COMPARISON OF DRYING CORN USING SODIUM AND CALCIUM

BENTONITE

K.C. Watts', W.K. Bilanski2, and D.R. Menzies3

'Agricultural Engineering Department, Technical University ofNova Scotia, Halifax, N.S. B3J2X4;
2School ofEngineering, University ofGuelph, Guelph, Ontario NIG 2W1; and JResearch Station, '
Agriculture Canada, Vineland, Ontario LOR 2E0
Received 31 December 1984, accepted 27 August 1985

Watts, K.C, W.K. Bilanski, and D.R. Menzies. 1986. Comparison of drying corn using sodium and calcium
tentonite. Can. Agric. Eng. 28: 35-41.

Bentonite is an abundant clay mineral that is suitable for adsorption drying ofvarious grains. However, bentonite isnot
uniform in composition, being predominantly composed of montmorillonite along with avariety of other minerals. Also
different bentonites have different naturally occurring adsorbed ions which markedly affect the water adsorption
characteristics. Corn drying experiments using two types of commercially available bentonite clay, calcium- and sodium-
based, are reported. Moisture isotherms for the two clays are deduced from the experimental results and compared to other
isotherm data. The experimental results showing the time history of the drying of the corn and wetting of the clay are
predicted well using existing relationships. The more common calcium-based clay absorbs more moisture than sodium-
based clay from corn at low-moisture contents in which the clay remains free flowing. Therefore, calcium bentonite is the
preferred clay for drying corn on the basis of moisture capacity.

INTRODUCTION tion results from different workers using Using corn with one type of sodium ben
Desiccant drying of grain has been different clays are complicated by several tonite and one type of calcium bentonite
studied by several workers in the past due factors (Zettlemoyer et al. 1955). Ben this study will determine which type of
to its many advantages. Danziger el al. tonite clays have exchangeable ion sites bentonite is better suited to grain drying.
(1972) and Tomlinson and Miller (1981) and the number of the sites is not constant The purposes of this paper are to exam
have used expensive buteffective silicagel per unit mass of clay. The type and number ine the theoretical base for moisture
to dry agricultural materials. Bern et al. ofadsorbed ions alterthemoisture adsorp adsorption by the two types of bentonite
(1981) used oven-dried corn as the dessi- tion and swelling properties of the clay. clay; to report some experimental results
cant. Sturton et al. (1981) and Graham et Different polar molecules in a swelling of drying corn using both sodium- and
al. (1983) have investigatedthe use of ben clay determine the amount of internal areas calcium-based clays; to demonstrate that it
tonites in the drying of grain. exposed. Also, there are variable amounts is possible to predict the experimental
In all desiccant drying using clay, the of impurities in each sample of clay. For results using existingrelationships for corn
wet grain and dry clay are intimately low clay moisture contents, calcium clay drying; and to draw conclusions as to the
mixed in a sealed container at ambient adsorbs more moisture and accomplishes best clay to use for drying grains.
temperature. The grain rapidlygives up its it faster than a sodium bentonite (Fig. 1).
moisture to the surrounding clay with little
or no temperature rise, as the heat of
hydration is negligible. This process dif SODIUM MONTMORILLONITE
fers from batch-in-bin drying which
causes overheating of some grain as the 6
dryingfront movesupwardfrom the hot air a
inlet. When using desiccants, the drying 5
S3 >-
occurs uniformly throughout the bin; no E <
o _J
regions in the grain remain wet for U) U
extended periods. The clay can be dried a 4
< >•
usinglow-costsolar energy during periods ac
of sunshine in the spring and summer, uj a
a:
thereby storing the energy in a readily 3
P u
accessible form for use when needed. Ben tz v.
in L3
tonites have one advantage not shared by
o c
the other desiccants: they have been shown
by workers at the University of Guelph to
reduce insect infestations in grain with
which they are combined, probably
because of the abrasive action of the clay
on the insect as it moves through the mix
ture. 10 100 1000
Bentonite clays are noted for their
hygroscopic properties. However, the TIME (MINUTES)
comparison and interpretation of adsorp Figure 1. Temporal moisture adsorption curvesfor calcium and sodium bentonite.

CANADIAN AGRICULTURALENGINEERING, VOL. 28, NO. 1, WINTER 1986 35

 MOISTURE ADSORPTION BY 300
BENTONITES
Bentonite clays are thought to be of
volcanic origin (Grim 1968), although a
at c
NUMBERS REFER fi/
clays having a similar structure and proper TO REFERENCES /
ties have been found which were of a a: <
hydrothermal source. There is, therefore, § u 200
a very large variation in composition. The
basic composition of bentonite is about UJ •
J'*
55% silicon dioxide and 20% aluminum
trioxide with smaller varying amounts of % u
p. N.
iron, magnesium, calcium, potassium, tn g
and sodium oxides. The dominant clay S * 100 -
mineral material is smectite with varying
amounts (sometimes up to 50%) of illite
and kaolinite and sometimes crystobalite.
Rarely is there less than 10% of non-clay
minerals.
The hygroscopic characteristics of ben
tonites are due largely to the smectite com .2 .4 .6 .8 1.0
ponent. However, the composition of
smectite varies greatly in different ben RELATIVE HUMIDITY P/Po
tonites, both in respect to the smectite lat Figure 2. Variation of reported values of moisture adsorption isotherms for sodium bentonite.
tice and the nature of the exchangeable
cations. The most abundant exchangeable
cation is Ca++ (Grim 1968), e.g. the Mis
sissippi bentonite. Although few carry NUMBERS REFER TO REFERENCES
Na+ as the dominant ion, the Wyoming
bentonite is a well-known example. These
250
two bentonites vary both in the structure of
the smectite lattice and adsorbed ions;
hence their properties and end uses are
considerably different.
Some workers (Keren and Shainberg 8^ 2D0 i 0 ^ 37.5 C
1979) have assumed that it is possible to
change a sodium bentonite to a calcium ° _j
in o
bentonite and vice versa by saturating the 2 150
clay with the appropriate ion. Grim (1968)
indicates that such treatment does not UJ r-%
DC u
bring a complete change in properties P O
which suggests that the smectite lattice &£ 100
structure must be different.
The hydration of bentonite is a complex
55
process. It appears that at very low
50
moisture contents the water is adsorbed on
the edge of the clay platelets and slowly
moves between the platelets where the
hydration of the interlayer cations takes
place. The calcium ions tend to hydrate 0 .2 .4 .6 .8 1.0
more readily forming a skeletal double
water layer corresponding to the RELATIVE HUMIDITY P/Po
octahedral coordination of hydration water
Figure 3. Variation of reported values of moisture adsorption isotherms for calcium bentonite.
around the calcium ion. Two complete Number refer to references. Note the large temperature effect seen in reference 10.
layers form for calcium clay for the rela
tive humidity (RH) range of 30-80%. For
sodium ions a first monolayer of water is
formed in the RH range of 40-70%. The TABLE I. CONSTANTS FOR POLYNOMIAL HYGROSCOPIC CURVES FOR DATA OF
KEREN AND SHAINBERG (1979)
second water layer is complete at a RH of
90%. This process was observed in X-ray EFES A B C D

studies (Zettlemoyer et al. 1955) but is not 0 -0.41303 38.3196 -260.8395 689.4663
0.1 -0.30857 31.1142 - 123.4283 231.4286
always evident because adsorption of 27.6160 -634.817 241.0623
0.2 0.049111
water on the outer surfaces and edges of the 0.4 -0.044101 55.4839 -292.0185 515.8957
platelets mask water uptake between the 0.6 2.23965 40.9714 -295.3833 617.1455
0.1 11.12086 -83.7160 + 351.4836 -564.1115
layers. It is because of the migration of

36 CANADIAN AGRICULTURAL ENGINEERING, VOL. 28, NO. 1, WINTER 1986

 300

EQUIVALENT FRACTION OF
EXCHANGEABLE SOOIUM
o

m P
£ < 200
to D
Q
< >-
ui a

P **
to o
100
35

Sfc^ x 4- X
0 .2 .4 .6 .8 1.0

RELATIVE HUMIOITY P/Po

Figure 4. Moisture adsorption isotherms for monoionic clays (Keren 1979).

*<?
~ 220 -
>-
<

u ^
>- ^
a 180
o

19
/
X

a J40

a:
o
in

a: FROM REFS. 5 & 11

I-
U)
•—•

60 "

.4 .6 .8 1.0

RELATIVE HUMIDITY P/Po

Figure 5. Moisture adsorption data for clays used in this study compared with several other
workers.

CANADIAN AGRICULTURAL ENGINEERING, VOL. 28, NO. 1, WINTER 1986 37

 DRYING CORN WITH SODIUM BENTONITE

• - EXPERIMENTAL GRAIN ADSORPTION DATA

CD
•3

UJ

8 .*
UJ
cc

\THEORETICAL PREDICTIONS
o

.1

x - EXPERIMENTAL CLAY AOSORPTION OATA

-L -L -L

10 20 30 40 50
TIME (HOURS)
Figure 6. Corndrying withsodium bentonite. Lines are theoretical predictions. •refers to the corn moisture content, x refersto the clay moisture
content.

water between the platelets that equi that the drying process may be analyzed as
librium is not instantaneous and that the a three-stage process with time as follows:
MR =4 [exp {-Kt) +i exp (-4Kt)] (1)
properties of the clay can change over the (1) moisture movement out of the grain where
period of an hour or so. At higher relative into the air which is assumed to surround
humidities, moisture condenses on the the grain; MR = (M - Meq)/(A/0 - Meq) (2)
external surfaces of the Na+ clay and (2) moisture adsorptionby the clay from 5023 \
osmotic swelling occurs. the air; and tfcom = 0.54 exp \--r (3)
The desorption isotherms for ben (3) a change in relative humidity of the
tonites are quite reproducible while the air due to the resultant moisture content of M = average decimal moisture content
adsorption isotherms are reproducible the clay and the grain. (db)
only after the first adsorption cycle and The equilibrium relative humidities of / = drying time (s)
then only with difficulty (Grim 1968). This the air corresponding to the initial M0 = original moisture content
implies an apparent change in clay struc moisture contents of the grain and clay ^eq = equilibrium moisture content
ture which is not returned to the original differ widely; therefore, there is a rapid 6abs = absolute temperature (°R) for
condition upon desorbtion. One can con initial transfer of moisture from the grain the constants specified
clude from the foregoing that there will be to the air and into the clay. As the clay The equilibrium moisture content of
a large variation in experimental data and adsorbs more moisture, the equilibrium grain M must be determined in order to
difficulty in developing theoretical predic relative humidities converge with time. use the above equation. However, A/eq is
tions. Since the relative humidity of the air itself a function of the relativehumidityof
surrounding the grain changes with time, the air surrounding the grain which varies
SUMMARY OF THE it is not possible to directly apply con during the drying process. Hence Meq for
MATHEMATICAL MODEL ventional drying equations which assume the corn must be continually evaluated dur
The mathematical model for sodium a drying medium with a constant relative ing the process using the Chung equation
bentonite is being developed and is humidity on entry. Hence, the process (American Association of Agricultural
reported here in summary form and must be analyzed numerically with time. Engineers 1982-1983).
expanded to include calcium bentonite. The grain-moisture relationship spec
It is assumed throughout this paper that ified for single kernel analysis of corn was Meq = 0.33872 0.05897
the clay and grain are intimately mixed and (Brooker et al. 1974) ln[- (T+ 30.205) In (RH)] (4)

38 CANADIAN AGRICULTURAL ENGINEERING, VOL. 28, NO. 1, WINTER 1986

 DRYING CORN WITH CALCIUM BENTONITE

• - EXPERIMENTAL GRAIN ADSORPTION DATA

m .3

UJ

S .2 """*•• * •
LU H
x^—
cn ^THEORETICAL PREDICTIONS
i—«

o
.1

x - EXPERIMENTAL CLAY AOSORPTION DATA

\ 1 i i i i i i i

10 20 30 40 50

TIME (HOURS)
Figure 7. Corndryingwithcalcium bentonite. Linesare theoretical predictions. •refers to the corn moisture content, x refers to the clay moisture
content.

where preferably be mathematically continuous RH = AX + BX2 + CX3 + DX4 (5)

M = equilibrium moisture, decimal over the range of relative humidities of where
(db) interest. The work of Keren and Shainberg X = the amount of moisture adsorbed
T = temperature (°C) (1979) was most suitable (Fig. 4). Their at pressure (P0)
RH = relative humidity, decimal data for low-moisture contents followed RH = relative humidity (P/PQ)
The moisture adsorption properties of the BET theory which implies wetting of P0 = saturation vapor pressure
claysare complex functions of a number of the first monolayer only. They also present No constant is included in the equation
variables as previously noted. This results data for the same bentonite clay saturated because a clay of zero moisture content can
in a wide scatter of results reported in the with varying amounts of calcium and only exist in an environment of zero rela
literature for predominantly sodium clays sodium ions. Because Keren and Shain tive humidity.
(Fig. 2) and predominantly calcium clays berg (1979) started with the same clay The values of the constants in Eq. 5 are
(Fig. 3). It is important to note the large (Wyoming bentonite) as a base material given in Table I as a function of the equiv
variationin hygroscopic properties of ben and changed the exchangeable ions, the alent fraction of exchangeable sodium
tonites since it dictates that the portion of lattice structure differences between natu (EFES), ie. EFES = 0 means a clay which
interest of the moisture isotherm of each rally occurring calcium and sodium ben is totally saturated with calcium ions.
clay must be determined if accurate esti tonites are not incorporated. However, the
mates of the drying capability of each clay continuous nature of the data is useful in
are to be made. With such large variations obtaining mathematical expressions for EXPERIMENTAL STUDY
in clay properties, prediction of grain dry the hygroscopicity curves. It was initially Commercially available sodium-based
ing using clay is somewhat involved. To thought that the BET equation could be bentonite (Black Hills, Wyoming) and cal
simplify the problem and to obtain func used to represent these curves, but above a cium-based bentonite (Dixie Bond, Mis
tional relationships for moisture isotherms relative humidity of 40% it is no longer sissippi) were used to compare the drying
for bentonites, one worker's results were valid. It was not possible to append another properties of clays with different ex
chosen as a standard and all clays were curve to the BET equation for higher rela changeable ions.
related to that standard. The standard tive humidities. Therefore continuous Nine experimental trials were con
hygroscopicity curves needed must be a polynomial curves were derived for all of ducted using intimate mixtures at 1:1 ratio
reasonable representation of most clays, Keren and Shainberg's data having the by weight (wet basis) of corn and clay. A
have a wide range of types of clay, and form 1:1 ratio was used since no extra storage

CANADIAN AGRICULTURAL ENGINEERING, VOL. 28, NO. 1, WINTER 1986 39

 (a)

8
5 3 #
d <sPCF
^
8
2 2
3
OS

<
ck

to
$^
< r J—~-
.1 .2 .3 .4

CORN MOISTURE CONTENT (OB)

4 -
ac

8
>-
<
3 -
U

8
on
S
3
2 -
8
Of

1 -
<
a:

ui
en
<

.1 .2 .3 .4
CORN MOISTURE CONTENT (DB)
Figure 8. Mass ratio of clay to corn for four different clays for two different desired final moisture contents, (a) 0.13 db (b) 0.17 db
volumeis required. Five experiments used remaining contents of the containers were relativehumidityof the air surrounding the
calcium clay and four used sodium clay. remixed and the containers were resealed grain and the clay should be identical. If
Rewetted corn was cleaned and its immediately after removing each sample. one also assumes that the hygroscopicity
moisture content was adjusted to various No corn fines werefound in the clay sam characteristics of the corn are well estab
values between 16.6 and 40% wb and ples and no clay was found on the corn lished and that the equilibrium moisture
allowed to equilibrate for a minimum of 3 kernals except when the initial corn content of the corn determines the relative
days. Three kilograms each of corn and moisture content was above 30% wb. humidity of the voids, it is possible to
clay were mixed together in air-tight con obtain the hygroscopicity characteristics
tainers at time zero. All tests were run at RESULTS (moisture contentvs. relative humidity) of
25°C. At each sample time, as notedon the The experimental data were validated by the clay.
graphs, three 0.2-kg subsurface samples confirming that all the moisture removed It was found that the commercially
were removed from the mixture in each from the grain entered the clay, i.e. the available calcium-based clay has better
container and the clay was separated from total moisture in a given bin must be con hygroscopicity characteristics (Fig. 5)
the corn on a no. 4 sieve. All corn samples stant for all times. than the commercially available sodium-
were dried at 105°C for a minimum of 48 h To estimate the hygroscopicity of the based clay. It was also apparent that dif
to obtain the final moisture content. The clays, one mustassume thateventually the ferent batches of the same clay (Sturton et
40 CANADIAN AGRICULTURAL ENGINEERING, VOL. 28, NO 1, WINTER 1986
 al. (1981) and Graham et al. (1983) vs. was assumed that the initial clay moisture dried desiccant. Agricultural Energy 1.
present study) have different charac content was 0. Two desired final moisture ASAE, St. Joseph, Mich.
teristics. This could be of some concern in contents were assumed, 12% wb (13.6% BROOKER, D. B., F. W. BAKKER-
specifying the characteristics of a given db) and 15% wb (17.7 percent db). The ARKEMA, and C. W. HALL. 1974. Dry
type of bentonite. required mass ratio of clay to corn is then ing cereal grains. AVI Publishing Co.,
In orderto predict the dryinghistory of given for two types of sodium- and cal Westport, Conn.
the intimate mixture of grain and clay it DANZIGER, M. T., M. P. STEINBERG, and
cium-based clays fordifferent initial grain A. I. NELSON. 1972. Drying of fieldcorn
was necessary to have a functional form of moisture contents. It can be seen that a with silica gel. Trans. ASAE (Am. Soc.
the hygroscopicity characteristics of the considerable saving in massof clay can be Agric. Eng.) 15(6): 1071-1074.
clay. This was done by relating the experi obtained by using the calcium-based clay GHATE, S. R. and M. S. CHHINNAN. 1984.
mentally determined hygroscopicity as compared to the sodium-based clay. It Adsorption characteristics of bentonite and
curves (Fig. 5) to Keren and Shainberg's can also be noted that this advantage is a its use in drying inshell pecans. Trans.
values (Eq. 5) by means of a term called function of the final desiredgrain moisture ASAE (Am. Soc. Agric. Eng.) 27:
"Purity" which is defined as content. 635-640.
GRAHAM, V. A., W. K. BILANSKI, and D.
Purity RH,Keren and Shainberg R. MENZIES. 1983. Adsorption grain dry
(mc = constant) RHci
^clay (6) CONCLUSIONS ing using bentonite. Trans. ASAE (Am.
(1) Calcium- and sodium-based ben Soc. Agric. Eng.) 26(5): 1512-1515.
The purity equations for the sodium-based
tonite clays commercially used in the steel GRIM, R. E. 1968. Clay minerology.
and calcium-based clays are given as fol McGraw-Hill, New York.
lows: industrycan be employedin desiccantdry KEREN, R. and I. SHAINBERG. 1979.
ing of grain. Water vapour isotherms and heat of immer
Black Hills: Purity = 0.31 + 1.03 RH - 0.48 RH2
(sodium-based) (7) (2) The desiccant properties of clays sion of Na+/Ca++ montmorillonite sys
vary considerably from one source to tems. II. Mixed systems. Clays Clay Miner.
Dixie Bond: Purity = 0.16 + 0.78 RH + 0.04 RH2 another and even between shipments of 27(2): 145-151.
(calcium-based) (8) clays from the same location. MOONEY, R. W. 1951. The adsorption of
(3) It is possible to predict the drying water vapour by the clay minerals, kaolinite
Numerically evaluating Eq. 1 at several time history of both sodium- and calcium- and montmorillonite. Ph.D. thesis, Cornell
time increments, and using Eqs. 2 - 8 as based clays. University, Ithaca, N.Y.
necessary, it is possible to obtain a time ORCHISTON, H. D. 1955. Adsorption of
(4) The more common type of bentonite
history of the drying of corn using Black water vapour. III. Homoionic montmoril-
clay, the calcium-based clay, adsorbs more lonites at 25°C. Soil Sci. 79: 71-78.
Hills and Dixie Bond clays. Results for moisture than the sodium-based bentonite
two experiments are shown in Figs. 6 and 7 SLABAUGH, W. H. 1959. Adsorption charac
at low moisture contents and is preferable teristics of homoionic bentonites. J. Phys.
in which the closeness of predicted and for grain drying applications. Chem. 63: 436-438.
experimental values are noted. STURTON, S. L., W. K. BILANSKI, and D.
One experimental condition was not R. MENZIES. 1981. Drying of cereal
modelled well in which experimental diffi ACKNOWLEDGMENT grains with the desiccant bentonite. Can.
culties had also been noted. At very high The experimental work was done at the Agric. Eng. 23(2): 109-112.
corn moisture contents 30% wb the clay University of Guelph under a contract TOMLINSON, E. A. and W. M. MIL
became sticky and it was not possible to 0652.0 from Agriculture Canada. LER. 1981. Modeling direct solar
separate the grain and the clay properly. regeneration of solids desiccants for surface
drying. Trans. ASAE(Am. Soc. Agric.
Also at very high moisture contents the
REFERENCES Eng.) 24(3): 770-776.
Chung equation (Eq. 4) is invalid, and ZETTLEMOYER, A. C, G. J. YOUNG, and J.
AMERICAN SOCIETY OF AGRICULTURAL
prediction was not good. ENGINEERS. 1982-1983. ASAE Year J. CHESSICK. 1955. Studies of the surface
The question as to which clay is best book. ASAE, St. Joseph, Mich. chemistry of silicate minerals. HI. Heats of
can be answered by considering the infor BERN, C. J., M. E. ANDERSON, and M. J. immersion of bentonites in water. J. Phys.
mation in Fig. 8. To derive this graph it WILCHE. 1981. Corn drying with solar- Chem. 59: 962-966.

CANADIAN AGRICULTURAL ENGINEERING, VOL. 28, NO. 1, WINTER 1986 41