+ All Categories
Home > Documents > The effect of cationic polymers on colloidal stability

The effect of cationic polymers on colloidal stability

Date post: 06-Jul-2018
Category:
Upload: jollyjummy5852
View: 220 times
Download: 0 times
Share this document with a friend

of 10

Transcript
  • 8/17/2019 The effect of cationic polymers on colloidal stability

    1/10

    The Effect of Cationic Polymers on the Colloidal

    Stability of Latex Particles

    J O H N G R E G O R Y

    Department of Civil and Municipal Engineering University College

    Gower Street London WC 1E 6BT England

    Received June 20, 1975; accepted December 12, 1975

    Using a simple turbidimetri c procedure, relative flocculation rates of a polyst yrene latex suspension

    with three cationic polym ers and a cationic surfacta nt have been determined. All four materials

    showed abo ut the same optim um flocculation concentration when expressed in equivalen t amounts,

    indicating the predominance of charge effects. In agreem ent with previous findings, a considerable

    enhancement of flocculation rate was observed with cationic polymers of moderately high molecular

    weight, which is consistent with the electro static pat ch model of polyel ectrolyt e adsorption. A t

    opti mum poly mer concentration, it was found th at the rate of flocculation after a two-stage add ition

    of suspension was just the same as th at after th e conventional one-stage addition. This, at first sight,

    surprising result also can be explained in terms of a simple patchwise adsorption model. However,

    the rate enhancement is considerably underestimated by this simple model, probably because the

    mutual orientation of colliding particles is not taken into account.

    INTRODUCTION

    The de s ta b i l i z a t ion a nd re s ta b i l i z a t ion o f

    ne ga t ive pa r t i c le s by c a t ion ic po lyme rs oc c urs

    pr ima r i ly by c ha rge ne u t ra l i z a t ion a nd c ha rge

    re ve r sa l , a s de mons t ra te d by the f a c t tha t

    the e le c t rophore t ic m obi l i ty o f pa r t i c le s a t the

    op t imu m f loc c u la n t c onc e n t ra t ion i s c lose to

    zero (1-4) . S l ight ly less direc t , t houg h n o less

    s ignif icant, evid ence is the f inding th a t ca t ionic

    po lyme rs , d i f f e r ing on ly in mole c u la r we igh t ,

    h a v e t h e s a m e o p t i m u m c o n c e n t r a t i o n f o r t h e

    f loccula t ion of nega t ive la tex par t ic les (5 , 6) .

    In o the r words , op t imum f loc c u la t ion oc c urs

    whe n a c er ta in a m oun t o f pos i t ive c ha rge ha s

    be e n a dsorbe d by the pa r t i c le s , i r r e spe c t ive o f

    the l e ng th o f the po lyme r c ha ins c a r ry ing the

    charge . Fur ther evidence a long these l ines wil l

    be pre se n te d in th i s pa pe r .

    The impl ic a t ion o f the se r e su l t s i s tha t

    b r id g ing o f ne ga t ive pa r t i c le s by ca t ion ic

    1 Presented at the 49th Nati onal Colloid Symposium,

    Potsdam, New York, June 16-18, 1975.

    po lyme rs ma y no t be o f g re a t p ra c t ic a l s ign i f i -

    cance . Needless to say, th is conc lus ion has no

    be a r ing on the be ha v io r o f

    anionic

    p o l y m e r s

    with nega t ive par t ic les , where br idging f loccu-

    la t ion re ma ins by fa r the mos t p la us ib le

    me c ha n ism (7 ) .

    I t s e e ms l ike ly tha t , whe n a dsorbe d on

    ne ga t ive pa r t i c le s , po lyc a t ions f a i r ly r a p id ly

    a dop t a r a the r f l a t c onf igura t ion be c a use o f the

    s t rong ion ic in te ra c t ion be twe e n oppos i te ly

    c ha rge d g roups on the pa r t i c le su r fa c e a nd

    a long the po lyme r c ha in . Ka spe r (8 ) ha s

    sugge s te d tha t suc h a c onf igura t ion would

    ver y much reduce the l ike l ihood of br idging

    f loccula t ion. There a re , however , two respec ts

    in whic h the s imple c ha rge ne u t ra l i z a t ion

    p ic tu re ma y ne e d re f ine me nt .

    F i r s t ly , the e qu i l ib r ium c onf igura t ion o f a n

    a dsorbe d po lyme r c ha in i s a c h ie ve d a t some

    fini te t ime (perhaps severa l seconds) a f te r

    in i t i a l c on ta c t , a nd dur ing th i s t ime , the

    d is t r ibu t ion o f po ly me r s e gme nts f rom the

    par t ic le surface is more extens ive than in the

    Copyright ~ 1976 by Academic Press Inc.

    All rights of reproduction in any form reserved.

    35

    Journal of Colloid and Interface Science Vol. 55 No. 1 Apr il 1976

  • 8/17/2019 The effect of cationic polymers on colloidal stability

    2/10

    36

    JOHN GREGORY

    equil ibrium state, so that bridging may be

    possible. Fro m the stan dp oin t of flocculation,

    the crucial factor is the particle collision

    frequency (and hence, the part icle number

    concentrat ion). In a concen trated sol a part icle

    may experience many collisions during the

    t ime in which the adsorbed polym er achieves i ts

    equil ibrium configurat ion and bridging could

    then be a significant destabilizing mechanism.

    In more dilute sols the collision frequency

    would be too low for this effect to be appreci-

    able. Recent results (6) with latex particles

    and high molecular weight cat ionic polymers

    indicate tha t this nonequ il ibrium floccula-

    t ion becomes imp ortan t with part icle concen-

    tratio ns of a bou t 10 n cm 3 or greater.

    The second effect arises when particles of

    fair ly low negative surface charge densi ty are

    neutral ized by highly charged cat ionic poly-

    mers. In such cases, elementary geometric

    considerat ions show that the part icle charge

    cannot be neutral ized

    uniformly

    by the ad-

    sorbed polycat ions but rathe r t hat the part icle

    surface will have a mo saic pat te rn of

    posi t ively and negatively charged areas. The

    importance of this electrostat ic pa tch effect

    in f locculat ion was f irst recognized by Kaspe r

    (8) and i t was also invoked by Gregory (5)

    to account for the enhanced rate of floccu-

    lat ion observed with cat ionic polymers of

    moderately high molecular weight . In this

    paper further experimental evidence is pre-

    sented on the electrostat ic pa tch effect ,

    and by varying the mixing procedure, i t is

    shown th at there is a close analogy with

    the heteroflocculat ion of opposi tely charged

    particles.

    I t is worth point ing ou t tha t the concepts of

    patch wise adsorption and the f lat configura-

    t ion of adsorbed polycat ions imply that the

    part icles are larger than the polymer mole-

    cules. Clearly, with v ery high mo lecular weig ht

    polymers and very small colloidal particles,

    this assump tion would no longer be val id and i t

    is not difficult to envisage a situation like

    that depicted by I ler (9) in which several

    part icles are at ta ched to one polym er chain.

    E XPE RI ME NT AL ME T HODS

    Polystyrene suspension This was prepared

    by emulsion polymerizat ion, using sodium

    dodecyl sulfate (2.5 raM) as emulsifier and

    sodium persulfate (2.5 mM) as ini t iator . The

    styrene was distilled prior to use and repre-

    sented about 10% by weight of the react ion

    mixture (45 g total) . Poly merizat ion was

    carried out overnight in a sealed tube rotat ing

    in a wate r b ath, at 70 ° . The result ing latex

    was extensively dialyzed in a Dow hollow-fiber

    beaker dialyzer , to remove residual ini t iator

    and react ion products and to reduce the level

    of emulsifier (it is extrem ely difficult to rem ove

    all

    adsorbed surfacta nt by dialysis (10)) . After

    dilution of the purified latex to 100 cm ~, the

    polys tyrene content was found to be 3 .92%

    (w/v), by evaporat ing a small port ion to

    dryness, heat in g at 110 ° a nd weighing the

    residue.

    The p art icle s ize of the latex was determ ined

    by electron microscopy. A part icularly effective

    method of specimen preparat ion was to

    immerse a coated copper grid in a di lute

    solut ion of cat ionic polymer (0.02% Polymer

    2, see below), rinse in water and then immerse

    for a few seconds in a diluted (1/100) latex

    sample. The negatively charged latex part icles

    adhere strongly to the adsorbed cat ionic

    polymer layer on the grid and appear to

    deposi t fair ly uniformly. Samples prepared

    in this way are ready for examination in the

    electron microscope after a few minutes

    drying on f i l ter paper.

    The mean part icle diameter was found to be

    296 nm with a s tandard deviat ion of 34 nm.

    Cationic polymers

    Three polymers were

    used in this work:

    1. Poly (d imethylaminoethyl methacryl -

    ate) of nom inal molecu lar weig ht 5 X 10~,

    ful ly quaternized with dimethyl sulfate.

    2. As polymer 1 but with a molecular

    weight of about 1.5 >( 105.

    3. Poly (1-ethyl 2-methyl 5-vinyl py-

    ridinium bromide) with a molecular weight of

    about 106

    Journal of Colloid and Interface Science Vol. 55 No. 1 April 1976

  • 8/17/2019 The effect of cationic polymers on colloidal stability

    3/10

    CATIONIC POLYMERS AND LATEX STABILITY 37

    Polymers 1 and 2 were described in (5)

    (there designated I and III, respectively).

    The preparation of polymer 3 was given in

    (6), where it was labeled B.

    Stock solutions of 0.1 polymer were pre-

    pared and these were diluted to working

    solutions of 10 or 20 ;~g cm 3.

    Reagents. Cetyl trimethylammonium bro-

    mide (CTAB) was BDH laboratory reagent

    grade and sodium nitrate was of Analar

    quality. Both were used without further

    purification.

    All solutions and suspensions were prepared

    with water that had been passed through a

    mixed-bed demineralizing column and then

    distilled from alkaline permanganate solution.

    From the very short persistence of bubbles

    after shaking it was assumed that the water

    was substantially free from organic impurities.

    When working with very dilute solutions of

    cationic polymers (in the present case down

    to a few nanograms per cubic centimeter) it is

    essential to eliminate interference from traces

    of organic matter, especially surfactants.

    Otherwise, reproducible results cannot be

    obtained.

    Flocculation procedure. Suspensions of poly-

    styrene particles at a concentration of 13 ug

    cm 3 (9.19 X 108 partic les cm 3) were floccu-

    lated with a cationic surfactant (CTAB) and

    the three polymers mentioned above, in the

    presence of 10 4 M NaNO3. The flocculation

    was carried out in a 4-cm spectrophotometer

    cell and the increase in turbidity was followed

    as a function of time. The output from the

    spectrophotometer (Unicam SP500 Mk2) was

    fed via a Quaylab log-lin converter to a chart

    recorder. The slope of the trace was then

    proportional to the rate of turbidity increase,

    and will also be assumed proportional to the

    rate of flocculation.

    The spectrophotometer cell was first charged

    with 8 cnl ~ of a 1.5 10-4 M NaNO3 solution,

    containing the appropriate amount of floccu-

    lant. Four cubic centimeters of a 39 ug cm ~

    polystyrene suspension was then rapidly added

    to the cell from a dispensing pipet te (E-Mil)

    and the turbidity record was commenced

    within a few seconds of mixing. For each

    flocculant, the range of concentrations was

    chosen so that the onset of flocculation and

    restabilization could be clearly established.

    When the optimum concentration of each

    flocculant had been established (i.e., that

    giving the maximum rate of flocculation), a

    two-stage mixing procedure was carried out

    as follows : To 8 cm3 of a 1.5 10 4 M NaNO3

    solution containing the optinmm amount of

    flocculant, 2 cm 3 of the stock latex suspension

    was added, followed, after a measured time,

    by a second 2-cm~portion. The rate of turbidi ty

    increase was measured after the second

    addition.

    RESULTS AND DISCUSSION

    One Stage Addition

    The results are presented in Fig. 1, where

    the ordinate is the slope of the recorded trace

    divided by the initial reading, thus giving the

    0 . 1 C

    O O e

    ~'~ 06

    E

    rl,~

    .04

    0 . 0 2

    0 5

    - - O C T A B

    - - - - x - - P O L Y M E R 1

    + 2

    D~S, -H-- 3

    \

    x

    / _ l\

    10 15 20 25 30

    F L O C C U L A N T C O N C E N T R A T I O N ( g e q /g l

    F i C . 1 . R a t e o f t u r b i d i t y i n c r e a s e f o r l a t e x s u s p e n s i o n

    flocculated with a cationic surfactant (CTAB) and

    three cationic polymers, plotted as a function of

    flocculant concentration. All results obtained in 10 4 M

    NaNO3 solution. Dashed horizontal line indicates the

    corresponding rate for the same suspension flocculated

    with 0.3 M Ca(NOa)2.

    Journal of Colloid and Interface Science Vol. 55, No. 1 , Apri l 1976

  • 8/17/2019 The effect of cationic polymers on colloidal stability

    4/10

    38 JOHN GREGORY

    rate of turbidi ty increase relat ive to the

    original value r0, or 1/ro)dr/dt, where r is

    the turbidi ty at t ime t af ter mixing. As dis-

    cussed previously (5, 6), the initial slope (i.e.,

    at t = 0) is not a reliable indicato r of the

    optimum flocculant concentrat ion, s ince ad-

    sorpt ion and rearrangement of polymer chains

    may st i l l be taking place during the f irs t few

    seconds after mixing, and this can lead to

    anomalous values of the rate. This effect is

    most pronounced wi th polymers of h igh

    molecular weight at concentrat ions above

    optimu m. Consequently, the slopes of the l ines

    have been taken over the interval between one

    and two minutes after mixing. The abscissa in

    Fig. 1 is the amount of flocculant, expressed

    in microequivalents , per gram of polystyrene

    part icles. The equivalent weight of the poly-

    mers is taken as the formula weight per

    quaterna ry n it rogen a tom. For polymers 1

    and 2 this is 283 and for polymer 3 the value

    is 228 . Th e dash ed horizo ntal line in Fig. 1

    represents the rate of turbidi ty increase when

    the particles are flocculated with 0.3 M

    calcium nitrate solut ion.

    I t is clear from Fig. 1 that the f locculat ion

    effects of a s imple cat ionic surfactant CTAB,

    and a low molecular weight cat ionic polymer

    1, are not great ly different . The maximum

    rate is achieved with essential ly the same

    amount of adsorbed posi t ive charge, 14

    /~equiv g-l, or a bou t 7 ~C cm 2. It is reasonable

    to assume from the work of Connor and

    Ottewil l (11) that quanti tat ive adsorption of

    C T A+ ions occurs on a polysty rene surface, at

    least up to the point of charge neutral izat ion.

    This wil l certainly be the case for polycat ions.

    Although, there is some scat ter in the resul ts ,

    the breadth of the f locculat ion region appears

    to be about the same in the two cases. I t

    seems very like ly tha t for C TAB and Polymer

    1, flocculation and restabilization are simply

    due to charge neutral izat ion and charge

    reversal by strongly adsorbed cat ionic ma-

    terials. Since the ionic strength of the solution

    is low, very small amounts of negative or

    positive charge are sufficient to stabilize the

    particles.

    With the higher molecular weight polymers

    2 and'3 , f locculat ion occurs over a som ewha t

    broader range , a l though the maximum ra te i s

    achieved with about the same equivalent

    amounts, so that the primary effect is s t i l l

    one of charge neutralization. It is difficult to

    accoun t for the broader f locculat ion regions by

    a bridging effect, since the considerable differ-

    ence in molecular weight between polymers 2

    and 3 appears to have no fur ther infuence .

    I t is just conceivable that , because of the

    strong interact ion, loop dimensions of the

    adsorbed polymer chains are independent of

    molecular weight above a certain value, but

    are still sufficiently large to prom ote b ridging

    between part icles. I t is more l ikely that the

    observed effects with increasing molecular

    weight can be explained by the electrostat ic

    patch model discussed earl ier . If the adsorbed

    positive patches are of sufficient size, en-

    counters with negative areas on othe r part icles

    will be sufficiently favorab le to cause atta ch-

    ment, even though the overal l charge on the

    part icles, i f uniformly distr ibuted, would st i l l

    be high enough to prevent f locculat ion.

    Fu rth er increase in the size of the pa tche s (i.e.,

    increasing molecular weight of adsorbed

    polycat ion) would necessari ly reduce the

    number of pa tches per par t ic le and tend to

    keep the chance of favorable contacts with

    other part icles about constant . Similar argu-

    ments apply to both the destabil izat ion and

    restabilization regions and so explain the sym-

    metrical broadening of the flocculation zone.

    Turning to the maximum rates of f loccu-

    lat ion observed, i t is appar ent tha t appreciably

    faster f locculat ion can be achieved with

    polymers of mod erately high molecular weight

    over that found by the addit ion of s imple

    salts. In the presence of 0.3 M Ca(NO3)2,

    double lay er repulsion is complete ly suppressed

    and the f locculat ion rate observed should be

    the rate of Brownian collisions between

    par tic les (modif ied by the hydr odyn am ic

    effect, see below). Th e results in Fig. 1 show

    that the ra te obta ined wi th polymers 2 and 3

    can be about twice this value, as reported

    previously for other suspensions_ (5, 6), Again,

    Jour nal of Colloid and Interface Science Vol. 55 No. 1 April 1976

  • 8/17/2019 The effect of cationic polymers on colloidal stability

    5/10

    CATIONIC POLYMERS AND LATEX STABILITY 39

    polymers 2 and 3 behave a lmost i dent i ca l ly

    in this respec t , in spite of sixfold differenc e in

    molecular weight . A patchwise adsorpt ion of

    polymer could probably expla in th i s l imi t ed

    rate enhancement , as shown in a later sect ion.

    The maxim um ra t e of f l occula tion found

    wi th polymer 1 i s about 50% higher t han tha t

    with salt and this is more difficult to explain

    by a patch model . I t i s not easy to see why

    posi t ive patches large enough to enhance the

    rate of f locculat ion appreciably, should not

    also cause a more substant ial broadening of

    the f locculat ion zone. However, even adsorbed

    CT A + ions app ear to enh ance the f locculat ion

    rate sl ight ly and in this case there should be

    no possibi l i ty of patchwise adsorpt ion, so that

    o ther fac tors , as ye t uncer t a in , m ay hav e to be

    t aken in to account .

    Two-Stage Addition

    By choos ing the opt imum polymer concen-

    t rat ion, b ut only adding hal f of the suspension,

    each part icle receives, on average,

    twice

    t he

    opt imum amount of f l oeculan t . I t can be seen

    from Fig. 1 that , for al l of the polymers

    employed, this would be sufficient to com-

    pletely restabi l ize the part icles (because of

    thei r excess posi t ive charge), and so essent ial ly

    no flocculat ion should occur after the addi t ion

    of the f i rst hal f of the suspension. When the

    second hal f of the

    will be l i t t le or no

    solut ion and i t i s

    suspension i s added there

    polymer remain ing in f ree

    safe to assume tha t t he

    adsorbed polymer on the f i rst part icles wi l l

    no t desorb to any apprec i ab le ex ten t . Thus ,

    the part icles in the second port ion wi l l remain

    negat ive ly charged and the two-s t age addi t ion

    procedure should produce a suspens ion con-

    t a in ing equal numbe rs of pos i ti ve and n egat ive

    part icles. In fact , this cond i t ion can be veri fied

    by di rect observat ion of such a suspension in

    microelect rophoresis cel l (Sheiham, unpub-

    l i shed observat ions, 1973). In the mixed

    suspension, the only coll isions result ing in the

    format ion of aggregates should be those

    between opposi tely charged part icles, which

    represent hal f the total number of col l i s ions.

    From th is s imple p i c ture i t might be pred ic t ed

    tha t t he ra t e of f locculat ion found af t e r a

    two-stage addi t ion of suspension would be

    just hal f that observed fol lowing the one-stage

    addi t ion procedure . Resul t s f ro m the two-s tage

    addi t ion exper iment s do not bear ou t t h i s

    predict ion, bu t do lead to a possible explana t ion

    of the effect of patchwise adsorpt ion on

    flocculat ion rates general ly.

    The fo rm of the tu rb id i ty versus t ime t races

    for one-stage and two-stage addi t ions i s

    shown schemat ical ly in Fig. 2 for the case

    of op t imum polymer concent ra t ion . Af t er t he

    one-stage addi t ion of the suspension at point

    A, the turbidi ty at tains i t s ini t ial value r0,

    and then ri ses at a rate that depends on the

    rate of f locculat ion under opt imum condi t ions.

    In a two-stage addi t ion, the f i rst hal f of the

    suspension i s added at point A and the turbid-

    i ty becomes

    ro/2.

    Apar t f rom a s l igh t non-

    equi l ibrium flocculat ion in the case of fai r ly

    high molecu lar weight polym ers (e.g. , 2 and 3),

    no change in turbidi ty occurs, as expected.

    The second por t ion i s added a t po in t B ,

    whereupo n the turb id i ty r ises to abou t to and

    then begins to increase at a measurable rate.

    Al though, the reproducibi l i ty of the resul ts

    i s not ent i rely sat i sfactory, sufficient experi -

    ment s have been conducted , us ing var ious

    >~

    £ ~o

    1

    STAGE

    /

    ¢

    ,,

    i

    . _ . . . . . . . . . . . J

    i

    S t

    t' ¢

    A B

    TIME

    2 S T A G E

    /

    /

    / /

    / /

    /

    /

    /

    FIG. 2. Schema tic llustration of the turbidity change

    with time after a one-stage (at A) or a two-stage (at

    A and B) addition of latex suspension to a solution

    containing the op timu m concentration of cationic

    polymer.

    Journal of Colloid and Interface ~¢ien¢¢ V o l . 5 5 , N o . 1 , A p r i l 1 9 7 6

  • 8/17/2019 The effect of cationic polymers on colloidal stability

    6/10

    40 JOHN GREGORY

    latex part icles and polymers, under di fferent

    condi t ions and by two di fferent experimental

    techniques, to make the fol lowing conclusions

    with some confidence:

    a) The t ime interval between the addi-

    t ions of the two port ions of suspension At has

    no infuence on the f loccula t ion ra t e a f t e r t he

    second addi t ion. V alues of At hav e bee n

    varied from a few seconds to many hours, wi th

    no change in the f locculat ion rate.

    b) Th e rate of f locculation after the

    addi t ion of the second hal f of the suspension

    is the

    same

    as tha t observed af t e r a one-s t age

    addi t ion of suspension, rather than hal f of

    the o ne-stage rate as exp ected on the basis of a

    simple heteroflocculat ion picture.

    Some typical resul ts for polymers 1 and 3

    are shown in Fig. 3, in which the po ints shown

    for At = 0 are those corresponding to a one-

    stage addi t ion of suspension.

    In the next sec t ion an a t t e mp t wi ll be made

    to account for these f indings in terms of the

    elect rostat ic patch model .

    Patchwise Adsorption and Flocculation Kinetic s

    In the case of one-stage addi t ion of suspen-

    sion, assume tha t a frac t ion f of the to tal

    surface area of the part icles i s covered by

    adsorbed ca tionic polymer , and hence , 1 - - f )

    i s the fract ion of negat ively charged surface.

    I t i s then possible to envisage three dist inct

    types of col l i sion, each characterized by a

    certain stabi l i ty rat io W as in the Fuchs

    theo ry of slow flocculat ion 12):

    a ) be tween negat ive a reas Wn ;

    b) between po si t ive areas W~2;

    c) between negat ive and posi t ive areas

    }V12.

    If i t i s further assumed that the relat ive

    frequency of each type of col l i sion is propor-

    t iona l t o the product of the appropr i a t e a rea

    fract ions, then the overal l s tabi l i ty rat io for

    the suspension Wt, can be wri t ten, fol lowing

    Hogg, Hea ly and Fuers t enau 13) :

    1 I 1 - - f ) 2 f f 2 f 1 - - f ) 1

    w , . E13

    0 1 0

    0 0 ~

    d

    ¢

    I ¢

    0 0 ~

    0 02

    []

    m

    POLYMER 3

    POLYMER 1

    X

    J

    At rnin)

    Fro. 3. Rate of turbidity increase after a two-stage

    addition of suspension, as a function of

    At

    the time

    interval between the additions of the two portions.

    The values at At = 0 are those for a one-stage addition

    of suspension.

    Journal of Colloid and Interface Science Vol. 55. No. 1. April 1976

    In the case of two-stage addi t ion, hal f of the

    part icles receive al l of the polymer so that a

    fract ion 2f of thei r surface consists of posi t ive

    patches. I t i s conv enient to assume f < ,

    i .e . , that there i s enough part icle surface to

    accom m oda t e t w i ce t he op t i m um am oun t o f

    polymer, al though exact ly the same resul t i s

    obtaine d i f f > ). A fract ion 1 -- 2f) of the

    surface of the f i rst part icles remains n egat iv ely

    charged. When the second hal f of the suspen-

    sion is added, al l of i t s part icle surface area

    remains negat ively charged since no cat ionic

    polymer i s avai lable for adsorpt ion. Assuming

    the same values of stabi l i ty rat ios

    Wn W22

    an d W12, and considerin g all six possible kind s

    of coll ision, the following expression for the

    overal l s tabi l i ty rat io, Wt , can be derive d:

    4 f 1 2 f ) q - - - + . [ 2 ]

    + W12 W n

    1 _ 1V 1- 2/ ) 2+ 4ff 1

    + - -

    W Wn

  • 8/17/2019 The effect of cationic polymers on colloidal stability

    7/10

    CATIONIC POLYMERS AND LATEX STABILITY 41

    Rearrangement of this expression gives:

    1 F 1 _ f ) 2 f 2

    2f(1 -- f)-] 1

    j - E3-]

    i .e . , exact ly the same stabi l i ty rat io as in the

    case of one-stage addi t ion. Hence, from the

    patch model , there i s good reason to expect

    tha t f l occula tion ra t e should be ind ependent

    of the mixing procedure.

    In solut ions of low ionic st rength, the inter-

    act ion between areas of l ike charge wi l l be

    high ly repulsive, i .e. , Wn and W22 will be

    very large. Conversely, because of the at t rac-

    t ion betw een are as of unlike charge, W12 < 1,

    i .e. , such coll isions should occur more fre-

    quen t l y t han expec t ed f rom t he B row n i an

    col l i s ion rate. Consequent ly, the overal l s ta-

    bi l i ty rat io

    Wt,

    should be de t ermined en t i re ly

    by the th i rd t e rm on the r igh t -hand s ide of

    Eq. [ -17 and hence , by the va lues o f f and W12

    Unfor tunate ly , ne i ther of t hese va lues can be

    est imated wi th much confidence.

    The f rac t iona l pa t ch area wi l l depend on

    the nature of the cat ionic polymer, especial ly

    on the charge den si ty and the f lexibi li ty of

    the cha in . However , t he va lue of t he t e rm

    2f (1 - f ) is no t very sens it ive to the va lue

    of f , e .g., over the rang e 0.2 < f < 0.8 this

    term varies between 0.32 and 0.5, wi th the

    ma ximu m value at f = 0.5. For a rou gh

    estimate of Wt i t wil l be sufficient to assume

    f= 0 .3 .

    The value of W12 requires rather more

    careful considerat ion.

    Calculation of [/V12

    In principle the stabi l i ty rat io W, for a

    suspension of spherical part icles can be calcu-

    l a t ed i f t he to t a l i n t e rac t ion energy be tween

    two spheres Vt , i s known as a funct ion of the

    distance of separat ion d.

    If the spheres are of equal radius a, the

    t reatment of Fuchs (12) leads to:

    f0 ~ exp

    V~/kT)

    W = 2 2 + u 2 du, [-4-1

    where k i s Bol tzmann's constant , T i s the

    absolu t e t empera ture , and

    u = d/a.

    Usual ly,

    the Fuchs method is appl ied to slowly floccu-

    lat ing systems, where

    Vt

    i s posi t ive at most

    separat ion distances and W > 1. If there i s

    no interact ion between the part icles

    Vt = 0),

    Eq . [4-] gives W = 1 and then , by definit ion,

    the f locculat ion rate i s governed ent i rely by

    the Brownian col l i s ion frequency of the

    particles.

    It is now recognized (14, 15) that Eq. [-47

    needs to be modified to take into account a

    hyd rody nam ic, or viscous, effect which tends

    to reduce the rate of f locculat ion. The effect

    arises essent ial ly from the need to squeez e

    ou t the l iquid between appro aching spheres,

    and this becomes increasingly di ff icul t as the

    gap narrows.

    A useful empirical approximat ion was given

    by Honig , Roebersen and Wiersema (15) ,

    which adequa te ly represent s t he hydrod yna mic

    effect and leads to the fol lowing, expression for

    the stabil i ty rat io, replacing Eq. [-4-]

    ~ (6u 2 + 13u + 2) exp

    V t / k T )

    W = 2 du.

    0 (6u 2 + 4u)( 2 + u) 2

    ES-]

    If Vt = 0, Eq . [-5-] gives W = ~ , i.e., no

    flocculat ion. Only i f there i s some at t ract ion

    betw een the pa rt icles (i .e. , Vt < 0 at close

    approach) can flocculat ion take place.

    Ignoring the possibi l i ty of ster ic effects ,

    t here a re two impor t an t cont r ibu t ions to the

    in t erac t ion energy , t he van der Waal s and

    electrical te rms :

    V, = V~dw + VE. [-6-]

    The v an der Waal s a t t rac t ion be tween equal

    spheres i s given by the Hamaker expression

    (16) :

    V v d w ~ - - - -

    2 2

    6 2 + 4u (u + 2) 2

    u 2 + 4u

    + In u + , [-7~

    where A i s t he Hamaker cons t an t , which , for

    Jour na l o f Co llo id and In terface S¢ ienc¢ Vol. 55 No. 1 A pril 1976

  • 8/17/2019 The effect of cationic polymers on colloidal stability

    8/10

    42 JOHN GREGORY

    the present case of polystyrene in water , wil l

    be as sum ed to be 5 X 10 21 J (17).

    Equat ion [ -7] does not take re ta rda t ion

    effects into account, nor have possible short-

    comings of the simple Ham aker approach (18)

    been considered. However, for the present

    case where the electrical term in Eq. [-6] is

    also at tract ive, the van der Waals at t ract ion

    is quite a minor contr ibution to Vt, so that no

    serious errors will arise by the use of Eq. [7].

    The VE term, presents some difficulties, for

    instance in the choice of surface potentials for

    the positive and negative regions. There is also

    the quest ion of whether the interact ion occurs

    at constant potential or constant charge (19),

    which is of crucial importance in the case of

    unlike surfaces, even the sign of VE can be

    different depending on the condit ion assumed

    (20). A nother problem is whethe r the posi t ive

    patch es are su fficiently large to justify the use

    of expressions based on the interaction of

    spheres, but s ince such an assumption is

    implici t in the previous derivat ion of the

    stability ratio Wt, it will not be discussed

    further .

    I t has been shown (21 ) tha t expressions

    based on the linear superposition approxi-

    mation (LSA) give resul ts intermediate be-

    tween those for the constant charge and

    consta nt potential condit ions and, s ince nei ther

    of these extreme conditions is likely to be

    appropriate in pract ical s i tuat ions, the LSA

    approach might be a reasonable one to adopt .

    A convenient expression for spherical particles

    of equal s ize but unequal potentials in a

    symmetrical (z-z) electrolyte solution is (22):

    128rrnk T a

    v~

    3 ,3 2 - - e ~p ( -- Kd ), l- S]

    K2 u q- 2)

    where n is the number of cations (or anions)

    per cubic meter , K is the Deby e-Hiicke l re-

    ciprocal length parameter (~2= 2e~nz2/~kT,

    where e is the electron charge and ~ the per-

    m itt iv ity of the med ium). 3'1 and 3 2 are fu nc-

    tions of the surface po ten tial s ~bl an d ~k2 of th e

    negative and posi t ive patches, respect ively:

    3 1 = tanh(ze~bl/4kT) etc.

    Str ict ly, Eq. I-8] only applies whe n the

    diffuse pa rt of the electrical double layer is

    fair ly thin compared to the radius of the

    particles (Ka > 10). For the latex particles used

    in the present work, a~---150 nm and, in

    10 4 M NaNO3, Ka is onl y abou t 5, but the

    errors involved in using Eq. [-8] are still no t

    serious.

    The appropriate values of ~bl and ~b2 ~tre

    not known, but real is t ic est imates can be

    made as follows. From electrokinetic studies

    on dialyzed latex suspensions (23), zeta po-

    tent ials in di lute electrolyte solut ions are

    often about --60 inV. After charge reversal

    by cat ionic surfactants (11) or polymers (2)

    the zeta potential reaches a maxim um value in

    the region +30 to 60 mV, al though local ized

    areas of higher potential probably exist under

    these conditions. Bearing these figures in mind ,

    the reduced potential ,

    y l (= ze~ l / kT) ,

    of the

    nega tive areas has been take n as --2 (i.e.,

    ~bl = -- 51 mV ) an d y2, for the positiv e pat ch es,

    has been given va lues of q- l , +2 , and + 4

    (i.e., ~b2 va ry ing betw een 26 and 102 mV ).

    After computing values of Vvaw and VE

    from Eqs. [7] and [.8], W12 has been obtai ned

    from Eq. 1-5] by a num erical integrat ion

    procedure. Results are shown in Fig. 4 for

    concentrat ions of 1- 1 electrolyte between

    10 5 an d 10 2 M, assum ing the radius of the

    part icles is 150 nm and the Hamaker constant

    is 5 X 10 21 J.

    Using Eq. [1] and assumingf = 0.3, values

    of W t may be calculated and these can be

    compared with the stabi l i ty rat io expected for

    the particles in the absence of electrical

    repulsion (e.g., in high salt concentration),

    W t°. By pu t t ing V~ = 0 and car ry ing out

    the numerical integrat ion procedure, W? is

    found to be 1.784, i .e., the flocculation rate

    should be about 56% of the Brownian coll is ion

    rate, in agreement with previous computat ions

    (15) of the h ydro dyn am ic effect .

    The ra t io W t ° / W , should then give the factor

    by which the flocculation rate is increased as a

    result of patchwise polymer adsorption. This

    ratio is plotted in Fig. 5 as a function of

    electrolyte concen trat ion and for various values

    Journal of Colloid and Interface Science Vol . 55 No. 1 Apri l 1976

  • 8/17/2019 The effect of cationic polymers on colloidal stability

    9/10

    CATIONIC POLYMERS AND LATEX STABILITY 42

    of y2 . I t i s a ppa re n t tha t the r e su l t de pe nds

    ve ry s t rong ly on the ion ic s t r e ng th bu t ve ry

    little on the value of y2.

    A s t rong de pe nde nc e o f ma ximum f loc c u-

    la t ion ra te on the ion ic s t r e ng th o f the so lu t ion

    ha s be e n re por te d p re v ious ly (5 ) a nd the

    resul ts in Fig. 5 shows a s imila r t rend. How-

    e ve r , the c om pute d re su lt s d i f fe r ma r ke d l y

    f rom e xpe r ime nta l f ind ings . Ac c ord ing to the

    resul ts in Fig. 5 pa tchwise adsorpt ion of

    c a t ion ic po lyme r shou ld

    decrease

    the ra te of

    f loccula t ion i f the sa l t concentra t ion is grea te r

    tha n a bou t 10 4 M, wh e re a s e xpe r im e nta l

    ra te s ha ve be e n shown to be e nha nc e d up to

    abo ut 10 2 M sa l t (5). The pres ent resu l ts in

    10 4 M NaN O3 show tha t th e f loccula t ion ra te

    c a n be ne a r ly doub le d wi th c a t ion ic po lyme rs ,

    whereas Fig. 5 indica tes a s l ight decrease in

    ra te a t th i s s a l t c onc e n t ra t ion .

    The c h ie f d i f f i c u l ty wi th the p re se n t mode l

    i s tha t a l thoug h the f a vora b le in te ra c t ion

    be twe e n pos i t ive a nd ne ga t ive a re a s l e a ds to a

    decrease in th e ca l cula ted va lue of W12, as

    20

    1

    11

    (35

    \\ \

    \ \ \ \

    - 4 - 3 - 2

    LOGloSALT CONCN ( IM)

    1 1o.

    5 . Th e r a t i o

    W?/Wt

    see t ex t ) as a func t ion o

    e l e c t r o l y te c o n c e n t r a t i o n . Co n d i t i o n s a s f or F i g . 4 .

    1 0

    W12

    O.

    / / /

    / / /

    /

    / /

    - 5 - 4 - 3 - 2

    L O G l o S A LT C O N C N ( M )

    FIG. 4. Calculated values of the stab ility rat io W ~,

    for the interacti on between o ppositel y charged spherical

    surfaces, as a function of the concentration of 1-1

    electrolyte. The reduced potential of the negative

    surface yt, is --2 and the numbers on the curves denote

    the values of y~, for t he po sitive surface.

    required, th is is usua l ly outweighed by th~

    small frac tio n of such collisions. I f f = 0.3

    on ly 42 o f c o ll i sions a re be twe e n pos i t iw

    and nega t ive a reas , according to Eq. 1-1]

    Thu s , the e ffec t ive s tabi l i ty ra t io W12, woulc

    have to be reduced by a fac tor of about 2 . . :

    to show a n ove ra l l r a te e nha nc e me nt . I t i ,

    very l ike ly tha t col l is ions a re not ent i re l)

    rand om, as assum ed in Eq. I -1] , but tha t ther~

    is a mu tua l o r ienta t ion e ffec t be tw een ap-

    proa c h ing pa r t i c le s , so tha t pos i t ive a nd ne ga -

    t ive a reas a re more l ike ly to col l ide than th~

    s imple f a c to r 2 f (1 - f ) would sugge s t . How.

    ever , a qua nt i ta t i ve t re a tm ent of this e ffec l

    would present cons iderable dif f icul t ies .

    REFERENCES

    1.

    TEOT,

    A. S.,

    Ann. N. Y. Acad. Sc i .

    155, 593 (1969)

    2. GREGORY, ., Trans. Faraday Soc. 65, 2260 (1969)

    3. DIx oN, J. K. AND ZIELYK, M. W.,

    Environ. Sci

    Technol.

    3, 209 (1969).

    4. SANDELL,L. S. AND LUNER,

    P. J . Appl . Polym. Sc i

    18, 2075 (1974).

    5. GREGORY, J. J. Colloid Interfa ce Sci. 42, 441

    (1973).

    Journ al of Colloid and Interface Science Vol.55,No. 1, April197~

  • 8/17/2019 The effect of cationic polymers on colloidal stability

    10/10

    4 4 J O H N G RE G O RY

    6. GREGORY, J. , AND SHEIHAM, I. Brit. Polym. Y. 6

    47 (1974).

    7. MICHAELS,A. S.,

    Ind. Eng. Chem. 46

    1485 (1954).

    8. RASPER, D. R., P h.D. Thesis, California Institu te

    of Technolog y, 1971.

    9. ILER,

    R. K. J. Colloid Interface Sc i.

    37, 364 (1971).

    10. VAN DEN HU L, H. J. AND VANDERHOFF J. W.,

    J. Colloid Interface Sci. 28, 336 (1968).

    11. CONNOR, P. AND OTTEWILL, R. H. J. Colloid

    Interface Sci.

    37, 642 (1972).

    12. OVERBEEK, J. TH. G.

    in

    Colloid Science, Vol. 1,

    p. 283. (H. R. K ruyt , Ed.), Elsevier, Amsterdam ,

    1952.

    13. HOG G, R. , HEA LY, T . W. , AND FUERSTENAU,

    D. W. Trans. Faraday Soc.

    62, 1638 (1966).

    14. SPIELMAN,L. A. J. Colloid Interface Sc i. 33, 562

    (1970).

    15. HON IG, E. P. , ROEBERSEN, G. J., AND WIERSEMA,

    P. H. J. Colloid Interface Sci.

    36, 97 (1971).

    16. HAMAKER,H. C., Physica 4 1058 (1937).

    17. GREGORY, J.,

    Adv. Collo id Interface Sci.

    2, 396

    (1969).

    18. PARSEGIAN,V. A. AND NINHA~I,

    B. W. ]. Colloid

    Interface Sci.

    37, 332 (1971).

    19. FR ENS, G. ANn OVERBEEK J. Th. G., ] .

    Colloid

    Interface Sci.

    38, 376 (1972).

    20. BELL , G. M. AND PETERSON,

    G. C. J. Colloid

    Interface Sci. 41, 542 (1972).

    21. GREGORY,

    . J. C olloid Interface Sci.

    51, 44 (1975),

    22. BELL , G. M. , L EVm E, S., AND MCCARTNEY,L. N..

    J. Colloid Interface Sci.

    33, 335 (1970).

    23. OT'rEWILL,R. H. AND SHAW,J. N. J. Electroanal.

    Chem. 37, 133 (1972).

    Journa l of Colloid and Interface Science. Vol. 55. No. 1, Apr il 1976


Recommended