 g soil (Table 4).The values decrease gradually with increasingpH. The low oxides/hydroxides and high phyllo-silicate content are reflected by the low pH ; the

TABLE 3The differences in negative charges in acid sulfate soils at pH 3 and 6

Series Horizon - VE Charges

3 6

Increase

Ap

C

Ap

C

5.55

9.77

10.93

11.52

meq/100 g soil

10.28

19.10

15.59

20.26

4.73

9.93

4.66

8.74

Jawa

Tongkang

Series

TABLE 4pH and positive charges on the soil at pH 3, 4, 5 and 6 of two acid sulfate soils

+ VE Charges (meq/100 g soil)

Ap

C

Ap

C

3

0.67

0.42

0.81

0.65

4

0.66

0.41

0.71

0.44

5

0.55

0.21

0.69

.24

6

0.27

0.09

0.48

0.21

4.4

2.7

2.2

2.6

Jawa

Tongkang



value of pH is less than 3 in the lower horizonsx o

(Table 4). Soils with high amounts of oxides/hydroxides register much higher pH values(Uehara and Gillman, 1981; Hendershot andLavkulich, 1983).

Comparing the negative charges (Fig, 4)and the positive charges (Table 4) at pH 3, 4, 5and 6, we see that the amount of negativecharges exceed that of positive. Even at pH closeto 0, negative charges are higher than positivecharges. Hence, it is not possible to determinethe PZNC of these soils. PZNC, as defined byGillman and Uehara (1980), is the pH at whichthe amounts of negative and positive charges ofthe soil systems are equal.

An interesting feature of these soils is thatthe soils are strongly buffered at pH 3 - 5 (Fig. 5;buffering curves of Tongkang Series are notshown here). Strong buffering at this pH rangecould be due to the action of jarosite, kaoliniteand Fe(OH)s as suggested by Carson et al.(1982). For the weathered, non-acid sulfate soilsin Malaysia, it has been shown (Shamshuddinand Tessens, 1983) that aluminium controlsbuffering action occuring below pH 5.5.

The titration curves for the samples werecompared to the standard curves of Al3 + (Fig. 6).The titration curves of the soil samples followclosely that of Al3+, but not that of Fe3+. The

• tt nO.I N KOH ml

Fig. 6: Standard titration curves of Fe *+ and Al *

effect of Fe3 + is probably minimal as the amountof iron in the soil is low (Table 1).

The bases (ml) needed to raise the soil pH to5.5 were estimated from the curves and theresults were compared to the clay content,organic matter and Al present in the soils.Jarosite was excluded in the regression analysis asit was difficult to determine its amount accurate-ly by X-ray diffraction or thermal analysis.

The relationships between base (ml) and thethree factors selected is given by the equation: —

OH =0.68R2 = 0.91, 3 3

Thus, it wasraise soil pH to 5.the clay contentmost importantaction was foundtion relating OHtion: —

.19clay% + 1.14OM + 0.21A1

3 = 10.58*

shown that the base needed to5 was significantly correlated to, organic matter and Al. The

factor controlling bufferingto be Al. The regression equa-

and Al is given by the equa-

0.1 N HO 0.1 N KOH ml

Fig. 5: Titration curves of soils o/Jawa (A) andTongkang Series (B)

OH " = 5.20 + 0.26 AlR2 = 0.88, F i 5 = 37.97**

The significance of Al in controlling the soilbuffering is shown in Fig. 5 where bufferingincreases with depth following the increase of Alwith depth (Table 1).



The bases needed to raise the soil pH from5.5 to 9.0 were also estimated. The multiplelinear regression study showed that there was apoor correlation between bases and the threevariable parameters selected. In an earlier studyby Shamshuddin and Tessens (1983), for non-acid sulfate soils, it was established that claycontent (mainly kaolinite) was highly correlatedto the base. Perhaps in acid sulfate soils, othercomponents such as jarosite and pyrite or evenFe(OH)g might be involved in the bufferingaction above pH 5.5.

General Discussion

Though the soils are dominated by mica,mica-smectite and smectite, which are tradi-tionally referred to as permanent chargeminerals (Uehara and Gillman, 1980), thenegative charges in the soils were found to in-crease significantly when the pH was increased to6. Negative charges at pH 7 are then consideredto be much higher than at soil pH. As such thetraditional method of CEC determination on pH7 does not reflect the true CEC value of the soil.

When the pH is raised, more OH areadsorbed onto the surface or H * are releasedinto the solution, causing an increase in negativecharges. These reactions are much more impor-tant above pH 5.5. Below pH 5.5, most of theOH " are used to neutralize Als + in the soil.

These reactions, most probably, take placeat the broken edges of the phyllosilicates. As forkaolinite, the increase in negative chargesbecomes important only at high pH; the pHQofthe broken edges of kaolinite is around 7.3(Rand and Melton, 1975). This particular pro-perty may have some effects on soil buffering(Fig. 3). '

When lime is applied, Al5 + is initiallyneutralized by the OH " produced by thehydrolysis of CO s". Soil pH will remain below 5until Als + is fully neutralized. Luckily, hydrolysisof jarosite is minimal, as it is very insoluble(Carson et al., 1982). Natrojarosite persists evenat high pH and in the presence of CaCO s

(Mermutt et al, 1985). This provides further

proof of the low solubility of the mineral. If themineral is highly soluble, its hydrolysis will causethe pH to remain below 4 until the reaction iscomplete.

From past experience, neutralizing the totalacidity by liming alone is not economical as itrequires too much lime. Besides, overliming maybe detrimental to plant growth (Van BreemenandPons, 1978).

CONCLUSION

The dominant clay minerals in the soilsstudied are kaolinite, mica, mica-smectite andsmectite. In increasing the soil pH from 3 to 6,the negative charges on the soil surface increasedtwice, although the soils are dominated by theso-called permanent charge minerals. The re-action is thought to take place mainly at thebroken edges of the phyllosilicates. The increasein charge with increasing pH implies that thetraditional CEC determination at pH 7 does notreflect the true CEC under field conditions. Thesoils*are also strongly buffered at pH 3 - 5 . Thestrong buffering at this pH range is due to thepresence of high amounts of Al3 + in the soil.

ACKNOWLEDGEMENTS

The authors wish to record their gratitudeto Universiti Pertanian Malaysia for financialsupport during the conduct of the research. Thehelp of Mrs. Aminah in TEM and SEM and thelaboratory staff of Soil Mineralogy Section of theDepartment of Soil Science, UPM, in soilanalysis, is greatly appreciated.

REFERENCES

BLOOMFIELD, C., COULTER. J.K. and KAMARIS-SOTIRIOU, R. (1968): Oil palm on acid sulfatesoils in Malaya. Trop. Ague. 45: 289 - 300.

BLOOMFIELD, C. and COULTER, J.K. (1973): Genesisand management of acid sulfate soils. Adv. inAgron. 25:265-320.

BLOOMFIELD, C. and POWLSON, D.S. (1977). Theimprovement of acid sulfate soils for crops otherthanpadi. Malays. Agric.J. 51(1): 62-76.



CARSON, CD. , FANNING. D.S. and DIXON, J.B.(1982): Alfisols and Ultisols with acid sulfateweathering features in Texas. In: Acid SulfateWeathering. Kittrick, J.A., Fanning, D.S.Hossner, L.R. (Ed.). Sod Set. Soc. Am. 127-146pp.

D E CONINCK, F. (1978): Physico-chemical aspects ofpedogenesis. I T C , RUG., Belgium.

GlLLMAN, G.P. and UEHARA, G. (1980): Chargecharacteristics of soils with variable and perma-nent charge minerals. II: Experimental. Soil Set.Soc. Am.J. 44:934-938.

HENDERSHOT. W.H. and LAVKULICH, L.M. (1983):Effects of sesquioxide coatings of surface chargeof standard mineral and soil samples. Soil Set.Soc. Am.J. 47: 1252 - 1260.

KANAPATHY, K. (1973): Reclamation and improve-ment of acid sulphate soils in West Malaysia. In:Dost, H (Edi.). Proc. Int. Symp. Acid SulphateSoils. Wageningen. ILRI Publ., 18(2): 383-390.

MERMUT. A.R., CARTIN, D. and ROSTAD, H.D.W.(1985): Micromorphological and submicrosco-pical features related to pyrite oxidation in aninland marine shale from east Central Saskat-chewan. SoilSci. Soc. Am.J. 49: 256-261.

PARAMANANTHAN. S. (1983): Soils of PeninsularMalaysia: Soils developed on marine, estuarineand brackish water deposits, UPM-MSSS., KualaLumpur.

SHAMSHUDDIN, J. and TESSENS, E. (1983): Potentio-metric titration of acid soils from PeninsularMalaysia. Pertanika, 6(1): 71 - 76.

TESSENS, E. and SHAMSHUDDIN. J. (1982): Charac-teristics related to charges in Oxisols of Peninsular Malaysia. Pedologie, 32: 85-106.

TESSENS, E. and SHAMSHUDDIN. J. (1983): Quantitativerelationships between mineralogy and propertiesof tropical soils. Serdang, Selangor, Malaysia.UPM Press.

UEHARA, G. and GILLMAN, G.P. (1980): Chargecharacteristics of soils with variable and perma-nent charge minerals. 1: Theory. Soil Sci. Soc.Am. / 4 4 : 2 5 0 - 2 5 2 .

UEHARA. G. and GILLMAN, G.P. (1981): The Minera-logy, chemistry and physics of tropical soils withvariable charge clays. Boulder, Colorado, West-view Press.

VAN BREEMEN, N. (1973): Soil forming processes inacid sulfate soils. In: H. Dost (Edi.). Acid SulfateSoils. Proc. Int. Symp. Wageningen, Nether-lands. ILRI Publ. 18(1): 66 - 130.

VAN BREEMEN, N. (1982): Genesis, morphology andclassification of acid sulfate soils in coastal plains.In: Acid Sulfate Weathering. Kittrick, J.A.,Fanning, D.S. and Hossner, L.R. (Edi.). Soil Sci.Soc. Am. pp: 9 5 - 108.

(Received 24 February, 1986)


mineralogy and surface charge properties of two acid ... papers/pert vol. 9 (2) aug. 1986... ·...

Documents