TECHNICAL PROGRESS REPORT 125 SEPTEMBER 1960

DISTRIBUTION OF GIBBSITE AND KAOUNITE

WITH DEPTH IN A GIBBSITIC SOIL ON KAUAI

W. E. Holmes,

Makoto Takahashi,

and

G. Donald Sherman

HAWAII AGRICULTURAL EXPERIMENT STATION, UNIVERSITY OF HAWAII

TECHNICAL PROGRESS REPORT 125 SEPTEMBER 1960

HAWAII AGRICULTURAL EXPERIMENT STATION, UNIVERSITY OF HAWAII

Honolulu, Hawaii

DISTRIBUTION OF GIBBSITE AND KAOLINITE

WITH DEPTH IN A GIBBSITIC SOIL ON KAUAI

w. E. Holmes, * Makoto Takahas.h.i,t and G. Donald Sherman+

Gibbsitic soils of Hawaii have been described by Shennan (2). Thesesoil s are located on all major islands, and they have received serious con­sideration by mining companies as possible sources of bauxite. Becausesuch soils may eventually be mined, the University of Hawaii was author­i zed by the 1957 Terri tori al Legislature to conduct experim-ents in the re­vegetation of a simulated stripmined area. An area was selected in theWailua Game Refuge on Kauai as an experimental area. In order to evaluatean appropriate depth at which to termin ate the simulated mining operations,analytical data were needed. These data were obtained from borings takenfrom the experimental area prior to starting the excavation. The purpose ofthis report is to describe the chemical and mineral composition oIbhebauxite deposits on the island of Kauai.

EXPERIMENTAL PROCEDURES

Sampling: Thirteen borings were made over the experimental area. Theborings were made after 1.0 to 1.5 feet of topsoil had been removed. A 3­inch diameter, hand-operated soil auger was used. Samples were taken at 1-

* Formerly Assistant Soil Scientist at the Hawaii Agricultural ExperimentStation and Assistant Professor of Agriculture, University of Hawaii.

t Assistant Agronomist at the Hawaii Agricultural Experiment Station.

:j: Senior Soil Scientist, Hawaii Agricultural Experiment Station; Head,Department of Agronomy and Soil Science, College of Tropical Agricul­ture; and Senior Professor 0 f Agriculture, University of H.awaii.

3

foot intervals to depths. ranging from 15 to 20 feet. A single boring wastaken to a depth of 35 feet, of which the last 20 feet were done after theexcavation of 14 feet. E a ch sample was placed in a plastic sack for ship­ment to the laboratory of the Department of Agronomy and Soil Science, 'foranalysis.

Upon arrival at the laboratory, each sample was thoroughl y mixed, anda small representative portion of the sample was obtained by, successivequarterin gs , This small sample was dried at 600C., ground to pass througha 60-mesh sieve, then placed in a labelled vial to be saved for differentialthermal analysis. The major portion of the initial moist sample was returnedto the plastic sack to be preserved for a sieving analysis.

Differential thermal analysis: A portable differential thermal apparatusdesigned for rapid anal yses was used. The procedure differed from that de­scribed by Norton (1) in that the rapid procedure made use of a fast, nonuni­form heating rate. The temperature range used was from room temperature toabout 650 0 C. This range permitted observing the characteristic endothermicreactions of gibbsite and kaolinite with the apparatus employed. A sampleof bauxite (gibbsite) designated as Dana 261 was used as a referencesample for estimating the percentages of gibbsite. The reference sample forestimating the perc en tage s of kaolinite was labelled Dana 492. The originsof the gibbsite and kaolinite reference samples were, respectivel y, LittleRock, Arkansas, and the McNamee Mine in South Carolina.

Sieving analyses: These analyses were conducted to determine thedistribution of gibbsite and kaolinite with respect to size of the soil parti­cles and agwegates. The moist samples saved after sampling for differentialthermal analyses were used. In order to have sufficient sample for sievinganalyses, the samples were composited such that each sample used forsieving represented increments of depth ranging from 3 to 6 feet.

For each si eving anal ysis a sampl e of some 500 to 600 grams of moistsoil was. used. From this, three samples ranging from 30 to 50 grams weretaken for a moisture determination. The moisture determination was neces­sary for making all cal culations on a dry weight basis. Prior to sieving,each sample was stirred in an aqueous suspension for 20 minutes with alaboratory stirring device. A series of sieves which included the 9-, 16-,32-, 60-, 115-, and 200-meshes per inch sizes were used, but only the dataof the 60-mesh sieve are used in this report. The material remaining on eachof these sieves was collected, dried at 1050 C., and weighed. The per­centages retained on each sieve were calculated using the estimated ovendry weight of the initial moist sample. This estimate was made using theinfonnation gained from a moisture determination. Small portions of eachdried sample were prepared for differential th erm al analysis in order toobserve the distribution of gibbsite and kaolinite in each size fraction.

"

RESULTS AND DISCUSSION

Variability in distribution of gibbs ite and kaolinite: The data obtainedfrom the differential thermal analysis are presented in table 1. Although thetwo sites studied are part of the same ridge and are essentially contiguous,being separated by a narrow neck about 300 feet long, gross inspection ofthe data between the two sites indicates distinct differences in the concen­tration and depth of gibbsite deposits. Subsequent stripping operations indi­cated some differences in the ore body. Some difference was noted in thecharacter of the weathered rocks. In the East site, nodules and fragments ofgibbsite were frequently encountered over the entire area but none was foundin the West site.

Because of the apparent differences between the two sites, data wererecorded and treated separately. Data on differential thermal analyses ofaverages of five borings taken from the East site and the averages of eightborings taken from the West site are presented in table 1. Since many of thereports on bauxite research present their data in terms of alumina (AI203)instead of gibbsite, the alumina equivalent has also been included in thetable. The following conversion factor was used: 100% gibbsite = 65%alumina.

In the East site the content of gibbsite increases with increase in depthup to the 8-foot depth and thereafter there is a steady decline with someminor fluctuations. On the other hand, in the West site gibbsite content re­mains at relatively uniform level up to the 8-foot depth and diminishes there­after with increase in depth. With the exception of the first foot of sampling,the East site has materially higher gibbsite content for all levels of depththan those of corresponding depths in the West si teo In nine of the foot-inter-­val sampl es, those from th e East site had doubl e th e gibbsite content ascompared to the corresponding samples from the \Vest site. The differencesin gibbsite content for corresponding depths for the two sites are presentedIn table 3. -

Individual borings highly variable: Marked variations were noted be­tween borings. The extent of variability between the different borings can bereadily seen by inspection of figure 1, in which th-e first three borings takenfrom the East site were plotted out individually. The case history of boringNo.3 is especially an interesting one. At the 6-foot depth it recorded thehighest gibbsite content of 52 percent out of a total 244 readings taken.However, from the 7-foot depth on, the content of gibbsite dropped abruptlyand none was recorded from the 12- to the 18-foot depth.

Because of the great vari ab i l ity in individual borings, for assaying anyprospective individual bauxite ore body site, a minimum of three boringsshould be taken.

5

52.5

45

315

WJ-(f) 30(D(D

(!)

~22.5zw(,)

a:: 15wQ.

~"'I

",I, ~, I, I, ,

I I, I

,r\ !/~I '{ \I !.I ,

/~ ,0'\ 'I '\ I, \ :I ! " :, \,, \,, \,,

nI \I \I \

\\ ~\ I \\ I \\ I \

\ r\ I\ I

~\\\\ p~

I3

BORING .. ,BORING #2BORING .#-3

I I6 9

DEPTH

Figure 1. Distribution pattern of gibbsite in the first three borings taken from theEast site.

Kaolinite-its distribution and relation to gibbsite: Examination of datapresented in table 1 clearly indicates that the content of kaolinite increaseswith depth. In general, kaolinite content increased very slowly for the first 8to 10 feet of depth but from 8 to 10 feet to the 18-foot depth it increasedprogressively with each increase in depth. Examination of the analyses ofgibbsite and kaolinite indicates high correlation. The precise rel ationshipbetween these two components was explored statistically by working out theregression equation. The results of the statistical analysis show a linearregression relationship between kaolinite and gibbsite.

6

TABLE 1. Tabulation of gibbsite content and percentage difference of two sites atvarying depths from 1 to 18 feet

Depth Percent gibbsite, Percent gibbsite, Di fference Percent

in feet East site West site between two increase orsites decrease in

gibbsitecontent - Eastover West site

1 15.6 18.6 - 3.0 - 10.62 23.8 20.6 + 3.2 + 15.5

3 24.8 20.0 + 4.8 + 24.0

4 33.4 25.5 + 7.9 + 31.1

5 29.8 20.2 + 9.6 + 47.3

6 37.0 18.6 + 18.4 + 98.3

7 39.4 21.0 + 18.4 + 87.6

8 29.4 14.0 + 15.4 + 102.3

9 24.6 10.7 + 13.9 + 129.1

10 30.8 13.5 + 17.3 + 127.9

11 17.8 10.3 + 7.5 + 73.4

12 17.0 10.1 + 6.9 + ss.o13 23.2 10.1 + 13.1 + 130.3

14 14.2 9.6 + 4.6 + 47.3

15 10.2 9.8 + 0.4 + 4.0

16 20.7 6.6 + 14.1 + 121.3

17 15.7 3.3 + 12.4 +254.9

18 14.0 5.3 + 8.7 + 96.3

Distribution of gibbsite and kaolinite in various size fractions: Datapresented in table 2 show the distribution of gibbsite and kaolinite in thefractions greater and less than 60-mesh in size. The data are from 4 of the13 borings made in the two excavation areas. The percentages of gibbsiteand kaolinite are weighted averages calculated from data presented in theappendix.

Within the upper 10 to 12 feet of the four borings, an average of 41 per­cent of the material was retained on a 60-mesh sieve. This average figuremay be somewhat high for samples from this area because of an unusuallyhigh value of 61 percent observed in hole No.6 of the West excavation site.The average of the other five borings is but 34 percent, The average gibb­site content in the greater than 60-mesh fraction was found to be 45 per cent,which compares to 25 percent observed in the unfractionated soil. Kaolinitein the greater than 6Q-mesh fraction was essentially the same as that of theunfractionated soil. An average gibbsite content of only 21 percent wasfound in the less than 60-mesh fraction, and the kaolinite content also wasnot essentially different from that of the unfractionated eoil; i.e., 2 to 3percen t,

7

T ABL E 2. Distribution of gibbsite and kaolinite in the size fractions greaterand less than 60-mesh

Dep th, feet Fraction Percentage 0 f

to tal sampl e

Percent

gibbsite

Percent

kaolinite

3 - 11

12 - 20

1 - 10

11 - 20

1 - 10

11 - 19

1 - 8

9 - 20

60 mesh60 mesh60 mesh60 mesh

60 mesh60 mesh60 mesh60 mesh

60 mesh60 mesh60 mesh60 mesh

60 mesh60 mesh60 mesh60 mesh

Hole No.1, East excavation

27672177

Hole No.4, East excavation

31662670

Hole No.5, East excavation

45552476

Hole No.6, West excavation

61365643

671843

7

3812

45

44336928

30

2084

oo

1216

6125954

oo16

oo

3230

Averages of four holes, fractions

1 - 10

10 - 20

60 mesh

60 mesh60 mesh60 mesh

41553267

45213111

2

32627

1 - 10

11 - 20

Averages of unfractionated soil

Entire soil 25

Entire soil 17

8

3

29

T ABL E 3. Che mical ana ly ses of a Kapaa si l ty clay profil e wi thin th e reclamationproj ect, Wai lua, Ka uai

Depth, inch es Si lic a (Si02), per c ent Alumina (AI203)' per cent

o - 10 6.58 29.7 610 - 15 1.85 3 1.4415 - 30 0.90 37.4430 - 69 0.69 42.3269 - 74 0.70 40.1674 - 77 0..73 4 1. 3277 - 93 0.67 42. 1693 - 103 0.72 40.40

103- 104 0.59 37.60104- 128 2. 18 29. 12128 - 130 4.27 31. 20130 - 136 5.7 1 28.48136 - 146 18. 50 29. 04146 - 150 8.85 40.48150 - 160 2 1.79 29.28160 - 170 9.0 2 45.9 2

Within the 10- to 20- foot depth, the average gibbsite content of thegreater than 60-mesh fraction was 31 percent, whi ch comp.ares to 17 percentfor the average of the unfractionated soil from th at depth. Kaolinite content,however, in the greater than 60-mesh fraction was found to be 26 percent,which is not essentiall y different from that of either the Iess than 60-meshfraction or the unfractionated soil.

Comparison of differential thermal data wi th those from chemical analy­

ses. Table 3 presents data from a single sampling in the same area fromwhich samples for the differential thermal data we re obtain ed. The samplesused for the anal yses represented in ta ble 3 are not the same as th ose usedfor the differential thermal data presented in table 1, but are taken from anadjacent site. In the differential thermal data of table 1 the percentages ofkaolinite and gibbsite are, respectivel y, measures of the percentages ofsilica (Si02) and alumina (AI203). Data in table 3 are expressed as silicaand alumina. For purposes of comparison, kao i in ite is 46 percent silica and40 percent alumina.

Keeping in mind that gibbsite is but 65 percent a lumina , a comparisonof the data of table 3 with those of table 1 shows that at depths up to 10feet the differential thermal data tend to underestimate th e percentage ofalumina. This conclusion that the differential thermal data tend to under­estimate the percentage of alumina appears to be justified because, withinthe top 10 feet, several of the samples of table 3 anal yzed 40 percent alu­mina. This corresponds to a gibbsite percentage of 60, which is higher than

9

any of the samples reported in table 1. The alumina contributed by kaoliniteis ne gl igi hl e at depths less than 10 feet. Also, the data of table 3 show alow silica content at depths less than 10 feet. This is in accord with thelow kaolinite perc en tage s shown at similar depths as shown in table 1.Thus, even though the differential thermal data differ from those of chemicalanalyses, the same trends are shown; namely, low silica or kaolinite ac­companied by high alumina or ~ibbsite at depths less than 10 feet andincreasing silica or kaolinite accompanied with decreasing alumina orgi bbsl te at lower depths. Subsequent investigation indicates that a largeproportion of the alumina exists as hydrated ferruginous-alumina gel and thusis not reflected in the differential analysis.

SUMMARY

Borings were made in a gibbsitic area within the Wailua Game Refuge ofKauai, The area was excavated for an experiment in revegetation of asimulated stripmined area.

Differential thermal analyses of soils from the borings showed thatwithin the first 10 feet, gibbsite predominates over kaolinite, but between 10and 20 feet, there is a gradual decrease in ~ibbsite which is accompanied bya sharp increase in kaolinite. Also within the first 10 feet, material retainedon a 60-mesh sieve is substantially higher in ~ibbsite than is either theless than 60-mesh fraction or the unfractionated soil. Within this depth (1 to10 feet) the content of kaolinite in the fractions and in the unfra ctionate dsoil averaged but 2 to 3 percent.

Within the 10- to 20-foot depth, ~ibbsite was higher in the grea ter than60-mesh fraction than in either the less than 60-mesh fraction or the un­fractionated soil. However, kaolinite content of this fraction did not differmaterially from that of the unfractionated soil.

Chemical analyses showed that the gi bbaite percentages obtained fromrapid differential thermal procedures underestimated the alumina present inthe 1- to 10-foot depths and that its values could not be used to evaluate theore for commercial purposes. These data indicate a large portion of thealumina is in a hydrated amorphous state, probably a ferrugin ou s-al uminou s~el.

LITERATURE CITED

1. Norton, F. H. 1939. Critical study of the differential thermal method forthe identification of clay minerals. Jour. Amer, Ceramic Soc. 22:54-63.

2. Sherman, G. Donald. 1958. Gibbsite-rich soils of the Hawaiian Islands.Hawaii AW. Expt. Sta, Bul. 116.

10

AP

PE

ND

IXLa

Tab

le1.

Dif

fere

nti

alth

enn

ald

ata

from

fiv

eb

ori

ng

sin

the

East

exca

vat

ion

site

Hol

eN

o.

1H

ole

No.

2H

ole

No.

3

Dep

th,

Sam

ple

Gib

bsi

te,

Ka o

lin

ite,

Sam

pIe

Gib

bsi

te,

Kao

lin

ite,

Sam

ple

Gib

bsi

te,

Kao

lin

ite,

feet

num

ber

per

cent

per

cen

tnu

mbe

rp

erce

nt

per

cen

tn

um

ber

per

c en

tp

erc

ent

I58

-1

811

058

-84

818

058

-9

88

110

258

-28

200

58-8

58

290

58-

9913

190

358

-3

816

058

-86

831

058

-10

08

310

458

-4

820

058

-87

84

10

58-1

018

210

558

-5

827

058

-88

837

058

-102

836

06

58-

68

290

58-8

98

420

58-1

03

85

20

758

-7

837

058

-90

B38

058

-104

817

08

58-

88

250

58-9

1B50

058

-105

814

09

58-

98

270

58-9

2B27

058

-10

68

1714

1058

-10

829

058

-93

B38

058

-10

78

158

1158

-118

1410

58-9

48

170

58-1

08B

935

12

58-3

98

1410

58-9

58

275

58-1

0 98

03

1....

13

58-4

08

465

58

-96

813

1358

-11

08

0.....

1458

-41

832

1058

-97B

1514

58-1

11

80

2215

58-4

2B20

1058

-11

213

05

016

58-4

38

1720

58-ll

3B0

40

1758

-44

812

2058

-114

B0

2418

58-4

513

1420

58-1

15

80

54

1958

-46

814

2020

58-4

7B15

40

21

58-4

88

05

222

58-4

98

74

32

358

-50

1311

2324

58-5

1B

1023

2558

-52

80

4326

58-5

3B

035

2758

-54

B0

5328

58-5

585

4229

58-5

6B8

4030

58-5

7B31

2131

58-5

8B35

18~2

58-5

9B5

7933

58

-60

83

4834

58-6

1B0

63

APPENDIX 1b

Table 1. (Continued)

Hol e No . 4 Hoi e No. 5

Depth , Sampl e Gibbs i te, Kaolinit e, Samp le Ci bbsite, Kaolinit e,

fe et number p erc ent p er cen t numb er p er cent p ercent

1 58-227B ~6 0 58-245B 12 02 58-228B 33 0 58- 246B 18 03 58-229B ' 23 0 58-247B 23 04 58-230B 42 7 58-248B 43 05 58-231B 24 12 58-249 B 25 06 58-232B 23 9 58-250B 39 07 58-233B 45 9 58-251B 60 08 58-234B 13 18 58-252B 45 09 58-235B 13 21 58-253B 39 0

10 58-236B 22 47 58-254B 50 011 58-237B 6 73 58-255B 43 012 58-238B 5 41 5~-256B 39 213 58-239B 24 69 58-257B 33 014 58-240B 0 9 5 58-258B 24 315 58-241B 3 44 58-259 B 18 016 58-242B 6 72 58-260B 39 617 58-2 43B 2 88 58-261B 33 1018 58-244B 3 4 8 58-262B 25 019 58-263B 26 16

12

APP

EN

DIX

2a

Tab

le2.

Dif

fere

nti

alth

erm

ald

ata

from

eig

ht

bo

rin

gs

inth

eW

est

exca

vat

ion

site

I--'

Col

j

Dep

th,

feet 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

Sam

ple

nu

mb

er

58-2

64B

58

-26

5B

58

-26

6B

58

-26

7B

58-2

68B

58-2

69B

58

-27

0B

58-2

71B

58-2

72B

58

-27

3B

58

-27

4B

58-1

93B

58

-19

4B

58-1

95B

58

-19

6B

58-1

97B

58

-19

8B

58-1

99B

58

-20

0B

Ho

leN

o.

Gib

bsi

te,

per

cen

t

23 24 27 46 23 29 23 17 19 24 15 21 14 3 11 3 3 7

Kao

lin

ite,

per

cen

t

o o o 2 o o 3 3 8 15 4 921 30 24 63 33 41

Sam

ple

num

ber

58-1

83B

58-1

84B

58-1

85B

58-1

86B

58

-18

7B

58-1

88B

58-1

89B

58-1

90B

58-1

91B

58-1

92B

58-1

75B

58-1

76B

58

-17

7B

58-1

78B

58-1

79B

58-1

80B

58-1

81B

58-1

82B

Hol

eN

o.2

Gib

bsi

te,

per

cen

t

15 13 31 18 16 21 17 12 14 17 6 1 2 3 8 6 19

Kao

lin

ite,

per

cen

t

o o o 8 21

2 9 3 10 24 29 22 56 65 31 69o

Sam

ple

nu

mb

er

58-3

19B

58-3

20B

58-3

21B

58-3

22B

58-3

23B

58-3

24B

58-3

25B

58-3

26B

58-3

27B

58-3

28B

58-3

29B

58-1

30B

58-1

31B

58-1

32B

58-1

33B

58-1

34B

58-1

35B

58-1

36B

58-1

37B

58-1

38B

Hol

eN

o.3

Gib

bsi

te,

per

cen

t

15 19 15 14 12 14 13 6 11 13 11 15 21 21 9 7 5 o o o

Kao

lin

ite,

per

cen

t

o o o o o o 2 21 15 22 16 29 21 9 49 25 53 51

AP

PE

ND

IX2b

Tab

le2.

(Con

tinu

ed)

Hol

eN

o.4

Hol

eN

o.5

Hol

eN

o.6

Dep

th,

Sam

ple

Gib

bsi

te,

Kao

lini

te,

Sam

ple

Gib

bsi

te,

Kao

lin

ite,

Sam

ple

Gib

bsi

te,

Kao

lini

te,

feet

num

ber

per

cen

tp

erce

nt

num

ber

per

cen

tp

erce

nt

num

ber

per

cen

tp

erce

nt

158

-308

B20

058

-286

B14

058

-275

B8

0

258

-309

B·

230

58-2

87B

170

58-2

76B

100

358

-310

B27

058

-288

B20

058

-277

B8

44

58-3

11B

360

58-2

89B

204

58-2

78B

270

558

-312

B21

058

-290

B30

058

-279

B8

36

58-3

13B

230

58-2

91B

190

58-2

80B

153

758

-314

B32

058

-292

B13

958

-281

B37

0~

858

-315

B23

058

-293

B5

658

-282

B19

0~

958

-316

B7

658

-294

B12

1358

-283

B0

510

58-3

17B

330

58-2

95B

114

58-2

84B

22

1158

-318

B16

1058

-296

B11

658

-285

B5

1612

58-1

66B

1725

58-1

48B

1514

58-1

57B

2539

1358

-167

B15

2058

-149

B10

4558

-158

B11

1458

-168

B23

1158

-150

B8

1558

-159

B20

15

1558

-169

B23

1358

-151

B4

1358

-160

B17

33

165

8-l

70

B23

1958

-152

B4

3958

-161

B11

19

1758

-171

B7

2158

-153

B-

-58

-162

B0

33

1858

-172

B19

3658

-154

B-

-58

-163

B7

13

1958

-173

B14

3758

-155

B5

3258

-164

B0

2058

-174

B0

4158

-156

B50

3'2

58-1

65B

033

._- - - - - ---

APPENDIX 2c

Table 2.(Continued)

Hole No.7 Hol e No.8

Depth, Sample Gibbsi te, Kaolinite, Sample Gibbsite, Kaolinite,feet number percent per cen t number p ercent p ercen t

1 58-297B 20 0 58-218B 34 02 58-298B 18 3 58-219B 33 03 58-299B 17 6 58-220B 33 04 58-300B 15 10 58-221B 15 45 58-301B 14 11 58-222B 36 56 58-302B 30 3 58-223B 17 137 58-303B 12 24 58-224B 17 108 58-30 4B 12 25 580225B 13 209 58-305B 14 22

10 58-306B 8 42 580226B 3 6711 58-307B 4 4412 58-139B 0 42 58-211B 1 8213 58-140B 0 6 4 58-212B 4 2514 58-141B 0 64 58-213 B 1 5715 58-142B 9 28 58-214B 3 7016 58-143B 0 20 58-215B 2 4417 58-144B 0 72 580216B 0 4718 58-l45B 0 63 58-217B 0 6019 58-146B 0 7520 58-147B 0 60

15

I.

UNIVERSITY OF HAWAIICOLLEGE OF TROPICAL AGRICULlURE

HAWAII AGRICULTURAL EXPERIMENT STATIONHONOLULU, HAWAII

LAURENCE H. SNYDERPresident of the University

MORTON M. ROSENBERGDean of the College and

Director of the Experiment Station