T,ONfl
'E C LOGICAL SUI
Bulletin No, 69
iaBsaaa
jsf^&tM^&Si,,.
JEMINA JiiXTRACTION
ILLINOIS STATE GEOLOGICAL SURVEY
3 3051 00000 1630
STATE OF ILLINOIS
DWIGHT H. GREEN, Governor
DEPARTMENT OF REGISTRATION AND EDUCATION
FRANK G. THOMPSON', Director
DIVISION OF THE
STATE GEOLOGICAL SURVEY
M. M. LEIGHTON, Chief
URBANA
BULLETIN NO. 69
AMENABILITY OF VARIOUS TYPES OF
CLAY MINERALS TO
ALUMINA EXTRACTION
BY THE LIME SINTER AND LIME-SODA 0^
SINTER PROCESSES G^°V°G *
R. E. Grim, J. S. Machin, and W. F. Bradley
Released for publication by the Office of Production Research and Development of the War Production
Board, Research Project NRC-523 under contract WPB-38 with hie University of Illinois
PRINTED BY AUTHORITY OF THE STATE OF ILLINOIS
URBANA, ILLINOIS
1945
ORGANIZATION
STATE OF ILLINOIS
HON. DWIGHT H. GREEN, Governor
DEPARTMENT OF REGISTRATION AND EDUCATION
HON. FRANK G. THOMPSON, Director
BOARD OF NATURAL RESOURCES AND CONSERVATION
HON. FRANK G. THOMPSON, Chairman
NORMAN L. BOWEN, Ph.D., D.Sc, LL.D., Geology
ROGER ADAMS, Ph.D., D.Sc, Chemistry
LOUIS R. HOWSON, C.E., Engineering
*WILLIAM TRELEASE, D.Sc, LL.D., Biology
EZRA JACOB KRAUS, Ph.D., D.Sc, Forestry
ARTHUR CUTTS WILLARD, D.Engr., LL.D.
President of the University of Illinois
GEOLOGICAL SURVEY DIVISION
M. M. LEIGHTON, Chief
^Deceased.
(82471— 4M— 5-45)
SCIENTIFIC AND TECHNICAL STAFF OF THE
STATE GEOLOGICAL SURVEY I) I V S I O N
100 Natural Resources Building, Urbana
M. M. LEIGHTON, Ph.D., Chief
Enid Townley, M.S., Assistant to the Chief
Velda A. Millard, Junior Asst. to the Chief
Helen E. McMorris, Secretary to the Chief
Effie Hetishee, B.S., Geological Assistant
GEOLOGICAL RESOURCES
Coal
G. H. Cady, Ph.D., Senior Geologist and Head
L. C. McCabe, Ph.D., Geologist (on leave)
R. J. Helfinstine, M.S., Mech. Engineer
Charles C. Boley, M.S., Assoc. Mining Eng.
Heinz A. Lowenstam, Ph.D., Assoc. Geologist
Bryan Parks, M.S., Asst. Geologist
Earle F. Taylor, M.S., Asst. Geologist
(on leave)
Rai ph F. Strf.te, A. M., Asst . Geologist
M. W. Pullen, Jr., M.S., Asst. Geologist
Robert M. Kosanke, M.A., Asst. Geologist
Robert W. Ellingwood, B.S., Asst. Geologist
George M. Wilson, M.S., Asst. Geologist
Arnold Eddings, B.A., Research Assistant
(on leave)
Henry L. Smith, A.B., Asst. Geologist
Raymond Siever, B.S., Research Assistant (on leave)
John A. Harrison, B.S., Research Assistant
(on leave)
Mary E. Barnes, B.S., Research Assistant
Margaret Parker, B.S., Research Assistant
Elizabeth Lohmann, B.F.A., Technical Assistant
Industrial Minerals
J. E. Lamar, B.S., Geologist and Head
H. B. Willman, Ph.D., Geologist
Robert M. Grogan, Ph.D., Assoc. Geologist
Robert T. Anderson, M.A., Asst. Physicist
Robert R. Reynolds, M.S., Asst. Geologist
Margaret C. Godwin, A.B., Asst. Geologist
Oil and Gas
A. H. Bell, Ph.D., Geologist and Head
Carl A. Bays, Ph.D., Geologist and Engineer
Frederick Squires, B.S., Petroleum Engineer
Stewart Folk, M.S., Assoc. Geologist (on leave)
Ernest P. Dubois, Ph.D., Assoc. Geologist
David H. Swann, Ph.D., Assoc. Geologist
Virginia Kline, Ph.D., Assoc. Geologist
Paul G. Luckhardt, M.S., Asst. Geologist
(on leave)
Wayne F. Meents, Asst. Geologist
James S. Yolton, M.S., Asst. Geologist
Robert N. M. LIrash, B.S., Research Assistant
Margaret Sands, B.S., Research Assistant
Areal and Engineering Geology
George E. Ekblaw, Ph.D., Geologist and Head
Richard F. Fisher, M.S., Asst. Geologist
Subsurface Geology
L. E. Workman, M.S., Geologist and Head
Carl A. Bays, Ph.D., Geologist and Engineer
Robert R. Storm, A.B., Assoc. Geologist
Arnold C. Mason, B.S., Assoc. Geologist
(on leave)
C. Leland Horberg, PhD., Assoc. Geologist
Frank E. Tippie, B.S., Asst. Geologist
Merlyn B. Buhle, M.S., Asst. Geologist
Paul Herbert, Jr., B.S., Asst. Geologist
Charles G. Johnson, A.B., Asst. Geologist
(on leave)
Margaret Castle, Asst. Geologic Draftsman
Marvin P. Meyer, B.S., Asst. Geologist
Robert N. M. Urash, B.S., Research Assistant
Elizabeth Pretzer, A.B., Research Assistant
Ruth E. Roth, B.S., Research Assistant
Stratigraphy and Paleontology
J. Marvin Weller, Ph.D., Geologist and Head
Chalmer L. Cooper, M.S., Assoc. Geologist
Petrography
Ralph E. Grim, Ph.D., Petrographer
Richards A. Rowland, Ph.D., Asst. Petrographer
(on leave)
William A. White, B.S., Research Assistant
Physics
R. J. Piersol, Ph.D., Physicist
B. J. Greenwood, B.S., Mech. Engineer
GEOCHEMISTRY
Frank H. Reed, Ph.D., Chief Chemist
Elizabeth Ross Mills, M.S., Research Assistant
Coal
G. R. Yohe, Ph.D., Chemist
Herman S. Levine, B.S., Research Assistant
Industrial Minerals
J. S. Machin, Ph.D., Chemist and Head
Delbert L. Hanna, A.M., Asst. Chemist
Fluorspar
G. C. Finger, Ph.D., Chemist
Oren F. Williams, B. Engr., Research Assistant
X-ray and Spectrography
W. F. Bradley, Ph.D., Chemist
Chemical Engineering
Harold W. Jackman, M.S.E., Chemical Engineer
P. W. Henline, M.S., Assoc. Chemical Engineer
James C. McCullough, Research Associate
James H. Hanes, B.S., Research Assistant (on leave)
Leroy S. Miller, B.S., Research Assistant (on leave)
Analytical
O. W. Rees, Ph.D., Chemist and Head
L. D. McVicker, B.S., Chemist
Howard S. Clark, A.B., Assoc. Chemist
William F. Wagner, M.S., Asst. Chemist
Cameron D. Lewis, B.A., Asst. Chemist
Herbert N. Hazelkorn, B.S., Research Assistant
William T. Abel, B.A., Research Assistant
Melvin A. Rluenstorf, B.S., Research Assistant
Marian C. Stoffel, B.S., Research Assistant
Jean Lois Rosselot, A.B., Research Assistant
MINERAL ECONOMICS
W. H. Voskuil, Ph.D., Mineral Economist
Douglas F. Stevens, M.E., Research Associate
Nina Hamrick, A.B., Research Assistant
Ethel M. King, Research Assistant
PUBLICATIONS AND RECORDS
George E. Ekblaw, Ph.D., Geologic Editor
Chalmer L. Cooper, M.S., Geologic Editor
Dorothy E. Rose, B.S., Technical Editor
Meredith M. Calkins, Geologic Draftsman
['.I i i \n Featherstone, B.F.A., Asst. Geologic
Draftsman
Willis L. Busch, Principal Technical Assistant
Portia Allyn Smith, Technical Files Clerk
Leslie D. Vaughan, Asst. Photographer
Consultants: Ceramics, Cullen W. Parmalee, M.S., D.Sc, and Ralph K. Hursh, B.S., University of Illinois
Mechanical Engineering, Seiciii Konzo, M.S., University oj Jllinoti
Topographic Mapping in Cooperation with the United States Geological Survey.
May, 1, 1945
Digitized by the Internet Archive
in 2012 with funding from
University of Illinois Urbana-Champaign
http://archive.org/details/amenabilityofvar69grim
CONTKNTS
Page
Introduction 9
Clay mineral concept 10
Objectives of the investigation 10
Acknowledgments 11
Materials studied 11
Clays 11
Selection of samples 11
Collection and preparation of samples 11
Analysis of samples 13
Limestone 13
Anorthosite 13
Lime-sinter process 14
Experimental methods 14
Preparation of sinter samples 14
Chemical analysis of sinters 15
Extraction of alumina from sintered material and analysis of extracts 15
Experimental results 15
Effect of type of clay minerals 15
Optimum yield of alumina 15
Lime-alumina ratio 16
Sintering temperature 19
Effect of time held at sintering temperature 27
Effect of furnace atmosphere 28
Dusting of sinters 29
Silica in the alumina extract 30
Effect of minor components present in the raw materials 30
General comments 30
Effect of magnesium 30
Effect of sulfur 32
Effect of phosphorus 32
Effect of iron 33
Effect of different types of limestone 34
Mineralogical analyses of sinters 35
Microscopic analysis 35
General comments 35
Effect of variations in lime content in sinters fired one hour at 1360° C 35
Effect of variations in sintering temperature on batches with a lime-to-alumina ratio
of 1.66 and held at sintering temperature for one hour 37
Effect of variations in sintering time of batches with a lime-to-alumina ratio of 1.66
fired at 1360° C 37
Effect of variations in kiln atmosphere 38
Effect of miscellaneous variations in composition 38
Effect of variation in the character of the limestone 38
X-ray analysis 39
General comments 39
Predominant phases 39
Low temperature syntheses 40
Sintering process 40
Differential thermal analyses 41
Summary and conclusions 44
CONTENTS — Continued
Page
Lime-soda sinter process 47
Experimental methods 47
Preparation of sinter samples 47
Chemical analysis of sinters 48
Extraction of alumina from sintered material and analysis of extracts 48
Experimental results 48
Effect of type of clay minerals 48
Optimum yield of alumina 48
Sintering temperature 50
Effect of time held at top sintering temperature 54
Effect of heating rate 57
Effect of furnace atmosphere 57
Effect of soda-to-alumina ratio 57
Effect of lime-to-alumina ratio 60
Silica in the alumina extract 60
Effect of minor components present in the raw materials 60
General comments 60
Effect of phosphorus 64
Effect of titania 64
Effect of iron 64
Effect of magnesia 64
Mineralogical analyses 64
Microscopic analysis 64
Effect of variations in top sintering temperature 64
Effect of variations in the ratio of soda to alumina 65
Effect of miscellaneous variations in composition 68
Effect of reducing atmosphere 68
Effect of variation in time batch held at top sintering temperature 69
Effect of variation in heating rate 69
Effect of variations in the ratio of lime to silica 69
X-ray analysis 70
General comments 70
The phases 70
The sintering process 70
Differential thermal analyses 71
Discussion and summary of study of phases present in the sinters 73
Summary and conclusions 74
References 77
ILLUSTRATIONS
Figure Page
1 A-E — Percent A1203 extracted versus mole ratio CaO/Al203 16—18
2 A-E — Percent A1203 extracted versus temperature at which sinter was held for
one hour 19—21
MgO
3 — Percent A1203 extracted versus 100 ., ., in sinter mix 31
4 — Percent A1203 extracted versus percent S03 in sinter mix 32
5 — Percent A1203 extracted versus percent P205 in sinter mix 33
6 — Percent A1203 extracted versus percent Fe203 in sinter mix 34
7 — Photomicrographs of lime-clay sinters 36
8 — Differential thermal analyses of lime-clay sinters 42
9 — Scale for determining the temperature differences recorded by peaks of the
differential thermal curves 43
10 A-B — Comparison of extraction values for lime-soda sinters prepared from clays
containing more than one clay mineral type with hypothetical mixtures 49
11 A-E — Percent A1203 extracted versus temperature at which sinter was held for
10 minutes 50-52
12 A-D — Percent A1203 extracted versus time sinter was held at 1100° C 55—56
13 A-D — Percent A1203 extracted versus mole ratio Na20/Al203 58—59
14 A-B — Percent A1203 extracted versus mole ratio CaO/Si02 61
15 — Typical record of a sinter heating schedule as traced from recorder chart 62
16 — Percent A1203 extracted versus percent P205 in mix 63
17 — Percent A1203 extracted versus Ti02 in mix 63
18 — Percent Al2Os extracted versus sinter temperature 63
19 — Percent A1203 extracted versus percent Fe203 in mix 63
% MgO
20 — Percent A1203 extracted versus 100 ~, . , ^ in sinter 64
21 — Photomicrographs of soda-lime-clay sinters 66
22 — Photomicrograps of lime-soda-clay sinters 67
23 — Differential thermal analyses of lime-soda-clay mixtures 72
21 — Equilibrium diagram of the system CaO— A1203— Si02 76
TABLES
1 — Common clay minerals 10
2 — Location and clay mineral composition of clay samples 11
3 — Chemical composition of clay samples 12
A — Location and texture of limestone samples 13
5 — Chemical composition of limestone samples 13
6 — Chemical composition of anorthosite sample 13
7 — Sinter composition and extraction data 22—26
8 — The effect of time held at sintering temperature 27
9 — The effect of furnace atmosphere on extractability 28
10— The effect of phosphorus on the extractability of lime-clay sinters 33
11 — Effect of different types of limestone on extractability of clay-lime sinters 34
12 — Effect of sintering temperature on percent A12C):; extracted with various types of clay 53
AMENABILITY OF VARIOUS TYPES OF
CLAY MINERALS TO
ALUMINA EXTRACTION
BY THE LIME SINTER AND LIME-SODA
SINTER PROCESSES
By
R. E. Grim, Petrographcr
J. S. Machin, Physical Chemist
W. F. Bradley, X-ray Analyst
INTRODUCTION
THE usual ore from which aluminum is
recovered is bauxite, an essentially
silica-free alumina hydrate. In order to meet
the normal peacetime needs of the United
States, it has been necessary to import
bauxite. With the outbreak of the war,
government agencies at once recognized the
seriousness of the situation that would
result if bauxite imports were halted and
began studies of the possible use of other
materials as sources of alumina. Particular
attention was directed to clays because they
are abundant, they contain considerable
quantities of alumina, and a considerable
amount of information concerning clays
was already in hand.
There are two general processes for ex-
tracting alumina from clays: An acid proc-
ess that uses an acid to selectively dissolve
the alumina, usually after the clay has been
roasted, and an alkaline process that uses
water or a dilute alkaline solution to selec-
tively dissolve the alumina from a sintered
mixture of lime and clay or of lime, soda,
and clay. Most past experimental work
with these processes had been limited to
high-alumina clays of the kaolin or bauxitic-
kaolin type. A study of the occurrence of
clays indicated that deposits of kaolinite
clays of sufficient purity and size for an
alumina plant are not now known. Kaolin-
ite clays are abundant, but in the main they
contain varying, although frequently small,
amounts of other clay minerals. Further,
other types of clay, notably the illite shales,
are widespread in enormous relatively uni-
form deposits, and some of them have
alumina contents not very much smaller
than the average in the large deposits of
kaolin.
It became obvious, therefore, that re-
searches were needed on the clay minerals
other than kaolinite as possible sources of
alumina. Accordingly in January 1943, the
War Metallurgy Committee of the Nation-
al Academy of Sciences suggested that the
Illinois State Geological Survey conduct
an investigation of the "amenability of clay
mineral types to lime sinter and lime-soda
sinter alumina processes" for the War
Production Board.
The Illinois State Geological Survey had
been and still is conducting extensive re-
searches on the clay mineral composition
and properties of the clays and shales of
Illinois. Since the properties of clay materi-
als in general are controlled largely by
their clay mineral composition, the Survey
has studied extensively the relation between
the clay mineral composition of all types of
clays and shales and their properties and
uses. The investigation undertaken for the
National Academy of Sciences was, there-
fore, essentially an extension of these re-
searches.
9 I
10
ALUMINA EXTRACTION
Clay Mineral Concept
Recent researches in several laboratories
have shown that natural clay substances
are composed of extremely small crystal-
line particles of members of any one or
more of a few groups of minerals known
as the "clay minerals." The "clay minerals"
(table 1) are hydrous aluminum silicates,
frequently with some replacement of alumi-
num by iron and magnesium and with small
amounts of alkalies and alkali earths. In
rare instances magnesium and iron com-
pletely replace the aluminum. In addition to
the "clay minerals," variable but usually
small amounts of quartz, limonitic material,
gibbsite, diaspore, organic material, pyrite,
feldspar, etc., may also be present in clay
materials. Grim1* has recently published a
summary of the composition, properties,
structure and occurrence of the various clay
minerals.
Table 1. — Common Clay Minerals
Name
Composition formula
Kaolinite group*. .
(OH)8Al4Si4O10
Illite groupb
(OH),Ky(Al4-Fe4-Mg4-Mgs)
(ol8-y - Aly)U20
Montmorillonite
group0
(OH)4Al4Si8O20-nH2O
Hydrated halloy-
sited
(OH)8Al4Si4Oio-2H20
Halloysite
(OH)8Al4Si4O,0
Attapulgitee
(OH2yOH)2Mg5Si8O20-4H2O
a Dickite and nacrite with the same composition also
belong to this group but their occurrence is extremely
rare. Anauxite with a slightly higher silica content is
usually classified in the kaolinite group — its occurrence
is also rare.
b As indicated by the formula the composition of the
illite group is variable. The y in the formula is frequently
equal to about 1. The illite group has not yet been divided
into distinct mineral species.
c The A1+++ of the montmorillonite is replaceable by
Fe+++ or Mg++. When the replacement is relatively complete
the resulting mineral species are nontronite and saponite,
respectively.
d Hydrated halloysite inverts to halloysite irreversibly
at about 60°C.
e In attapulgite A1+++ can replace Mg++ and Si+++ to a
limited extent.
The type of clay known as kaolin is com-
posed of minute ( 1 micron ± ) crystalline
particles of the mineral kaolinite. Some such
clays contain also small amounts of mont-
morillonite and illite. The ceramic ball
clays and fireclays are usually composed of
kaolinite with small amounts of illite. The
high-alumina clays are mixtures of kaolin-
*See references at end of report.
ite and diaspore or gibbsite. Large deposits
of halloysite clays are not known, but the
mineral may be present in small amounts
in many clays that are made up primarily of
other clay minerals.
Illite is the dominant mineral in all the
shales that have been studied mineralogical-
ly. Bentonites are composed primarily of
extremely minute particles of montmorillo-
nite, and attapulgite is the component of
certain fuller's earths. The clay mineral
composition of surficial soil materials is
variable depending on the conditions under
which the soil has formed and on the parent
material.
Objectives of the Investigation
Objectives of the investigation were as
follows :
1. To determine if any clays other than
those composed of kaolinite are promising
raw materials for alumina extraction by
the alkali methods, and if so how the con-
ditions governing sintering and extraction
of other clays differ from those governing
the sintering and extraction of kaolinite
types.
2. To provide some of the necessary
data for the solution of problems that will
arise in the large scale use of clays as a
source of alumina because of the inevitable
variations in clays. It is unlikely that a de-
posit of any clay, kaolin or some other type,
large enough for several years supply of
alumina, is to be found anywhere that is
composed throughout of a single clay min-
eral. Variations in kaolin deposits, for ex-
ample, may be expected by the sporadic
occurrence of small amounts of illite or
montmorillonite.
3. To throw light on the mechanism of
the extraction of alumina from clays by the
alkali methods. It seemed probable that the
study of a variety of materials might provide
clues that would not appear in investiga-
tions limited to one type of clay.
After the investigation was started it
was enlarged to include the study of the
effect on alumina extraction of certain non-
clay-mineral impurities apt to be found in
clays and of variations in the character of
limestone.
MATERIALS STUDIED
11
Acknowledge ents
'{'he authors wish to express their appre-
ciation to the War Metallurgy Committee
of the National Academy of Sciences and
the War Production Board for sponsoring
the project and their continued interest in
the work. They also wish to make acknowl-
edgment to the Project Committee which
supervised the project, particularly to its
chairman, Mr. John D. Sullivan, for many
helpful suggestions and comments.
MATERIALS STUDIED
CLAYS
Selection of Samples
On the basis of a large amount of analyt-
ical data in the files of the Illinois State
Geological Survey, 19 samples of clay
(table 2) were selected to represent the
important types of clay minerals and
mixtures of these minerals. Particular at-
tention was given to clays containing some
kaolinite because such clays tend to have the
highest content of alumina. The samples
selected to represent mixtures illustrate the
combinations of clay minerals that are apt
to be encountered in large deposits of the
common types of clays. Further consider-
ation was given to the selection of samples
that would represent important types of
clay that occur widely in large deposits.
Collection and Preparation
of Samples
All of the samples except 865, 868, 881,
882, and 883 were collected personally by
Drs. Grim and Bradley. About 100 lbs.
of each sample was obtained in lump form.
Table 2. — Location and Clay Mineral Composition of Clay Samples
Sample
No.
Type of clay
Location
Mineral composition
865
866
Kaolin Ringgold, Tenn.
Underclay Grundy County,
Kaolinite
111 Illite, kaolinite (20%±), quartz (15%=b), trace
of pyrite, gypsum, and calcite
Halloysite clav Eureka, Utah Hallovsite and hydrated halloysite, gibbsite
(10%±)
Bentonite Clay Spur, Wyo Montmorillonite, quartz (10%±)
Kaolin Union County, 111 Kaolinite with unusual lattice structure proper-
ties, quartz (10%±)
Shale Menard County, 111 Illite (high iron variety), quartz (15%±), trace
of pyrite, limonite, kaolinite
Plastic fireclay Mexico, Missouri Kaolinite with small amount of illite, montmoril-
lonite and organic material
Diaspore Swiss, Missouri Diaspore, trace of kaolinite, anatase, rutile
Flint clay New Florence, Missouri. Kaolinite, trace of anatase and montmorillonite
(?)
Bauxitic kaolin Dry Branch, Ga Kaolinite with small amount of gibbsite, trace of
anatase
Bauxite Irvington, Ga Gibbsite with small amount of kaoinlite, trace of
anatase
Hard kaolin Gordon, Ga Kaolinite, very fine grain size, trace of montmoril-
lonite
Soft kaolin Dry Branch, Ga Kaolinite, medium grain size
Plastic kaolin Dry Branch, Ga Kaolinite, montmorillonite (25%±)
Ball clay Whi dock, Tenn Kaolinite, illite, quartz (10%±), organic ma-
terial
Ball clay Atwood, Tenn Kaolinite, illite (?), gibbsite (?)
Kaolin Aiken, S. Carolina Kaolinite, trace of alunite
Fuller's earth Quincy, Florida Attapulgite, montmorillonite (10%±)
Kaolin Hobart Butte, Oregon. . . Kaolinite with low degree of crystallinity
881
882
883
12
ALUMINA EXTRACTION
A
o 2
i,
©O©©© -h o o o o ooooo ooo ©
© o\ o o © ooooo ooooo ooo ©
nt^1^^
\0 0\COnN
<N CO © CN
CN CO CN "^ -— '
co cn r- o
CN i-i
CN CN OO CO "^
'-H O CN OO CO
(N CO >-0 CN
O ©CN i-h
>0 CJ\ vn CO CN
ooomo
oo © ©© i-
CO CN rt< CN CO
Tf< CN CN CN r-H CN -— i i— i CN.— i
s
LTi^r^oooN © -— i cn co ""f
vovcovo^ r^- r^ r- r-~ r--
oooooooooo oooooooooo
LO'xor^-oo^ ©^-<cn
r-t^-r^-r^r-- oocooo
oooooooooo cooooo
MATERIALS STI'DIKP
13
Table 4. — Location and Texture of Limestone Samples
Sample No.
Aye and Location
Size of calcite particles
R7 St. Louis limestone, Alton, Maximum diameter 5 microns; most particles less than 3 mi-
lllinois crons
R2 St. Louis limestone, Alton, Maximum diameter 15 microns; particles range from 3 to 15
Illinois microns
MC Burlington limestone Maximum diameter 70 microns; most particles larger than 20
Marblehead, Illinois microns
A small quantity of the lump material of
each sample was retained for reference, and
and the remainder was ground to minus
10-mesh in a disk pulverizer in order to
obtain thorough mixing. Ten-pound lots
of each sample were ground further to
minus 80-mesh in preparation for chemical
analysis and sinter experiments.
Analysis of Samples
The clay mineral composition of each
sample was determined by X-ray diffrac-
tion, differential thermal, and optical analy-
ses with the results given in table 2. Chemi-
cal analyses of all samples are given in table
3.
LIMESTONE
Precipitated CaCOs was used in prepar-
ing all of the sinters except those especially
made to investigate the effect of variations
in the textural character of the limestone
on alumina extraction. For this purpose
three limestones were selected. Table 4
gives the location of the limestones and the
size of the particles of calcite that make up
each of them. Chemical analyses of the lime-
stones are given in table 5.
ANORTHOSITE
Anorthosite was included in the part of
the investigation concerned with lime-soda
sinters. A sample was supplied by Mr. D.
R. Williams of the Monolith Portland
Midwest Company from a deposit near
Laramie, Wyoming. A chemical analysis of
the sample is given in table 6. Mineralogi-
cally the anorthosite is composed of a
plagioclase feldspar of about labradorite
composition.
Table 5. — Chemical Composition of Limestone
Samples
(In percent)
R7
Si02 2.59
Ti02 02
AL03 55
Fe203 14
MgO 33
CaO 53.77
Na20 06
K2O 07
P205 00
SO3 14
Loss on Ign 42.53
Total 100.20
H2O-(<110°C). .24
C02 42.30
Table 6. — Chemical Composition of Anortho-
site Sample
(In percent)
Si02 52.50 MgO... 0.21 P2Os... 0.08
TiOs 23 CaO... 11. 71 Loss on
Ign 32
Fe203 2.28 Na20....3.94
AI2O3. . . .28.48 K2Q 43 Total. .100. 16
R2
MC
0.46
0.35
.02
.00
.27
.14
.14
.07
.23
.27
55.36
55.65
.05
.08
.03
.00
.00
.00
.06
.04
43.67
43.72
100.29
100.32
.08
.11
43.51
43 . 69
14
ALUMINA EXTRACTION
LIME-SINTER PROCESS
This process in brief consists in mixing
ground siliceous alumina bearing ore with
ground calcium carbonate in the proportions
to give a mixture corresponding approxi-
mately in percentage composition (ignoring
volatile constituents present) to a mixture
of dicalcium silicate and pentacalcium tria-
luminate. This mixture is then heated at
a suitable temperature (of the order
1375°C.) and for such period of time as
may be necessary to develop dicalcium sili-
cate and some calcium aluminate compound
or compounds which can be acted upon by
dilute alkali carbonate solution in such
manner as to dissolve alumina and leave
undissolved all or nearly all of the silica,
lime, and any other materials present in
the sintered material. The dicalcium silicate
plays an extremely important dual role.
First, its development is complete enough
to tie up nearly all of the silica present in
a form not soluble in the leach liquor.
Second, it undergoes a crystallographic
transformation on cooling below about
675 °C. with an increase in volume which
results in reduction of the sintered mass
to a powder, making the soluble alumina
compounds easily accessible to the action of
the leach liquor without grinding. This
phenomenon is commonly called "dusting."
The clay samples used in these experi-
ments after mixing with CaCOs and heating
gave sintered materials in which the com-
position ranged about as follows in most
cases :
AhO 10 to 30 percent
CaO 52 to 62 percent
SiO 14 to 28 percent
The fusion points of mixtures represent-
ed by such ranges in composition are well
above 1375°C. so that in general, it is to
be expected that a state of chemical equilib-
rium will be approached but not reached
unless the mixtures are kept at the sintering
temperature for very long periods of time.
The longer the period of time and the high-
er the temperature, the closer will be the
approach to a state of equilibrium. The
presence of varying, even though minor,
amounts of alkalies, iron, magnesia, phos-
phorus, and titania also might be expected,
through fluxing action or otherwise, to in-
fluence the final state reached. Finally the
various clay minerals, with their differences
in reactivity and refractoriness might be ex-
pected to exert a major effect, at least up
to some critical temperature which would
probably be different for each clay mineral
type. This critical temperature might be
the fusion point of the sinter mixture or
might be a lower temperature. The data
presented in this report were collected from
experiments designed to study the influence
of these factors on the lime-sinter process.
The equilibrium diagram of the ternary
system CaO- AI2O3— SiC>2 of Rankin and
Wright as revised by Schairer9 (page 76) is
included for the convenience of the reader
in visualizing the compositions of the sin-
ters discussed in the lime-sinter section of
this report. The shaded section includes the
range of compositions included in our ex-
periments.
EXPERIMENTAL METHODS
Preparation of Sinter Samples
All of the clay samples except 865 and
872 were calcined at 800°C. Sample 872
was not calcined because calcination of
diaspore clay results in crystal changes that
take place below 800° C. which it was
desired to avoid. Sample 865 was the clay
used most extensively in certain experiments
at the T. V. A. laboratory,3 and it was de-
sired to follow their experimental method.
In this instance the calcination temperature
was 900° C.
Calcined clays were ground to pass 80-
mesh sieves and then mixed thoroughly
with the desired quantity of precipitated
calcium carbonate and enough water to
make a good briquet. The mixtures were
pressed into cylindrical briquets about \j4,
inches in diameter under a pressure of
LIME-SINTER PROCESS
15
approximately 5000 pounds per square inch.
These briquets were dried, placed in a cold
Globar-heated furnace, heated up to the
desired temperature, and held for a chosen
period at that temperature. The cooling
rate was such that approximately 30 min-
utes elapsed while the furnace cooled down
to 1100°C After this the power was turned
off. The sintered briquets were removed
from the furnace when the temperature had
dropped to about 700°— 800°C and were
stored in closed bottles as soon as they were
cool enough to handle. Usually the briquets
dusted in the cooling process.
Most sinters prepared as described dusted
to a powder which easily passed a 200-mesh
screen with one percent or less oversize. A
few had considerable plus 200-mesh mate-
rial. Some of those which did not so disinte-
grate were ground to pass 200 mesh and
subsequently extracted in the same manner
as the others.
Chemical Analysis of Sinters
The sintered material was analyzed by
gravimetric methods for SiOa, Al^O^, CaO,
FesOs and TiCX About three quarters of
all sinters were so analyzed.
Extraction of Alumina from Sintered
Material and Analysis of Extracts
The extraction procedure used was basi-
cally similar to that used by Walthall 2 and
coworkers at the T. V. A. laboratory, Wil-
son Dam, Alabama. Ten-gram samples of
each sinter were extracted with three per-
cent sodium carbonate solution using me-
chanical stirring sufficiently vigorous to pre-
vent settling. The extraction time was 15
minutes and the temperature was 65 °C.
Sufficient sodium carbonate solution was
used to provide 1.66 moles of Na;CO< for
each mole of Al^Oa calculated to be present
in the 10-gram sample. The extract was
immediately filtered off and the residue
washed. The extract was analyzed for
alumina and silica using gravimetric meth-
ods for alumina and colorimetric methods
for silica.
EXPERIMENTAL RESULTS
Effect of Type of Clay Minerals
optimum yield of alumina
The factors which may be varied at the
will of the operator of a plant producing
alumina by the lime-sinter process are the
lime-to-clay ratio, the temperature, the du-
ration of the heating period, and to a much
lesser extent the furnace atmosphere. Proper
adjustment of these variables with relation
to one another might be expected to result in
an optimum yield for a given raw material,
assuming good extraction practice. Data
are presented in graphic and tabular form
in figures 1 and 2, and in tables 7, 8, and 9
which bear on the effect of these operating
variables when the various types of clays are
used in the sinter mix under laboratory con-
ditions. Variables connected with the ex-
traction process were not considered in these
experiments except that care was taken to
keep the extraction conditions constant.
The data indicate that gibbsite-kaolinite
clays and kaolinite clays may yield above
(some well above) 90 percent of their
alumina with little to choose between them.
Note data on clay samples, 865, 869, 873,
874, 875, 876, 877, 880, and 881.
Clay sample 883 did not quite measure
up to the other kaolinite clays in yield of
alumina. Two possible causes of the lower
yields from this clay can be suggested; the
first and more probable is that it differed
considerably from the typical kaolins insofar
as its content of true kaolinite is concerned ;
second, the titania content was rather high.
Clay sample 865 also had rather high titania
content and gave slightly lower yields than
some of the best kaolinite clays. It is not
clear, however, why titania should cause
lower yields. The presence of calcium titan-
ate in the sinters (see page 38) suggests
that better yields might be expected with the
higher lime-to-clay ratios, but this sugges-
tion is not confirmed experimentally. Impure
kaolins, containing illite and/or montmo-
rillonite, gave lower yields roughly in pro-
portion to the amounts of these clay miner-
als present.
The illite and montmorillonite clays
gave lower percentage yields than the
16
ALUMINA EXTRACTION
kaolinite clays although it is not proved that
this was due to the clay mineral type. Sin-
ters made with such clays tended to dust
slowly and poorly, possibly because of their
high alkali content. That is, the dusted sin-
ter contained larger amounts of coarse
sandy material. The iron, as indicated in
another part of the report, appeared in the
sinter as tetracalcium alumino ferrite from
which alumina is not readily extractable
with dilute sodium carbonate solutions. The
illites reacted adversely to low lime ratios
(fig. 1, samples 866 and 870), probably be-
cause of their high iron content.
Halloysite clay gave low yields except at
very high temperatures. This clay contained
remarkably small amounts of oxides other
than alumina and silica. The sinters pre-
pared from it dusted fairly quickly but in-
completely and contained much coarse sandy
material, indicating either incomplete in-
version of dicalcium silicate, incomplete re-
action between the clay and the lime, or
the presence of stable glass.
Diaspore clay gave very unsatisfactory
sinters. Attempts to sinter it with lime at
temperatures above 1300°C. resulted in par-
tially fused glassy masses which did not
dust. Upon grinding through 200 mesh
these sinters yielded about 70 percent of
their alumina. Both the fusion and the
failure to dust were expected. The sinters
prepared from this clay contained 6 to
7 percent silica, about 40 percent alumina,
48 to 50 percent lime, and the balance iron,
titania, and alkalies. The fusion point of
such a composition should be below 1335°
C, and it is not surprising that the relative-
ly small amount of dicalcium silicate that
could be formed should be insufficient to
cause dusting.
Attempts to prepare sinters from the
attapulgite clay resulted in cinder-like mass-
es which disintegrated slowly into coarse
sand-like particles. These sinters were so
unpromising that no further work was done
on this clay.
LIME-ALUMINA RATIO
Data presented graphically in figure 1
show the effect of varying the CaO/ALO
ratio on the percentage of alumina extracted
from different types of clay. The CaO/Ab
Os ratios consider only that part of the total
CaO above that required to react with the
silica to produce dicalcium silicate. These
same data are listed in table 7 together with
the analyses of the sintered materials and of
the clays from which they were prepared.
The data show that in general pure
kaolinite clays are not very sensitive to
moderate variation in the CaO/ALOs ratio.
Ratios between 1.5 and 1.8 appear to be
favorable for most kaolins, whereas ratios
outside these limits usually result in lower
yields of alumina.
Illite clays, probably because of high iron,
sometimes show considerable improvement
in yield with increased lime (note sample
870 in figure 1). Illite sample 866, which
gave a low yield with a CaO/AbOs ratio
of 1.4, showed an increase in yield when the
ratio was increased to 1.5, but no further
increase with larger amounts of CaO.
Neither the halloysite clay, nor the mont<
morillonite clay showed any decided sen-
sitivity to the CaO/AbOs ratio. The data
were somewhat erratic. Such erratic ex-
traction values are common if the dusting
is incomplete, as it was with most sinters
prepared from these two clays.
Gibbsite-kaolinite (bauxitic) clays be-
haved very much like the pure kaolinites.
Like these clays, they reacted unfavorably
to much excess lime.
1- A
Q
U
<
N
>
X
KAOLIN
Ld
"> en
hobart .butte
Ckaolinit
OREGON
O 60
(\1
E)
< sn
QQT
o^°
^ OOo ^
1.4 1.5 1.6 1.7 1.8 1.9 2.0
CaO/AI203
Fig.
1.— p
art A
Percent A1203 extracted versus mole ratio CaO/
A1203 (considering only that part of the
CaO beyond that required to react with the
silica to form 2CaO • Si02). Sinter temper-
ature 1360°
LIME-SINTER PROCESS
17
40
l-B
BO
— •-
HALLOYSlTE
(HALLOYSITE )
/
r
867
70
no
i
r
50
UNDERCLAY
C KAOLINITE, ILLITE )
s*
40
866
i
r
^Nr-^"
90
flO
*
f
(
L_
i
>
Nr"
ftO
ringgold
Ckaolinite )
BENTONITE
( MONTMORILLONITE)
«,n
86
5
86
8
40
1.3
90
D
H 50
<
or 40
70
50
1-9 2.0 J. 4 |.5
CaO/AI203
j
i
i-c
/
y
/
"•r
<
\ 4
V
■^" — i
SHALE
ClLL ITE)
FIRECLAY
(KAOLINITE , ILLITE, MONTMORILLONITE)
870
i
6"
f\
N
i
KAOLIN, ILLINOIS
(KAOLINITE)
FLINT CLAY
(KAOLINITE)
Qi
>9
81
r3
1.4 1.5 16 1.7 18 19 2 0 1.4 1.5 1.6 17 1.8 19 2.0
Ca O /Al2 03
Fig. 1.— Pan, B and ('
Percent Al2Oa extracted versus mole ratio CaO/Al^Os (considering only that part of the CaO
bevond that reauired to react with the silica to form 2CaO.SiOv). Sinter temperature 1360°C.
18
ALUMINA EXTRACTION
^
l-D
>-
i
r
4
if
^^"^v
\
1
>
■-%
>
BAUXITE
(GIBBSITE, KAOLINITE)
HARD KAOLIN
(KAOLINITE )
8"
?5
876
<
j
%
A-
/I
>
^-^
\
1
r-
\
>
N>
BAUXITIC KAOLIN
(KAOUNITE.GIBBSITE)
SOFT KAOLIN
(KAOLINITE )
87
'4
877
1.
4 1
5 1
6 1
7 1
8 1
9 2
O
100
90
60
o
to
K so
o
<
(T 40
J!* 90
<
£ 80
70
CaO/AI203
~ •-
-•^
l-E
<
r*
N.
^
>
(
BALL CLAY
KAOLINITE, ILLfTE)
c
BALL CLAY
KAOLINITE, ILLITE
)
8"i
>9
8e
50
- #<
i
r
* "•
1
>
PLASTIC KAOLIN
(KAOLINITE,MONTMORILLONITE)
KAOLIN, AIKEN.S C
(KAOLINITE )
8";
rQ
86
w
9 2.0 \.d
Ca 0/AI2 03
2.0
Fig. 1.— Parts D and E
Percent A1203 extracted ver us mole ratio CaO/Al203 (considering only that part of the CaO
beyond that required to react with the silica to form 2CaO.Si02). Sinter temperature 1360°C.
LIME-SINTER PROCESS
19
SINTERING TEMPERATURE
The data presented in table 7 and figure
2 indicate that for most clays sintering tem-
peratures between 1340° and 1380°C. are
favorable. Higher temperatures are with
few exceptions not necessary unless for the
purpose of shortening sintering time.
Kaolinite clays usually give good yields
when sintered at 1360° to 1380°C. There
was a pronounced tendency to show a de-
creased yield with lowered sintering tem-
perature which sometimes was manifest at
1340°C. (869, 873, 876, 877, 883), and
sometimes not until somewhat lower tem-
peratures were reached (865, 871, 874, 878,
879, 880, 881). Sinters prepared from pure
kaolinite clays showed little or no tendency
to overburn.
Sinter mixes prepared from gibbsite-
kaolinite and diaspore clays, due to their
high alumina and low silica content, were
apt to melt. The composition of sinters pre-
pared from sample 875 — if only alumina,
lime and silica are considered — should not
have melted much below 1600°C. However,
the small amounts of other oxides present
lowered the fusion point of the mixture
considerably. All sinters from sample 875
melted partially when the sintering tem-
peratures were only as high as 1360°C. As
stated above, sinters from the diaspore clay
melted at even lower temperatures.
Sinters prepared from illite and mont-
morillonite clays were somewhat sensitive
to sintering temperature. Thej appeared to
give lowered yields with temperatures on
either side of an optimum which was in the
range 1360° to 1380°C. They showed a
definite tendency to overburn at higher
temperatures.
Sinters prepared from halloysite clay had
to be heated to high temperatures in order
to give even moderately good yields. Even
when heated to 1450° (J. where fusion or
semifusion took place the disintegrated sin-
ter still felt somewhat gritty.
® INDICATES THE ATMOSPHERE
WAS REDUCING
% INDICATES OXIDIZING OR
NEUTRAL ATMOSPHERE
IOO
LU
MOLE RATIO CaO/AI203 = 166
SHALE
2 -A
(ILLITE)
% 60
cr
h
X in
!
fi
5
870
1
1
n i
o
■
vl
<
>° 50
1
>
1
1340 1380 1380 1400
TEMPERATURE DEGREES C
Fig. 2.— Part A
Percent A1203 extracted from sinter versus tem-
perature at which sinter was held for one
hour.
20
ALUMINA EXTRACTION
o
< 30
on
h
X
u ioo
<
*? 8C
i\
2-B
- ^K
4k
1*
G3
o /
/ i
>
KAOLIN
, ILLINOIS
CK/
\OLIN
TE,
Fl
^ECLAY
)RILLONITE)
(KAOLINITE)
1 869 .
LLITE,MONTMC
I 871
g
5
1
►
4
1
<
\ ^
x^<
>
/
K"^
Rll>
JGGOLD
/
r
rLINT
CLAY, M
SSOU
Rl
/
(KA
OLIN
IU
/i
►
(KAOLINITE)
/
,/
<
/
1400 1320 1340
TEMPERATURE DEGREES C
8
\
k
J
2-C
lb-
|
o
kl
/
V
-J
^ (GIQBSn
BAU>
-e, k
CITE
s^OLINITE) i
(
HAR[
(KA
) KA
OLINI
876
DLIN
TE)
l>
875
A
1
k 1
<
i
h
r
i
■ <
1 L^
/
/^
/
BAUXITI
CKAOLINIT
: ka
E,GIE
74
DLIN
BSITE)
. /
/
SOFT
(KA
KAC
DLINI
877
>LIN
TE)
8
<
'/
<
>
f
1400 1320 >340
TEMPERATURE DEGREES C.
Fig. 2.— Parts B and C
Percent A1203 extracted from sinter versus temperature at which sinter was held for one hour.
LIME-SINTER PROCESS
21
<
\
%^^
— \r~
♦
2-0
•
80
70
60
<
50
/
/
'
/
/
BALL CLAY
(KAOLINITE, ILLITE)
/
BALL CLAY
(KAOLINITE, ILLITE)
/
1
87<
d
i
/
880
•
J
k
— |k
^r
i
> S>
^^i
I
i
\
i
\^*
70
60
CK/
PLASTIC KAOLIN
TE)
/
KAOLIN, AIKEN,
s.c
/
1 | 878
/
881
/
X
/
"?J
lo
13
40
• 3
60
80
14
00
13
20
13
40
1360
1380
1400
TEMPERATURE DEGREES C
r,
50
I. J
h
( >
<
40
(X
h-
X
LJ
IOC
m
O
<\j
90
2-E
1
>
*y
i
1
,X1
>
T
tr
u
CILLI"
NDERCLAN
rE, KAOLIh
JITE)
H
(H
ALLOYSITE
ALLOYSITE
)
866
A
/
867
0
&
J
l__
>
f
|^
i
i
\^
i
KAOLIN,
OREGC
HOBART
)N, (KAOL
BUTTE,
NITE)
(MO
BENTONITE
MTMORILLONIT
E)
883
868
1400 1320 1340
TEMPERATURE DEGREES C
Fig. 2.— Parts I) and E
Percent A1203 extracted from sinter versus temperature at which sinter was held for one hour.
22
ALUMINA EXTRACTION
o g
c75
OS ">
lo _ B
• £ 2
£ 2 fcb
^ a
Qo
1 _ <^
OkO
p6s
ifcS
<
CO
0
OOOOO
O
OOOO
vO ~ - ~
- (N-tOOO^O
sD „ ~ -
~ CN i*< GO O
CO" ~ ~
~ CO CO CO -f CO
CO "" ~* ~
" CO CO CO if
00
^ O m
- r*4 1*4
Tf in \0 CO vO vD
^oocM^h-r^
O i*< O -f —1 i*< (N CO O O
co T-t 01 1— 1 CN cN co CN co co
OOOO 000000
in co O 00
CN 1-1" i-H
t^ OO ON ~f CO
i-H CN CM CN CM
MD O 00 -sO co ^O 00 O ^O
i-l ^h o O '-i '-' '-i <N ^H
OOOO OOOOO
r^ 00 cN r^ m CN
CN t-i CM <— < CN cN
CN CN CN CN co CO
OOOOOO
9 °.o
vcs 3 ~- CN -f
CO co CO
OOvDO\C\£)
■f m VO O vO vO
r-*rH,H(NrH^
O 00 CN CO co On
vCtOOwOoO
OOOOOO
ON -f so CN 1— 1 vO '
r- if 00 as on on
CO co CO co co co
VO •sO *0 -O \D
in CN CN CN CN
vO vO vO vO vO
CN CN CN CN CN
CO t*> -f if if rp if
OO CN CN CN CN CN CN
vo CO CO CO CO CO co
00 VO CO CO CO
CO 1-1 i-I i-H 1-4
CO ^f CO V) \o \£> vO lo in in in
OOOOOOO
l-» CO i*1 if CO COCOCOCOCOCO oOrHHCS^ O O >— ihh vO CO CO CO CO CO co M VO vO ^O \0 >fl >o
O ifcN'CNCNcN CNcNcNCN CN eO t^ 1-1 i-h t-i i-h 1-1
^HCNOM^-T-i cor-r^-or^r-. oovoifvo co co \o co co <o — i cn co ^ .— i ,—< o> vo o cn m >-i ^
Ovovor^-co on r^ r^- 1^- r^ r~~ >— icor^-coco Ooooooo r^r^ovooooo ooovav^oo
in in in >-n 1^ iT) in 10 10 id 10 in «-n in vOvcivnvOvO in 10 m >n vO in inin'nvo vo vo
O O VO CN CN OO t^ 0> t^ >0 >-0 CO(JlrHrH
"n^cnirun
On On ON On On
if vO vO O if CO ON
r^ if r^ (N 00 00 co
1—1 On on On 00 00 00
h^ovoh ONvooNoor^-r^ co on co in if r^vocococo if 00 on in — 1 in if cN in r~~ ^o co
1-4 CSCNrtrH O ~H r- I H <-4 ,-H
in CN CN CN CN CNCNCNCNCNCN
c if if cn co m ^o co r^- 00 co
g^cN — (^Hr-H cn m co co r^ on
C O in
vo r-
vor- r» r- on
cn in o -— 1 r-
c -h ^ m n n ^
{^r^ CN CN CN ON O bfl
C CN CO ON CO CO CO
,r- i-h r- 1 cn m o
J5 c
ULo
Uw
is c- * « «
Uv3
J3 c "
Uc/3
LIME-SINTER PROCESS
23
o o o
©
o o o o
o
^O oo o
no - ~ -
- CNTfOOO
NO
nrTt
CO
nnn*
CO
o o
o
o o o ©
©
CN O
sO - . .
- ~ rirfwo
sO
CO TjH
CO - "* -
" co CO CO Tf
CO
© VsO o no
© \o nO NO
1—1 T-i _| (N
oo On *0
(NMrH ^ r» CS r-H r-H
l^©cow-i r^r-i(J\oo
nnrics cn cn cn
co on © oo oo co co © © on
CN »-h r-* rl n rl n -i
CN ON r-H |
©©©
Oh-oO'O no no no ON ©
cool-— i© t- KN(Nth(N
©©©© ©©©©©
Ob-vOvO NO NO OO -"f
© © © © © © © ©
CO i— i CN ON OO Tt< CN CN CO i— <
(NCSrnO © CN <N CN CN CN
©©©© ©©©©©©
vOMO\(N
CO CO — « —•
©©©©
CN CN CN lo on ON ON On ONONONONON On CN CN CN CN CN CN CN CN OO ^ Th ^ Tt< rf *f ^ Tf tJ< rfi t*< rf Tf rp rf<
COCOCO T-i NOCOCOCOCO COCOCOCO CN ^H r-H t^ ,-H T-i T-4 T-i T^ T-1 T-i
>-h t— i i-h no r*~ r~- r~- 1~- no r^~ t~-~ i^~ i^~ co w-> u-i to i_r> irnovoio t— i oo t — t--t^ i^- r~-- r~-- on r^~ on ooco©©©
,-4" r-4 CN CN t-i — I T-i t-h
•-onDso no 00 1— no r^
r~- r- r- r-
1— I Tfrl Wi \£> CO
00 co co co co
uo <-n u-» fX
CO CO CO CO
ur> t— 1 ,-h ,— 1 ,_i ©r-~©
NO CN CN CN CN
OO 'f CO O OO CN CO ON ON CO r-- NO CN OO O
-*1 1— < ON ON ^f OO-tOOt^OM^ CNOOCNu-i
OOCNCN t^wwivO^l OOcOCOcOcO cO-^llOl^cO ON CO 1— 1 OO t^NO^ocOcO 00 ON CO NO CN nO ON r^- © CO ©
OO ON ON rf CO t^i ^ Tf CO ^t1 ^ Th -of t-h ON ON ON ON OO ON ON OO ©COCOCOCO CN CN CO CN CO CN CO ON © ON ON
r)^,^^^ ^-1 T-i T-i T-i r-^ CN MHr- i T-i -^ ,— I ^H _l ^ ,-H ,-H ^ ^ f>J ,_ _H
-Hoooo rici © 00 — ( on no no oo oo co -^ on no r-- oo co co "f oo i-< no cn on ©oncnconoco cooncoonno
f^ r» r» 0s. lo vo Tt< tJt1 CO -<*< tF Tf ■<*< 1— 1 r^- NO NO nO NO t^- nO nO O nO w-> un t*i ^utow^ioio HCSMrn,-
CN CN CN 1^ (N M M CN CNCNCNCNCN sO CN CN CN CN CNCNCNCN vOCNCNCNcN CNCNCNCNCNcN ^O CN CN CN CN
c — ^o o r-
Kfloo '^uiin
ON ON ON rf t-h r- <"> ^ NO r- ©NO©CN
CN 10 © CN OO ££,!>■ CN CN CN CO ON sO OO
g Tf OO ON ©
-h r- .-h © ON CO c NO -h CN CO
CO ON NO i-i T-i OO bfll--- CO CO CO
J3 c- - ■■
uc/3
Ox
_2 c!
U'c/3
Ucyi
24
ALUMINA EXTRACTION
Ih
u
c
c/5
SO
4J 0
H
Mole
Ratio
CaO/Al203
d
X
o
o
<£
I) <"
O c
c/3"""
o2 *
£ 2 &
-a
3
a
Qo
OS >J-1 ^
on
r s c
c
o
U
T
w
Ofe?
<!
3s
to
3^
<
c/)
OOOOO
fN \o Tfi O —■ i
fN (N CO fN CO
o o o o o
co m n v^ n
oo rfi >sO ON On
o o o o
o
o o
fN -f oo ©
^O^ - ^
- CSvO
CO CO co rf
CO
CO CO
o o o o
- (S'tooO
" CO CO co Tf
O vO vO ^O 'sD
O ^O -sO
O O -sO
oo sO co On
(N r-- OS fN rf<
tS (S rn' -h ^
fN fN CO ON
CO rJH rf
fN \0 O © •sOTfioO^OTh
-^ CO CO fN co-^rOCNco
oooo o o o o o
OnOnONOn l^f^OOONON
't'vDiO'* O fN fN
CO co CO fN co •<*< CO
OOOO o o o
On <-o t— i ro *o
OOOOfNON CO l^ OO ON fN
(NcSr-iO ri n ri >-i fN
OOOO OOOOO
oo >~n ^o oo
oo On On oo
oo r-. co
OO ON ON ON
Oro^"*n
oo vO oo On On
■* Tf Tfl Tf Tf CO^f-*1^
•^^f-^-^Tt^ vO'O^'JTO u-> u-> un coOOOO OOOOO
fN fN fN fN fN w->
OOOO OO ON t— i ON O HiOl1^^ TfW-)CO t-h ON OO OO OO oo r^ On r^ oo
,— I ,— I ,-H ^H CO ,-H ,-H ,-H ,-H ,-< ,_( ,-H fN
fNcOCOCOfN TffNfNfNfN fNfNfNfNfN ^ ^O OO t^ VO r^- t-- t^- * VO ȣ) vO ^D sOvO^O^OvC
OOU-irJH N M ""t vO VT)
loOco sCcoo fNoo^^O oo on co r^ »-o
O^i^w^u-i hHHrfi^ cOOON-^cN Or^-vOvOh fSt^^i cOvD^ffNcO CO ^n ON On r-
O00O0O0OOO fNTf-^COfN fNcOfNcOfN rl O O OO CO ONONON 00\0\On£>u-> in\Ov>WNUl
^h ^ ,_ ,_, — h wofNfNfNfN fNfNfNfNfN NO CO CO fN fN fNfNfN CO i— I i— I i— 1 r- 1 ,-h ,-t ,— i ,-h t-h
cooooooooo *— i r^ tjh fN on fN oo © ,— i oo O rl h ^ rl on >-o © hoos'OM oo \o lo r^ >-n
,_ ,_, ,— > ,-h ,-^ coonOnOnoo ooooononoo rHin^rtiTt* cO-^lo ^D-^cocOcO fN co fO CO CO
(N (N rl M (N rf^Hr-^r-M^ ,_< _ ^ ,_ ,_ co -* -h —< --h ,-h ,-h ,-h u-> eN fN fN fN fNfNfNfNfN
CI^Or-l(N
<N \J,~ ™
r^^'*Tf
co co fN © u~» coor)i*vi
CO ^O i—i fN OO
,r- rf Tf r^
c on ^o r^ oo
{^j-j t ~ "^ ^t4 "^
>-o io co i— ' O
CO vO i— i fN 00
WHClHH
W(7j
Uv5
vtf
LIME-SINTER PROCESS 25
o
so ~ ~ ~
ooo o
- (N t^i oo O
O
SO - - ~
o o o o
8. s a
- CN t*< oo O
CO ** ~ ~
~ CO CO CO -<f
co~ ** ~
"* CO CO CO rf
CO - •* -
co CO CO -*1
O O sO CO
— < On -f -f oo^w^^in CO On so r-i ONO>JT* -"f r^ rt< CN DoOvO-hvO oo r^ u-> -h vO X •* O ■*
r) ^- ^_ (SMCSpHrH CSHHH r-l CO CN i r^ -H rJ _H r^ r^ rt (S -H -i PH H C»l r-H 1-H ' C< »■* 1-1 »H
oo O -O sO O <N CN so sO ■* oo "*f so COCNOOCNO ^O -f O "O ^vOMvOrM OO Tf t*i ^n On SO CN r- rs
co co 01 o 't n n ri n co cn cn -h *-< <n cn cn cn co cn cn r^ — i cn cn -h cn cn co cn r-« o cn cn r^ cn
oooo ooooo oooo ooooo oooo ooooo oooo ooooo
OO CO CN OO OOONt^^^ CO r-n CO CN ON ON CO O sO *-h r-i CN ON OO sO rh r-< so t~-~ CO CN CN ^f CN CN t^ CO
OO ON ON OO f-- SO SO ON ON OOOOGOOO r-~ CO OO ON OO ON OO ON OO 1^ V) OO Q\ OO OO ON ON ON CO »0 ON On On
co -^f ^ "^f -^ Tt<ThT^Tt<Tfi OhhhN r^r^r^r^-r- ^ '^ vo \o vo so so so so so coooooooco ococcoccc©
«o r- O On on On on o oo O on oo r- oo r^ oor--oosooo oo j^. t^ co so oo^Ob-r-oo co On on on co on oo oo on o
-<t< CN CN CN CN CN CN CN CN CN "<f CN CN CN CN CN <N CN CN CN Hi^ioioin uhowmo^ oOTfTfTfTf rt< "ti -rf rfi -f
r~- CN "+1 OO ^u-iONOlON ON OO CN u-> cN On CN so sO rf r-< ,-h v/~i CO CO -< ON -< oo ^- _ »^> CO On rt< r^-cO^OOCN
o o
or^ooco ONr^-r^-oor^ i^ooooo\ Ononooooco ooh-coa Or^ONcooo r^ oo oo on Onoooooooo
u-i ON r-> CN On oo oo so On so CN On so co oo >Oh^«ooo O r~- co so ■* "f i— i r- so ■<f -^sOo^ro to u-> oo ri r-
rf on on on oo r-^ oo oo oo oo or^r-i^-so sor^-sososo t^ so so u-> u~> iosOio^io Or^-r^-r^sc or^ or^vo
,*,— lT-l,— I,— I r-l -H r-H r-l r-l -rj< r-H rH r-l r-H t-H r-l r-l r-l r-H COr-lr-lr-lr-( r-H r-H r-H r-l r-H "f i— I r-H r-H r-l r-l r-H r-l r-l r-l
On so -^ ON so CN r-~ On O ON oo oo so CO ON CO CN ^ -^ h On\OhH CosOt^vo ^^coOnO ^rnOrnO
,-^OlCNr-ir-i r^r-^r-i(Nr-H ^ co co co CN CNcocOcoco so t$< co CO co cOcOcococo ""f co co CO CN CN CO co co CO
w-> CN CN CN CN CNCNCNCNCN »0 CN CN CN CN CNCNCNCNCN *0 CN CN CN CN CNcNcNcNCN ^ CN cN CN CN CNCNCNCNCN
-§
^ O ON O r-l SO sO "<f CN r- CCNCNCO-*1 r-l f^ OO CO OO C CO OO ON O CNOOONCNON r-^rHCSCO coonOcoo
kJjOC Tfinvo CO O -h cN OO oxjOO unoio OsOcOCNOO tjjOO iovosO OsOcO-^CN bflOO sOsOsO O sC "f "f ON g
J]
oo
r-
oo ^
_rtj-SS^ S 3 S S S _rf J- - ~ ~ 3 3 S S S J* C ~ ~ ~ 33SS3 J*. C - - 3
Uc/5 Uc73 O'c^ Uc/3
26
ALUMINA EXTRACTION
cof-i
q
<U O r-l
*— 1 « ^^
u
'Jo
9 ,1
'-J X U
rt ^ >r-j
Z M
:&S
Is?
3*
cow
s« „ „
OQOO
*•* *• rf-^N *-^\ **#-\ U4
^O 'O ^O ^O v^
?.,,
9
O — i ■* *o oo co ■<*< so O so
CN CN y-t rHrHCNrHrtrH
^oot(*oo Tf" cn r- so oo so
tStOMO CN <N CN CN T-t CN
oooo oooooo
ntNONvD OO so O CN
DMrHrt r-l CN CN CN
CN CN OO CN CN CO OO O
CO CO CN CN CN CN CN CO
OOOO o ooo
OVOCOW-) ->f .CN-^cN
oooooooo r^Totooooo
or^t^r^-r^ t^-t^r^r^t^-t^- »-o v-> w-> u-> io lovun^io
CN
t^-oOh-C^XCA
vo •* w-) cO
CO t-H r-i r-l
CN ^ ^ tJh tJh
r- oo oo oo oo r- oo r- oo oo oo ^n co cn co cs cn co co co co
ON CO cn o ■* cn u-> o -"f oo
t^- Tfi SO SO -* CN CN CN CN
^in^-to O0000\0\wi rf" CN t^- C\ CO CN^-^^tJh
i—toor^-oooo ^o oo oo r^ t^ Is- On t^- so so so so so so so so
^ r-H r-l r-l r-H r-l rM r-l r-l t-H r-l CO r-l r-l r-H r-l H r-l r-l r-l r-l
""f O SO "** O t^t^i^cOiOTji r-i vo lo o r-^ ■"f CN CN CN CN
CN CO CN CN CN — i CN CN CN CN CN 10 CO CO CO CN CN CO CO CO CO
w-> CN CN CN CN CNCNCNCNCNCN *n CN CN CN CN CN CN CN CN CN
Jjoo so so so
c' u-> so r^- oo
^j-j-^l T^l tJ-I -rjH
OO ,*
Geo
Olo
LIME-SINTER PROCESS
27
EFFECT OF TIME HELD AT SINTERING
TEMPERATURE
A number of sinters were prepared using
selected clay samples for which the time
that the sinter was held at maximum tem-
perature was varied. For all of these sinters,
the lime-alumina ratio was 1.66 and the
temperature was 1360°C. The data are re-
corded in table 8.
Kaolinite clay sinters all showed reduced
extractability when the time held at 1360°
C. was only 10 minutes, but there appeared
to be no advantage in heating them for
periods greater than one hour.
Halloysite clay sinters showed better
extractability with increasing time. This
clay is not well suited to treatment by the
lime-sinter process. The highest temperature
used and the longest heating periods were
insufficient to give good yields from sinters
made with this clay.
The extractability of the gibbsite-kaolin-
ite clay was not affected by variations in the
sinter time within the range of times used.
The montmorillonite clay and the illite
clay (sample 870) appeared to be rather
critically sensitive to overtime heating. The
reason for the bad effect of long time and
high temperature heating is probably related
Table 8. — The Effect of Time Held at Sintering Temperature
Sinter
No.
Time at 1360°C.
in minutes
AI2O3 extracted
in percent
Clay
Sample No.
Type of
clay
201 .. .
12...
225...
10
60
240
32]
88 865
90j
kaolinite
290...
288...
291...
10
60
240
771
78 866
79j
illite,
kaolinite
22...
203...
233...
60
120
240
451
61 867
70j
halloysite
216...
19...
295...
234. . .
10
60
150
240
611
87 1 868
67 f
56j
montmorillonite
217...
56...
226...
10
60
240
571
88 869
90J
kaolinite
218...
26...
227...
10
60
240
721
71
59
870
high iron illite
219...
32...
228...
10
60
240
321
88 873
93j
kaolinite
(flint clay)
296...
41...
299...
10
60
240
921
95 874
92j
gibbsite, kaolinite
(bauxitic kaolin)
221...
47...
230...
10
60
240
491
91 876
92J
kaolinite
222...
59...
231...
10
60
240
451
91 879
90J
kaolinite
(ball clay)
202...
65...
232...
10
60
240
46^
93
93;
881
kaolinite
28
ALUMINA EXTRACTION
to the relatively high content of alkali in
both clays. Although the alkali would tend
to be lost by vaporization from the sinter
at high temperatures and with prolonged
heating periods, its presence no doubt in-
directly inhibits inversion of dicalcium sili-
cate by causing the formation of appreciable
amounts of stable glass. The other illite clay
(sample 866) did not show this sensitivity
to sintering time. The reason for the differ-
ent behavior of this illite clay is obscure
although it may be because it contains an
appreciable amount of kaolinite and con-
siderably less iron than sample 870.
EFFECT OF FURNACE ATMOSPHERE
In a great many tests made by the writers
and also by others, the briquetted sinter
mixes were placed upon a graphite block
because graphite is the only material so far
found to which the sintered briquets do
not stick. It was noted on many occasions
that the end of the briquet which was ad-
jacent to the graphite block was different
in color from the opposite end and that the
color changed more or less gradually as
the distance from the graphite increased.
Furthermore, it was noticed that in some
cases there was a pronounced odor of hy-
drogen sulfide when the sodium carbonate
extracts were acidified in preparing to de-
termine their alumina content. When the
iron content of the sinters was considerable,
the sintered material usually showed a gain
in weight when ignited in oxidizing atmos-
pheres if the briquet was sintered on a
graphite block. All of these phenomena in-
dicated that more or less reduction was
taking place during the sintering process.
It was therefore decided to determine, if
Table 9.
-The Effect of Furnace Atmosphere on
extractability
CaO/AhOs
Mole Ratio
Atmosphere
Oxidizing
Reducing
Clay Sample
No.
Graphite
Crucible
Natural
Gas
Alumina extracted in percent
865
kaolinite
35
1.50
1 . 66
1.83
866
1.40
illite
1.50
1.66
1.83
868
1.40
montmor
llonite
1.50
1 . 66
1.83
870
1.40
illite
1.50
1.66
1.83
871
1.40
kaolinite
illite
1.50
1.66
montmor
llonite
1.83
83
92
88
84
63
78
81
77
83
77
87
5-s
57
71
69
73
80
85
83
83
89
89
92
59
67
77
76
70
80
86
84
63
69
81
82
62
73
79
64
83
93
71
88
59
64
71
78
77
77
84
67
72
81
83
85
81
70
58
LIME-SINTER PROCESS
29
possible, whether furnace atmosphere exert-
ed any considerable influence on the amount
of extractable alumina. Tests were made
with five different clays.
Oxidizing atmospheres were maintained
by placing the briquets in an electrical 1\
heated muffle furnace which contained no
combustible material whatever. Reducing
atmospheres were maintained in two ways :
(1) the briquets were sintered in covered
graphite crucibles; (2) the briquets were
placed in the furnace on a graphite block
and a slow current of natural gas was passed
into the furnace during the sintering period.
The sintering temperature wTas 1360° in
all cases.
The clays chosen for test were numbers
865, 866, 868, 870, and 871. Sinters with
lime-to-alumina ratios varying from 1.4 to
1.83 were made with each of the five clays
using oxidizing atmospheres and also re-
ducing atmospheres. The extraction data
are presented in table 9.
With clay 865 (kaolinite) there was no
significant difference in the amount of
alumina extracted which could be correlated
with furnace atmosphere. With clay 871
(kaolinite, illite, and a little montmorillo-
nite), wThich was chosen because it con-
tained considerable iron, the data are
erratic. The high-lime sinters gave less
satisfactory yields in reducing atmospheres.
Sinters made from clay 866 (largely illite
high in iron) gave slightly better extraction
in oxidizing atmosphere when the lime-to-
alumina ratio was low, but there was no
significant difference when higher lime-to-
alumina ratios wTere used. Sinters prepared
from clay 870 (illite clay very high in iron)
gave definitely better yields of alumina from
sinters burned in reducing atmospheres.
There were no significant differences in
alumina yields from sinters prepared from
clay 868 (montmorillonite) which could be
correlated with atmosphere.
DUSTING OF SINTERS
The speed of dusting is erratic even for
sinters made with the same type of clay and
fired to the same temperature. The com-
pleteness of dusting is occasionaly erratic
especially for sinters made with clays con-
taining alkalies and or considerable iron.
High temperatures and long heating periods
favor rapid and complete dusting. However,
it was onl\ at the highest temperature
(1400°C.) used that the dusting was notice-
ably faster. The average dusting time for
sinters heated to 1360°, 1340° and 1320 C.
was nearly the same for the same type clays.
There were some sinters prepared at all of
these temperatures which required several
days to dust.
Sinters prepared with kaolinite clays
usually dust fairly completely in 5 to 15
minutes. None of the sinters prepared with
kaolinite clays 869, 871, 873, 876, 878, 879,
880, and 881 required more than 30 min-
utes to dust. A few sinters prepared with
kaolinite clays 865 and 877 required from
2 hours to overnight to dust. Only one sin-
ter prepared from a kaolinite clay and sin-
tered at a temperature of 1320° or above
did not dust at all.
The dusting characteristics of sinters
prepared with gibbsite-kaolinite clays were
much like those of sinters prepared with
kaolinite clays. Sinters prepared from hal-
loysite clay 867 dusted quickly but contained
considerable gritty material. Sinters pre-
pared with illite and montmorillonite clay
frequently dusted slowly and incompletely.
Although it is usually true that quick
dusting sinters are apt to extract well, they
do not always do so. For example, sinter 16
from clay 866 was incompletely dusted after
standing overnight yet yielded 81 percent
of its alumina. Sinter 194 from the same
clay dusted completely in 10 minutes and
yielded 77 percent of its alumina. Another
sinter from this same clay dusted in 4 min-
utes but yielded only 50 percent of its
alumina. Examples of less extreme varia-
tion with all types of clay are common. In
some instances the explanation is obvious.
The dusting may be quick but incomplete.
In other instances there seems to be no ob-
vious reason why one sinter dusts in 10
minutes and another requires 24 hours al-
though both give good extraction yields. In
some cases dusting time is related to the
presence of iron or magnesia, as described
on pages 32, 33.
30
ALUMINA EXTRACTION
As might be expected, gritty particles in
the sinter dust are not associated with the
best yields of extractable alumina. Such
particles are sometimes glassy and some-
times resemble particles of burnt clay,
suggesting that the clay and lime have not
always been sufficiently mixed, even though
all batches were tumbled for several hours
in mixing jars, and in most cases were
sieved afterward.
SILICA IN THE ALUMINA EXTRACT
Under the extraction conditions which
prevailed in these studies, sinters from
kaolinite clays gave alumina extracts in
which the average values of the ratio
(100xSiO»)/(Al.O.+SiO0 usually ranged
from 1.5 to 1.7. The weight of silica ex-
tracted from 10 grams of sinter averaged
from 0.02 to 0.03 gram. High sintering
temperatures resulted in lowered solubility
of silica, and low temperatures favored in-
creased solubility of silica. The amount of
soluble silica decreased with increasing lime-
to-alumina ratio, except that some sinters
appeared to have more soluble silica when
the CaO/AbOs ratio was increased beyond
1.83.
Sinters made with gibbsite-kaolinite clay
behaved like those made with kaolinite
clays, so far as silica extraction was con-
cerned, except that the lime-to-alumina
ratio did not clearly affect the amount of
soluble silica.
Sinters prepared with illite and with
montmorillonite clays gave extracts in
which the value of the ratio (lOOXSiO)/
(AbOi+SiO.) was 2-0 to 2.5. This high
value is the result of low soluble alumina
rather than high soluble silica. The average
weight of silica extracted from 10 grams
of sinter was only 0.015 gram. The response
to changes in lime-to-alumina ratio was
similar to that noted for kaolinite clays. It
could not be observed that the amount of
soluble silica varied with sinter temperature.
Extracts from sinters prepared with the
halloysite clay had about the same soluble
silica content as those prepared with kaolins.
There was no definite variation of soluble
silica with either temperature or lime-alu-
mina ratio.
Effect of Minor Components Present
in the Raw Materials
general comments
All clays and limestones likely to be
chosen as raw materials for the lime-sinter
process are apt to contain magnesia, sulfur,
phosphorous, and iron in varying quanti-
ties. It was decided, therefore, to make some
tests for the purpose of determining wheth-
er these elements are likely to exert any
important influnce on the efficiency of the
extraction. The study was limited to one
kaolinite clay. Only in a pure mono-mineral
clay could the effects be easily detected. Clay
877 was chosen because it was the purest
and cleanest kaolinite clay available.
Sinter mixes were prepared using precipi-
tated calcium carbonate as a source of lime.
The mole ratio of lime-to-alumina was 1.66
in all test batches. Quantities of iron, sulfur,
phosphorous and magnesia were added to
these batches in the form of Fe2C>3 (rouge),
CaSCK (precipitated), Cas (PO*)2 (pre-
cipitated) and MgCOs (precipitated), and
in amounts calculated to cover the ranges
most likely to be encountered in practice.
The batches were briquetted and sintered
for one hour at 1360°C. (except as other-
wise noted below).
EFFECT OF MAGNESIUM
Other workers 3 have reported the ad-
verse effect of MgO on the amount of
extractable alumina in lime-clay sinters.
Since MgO is a common impurity in both
the limestone and clays likely to be used
in the lime-sinter process, it was given more
attention than the other minor components.
Precipitated magnesium carbonate was
added to the sinter mixes in such quantities
that the amount of MgO (calculated)
varied from about 1.5 percent to about 37
percent of the weight of AUOs in the fin-
ished sinter. One series was sintered at
1360° and another at 1400°C, both for one
hour. The results are shown graphically in
figure 3, which gives percent of alumina ex-
tracted as a function of the ratio MgO/Ah
Os expressed as weight percent. One sinter
not shown on the figure contained MgO
LIME-SINTER PROCESS
31
1 1
^^-^—
3
*J»
•
1360°
•
V
•,
^s
1
® 1400°
®
<8>
0
%^s
8
• SINTER TEMPERATURE I360°C
® " " I400°C
Ca 0/ai203 1.66
CLAY 877
<8>
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40
Al203
MgO
Fig. 3. — Percent A1203 extracted versus 100 A, ~ in sinter mix.
AI2U3
equal to 242 percent of the weight of alu-
mina. This sinter was made at 1400/C.
Sixty-three percent of the alumina it con-
tained was extracted.
The data show that the alumina yield de-
creased as the MgO increased up to about
14 percent MgO. With further increases
in MgO the yield increased until about 28
percent MgO was present. Larger amounts
of MgO caused no further change in
alumina yield. The scattering of points
about the 1400°C. curve on figure 3 is due
in part to the fact that it was necessary to
make two heats in collecting the data for
this curve. If the data from each heat are
considered separately, there is much less
scattering of the points. This suggests that
some phenomena which is very sensitive to
slight changes in conditions may be involved.
It is of interest to note that the effect of
MgO on the extractability was greater at
the higher temperature. This suggests that
the phenomenon may be related to the solu-
bility of periclase in the liquid portion of
the sinter. Presumably, there is more liquid
in the 1400°C. sinter and hence more
dissolved MgO. Possibly one reason why
more AhOs is extractable in the presence
of excess MgO than in the presence of
smaller quantities is that this excess MgO
inhibits supersaturation of the liquid part
of the sinter with MgO. This would mean
that the saturation point is reached near
the minimum on the curve where the MgO
is about 14 percent of the alumina present.
Reports of the optical examinations, how-
ever, note the presence of periclase in sinters
on both sides of the minima. This would not
rule out the supersaturation hypothesis,
however, unless equilibrium conditions pre-
vailed, and it is most unlikely that complete
equilibrium was reached in these sinters.
Reports of the X-ray examinations indicate
the presence of an unidentified compound in
some but not in all of these sinters. At
1360°C. approximately 65 percent of the
32
ALUMINA EXTRACTION
Q 70
< 60
<
^3C
<» — i
> — i
>
— i
|- —
4
!=*•=
S03 ADDED AS Ca S04
Ca o/a^Oj 1.66
TEMPERATURE 1360°
CLAY 87 7
.2 .3 .4 .5 .6 .7 .8
% S03 IN SINTER MIX (CALCULATED TO IGNITED BASIS)
Fig. 4. — Percent A1203 extracted versus percent S03 in sinter mix.
alumina is extractable in the neighborhood
of the minimum. The unextractable alu-
mina and the MgO present are, therefore,
in the proper proportions to form the spinel
MgO • AI2O3 on that part of the 1360°C.
curve to the left of the minimum.
However, if the formation of MgO.Ab
Os was the whole explanation, we should
not expect the extractability to improve
when furtherMgO is added nor should we
expect the increased temperature to affect
the extractability so markedly, although
the increased temperature might promote
solution of alumina in MgO.AUOs, and
explain the portion of the curves to the
left of the minima. Similar objections might
be raised to any explanation based on the
hypothesis of a phase composed of any
compound involving aluminum and mag-
nesium oxides.
Whatever the explanation of the peculiar
behavior of sinters containing magnesium,
it seems clear that the presence of mag-
nesum in the quantities most likely to be
encountered is bad and that its effect is
aggravated at higher temperatures.
It was observed in the course of this series
of tests that when a group of sinters was
prepared under precisely identical con-
ditions, increasing the quantity of magnesia
shortened the time required for the sintered
briquet to "dust."
EFFECT OF SULFUR
Precipitated calcium sulfate was added
to the mixtures to produce sinters contain-
ing up to one percent of sulfur as S(X The
six sinters prepared gave yields of extract-
able alumina which were the same within
the precision of the work. Sulfur added as
calcium sulfate apparently has little or no
harmful effect. The test data are presented
in figure 4.
EFFECT OF PHOSPHORUS
Phosphorus is present as an impurity in
many limestones and in some clays, and
because it has been reported 4'5 to inhibit
the transformation of p— C2S to the
y-form, it was desirable to obtain some
information concerning its influence on the
lime-sinter process.
Six sinters were made with added Cas
(PCX) 2. The P2O5 content of the sinters
(ignited basis) varied from zero to one
percent by weight. The results presented
in table 10 and figure 5 show that the ex-
tractability falls off as the phosphorus in
the sinter increases.
The effect of phosphorus was worse than
the numerical data indicate. Sinter 312 was
very poorly dusted and contained much
coarse sandy material. Sinter 313 was a
hard glasslike lump that showed no evidence
LIME-SINTER PROCESS
33
Table 10. The Effect oi Phosphorus on rHi
Ex ik \r r \uiu in OF 1 .IME-Cl 11 SlN rERS
Sinter
No.
%P20, in
Sinter
% A120.;
E xtracted
Sinter
Temp.
Dusting
Time
303
309
310
311
312
313
.01 94 1360 1.5 hours
.05 93 1360 1.5 hours
.10 88 1360 overnight
.29 91 1360 overnight
4(; 62 1360 several davs
1.00 72 1360 did not dust
100
<
90
80
• \ j
4
5
•
1
<>
UJ
1-
u
< 60
a.
t-
4
>
uJ 5 0
O
<
P205 ADDED AS Ca3(P04)_,
v°
CaO/AI203 1.66
TEMPERATURE I360°C
10
CLAY 877
.2 .3 .4. .5 .6 .7 .8
% Pa05 IN SINTER MIX (CALCULATED TO IGNITED BASIS)
Fig. 5. — Percent ALOa extracted versus percent P^O-, in sinter mix.
of dusting after several days. It was
necessary to crush and grind it through a
200-mesh sieve before it could be extracted
at all. The tests indicate that no limestone
containing phosphorus should be used as
raw material in the lime-sinter process
without very careful testing.
EFFECT OF IRON
The amount of iron present in good
kaolinite clays usually does not exceed one
or two percent as Fe^O, but may run much
higher in other clays and shales. Examina-
tion of figure 6 indicates that the effect of
iron in the form added is not serious unless
the quantity present in the sinter is greater
than two or three percent. The optical
examination usuallv shows the presence of
the compound 4Ca().AUO<.Fe;0.< in clay-
lime sinters which contain appreciable quan-
tities of iron. This compound does not
easily yield its alumina to dilute sodium car-
bonate solutions so that it is not surprising
that iron reduces the yield of alumina from
clay-lime sinters. The reduction in yield
appears to be somewhat greater than is ac-
counted for by the quantity of alumina in-
volved in tetra-calcium-alumino-ferrite. The
difference may be partly due to the fact that
considerable lime is also tied up in the com-
pound. This, of course, would have the
effect of changing the lime-alumina ratio.
It was observed that increasing the iron
accelerated the dusting rate when con-
ditions under which the sinters were pro-
duced were otherwise identical.
34
ALUMINA EXTRACTION
90
Q 70
y
< 60
CC
h-
X
LU SO
o
_N40
<
20
6
Fc203 ADDED AS SUCH
Ca 0/AI203 1.66
TEMPERATURE I360°C
CLAY 877
12 3 4 5 6
Vo F«203 IN MIX (CALCULATED TO IGNITED BASIS)
Fig. 6. — Percent AljO:; extracted versus percent Fe-03 in sinter mix.
Effect of Different Types of
Limestone
Varying results have been reported" when
different limestones were used as sources of
lime in the lime-sinter process. It was
thought that the differences might be due to
the size of the ultimate crystals of calcite
of which the limestone is composed. Ac-
cordingly, three samples of limestone were
selected which are designatd R7, R2, and
MC. R7 is fine-grained, R2 is intermediate-
grained, and MC is coarse-grained. R7 and
R2 were from the St. Louis limestone taken
from a quarry near Alton, Illinois. MC was
from the Burlington limestone taken from
a quarry near Marblehead, Illinois. The
analyses of the limestones are given in table
5 on page 13-
The limestones were ground so that 95
percent passed through a 200-mesh sieve.
Sinters were prepared using illite clay 870
and kaolinite clays 877 and 880. The lime-
alumina ratio chosen was 1.50. This low
ratio was selected so that small or moderate
effects would not be obscured by the
presence of excess lime. The sinters were
burned at 1360° C. for one hour following
which they were extracted in the usual
manner.
The results presented in table 1 1 indicate
no basis of choice between the three lime-
stones. For a given clay any one of the
limestones or a precipitated calcium car-
bonate which is extremely fine-grained gives
equally good results within experimental
error. The sinters all dusted well and with-
in a reasonable time.
It must be concluded that differences in
limestone insofar as they affect the efficiency
of the lime-clay-sinter process are due to
factors other than the grain size.
Table 11. — Effect of Different Types of Lime-
stone ON EXTRACTABILITY OF CLAY-LlME
Sinters
Clay Sample
Limestone
Sinter
Percent
A1203
extracted
No.
Sample No.
No.
870
R2
314
58
illite
R7
315
59
MC
316
54
Fpt CaC03
25
57
877
R2
317
92
kaolinite
R7
318
88
MC
319
90
Ppt CaC03
49
93
880
R2
320
91
kaolinite
R7
339
86
MC
340
89
Ppt CaCO,
61
93
LIME-SINTER PROCESS
35
MINERALOGICAL ANALYSES
OF SINTERS
Microscopic Analysis
Microscopic studies were made with the
petrographic miscroscope on mounts pre-
pared by immersing the sinter material in
liquids of known refractive index. Magni-
fications up to 900 X were used.
GENERAL COMMENTS
The principal phases developed in the
sinters were CS(2CaO.SiO->) and C5A3
(5Ca0.3Al20;:). The former was present
in euhedral grains and the latter as inter-
stitial material between grains. In some
sinters prepared with high alumina (gibbs-
ite-kaolinite) clays, a small amount of
euhedral CAs could be identified.
In sinters prepared with kaolinite clays
the C2S was usually in grains 5 to 10 mi-
crons in diameter (fig. 7B). The C2S tended
to be smaller in sinters prepared with clays
composed of well crystallized kaolinite than
in those prepared with clays composed of
poorly crystallized kaolinite. In sinters pre-
pared from clays composed of other clay
minerals the same component was frequent-
ly 20 to 60 microns in diameter (fig. 7A).
The OS was usually in the y-form al-
though sinters prepared from all clay
samples contain some uninverted ft— CS.
Sinters prepared from all types of clay
contained a yellow-brown pigmentary ma-
terial that was present usually in particles
less than 2 microns in diameter. It was im-
possible to be certain of the identification
of all of the pigmentary material, but a
considerable part at least was 4CaO.Al*03.
Fe^Os. The amount of this material was
small in sinters prepared with halloysite
or kaolinite clays because such clay minerals,
unlike illite and montmorillonite, do not
have appreciable amounts of iron in their
lattice structure.
Sinters prepared with some clays at low
temperatures contained some material that
showed little or no development of the
new phases. This material was isotropic in
appearance which suggests that the reaction
of the components had not been carried to
completion.
EFFl.ci OF VARIATIONS l\ LIME CONTEN1
IN SINTERS FIRED ONE HOUR AT 1360°C.
Kaolinite clays. — The inversion of C»S
to the y-phase was most complete (90'/ ±)
in sinters with a lime-to-alumina ratio of
1.83. Sinters prepared with a lime-to-alu-
mina ratio less than 1.66 frequently con-
tained small amounts of CA - the amount
increasing as the ratio decreased below 1 .66.
A component identified as CA was present
in small amounts in sinters with a lime-to-
alumina ratio of 1.83.
In general the amount of C A> was about
the same in sinters prepared with lime-to-
alumina ratios of 1.66 and 1.83. The
amount of this component decreased as the
ratio decreased below 1.66.
The amount of pigmentary material in-
creased as the lime content decreased.
Illite and montmorillonite clays. — Sinters
prepared with such clays showed about the
same general effects of changes in lime con-
tent as those prepared with kaolinite cla\
except that they had less tendency to develop
CA or CA. Also the variation in the
amount of inversion of CS to the y-phase
with changes in the lime content was great-
er for illite and montmorillonite clays than
for kaolinite clays. Sinters of the former
type clay with high or low lime content fre-
quentlv contained as much as 70 percent
p-OS.
Diaspore clays. — Sinters prepared with a
lime-to-alumina ratio of 1.66 were composed
largely of CsAs. A small amount {25% ±)
of indistinct material, which may be incom-
pletely reacted, was present also. Sinters
with a lime-to-alumina ratio of 1.83 seemed
to be about the same as the foregoing, but
the indistinct material became more prom-
inent as the lime-to-alumina ratio decreased
below 1.66.
Gibbsite-kaolinite clays. — Sinters pre-
pared with varying lime-to-alumina ratios up
to 1.83 all appeared about the same, being
composed almost wholly of CS and C«A».
Sinters with a higher ratio were somewhat
different because of the presence of indis-
36
ALUMINA EXTRACTION
V
•"^i
--i*;a:
few'
cj !#
.'# t
^
«#
*
,o
Sg^ \* -
Fig. 7. — Photomicrographs of
A. Illite-clay sinter fired to 1360° C, show-
ing distinct development of large parti-
cles of 7C2S, and pigmentary ferrite
(dark material).
B. Kaolinite-clay sinter fired to 1360° C.
showing distinct development of small
particles of 7C2S. The C5A3 is interstitial
between the silicate grains and is not
evident on the photomicrograph.
lime-clay sinters, 300 X.
C. Halloysite-clay sinter fired to 1360° C.
showing the development of some C^S,
and large aggregate masses of material
that appear to be incompletely reacted.
D. Kaolinite-clay sinter fired to 1320° C.
showing the distinct development of
small particles of C2S, and large aggre-
gate masses in which the new phases
are less clearly developed than in B.
LIME-SIN 7 ' E R PROC i SS
37
tinct material which seemed to be incom-
pletely reacted.
Hal lay site clay. — All sinters fired at
1360°C. contained a large amount (40' ,
±) of material which appeared to he in-
completely reacted.
EFFECT OF VARIATIONS IN SINTERING
TEMPERATURE OX BATCHES WITH \
LIME-TO-ALUMINA RATIO OF 1.66 AND
HELD AT SINTERING TEMPERATURE
FOR ONE HOUR
Kaolinite clay. — Sinters fired at 1320°C.
contained a considerable amount of material
(40% — ) that suggested poor development
of new phases (fig. 7D). The amount of
this material decreased as the temperature
increased until in sinters fired at 1360°
C. almost the entire mass was composed of
rather distinct new components, i.e., C->S
and CsAa.
An increase in sintering temperature was
accompanied by an increase in size of the
y— C2S grains from 5 to about 15 microns
and a decrease in the amount of /?— C2S. In
version of C2S to the y-phase was about
complete in sinters prepared at 1400°C.
Illite and montmorillonite clays. — Sinters
fired at 1320°C. were composed entirely of
new reaction products. As the sintering tem-
perature increased the size of the individual
units of y— C2S increased from about 20 to
60 microns, and the amount of (3— C2S de-
creased slightly. The presence of illite or
montmorillonite in kaolinite clays lowered
the temperature required for complete re-
action and increased the size of the C-S
formed at a given temperature.
Diaspore clay. — Sinters fired at 1320°C.
like those fired at 1360°C. were composed
of OAs with a relatively small amount
(25% ±) of indistinct material that could
not be positively identified.
Gibbsit e-kaolinit e clay. — Sinters m ide
from clays of this type reacted to variation
of sintering temperature in much the same
manner as did those made from kaolinite
clays except that the new phases showed
slightly better development at 1320°C and
1340°C.
Halloysite clay. — Sinters fired at 1360°
C. were composed of about 50 percent 01
isotropic material that suggested incomplete
reaction between components (fig. 7C).
The remaining 50 percent was C-S and
CnAs. About 20 percent of the sinter had
the appearance of incomplete reaction after
firing to 1400°C. and only after firing to
1450°C. were the phases about completely
developed. The presence of halloysite in a
clay would increase the necessary sintering
temperature.
EFFECT OF VARIATIONS IN SINTERING
TIME OF BATCHES WITH A LIME-TO-
ALUMINA RATIO OF 1.66 FIRED AT
1360°C.
Kaolinite clays. — Sinters held at 1360° (J.
for only ten minutes contained a consider-
able amount (40% ±) of material showing
poor development of new phases. The re-
maining 60 percent was made up of distinct
particles of C2S and interstitial OAa.
Sinters held at 1360°C. for one hour
were composed almost completely of new
phases, and those fired four hours at 1360°
C. differed from the one-hour sinters only
in having more complete inversion of the
C2S to the y-form and slightly larger indi-
vidual units of C2S.
Illite and montmorillonite clays. — Sinters
fired for ten minutes and 60 minutes at
1360°C, both showed nearly complete de-
velopment of new phases.
Gibbsit e -kaolinite clays. — Aggregates of
isotropic material without the distinct de-
velopment of new phases entirely composed
sinters fired for only 10 minutes. Curiously
the sinter of this clay (unlike those of the
kaolinite clays) fired for 10 minutes, gave
high extraction yields. Sinters fired one
hour and four hours were both completely
composed of new phases — the four-hour
sinter contains a much larger portion of the
C2S inverted to the y-form.
Halloysite clay. — Sinters fired at 1360°
C. for four hours were composed almost
entirely of new phases whereas those fired
at this temperature for one hour contained
about 50 percent of material with incom-
plete new phase development.
38
ALUMINA EXTRACTION
EFFECT OF VARIATIONS IN KILN
ATMOSPHERE
Kaolinite clays. — Sinters with a lime-to-
alumina ratio of 1.66 and fired in reducing
atmospheres differed from similar batches
fired under definitely oxidizing conditions
by showing more complete reaction at a
given firing temperature, more complete in-
version of C2S to the y-form, and less pig-
mentary material. The differences between
the sinters fired under the two sets of con-
ditions were, however, very slight, and no
new phases were found in the sinters fired
in reducing atmospheres.
The foregoing statement does not hold
for sinters with high lime contents. For ex-
ample, sinters with a lime-to-alumina ratio
of 2.00 fired under oxidizing conditions
showed more complete reaction, more in-
version of C2S to the y-form, and less pig-
ment than those fired in a reducing atmos-
phere.
Illite and montmorillonite clays. — Sinters
with a lime-to-alumina ratio of 1.66 fired
in reducing atmosphere at a given temper-
ature showed better development of new
phases, more inversion of C2S to the y-form,
and less pigment than those fired under
oxidizing conditions. The differences be-
tween sinters fired under the two sets of
conditions were pronounced although no
new phase could be found in the sinters fired
in reducing atmosphere.
Like sinters composed of kaolinite clays,
the foregoing statements do not hold for
sinters of either illite or montmorillonite
clays when the lime-to-alumina ratio is high,
e.g., 2.00. In such sinters, there was better
phase development, more inversion of OS
to the y-form, and less pigment when the
firing was done under oxidizing conditions.
Gibbsite-kaolinite clays. — A comparison
of sinters fired under oxidizing and reducing
conditions gives the same results as those
noted above for kaolinite clays.
EFFECT OF MISCELLANEOUS VARIATIONS IN
COMPOSITION
The following data were obtained on
sinters prepared with kaolinite clay 877 to
which various components were added. The
batches had lime-to-alumina ratios of 1.66
and were held for one hour at 1360°C.
Titanium. — No distinct TiO* phase
could be found in any sinters even when a
large percentage of titania had been added
to the sinter batch. X-ray diffraction data,
however, indicate perovskite (CaO.TiOz)
in sinters prepared with added T1O2.
Phosphate. — Sinters to which 0.05 per-
cent P2O5 had been added showed consider-
able reduction in the amount of C2S invert-
ed to the y-form. As the amount of P2O5
in the sinters increased, the amount of GS
inverted to the y-form decreased until in
sinters with one percent P2O5, all of the C2S
remains in the /?-form, and therefore no
dusting took place.
Iron. — In general the only difference that
can be detected in the sinters as the content
of iron increased is an increase in the
amount of the compound 4CaO.Al2O3.Fe2
Os. In clays with a low iron content this
compound was present in minute (2/ndz)
pigmentary units whereas with larger
amounts of iron (6%±Fe203) it occurred
in larger masses interstitially between the
C2S grains.
Magnesia. — Sinters prepared from batch-
es with added MgCOs frequently contained
periclase (MgO) in distinct euhedral
grains with a diameter of about 60 microns.
The calcium silicate in such sinters often
had optical characteristics somewhat un-
usual, such as higher birefringence, suggest-
ing some solid solution phenomenon. It
has not been possible, however, to establish
any definite correlation between extraction
results and the character of such sinters as
seen under the microscope.
EFFECT OF VARIATION IN THE CHARACTER
OF THE LIMESTONE
No differences in character of compo-
nents could be detected in sinters prepared
with either coarse, medium, or fine crystal-
line limestone. Further, no difference could
be detected between such sinters and those
prepared with precipitated calcium carbon-
ate. The foregoing statements appear to
hold regardless of the clay mineral composi-
tion of the clay used in preparing the sinters.
LIME-SINTER PROCESS
39
The only conclusion justifiable on the
above data is that the size of the calcite
units composing a limestone does not deter-
mine its suitability for the lime-sinter proc-
ess.
X-ray Analysis
general comments
The method of examination consisted of
the registration at room temperature of
powder diffraction diagrams of representa-
tive whole sinter preparations. Such patterns
consist of superpositions of the diffraction
patterns of the several individual compo-
nents, each developed in proportion to the
proportion of that component in the sinter.
The ease and accuracy with which a
given component can be recognized depends
not only on the component itself, but on the
accessories which happen to be present in a
particular instance. For phases like CaO
and MgO, 2 or 3 percent may be clearly
evident; for phases like the cubic alumi-
nates, 5 or 10 perecent may be necessary for
identification, and for phases like the lime
silicates, 20 or 30 percent may be somewhat
ambiguous- Under such circumstances, little
value can be attached to the examination
of any one sinter, yet generalizations drawn
from a sufficiently varied group of sinters,
viewed collectively, may be taken with some
confidence.
All of the important phases encountered
in the lime sinters are relatively well known
from the extensive literature on the con-
stitution of Portland cement.
PREDOMINANT PHASES
Calcium orthosilicates. — Of the three di-
calcium silicates, only /3-GS and y— OS
are of significance in the lime-sinter process.
Of the many studies of the /3 to y inversion
(and its failure) the most recent one'' ad-
vances the idea that the inversion may be
inhibited chemically, on one hand, by
foreign ions in the crystal lattice, or, on
the other hand, physically, by the isolation
from each other of enormous numbers of
fine-grained /?^OS particles.
i lie behavior of the lime-sinter prepara-
tions seems to be consistent with such a
concept. For sinters prepared at 1360°C,
those made with kaolinite clays usually in-
verted fairly rapidly, and to about the same
degree. A small amount of /3~CS invariabh
remained uninverted, but it did not keep
the sinters from appearing to be completely
"dusted." For sinters prepared with clays
composed mainly of montmorillonite or
illite, however, dusting usually was slow,
and occasionally failed altogether. In these
sinters the fi~ C2S grains were comparatively
large, and in those in which y— C2S was
developed it was also clearly larger than
in sinters prepared with kaolinite clays.
One is led to the conclusion that the more
abundant impurities in the latter two types
of clays included ions with inhibiting effect.
When the sintering temperature was re-
duced below about 1300°C, dusting first
slowed up, then stopped, even for the
kaolinite sinters. In these sinters f$-C&
was abundantly developed, but the primary
particle size was much smaller (by a factor
of 100 or more). One can hardly say
whether the composition was typical, but the
physical condition was clearly conducive to
physical inhibition. The material reported
by microscopic analysis as showing incom-
pleted development of new phases probably
was composed to a considerable extent of
such fine particles.
For both the y— C2S, whose crystallization
is of the olivene type, and for the /?— C2S,
whose crystallization is not known, minor
variations in the cell dimensions were ob-
served which could be attributed to slight
variation in chemical composition, but no
attempt was made to correlate or interpret
such variations. It seemed quite possible
that the failure of low temperature (i~~ OS
preparations to invert might be due to small
departures from the ideal lime-to-silica ratio
as well as to physical inhibition.
Calcium a! u mi nates. — Three stable alu-
minates have compositions such that they
would be anticipated in materials of the
compositions of the various sinter prepara-
tions. They are the tri-, penta-, and mono-,
referred to respectively as OA, GA* and
CA. It is commonly claimed that the CeAa
40
ALUMINA EXTRACTION
phase actually runs slightly higher in lime
than the formula indicates.
In the region of the compositions em-
ployed, there are two eutectic compositions ;
one involving C-'S, C>A», and CA, another
C*S, C5A3, and OA. Both of these melt
below 1360°C, the normal sintering tem-
peratures used in our studies. In any sinter
then, along with the C2S and CoAs either
some CsA or some CA should also be
present. Probable amounts, though, are only
from about 10 down to 1 or 2 percent, and
detection by the X-ray method is erratic.
The CsA phase was clearly observed only
in one high-lime and high-alumina sinter,
and the CA phase was apparent in only
about half of the sinters examined (those
sinters which were low in lime).
Comparison of this trend with the trend
of extraction data suggests that batch com-
positions had best be made up with residual
lime-to-alumina ratio in the range 1.55 to
1.66 (which includes the C2S— C5A3— CA
eutectic) rather than with a higher ratio
(which would include the CaS- CsAs— CaA
eutectic).
LOW TEMPERATURE SYNTHESES
The course of sinter reactions in the
temperature range between the decarbona-
tion of limestone and the normal finishing
temperature is best followed by diffraction
methods. In order to study these reactions,
powder diffraction diagrams were registered
for a variety of synthetic mixtures fired at
different temperatures. The following con-
clusions are the result of this work.
Lime-silica. — Lime-silica mixtures, in the
ratio of C2S show extensive reaction at
1000°C. The chief phase developed is ft-
C2S which is fine-grained and not subject
to inversion. An additional phase is appar-
ently wollastonite, the /3~CS, but the
amount present does not permit positive
identification. Unreacted lime does not ex-
ceed 2 or 3 percent. Even as low as 800°C.
some reaction is evident, the new phase also
apparently being wollastonite.
Lime-alumina. — Lime-alumina mixtures,
in the ratio of CsAs react slowly over a
range from 1000° or 1100° on to 1300°C.
or possibly higher. At 1100° it has been
possible to observe diffraction effects from
C3A, CsAa, and CA in a single sinter. Con-
tinued temperature rise increases the amount
of GA.t at the expense of the other alumi-
nates.
The source of alumina is a highly sig-
nificant factor. Natural aluminous clays
resist attack by lime more than does chem-
ically prepared hydrated alumina. Diaspor-
ite, for example, dehydrates to corundum
at about 500° C, and some unattacked
corundum is still evident in lime sinters
fired to 1200°C. Other aluminous clays
dehydrate first to an "active" y-oxide which
either may or may not invert to corundum
before it can be attacked by free lime.
Chemically prepared alumina and hydrated
alumina remain active to 1200°C. or higher.
Lime-kaoliniie. — In the critical tempera-
ture range from about 900° to 950°C, de-
hydrated kaolinite is not a simple mixture
of silica and alumina, but possesses a certain
degree of association. This material has
been called "metakaolin." Prepared lime-
kaolinite mixtures in sinter proportions were
fired to a series of intermediate tempera-
tures. At 1000°C. gehlenite is an abundant-
ly developed phase and unreacted free lime
is prominent. Little ft— C2S has yet been de-
veloped, and the presence of any aluminate
is doubtful. From 1000° to 1300° or 1350°
C, free lime and gehlenite gradually dis-
appear and the ft— C2S, CA. and CsAs develop
in the order named. The ft— C2S does not be-
come subject to inversion until the firing
temperature exceeds 1300°C, and apparent-
ly not until the aluminate phases have ad-
justed their proportions to the sinter com-
position. It is suggested that the composition
of ft— C2S may also be undergoing some re-
adjustment in this temperature interval.
SINTERING PROCESS
The conditions for the preparation of the
standard lime sinters, and of the deliber-
ately under-burned sinters described above,
differ from the conditions under which
thermal curves are registered in that more
time is available for progress of the slower
reactions. It is apparent in a qualitative way
on the differential thermal curves (see page
LIME-SIM ER PROCESS
41
42) themselves where the thermal effects on
reactions involving only one mineral are
generally much more prominent than are
those involving two solids.
High alumina and kao Unite clays. — The
endothermic effects up to about 95()°C. are
all known effects ascribable to individual
components.
The intensity of the exothermic effect
near 950° C. seems clearly to be correlated
with the abundance of kaolinite minerals in
the clay, and in turn with the abundance of
gehlenite present in sinters finished off at
only slightly higher temperatures. It seems
not unlikely that the formation of gehlenite
from lime and "metakaolin" could be a
sufficiently fast process to become apparent
in the thermal curves. The even more prom-
inent exothermic effect in the thermal curve
of pure kaolinite ' at this temperature is
considered by the authors to represent the
collapse of "metakaolin" to mullite. Com-
parisons of the respective compositions also
suggest that comparable heat effects would
result from the transition from metakaolin
to mullite, or to gehlenite, respectively. It
is not possible to say whether this same ex-
othermic effect may also include the forma-
tion of some wollastonite.
The further attack of free lime on the
gehlenite, or any extraneous material, is a
slow, gradual process for which no effects
are observable in the curves.
Endothermic effects near the finishing
temperature probably represent the melting
of some of the material.
lllite and montmorillonite. — Curves for
the clays composed of these minerals differ
from those composed of kaolinite in the
lesser significance of the exothermic effect
around 950°, and in the presence of prom-
inent effects in the neighborhood of 1200°.
Examination of the sinters reveals that
the (i~CS developed in them is remarkably
coarse-grained. It is suggested that the small
alkali (and other impurity) contents of
these materials exercise enough mineralizing
influence to render the formation of /3~OS
(and possibly the aluminates) evident in
these curves.
DIFFERENTIAL THERMAL
ANALYSES
Differential thermal analysis consists oi
heating a sample at a constant rate of in-
crease of temperature and recording the
temperature at which thermal reactions take
place and their intensity. The minerals in
a clay may be identified usually from a
differential thermal curve,7 and the develop-
ment of phases that form when a clay-lime
mixture is heated may be studied by the
differential thermal curve of such a mix-
ture.
In figure 8 differential thermal curves are
given for batches of various types of clay
with lime equal to a lime-to-silica ratio of
2.0 and lime-to-alumina ratio of 1.66. The
temperature in the analyses was carried to
1400°C. with a rate of increase of about
10°C. per minute. The downward deflec-
tions of the curves indicate endothermic
reactions, and upward deflections indicate
exothermic reactions. A vertical scale for
determining the temperature difference in-
dicated by the deflections of the curve is
given in figure 9.
Diaspore clay 872, sinter mixture CS35.
— The endothermic reaction between about
500°C. and 600°C. corresponds to the loss
of lattice water (OH) from the diaspore.
The endothermic reaction between about
800° C. and 970° C. is due to the loss of
CO2 from the lime carbonate.
Except for a very slight endothermic re-
action at about 1325°C, which might be
partial fusion, the curve between 970° C.
and 1400° C. shows no thermal effect. The
reaction at 1325°C. is of such low intensity
that it is unlikely that it corresponds to a
major reaction of components. Further, if
the formation of new phases was accompan-
ied by a distinct thermal reaction, it would
be expected to be exothermic rather than
endothermic. X-ray and other data show
the presence of new phases in material heat-
ed to no more than 1360°C, and therefore
since the thermal curves do not show dis-
tinct reactions, it must be concluded that
the new phases form from the lime and de-
hydrated diaspore without a thermal effect
showing on the thermal curves. It would
42
ALUMINA EXTRACTION
Fig. 8. — Differential thermal analyses of lime-clay sinters. See also fig. 9.
LIME-SINTER PROCESS
43
LU
—
o °
■
Q |2
H ,6
z
U 20
DC
u_ 24
L.
Q 28
UJ
5 32
$
cr
LjJ
Q.
UJ
1-
200
500 600 700 800 900
TEMPERATURE - DEGREES C
1200 1300
Fig. 9.
-Scale for determining the temperature differences recorded by peaks of the
differential thermal curves (figs. 8 and 23).
seem that the velocity of such a reaction
must be low.
Kaolinite clay 877, sinter mixture CS50.
— The endothermic reaction with a peak at
600° C. is due to the loss of lattice water
(OH) from kaolinite. The endothermic
reaction between about 800° C. and 970° C.
corresponds to the loss of CO2 from the
lime carbonate.
The reaction due to loss of CO2 is followed
immediately by an exothermic effect in-
dicating that in this clay there is a reaction
between components as soon as CaO is
formed. X-ray data indicate that this re-
action is the formation of gehlenite. Again
in this material there is no thermal effect
shown by the curves corresponding to the
formation of CsAs and C2S and it follows
that these new phases are probably formed
gradually and slowly.
Gibb site -kaolinite clay 875, sinter mix-
ture CS44. — The initial endothermic peak
is due to loss of adsorbed water and sug-
gests that halloysite as well as kaolinite is
present in this clay. The endothermic peaks
at about 340° and 590°C. are the result
of loss of lattice (OH) water from the
gibbsite and kaolinite (and halloysite),
respectively.
The endothermic reaction between 800°
C. and 970°C, due to loss of CO* from
the lime carbonate, is followed immediate-
ly by an exothermic effect which is inter-
preted as resulting from a reaction between
the dehydrated kaolinite component and the
CaO leading to the formation of gehlenite.
The intensity of this exothermic reaction
fits with the intensity of the kaolinite de-
hydration reaction. If this exothermic effect
represented a reaction between the CaO
and both the dehydrated kaolinite and gibbs-
ite, it would be expected to have greater in-
tensity. Sinters made with this type of clay
show the development of C*S and C5A3 at
least by 1320°C, and since the curves show
no thermal effects corresponding to these
reactions it would seem that the new phases
form gradually and slowly. The endother-
mic effect at the end of the curve may
represent partial melting.
Illite clay 870, sinter mixture CS26. —
The initial endothermic reaction is due to
the loss of adsorbed water, and the endo-
thermic peak at about 575 °C. corresponds
to the loss of lattice (OH) water from the
illite.
The endothermic reaction corresponding
to loss of CO* is not followed by a distinct
exothermic reaction suggesting that dehy-
drated illite, unlike dehydrated kaolinite,
does not react with lime as soon as the CO*
is driven off. The thermal reactions between
1150°C. and 1225°C. are probably the
result of the formation of (3— C2S and C5A3.
It would seem that the kind of reaction
and temperature of the reaction is different
44
ALUMINA EXTRACTION
for the CaO and dehydrated kaolinite, de-
hydrated illite, dehydrated gibbsite, and
diaspore.
Kaolinite -illite {ball) clay, sinter mix-
tare CS59. — The initial endothermic peak
corresponds to the loss of adsorbed water
by the illite, the exothermic reaction be-
tween 200°C. and 500°C. is the result of
burning off of organic material, and the
endothermic peak at about 600° C. is due
to the loss of lattice water (OH) from the
kaolinite and illite.
The endothermic reaction due to the loss
of CO* is followed immediately by a small
exothermic effect which is interpreted as
the result of a reaction between the de-
hydrated kaolinite portion of the clay and
lime to form gehlenite. The exothermic re-
action just above 1200°C. is interpreted as
the result of a reaction between the de-
hydrated illite and the lime. This clay com-
posed of two clay minerals, kaolinite and
illite, affords an excellent illustration of the
different temperatures required for the re-
action between the various clay minerals and
lime.
Montmorillonite clay 868, sinter mixture.
— The endothermic peaks at about 150°C.
and 700° C. are due to loss of adsorbed
water and lattice water (OH), respectively,
from the montmorillonite.
Like the illite clay, the reaction due to
loss of CO* is not followed immediately by
an exothermic reaction. Rather there are
exothermic reactions at about 1200°C,
suggesting that this temperature is required
for the reaction of components and the de-
velopment of new phases.
Montmorillonite is like illite and unlik^
either kaolinite, gibbsite, or diaspore in the
temperatures at which dehydrated compo-
nents react with the lime to form new pha-
ses. The higher temperature to bring about
the reaction is probably due to the higher
temperature required for the complete
destruction of the montmorillonite and illite
lattice as compared to the kaolinite lattice.
This will be discussed in more detail for
the lime-soda-sinters where a similar corre-
lation of reaction temperature and clay min-
eral type prevails.
SUMMARY AND CONCLUSIONS
Kaolinite clays and gibbsite-kaolinite
(bauxitic) clays gave higher percentage
yields of alumina than clays composed of
other clay minerals. When such clays were
composed of poorly crystallized kaolinite,
the yield of alumina was reduced slightly.
Halloysite clays gave yields comparable to
those for kaolinite clays only if they were
heated to unusually high temperatures. The
diaspore-clay sinters fused and did not dust
so that they are less satisfactory for this
process as commonly practiced.
It is probably significant that the more
desirable sinters all resulted from those two
clay minerals which yield loose active
products on roasting, i.e., metakaolin from
kaolinite and y— AhOs from gibbsite.
Within the range of variations in the
CaO/ALOs ratios from 1.5 to 1.8, only the
illite clays, perhaps because of their high
iron content, showed considerable variation
in the yield of alumina. Illite clays frequent-
ly gave considerably higher yields with in-
creased amounts of CaO. In the case of pure
kaolinite clays, however, the lime in the
mixture can be reduced considerably below
the quantity required to give the CsAa com-
position without seriously affecting extract-
ability.
Kaolinite, illite, and montmorillonite
clays all showed optimum yields when sin-
tering temperatures were around 1360°
to 1380°C. Sinters made with illite and
montmorillonite clays were more sensitive
to overburning than others. At 1360°C.
sinters made with diaspore and gibbsite-
kaolinite clays fused whereas those made
with halloysite clays required sintering tem-
peratures of 1400°C. and higher for mod-
erately good yields.
The yield of alumina from kaolinite and
montmorillonite clays was not affected by
furnace atmosphere. The effect of variations
in the furnace atmosphere on illite clays
was erratic except that high iron illite clays
gave improved yields when sintered under
reducing conditions.
Sinters prepared with illite and mont-
morillonite clavs tended to dust less com-
LIME-SINTER PROCESS
45
pletelv and less rapidly than sinters made
with other types of clay.
The values for (100XSiO«)/(Al»Oa+
SiOa) were greater for ill i te and mont-
morillonite clays than for clays of other
types, primarily because of low soluble
alumina rather than high soluble silica.
Only in the case of kaolinite clays was the
silica content decreased by increasing the
sintering temperature. Varying the CaO/
Al-O* ratio caused small variations in the
silica in the extracts from kaolinite, illite,
and montmorillonite clay sinters.
In kaolinite-clay sinters fired at a given
temperature the yield of alumina decreased
as the MgO increased up to about 14 per-
cent of the alumina. With further increases
in MgO up to about 28 percent, the yield
of alumina increased. Larger amounts of
MgO caused no change in alumina yield.
The adverse effect of MgO on extract-
abilitv was greater in sinters fired at 1400°
C. than at 1360°C.
No change in the amount of extractable
alumina was encountered when sulphate
was added in amounts up to one percent
sulphur as SO^.
Phosphate even in small quantities in-
hibited dusting and sharply reduced the
yield of alumina. Iron in the form of Fe^O-
did not reduce yields of alumina much un-
less it exceeded one or two percent. In
larger amounts iron caused considerable
reduction in alumina yield.
Differences in limestone, insofar as they
affect the efficiency of the lime-clay-sinter
process, are due to factors other than the
grain size of the calcite units of which they
are composed.
The principal phases developed in the
sinters were GS (2CaO.SiO0 in euhedral
grains and OA^CaOJAhOO as inter-
stitial material. The GS was usually in the
y-form and attained a maximum size of
about 10 microns in kaolinite-clay sinters;
60 microns in illite- or montmorillonite-
clay sinters.
The compound 4CaO.Al-XXFe^ was
present in sinters made with clays contain-
ing appreciable iron. The amount of ferrite
increased as the lime content increased and
was greater in sinters fired under oxidizing
conditions than under reducing conditions.
The compounds CaO.Al^Oa and 3CaO.Al2
Os were identified in some sinters in which
the lime-to-alumina ratio varied below and
above 1.66.
Kaolinite-clay sinters fired at 1320°C
contained a considerable amount (40% ±)
of material that suggested poor development
of new phases. At 1360°C. the new phases
were completely developed, but on increas-
ing the sintering temperature to 1400°C
the units of GS became larger and more
completely inverted to the y-form. New
phases were completely developed in illite-
and montmorillonite-clay sinters fired to
1320°C Higher sintering temperatures
served only to increase the size of the GS
units. New phases were completely developed
also in diaspore clay sinters fired at 1320°
C. Gibbsite-kaolinite-clay sinters showed
about the same relation to sintering tem-
perature as kaolinite-clays except that there
was slightly better phase development at
the lowrer temperature. In the case of halloy-
site-clay sinters, extensive new phase de-
velopment was not attained until about
1450°C. The amount of material with poor
phase development in sinters made with
kaolinite and halloysite clays was reduced
by increasing the time the sinter was held
at the top temperature.
X-ray diffraction analysis indicates that
the material suggesting poor development
of new phases was largely /}— GS in ex-
tremely fine units in which minor variations
in cell dimensions have been observed that
could retard inversion to the y-form.
X-ray and differential thermal analyses
suggest that in diaspore clay sinters the re-
action of components to form new phases
began as soon as the CO^ was driven off
and continued slowly without pronounced
thermal effects. In the case of kaolinite clays
the loss of CO^ was followed immediately
by the formation of gehlenite which was
accompanied by a sharp endothermic re-
action at about 950° C. In such mixtures
(kaolinite clay) fired 1000°C, gehlenite
and free lime were prominent phases. As the
temperature was carried from 1000° to
1300°C free lime and gehlenite disappeared
gradually and the /8-CaS, GA and GAs
46
ALUMINA EXTRACTION
developed in the order named, the latter at
the expense of CsA.
X-ray and differential thermal analyses
for mixtures of illite and montmorillonite
clays did not indicate the formation of
gehlenite. Apparently there was little re-
action of components until about 1200°C.
when the f3~CS and possibly the aluminates
developed sharply with a pronounced ther-
mal effect. The higher temperature required
for the reaction of illite and montmorillo-
nite with lime than for kaolinite and lime
was probably the result of the higher tem-
peratures required for the destruction of
the lattice structure of the former clay
minerals.
LIME-SODA-SINTER PROCESS
47
LIMK-SODA-SINTER PROCESS
From the chemical viewpoint the lime-
soda-sinter process is not fundamentally
different from the lime-sinter process. The
ground siliceous alumina-bearing ore is
mixed with calcium carbonate and sodium
carbonate in the proportions to give a mix-
ture corresponding approximately in per-
centage composition (ignoring volatile con-
stituents) to a mixture of dicalcium silicate
and sodium aluminate, NaAlO--. The mix-
ture is then heated to such temperature and
for such a period of time as may be neces-
sary to render a maximum proportion of
alumina and soda and a minimum propor-
tion of other materials present soluble in
dilute alkali-carbonate solution. The tem-
perature necessary is not so high and not so
critical as for the lime-sinter process. As
in the lime-sinter process the primary
function of the lime is to tie up the silica
in a form not soluble in the leach liquor.
The transformation of dicalcium silicate
when or if it occurs does not result in dust-
ing of the sinter. The sintered mass is
usually considerably less dense than that
produced when the lime-sinter method is
used. It is rather friable, porous, and easily
ground when not overburned.
The clay samples used in these experiments,
after mixing with CaCOs and NaaCOs and
heating, gave sintered materials in which
the compositions in percent for the im-
portant oxides were about as follows in
most cases.
II lite and
Kaolinite clays Montmorillonite clays
(percent) (percent)
SiO^ 23 26
Al2Oa 20 9
CaO 44 53
Na20 + K,0 12 6
Sinters made from the very high alumina
clays and bauxites, of course varied con-
siderably from the above values. Halloysite
clay sinters were similar to kaolinite clay
sinters in composition.
Equilibrium data on the four-component
system AUO^-SiO^-Na^O-CaO are not
available on which estimates of fusion points
can be based, but experiment has shown that
most of these sinters are only partially
fused at 1300°C. Hence, it is to be predict-
ed that an equilibrium state will probably
not be reached at lower temperatures un-
less the mixtures are held at those temper-
atures for considerable periods of time. The
presence of iron, magnesia, phosphorus, and
titania, even in minor amounts, cannot
safely be presumed to be without influence
on the final state reached under a given set
of operating conditions. The nature of the
various clay mineral types, which react
differently when heated, might be expected
to have an important effect on the final
state reached when such mixtures are heat-
ed unless the heating is carried far enough
to produce complete or nearly complete
fusion of the sinter mix. The data presented
were collected from experiments designed
to study the influence of some of these
factors in the lime-soda-sinter process.
EXPERIMENTAL METHODS
Preparation of Sinter Samples
Clay samples were ground to pass 80
mesh, calcined at 800°C. (Sample No. 872
diaspore was not calcined, see page 14),
mixed carefully with the appropriate quan-
tities of precipitated CaGO and C. P. grade
anhydrous NaaCCX The batches were then
pressed into cylindrical briquets 1^ inches
in diameter under a pressure of 5000 pounds
per square inch.
Except in work planned to study the
effect of varying amounts of lime or soda,
batches were made up with sufficient CaCO-
to furnish two moles of CaO for each mole
of SiOa and enough Na*CO;= to supply one
mole of Na-^O for each mole of AUO:. In
calculating the amount of NaaCOa to be
added, account was taken of NaaO and JGO
occurring naturally in the clay. IGO was
considered equivalent to Na*0 mole for
mole.
The briquets were placed in a cold
electrically-heated muffle furnace, heated in
a manner indicated bv the recorder trace
48
ALUMINA EXTRACTION
(fig. 15) except as noted otherwise in
special instances. The briquets were placed
in tightly closed bottles as soon as they
became cool enough to handle.
A few sinters were made with the
briquets resting on a graphite block which
in turn rested on the floor of the muffle.
The color of such briquets indicated that
the furnace atmosphere was reducing under
these conditions. Since oxidizing conditions
are normal in commercial kilns, measures
were taken to insure oxidizing conditions in
the test furnace except as otherwise noted
for studies on the effect of furnace atmos-
phere. The briquets thereafter were sup-
ported in small clay dishes with a thin
briquet of composition identical with that
of the test briquet interposed between the
dish and the test briquet. After sintering,
the thin briquet was discarded. No trouble
was encountered until the temperature
1300°C. was reached when the corrosive
action of the briquet on the clay dishes
began to be serious with briquets made from
most of the clays. Briquets from clays con-
taining considerable iron showed signs of
incipient fusion at 1300°C. Briquets from
fairly clean kaolins could probably be heat-
ed to temperatures somewhat higher than
1 300° C. without much distortion if support-
ed on graphite or otherwise protected from
the slagging action of the clay dish. As a
rule the sinters shrank during the heating
process, although some prepared at the
lower temperatures swelled noticeably.
Chemical Analysis of Sinters
The sintered material was analyzed for
silica and R-Os by standard methods. The
values for alumina were obtained by cor-
recting the R*Oa values for Fe2C>3 and TiO*
as calculated from the batch.
Extraction of Alumina from Sintered
Material and Analysis of Extracts
After some preliminary experimenting
the method of extraction described below
was adopted as the standard procedure for
this series of tests. The sintered briquet was
ground to pass 65 mesh. A 10-gram sample
was weighed out, and the quantity of AhO
present in the sample was calculated from
the batch composition. A quantity of Na2
COs (in the form of a standard solution)
was taken, sufficient so that the total Na20
available including that in the sinter
amounted to 1.8 moles Na20 per mole of
AhOs in the 10-gram sinter sample. This
NasCOs solution was diluted to 100 cc,
heated to 65 °C, and maintained at this
temperature while the ground sinter was
introduced in such manner that no lumps
were formed. The suspension was stirred
with a motor stirrer for 15 minutes, at a
rate which would prevent sedimentation,
and filtered immediately with suction. The
solution was made up to 250cc. and ana-
lyzed gravimetrically for ALOs and colori-
metrically for SiCX
EXPERIMENTAL RESULTS
Effect of Type of Clay Minerals
optimum yield of alumina
The factors which may be varied at the
will of the operator of a plant producing
alumina by the lime-soda sinter process are
the lime-clay and soda-clay ratios, the tem-
perature, the duration of the heating period,
and to a lesser extent the furnace atmos-
phere. Proper adjustment of these variables
with relation to one another might be ex-
pected to result in an optimum yield for a
given raw material, assuming good ex-
traction practice. Data are presented in
graphic and tabular form in table 12 and
figures 11, 12, 13, and 14, which bear on
the effect of these operating variables when
various types of clays are used in the sinter
mix under laboratory conditions. Variables
connected with the extraction process were
not considered except that care was taken
to keep the extraction conditions standard-
ized.
The data indicate that the kaolinite clays
and gibbsite-kaolinite (bauxitic) clays may
yield more than 90 percent of their alumina
when clean, that is, when such clays are
relatively free from iron and clay minerals
other than kaolinite and gibbsite. Halloy-
site and montmorillonite clays and anortho-
LIME-SODA-SINTER PROCESS
49
lOA
• — EXTRACTION VALUES
ACTUAL SINTERS
— $ — EXTRACTION VALUES
HYPOTHETICAL SINTERS
)
1
90
A
—»
- -<
5>
_— «>
5^
•*
1
80
s^
•- 878
y
ly _
-•- 87<
./>/
.'
^
\^Z-
ESTIMATED
/
ESTIMATED
/
25 % ILLITE
60
REST - KAOLINITE
0
H--
* - T
75 % KAOLINITE
1300 1000 1100
TEMPERATURE- DEGREES C
00
.--Jj
J>
10-B
9 0
^-c
9
>
/
t>~
ao
t
>--
/ y
'/
r
/I
//
1
-— ^
V
J-"
bO
/ /
/
bo
/
>—
^66
>— 8
71
40
1
JC
— / —
ESTIMATED
80% ILLITE
20% KAOL 1 M 11
20%
ESTIMATED
ILL ITE & MON
JTMO
RILLO
NITE.
3 0
C
L >'
r
-~ —
'E
(MOSTLY ILLITE)
80% KAOLINITE
20
1 0
i
300 1000 1100
TEMPERATURE - DEGREES C
Fig. 10.— Parts A and B
Comparison of actual extraction values for lime-soda sint:rs prepared from clays containing more
than one clay mineral type with hypothetical mixtures of clay: 877 kaolinite, 870 illite, and
868 montmorillonite.
50
ALUMINA EXTRACTION
site yielded nearly as great a percentage of
their alumina as the kaolins, but it was
somewhat more difficult to get top yields.
The top yields from illite clays ran 10 or
12 percent below those from the other
clays. The single diaspore clay studied did
not yield quite as high a percentage of its
alumina as most of the kaolins.
Clays which are composed of mixtures of
clay minerals behaved about as would be
predicted from a consideration of the be-
havior of the nearly pure types. The graphs
in figure 10 are comparisons of actual yields
from clays which contain considerable
amounts of more than one type of clay
mineral with hypothetical mixtures of pure
types. For example, clay 879 was estimated
to contain 25 percent illite and 75 percent
kaolinite. The yield from clay 877 (pure
kaolinite) at 1200°C. was 94 percent; that
from clay 870 (nearly pure) illite was 76
percent. Seventy-five and twenty-five per-
cent of these values were 70.5 and 19
respectively. The sum of these is 89.5 which
is taken as the yield of a hypothetical mix-
ture containing three parts by weight of
clay 877 and one part by weight of clay
870.
Hypothetical yields from mixtures of 877
(kaolinite) with 870 (nearly pure illite)
and with 868 (nearly pure montmorillo-
nite) are plotted on the same diagrams (fig.
10) with actual yields from clays 866, 871,
878, and 879. The agreement is considered
good, taking into account the difficulty of
making accurate estimates of the relative
amounts of the clay minerals in a natural
mixture and that there are factors other
than the clay mineral types which are diffi-
cult or impossible to evaluate.
SINTERING TEMPERATURE
Fourteen representative types of clay
materials and one anorthosite comprising
samples 865, 866, 867, 868, 869, 870, 871,
872, 875. 877, 878, 879, 881, 883, and 884
II- A
MOLE RATIO Na20/A|203=l
ii Ca 0 / Si 02 =2
SHALE
(ILLITE )
870
^\
\
/
/
4
/
1
A
t
/
1
UJ
1
/
<
CL
/
/
/
\-
X
UJ
4
t
i
\^
/
/
>
/
o
_f\J
<
Kfi
OLIN , 1 LL IN 01
KAOL 1 NITE)
s
5L_
**
.
869
4
> —
EOO 1300 (000
TEMPERATURE -DEGREES C
Fig. 11. — Part A
Percent AI2O3 extracted versus temperature at which sinter was held for ten minutes. See pp. 52.
LIME-SODA-SINTER PROCESS
51
i_
ll-B
*~~\
N,
DIASPORE
( MISSOUR 1 )
872
BAUXITE
(GIBBS ITE. KAOLINITE)
875
>"*
)
— *
S
A
i
i
t- ■*
X"
\'
— i
IT
( KAOLINITE
FIK tt-LAY
, ILLITE, MONT
MORILLONITE)
KLAb 1 IC KAULI N
(KAOLINITE, MONTMORILLONITE
a 1 1
878
1
1000 MOO 1200 1300 1000 1100 1200 1300
TEMPERATURE - DEGREES C
<
^ ^
pn- C
/.
— ^
//
//
//
//
UNDERCLAY
( HALLOYSITE)
i
1,
V M
LL ITE, KAOLINITE)
867
1
V
866
-
f
/
^i
>- "~
i
>- ^
Q
ID
o
Q
Id
h
o
i
a.
X
UI
<
f-
X
u
U
z
o
IU
z
o
KAOLIN, RINGGOLD, TENN.
1 BENTONITE
f k aoi initf )
( MDNTMORII 1 ONITF )
865
868
/
/
1000 I I 00 1200 1300 1000 1100 1200
TEMPERATURE - DEGREES C
Fig. 11.— Parts B and C
Percent AI2O3 extracted versus temperature at which sinter was held for ten minutes
52
ALUMINA EXTRACTION
LU
h-
U 50
<
CE
h-
X
Id
<*>
° 90
(M
<
K80
lll-D
J
i
r^
1
KAOLIN AIKIN,S.
Ckini i mitp 1
c
SOFT KAOLIN
( KAOLINITE )
877
88 1
^ -
- - 1
>- ^
>
x
/
►
/
">
>
ir
t^
1
r
BALL CLAY
( KAOLINITE . ILLITE1
879
KAOLIN, HOBART BUTTE, ORE.
( K AOI IN ITF ")
883
i
1200 1300
TEMPER AT U
were mixed with CaCOs and Na2C(X The
mole ratio of soda to alumina was one and
that of lime to silica was two. Briquets
made from each were heated up to top
temperatures ranging from 1000°C. to
1300°C. as indicated in table 12, held there
for ten minutes, allowed to cool, ground,
and extracted. The data are presented in
table 12 and in figure 11.
It will be noted that the curves showing
the behavior of clays of a given type are
quite similar in shape although the yields
of soluble alumina vary considerably. For
some clays the yields from sintering at
1000°C. and 1100°C. are about the same
but there is a definite increase in yield when
the temperature is raised to 1200°C. and
higher. The reason for this break is partly
that, since it required longer to heat to the
higher temperature, the effect involves both
a time and a temperature factor, which
circumstance tends to magnify any increase
in extractability due to increased tempera-
ture. In some cases where the variation was
most extreme a new series of sinters was
made for which the time that the sinters
were kept above 1000°C. was held constant
RE
1000 1100
DEGREES C
<
cr so
ll-E
ANC
)RTH0SIT
E
/
>
884
/
>
/
1000 1100 1200 1300
TEMPERATURE - DEGREES C
Fig. 11.— Parts D and E
Percent AI2O3 extracted versus temperature
at which sinter was held for ten minutes.
and equal to three hours, as nearly as this
could be done. This was done for clays 865,
866, 870, 871, and 879, and the resulting
data appear on figure 11 as dotted line
curves.
Kaolinite clays, if quite pure, gave nearly
the same percentage yield of alumina for all
temperatures in the range 1100°C. to 1300°
C. when the time was the same. The opti-
LIME-SODA-SINTER PROCESS
53
mum temperature is probably about 1200°
C, Yields ranging upward from 75 percent
of total alumina are obtainable from good
kaolinite clays even at 1000°C. Imperfectly
crystallized kaolinite clays such as 869 and
883 did not yield so well, especially at the
lower temperatures.
Sinters prepared from gibbsite-kaolinite
clays are less sensitive to temperature varia-
tion than those from kaolinite clays. The
yields were about the same at 1000°C. as
at 1300°C despite the fact that the heating
period was longer at the higher temperature.
Sinters prepared from diaspore clay be-
have toward variation of temperature much
like those from gibbsite-kaolinite clays. The
small rise in yield with increasing temper-
ature noted for 872 was probably mostly
due to the longer period required to reach
the higher temperatures.
No alumina could be extracted from
sinters prepared from montmorillonite clay
at temperatures up to 1100°C. Even at
1200°C. only half of the alumina was made
soluble. The extraction yield went up to 90
percent at 1300°C.
Table 12 — Effect of Sintering Temperature on Percent AI2O1 Extracted with Various Types of
Clay.
Time at top temperature — 10 minutes
Na20/Al203 = 1; CaO/Si02 = 2; see Figure 14
Clav Sample
No.
Sinter
Sinter Temp.
Degrees C.
Percent AI2O3
Percent AI2O3
100 Si02
No.
in Sinter
Extracted
AhOi+SiOj
in extract
865
51
1000
19.7
74
1.2
37
1100
«
76
2.4
70
1200
«
87
2.1
82
1300
«
90
1.8
866
52
1000
9.7
52
0.2
38
1100
ft
44
1.4
71
1200
«
75
2.7
83
1300
ft
85
1.7
867
53
1000
21.2
71
1.8
39
1100
«
72
2.0
72
1200
«
83
1.9
84
1300
«
89
1.7
868
54
1000
9.2
40
1100
ft
—
—
73
1200
«
44
1.5
85
1300
ft
89
1.7
869
55
1000
15.0
69
1.3
41
1100
«
69
.2.5
69
1200
«
82
2.4
86
1300
»
90
2.2
870
56
1000
9.5
13
2.4
42
1100
«
22
0.0
75
1200
«
75
3.8
87
1300
«
79
1.8
871
57
1000
13.3
65
2.3
43
1100
«
64
2.5
76
1200
«
79
2.2
88
1300
«
91
1.8
872
58
1000
46.1
80
2.6
44
1100
«
81
2.1
66
1200
«
82
2.1
89
1300
«
86
1.6
54
ALUMINA EXTRACTION
Table 12 — Continued
Clay Sample
No.
Sinter
Sinter Temp.
Degrees C.
Percent AI2O3
Percent AI2O3
100 Si02
No.
in Sinter
Extr acted
AhOa+SiO,
in extract
875
59
1000
32.1
85
2.5
45
1100
32.2
87
1.8
77
1200
32.2
92
2.2
90
1300
34.3
85
2.0
877
64
1000
19.4
80
2.9
50
1100
19.6
82
2.3
65
1200
19.9
94
1.9
95
1300
20.2
96
1.8
878
60
1000
17.7
73
2.7
46
1100
17.7
76
2.3
78
1200
17.9
86
1.9
91
1300
18.0
91
1.9
879
61
1000
16.5
76
2.0
47
1100
16.5
74
2.1
79
1200
16.7
90
2.0
169
1300
16.7
89
1.9
881
62
1000
18.7
77
2.3
48
1100
18.6
76
2.2
80
1200
19.0
86
2.3
93
1300
19.1
93
2.0
883
63
1000
17.0
62
2.2
49
1100
17.4
64
2.1
81
1200
17.5
82
1.8
94
1300
17.7
87
1.8
Sinters prepared from illite clays yielded
poorly at temperatures up to 1100°. The
yield was much higher at 1200°C. but little
was gained by increasing the temperature
above 1200°C.
EFFECT OF TIME HELD AT TOP SINTERING
TEMPERATURE
A series of briquets was prepared in
which the mole ratio of soda to alumina
was one and that of lime to silica was two.
These briquets were then heated to 900 °C.
at a rate of about 50° per hour. The rate
was then increased to approximately 250°
per hour and continued until the temper-
ature reached 1100°C. The temperature
of 1100°C. was held for 10 minutes with
one set of briquets, 60 minutes for another
set, and 120 minutes for a third set. The
temperature of 1100°C. was selected be-
cause the extraction values for short sinter-
ing periods at this temperature were low
enough to allow improvement with more
extended periods.
The results obtained from this series of
tests are presented graphically in figure
12. All samples except the montmorillonite
clay showed some improvement in yield of
alumina with increased time of heating.
Anorthosite showed more improvement than
any of the clays. This is probably because
of the relatively large size of the anortho-
site particles in comparison with the clay
mineral particles that make up clays. Al-
though the anorthosite was ground so that
90 percent passed the 200-mesh sieve, in-
dividual particles are relatively large. Each
200-mesh particle of anorthosite can be
visualized as a single fragment of this min-
eral. On the other hand, a 200-mesh par-
ticle of clay is composed of a large number
of smaller single particles aggregated to-
gether. Furthermore, the breakup of the
clay structure on heating might be expected
to result in a porous material which pre-
LIME-SODA-SINTER PROCESS
55
100
90
80
12-A
Ci
UNDERCLAY
LLITE, KAOLINITE
866
0
MOLE RATIO Na2o/AI203=l
" ii CQ0A1O2 -Z
O *n
h-
o
< 50
a
\-
X
UJ
(0
° on
<
>
N90
<
•
p
^AOLI
N
(KA
OLIN
865
TE)
Challoysite)
OD /
40
20
20 20 60
TIME (MINUTES) HELD AT I I00°C
PO
12-B
(1
HAL
LLIT
:)
W*
1
KAOLIN
870
(K/
UDLIN
869
TE) -
60
70
c
\-
<
a
>
- L
u
\
1
c
b
1
j
c
h
>
J
:
c
K
X
<
er
1-
X
I—
1
2
0
2
UJ
2
O
2
50
r
MON-
BEN
l"MOR
TONITE
ILLONITE
)
(KAOLI
MITE,
FIRECLAY
ILLITE, MONTK
rfORIl
.LONI
TE)
868
e
>7I
60 120 20 60
TIME (MINUTES) HELD AT IIOO°C
Fig. 12.— Parts A and B
Percent A1203 extracted versus time sinter was held at 1100°C.
56
ALUMINA EXTRACTION
100
!
»
_____
12-C
•
4
>
BAUXITE
SOFT KAOLIN .
Q _~
(Gl
3BSITE, KAOLINI1
rE)
(KAOLINITE)
ail
UJ
1-
o
O / _>
< 50
or
\-
X
UJ
<
f
i
4
*
J> 80
70
D
IASPC
872
)RE
PLASTIC KAOLIN
(KAOLINITE, MONTMORILLONITE")
878
40
120 20 60
TIME (MINUTES') HELD AT I 100-C
100
12-D
^ — <
, (
i
•^
■"
K
AOLIt
g
(KA(
DLINI
881
rE)
Id
o
AOLIf
883
*JITE
o.
\-
X
_ i
° SO
<NJ90
<
^ 80
r
BALL CLAY
^AOLI MITF. II 1 IT
fO
879
CA
SJORTHOSIT
884.
E)
40
60 12.0 20 60
TIME (MINUTES) HELD AT I 100-C
Fig. 12.— Parts C and D
Percent A1203 extracted versus time sinter was held at 1100°C.
LIME-SODA-SINTER PROCESS
57
sumabh could absorb molten sodium car-
bonate so as to allow relatively quick re-
action all the way to the center of the
particles. The less porous anorthosite unit
would be exposed to the action of soda and
lime only at the surface of the particle.
Sinters prepared from pure kaolinite
clays showed continuous though moderate
improvement with increasing time. The
yield from those prepared from imperfectly
crystalized kaolinite clays (869 and 883)
improved more with increasing time than
did the others.
Sinters prepared from diaspore clay were
less affected by the time of sintering than
those prepared from kaolinite clay.
Sinters prepared from gibbsite-kaolinite
clay behaved much like those prepared from
kaolinite clays toward time variation ex-
cept that the improvement in alumina yield
was somewhat less.
Sinters prepared from illite clay 870
showed great and nearly linear improve-
ment with increased sintering time but
those from clay 866 (estimated to contain
80 percent illite) improved only moderate-
ly with increased time. Possibly the reason
for the difference is related to the fact that
the iron content is very high in clay 870.
No alumina could be extracted from
sinters prepared from the montmorillonite
clay heated to 1100°C. even after sintering
for 120 minutes at this temperature.
HFFECT OF HEATING RATE
A series of briquets, covering the same
samples and made up to the same specifica-
tions as those used in the series discussed
in the preceding section were heated to
1 100°C on a heating schedule in which the
rate of increase averaged about 50° C per
hour over the range from 900° to 1100°C.
The part of the heating schedule below
900° was similar to that illustrated in
iigure 15. The yields of soluble alumina
were almost exactly the same in nearly all
cases as those shown on figure 12 for
the period in which the sinters were
held at 1100°C. for 60 minutes. These
sinters (figure 12, 60 minutes) were above
1000°C. for about 95 minutes, whereas
the sinters heated at the slower rate were
above 1000° (J. for about 120 minutes.
There was, therefore, no considerable effect
on the yield which could be ascribed to the
slow rate of heating. The character of the
briquets was, however, noticeably different
in that the) were more friable than similar
briquets heated to the same top temperature
at a more rapid rate. Some of them had
increased in diameter as much as \/% inch.
Others, although showing some shrinkage,
did not shrink as much as similar briquets
which had been heated more rapidly.
EFFECT OF FURNACE ATMOSPHERE
Sinters were made at 1200°C. using the
same composition for the sinter mixes as
in the tests described in the two preceding
sections with the furnace atmosphere ren-
dered strongly reducing by passing a current
of natural gas into the furnace. Comparison
of the extraction yields from these sinters
with those from similar sinters burned un-
der the normal oxidizing conditions at the
same temperature disclosed no differences
which could certainly be correlated with
atmosphere. There wrere twTo cases in which
considerably better yields were obtained
with reducing atmospheres than with oxidiz-
ing atmospheres. These were from sinters
prepared with the montmorillonite clay
and with the anorthosite. However, it is not
considered that the small amount of data
justify more than a tentative conclusion in
these two instances.
EFFECT OF SODA-TO-ALUMINA RATIO
A series of sinters was made in which the
mole ratio of soda to alumina was 0.8, 1 .0,
and 1.2. The results obtained from these
tests are presented in figure 13. In all cases
the effect of increasing the Na*0/Al-0
ratio was an improvement in the yield of
soluble alumina. The degree of improve-
ment, however, varied greatly. The data
are erratic in some cases. This frequently
happened when some of the sinters were
rather completely vitrified. When such a
sinter is ground to pass 65-mesh in prepa-
ration for extraction it will contain more
particles close to 65-mesh size and fewer
58
ALUMINA EXTRACTION
90
eo
70
60
to
40
3-A
t
>^
MOLE RATIO Ca O/Si 02 = 2
® REDUCING ATMOSPHERE
/
^
UNDERCLAY
(1 LL ITE. KAOLINI
rE)
866
90
70
60
50
40
1
r
\
(KA
<AOU
OLIN
M
TE)
(HAL LOYSITE)
Rfi7
865
MOLE RATIO Na20/Al203
U 50
<
a:
.eo
13-B
<
$ ^^
Cka
(AOLI
OLIN
N
TE)
<
/'
<
dHAL
LLI T
87C
E
869
<s
>
/
i
Dl
ASPO
872
RE
/
(M
BEN
ONTN
TONI
/lORIL
TE
LONI"
m
N
1
368
MOLE RATIO Na20/Al203
Fig. 13.— Parts A and B
Percent A1203 extracted versus mole ratio Na20/Al203
Sinter temperature 1200° C. ; sinter time 10 minutes at top temperature.
LIME-SODA-SINTER PROCESS
59
100
1
13- C
i
<
A
^ i
t
80
SOFT KAOLIN
1/
7A-
70
(KAOLIN
A77
ITE)
<
60
O
u
h
(KAC
)LINITE ,
MONT
878
MORILLONITE)
U 50
<
cr
Ul
«
i
<
/
^
!
K 8U
" (Gl
B/
\UXITE
IITE")
«
r
70
BBSIT t, kaulit
875
1 BALL CLAY
(KAOLINITE, ILLI
Tt)
60
879 "
50
40
.
MOLE RATIO Nd20/Al203
too
90
80
13-D
6
a
KAOLIN
KAOLINITE'
60
(r
;
86
*3
ANORTHOSITE
884
90
80
70
60
^
►
B
i ^^.
«
1
<
r^"
KAOLI
M
(KA
DLINI
8RI
TE)
PLA
Ckao
MON
5TIC F
LINIT
TMOR
Q-
IRECL
:,ILLI
ILLON
AY
ITE)
40
/
O / 1
MOLE RATIO N<i20/Al203
Fig. 13.— Parts C and D
Percent A1203 extracted versus mole ratio Na20/ALO:t
Sinter temperature 1200°C. ; sinter time 10 minutes at top temperature.
60
ALUMINA EXTRACTION
smaller particles than would a more friable
material ground to pass the same sieve.
The typical kaolinite clays show some-
what less variation in alumina yield with
increasing Na^O/ALO^ ratio than do im-
pure kaolinite clays 871, 878, 879 or clays
with imperfectly crystallized kaolinite, such
as 869 and 883.
The illite clays reacted to variation in
soda-to-alumina ratio in much the same
way as did kaolinite and gibbsite-kaolinite
clays.
The data for montmorillonite clay were
erratic but it is probable that more exten-
sive tests would give results similar to those
obtained with illite clay.
Sinters prepared from the diaspore clay
were rather insensitive to variation of Na*0
/AbOa ratio insofar as extraction yields
were concerned.
Anorthosite showed considerable im-
provement in extractability with the high
NasO/AhOs ratio.
EFFECT OF LIME-TO-ALUMINA RATIO
Data are presented graphically in figure
14 which indicate the variation of the per-
cent alumina extracted with CaO/SiO*
ratios 1.8, 2.0, and 2.2. The Na«0/A1.0.
ratio used was 1.0 in all cases except for
kaolin 877 where it was 1.2. These sinters
were burned at 1200°C. according to the
schedule shown in figure 15.
Sinters prepared from kaolonite or illite
clays give sharply reduced yields of alumina
if the CaO/SiO* ratio is less than two, but
the yields are not improved by increasing
the value of this ratio above two.
Sinters prepared from the montmorillo-
nite clay showed improvement with excess
lime. It is not clear why montmorillonite
should react so differently as compared to
other clays toward excess lime.
SILICA IN THE ALUMINA EXTRACT
Sinters from kaolinite clays gave alumina
extracts in which the average values of the
ratio 100XSiO/(Al2O3+SiO<) lie in the
range 2.0 to 2.3. The weight of the silica
extracted from 10 grams of sinter averaged
about 0.037 gram for pure kaolinite clays
and somewhat less for imperfectly crystal-
line kaolins such as 869 and 883. Sinters
from kaolinite clays which contained appre-
ciable amounts of illite or montmorillonite
showed less silica in their extract than did
pure kaolins. High lime-clay ratios decreased
the silica in the extracts from sinters pre-
pared from kaolinite clays.
Sinters prepared from illite clays and
montmorillonite clays gave extracts in
which the average values of the ratio 100X
SiO2/(Al.O.+SiO0 were 2.3 to 3.2. The
weights of silica extracted from ten grams
of sinter averaged about 0.014 gram. High
lime-clay ratios decreased the weight of
silica in the extract in most cases.
Sinters prepared from halloyiste, dia-
spore, and gibbsite-kaolinite clays gave ex-
tracts in which the silica-to-alumina ratio
was about the same as for kaolins.
Anorthosite was also much like kaolinite
clays insofar as the amount of soluble silica
in sinters prepared with it was concerned.
Effect of Minor Components Present
in the Raw Materials
general comments
Practically all clays and limestones con-
tain compounds of some or all of the el-
ements iron, magnesium, titanium and phos-
phorous. Sinter mixes were prepared, burned,
and extracted as described above with
the exception that quantities of Fe^Os, Cas
(PO.) 2, TiO, and MgCO were added.
The quantities were chosen so that the
ranges likely to be encountered in naturally
occurring clays or limestones would be
covered. Clay 877 was chosen for the tests
for the reason that it contained unusually
small amounts of all of these substances
except TiOa and because it contained no
clay mineral other than kaolinite. The tests
were carried out at 1150°C.top temperature
with a firing schedule as illustrated in figure
15. The data are presented graphically in
figures 16, 17, 19, and 20.
LIME-SODA-SINTER PROCESS
61
14-A
BENTONITE
MOLE RATIO N a2 0 /Al 2 03 = 1 .
c
MONTMORILLONITE )
868
\
KAOL IN
- 70
(K
AOLINITE
877 -
/Al203 =
)
Na20,
1. 2.
50
2.2 1.6
Ca O/Si 02
1
I4-B
^
^
>
i \*
k
CAOL
N
k
vAOLI
N
i
K
DRY
OLIN
BRA^
ITE )
jru r
»A
(f
(AOL
p
NITE
RE C
CO ;
87I
, ILL
LAY
MO
TE)
878
MEX
-4
>
UNDERCLAY
SHALE
866
4
k
H
IGH
IRO
70
N
8
.8 2.0 2.2 I .8 2.0
CaO /Si 02
FIG. 14.— Parts A and B
Percent Al-Oa extracted versus mole ratio CaO/SiU^
Sinter temperature 1200°C. ; sinter time 10 minutes at top temperature.
62
ALUMINA EXTRACTION
10
^
i
"^Sintered t
from furn
between
riquet
ace at
fOO° a
i
s removed
temperatun
id 800°
15
i
Pow
er of
r
10 m
int
nute
srval
f~T
Po
wer
ncre
ased
af «"
a*
3UU
v^.
0 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500
TEMPERATURE-DEGREES C.
Fig. 15. — Typical record of a sinter heating schedule as traced from recorder chart.
LIME-SODA-SINTER PROCESS
63
H
<\J
o —
-J
o
1
O
2
DC
UJ
1 M
Eo°
oe
o
u.
a
UJ
8
•
1150
.LOWE
O
-l
<
1
(
i
« £ ~ w 2 o
u H ° o ° Z r>
tis
I
<
<
« 1 o o o 5 >
3 2 * - z - J
T
\
- \- 2
o O
® o
(0
t4
/"N
6V
y~> O
*
5 u - tVi
c l> II
Q fl O «,
r-
"J < rg O f
Q a. Zi <o
< £ > \ v
•n 3 O o <
°, « S * -1
a r- z o O
»
•
2
Z
«o ^
«0
<
<n ><
* 2
■a
O
N
ro a
p
tV^ <D
0>
7
i
o
CO
(0
<
a
UJ
a
Q
<
O
u.
o
o
/
f
UJ II
S »:
5 ; o t
S N B •
a. o x >
5 N O <
»
>
i-
Z O
o
GO
at
\
\
<fl <« -I
Z U U
\.
X
*\
f
\
•,
o °
tt
UJ
CL
o 2
o Ul
Z I-
ti c
X -
* Oq
Q310VaiX3 r0 ?IV %
CI3.LOVH1.X3 ro2iv o/q
64
ALUMINA EXTRACTION
1 00
60
50
40
*
A
•
20
m
IP
MgO ADDED AS M9 COj
SINTER TEMPERATURE 1150 °C
Na£ 0 /Al2 Od =1
Ca 0 /SI 02 * 2
CLAY bU
9
100 %M*°
%AI203
12
TEMPERATURE
16
I50°C
20
24
Fig. 20. — Percent A1203 extracted versus 100 c, ., ^
7r MgO .
2Oa
EFFECT OF PHOSPHORUS
(Figure 16)
The data indicate that the presence of
phosphate in quantities ranging up to one
percent of the weight of sintered material
exerts no important effect on the amount
of AbOs made extractable by lime-soda
sintering. This is an expected result but
is in contrast with results obtained when
the lime only is used as the alkali in the
sinter mix.
EFFECT OF T1TANJA
(Figure 17)
The results of these tests indicate that
titania in the sinter is not harmful and may
possibly have a small beneficial effect. There
is a suggestion that it may be desirable to
allow an extra mole of Na^CO for each
mole of TiOa present. The variations in
yield are, however, of the same order as the
probable precision of the data.
EFFECT OF IRON
(Figure 19)
Iron apparently has a moderate adverse
effect on the yield of extractable AkCX This
effect is not serious, however, unless three
percent or more FeaOs is present in the
sinter.
EFFECT OF MAGNESIA
(Figure 20)
The data indicate no effect of magnesia
on the percent of extractable alumina even
when the amount of magnesia present is as
much as 26 percent of the total alumina
present in the sinter. This is markedly
different from the results when the lime-
sinter method is used.
MINERALOGICAL ANALYSES
Microscopic Analysis
effect of variations in top sintering
temperature
Mixtures with a lime-to-silica ratio of
2.0 and a soda-to-alumina ratio of 1.0 were
fired for 10 minutes under oxidizing condi-
tions according to the schedule in figure 15.
Kaolinite clay. — New reaction products
were very indistinctly developed in sinters
fired to 1000° C, and it appeared that the
components had only started to react and
that new phases had only begun to separate
out (fig. 21C). Most of the material was
an aggregate mass of faintly anistropic
material with an index of refraction about
equal to 1.650. Some material, also slightly
anistropic, with an index of refraction of
about 1.590 was present also.
Sinters fired at 1 200 °C. were composed of
j8— C2S in grains with a maximum diameter
of 3 microns, and a lesser amount of a
slightly anistropic material with a mean
index of 1.590 that was interpreted as CsAa
with considerable soda in solid solution.
LIME-SODA-SINTER PROCESS
65
The sinters fired to 1300 C. (fig. 21 1) )
were similar to those fired to 1200°C. ex-
cept that the ft— C*S had a "cleaner" appear-
ance. A possible interpretation is that £— C»S
contains solid solution components at 1200°
C. which separate out when the temper-
ature is increased to 1300°C.
Halloysite clay. — Sinters prepared with
this type of clay had the same characteristics
as those made with kaolinite clay.
Diaspore clay. — Sinters fired at 1000°C.
were composed largely of NA in irregular
aggregate masses of individual units less
than about 1 micron in diameter (fig. 21 A).
A small amount of /3— OS in grains with a
maximum diameter of 3 microns was present
also (the sample contained a little kaolinite
and therefore some silica as an impurity).
Sinters fired to 1100°, 1200°, and 1300°
C. had the same characteristics as the
1000°C. sinter (fig. 21B).
Gibbsite-kaolinite {bauxite) clay — NA
(Na^O.AbOs) wTas present in rather in-
distinct extremely minute grains, and some
poorly developed /3— C2S was found also in
sinters fired to 1000°C.
Sinters fired to 1100°, 1200°, and 1300°
C. all showed well developed NA in aggre-
gates of units less than about 1 micron, and
ft— C2S in distinct grains with a maximum
diameter of 5 microns.
Illite clay. — A mixture of particles of
lime and dehydrated illite composed sinters
fired at 1000° and 1100°C. (fig. 22A).
In sinters fired to 1200°C. the compo-
nents were ft— C2S in units with a maximum
diameter of 15 microns, a pigmentary mate-
rial with the characteristics of -J-CaO.AbCX
FezOs, and a colorless faintly anistropic in-
terstitial material with a mean index of re-
fraction of 1.60. The latter component was
interpreted as CbAs with soda in solid solu-
tion (fig. 22B).
Sinters fired at 1300°C. were like those
fired at 1200°C. except that the /J-OS had
a "cleaner" appearance as if some solid
solution components had separated out (fig.
22C ).
Montmorillonite clay. — A mixture of
particles of lime and dehydrated montmo-
rillonite composed sinters fired at 1000° and
1100°C.
Sinters fired at 120" C. were composed
chiefly of j3— OS in units with a maximum
diameter of 15 microns. A small amount of
pigmentarj material with the characteristics
of 4CaO.AK'03Fe20< was present also. A
component approaching CbA« ma\ have been
present, but it could not be positively identi-
fied.
C^S, chiefly fi-ionw (there was some in-
version to y-form), in units with a maxi-
mum diameter of 15 microns made up about
all of the 1300°C. sinter. The ferrite ap-
peared to be less abundant, and the alumi-
nate was less certain than in the 1200°C
sinter.
Attention should probably be directed to
the fact that these and ensuing mineralogical
comments relate only to crystallized ob-
servable phases. In many cases it is obvious
that the lime content is insufficient to de-
velop both a silicate and an aluminate for
the entire silica and alumina contents with
even the most generous allowances for soda
solution.
EFFECT OF VARIATIONS IN THE RATIO OF
SODA TO ALUMINA
The soda-to-alumina ratio was varied
from 0.8 to 1.2, in sinters writh a lime-to-
silica ratio of 2.0 and fired to 1200°C. for
10 minutes under oxidizing conditions ac-
cording to the schedule in figure 15.
Kaolinite clay. — Variations in the soda-
to-alumina ratio from 0.8 to 1.2 caused no
essential variations in the character of the
sinters.
Diaspore clay. — Sinters prepared with a
soda-to-alumina ratio of 1.2 were composed
essentially of NA. They also contained a
considerable amount of green-brown pig-
mentary material that has many of the
characteristics of the ferrite, and it is sug-
gested that the excess soda is held in some
such compound. A small amount of (3— C2S
was present also.
An isotropic glassy material with an index
of refraction roughly equal to the mean
index of NA was the essential component
in sinters prepared with a soda-to-alumina
ratio of 0.8. A deficiency in soda appeared
to hinder the crystallization of NA.
66
ALUMINA EXTRACTION
...
w * I
n '%
5» %
# #•
:3- 4" *'*
^$ I y'' ^^ "#
#"
.,, *''fc'
<»
«?:■ A^"<i*fc
Fig. 21. — Photomicrographs of soda-lime-clay sinters, 300 X.
A. Diaspore-clay sinter fired to 1000°C. es in which the formation of new phases
showing aggregate masses of aluminate. is indistinct.
B. Diaspore-clay sinter fired to 1200°C. ^ ^ ,. . , . x , . ioaao/-.
showing aggregate masses of aluminate D" Kaolimte-clay sinter fired to 1300 C
somewhat better developed than in A. showing the distinct development of
C. Kaolinite-clay sinter fired to 1000° C. sma11 particles of C2S (mostly fl-form).
showing the development of some parti- The aluminate is interstitial and indis-
cles of new phases, and aggregate mass- tinct.
LIME-SODA-SINTER PROCESS
67
'j § * ,
Fig. 22. — Photomicrographs of lime-sod
A. Illite-clay sinter fired to 1000°C. show-
ing the absence of new phase develop-
ment. Aggregate masses of dehydrated p
clay and lime are the components.
B. Illite-clay sinter fired to 1200° C. show-
ing the development of small particles
of #CvS and pigmentary ferrite. Large
aggregate masses in which there is in-
a-clay sinters, 300 X.
distinct development of new phases are
also present.
Illite-clay sinter fired to 1300°C. show-
ing the excellent development of large
particles of /3C2S and pigmentary fer-
rite(dark material). An aluminate phase
is not evident in the photomicrograph.
68
ALUMINA EXTRACTION
Gibb site -kaolinite (bauxite) clay. — Sin-
ters prepared with a soda-to-alumina ratio
of 1.2 were composed of NA and ft— C2S.
In sinters prepared with a soda-to-alu-
mina ratio of 0.8, the component in addition
to f$— OS was an isotropic glassy material
with a mean index of refraction about equal
to that of NA. Again the crystallization of
NA appeared to be retarded by a deficiency
of soda.
lllite clay. — An increase in the soda-to-
alumina ratio from 1 to 1.2 caused the
aluminate to be more prominent and to
have properties more nearly approaching
NA. Further, the OS became "dirty" as if
it had taken something into solid solution.
In sinters prepared with soda equivalent
to 0.8, ft— C2S, dirty in appearance, seemed
to be about the only component. Aluminate
could not be identified positively.
Montmorillonite clay. — Increasing the
soda-to-alumina ratio from 1 to 1.2 caused
a distinct increase in the prominence of the
aluminate and gave it properties more
nearly approaching NA. Further, the C2S
in such sinters was "dirty" in appearance.
A decrease in the soda-to-alumina ratio
from 1 to 0.8 was attended by a decrease
in the abundance of aluminate and a trend
toward material with the characteristics of
CsAa. Further, the C2S was distinct and
"clean" in units to 10 microns in diameter.
EFFECT OF MISCELLANEOUS VARIATIONS
IN COMPOSITION
Varying amounts of Fe^Oo, MgCCX
TiC>2, and Ca3(PO02 were added to batches
prepared with kaolinite clay 877 with a
lime-to-silica ratio of 2, and soda-to-alumina
ratio equal to 1. The sinters were fired at
1150°C. for 10 minutes under oxidizing
conditions.
The sinter with the added FesOs con-
tained a considerable amount of minute dis-
crete particles of iron oxide. There was no
suggestion that the iron had reacted with
any of the other components.
No distinct phase containing TiO* could
be detected in the sinters to which this com-
ponent was added. However, the C*S and
aluminate phases were much less distinct,
suggesting that the TiO* had hindered their
development.
The sinter with added magnesia up to
26 percent of total alumina, and phosphate
up to 1 percent of total sintered material
were like those without these added com-
ponents.
EFFECT OF REDUCING ATMOSPHERE
Sinters with a soda-to-alumina ratio of
1 and a lime-to-silica ratio of 2 were fired
to 1200°C. under reducing conditions ac-
cording to the heating schedule given in
figure 15.
Kaolinite clays — Sinters fired under re-
ducing conditions were essentially the same
as those fired under oxidizing conditions
except that under the former conditions the
new phases were better developed. Further,
when any iron compound was present in
the clay as an impurity, the aluminate in
the sinter fired under reducing conditions
had a dirty yellowish appearance and a
higher index of refraction indicating that
iron was held in solid solution.
Gibbsite and diaspore clay. — If the clay
contained any iron component as an im-
purity, the aluminate in sinters prepared
under reducing conditions was dirty in
appearance and had a higher index of re-
fraction, indicating the presence of iron in
solid solution.
lllite clay. — The sinters fired under re-
ducing conditions exhibited an absence of
any material that appeared unreacted and
an enhanced development of new phases.
The (3— C2S developed in units with a maxi-
mum diameter of 20 microns as compared
to 10 microns in sinters fired under oxidiz-
ing conditions. Further the aluminate was
more distinct and had characteristics sug-
gesting that iron had been taken into solid
solution.
Montmorillonite clay. — Sinters prepared
under reducing conditions showed a larger
percentage of definitely reacted material
and better development of new phases. The
(3— C2S was present in cleaner and larger
units (20 microns ±), and the aluminate
appeared to have taken iron into solid solu-
tion.
l./MI -SOD.I-SINTER PROCESS
69
EFFECT OF VARIATION IN TIME BATCH HELD
AT TOP SINTERING TEMPERATURE
Sinters prepared with a soda-to-alumina
ratio of 1 and a lime-to-silica ratio of 2
were heated according to the regular firing
schedule shown in figure 15 to 1100°C.
One batch of sinters was held at 1100°C.
for 10 minutes, another batch for 60 min-
utes, and a final batch for two hours.
Kaolinite clay. — Only a slight difference
could be detected between the sinters fired
for 10 minutes and those fired for two
hours. The 10 minute sinter was composed
of a small amount of material that seemed
unreacted and material that appeared to
be in the initial stage of reaction. The two
hour sinters showed a more advanced de-
gree of reaction writh better development
of new phases.
Gibbsite and diaspore clays. — Sinters
fired at 10 minutes and at two hours both
showed complete reaction of components
with the development of new phases.
Illite clay.— The sinter held at 1100°C.
for two hours was composed of unreacted
components with a small amount of material
that might have been in the initial stages
of reaction. The 10 minute sinter was com-
posed almost entirely of unreacted material.
Montmorillonite clay. — Sinters fired at
10 minutes and at two hours were both
composed of essential}7 unreacted material.
EFFECT OF VARIATION IN HEATING RATE
The following data were determined on
sinters with a lime-to-silca ratio of 2 and
a soda-to-lime ratio of 1 fired under oxidiz-
ing conditions, with a heating schedule of
50° C. per hour over the range from 900°
to 1100°C. instead of the more rapid and
standard rate shown in figure 15.
Kaolinite clay. — The C2S and aluminate
appeared to be more distinct and in better
developed units in the sinters prepared at
the slower heating rate. The aluminate
showed more distinct anisotropism, and the
C2S was more definite and "cleaner", sug-
gesting less solid solution effects. Also there
seemed to be more inversion of C2S to the
y-form.
Gibbsite and diaspore clays. — The sinters
prepared at the slower heating rate were
usually about the same as those prepared
at the regular heating rate. In a few in-
stances there was a suggestion of better
phase development with the slower heating
rate. Also, any C*S present, because of the
presence of a silicate mineral in the clay,
showed more inversion to the y-form.
Illite and montmorillonite clays. — No
difference could be detected between sinters
prepared at the regular and at the slow
rate.
The more friable character of many of
the sinters (see page 57) prepared with the
slower heating rate was in accord with the
slightly better phase development and the
greater degree of inversion of OS to the
y-form.
EFFECT OF VARIATIONS IN THE RATIO OF
LIME TO SILICA
The lime-to-silica ratio was varied from
1.8 to 2.2 in sinters with a soda-to-alumina
ratio of 1, and fired to 1200°C. for 10
minutes under oxidizing conditions accord-
ing to the heating schedule given in fig. 23.
Kaolinite and illite clays. — The sinters
showed that the best phase development
took place when the lime-to-silica ratio was
1.8. In such sinters the /?— C2S was in distinct
units, and the aluminate, which approaches
OAs in optical properties, was plainly vis-
ible.
As the lime content increased, there was
a decrease in the degree of phase develop-
ment. For example, sinters with a lime-to-
silica ratio of 2.0 frequently contained a
considerable amount of material in aggre-
gate masses of indistinct units with about
the optical properties of fi— C2S but with few
distinct grains of /?— C2S. Also the aluminate
was less distinct and closer to C»Aa in op-
tical properties.
With an increase in the lime-to-silica
equal to 2.2 there was a further slight
decrease in the degree of development of
phases.
Montniorillonite clay. — In sinters pre-
pared with clays containing montmorillo-
nite, there appeared to be little or no dif-
70
ALUMINA EXTRACTION
ference in characteristics when the lime-to-
silica ratio was varied from 1.8 to 2.2.
X-ray Analysis
general comments
In the application of the X-ray diffraction
method to the lime-soda sinters, two un-
satisfactory situations must be admitted:
first, it is uncertain whether all of the
phases in the four— component system are
known, and second, the extent to which
various solid solution possibilities may be
realized cannot be established. Diffraction
patterns are frequently not precisely typi-
cal for a given phase, and should therefore
not be relied upon for estimates of abun-
dance of the phase.
THE PHASES
Calcium or tho silicate. — The lime-soda
sinters do not dust. In none of the sinters
examined did enough orthosilicate invert
to the y-form so that the y— C2S could be
found by diffraction methods. The develop-
ment of /3-C2S was essentially parallel to
the development outlined in the lime-sinter
discussion. In the lower temperature sinters
(1000°C.) a considerable amount of re-
action products was evident, the material
being of extremely small particle size, and
the diffraction pattern of that resulting
phase which we consider to be the silicate
departed from the typical /3— C2S to such a
degree that in ascribing the pattern to
fi— C2S we could be in error. For the kaolin-
ite sinters, development of this phase was
gradual, and by 1200°C. the pattern was
typical for (3— C2S whose particle size is
about of the order of a micron. In the illite
and montmorillonite clay sinters the same
material developed abruptly above about
1200°C. and by 1300°C. attained a par-
ticle size of 25 to 50 microns.
The large fi— C2S grains, which did not
invert, no doubt held some extraneous ions
in solution. The low temperature {3— C2S
could vary from ideal composition to a
much greater degree.
Sodium aluminate. — The sodium alumi-
nate, NaAlO, is a well-established phase
for which the powder diffraction diagram
is readily recognized but the crystallization
has not been worked out. The phase was
clearly developed and typical only in the
sinters prepared with the highest alumina
clays. In other cases, where the phase was
observed at all, some lines had anomalous
weak intensities or were missing, and the
apparent abundance was not commensurate
with the alumina content. It is our un-
supported opinion that the typical NA is
a variation of the cristobalite structure, and
that the non-typical material is a combina-
tion of this with y-AbOs. In some cases
the development of NA was promoted by
increased addition of Na2COs.
The complex aluminate. — A prominent
feature in the diffraction diagrams, especial-
ly of the kaolinite sinters at 1000° and
1100°C, was a pattern rather closely re-
sembling that of CbAs. At higher temper-
atures, after the full development of 0-CaS,
this material was no longer apparent. In
the CaO-Al203-Na20 system, according
to Brownmiller and Bogue,8 OA3 does not
dissolve soda. However, the published CsAa
structure is of an open type, described as
based on garnet, and it seems quite possible
that soda and silica together could enter
this crystallization. This material is assumed
to be the nearly isotropic phase observed
optically with the index 1.590.
The prominence of the above described
pattern declined in sinters fired at 1200°C.
and was not apparent at 1300°C. although
it was still observed microscopically. One is
led to the conclusion that the phase softened
to a glass.
THE SINTERING PROCESS
In the lime-soda sinters, reaction is initi-
ated at somewhat lower temperatures than
in the lime sinters. The softening of the
sodium-calcium double carbonate (appar-
ently represented in the thermal curves by
a flexion near the shoulder of the decarbon-
ization peak) provides a liquid agent for
attack on the clay. High alumina materials,
even the highly resistant a—AhOs, react
readily with the lime and soda, kaolinite
LIME-SODA-SINTER PROCESS
71
reacts less readily, and illite and mont-
morillonite react relatively little.
It was not possible to identify the first
reaction product in sinters made with high
aluminous clays, but at 1100°C., NA had
developed. On further heating, the NA
grew in grain size and had a typical diffrac-
tion pattern. Only a little /?— C2S devel-
oped, as would be expected from the low
silica content of these clays.
Kaolinite clay sinters have reacted at
1000°C. to form an extremely fine-grained
mixture of the complex aluminate, "CsAa,"
and non-typical /?— C2S. Continued heating
gradually developed more ft— C2S, which by
the time the temperature reaches 1200° C.
gave a typical diffraction pattern, while the
crystalline structure of the complex alumi-
nate was destroyed. The transition from the
complex aluminate to the noncrystalline
material was considered to be reflected in
the greater extractability of aluminum in
the higher temperature (or longer heated)
sinters. The development of NA was either
doubtful or unimportant in the sinters of
kaolinite clays. When small amounts were
indicated they appeared as well or better
developed in the low temperature sinters as
in the high.
In illite and montmorillonite clay sinters
the complex aluminate, "GAs," was less
prominent. The (3-C& development, as in
the case of the lime sinters, showed a
sudden marked growth to large particle
size when temperatures of 1200° or 1300°
C. were reached. This growth was probably
not significant in itself from the stand-
point of extraction, but apparently reflected
the greater difficulty of attack on these
more stable clays.
DIFFERENTIAL THERMAL
ANALYSES
Differential thermal analyses of batches
of various types of clay with a lime-to-silica
ratio equal to 2 and a soda-to-alumina ratio
equal to 1 are given in figure 23.
The differential thermal curves show the
relative intensity and the temperature of
the reactions that take place when sinter
mixtures are heated up to 1300°C. at a
uniform rate of approximate^ 10 C. per
minute. The downward deflections of the
curve indicate endothermic reactions and
the upward deflections indicate exothermic
reactions. A vertical scale for determining
the temperature difference indicated by the
deflections of the curve is given in figure 9.
Diaspore clay 872, sinter mixture SS89.
— The endothermic reaction between 500°
C. and 600° C. corresponds to the loss of
lattice water (OH) from the diaspore. The
part of the curve between 700° C. and
875 °C shows an endothermic reaction due
to the loss of CO2 from the carbonates plus
an exothermic reaction which indicates the
formation of a new phase. The curve char-
acteristics that result when loss of CO* is
the only reaction taking place in this tem-
perature interval can be seen in curves for
SS85 and SS87.
The curve for sinter mixture SS89 shows
that a reaction is taking place between the
components of the diaspore clay and the
carbonates, and that a new phase or phases
develop at about 800° C. The formation of
the new phase begins before the reaction
corresponding to the loss of CO* is com-
plete.
The portion of the curve above 900° C.
cannot be interpreted in detail, but it prob-
ably signifies changes in the form of the in-
itial phases and/or the separation out of
material held in solid solution.
Gibbsite-kaolinite {bauxite) clay 875,
sinter mixture SS90. — The initial endother-
mic peak is due to loss of adsorbed water
and suggests that halloysite as well as kao-
linite is present in this clay. The endother-
mic peaks at about 340° C. and 590 °C. are
the result of loss of lattice water (OH)
from gibbsite and kaolinite (and halloysite),
respectively.
The endothermic peak between about
790°C. and 840° C. is the result of loss
of CO* from the carbonates. However, if
loss of CO* was the only reaction in this
temperature interval, the curve should be
like those of SS85 and SS87 in the same
temperature interval. It can be concluded,
therefore, that an exothermic reaction due
to the formation of a new phase or phases
72
ALUMINA EXTRACTION
C 1300" C
Fig. 23. — Differential thermal analyses of lime-soda-clay mixtures. The curves
posits from data for the clay alone and for sinter mixtures. (See fig. 9.)
LIME-SODA-SINTER PROCESS
73
begins soon after the start of loss of CO*,
i.e.. about 800°C.
The curve above 950° C. is probably the
result of changes in the initial phases and/
or the separation out of material held in
solid solution. It would seem that such
changes take place without appreciable
thermal effect. There appears to be no
distinct thermal effect corresponding to the
formation of the disilicate or the aluminate.
Kaolinite clay 877, sinter mixture SS67.
— The endothermic reaction at about 600°
C. corresponds to the loss of lattice water
(OH) from the kaolinite. Again the endo-
thermic reaction between about 750° C. and
875 °C. is not as large as would be expected
if loss of CO2 were the only reaction during
this temperature interval. It is probable
that C2S and perhaps other phases begin to
form before the destruction of the carbon-
ate is complete. The new phases seem to
develop at slightly higher temperatures in
sinter mixes containing kaolinite clays than
in those prepared with diaspore or gibbsite
since the reaction due to loss of CO2 is less
affected in such sinter mixes.
The curve above 950 °C. is like that for
batch SS90, and the same explanation is
offered.
Kaolinite-illite {ball) clay 879, sinter
mixture SS92. — In this curve the initial
endothermic peak corresponds to the loss
of adsorbed water by illite, the exothermic
reaction between about 200° C. and 500° C.
is the result of the burning off of organic
material, and the endothermic peak at about
600 °C. is caused by the loss of lattice water
(OH) from the kaolinite and illite.
The portion of the curve between 750°
and 1000°C. is about like that for the
preceding sinter batch SS67 and the same
ueactions are indicated. The curve above
1000°C. has some slight differences which
are like those in the curve for the batch
containing illite (SS87), and show the in-
fluence of the small amount of illite in this
clay.
Illite clay 870, sinter mixture SS87. —
The broad initial endothermic reaction is
due to loss of adsorbed water, and the endo-
thermic peak at about 575° C. corresponds
to loss of lattice water (OH) from the
illite.
The endothermic reaction between about
700°C. and 925° C. appears to be the result
of a single reaction. Unlike the curves for
the previous mixture, the removal of CO*
from the carbonate seems to be the only
reaction taking place in this temperature
interval.
The sharp exothermic reaction at about
1200°C. represents the formation of C»S
and perhaps other new phases. The temper-
ature at which new phases form is several
hundred degrees higher in sinter mixtures
containing illite clays than in those made up
of kaolinite, gibbsite, or diaspore clays. This
is in accord with the extraction data which
show very little alumina is extractable from
illite clay sinters until firing temperatures
reach 1200°C, whereas a high percentage
of alumina is extractable from sinters con-
taining kaolinite, gibbsite, or diaspore when
the firing has been carried to only 1000°C.
Montmorillonite clay 868, sinter mixture
SS85. — The endothermic peaks at about
150°C. and 700°C. are due to loss of ad-
sorbed water and lattice water (OH),
respectively, from the montmorillonite. The
portion of the curve above 750° C. is simi-
lar to that for the preceding mixture con-
taining illite clay and the same reactions
are indicated.
Like illite mixtures, batches containing
montmorillonite do not yield new phases
until a temperature of about 1200° C. is
reached, and appreciable alumina is not ex-
tractable until sinters have been fired to
this temperature.
DrscussiON and Summary of Study of
Phases Present in the Sinters
The X-ray and optical data concurred
in the conclusion that in sinters with a lime-
to-silica ratio of 2 and a soda-to-alumina
ratio of 1, the compound Na*O.Al-Os as
described by Brownmiller and Bogue8 is
well developed only in sinters prepared with
high alumina clays (containing diaspore or
gibbsite). In kaolinite clays the aluminate
appeared to be 5Ca0.3AU0: with soda and
perhaps silica in solid solution. In illite and
74
ALUMINA EXTRACTION
montmorillonite clays the development of
aluminate was very poor, but the compound
again seemed to be more nearly like C5A3.
Soda in excess of that required for a
ratio of soda to alumina equal to one caused
an enhanced development of aluminate with
characteristics more like NA. A decrease in
the soda content caused a decrease in the
development of aluminate and a trend
toward characteristics like those of C5A3.
The C2S occurred in distinct units only
a few microns in diameter in sinters made
with kaolinite clay. In sinters made with
illite or montmorillonite clays, the C2S
attained a maximum diameter of over 20
microns. The C2S was in the /?-form with
little inversion to the y-form.
The aluminate was found in irregular
aggregates and interstitial masses composed
of indistinct individual units less than one
micron in diameter.
In sinters prepared with gibbsite or dia-
spore clays new phases began to develop as
soon as loss of CO2 released CaO and Na20,
or perhaps sooner, under the attack of the
fluid double carbonate, hence a large per-
centage of alumina was extractable in sinters
made at low temperatures, e.g., 1000° C.
Again in sinters made with kaolinite clays,
new phases began to form before the car-
bonates were completely broken down al-
though it would seem that the temperature
was slightly higher than for the gibbsite or
diaspore clays. As a consequence sinters
containing kaolinite clay and fired at low
temperatures had high percentages of ex-
tractable alumina. In the kaolinite and
gibbsite clays the phases appeared to con-
tinue to develop throughout a temperature
interval of several hundred degrees, and
without sharp thermal effects.
When sinters were prepared containing
illite or montmorillonite clays, new phases
did not develop extensively until about
1200°C. and consequently little alumina
was extractable unless the firing was carried
to this temperature. Further there was a
sharp thermal reaction which began at about
1175° corresponding to the formation of
the new phases in illite and montmorillonite
clay mixtures.
A plausible explanation for the difference
in temperatures required for new phase
development is as follows: In the case of
gibbsite and diaspore the loss of (OH)
water at about 325° C, and 525° C, respec-
tively, results in the formation of free
alumina which appears to be particularly
susceptible to attack by fluid double carbon-
ates. As a consequence new phases form at
a very low temperature. The loss of (OH)
lattice water from kaolinite at about 575°
C. produces the active, but still crudely com-
bined alumina and silica of metakaolin, and
new phases are formed slightly later, after
considerable development of free lime.
In the case of montmorillonite and illite,
the loss of (OH) water at 575° and 675°
C, respectively, is not accompanied by a
destruction of the lattice of these minerals.
A definite structural configuration is re-
tained until a temperature is reached several
hundred degrees in excess of that required
for loss of (OH) water. It would be ex-
pected that alumina and silica locked up in
a definite structural configuration would be
relatively unreactable, and that new phases
would not develop until such a configuration
is destroyed. New phases, then, would not
be expected in sinters prepared with illite
or montmorillonite clays until a higher
temperature is reached (or until after a
more protracted heating interval) than is
required for similar development in batch-
es containing the other types of clay mate-
rials. It follows, of course, that in plant
practice the presence of any illite or mont-
morillonite in a clay would raise the sinter-
ing temperature necessary for the best ex-
traction.
All the data suggest that the new phases
formed at low temperatures in the kaolin-
ite, diaspore and gibbsite batches undergo
changes as the temperature is raised. The
)8— C2S can be seen under the microscope to
become "cleaner" in appearance in sinters
fired to higher temperatures, as if the ft— C2S
were being freed of material in solid solu-
tion. The diffraction pattern of ft— C2S is
typical only in higher temperature sinters.
In lower temperature sinters the pattern
departs considerably from that typical for
LIME-SOD A-S1NTER PROCESS
75
this compound. The prominence of the pat-
tern of the complex aluminate is reduced
as the temperature of sintering is increased,
suggesting; that the material is reduced to
a kind of glassy suhstance.
SUMMARY AND CONCLUSIONS
The introduction of soda into the alkali
extraction processes as practiced in the lime-
soda method resulted in attack of the clay
at temperatures much lower than when lime
alone wras used. When the clays were of
the high alumina variety (diaspores and
bauxites) fairly good extractions were ob-
tained with sintering temperatures of 1000°
C. Kaolinite-clay sinters required slightly
higher temperatures, whereas sinters pre-
pared from illite or montmorillonite clays
had to be heated to temperatures in the
range 1200° to 1300° C. to give satisfactory
yields. Unless raw materials containing
sufficient soda to carry the process are avail-
able there is little advantage in using the
lime-soda as compared to the lime process
with the two latter types insofar as sinter-
ing temperature is concerned.
Kaolinite clays and gibbsite-kaolinite
(bauxitic) clays may yield above 90 percent
of their alumina by the lime-soda process
when they are relatively free from iron and
other clay minerals. Halloysite clay, mont-
morillonite clay, and anorthosite may also
yield over 90 percent of their alumina, but
it is more difficult to get top yields. The
highest yields for illite clays were 10 to 12
percent below those for the other clays.
Yields for clays composed of mixtures of
clay minerals were in agreement with
results predicted from pure types.
Diaspore clays and gibbsite-kaolinite
clays yielded nearly as much alumina on
sintering to 1000°C. as to 1300°C. Pure
kaolinite clays showed no increase in yield
when sintered above 1100°C. other factors
being the same, and yielded about 75 percent
of their alumina when sintered to 1000°C.
Montmorillonite clavs vielded none of their
alumina on sintering to 1100°C, 50 per-
cent on sintering to 1200°C, and 90 per-
cent on sintering to 13()0°C. Like mont-
morillonite clay, illite clay required sinter-
ing above 1200°C. for good alumina yields.
In sinters fired to 1100°C, increasing the
time held at top temperature caused onl>
slight increase in yield of alumina from
diaspore clays, only moderate increases foi
clays composed of kaolinite, gibbsite, and
halloysite, and very great improvement in
the yield from illite clays. Regardless of
sintering time, montmorillonite clays yield-
ed no alumina on firing to 1100°C.
Sinters of the various clays fired under
strong reducing conditions gave the same
yield as similar sinters fired under oxidizing
conditions.
Although some of the results are erratic,
the data show definite improvement in the
yield of alumina from all types of clay
except the diaspore as the Na20/Al20s
ratio increased from 0.8 to 1.2.
Sinters of kaolinite clay and illite clay
gave sharply reduced yield when the CaO/
SiOa ratio was less than two, but not im-
proved yields with a ratio above two. Mont-
morillonite clay showed greatly increased
alumina yields with CaO/SiO^ ratio great-
er than two.
Extracts from illite clays and mont-
morillonite clay contained a slightly larger
percent of silica than the extracts from other
clays. Kaolinite clays which contained ap-
preciable amounts of illite and montmo-
rillonite gave extracts with lower silica con-
tents than pure kaolinite clays, but the
difference was small.
Addition of phosphate (Ca:>(PO02) or
magnesia to a kaolinite clay batch does not
affect the amount of extractable alumina.
Added TiOa also was not harmful, but the
data suggest that an extra mole of Na^O
ought to be allowed for each mole of TiOz
present. Added Fe*Os had a moderate ad-
verse effect that was serious when the
amount was more than three percent.
76
ALUMINA EXTRACTION
CoO.S.02
1544
3Co02S.02
2CoO.Si02 f0
2'30
3CoO.S.02
3AltO,2SiOt
Fig. 24. — Equilibrium diagram of the system CaO — A1203 — SiCK.
Diagram of Rankin and Wright revised by Schairer (see reference 9) and reproduced with his
permission. Shaded area includes compositions of all lime sinter batches discussed in this
report. Numbers represent temperatures in degrees C.
REFERENCES 11
REFERENCES
1. Grim, R. E., Modern concepts of clay mate-
rials. Jour. Geol. 50, 225-275 (1942) ;
Rept. of Inv. 80, 111. Geol. Survey (1942).
2. Walthall, J. H., Personal communication.
3. Copson, R. L., Walthall, J. H., and Hignett,
T. P., Final report on the extraction of
alumina from clay by the lime-sinter
modification of the Pedersen process.
Serial W-103 War Metallurgy Committee
of National Academy of Sciences (1943).
4. Bates, P. H., and Klein, A. A., Properties of
calcium silicates and calcium aluminates
occurring in normal Portland cement.
Tech. Paper 78, U. S. Bur. of Standards
(1917).
5. Barrett, R. L., and McCaughey, W. J., The
system CaO-Si02-P205. Am. Min. 27,
680-695 (1942).
6. Zerfoss, S., and Davis, H. M., Observations
on solid-phase inversions of calcium or-
thosilicate, constituent of dolomite-silica
brick. Jour. Am. Cer. Soc. 26, 302-307
(1943).
7. Grim, R. E., and Rowland, R. A., Differential
thermal analyses of clay minerals and
other hydrous materials. Amer. Mineralo-
gist 27, 746-761, 801-818 (1942).
8. Brownmiller, L. T., and Bogue, R. H., The
system CaO— Na20— Al2Os. Am. Jour. Sci.
XXIII, 501-524 (1932).
9. Schairer, J. F., The system CaO- FeC—Al20:!
— Si02: 1. Results of quenching experi-
ments on five joins. Jour. Am. Cer. Soc.
25, 241-274 (1942).
Illinois State Geological Survey
Bulletin No. 69
1945