Effects of Inorganic and Organic Additives on the Adsorption ofNonionic Polyacrylamide on Hematite

L. T. LEE I AND P. SOMASUNDARAN2

Columbia University. New York. New York 10027

Received October 18. 1989; accepted September 19. 1990

The effects of inorganic anions such as chloride, sulfate, and phosphate on the adsorption of non ionicpolyacrylamide on hematite are studied. Adsorption of the polymer is reduced in all cases, with completedepression by phosphate. The relative eff«tiveness of these anions in reducing polymer adsorption in~in the order chloride < sulfate < phosphate. a trend which corresponds to the relative stabilities of theanion-Fe)+ complex formation. The eff«tiveness of sulfate and phosphate in competing with the polymerfor surface sites is explained in terms of the strong interaction of these anions with FeOH~ and FeOHby ligand exchange, replacing the surface hydroxyl groups. Since at lower pH there is higher interactionof these anions with the surface, the effect of pH on polymer adsorption is reversed. Urea is also foundto decrease polymer adsorption. In this case, the phenomenon is due to H-bonding of the urea with thesurface hydroxyls, depleting adsorption sites for the polymer. c 1991 Academic PJess, Inc.

INTRODUCfION

The adsorption of water-soluble polymerson hydrophilic solid surfaces is predominantlygoverned by electrostatic interactions, H-bonding, and solvation forces. For chargedpolymers, electrostatic interaction is significantand parameters like pH and ionic strength ofthe solution become determining factors ofadsorption. If the surface and the polymercarry opposite charges, both electrostatic andH-bonding forces favor the adsorption process.If, however, they carry similar charges, thenadsorption is determined by a competition ofattractive H-bonding force and repulsive elec-trostatic interaction. For uncharged polymers,only H-bonding (1-8) and solvation forces,the latter of which always act against adsorp-tion, are considered to be important. In spiteof that, adsorption of nonionic polymers onoxide minerals has been found to be pH de-pendent (5,8). This has been studied system-

atically for adsorption of polyacrylamide ondifferent oxides and the pH dependence ob-served has been attributed to the site depen-dence ofH-bonding (8). Although H-bondingis not considered to be an electrostatic typeinteraction, the functional group of the poly-mer ( C = 0 ) which is engaged in the bondingmechanism does carry a net electronegativity,and interaction of it with surface sites woulddepend on the relative electropositivity of thelatter, which have been proposed to be theMOH! and MOH groups. Such a dependencethen manifests itself in pH dependence of H-bonding.

In this study, the effects of additives whichcan adsorb onto and compete with the poly-mer for these adsorption sites are investigated.The competing agents chosen for this purposeare negative inorganic anions varying incharges and affinities for the oxide surface suchas chloride, sulfate, and phosphate. The influ-ence of pH on polymer adsorption in the pres-ence of these anions is also studied. The effectof an organic compound, urea, is also inves-tigated in view of its proposed character as aH-bond breaker.

I Present address: laboratorie Leon Brillouin, CEN.

Saclay, 91191 Gif-sur-Yvette Cedex, France.2 To whom correspondence should be addressed

4700021-9797/91 $3.00~I 0 1991 by A£ademic PI-. !DC.AU riIIIts of ~ in any form I. -'-Mi O/CoI1oitI8Id1.-I-~. va 142. No. 2. M8rd1IS, 1991

471AI:s:)RPTJON OF POLYACRYLAMIDE

EXPERIMENTAL

The hematite used is a synthetic oxide fromAlfa Chemicals. It is in powder form with aspecific surface area of 8.4 m21 g as measuredby BET technique using N2 as adsorbate. Thepoint of zero charge is measured to be 1.5 usingthe zeta-meter.

The nonionic polyacrylamide is synthesizedusing radiation induced precipitation methodwith a ~o source. It is labeled with 14C asradio-tracer. From viscometry, the averagemolecular weight is determined to be 0.1 mil-lion Da. By adsorbing the polymer onto amineral support and measuring the zeta p0-tential of the polymer covered mineral, it isverified that the polymer is nonionic in na-ture (8).

The experimental procedure consists of apreconditioning step and the conditioning ofthe mineral with polymer. In preconditioning,the mineral is added to the solvent containingthe necessary salts and adjusted to the requiredpH. This is a wetting stage to ensure complete

He"-~e

>-I-inz~z0i=

~0U)QC

0 100 200 300 400 500 600 700 800RESIDUAL POLYMER. mq/kq

fK;. I. Effects of chloride. sulfate. and phosphate on PAM adSOrption on FeA at pH 6.6.

1--' t6"~ -,~ ~ Vol 142, No. 2, M8dt I'. 1991

dispemon of the fine mineral particles. Poly-mer solution adjusted to the same final con-ditions of the solvent is then added. The pre-conditioning time, determined from the finalconstant pH value, and the polymer adsorp-tion time, determined from kinetic studies, are3 and 4 h, respectively. Mixing in both casesis facilitated by a mechanical wrist-actionshaker. In all cases, the floal solid/liquidweight ratio of the suspension is 0.03. Theamount of polymer adsorbed is determinedby the depletion method. The polymer con-centrations before and after adsorption aredetermined from the radiations of the 14C la-beled polymer samples using a scintillationcounter.

RESULTS AND DISCUSSIONS

Effect of Anions

The effects of chloride, sulfate, and phos-phate on polyacrylamide adsorption on Fe203

472 LEE AND SOMASUNDARAN

at pH 6.6 are shown in Fig. I. The concentra-tions of these anions, with sodium as thecounterion, are 3 kmol/m3, I klnol/m3, and0.3 klnol / m3 so that comparison of these ionsis made at a constant ionic strength. At pH6.6 where the Fe203 has a net positive charge,the presence of anions reduces polymer ad-sorption significantly in the order chloride< sulfate < phosphate, with complete depres-sion of the PAM adsorption by phosphate.This is due to interactions of the anions withadsorption sites on the Fe203 thus reducingthe number of adsorption sites available forthe polymer.

The order by which the anions depress ad-sorption of PAM on Fe203 follows the se-quence of affinities of the anions for Fe203,namely, chloride < sulfate < phosphate. Asexpected, chloride, being monovalent, has alower affinity for Fe203 than its higher valencecounterparts, sulfate and phosphate. However,it should be noted that even though electro-static interaction contributes to the anion ad-sorption on Fe203, it is not the sole factor.Phosphoric acid has three dissociation con-stants and at pH 6.6, even though 80% of thetotal phosphate is in monovalent formH2PO;, and 20% is in the bivalent formHPO~-, phosphate interacts more stronglywith Fe203 than sulfate, 100% of which is inthe bivalent fonn SO~-. Even though the ad-sorption of both sulfate and phosphate basbeen shown to be of "high affinity" type (9,10), the higher (9 ) and stronger ( II ) adsorp-tion of phosphate over sulfate has in fact beenshown elsewhere. In addition, it has also beenshown that phosphate can be used as an agentto displace adsorbed sulfate from soils ( 12).The unusually high affinity of phosphate foroxide minerals and clays is thought to be dueto the geometric compatibility between thephosphate ions and the oxide surface structure( 13), and the effective role of phosphate indesorbing polymers from clays has also beenstudied (14). It has been proposed also thatnot only does phosphate adsorb on the positiveand neutral sites but also on the negative sitesas described by the equilibria ( 15)

\-

'--"~~~Sci8I:r. Vol 142. No , N8da IS. 1991

FeOH! + H2PO. +-+ FeH2PO4 + H2O

FeOH! + HPO~- +-+ FeHPO. + H2O

FeOH + H2PO. +-+ FeH2PO4 + OH-

FeOH + HPO~- +-+ FeHPO. + OH-

FeO- + H2PO. +-+ FeHPO. + OH-

FeO- + HPO~- +-+ FePO~- + OH-

Similarly, for sulfate adsorption, we propose

FeOH! + SO~- +-+ FeSO. + H2O

FeOH + SO~- +-+ FeSO. + OH-

The effectiveness by which sulfate andphosphate ions compete with polyacrylamidefor adsorption sites on Fe20J merits furtherdiscussion, in particular, that which concernstheir mechanisms of adsorption. In contrastto the nonspecific type of adsorption bymonovalent ions such as a - and NO.1 , sulfate

and phosphate bond to the Fe203 surface byspecific adsorption, and when present in excesscause charge reversal of the mineral. In fact.under certain conditions, phosphate can beconsidered to be a potential determining ionfor hematite and other oxides (10). It has beenproposed, from infrared spectroscopic studies,that both sulfate and phosphate ions adsorbby ligand exchange with exposed surfaceFeOH and FeOH! groups to form binuclearcomplexes (9, 16). Thus, formation of thesecomplexes involves the exchange or displace-ment of two adjacent surface hydroxyls or wa-ter from the Fe3" coordination. This modelstrongly supports the proposed hypothesis that,for an oxide mineral. the MeOH andMeOH! surface grou~ are the adsorptionsites for polyacrylamide, and that polymer ad-sorption takes place through H-bonding withthese surface hydroxyls (8).

Effect of pH

Interactions of the above anions with FeiOJare stronger at lower pH where the charge dif-ference between the anion and the Fe203 ishigher. Consequently, depression of polymer

ADSORPTION OF POLYACRYLAMIDE 473

.

'"e

"-~e

>-"t-U)ZI4J Ia

z0~Go 4~0cna0{ I

0 100 100 300 400 500 600

RESIDUAL POLYMER, mQ/kQ700 800

b 1.6

1.4

NE

""-~E

~-t-in2iiia20~a.~0Ina..r

~

00 100 200 300 400 500 600 700 800

RESIDUAL POLYMER. mq/kq

FIG. 2. (a) Effect of 0.5 kmol/m3 Na2SO. on PAM adsorption on Fe~3 at pH 7.3. (b) Effect of 0.5kmol/m3 Na2SO. on PAM adsorption on FC203 at pH 3.6. (c) Effect of pH on PAM adsorption in I kmol/m3 Na2SO..

J 1I.t cfCeRl1ld -' I--~. Vol. \42, No. 2, MudlI~. \99\

1.2

1.0

0;1

0..

0.4

474 1.E:E AND SOMASUNDARAN

c

Ne"-~e

>-I-~Z1&10

Z0~

~0cn0C

The pH dependence of anion-Fe20J inter-action and the resultant dep~on ofpoiymeradsorption thus inverts the trend by whichpolymer adsorption changes with pH. In thepresence of nonspecifically adsorbing anionslike chloride, the original trend ( PAM adsorp-tion increases with dccrease in pH (8» ismaintained, as can be seen from Fi~. 3aand3b.

The preferential adsorption of PAM onpositive and neutral sites does not imply thatpolymer cannot adsorb on the negative sitessince, under basic conditions where the surfaceis negatively charged, adsorption can still takeplace. H-bonding in this case is probably be-tween the weakly acidic NH2 group on thepolymer and the basic MO - on the oxide. Ad-

sorption is thus lower due to the weaker H-bond and to electrostatic repulsion betweenthe negatively charged surface and the polymerif the latter had undergone some hydrolysisunder basic conditions. Under low adsorptionconditions at pH 12, addition of 0.5 kmol/mJsulfate does not reduce polymer adsorption

~

adsorption due to competition from these an-ions is expected to be more significant at lowerpH. This is indeed observed upon comparisonof results obtained when 0.5 kmol/ m3 sulfateis added at pH 1.3 and pH 3.6 (Fig. 2a and2b, respectively): the percent reduction inpolymer adsorption at pH 3.6 exceeds that atpH 7.3 by the higher interaction of sulfate withthe oxide surface at lower pH. In this case, notonly is adsorption of the anion weaker at neu-tral pH but the higher proportion of neutralsites a}S(j serves to limit the amount of polymeradsorption that is reduced. However. inter-action of sulfate with neutral sites is enhancedat higher sulfate concentration ( 1.0 klnol/ m 3),decreasing significantly polymer adsorptioneven around neutral pH (Fig. 2c). One caninfer that the addition of phosphate at low pHwill also result in complete depression of PAMadsorption since adsorption of phosphate in-creases with decrease in pH ( 10). The adsorp-tion densities of sulfate and phosphate onFe203 at pH 3.5 have in fact been measuredto be 85 and 110 pmol/g, respectively (9).

J ti'~_I--~. Vol 142.No.2.~ I~ 1991

ADSORPTION OF POLYACRYLAMIDE 475

a t.-

1.4

HE

~E

>-'~UiZ&&J0

Z

0

~0..C0In0C

0.2

00 100 200 300 400 500 600

RESIDUAL POLYMER, mg/kg700 800

b 1..

1.4

N

E"

00

E

>-'

I-

Ui

Z\oJ I

a

z

0

~Q. I

cr

0

In

0

-t I

0.1

00 100 200 300 400 500 600 700 800

RESIDUAL POLYMER. mq/kq

FIG. 3. (a) Effect of pH on PAM adsorption in 0.03 kmol/m3 NaO. (b) Effect of pH on PAM adsorption

in 3 kmol/m3 NaCl

~..~-~sa-., Vol 142, No. 2,Man:Io 15,1991

1.2

1.0

0",

0.6

0.4

1.2

1.0

0.8

0.6

0.4

476 LEE AND SOMASUNDARAN

1.6

Fe203

'"

~"-.e>-t-inz~z0f:-a.~0cn0~

~

pH0 6.6

0 3.6Cr . 500mg/kg

~'\\

'\\

\t

0.2

0

"""~aII41~Sd41Iee. Vol 142, No. 2, MaIdII5, 1991

t~a

1.0

0.8

0.6

0.4

"~-- " "

- "~0 CHLORIDE SULFATE PHOSPHATE

FIG. 5. Summary of effects of chloride, sulfate, and phosphate on PAM adsorption on F~O3 at pH 3.6and 6.6.

477AI:S)RFnON OF POLYACRYLAMIDE

constants from Sillen and Martell ( I 7) showthat formation of these complexes is insignif-icantly small due to the low solubility of theoxide. These calculations also show that therelative stabilities of the Fe3+-phosphate,Fe3+ -sulfate, and Fe3+ ~hloride complexesdecrease in the order by which the corre-sponding anions reduce polymer adsorptionon Fe203. If the stabilities of these complexesare maintained at the interface, then the rel-ative effectiveness by which the anions depresspolymer adsorption is comprehensible.

Mea of Urea

since the sulfate does not adsorb on the neg-atively charged surface (Fig. 4). On the con-trary, a slight increase in adsorption of thepolymer is observed. This may be due to "salt-ing-out" effect of the polymer by the sulfatewhich is observed when polymer-surface in-teraction is weak. At a lower but still basic pH(9.6) where there are still some neutral siteson the oxide surface, addition of 0.5 kmol/m3 Cao2 does not alter the adsorption levelof the polymer even though the calcium ionsdo adsorb on the negative sites. This suggeststhat polymer adsorption in this case is stilltaking place preferentially on the neutral sites.It is not totally ruled out, however, that thecations can replace some polymers from thesurface but that this effect is undetected and/or is counter-balanced by salting-out effect ofthe polymer by the cation. Interestingly, anenhancement in polymer adsorption by cationactivation is not observed.

From the above results, it is evident thatcomplexing agents such as sulfate and phos-phate compete very successfully with polymerfor adsorption sites on the Fe203 and dra-matically reduce the level of polymer ad-sorbed. A summary of results is given in TableI and graphically represented in Fig. 5.

The addition of anions may cause com-plexation between the anions and the mssolvedspecies from the oxide. Calculations of suchcomplexation for the species Fe3+ and theseanions in the bulk solution using equilibrium

TABLE I

Summary of E~ of OIloride, Sulfate, and Phosphateon Polyacrylamide Adsorption on FeP3 (Residual Con-~ntration ~ SOO mg/kg)

~ ill

~

pHJ.6 pH 6.6~

1.0 kmoI/m' chloride3.0 kmol/m' chloride0.5 kmoVm' sulfate1.0 kmoVm' sulfate0.3 kmoI/m' phosphate

6.319.732.364.6

100

-19.377.976.6

It has been shown above that competingagents which adsorb specifically on the mineralsurface and thus deplete adsorption sites arehighly effective in depressing polymer adsorp-tion. Another way to reduce the level of poly-mer adsorbed is by the use of molecules whichcan affect the bonding mechanism, namely H-bonding between the polymer and the surfacehydroxyls. For this purpose, urea is chosensince it is believed to be a H-bond breaker byvirtue of its own tendency to form H-bond.The effect of urea on H-bonding in other sys-tems have been studied elsewhere (18, 19).

In solution, urea acts as a "structure-maker"(20) which implies that it can salt-out poly-mer, a phenomenon which generally leads toincrease in polymer at the interface. But theresults in Fig. 6 show that the polymer ad-sorption is decreased in the presence of urea.This is attributed to the overriding effect ofurea at the interface. Since urea is also anamide (carbamide), adsorption of it on theoxide surface is not surprising. In fact. it isbelieved that urea is a stronger H-bonding basethan other amides due to resonance stabili-zation (21), therefore interaction of it withoxide surface hydroxyls can be expected to Ixslronger than that of polyacrylamide.

An important point worth noting here isthe mechanism by which polymer adsorptionis reduced by the various competing agents.While sulfate and phosphate deplete adsorp-tion sites by ligand exchange with FeOH! and

J--ifCJJIIrUIDIdI--~. Vol 142, No.2,~ J~. 1991

478 LEE AND SOMASUNDARAN

1.6

1.4

...e"GO

e

,:I-~Z

~Z0

~Go~

0cn0.

0.2

00 1.0 2.0 3.0

UREA. kmol/m3

Fki. 6. ~ olPAM 00 F~ in the presena: of~ at pH 7.

4.0

FeOH replacing the surface hydroxyis, ureabinds with the surface hydroxyls by H-bondingat the expense of polyacrylamide. In this case,the level by which polymer adsorption is re-duced is found to be independent of the modeof addition of urea; irrespective of whetherurea is added before, together with, or afterthe polymer, the reduction in polymer ad-sorbed is the same. This is a consequence ofthe reversibility polymer in adsorption.

lower pH due to higher interaction of anionswith the more positively charged surface.Thus, the effectiveness by which these anionsreduce polymer adsorption follows, in in-creasing order, chloride < sulfate < phosphate.This is also the order of relative stability ofthe anion complexation with Fe3+.

Urea is also found to reduce polyacrylamideadsorption on Fe203. Contrary to the case ofsulfate and phosphate where surface hydroxylsare replaced, this is attributed to H-bondingof urea itself with the surface hydroxyls thusdepleting adsorption sites for the polymer.

CONCLUSIONS

Polyacrylamide adsorption on Fe203 is de-creased in the presence of competing anionssuch as chloride, sulfate, and phosphate. Whilethe effect of chloride is less marked, significantreduction by sulfate and complete depressionby phosphate are obtained. This can be ex-plained by taking into consideration the spe-cific interactions of sulfate and phosphate withFe203 by ligand exchange, replacing surfacehydroxyls which are adsorption sites for thepolymer. These effects are more significant at

J-.-ti"c MdI-nj8I:rSc'-r. Vol 142. No. 2. MaIdIIS. 1991

ACKNOWLEDGMENT

The autbolS are grateful for ftnaocial suppon from theNational ScjeD(X Foundation (PmBIams: Auid, Particulateand Hydraulic Processing &. Interfacial, Transpon, andSeparation ~), Naloo. B. P. America, and Engel-hard

'V'

REFERENCES

I. Emerson, W. W., and Raupach, M., AUSlraJ. J. SoilSci. 2, 46 (1964).

1.1

1.0

0.8

0..

0.4

ADSORPTION OF POLYACRYLAMIDE 479

2. Griot, 0., and Kitcbener, J. A., Trans. Faraday Soc.61.1026(1965).

3. Kavanagh, B. V., Posner, A. M., and Quirk, J. P., J.Soil Sci. 27.467 (1976).

4. Der, R. K., "The Chemistry of Silica." Wiley, NewYork, 1979.

5. Gerbhardt, J. E., and Feurstenau, D. W. in "1ntelfacialPhenomena in Mineral Processing" B. Yarar andD. J. Spottiswood, Eds.), p. 175. The EngineeringFoundation, New York, 1982.

6. Pefferkom, E., Nabzar, L., and Carmy, A., J. ColloidInterface Sci. 106. 94 (1985).

7. Pefferkom, E., Nabzar, L, and Varoqui, R., ColloidPolym. Sci. 265. 889 (1987).

8. Lee, LT., and Somasundaran, P., Langmuir 5,854( 1989).

9. Parfitt, R. L, and Smart, R. St. C., Soil Sci. Soc. Amer.J. 42. 48 (1978).

10. Breeuwsma, A., and Lyklema, J., J. Colloid lnterfaceSci. 43. 437 (1973).

II. Hin~n, R. J., Posner, A. M., and Quirk, J. P., J.Soil Sci. 23, 177 (1972).

12. Barrow, N. J., Soil Sri. 104. 242 (1967).13. ~,G., "The Surface Olemistry of Soils. " Oxford

Univ. Press, New York. 1984.

14. Dodson, P., and Somasundaran, P., J. Colloid Inter-face Sci. 97,481 (1984).

15. Breeuwsma, A., Thesis. Agricuhurat Univenity,Wageningen, The Netherlands, 1973.

16. Parfitt, R. L., Atkinson, R. J., and Sman, R. St. C.,Soil Sci. Soc. Amer. Proc. 39, 837 (1975).

17. Sillen, L. G., and Marten, A. E., "Stability Constantsof Metal-Ion Complexes," Special Publication, No.17. Chemical Society, London, 1964.

18. Guy, R. H., Honda, D. H., and Aquino, T. R. II, J.Colloid Interface Sci. 87, 107 (1982).

19. Suzawa, T., Shirabama, H., and Fujimoto, T., J. Col-loid Interface Sci. 93, 498 (1983).

20. Lee, L. T., Thesis. Columbia University, New York,1986.

21. Pimental, G. C., and ML<:IeUan, A. L, "The HydrogenBond." Freeman, San Francisco/London, 1960.

--cfc l--~ Vol 142, No. 2, M8IdIIS, 1991