The TiO2/electrolyte solution interface I. Influence of pretreatment conditions and of impurities

(1)

The TiO2/electrolyte solution interface I. Influence of

pretreatment conditions and of impurities

Citation for published version (APA):

Janssen, M. J. G., & Stein, H. N. (1986). The TiO2/electrolyte solution interface I. Influence of pretreatment conditions and of impurities. Journal of Colloid and Interface Science, 109(2), 508-515.

https://doi.org/10.1016/0021-9797%2886%2990330-9, https://doi.org/10.1016/0021-9797(86)90330-9

DOI:

10.1016/0021-9797%2886%2990330-9 10.1016/0021-9797(86)90330-9

Document status and date: Published: 01/01/1986

Document Version:

Publisher’s PDF, also known as Version of Record (includes final page, issue and volume numbers)

Please check the document version of this publication:

• A submitted manuscript is the version of the article upon submission and before peer-review. There can be important differences between the submitted version and the official published version of record. People interested in the research are advised to contact the author for the final version of the publication, or visit the DOI to the publisher's website.

• The final author version and the galley proof are versions of the publication after peer review.

• The final published version features the final layout of the paper including the volume, issue and page numbers.

Link to publication

General rights

Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights.

• Users may download and print one copy of any publication from the public portal for the purpose of private study or research. • You may not further distribute the material or use it for any profit-making activity or commercial gain

• You may freely distribute the URL identifying the publication in the public portal.

If the publication is distributed under the terms of Article 25fa of the Dutch Copyright Act, indicated by the “Taverne” license above, please follow below link for the End User Agreement:

www.tue.nl/taverne

Take down policy

If you believe that this document breaches copyright please contact us at:

openaccess@tue.nl

providing details and we will investigate your claim.

(2)

T h e T i O 2 / E l e c t r o l y t e S o l u t i o n I n t e r f a c e I. Influence of Pretreatment Conditions and of Impurities

M. J. G. JANSSEN AND H. N. STEIN Laboratory of Colloid Chemistry, Eindhoven University of Technology,

P.O. Box 513, 5600 MB Eindhoven, The Netherlands Received March 25, 1985; accepted June 21, 1985

The influence of pretreatment of TiO2 in oxidizing or reducing atmospheres on surface charge, charge behind the electrokinetic slipping plane, and stability toward coagulation has been investigated. No influence was found on pure TiO2, which does not show any specific adsorption of K +, CI-, or NO~ ions. Impurities in the TiO2, however, lead to effects suggesting specific adsorption of anions (SOl-). Pretreatment of impure TiO2 under reducing conditions enhances this effect (reduction of SO 2- to S 2-, which is strongly specifically adsorbed). However, the effect is compensated behind the electrokinetic slipping plane by additional adsorption of anions or desorption of cations. © 1986 Academic Press, Inc.

INTRODUCTION

Oxides and oxidic materials are among the most common solid phases in dispersions. For colloid chemical investigations, however, they offer more difficulties than, e.g., AgI because the low conductivity of oxides makes changes in the overall potential drop across a solid/ electrolyte solution interface not accessible to direct measurement.

This difficulty can be circumvented by em- ploying semiconducting oxides; here a change in flat band potential can be measured. The flat band potential is that overall potential difference between the solid and a reference elec- trode, at which there is no potential difference between the bulk solid and a position near the phase boundary (but still within the solid). The use of semiconducting oxides for bridging the gap between model substances (such as AgI) on the one hand, and solids encountered in practical situations on the other, raises the question, whether a change in solid-state properties influences colloid chemical phe- nomena.

The present study aims at elucidating this question for TiO2. This is an n-type semicon- ductor, with a band gap of 3.3 eV for anatase

0021-9797/86 $3.00

and of 3.1 eV for futile (1, 2). Thus, pure TiO2 will behave at room temperature as an insu- lator. Its conductivity, however, can be increased by heating in hydrogen (3, 4). During reduction, titanium interstitials are introduced as multiple donors according to

2H2 + 20~0 + Ti-~i ~ Ti~ + 3e' + 2H20 [1] In Eq. [1], the defect notation of K_rrger (6) has been used.

In this study, we measured the colloid chemical properties of powdered TiO2, with regard to the influence of reduction and of impurities on surface charge, electrophoretic mobility, and stability against coagulation, of two powdered TiO2 samples of different degree of purity.

EXPERIMENTAL

Materials

Ti02: (a) ex Merck pro analysi (M 808), consisting of XRD-pure anatase; (b) ex De- gussa (DP25), consisting of a 80:20 mixture of anatase and rutile (estimated from XRD).

Pretreatment consisted of heating under oxidizing or reducing conditions. The oxidiz-

508

(3)

TiO2/ELECTROLYTE INTERFACE, I 509 ing pretreatment involved heating for 20 h in

a stream of oxygen at 600°C for TiO2 M808 and at 530°C for TiO2 DP 25, while cooling was performed in a stream of oxygen. The temperatures were chosen so as to prevent pronounced sintering. The reducing pretreatment involved heating for 19 ½ h in a stream of oxygen, followed by 30 min heating in a stream of nitrogen and 15 min heating in a stream of hydrogen, while the H2 stream was continued during cooling. The heating temperatures again were 600°C for TiO2 M808 and 530°C for TiO2 DP 25. The gases were used at atmospheric pressure. After heating, the BET surface area was 7.4 m E g-i for TiO2 M808, and 42.4 m 2 g-i for TiO2 DP 25, irre- spective of the atmosphere during heating. The samples pretreated under oxidizing conditions are indicated as TIO2/O2, while those pretreated under reducing conditions are men- tioned as TiO2/H2. The heating pretreatment was preceded in some cases by Soxhlet extraction using water as a solvent.

All other materials were of analytical grade. The water employed had been twice distilled.

Methods

Surface charge measurements

were per-

formed by registering the H ÷ or OH- consumption necessary to keep the pH at a preset value, by means of two Radiometer TTT 80 titrators and two Radiometer ABU 80 Auto- matic Burettes (one for acid and one for alkali addition), in combination with two Radiome- ter REC 80 recorders extended with a REA 270 pH stat module. The resulting pH was followed using a Radiometer pH M84 Re- search pH meter and a Kipp BD 40 recorder. The measurements were performed in a 400- ml thermostated beaker at 20.0 + 0.1 °C, containing 300 ml aqueous electrolyte solution. In order to eliminate possible effects of light on H ÷ and OH- consumption as reported by Trimbos and Stein (7) for the case of ZnO, the beaker was painted black from the outside. The solution was intensively stirred by means

of a magnetic stirrer. For more details see Ref. (8).

Electrophoretic mobilities

of suspended

TiO2 particles were measured with a Rank Brothers Mark II Microelectrophoresis appa- ratus, using a fiat cell. After termination of a pH stat experiment, 100 ml of the suspension was centrifuged (20 min at 3000 rpm). Part of the sediment was redispersed in 50 ml of the supernatant, the very dilute suspension formed was used for microelectrophoresis. No signif- icant difference in pH was found between the supernatant and the suspension during the pH stat experiment. Thus, the suspension effect (9) can be neglected. Electrophoretic mobilities were determined at 25.0 + 0.1 °C; average values of up to 20 particles were recorded. Zeta potentials were calculated by the Helmholtz- Smoluchowski equation (10). It is true that the primary particles were too small to comply with the requirement ra >> 1 under all circum- stances met in this investigation; however, in the microelectrophoresis cell, flocs are followed which are large enough for this demand, and the floes' dimensions determine the hy- drodynamic friction.In most cases the shape of the flocs could be determined through the microscope. In others only vague light scat- tering objects could be discerned; the intensity, however, was such as to justify the supposition that the objects themselves were only slightly beyond the resolving power of the microscope.

Electron spin resonance

was measured at

liquid-nitrogen temperature, using an X-band Varian E- 15 spectrometer.

Stability toward coagulation

was measured

by a comparative method. A suspension of 0.2 g of TiO2 in 90 ml of KC1 or KNO3 solution was prepared by stirring magnetically for at least 24 h 0.1 ml of this suspension was diluted with 24 ml electrolyte solution, and the pH was adjusted to the desired value. The suspension was kept in the dark for 24 h; during this time the pH was adjusted if necessary. Five milliliters suspension was transferred to a coagulation tube, which was closed air tight by means of a snapcap. All these actions were

(4)

510 JANSSEN AND STEIN performed in a CO2-free glovebox. The dilute

suspension in the coagulation tube was subjected to ultrasonic treatment in a Sonicor Type SC-50-22 stirring bath during 20 min.

Light extinction (X = 440 nm) was measured immediately after this treatment (Eo) and 2 h later (E2); the ratio

E2/Eo

was taken as a measure of the stability of the suspension. More elaborate procedures, employed in the analogous cases of aqueous SiO2 or ZnO dispersions (11-14) showed that changes in light extinction is connected, in the dispersions concerned, with coagulation.

RESULTS

ESR

TiO2 DP 25/02 did not give a detectable ESR signal; TiO2 DP 25/H2, however, showed a broad signal at g = 1.96 corresponding to Yi 3+ (3d 1) ions in the TiO2 matrix (15). With TiO2 M808, however, additional signals were found: for TiO2 M808/O2 at g = 2.004, 1.99, 1.97, and 1.942. Treatment with H2 resulted for TiO2 M808 in a reduction of the signals at g = 2.004, 1.99, and 1.942, while again a broad signal at g = 1.96 appeared. The origin of the additional signals in TiO2 M808 is not clear; but these signals indicate a larger amount of impurities in this sample than in TiO2 DP 25.

Adsorption of Potential Determining Ions

and Zeta Potentials

The net consumption o f H + or OH- by TiO2 M808/O2 (designated by ron-.n,) in 10 -2 M KC1 solution is shown in Fig. I. At pH i> 7, a fast consumption of H + or OH- is followed by a slow OH- consumption. Similar phe- nomena have been reported by B6rub6 and De Bruyn (16) and by Schindler and Gams- jager (17). This slow process is not due to silica dissolution from the glass beaker, since the same phenomenon was observed in a PTFE beaker. IR absorption data indicate a possible CO2 surface contamination on all TiO2 samples; reaction by CO2 would explain the pH

dependence of the slow process. In addition, Morterra c.s. (18) showed that C O 2 adsorption

on anatase is only slightly dependent on the sample pretreatment temperature. In the case of ZnO, a slow process caused by adsorbed CO2 was found (12, 13) (in this case the CO2 can be expelled by heating). For these reasons, we ascribe the slow OH- consumption by TiO2 at high pH values also to CO2; in the rest of this paper, we treat ron-.on+ values found by extrapolation to time = 0 if a slow process is present.

From I ~ O H - . H + , surface charges (~r0) were

calculated; the Figs. 2-4 show a0 (pH) in different solutions, for TiO2 M808/O2 and TiO2 M808/H2.

These figures (2-4) show in all cases the PZC at higher pH values than the IEP, indicating pronounced chemisorption of anions (20, 21). Other investigators (22-30) report both PZC and IEP for TiO2 at pH 5.8-6.5; only Fukuda and Miura (19) found in some cases an IEP at pH 3.5, consistent with the ~'(pH) curves in the present work.

This suggests that anionic impurities are responsible for the difference between PZC and

0 -10 -20 -30 ~OH:F~xIO? MnI2 pH f 1 - - 8 2'0 z,'O 60 80 ~ f{hours}

FiG. '1. Net consumption ol~H+/OH - as a function of time, for dispersions of TiO2 M808/O2 in 10 -2 M KC1.

(5)

TiO2/ELECTROLYTE INTERFACE, I 51 1 .3{ Oo(C~ z) .20 40 o i -.10 --2(

\

> pH

\

$ o\ ~(mV) "60 -2O

"°'I I

\

-20 .10 S 6 7 [ (mY) O pH 9 10 -60 -20

FIG. 2. Zeta potential (O) and surface charge d e n s i t y

(O) as functions o f p H for TiO2 M808/O2, in 10 -3 MKC1.

\

Oo(C i~ 2) ~

/

S 6 -.20 ~(mV)

t

. . ~)~ -20

,o

0

FIG. 3. Zeta potential and surface charge as functions o f p H for TiO2 M808, in 10 -2 M KI. (O) TiO2 without pretreatment, (©) TIO2/O2, ([3) TiO2/H2.

FIG. 4. Zeta potential and surface charge as functions o f p H for TiO2 M808 in 10 -2 M K N O 3 . (©) TIO2/O2, (O) TiO2/H2.

IEP found by us. We investigated the influence of impurities on surface charge and electro- kinetics by two methods.

(a) TiO2 M808 was subjected to Soxhlet extraction for various periods before the heat- ing pretreatment.

Figure 5 shows that both #0 and the elec- trophoretic mobility at pH 5 in 0.01 MKNO3 solution decrease on Soxhlet extraction; the decrease of the former is particularly pro-

0 I -1. -2. -3. • 1 .2 1.1 (p.m.t'-/$.V) ,,, ' , .3 :,., Oo((~ 2)

FIG. 5. Mobility as a function o f surface charge for TiO2 M808 at pH 5 in 10 -2 M KNO3: influence o f duration o f

S o x h l e t extraction (in hours).

(6)

512 JANSSEN AND STEIN

nounced. Figure 6 shows the figure compa- rable to Fig. 4, after 72 h of Soxhlet extraction; IEP and PZC are shifted toward each other, but do not yet coincide. In Fig. 7 we plotted ao vs ~re, the net charge between the phase boundary and the electrokinetic slipping plane. The latter was calculated by

with

cr~ = a t - a0 [2]

2 e o e r k T . , [ z e ~ " t

,rt - - - z e ~ s m n ~ 2 ~ / [3]

for Ka > 100; for lower Ka values, the approx- imate Loeb-Wiersema-Overbeek formula (31) was used. In Eq. [3], ~o is the permittivity of vacuum; Er the relative dielectric constant; e is the absolute value of the charge of an electron; z the valency of the counterions; k the Boltzmann constant.

From Fig. 7, we see that at the PZC, before Soxhlet extraction, ae is < 0; after Soxhlet extraction, a~ is at the PZC almost = 0 but still

.IC .05 0 ~OS -40 -.15 -.20 -,25 o.([~)

)

O

° I---i-

/ ~ o / °

0 ~ c 9 ( c}' ~ (my)

T

" ~ , ~ ~ pH

o

0 -50 -30

FIG. 6. Zeta potential and surface charge as a function o f p H for TiO2 M808 in 0.01 M KNOa solution, after 72 h of Soxhlet extraction. (O) TIO2/O2, (e) TiO2/H2.

32 F

\

o 2t+ ° 42 -2'4 -I'6 -8 -1 -32 Oo (C r62)-I02 -~;-- O~ (On-2).I02 8 16 24

FIG. 7. Surface charge (a0) vs charge in the Stern layer (%), before (O) and after (e) Soxhlet extraction.

< 0 (cf. also Fig. 6). This shows that impurities in TiO2 M808, which are removed only with great difficulty, are responsible for the difference between PZC and IEP. The effect could be due to specific adsorption of anions, or to desorption of specifically adsorbed cations.

In the wash water obtained during the Soxhlet extraction, SO24 - could be detected. The most probable contaminating cation is Fe 3+, since TiO2 is produced commonly from Fe-containing sources and Fe 3+ is the most widely occurring highly charged cation, sep- arated entirely from TiO2 only with difficulty. Fe 3+, however, was not detected in the wash water by atomic absorption spectroscopy (de- tection limit: 2 × 10 -6 M).

(b) TiO2 from another source (TiO2 DP 25), showing less impurity influence in the ESR spectrum, was employed. Here PZC and IEP coincided after 32 h Soxhlet extraction both in KNO3 solutions (Figs. 8 and 9) and in KC1 solutions, showing absence of specific adsorption in the pure TiO2. No systematic difference is found between TiO2 DP 25/02 and TiO2 DP 25/H2 with regard to ao (see Fig. 8). Neither was there a shift of the a0 (pH) or the ~'(pH) curve on Soxhlet extraction (Figs. 10, 1 l).

In the wash water of the Soxhlet extraction Zn ions (2.4 × 10 -5 M) could be detected but no SO ]-, CI-, or Fe 3+.

(7)

TiO2/ELECTROLYTE INTERFACE, I 513 -'20 -.16 -'1; -'0~ -'Or 0 -OZ~ -08

Oo(C.m-z3

• 0 2 / " 0104M 0 H2 /

(

,jc~~i(1~M

pH U 7

B

9 10 .O6 "02 0 -.02 -'06 -'10 -'lk: -'1~ -'2;

Oo(C,m-2)

t

, ?',,, 5 6 \ ~ 8 9 o o > pH 15

FIG. 8. Surface charge as a function of pH in KNOa

solutions for TiO2 DP 25. Concentrations indicated as parameters to the curves. (O) TiO2/O2, (O) TiO2/H2.

FIG. 10. Surface charge vs pH in 10 -2 M KNO3, for

TiO2 DP 25 (O) after Soxhlet extraction; ( e ) after Soxhlet extraction and heating in oxygen.

Stability toward Coagulation

Figure 12 shows values

ofE2/Eo

vs pH for TiO2 DP 25/02 (Fig. 12a) and TiO2 DP 25/

HE

(Fig. 12b). No distinct difference is found.

DISCUSSION

The most striking result of this study ap- pears to us the absence of any influence of reduction by H: on a0 and on a s, if pure TiO2 is employed. The purity of the TiO2 DP 25 is

30L I ~

"°:~ ~H °H2

, J , "~k ~ ' ' > pH

0 ~ S 6 ~ . 7 8 9 10

-10 ~ ' o ~ ~

-30 "~ "q'

FIG. 9. Zeta potential vs pH for TiO2 DP 25. KNO3

solutions; concentrations indicated as parameters to the curves. (O) TIO2/O2, (O) TiO2/H2.

attested both by ESR data and by the coinci- dence of PZC and IEP after Soxhlet extraction.

The stability toward coagulation (Fig. 12) shows, that the Hamaker constant is not in- fluenced either by the difference in pretreat- ment; at least not sufficiently to lead to dis- tinctly different coagulation kinetics. This is remarkable because TiO2/H2 contains a con- siderably increased donor concentration in comparison with TIO2/O2 (cf. the ESR data). The increased donor concentration in TiO2/

H2

could be expected to lead to a larger po- larizability of TiO2, which in turn would lead to a larger Hamaker constant; such an influ- ence, however, was not found. Thus we con-

t,O 30 20 10 0 -10 -20 -30

(mY)

% ~ > pN

FIG. 1 1. Zeta potential vs pH in 10 -~ M KNO3. Con- ditions as in Fig. 10.

(8)

514 JANSSEN AND STEIN E2iE° 1-2 / \° o ,~ I E P_~ H 0 ~ ' 6 8 0.8 j 0.4 0 . . . . l :p,", L, 6 8 10

FIG. 12. Stability toward coagulation in 10 -2 M KNO3. (a) TiO2 DP 25/02, (b) TiO2 DP 25/H2.

clude that the colloid chemical parameters in- vestigated are insensitive toward changes in solid state character as effected by pretreat- ment in 02 or Hz.

The colloid chemical parameters are de- pendent, however, on impurities. This refers predominantly to the surface charge on TiO2 M808 (Figs. 2-4) showing a shift toward more positive values on reduction. A possible ex- planation is that SO24 - ions are reduced upon treatment with H2, to S 2- which are more strongly specifically adsorbed than the SO4 z- ions. Another explanation, viz. that the effect is due to foreign cations (e.g., Fe 3÷) which are reduced to a lower valency state (Fe 2÷) during H2 treatment appears to be less attractive; it would involve the assumption that Fe 3+ is de- sorbed to a larger extent than Fe 2+.

The influence of reduction on the foreign ions is apparent only in or0, not in ~rr. Clearly, any effect of differences in specific adsorption is compensated by additional adsorption of counterions, or desorption of coions behind the electrokinetic slipping plane.

On pure TiO2, neither in KC1 nor in KNO3 solutions any net specific adsorption occurs (see Figs. 8 and 9). The simplest explanation

is that neither K ÷ or C1- nor NO~ is adsorbed on TiO2 to an extent visible in the colloid chemical properties.

CONCLUSION

Surface charge, charge behind the electrokinetic slipping plane and H a m a k e r constant are insensitive toward changes in donor concentration in TiO2. Any effect produced by an oxidizing or a reducing pretreatment appears to be due to the reduction o f impurities.

ACKNOWLEDGMENTS

The authors gratefully acknowledge the very valuable contributions of W. Smit to this investigation, both with regard to assistance in experiments and with regard to discussions. F. N. Hooge, Th. G. M. Kleinpenning, and L. K. J. Vandamme also added valuable contributions to the discussions.

REFERENCES

1. Kohnke, E., J. Phys. Chem. Solids 23, 1557 (1962). 2. Cronemeijer, D. C., Phys. Rev. 87, 876 (1952). 3. Breckenridge, R. G., and Hosler, W. R., Phys. Rev.

91, 793 (1953).

4. Harris, L. A., and Schumacher, R., J. Electrochem. Soc. 12, 1186 (1980).

5. Chester, P. F., J. Appl. Phys. 32, 2233 (1961). 6. Kr6ger, F. A., "The Chemistry of Imperfect Crystals."

North-Holland, Amsterdam, 1964.

7. Trimbos, H. F. A., and Stein, H. N., 9". Colloid Inter- face Sci. 77, 386 (1980).

8. Janssen, M. J. G., "The Titanium Dioxide/Electrolyte Solution Interface," Ph.D. thesis. Eindhoven, 1984. 9. Bates, R. G., "Determination ofpH," p. 322. Wiley-

Interscience, New York, 1973.

10. Hiemenz, P. C., "Principles of Colloid and Surface Chemistry," p. 462. Dekker, New York/Basel, 1977.

11. Healy, T. W., and Jellet, V. R., J. Colloid Interface Sci. 24, 41 (1967).

12. Logtenberg, E. H. P., "The Relation between the Solid State Properties and the Colloid Chemical Behav- iour of Zinc Oxide," Ph.D. thesis. Eindhoven,

1983.

13. Logtenberg, E. H. P., and Stein, H. N., J. Colloid In- terface Sci. 104, 258 (1985).

14. van Diemen, A. J. G., and Stein, H. N., J. Colloid Interface Sci. 96, 150 (1983).

(9)

TiO2/ELECTROLYTE INTERFACE, I 515 15. Iyengar, R. D., and Codell, M., Adv. Colloid Interface

Sci. 3, 365 (1972).

16. B6rub6, Y. G., and de Bruyn, Ph. L., J. Colloid In- terface Sci. 27, 305 (1968).

17. Schindler, P. W., and Gamsjager, H., Kolloid Z. Z. Polym. 250, 759 (1972).

18. Morterra, C., Chiorino, A., and Boccuzzi, F., Z. Phys.Chem. Neue Folge 124, 211 (1981). 19. Fukuda, H., and Miura, M., J. Sci. Hiroshima Univ.,

Ser. A: Math. Phys. Chem. 36, 77 (1972). 20. Lyklema, J., Croat. Chem. Acta 43, 249 (1971). 21. Lyklema, J., Discuss. Faraday Soc. 52, 302 (1971). 22. Hunter, R. J., "The Zeta Potential in Colloid Science,"

p. 70. Academic Press, New York, 1981. 23. Furlong, D. N., and Parfitt, G. D., J. Colloidlnterface

Sci. 65, 548 (1978).

24. Herrmann, M., and Boehm, H. P., Z. Anorg. Chem. 368, 73 (1969).

25. Tschapek, M., Wasowski, C., and Torres Sanches, R. M., J. Electroanal. Chem. 74, 167 (1976). 26. Nechaev, E. A., and Shein, V. N., Kolloidn. Zh. 41,

361 (1979).

27. Cornell, R. M., Posner, A. M., and Quirk, J. P., J. Colloid Interface Sci. 53, 6 (1975).

28. Ahmed, S. M., and Maksimov, D., J. Colloid Interface Sci. 29, 97 (1969).

29. Maraburi, A. M., and Barabaro, M., Int. J. Miner. Process. 7, 159 (1980).

30. Foissy, A., M'Pandou, A., Lamarche, J. M., and Je- ferezic-Renault, N., Colloids and Surfaces 5, 363 (1982).

31. Loeb, A. L., Wiersema, P. H., and Overbeek, J. Th. G., "The Electrical Double Layer around a Spherical Particle." MIT Press, Cambridge, Mass.,

1961.