Western Michigan University

ScholarWorks at WMU

Paper Engineering Senior Theses Chemical and Paper Engineering

4-1990

The Role of Degree of Substitution of Cationic Starch on

Optimizing Retention
Christopher C. Gross
Western Michigan University

Follow this and additional works at: https://scholarworks.wmich.edu/engineer-senior-theses

Part of the Wood Science and Pulp, Paper Technology Commons

Recommended Citation
Gross, Christopher C., "The Role of Degree of Substitution of Cationic Starch on Optimizing Retention"
(1990). Paper Engineering Senior Theses. 198.
https://scholarworks.wmich.edu/engineer-senior-theses/198

This Dissertation/Thesis is brought to you for free and

open access by the Chemical and Paper Engineering at
ScholarWorks at WMU. It has been accepted for inclusion
in Paper Engineering Senior Theses by an authorized
administrator of ScholarWorks at WMU. For more
information, please contact wmu-
[email protected].
THE ROLE OF DEGREE OF SUBSTITlffION OF

CATIONIC STARCH ON OPTIHIZING RETENTION

BY

CHRISTOPHER c: GROSS

A THESIS SlIBHITTED

IN PARTIAL fULfILLHENT OF

THE COURSE REQUIREHENTS FOR

THE BACHELOR OF SCIENCE DEGREE

DEPARTHENT OF PAPER AND PRINTING

SCIENCE AND ENGINEERING

WESTERN HICHIGAN UNIVERSITY'

KALAHAZOO, 11ICHIG.AN

APRIL, 1990
 I

TABLE OF CO:rrt:RTS

ABS'I'RACT.............................................. 1
INTRODUCTION.......................................... 2
ANM,YSIS OF LITERATURE................................ 2
BACKGROIDID.........................................2
FINES.............................................. 2
l1ECHANISH OF FLOCCULATION............................. 3
CHARGE NEUTRM,IZATION.............................. 3
ELECTROSTATIC PATCH FORHATION...................... 4
BRIDGING THEORY
FLOCCULATION TYPE..................................4
ADSORPTION......................................... 5
OPTI11Ul1 POLYl1ER CONCENTRATION...................... 5
HOLECUL.AR WEIGHT................................... 6
FAC'roRS EFFECTING FLOCCULATION........................ 7
ELECTROKINETICS....................................... 8
CH.ARGE AND WET END CHEHISTRY.........................12
CHARGE DETERHINATION..............................12
CATIONIC ST.ARCHES....................................13
CONCLUSION........................................... 15
STATEl1ENT OF PROBLEH.................................16
OBJECTIVES........................................... 17
EXPERil1ENTM, ......................................... 17
TESTING l1ETHODS...................................17
HATERIM,S.........................................19
FINES RETENTION................................... 19
COLLOID TITRATION RATIO........................... 21
RETENTION CM,CULATIONS............................ 21
PRESENTATION AND DISCUSSION OF RESULTS............... 24
STATISTICM, ANM,YSIS.............................. 24
COLLOID TITRATION RATIO........................... 35
CONCLUSIONS ......................................... 37
 II
SUGGESTIONS FOR FURTHER WORK......................... 37
LITERATURE CITED..................................... 39
APPENDICES........................................... 40
 ABSTRACT

The objective of my work was to show experimentally with the

use of statistics whether it is more beneficial to optimize the
degree of substitution of a cationic starch rather than the addition
rate to maximize retention.

Fines retention, as measured using the Dynamic

Retention/Dra.inage Jar (Britt Jar), served as the method of
experimental evaluation. Three starches were studied ranging in
degree of substitution from 0.014 to 0.0833 at a constant molecular
weight. Calcium carbonate was used as the filler fraction because
it can be quantitatively analyzed via an EDTA titration. Starch
furnish used was four, eight, and twelve pounds per ton along with a
30% filler level. Colloid Titrations were used in determining the
ionic demand of the system.

As shown by aultiple regression analysis of the data, it is

equally beneficial to either optimize the degree of substitution or
the amount added to maximize retention. In addition, the Colloid
Titration Ratio was not a good predictor of the ionic demand of the
system.
 2

IBTRODUCTIOB

The subject of this thesis is to determine if there is any

difference in retention achieved from varying the degree of
substitution rather than the addition rate of a high molecular
weight cationic starch. In the following review many topics are
discussed including the definition of fines and fines retention and
the factors effecting retention. In addition, the mechanisms of
flocculation as they pertain to retention will be considered.

llALYSIS or LITERATURE

BACIGROURD

When dealing with wet end chemistry it is comm.on to talk about

dispersed solids in water. The particles in this system are usually
defined in terms of their size alone, which are: gross dispersion,
colloidal dispersion, and ionic dispersion. (1) For optimum results
in retention it is necessary to control the electrokinetics of the
papermaking system. The main mechanism of controlling the fines
loss is by flocculation.

FIRES

The definition used for this experiment for fines is suggested

by Britt: (2) fines are all cellulosic and mineral material passing
through a 200 mesh (75 µm diameter hole) screen. Retention of these
fines is important for several reasons: (3)
1. Recovery of material costs.
2. Improvement of paper properties.
 3

3. Reduction of BOD.
4. Reduction of fiber depositing.
5. Increased drainage rate due to
reduction in hydrodynamic surface
area brought by fines and filler
area.
However, Britt states that maximum retention is not always a good
papermaking practice. (2) There are many unwanted effects from
increased retention such as drainage problems, poor formation, and
inefficient utilization of pigment. <3)

nICH..UIISn or FLOCCOLATIOB

It is generally accepted that particle aggregation can occur

via three basic mechanisms. These are:
1. Charge neutralization.
2. Electrostatic patch formation.
3. Bridging.

CHARGE BEUTRALIZATIOB

This can be defined as coagulation because it accomplishes the

reduction of interparticle repulsion and can be produced by the
following methods: (6)
1. Introduction of an ion of opposite
charge which is specifically adsorbed.
t... Change in pH.
'")

3. Compression of the diffuse layer by

increasing the concentration of the
counterions.
 4

Rapid coagulation occurs on or around the isoelect.ric point and any

further addition of electrolyte beyond that required for this point
does not increase coagulation.

ILICTROSTATIC PATCH FORJlATIOB

The electrostatic patch model involves partial neutralization

of the surface charge of the particles. (?) The result of
flocculation is caused by the attraction between the positive and
negative charge sites on the particles.

BRIDGIBG THIORY

FLOCCULATIOR TYPI

Flocculation is considered a special case of coagulation. The

bridging action of polymers forms a loose, three dimensional floe.
Britt and Unbehend suggest that the only difference between
coagulation retention aids and flocculation aids is that the latter
have a molecular weight that allows one molecule to be attached
simultaneously to several particles. ( 9 )

Two basic types of floes have been proposed. ( 1 0, 11, 1 2)

One type is called a "hard floe" which resists redispersion by
shearing action. The other is called a "soft floe" and is easily
dispersed by shearing action. Holecular weight seems to play an
import.ant. role in resistance of floes to shear. < 1 3, 19 ) The higher
the molecular weight, the more shear resistant the floes that are
formed. (6, 9)
 5

ADSORPTIOll

Bridging is accomplished basically by segments of a polymer

molecule adsorbing onto the surface of a particle and the other
segments of the molecule adsorbing onto uncovered surfaces of a
second particle. (8, 14 ) Strazdins suggested that the long polymer
cha.ins gather the fines by the bridging mechanism and deliver them
to the fibers. < 14 )

Adsorption of the polymer cha.in onto the surface is dependent

on the shear conditions at the interface. (8) The configuration of
the polymer adsorbed onto the surface is determined by the charge
density of the polymer. (8, 1 5, 1 6, 1 7) Flatter configurations are
assumed by polymers of higher charge densities due to the larger
interaction energies with the surface. (6)

According to Arvela, Swanson, and Stratton, molecular weight

has a secondary effect on adsorption. ( 18) they suggested that the
dependence on molecular weight is a function of the charge of the
molecule and is usually negligible.

OPTifflm POLYDIR COllCEBTRATIOR

Latler and Healy have found that there is an optimum. polymer

concentration for flocculation that is governed by two competing
processes. (8)As the polymer concentration increases there is more
polymer to react, however, as this reaction takes place more and
 6

more a,,ailable reaction sites are taken up thus decreasing the

effectiveness of adding so much polymer to begin with. Optimum.
polymer concentration has been found to be independent of molecular
weight but dependent upon agitation time and agitation intensity.
(15)

The requirem.ents for flocculation by bridging are as follows:

(6)

1. Extended segm.ents must be available for bridging.

2. These segments must be sufficient length and
number to form stable floes by bridging.
3. Free surface area must be available to act as
bridge sites.
A flocculated system can become or remain dispersed if any of the
follo\ring occurs: (6)
1. The surface of the particles become so
saturated that bridging cannot occur.
2. The extended segments physically interfere
with one another and prevent bridge
for:ma.tion.

DOLECULll ¥EIGHT

Longer chains will result in longer more num.erous loops

available for bridging. (6) Yorke found that the optimum. molecular
weight depends on the size of the particles to be flocculated. (19)

According to Dobbins, the ideal system would be to floe the

fines/filler portion to improve retention and not to floe the long
 7

fibers in the system to preserve formation and the final physical

and optical properties of the sheet. (4) However, this system
cannot be obtained as noted by the author. The order of adsorption
of particles by the cationic polymers are those with high surface
area, high charge density first, and those with decreasing surface
area and charge density in succession. Hence, there is no selective
adsorption and the long fibers flocculate along with the fines. The
optimllill amount of flocculation, controlled by polymer addition, lies
somewhere near the point where the retention of fines has been
maximized, but where the flocculation of long fibers has not yet
begun. (3)

FACTORS EFFECTIBG FLOCCULATIOB

King and Williams point out that polymer bridging must be able
to withstand shear forces to remain effective. (5) Other factors
which influence the effectiveness of floe formation and stability
include polymer type, polymer concentration, electrolyte type,
electrolyte concentration, fines addition level, agitation level,
and pH. (3) Studies have shown that a high molecular weight, low
charge density polymer under minimal shear will give the best
retention.

The characteristics of the pulp furnish used can effect

retention. The average fiber length, fiber length distribution,
fiber flexibility, fiber surface roughness have been found to effect
retention by affecting fiber/fines flocculation. <23) In addition,
due to more surface area generated by refining, furnishes that are
refined more retain more fines.
 B

The machine variables used during the papermaking process

directly affect retention. Due to the mechanical retention,
reducing machine speed, reducing the vacuUJn in the suction boxes,
and using hydrafoils instead of table rolls all increase retention.
(6) Also, machine speed increase reduces retention due to more
shear generated which degrades the floes that were formed.

The soluble fraction of a furnish also effect retention.

Again as seen in Table 1, there are a variety of soluble species
which contribute to the cationic demand of the system. Another
factor affecting retention is the protective colloid effect. This
occurs when: (6)
1. The surface of the colloid is partially
soluble,forming a gel like structure.
2. When various white water components absorb
onto the surface of the particles.

ELECTROIIRETICS

The repulsive forces that are present in papermaking systems

and that prevent flocculation without the addition of chemicals are
of an electrical nature. (6) The charge on the surface of the fines
is responsible for the repulsion since like charges repel one
another. The surface charges occur due to: (20)
1. Ionization of molecules on the surface, or
2. Selective adsorption of specific ions from
solution,or
3. Faults or defects in the crystal lattice, or
 9
4. Partial solution of a crystal lattice.

The zeta potential is an indication of the repulsion of the

suspended particles in the paperm.aking system. The zeta potential is
defined in Figure 1. It is reflection of a system of adsorbed ions
and counterions. In terms of paper fiber, which inherently carries
a negative charge, opposite charged ions are attracted to and near
the surface of the solid. The next layer, known as the Stern Layer,
contains imm.obilized counterions. The Gouy Layer, which is adjacent
to the Stern Layer is made up of the rest of the cour1ter ions.
There is a definite thickness of surrounding liquid bound to the
particle. The electrical potential at the plane of shear between
the bound liquid and the free liquid is the zeta potential.

As seen in Table 1, there are many soluble and fiber/fines

associated charged species all of which contribute to the zeta
potential of a system. When the zeta potential has been
neutralized, the iso electric point of the particle has been reached
which makes them susceptible to flocculation. Hence, due to low
repulsion of particles in the system. high retention is
favored.Valette proposed that the zeta potential, and thus the
repulsive energy potential is sensitive to both the nature of the
water and the pH of the solution. <21) In addition, fiber type,
filler type, alum dose, polymer addition and wet end additives also
have an effect on zeta potential. Arno, Frankle, and Sheridan
proposed that the zeta potential of a furnish is a com.posite of the
zeta potentials of all the particle types in the system; each
particle type has a unique zeta potential dependent on the shape of
the particle. <22)
 10

Table 1

Soluble and Fiber/Fines Associated Charged Species

Effect on Wet End Contribution to

Substance Chemistry Charge determ.ina. tion
sodiUill very slight none
calciUill modern.te at high cone. slight
magnesium moderate at high cone. slight
alUillina. large moderate
acid pH large small
sulfate moderate at high cone. none
chloride none none
lignin large large
tannins large large
dispersants large large
coagulants large large
cellulose large large �
filler large large �

�=These effects are seen in determination on total cationic

demand.
 l l

POTENTIAL
DETERMINING • SHEAR
ION LAYER PLANE

Y:
I
I
I
I STERN GOUY-CHAPMAN
I REGION /EGION

I• _., ..,..._______►
I
I I

I DISTANCE FROM
0 I SURFACE
ELECTRIC
POTENTIAL

ZETA
POTENTIAL

FIGURE 1 - ZETA POTENTIAL SCHEME

 12

CHARGE ARD ¥ET EBD CBEDISTRY

Papermaking furnishes contain many charged materials which all

have a profound influence on the retention of fine solids and
chemical additives. The Charge Determination technique can be used
as a powerful diagnostic tool since it measures the sign and
mc1.gnit.ude of the charge. Charge Determination measures directly the
combined total of soluble charge and particle charge. As shown in
Table 1, not all soluble and fiber/fine charged species are detected
in Charge Determination nor are all of them a factor in wet end
chemistry as noted in the Table.

CHARGE DETP.:RDIBATIOB

The water from papermaking furnishes can be titrated much like

an acid base titration to an end point of zero charge. This is what
is called Charge Determination. The value and sign of the charge
depends on the quantity of titrants used to reach the end point.
This quantitative charge is useful in diagnosing retention studies.

The Charge Determination method measures directly the water

soluble charge only of the furnish. By appropriate experimentation
the charge demand of the entire furnish can be determined by taking
each fraction separately and measuring its charge. However, it is
important to note the Charge Determination value is not identical to
zeta potential. Charge Determination measures the amount of
cationic flocculant needed to satisfy the actual and potential
anionic charge of the water/fiber system. The suspended solids in a
furnish can continue to adsorb the cationic polymer past the point
of zero zeta potential. This is a function of the ion exchange
 13

capacity of the pulp and filler. Polymers can adsorb against.

electrical charge barriers even onto neutral surfaces.

The total amount of cationic polymer that a furnish will

adsorb is the sum of the following charge demands:
1. Soluble anionic charge.
2. Actual anionic charge on the surfaces of
fiber and fines.
3. Ion exchange potential of fiber and fines.
Charge Determination measures directly only the first. The sum of
2. and 3. is determined indirectly through a series of tit.rations
described in the experimental plan.

CATIOBIC STARCHES

The basic repeating unit of starch is an anhydroglucose unit.

These units polymerize through alpha (1,4) or alpha (1,6) linkages
to form amylase or amylopectin respectively. A common starch
consists of both polymers and some respective percentages are seen
in Table 2.

Table 2.

Amylose and Amylopectin Content of Various Starches

Starch type ".Amylopectin ,.: hylose
Regular Corn 73 27
Potato 80 20
Waxy 11aize 100 0
High Amylose 20 80
 14

Variations in the ratio of amylose to amylopectin will result in

variations in viscosity, molecular weight, and charge distribution.
(24)

Cationic starches can be prepared by a variety of methods with

three of these being:
1. Reaction of starch with beta-halogenated
amines in the presence of NaOH.
2. Reaction of starch glycidyl tertiary amines
in the presence of NaOH.
3. Reaction of starch with epoxypropyltrimethyl­
a:mm.onium chloride.

The third method prepares a quartenary amine starch which could be

represented by the following formula.
Starch-R-O-CHz-CH(OH)-CHz-N+ -(CH3)3 + c1-
Cationic starches prepared using primary, secondary, or tertiary
amines must be used in acid systems to attain positive charge.
"Onium starches" are starch derivatives bearing fully substituted
radicals such as qua.rtenary amines, qua.rtenary phosphates, or
tertiary sulfonates. These starches carry a full positive charge
over the entire pH range.

Unmodified starches are not retained well in the sheet and the
retention that does occur is reversible; over 80% of the starch
adsorbed by paper can be removed by heating the paper in water. <25)
Cationic starches are strongly adsorbed by fibers and this retention
is irreversible; 85% of the adsorbed starch remains when the paper
is placed in boiling water and can be redissolved only by washing
 15

with strong acid. <25) As such, cationic starches vn.11 remain

adsorbed to the fibers during processing and reuse of paper mill
broke, will not contribute to the BOD of the mill effluent as
unmodified starches do. <25) J. 11a.rton and T.11a.rton <25)have
studied cationic starch adsorption by cellulose fibers and reported
the followi1)g:
1. Cellulose fines retain more cationic starch
than do fibers.
2. Increased refining increases the adsorption
of the cationic starch.
3. Soft.wood and hardwood fibers retain cationic
starch equally well.
4. Dried pulps retain less cationic starch than
never dried pulps.
5. The adsorbing capacity of fibers increases as
carboxyl content increases.
6. Alumina precipitated on a fiber reduces its
adsorptivity of cationic starch.

COHCLUSIOH

The use of cationic starch as a retention aid has been used

successfully in many mill applications. Reduced rosin consumption,
the use of cheaper filler with no appreciable loss in strength or
printing performance, and increased ma.chine speeds due to increased
drainage rate allowed production cost at one mill to be reduced when
cationic starch was used as a retention aid. <27) Van der Burgh
found that the other effects os using wet end cationic starch
included : ( 28)
1. Reduced picking at the press rolls.
 16

2. Faster machine equilibration after grade

changes due to improved one pass retention.
3. Better color retention due to increased
retention of pigmented dyes and the fines to
which these dyes are adsorbed.
4. Less sheet two sidedness due to more uniform
retention throughout the sheet.
5. Less paper shrinkage during drying.
6. Change in sheet porosity.

STAT!:m::rr o:r PROBL:[D:

A discussion of the complexity of fines with relation to their

retention has been included in the previous literature search. It
is the intention of this study to experimentally show the effects of
degree of substitution at constant. molecular weight and through the
use of statistics develop a model for prediction of the optimum
degree of substitution.
The primary purpose of this work is to show if it is more beneficial
to optimize the degree of substitution rather than the addition rate
to maximize retention of fines. This comparison will be done by
using mathematical models generated by regression analysis. The
dependent variable in the model will be retention values(First Pass,
Total Fines, and CaC03), and the independent variables will be the
addition rate and the degree of substitution. The Colloid Titration
Ratio test will be used to determine the cationic demand of each
system and the DRJ will be used to determine the retention achieved
by each chemical system.
 17

The relationship between degree of substitution of cationic

polymers and fines retention bas been studied previously. However,
no work bas to the author's knowledge been done to see if it is
beneficial to optimize the degree of substitution per system rather
thar1 the addition rate. Therefore the results of this research
could prove beneficial in the approach that papermills take in
selection of a cationic starch that is to be used as a retention
aid.

OBJECTIVES

1. To experimentally establish a relationship between the

degree of substitution of a cationic starch and its
fines retention capability.
2. To develop statistical models via regression analysis
of results to show effects of degree of substitution
and addition rate.
3. To show if it is better to optimize the degree of
substitution rather than the addition rate to maximize
retention of fines.

EXPERill!JITAL

The overall experimental approach is to use a Dynamic

Retention/Drainage Jar (DRJ), a Colloidal Titration and a hardness
titration to generate results for comparison of various starches in
 18

determining which variable affects retention the most.

The DRJ, developed by Britt and Unbehend measures the effect
of turbulence on colloidal retention (retention due to van der Waals
and electrostatic forces) independent of the mechanical variables
that effect retention. (9) Isolation and elimination of the effect of
mechanical factors in the DRJ is accomplished:
1. Preventing mat formation.
2. Controlling screen mesh size by using
electrodeposited screens having precisely
controlled hole size.
3. Using an agitator speed system to control
turbulence.
The DRJ attempts to simulate the headbox slice and the wire section
that iuediately follows where there is little retention due to
entrapment in the fiber mat as the mat is in dynamic state. A large
fraction of the unretained fines are lost to the white water at the
slice and in the initial forming area. Thus, the DRJ is a useful
method to examine retention of a paper:m.a.chine in the laboratory.
Data from the DRJ can be correlated to an actual paperma.chine by
determining the turbulence level in the jar that matches that on the
:ma.chine. This is accomplished by :matching the one pass retention
level of the :ma.chine with the turbulence in the jar that gives the
same retention level using a "flocculation index" curve. (9)

In carrying out the experimentation necessary to accomplish

the objectives, the following was used:
Fiber
-80% Bleached Hardwood Kraft
-20% Bleached Softwood Kraft
 19

Filler
-30% CaC03
The fiber fraction was refined at 1.57% consistency in the
laboratory Valley beater to approximately 400 ml. CSF. An
appropriate amount of filler was added to achieve the 30% target.
The resultant furnish was then diluted to 0.50%.

llATERIALS

The starches to be used were experimentally developed by a

starch manufacturing company using amylopectin as the parent
molecule. The st.arches had been precipitated from the paste form in
which they were converted using methanol. The degree of
substitution data, obtained from the Kjeldahl procedure for nitrogen
determination, was provided by the starch supplier.
The following starches were used.
11olecular Weight High
Degree of Substitution . 0226
. 0307
. 0833
The starches were cooked at 20.0% solids and then diluted to 0.50%
for application.

FIRS RETEBTIOB PROCEDURE

Three various degrees of substitution along with three various

addition rates and corresponding blank runs were studied. These
runs were done in triplicate to yield a total of thirty different
experiments. A random number generator was used to designate the
order the experiments. This was done to elimir.18.te any systematic
errors that could have occurred in the experiment.
 20

The DRJ was used to determine the retention characteristics of

the system. A 200 mesh (75 Jlm) screen was used along with a vaned
drainage chamber (see Figure 2). The fiber furnish consisted of a
blend of 80% bleached hardwood kraft and 20% bleached softwood kraft
along with 30% Caco3 based on dry fiber. The fibers were then
refined to 400 ml CSF using a laboratory Valley beater, filler was
then added, and the resultant pulp the diluted to 0.50% consistency.
The procedure was as follows.
1. Heasure out 500 ml.
2. Pour into DRJ and e.gitate at 400 RPH
3. Add starch - mix for 1 min.
4. Turn up RPH to 750 - mix for 1 min.
5. Collect sample in pre-tared 4 oz. plastic bottle w/lid
6. Take excess of 50 ml. of Hbox stock for Colloid
Titration
7. Weigh sample bottle.
8. Add 5 ml. of 5N HCl
9. Wait 1/2 hour
10.Filter fiber on tared filter paper and collect
filtrate
11.Heasure out with pipette 25.0 ml. aliquot and place
in 100 ml. beaker. Add 1 - 2 ml. of hardness buffer
and enough indicator for medium. pink color.
12.Titrate with .02N EDTA to blue end point. Record Titer
13.Take 500 ml. of Hbox* furnish and repeat steps 7-12
14.Heasure out 100 ml. samples for fines analysis. Pour
sample into DRJ. Dilute to 500 ml., and turn on
agitator to 1000 RPH. Add 5 ml. of Ta.mol and let mix.
Drain, rinse and repeat several times. Rinse off
 21

fiber from screen onto tared filter paper. Dry and

weigh.
% fines= f (Hbox cons.)- (wt. of fiber)l100
(Hbox cons.)
15. Take process water samples for Hardness Titration.

COLLOID TITRATIOB RATIO

To determine the colloid titration ratio a 50 ml. sample of as

is stock is diluted to 500 ml. Then four 100 ml. aliquots are then
taken and pour into 250 ml. beakers. A :magnetic stir bar is then
used and the test is then run according to Fig. 3 . It should be
noted that after the addition of 10m.ls. of either polymer the
solution was allowed tom.ix for one minute. DDPH is the cationic
polymer and PVS.AK is the anionic polymer and both are 0.1N.

R!:"l"EBTIOR C.I.LCULATIORS

To determine the percent Caco 3 retention, a Hardness Titration

was done.

*Hbox is an abbreviation for Headbox.

A 25,0m.l. aliquot was transferred via a pipette to a 100m.l.
beaker. 5m.l. of 5N HCl was added to dissolve the Caco 3 via the
following reaction:
CaCO3 + 2 HCl = CaC12 + H2co 3 (soluble salt)
1 - 2m.ls. of hardness buffer was added along with
 22
Eriochrome Black T indicator to give a medium pink color. A blank
was run on the process water used in actual caco3 retention. The
percent Caco3 retention was calculated as follows.
wt. CaC03 = (ml EDTA smpl. - ml EDTA blank)
10 (ml of sample)

% caco3 retained= (wt. eaco3 Hbox - wt. caco3 smpl. )100

wt. CaC03 Hbox

% First Pass = ("wt. Hbox - wt. siapl. )100

wt. Hbox

. *
% Total fines = (.38(wt. Hbox) - wt. smpl. )100
.38(wt. Hbox)

*.38 IS FROH 'roTJJ.. FINES CALCULATED BY BRITT DEFINITION.

 23

FIGURE 2 - COLLOID TITRATIOH R!TIO

CATIONIC ANIONIC
DEl1AND DE11AND
DETEPJITNATION DETERHINATION

' • ' •
�� ¥/��
ADD 10 111. FILTER ADD 10 111. FIL'Il:R
DDP11 PVSAK

• ' ' '

ADD 10 l:1L. .ADD 10 HL.
FILTER DDPH FILTER PVSAI

ADD 6 ADD 6 ADD 6

•
ADD 6

•
DROPS TBO DROPS TBO DROPS 'I'BO DROPS TBO

•
(BLUE) (BLUE) (PINK) (PINK)

♦
BACK TITRATE BACK TITRATE BACK TITRATE BACK TITRATE

' •
WITH PVS.AK WITH PVS.AK WITH PVSAK YHTH PYS.AK

• •
(PINK) {PINK) (BLUE) (BLUE)

CATIONIC DEllJJID CATIONIC DEltAND ANIONIC DDWID ANIONIC DEl1AND

OF SAHPLE OF BLANK OF SAHPLE OF BL.Al-lK
(A) (B) (C) (D)

(B-.A) = CATIONIC DEl1A.ND OF S'IOCK

(D-C) = JJUONIC DEJ.1AND OF S'K•CK

(D-C) ·= COLLOID TITRATION RATIO

(B-A)
 24

PRESEBTATIOH AJID DISCUSSIOH OF RESULTS

The results of these experiments will be eva.luated and

presented on the basis of calcium carbonate, total fines and first
pass retention. The retention is relative to degree of substitution
and addition rate. The results will be discussed as they relate to
one another in a statistical model.

STATISTICAL llALYSES

It was the intention of this study to analyze the data with

the use of regression analysis and analysis of variance. The
objective of regression analysis is to find a model which is both
simple ( relatively few parameters) and provides a good fit to the
data. The first step taken in regression analysis is to find out
which model actually fits the relationship between the dependent and
independent variable. In this experiment there were two independent
variables (degree of substitution and addition rate) and one
dependent variable {retention va.lues). Therefore, plots of
retention versus addition rate and retention.versus degree of
substitution were generated. The purpose of this was to see wha.t
type of mathematical relationship existed. Figures 3 - 8 show tha.t
a quadratic provide the best fit to the data.Each point on the plots
is the mean of three determination of retention. The error bars
sho�1ll. represent one standard deviation above and below the mean of
three determinations. Hence, a quadratic regression model was used
for comparison purposes.
 25
FIGURE 2 - PROOF FOR QUADRATIC REGRESSION
MODEL FOR TOTAL FIN-ES RETENTION
60.0%
z
. 0
I- 50.0%
z
w
I-
w 40.0%
C:
CJ)
w 30.0%
z
u..
...J
20.0%
<t
I-
0 10.0%
I-

0.0%
0 5 10 15
ADDITION RATE (#/TON)

FIGURE 3 - PROOF FOR QUADRATIC REGRESSION

MODEL FOR TOTAL FINES RETENTION
60.0% -----------------
z
0
1-
50.0%
z
w
1-
w 40.0'%
C:
CJ) 30.0%
w
z
u.. 20.0%
...J
<t
l­
o 10.0%
I-
0.0% ___________________.
0.0000 0.0200 0.0400 0.0600 0.0800 0.1 000
D.S.
 26
FIGURE 4 - PROOF FOR QUADRATIC REGRESSION
MODEL FOR FIRST PASS RETENTION
8b.U "'/o ---------------------,

z
0
i-
z 80.0%
w
i-
a:
Cf)
75.0%
Cf)

i­ 70.0%
(/)
a:
u..
65 .0% ---------.------.---------------.---
0 5 1 0 15
ADDITION RATE (#/TON)

FIGURE 5 - PROOF FOR QUADRATIC REGRESSION

MODEL FOR FIRST PASS RETENTION

z
0
i-
z 80.0%
w
i-
a:
Cf) 75.0%
Cf)

i- 70.0%

u..
65.0% -t--�"--T---,.--------------,---t
0.0000 0.0200 0.0400 0.06 00 0.08 00 0.1000
D.S.
 27
FIGURE 6 - PROOF FOR QUADRATIC REGRESSION
5 MODEL FOR CALCIUM CARBONATE RETENTION
60.u-;o
1-
z
w
1-
w
a:
w 50.0%
1-
<t
z
0
m
C:
<t 40.0%
C)
�
::,
C)
...J
<t 30.0% -+--..----.----.----.---,,---..---,,--,-------.----.----.----,,---�
C)
0 5 10 15
ADDITION RATE (#/TON)

FIGURE 7 - PROOF FOR QUADRATIC REGRESSION

MODEL FOR CALCIUM CARBONATE RETENTION
z
0 60.0% ------------------
1-
z
w
1-
w
a:
w 50.0%
1-
<t
z
0
m
C:
<t 40.0%
C)

-C)
�
::,
...J
<t 30.0% -------------...-------
C)
0.0000 0.0200 0.04 00 0.0600 0.0800 0.1 000
D.S.
 28

There are several statistical tests that can be used to test

the utility of the model that was generated through regression
analysis. These tests are all a part of analysis of variance. The
tests used were the t-ratio, the F test and the p test.

The t-ratio is a test to see if the constant, linear, and

quadratic terms were necessary in the regression model.

H0 : n2 = o, Ha_ : n2 $ o
H 0 here says that the term is not necessary. Hence, for the model
to be used H 0 must be rejected. For 95% confidence H 0 is rejected
if t.05,7 � 2.179 or t.o5,7 � -2.179. As seen in Table 4 each
individual regression term for every model is necessary to describe
the relationship that existed.

The F test is used to verify if the quadratic regression model

explains most of the variation in he observed values. The F value
is actually the ratio of explained variation to unexplained
variation. The null hypothesis for this test is as follows:
H 0 : n 1 = n2 = O, Ha_ : @ least one n $ o
For a 95% confidence F.o5,2,7 must be> 4.74. If the explained
variation is high relative to unexplained, H 0 would naturally be
rejected and this would confirm the utility of the model. As seen
in Tables 5 - 7, the F values for each individual model were greater
than 4.74.
 29

T.ABLE 4 - t-ratio TEST FOR QUADRATIC REGRESSION

FIRST PASS RETEJITIOB

x = I/TON x = deqree of substitution
PREDICTOR t - ratio PREDICTOR t - ratio
constant 33.56 constant 38.61
X 3.87 X 5.04
x"'2 -3.63 x"'2 -4.75

CALCIUn CARBOllTE RETDTIOB

x = I/TON x = degree of !ub!titution

PREDICTOR t - ratio PREDICTOR t - ratio
constant 14.14 constant 17.77
X 3. 92 X 5.91
x"2 -3.43 x"2 -5.92

TOTAL FIJIES ltETEftIOR

x = I/TON x • degree of substitution
PREDICTOR t - ratio PREDICTOR t - ratio
constant 3.88 constant 3.76
X 3. 87 X 5.00
x"2 -3.63 x"2 -4.72
 30

TABLE 5 - FIRST PASS REGRESSION ANALYSIS AND ANOVA

y = .701 +.0265 x-.00177x A 2 y = RETENTION

X = #/TON
dy/dx = 0 leads to x = 7.5 #/TON RA 2 = 68.4%
= 7.58
y(f) = 80.0 ± 5.6 % = .02

s = .02248 s(f) = .00998

y = .687 + 4.93 x- 46.2 X A 2 y = RETENTION

x = D.S.
dy/dx = 0 leads to X = .0534 R "'·2 = 78.9%
F = 13.1
y(f) = 81.9 I 4.6 % p = .00

s = .0J..837 s(f) = .00826

 31

TABLE 6 - CaCO3 REGRESSION ANALYSIS AND ANOVA

y = .381 +.0346 x-.00217x A 2 y = RETENTION

= #/TON
A
dy/dx = 0 leads to x = 8.0 #/TON R 2 = 70.8%
= 8.5
y(f) = 51.9 ± 3.2 % = .01

s = .02889 s(f) = .01287

y = .376 + 6.86 x - 68.4 x A 2 y = RETENTION

x = D.S.
dy/dx = 0 leads to X = .0501 RA2 = 78.9%
F = 13.1
y(f) = 54.8 ± 2.4 % p = .00

s = .02182 s(f) = .00981

 32

TABLE 7- TOTAL FINES REGRESSION ANALYSIS AND ANOVA

y = .213 +.0699 x-.00468 x A 2 y = RETENTION

X = #/TON
dy/dx = 0 leads to x = 7.5 #/TON RA2 = 68.4%
= 7.58
y(f) = 47.4 ± 6.5% p = .02

s = .05924 s(f) = .0263

y = .178 + 13.0 x -122.0 x A 2 y = RETENTION

X = D.S.
dy/dx = 0 leads to X = .0533 RA 2 = 78.7%
= 12.90
y(f) = 52.4 ± 5.3 % p = .00

s = .04780 s(f) = .0219

 33

.Another test that was used to test the utility of the model was the
p test. The null hypothesis for this test is as follows:
H0 : n2 = o, Ha : n 2 $ o
If H0 is rejected this shows that the complete model is a very
useful tool for prediction. For 95% confidence H0 is rejected if p
< .05. Tables 5 - 7 show that all p values are below .05.

The coefficient of multiple determination (R A 2) is another

number generated through the analysis of variance. This number
indicates what percentage of observed �riation can actually be
explained by the quadratic regression model. For example, as seen
in Table 5 R A 2 is 68.4%, so this states that 68.4% of the observed
variation were explained by the regression equation.

As seen in Table 8 the quality of results indicated by most of

the standard deviations are very good. However, some standard
deviations were high which led to the insignificant levels for the
coefficient of multiple determination.

To compute the maximum values that could be obtained from the

basis of the experimental results the first derivative of the
regression equation was taken and set equal to zero and solved for
x(degree of substitution or addition rate). This led to the maximum
value of y ( the particular retention being investigated). For
comparison of maximum predicted retention values for each situation,
95% confidence interval were generated using the standard deviations
of the observed
 34

TABLE 8 - SUMMARY OF RETENTION RESULTS

AVERAGE AVERAGE AVERAGE

( .tt /TON)/D.S. FIRST PASS CALCIUM CARBONATE TOTAL FINES
69.1% 38.0% 18.7%
BLANK ±.2 ±.7 ±.7
78.9% 48.5% 44.5%
4/.0226 ±.6 ±.8 ±1.6
79.1% 51.3% 45.0%
4/.0307 ±.3 ±.7 ±.9
78.7% 45.9% 43.9%
4/.0833 ±.1 ±.6 ±.2
78.3% 49.9% 42.9%
8/.0226 ±.2 ±.3 ±.5
81.2% 56.8% 50.6%
8/.0307 ±.2 ±6 ±.4
77.4% 48.7% 40.5%
8/.0833 ±.6 ±1.1 ±1.4
73.5% 48.1% 30.2%
12/.0226 ±.4 ±.3 ±1.2
79.8% 50.3% 46.8%
12/.0307 ±0 ±0 ±0
76.9% 46.9% 39.1%
12/.0833 ±.9 ±.7 ±2.3

NUMBERS BELOW THE PERCENTAGES ARE± ONE STANDARD DEVIATION

 35

and predicted future value. The equation tor the confidence

interval is as follows:
y(f) ± t .025,9(�s 2 +s(f) 2)
A A

The results can be seen in Tables 5 - 7.

COLLOID TITRATIOB RATIO

As shown in Table 9 no trend existed as to what actually

ha.ppened to the wet end chemistry for each system. The standard
deviations were much too great which makes the numbers generated
insignificant. A possible explanation could be that not all
titrations were done on the same day. Also, since the color change
is observed by eye, human error could be high. Hence,this is not a
very good research tool to be used unless an instrument can be used
to indicate color change.
 36

TABLE 9 - SUHHARY or COLLOID TITRATION

AVERAGE AVERAGE AVERAGE

{1/TOB}/D. S. llIOBIC DEIWID CATIOBIC DE1[ll]) CTR

BLAH 0.13 2.57 0.05

:1:0.06 :1:0.21 :1:0.02
4/.014 0.63 2.10 0.32
:1:0.49 :1:0.35 :1:0.27
'l/.0307 0.47 1. 77 0.27
:t0.15 :t0.67 :t0.09
'l/.0833 1.60 1.03 3.02
:t0.78 :t1.01 :t2.i6
8/.01'1 0. 77 1.77 0.44
tO.31 :!:0.21 :!:0.19
8/.0307 0.90 2.67 0.31
:t0.76 :!:1.33 :!:0.19
8/.0833 0.90 1.57 0.66
:1:0.46 :1:0.21 :1:0.21
12/.01'1 0.70 1.63 0.63
:1:0 :1:0.83 :1:0.47
12/.0307 1.10 2.13 0.62
:t0.26 :1:0.12 :1:0.12
12/.0833 1.10 1.77 0.62
:1:0.26 :1:0.36 :1:0.06
 37

CORCLUSIORS

On the basis of data collected and analyzed the following

conclusions were drawn:

1. Based on comparison of the values and their

confidence interval it was equally beneficial to
either optimize the degree of substitution or tt1e
a.ddition rate of the cationic starch.

2. The Colloid Titration Ratio was not a good test for

this research.

3. The use of regression analysis is an excellent tool

for this type of analysis.

4. Cationic starch is a very good retention aid in a

naturally buffered furnish which contains calcium.
carbonate.

5. The Dynamic Britt Jar was a good research tool for

comparing retentions given by various polymer systems.

SUGGISTIORS FOR FORTBIR YORK

1. Identical trials can be done but with more

repetitions.

2. Handsheets can be made to see whether it is more

 38

beneficial to optimize the degree of substitution

rather than the addition to optimize strength.
 39

LITERATOR� CITED

1. Sennett, P. and Oliver, J.P., Interfaces Symposium, June

15 & 16, 1964, Chemistry and Physics of Interfaces,
J\merican Chemical Society Publications, Wast.1ington
D.C.,75 p.

2. Britt, K.W.,Tappi 56(10): 46(1973).

3. Goulet, H. "The Effect of Holecular Weight and Charge

Density of a Cationic Starch on Retention", B.S. Thesis.
Western Hichigan University, 1983.

4. Walkush, J. C., a.nd Williams, D.G., Tappi 57(1): 112

(1974).

5. Casey, J.P., "Pulp and Paper". Vol. II, Interscience

Publishers Inc., New York. 1959 Chapter 2.

6. Bale, H.D."The Effect of Holecular Weight and Charge

Density of a Cationic Starch on Retention"B.S. Thesis,
Western 11ichigan University, 1979

7. Deyaguin,B., Tran. Faraday Soc. 36(203): 730(1940)

8. Lamer, V.K., and Healy, T.W., Review of Pure and Applied

Chemistry 13: 112(1963).

9. Britt, K.W.,and Unbehend. J.E. Tappi 57(10): 81(1974).

10. Britt, K.W.,Tappi 56(10): 46(1973).

11. Unbehend, J.E.,Tappi 56(10): 74(1973).

12. Britt, K.W., Tappi 56(10): 83(1973).

13. Pummer, H., Das Pa.pier 27(10): 417(1973)

14. Strazdins,E.,Tappi 53(1): 80(1970).

15. Goosens, J.W.S., and Luner, P., Tappi 59(2): 89(1976).

 40

16. Institute of Paper Chemistry, The Role of Polymers in

Papermaking, J:1arch 7, 1977 and April 4, 1977.

17. Strazdins, E., Preprint: First International Seminar in

Hill Chemistry, .Amsterdam, Netherlands, Sept. 11-13,
1977.

18. Arvela, P.,Swanson, J.W., and Stratton, R.A., Tappi

58(11): 86(1975).

19. Yarke, H.A.,In "Water Soluble Polymers", New York, Plenum

Publishing Co.., 1973, pp. 93-104.

20. Williams, D.G., and Swanson, J.W: on "Retention of Fine

Solids during Paper Hanufacture", T.Af•PI CA report No. 57,
Sept. 1, 1975.

21. Valette, P., Revue A.T.I.P. 29(1): 1(1975).

22. Arno, J.N., Frankle, W.E., and Sheridan, J.L. Tappi

57(12): 97(1974).

23. Ha.son, S.G., Tappi 3(7): 440(1950).

24. Bale, H., Harvey, R.D., Hubbard, E.D., and Klem, R.E.
Cationic Starches in Papermakinq Applications, 1980 Tappi
Retention and Drainage Course.

25. Rogols, S. and High, R.l.,Tappi 46(7): 424(1963).

26. Ga.spar, L. in"Retention of Fines Solids during Paper

11anufacture", T.APPI Call Report No. 57, Sept. 1, 1975.

27. Stephenson, W.D., and Hoeller, H.W. Pulp and Paper

41(47): 41 (1967).

28. Van der Burgh, L.F., Pulp and Paper of Canada 71(18):
T403(1970).
 41
X
0
z
w
Q_
Q_
<(
 APPENDIX I - COLLOID TITRATION DATA

ml. DDPN ANIONIC AVERAGE ml.PVSAK CATIONIC AVERAGE CTR AVERAOE

•/TON -D.S. BLANK SAMPLE DEMAND DEMAND BLANKSAMPLE DEMAND DEMAND IONIC RATIO CTR
BLANK 10.0 9.8 0.2 10.0 7.2 2.8 0.07
BLANK 10.0 9.9 0.1 0.13 10.0 7.5 2.5 2.57 0.04 0.05
1
BLANK 9.0 8.9 0.1 0.06 10.6 8.2 2.4 0.21 0.04 0.02
4/.014 10.0 9.7 0.3 9.7 7.8 1.9 0.16
4/.014 10.2 9.8 0.4 0.63 10.4 7.9 2.5 2.10 0.16 0.32
4/.014 11.1 9.9 1.2 0.49 10.7 8.8 1.9 0.35 0.63 0.27
4/.0307 10.0 9.7 0.3 9.8 8.5 1.3 0.23
4/.0307 9.9 9.3 0.6 0.47 9.9 8.3 1.6 1.77 0.38 0.27
4/.0307 10.0 9.5 0.5 0.15 10.1 7.7 2.4 0.57 0.21 0.09
4/.0833 10.1 8.2 1.9 9.6 9.1 0.5 3.80
4/.0833 10.3 8.3 2 1.50 10.1 9.7 0.4 1.03 5.00 3.02
4/.0833 10.2 9.6 0.6 0.78 10.2 8.0 2.2 1.01 0.27 2.46 �
N
8/.014 9.2 8.7 0.5 9.6 8.0 1.6 0.31
8/.014 10.1 9.0 1.1 0.77 9.8 8.1 1.7 1.77 0.65 0.44
8/.014 9.8 9.1 0.7 0.31 9.9 7.9 2 0.21 0.35 0.19
8/.0307 10.2 9.4 0.8 10.0 8.2 1.8 0.44
8/.0307 9.9 9.7 0.2 0.90 10.1 8.1 2 2.67 0.10 0.31
8/.0307 9.9 8.2 1.7 0.75 12.2 8.0 4.2 1.33 ·' 0.40 0.19
8/.0833 10.0 9.5 0.5 9.8 8.4 1.4 0.36
8/.0833 10.1 9.3 0.8 0.90 10.0 8.5 1.5 1.57 0.53 0.56
8/.0833 10.2 8.8 1.4 0.46 10.2 8.4 1.8 0.21 0.78 0.21
12/.014 11.7 11.0 0.7 10.0 8.2 1.8 0.39
12/.014 9.8 9.1 0.7 0.70 10.1 9.5 0.6 1.53 1.17 0.63
12/.014 11.2 10.5 0.7 0.00 10.2 8.0 2.2 0.83 0.32 0.47
12/.0307 10.0 9.0 1 10.0 8.0 2 0.50
12/.0307 10.2 9.3 0.9 1.10 10.0 7.8 2.2 2.13 0.41 0.52
12/.0307 10.5 9.1 1.4 0.26 10.7 8.5 2.2 0.12 0.64 0.12
12/.0833 10.7 9.8 0.9 9.7 8.3 1.4 0.64
12/.0833 10.7 9.7 I 1.10 9.8 8.0 1.8 1. 77 0.56 0.62
12/.0833 9.9 8.5 1.4 0.26 10.0 7.9 2.1 0.35 0.67 0.06
z:v
 APPENDIX II - BRITT JAR VALUES

WET WEIGHT DRY WEIGHT WT. OF FIBER

D.S.--•/TON TARE WT. GROSS WT. NET WT. TARE WT. GROSS WT. NET WT. g/ 100ml
.0226--12 19.06 122.22 103.16 0.434 0.482 0.048 0.047
.0226--12 19.21 125.75 106.54 0.436 0.483 0.047 0.044
.0226--12 19.11 120.95 101.84 0.500 0.543 0.043 0.042
.0226--4 19.31 118.61 99.3 0.474 0.490 0.016 0.016
.0226--4 19.18 112.68 93.5 0.511 0.530 0.019 0.020
.0226--4 19.29 117.82 98,53 0.475 0.494 0.019 0.019
.0226--8 19.24 120,74 101.5 0.444 0.468 0.024 0.024
.0226--8 19.35 113.07 93,72 0.438 0.459 0.021 0.022
, .0226--8 19.27 116.86 97.59 0.438 0.462 0.024 0.025
· .0307--12 19.38 121.87 102.49
.0307--12 19.19 112.4 93.21 0.440 0.456 0.016 0.017
.0307--12 19.02 108.84 89.82 0.477 0.492 0.015 0.017
.0307--4 19.17 122.72 103.55 0.358 0.380 0.022 0.021
.0307--4 19.27 112,85 93.58 0.351 0.373 0.022 0.024
.0307--4 19.22 117.41 98.19 0.501 0.522 0.021 0.021
.0307--8 19.37 120.93 101.56 0.352 0.373 0.021 0.021
.0307--8 19.05 112.5 93.45 0.440 0.459 0.019 0.020
.0307--8 19.28 117. 74 98,46 0,352 0,373 0.021 0.021
.0833--12 19.13 118.77 99.64 0.465 0.489 0,024 0.024
.0833--12 19.24 119.27 100.03 0.498 0.528 0.03 0.030
.0833--12 19.21 118.8 99.59 0.506 0.529 0.023 0.023
.0833--4 19.27 113.99 94,72 0.441 0,454 0.013 0.014
.0833--4 19.23 111.29 92.06 0.479 0.493 0,014 0,015
.0833--4 19.46 123,49 104,03 0.343 0.360 0.017 0.016
.0833--8 19.32 112.4 93.08 0.442 0.467 0.025 0.027
.0833--8 19.46 116.45 96.99 0.439 0.463 0.024 0.025
.0833--8 19.17 108.03 88.86 0,200 0.223 0.023 0.026
BLANK 19.09 118.62 99.53 0,504 0.552 0.048 0.048
BLANK 19.07 112.83 93.76 0.516 0.562 0.046 0.049
BLANK 19.22 117.65 98.43 0,438 0.487 0.049 0,050
 APPENDIX 111 - HARDNESS TITRATION VALUES

SAMPLE WATER REAL TITE SAMPLE TITER wt. of CaC03 HBOX wt. of CaC03
D.S.--•/T0N TITER ml EDTA TITER ml EDTA ml EDTA m 1.EDTA/m I filtrate SAMPLE ml EDTA HBOX
.0226--12 24.0 2.4 21.6 0.864 0.086 43.7 0.165
.0226--12 23.6 2.4 21.2 0.848 0.085 43.7 0.165
.0226--12 23.9 2.4 21.5 0.86 0.086 43.7 0.165
.0226--4 23.1 2.4 20.7 0.828 0.083 43.7 0.165
.0226--4 23.8 2.4 21.4 0.856 0.086 43.7 0.165
1 .0226--4 24.0 2.4 21.6 0.864 0.086 43.7 0.165
.0226--8 23.1 2.4 20.7 0.828 0.083 43.7 0.165
.0226--8 23.2 2.4 20.8 0.832 0.083 43.7 0.165
.0226--8 23.0 2.4 20.6 0.824 0.082 43.7 0.165
.0307--12 0.165
.0307--12 23.0 2.4 20.6 0.824 0.082 43.7 0.165
.0307--12 23.0 2.4 20.6 0.824 0.082 43.7 0.165
.0307--4 22.1 2.4 19.7 0.788 0.079 43.7 0.165
.0307--4 22.3 2.4 19.9 0.796 0.080 43.7 0.165
.0307--4 22.9 2.4 20.5 0.82 0.082 43.7 0.16S
.0307--8 20.5 2.4 18.1 0.724 0.072 43.7 0.1�
.0307--8 20.3 2.4 17.9 0,716 0.072 43.7 0.165
.0307--8 20.0 2.4 17.6 0,704 0.070 43.7 0.165
.0833--12 24.5 2.4 22.1 0.884 0.088 43.7 0.165
.0833--12 24.6 2.4 22.2 0.888 0.089 43.7 0.165
.0833--12 24.0 2.4 21.6 0.864 0.086 43.7 0.165
0.896 0.090 43.7 ' 0.165
.0833--4 24.8 2.4 22.4
.0833--4 25.0 2.4 22.6 0.904 0.090 43.7 0.165
.0833--4 24.4 2.4 22 0.88 0.088 43.7 0.165
.0833--8 24.0 2.4 21.6 0.864 0.086 43.71 0.165
.0833--8 23.0 2.4 20.6 0.824 0.082 43.7 0.165
.0833--8 24.0 2.4 21.6 0.864 0.086 43.7 0.165
BLANK 28.0 2.4 25.6 1.024 0.102 43.7 0.165
BLANK 28.5 2.4 26.1 1.044 0.104 43.7 0.165
BLANK 27.7 2.4 25.3 1.012 0.101 43.7 0.165
 APPENDIX IV - RETENTION VALUES

Z SAMPLE Z HB0X Z F.P. AYO Z F. P. ZCoC03 AYO ZCoC03 Z FINES AYO Z FINES
D.S.--•/TON CONS. CONS . RETN. RETN. RETN. RETN. RETH. RETH.
.0226--12 0.133 0.49 72.9% 47.9% 28.6%
.0226--12 0.129 0.49 73.7% 73.5% 48.5% 48.1% 30.7% 30.2%
.0226--12 0.128 0.49 73.9% 0.4% 47.9% 0.3% 31.3% 1.2%
.0226--4 0.099 0.49 79.8% 49.7% 46.8%
.0226--4 0.106 0.49 78.4% 78.9% 47.9% 48.5% 43.1% 44.5%
1
►,0226--4 0.105 0.49 78.6% 0.6% 47.9% 0.8% 43.6% 1.6%
.0226--8 0.107 0.49 78.2% 49.7% 42.5%
.0226--8 0.105 0.49 78.6% 78.3% 49.7% 49.9% 43.6% 42.9%
.0226--8 0.107 0.49 78.2% 0.2% 50.3� 0.3% 42.5% 0.5%
'
.0307--12 0.000
.0307--12 0.099 0.49 79.8% 79.8% 50.3% 50.3% 46.8% 46.8%
.0307--12 0.099 0.49 79.8% 0.0% 50.3% 0.0% 46.8% 0.0%
.0307--4 0.100 0.49 79.6% 52.1% 46.3%
.0307--4 0.104 0.49 78.8% 79.1% 51.5% 51.3% 44.1% 45.0% .t,.
.0307--4 0.103 0.49 79.0% 0.3% 50.3% 0.7% 44.7% 0.9% U1
.0307--8 0.093 0.49 81.0% 56.4% 50.1%
.0307--8 0.092 0.49 81.2% 81.2% 56.4% 56.8% 50.6% 50.6%
.0307--8 0.091 0.49 81.4% 0.2% 57.6% 0.6% 51.1% 0.4%
.0833--12 0.112 0.49 77.1% 46.7% 39.8%
.0833--12 0.119 0.49 75.7% 76.9% 46.1% 46.9% 36.1% 39.1%
.0833--12 0.109 0.49 77.8% 0.9% 47.9% 0.7% ' 41.5% 2.3%
.0833--4 0.104 0.49 78.8% 45.5% 44.1%
.0833--4 0.105 0.49 78.6% 78.7% 45.5% 45.9% 43.6% 43.9%
.0833--4 0.104 0.49 78.8% 0.1% 46.7% 0.6% 44.1% 0.2%
.0833--8 0.113 0.49 76.9% 47.9� 39.3%
.0833--8 0.107 0.49 78.2% 77.4% 50.3% 48.7% 42.5% 40.5%
.0833--8 0.112 0.49 77.1% 0.6% 47.9% 1.1% 39.8% 1.4%
BLANK 0.150 0.49 69.4% 38.2% 19.4%
BLANK 0.153 0.49 68.8% 69.1% 37.0% 38.0% 17.8% 18.7%
BLANK 0.151 0.49 69.2% 0.2% 38.8% 0.7% 18.9% 0.7%