CROATICA CHEMICA ACTA CCACAA 60 (3) 383-393 (1987)

YU ISSN 0011-1643UDC 541.18:537.24

Conferenee Paper (Invited)

Electrochemical Characterization of Aerosil OX 50 Dispersions*

CCA-1737

H. Sonntag, V. Itschenskij, and R. KoLeznikovaInstitut for Physical Chemistry, ADW DDR, 1199 Berlin

Received December 15, 1986

The interface between the Aerosil OX 50 colloidal dispersionand the electrolyte solution was studied using three indepen-dent methods: titration for surface charge density determination,conductivity, and electrophoretic mobility. It is shown that allthree methods give good agreement for the surface charge densityin 10-2 rnol/dm" KCl solutions, but less so for 10-3 mol/dm".

INTRODUCTION

The properties of colloidal dispersions are mainly determined by the inter-action between their particles. One component of the interaction is the electricdouble layer repulsion, which is the result of the overlapping of the diffusepart of the double layers of approaching particles during Brownian encounters.Knowledge on the structure and properties of double layers is therefore veryimportant in colloid stability.

EXPERIMENTAL

We investigated aerosil OX 50 dispersions of 2.5 respectively 5 weight percent.OX 50 was first dried at 420 K for two hours and the n dispersed either by ultra-sonication (45 minutes) or using an ultraturrax. The electrochemical properties ofthe aerosil particles are determined by the silanol groups, On the average thereare 5 silanol groups at 1 nm" at the surface. The number of silanols dissociated,that means the surface charge density, depends on the pH of the solution accordingto the reaction

The reactionSiOH + Hp+ ~ SiOHt + Hp

was not observed in aerosil OX 50 dispersions.

RESULTS AND DISCUSSION

A typical titration curve is shown in Figure 1; curve 1 for the dispersionand curve 2 for the solution. From IlVorr the surface charge density wascalculated according to eq. 1

F(10 = F tI'H30+- r ow) =S (~VH30+. CH3o+- ~ vorr . Cow) (1)

* Based on an invited lecture presented at the 7th »Ruđer Bošković" Institute'sInternational Summer Conferenee on the Chemistry of Solid/Liquid Interfaces, RedIsland - Rovinj, Croatia, Yugoslavia, June 25-July 3, 1986.

384 H. SONNTAG ET AL.

F Faraday constant, T adsorbed amount (mol m-2), S surf'ace, 6.VH,O+, 6.VOIrvolume difference of hydronium respectively hydroxyl ions between the pureelectrolyte solution and the dispersion to obtain the same pl-I-val ue.

ml 0,1nKOH

0,5

ml 0,1 nHel

o0,1

30,1

0,2

0,3

0,4

Figure 1. Titrations curves for the electrolyte solution (2) and a 2.5% aerosilOX 50 dispersion (1) in 10-3 normal K' solution.

The first reaction is dependent on the kind and concentration of counte-rions. In Figure 2 this is shown for 3 different potassium chloride concen-trations. Then higher the potassium chloride concentration, then highen thesurface charge density. From this, it follows, that the surface charge densitymay not be varied simply by alkaline addition. On the other hand the experi-ments have shown that chloride ions are not adsorbed at the silica interface.So the pR was varied by adding hydrochloride acid to aerosil dispersions ofconstant alkali concentration. With this method surface charge density mea-surements can be performed only with soluble hydroxides. For unsolublehydroxides a described coulometric method was used.1 The hydroxyl ionswere produced electrochemically at a platinized platinum electrode. In expe-riments in which hydrogen can influence on the sur:face charge of the dispersedparticles a palladium cathod was used in which the gaseous hydrogen isabsorbed. The anode was made from silver, and' the silver ions are settledas silver chloride in an agar-agar jelley. A porous filter prohibited agar-agarmolecules to diffuse into the electrolyt solution. A stirrer, a glass electrodeand a calomel electro de are placed within the cel!.'.Argon flows through thecell during the experiment to avoid adsorption of carbon dioxide. Two celles

· 6/-UC- o cm2

7

6

5

4

3

2

1 .-o2 3 04- 5 6

AEROSIL OX 50. (385

!10-2 m KCl

!10-3 rn KCl

7 e 9 pH

Figure 2. Surface charge density pR curves for 2.5% aerosil OX 50 dispersionsat three different potassium chloride concentrations.

6/~ t- o cm:5

4

3

2

1

O2 3 4 S 6 7 8 pH

Figure 3. Surface charge density pR curves for 2.5% aerosil OX 50 dispersionsin 10-2 M alkaline chloride solutions cl i.ici O KCi X RbCi (coulometrically)

D KCi potentiometrically.

were used in series one contained the dispersion the other only the electrolytesolution. The results are shown Figure 3. for lithium (.), potassium (O) andrubidium chloride (X). For comparisson the curve for potassium chloride ob-tained by titration (O) is included in Figure 3. This lyotropic row is explainedby adsorption of hydrated cations and the decrease of hydration from lithium

JAC-C"o -2cm11

10

9

8

7

6

5

4-

3

2

1

O2 3 4 5 6

386 H. SONNTAG ET AL.

CaCLZ

KCl

7 8 9 pHFigure 4. Surface charge density pH curves for 2.5~/o aerosil OX 50 dispersions

in 0.01 M KCl, 0.005 M CaCh and BaC12.

to rubidium, that means it is an effect of double layer capacity. We alsoinvestigated the influence of barium and calcium counter ions in comparisonwith potassium ions (Figure 4). The results showed that there is no shift ofthe point of zero charge and also no specific adsorption of these kind of ions.From the difference in the slopes of the curves for potassium, calcium andbarium chloride the same conclusion can be drawn, namely the surface chargedensity is highest with Ba2+ ions, that means these ions are adsorbed withtheir hydration shells. The same conclusions were already made by Tadrosand Lyklema with precipitated silica.š

Potentiometric titration experiments alone do not allow the decisionwhether there is specific adsorption of earth alkaline ions or not. This waspossible by measuring electric light scattering in the low frequence region.It was shown" that above pH 8 there was a small specific adsorption of Ca2+and Ba2+ at the aerosil surface. According to the Gouy-Chapman theory thereexists arelation between surface charge density and surface potential. On theother hand it was shown by Martynov! that the difference between the therrno-dynamic potential ('./10) and Stern potential is negligible, if the surface chargeis formed by adsorption of ions. The Stern potential c/Js is related to surfacecharge density by the following equation

AEROSIL OX 50 387

(2)

El dielectric constant, Eo influence constant, x-1 Debye parameter. The dimension-

less potential is defined, ifJ = zieifJlkT.

The resuIts for ifJ" as a function of IJo are shown for 3 different potassiumchloride concentrations in Figure 5. At low surface charge densities there is alinear dependence between ifJ" and IJO.

100

o L--L__L-~ __~~ __~~~~ __~

O 1 2 5 6 7

150

Figure 5. Stern potential surface charge density curves at different KCl eon-centrations calculated according to eq. 2.

Electrophoretic Mobility

The electrophoretic mobility of aerosil particles cannot be measured bysingle partic1e electrophoresis due to the low partic1e radii. We used masstransport electrophoresis. The method was described."

The electrophoretic mobility was found to be dependent on the surfacecharge density and on the concentration of the indifferent ions too. Withincreasing electrolyte concentration the electrophoretic mobility became lower(Figure 6) this is in aggreement with the Gouy-Stern-theory namely thenhigher the concentration of the electrolyte then higher the counterion eon-

388 H. SONNTAG ET AL.

4

3

2

1

345 6 7 .8 9 10 pHFigure 6. Electrophoretic mobility as a function of p'H in 10-4 molar and 10-3 M KCl.

centration near the surface. If one compares electrophoretic mobility in. thepresence of different alkaline ions (Figure 7) at the- same electrolyte concen-tration no influence was found. But if we compare electrophoretic mobilityof potassium, calcium and barium ions at the same concentration of 10-3normal different (Figure 8) valu es are obtained. The reason is the increase ofadsorption from K+< Ca2+<Ba'". From electrophoretic mobility the Š -potential was calculated using Wiersema-Loeb-Overbeek-theory.6 The š-poten-tial-surface charge density curve is compared with the Stern potential-surfacecharge density curve, calculated with eq. 2 from surface charge density (Fi-gure 9). Unexpectedly the experimental values of the š-potential are higherthen calculated Stern potentials from surface charge density at low (Jo. In10-2 mol ar KCl solutions both curves are identical.

Conductivity Experiments

The specific conductivity of colloidal dispersions depends on the specificconductivity of the dispersion medium, and of the dispersed phase and on thesurface conductivity. The last term was first considered by von Smoluchow-skF later by Henry" and Booth." The influence of the polarisation of the

AEROSIL OX 50 389

U'104 J.cm2./ V'S

4

3

2

1

3 4 5 6 7 8 9 pH

Figure 7. Electrophoretic mobility as a function of pH in 10-3 normal solutionsof different alkaline ions.

double layer on the electric conductivity of colloidal dispersions for x a » 1was first developed by Duchin and Schilov.t? According to Duchin the fol-lowing relation is valid between the specific conductivity of the dispersion (k),the specific conductivity of the dispersion medium (k), the volume fraction pof the dispersed phase and the parameter Rej, which characterizes the surfaceconductivity (k<1) and the polarization of the double layer:

k = 1_ ~ p (1 3 Rei ) (3)k 2 1 + 2 Rei

~ ~Rei = :: = x

1a[ 4 (1+ 3m) sin h2 (+) + 2 (cos h (1f2'o ) - cos h (~)] (4)

m -= kB2 T2 S Co

e2 6 JI: 'Y) D

kB Boltzmann constant, T temperature, e riielectric constant, Eo influence con-stant, e elementary charge, r] viscosity, D diffusion coefficient.

390 H. SONNTAG ET AL.

4

co++

Ba++

3

2

1

o2 3 4 6 7 8 9 pU5

Figure 8. Electrophoretic mobility in 10-3 M KCl, 5 X 10-4 M CaC12and BaC12atdifferent pH.

If Rei increases and becomes 1 the term in brackets in eq. 3 becomes zero.Under such condition dispersion and dispersion medium have the same con-ductivity (isoconductivity). In Figure 10 the conductivity of the dispersion (.)and centrifugate (X) is shown in 10-2 M KCl as a function of pR and in Figure11 as a function of surface charge density. In Figure 12 the conductivity isshown in 10-3 M KCl (. dispersion, + centrifugate) as a function of pR. Onecan see that with increasing surface charge density k, becomes higher than k(enhanced conductivity). The position of the point of isoconductivity dependson the electrolyte concentration and moves to higher pR values with increasingelectrolyte concentration.

In Figure 13 the ratio k/k is shown as a function of the volume fraction pfor different electrolyte concentrations. All the straight lines cut the ordinateat the point of isoconductivity. The drawn line corresponds to the equationk/k = 1 - 3/2 p.

The experimental result that with higher electrolyte concentrations thecurves are lower than the theoretical curve may be caused by formation ofaggregates. If we now compare the calculated Stern potentials in 10-2 and

AEROSIL OX 50 391.

150

/..~ 10-4 mKCl

10-3 mKCL

oO 1 2 3 ~ 5 S '_~/t~2

Figure 9. Comparison of the Zeta- (X) and the Stern- (e) potential calculatedfrom o; (eq. 2).

K,K·J03

~~cm

2p

\.9

·\8

1,7

1.6

1,5

1,4

1,3

\2

4 6 8 9 P~

Figure 10. Specific conductivity of the 2.50/0 aerosil OX 50 dispersion e and of thedispersion medium X in 10-2 M KCl as function of o.;

392 H. SONNTAG ET AL.

1,7

1,6-

1,3

1,2

1,1 I.-.--I_.......L_-L.._..L.-_..I.-_L.:.---I_---.-

O 1 3 4 5 6 /"uC-"O cm22

Figure 11. Specific conductivity of the 2.5% aerosil OX 50 dispersion • and ofthe dispersion medium x in 10-3 M KCl as a function of pH.

K,K·IO·

[,~Icmll

3.:,

\3,e

2,6

-,\

1,8

I"

1,0

pH

Figure 12. Specific conductivity of the 2.50/0 aerosil OX 50 dispersion • and ofthe dispersion medium x 10-3 M KCl as a function of pH.

10-3 M KCl solution obtained from surface charge density, zeta potential andconductivity the following results are obtained:

r

AEROSIL OX 50 393

lJ!o/k/k (mV)

10570

11475

10855

1,1 ----- -..---c-__--- .-- --6--

-:::::---.-------~----- 1 221'O!"-'~~;;;:c=-~~"___;~

~~

R/K1,2

O~

Figure 13. Relative conductivity k/k as a function of the volume fraction indifferent KCl solutions O 5 X 10-5 M, X 10-4 M, 6. 1.5 X 10-4 M, \l 5 X 10-4 M,

O 10-3 M, + 10-2 M. The drawn line corresponds to eq. 5.

The aggreement between these values in the 10-2 mol ar solution is rather good,but in 10-3 mol ar solution the calculation from conductivity according toDuchin-Schilov gives too low value. May be the limiting condition for theD-S-Theory (x a » 1) is no more valid (for 10-3 M x a ~ 2.6).

REFERENCES

1. H. P i 1g rim m and H. S o n n t a g, CoHoid PoLymer Sei. 258 (1980) 471.2. Th. F. Tad r o s and J. L y k 1e m a, J. ELeetroanaL. Chem. Interf. ELeetroehem.

17 (1968) 267, ~2 (1969) 1.3. S. Lev i n e and G. M. B e l l, Diseus. Far. Soe. 42 (1966) 69.4. G. A. Mar t y n o v, KoUoidny ZurnaL 41 (1979) 1105.5. H. P i 1g rim m, W. Kat z, and H. S o n n t a g, CoHoid PoLymer Sei. 257

(1979) 320.6. A. L. Loe b, P. H. W i e r se m a, and J. Th. G. Ove r b e e k, The ELeetrieaL

DoubLe Layer Arround a SpherieaL CoUoid ParticLe, M. 1. T. Press Cambridge,Mass. 1961.

7. M. von S mol u c how s k i in G r a e t z, Handbuch der ELektrizitii.t und Ma-gnetismus, II, Leipzig, 1914.

8. D. C. He n r y, Proe. Roy. Soe., London A 133 (1931) 106.9. F. B o o t h, Proe. Roy. Soe. A 203 (1950) 534.10. S. S. D u c h i n and V. N. S c hil o v, DieLektri(;eskie javlenija dvoinoj sLoj

v dispersnyeh sistemaeh, Kiev Naukova Dumka (1975).

SAZETAKElektrokemijska karakterizacija Aerosil OX 50 disperzija

H. Sonntag, V. Itsehenskij i R. KoLeznikova

Primijenjene su tri neovisne metode na određivanje polarizacije u električkomdvosloju na granici faza Aerosil OX 50/otopina elektrolita. Pokazano je da svetri metode - vodljivost, titracija za određivanje površinske gustoće naboja ielektroforetska pokretljivost - daju vrlo sukladne rezultate za KCl koncentracije10-2 mol/dm", ali i značajno veća neslaganja za koncentraciju 10-2 mol/". Razlogje tome što za nižu koncentraciju nije ispunjen uvjet x a » 1.