Deposition of colloidal hematite onto mercury from water-ethanol mixtures

ARTICLE

Deposition of colloidal hematite onto mercury from water-ethanol mixtures

Estela María Andrade^a, Fernando Víctor Molina^{a *} , and Dionisio Posadas^b

^aINQUIMAE, Departamento de Química Inorgánica, Analitica y Química Física, Facultad de Ciencias Exactas y Naturales, UBA, Ciudad Universitaria, Pabellón II, 1428, Buenos Aires Argentina

^bDepartamento de Química, Facultad de Ciencias Exactas, Instituto de Investigaciones Fisicoquímicas Teóricas y Aplicadas (INIFTA), UNLP, Suc. 4, CC 16, 1900 La Plata, Argentina

Received: July 25, 1996

A aderença de partículas de hematita (a Fe₂O₃) sobre eletrodos de mercúrio foi estudada em misturas de água-etanol através de contagem sobre imagens de um microscópio ótico. Quando o conteúdo de etanol é pequeno, o número de partículas aderidas decresce com o aumento da concentração de etanol. Para uma fração molar de etanol de 0,2, o número de partículas passa por um mínimo, aumentando para maiores concentrações de etanol. Modificando o potencial eletródico, observa-se que as curvas número de partículas vs. Fração molar de etanol pode ser explicada com base na dependência com a composição do solvente das energias de interação partícula/solução e metal/solução.

The adherence of hematite (aFe₂O₃) particles onto mercury electrodes in water-ethanol mixtures has been studied by counting using optical microscope images. The number of attached particles, when the ethanol content is small, decreases as the ethanol concentration increases. At a mole fraction of ethanol near to 0.2, the number of particles goes through a minimum and then it increases with ethanol concentration. When the electrode potential is modified, curves of the number of particles vs. ethanol mole fraction with the same shape are found, but which cross each other.

The dependence on the ethanol concentration can be explained based on the dependence of particle/solution and metal/solution interaction energies with the solvent composition.

Keywords: hematite adhesion, metal surface, water-ethanol mixtures

Introduction

The attachment of colloidal particles onto metallic surfaces is a field relevant to many applied problems, such as corrosion protection, biofouling and semiconductor manufacturing, to mention only a few. However, it has not received much attention in fundamental studies. There are some published papers on adherence of particles to surfaces (usually non conducting)^1-6 and to different particles^7-12, where the surface electric potential cannot be easily modified. In our group we have studied the attachment of hematite particles onto mercury and silver surfaces under different conditions^13-16. Theoretical work has been mainly devoted to extension of the DLVO theory of colloid stability¹⁷ to heterocogulation (coagulation of dissimilar particles)^18-20 or to the kinetics of deposition^21-23.

Mixed solvent media has been used to study the ionic equilibrium at the hematite/solution interface²⁴. In this work the influence of the solvent media is studied, by measuring the number of hematite particles attached onto mercury surfaces from suspensions in water-ethanol mixtures, using NaClO₄ as the supporting electrolyte. The number of particles was measured by counting using video images, as a function of ethanol concentration at different applied potentials.

Experimental

Preparation of hematite suspensions

Hematite was synthesized as previously described^13-15. The particle surface charges in water/ethanol media were determined by potentiometric titration^24-25. A fixed amount of HClO₄ was added to a suspension of 1.5 g dm^-3 hematite, and titrated with NaOH. The resulting curve was compared with that of the blank electrolyte in order to obtain the isoelectric point and the charge - pH curve.

Oxide particles (80 mg dm^-3) were suspended in 0.01 M NaClO₄ solutions in water/ ethanol mixtures for the deposition experiments. All reagents used were analytical grade. Water was obtained from a Milli-Q system and subsequently distilled with a sub-boiling point apparatus.

Cell and apparatus

A cell with three compartments was employed for the deposition experiments on mercury film electrodes which allowed the separate placement of a reference, an auxiliary, and the working electrode. A saturated calomel electrode (SCE) was used throughout as reference. Potential values in this work are refered to the SCE. A PAR Model 273 Potentiostat was used for the experiments.

Hematite on mercury

Mercury films were supported on silver discs of 0.5 mm diameter as it has been described before¹³. The colloid deposits were performed in water/ ethanol mixtures as follows:

i. The electrode potential was held for 30 min at -1.0 and -0.3 V; it has been found¹⁴ that these potentials correspond to the cathodic maximum and to a minimum of hematite deposition from water suspensions at pH = 5. Furthermore, as the zero charge potential of mercury in aqueous media is around -0.5 V³³, the potential values chosen should correspond to negatively and positively charged metal surfaces, respectively. All the experiments were done at pH = 5, as referenced to aqueous solution. In water-ethanol mixtures, the correction for the Born free energy of transfer between solvents was applied, as detailed by Heisleitner et al.²⁴. This assured that the particles were in comparable charge states in the different mixtures.

ii. The electrode was placed "face up" at the bottom of the cell. When the deposition time was over, it was removed from the cell, rinsed carefully and dried. It was then observed under an optical microscope, Leitz DM RX, equipped with a video camera. Images were captured and processed using the Jandel Scientific MOCHA image analysis software.

Results

Figure 1 shows a typical titration curve of hematite in water-ethanol medium. It is in general agreement with those obtained by Hesleitner et al.²⁴. The isoelectric point, i.e.p., was determined from the s vs. pH plot, and found to be 7.4, the same as in the absence of ethanol¹³. The relative independence of the i.e.p. with ethanol concentration was also found by Hesleitner et al.²⁴.

In Fig. 2 typical video images obtained in the experiments are shown; in each experiment, several images were processed and their results averaged. Figure 3 presents the results obtained for the number of particles, N_d, as a function of ethanol mole fraction, x_Et at the two potentials studied. N_d spans a wide range and therefore is plotted logarithmically. It is observed that, as the ethanol content increases, a strong decrease in N_d occurs initially followed by a slower increase after reaching a minimum. At the two potentials the behavior is similar, but in the region of the minimum N_d an inversion is observed: the N_d at -0.3 V is higher whereas in the extremes the opposite holds. It should be noted that at the two potentials studied the metal is expected to have charges of different sign, but no significant difference in the number of particles is found, indicating that the electrostatic interactions should not play a significant role in the observed behavior.

Discussion

The experimental results show that the attachment of hematite particles onto mercury surfaces has a relatively complex behavior. It shows an increase in the number of attached particles above an ethanol mole fraction of about 0.2. This fact can be qualitatively explained on the basis of the changes of metal/solution and particle/solution interactions with x_Et. To that end we will assume, in a general way, that for particle/metal adhesion, the number of attached particles will increase as the (electrochemical) free energy of adhesion, DG_adh, decreases. Then, we will consider its dependence with solvent composition.

The free energy change per unit area when two different surfaces (in this case metal, M, and particle, P) immersed in a third medium (solution, S) attach to each other is given by²⁶:

where g_MP, g_PS and g_MS are, respectively, the metal/particle, particle/solution and metal/solution surface energies. Each of these is, in turn, the sum of several contributions, namely Van der Waals (VW) forces (including dispersion and polarization), diffuse layer electrostatic interaction (el), chemical (chem), and image forces (im)^{26, 27}, so that

where i = MP, PS or MS. Then, DG_adh should, in general, be affected by all these contributions. The VW and electrostatic interactions are well known because they are considered in the colloid homo and heterocoagulation theories^17,18,20. When the solvent is changed, all the contributions to g_PS and g_MS are expected to be modified, whereas g_MP (the interfacial tension of metal and particle in close contact) can be regarded as constant. Then, the dependence of DG_adh on the ethanol mole fraction can be expressed as

where DG'_adh(x_Et) is the solvent dependent part of DG_adh(x_Et). We will now consider the contributions to DG'_adh(x_Et) to estimate its variation with x_Et. The behavior of g_MS is well known for mercury/aqueous solution interfaces. It has been studied for water-ethanol mixtures by Ockrent²⁸ using NH₄NO₃, and it shows the usual inverted parabola shape, with the maximum being lowered and shifted anodically as the ethanol concentration increases.

On the other hand, g_PS is less known. Following Israelachvili²⁶, it can be considered as the free energy change when a unit area of particle-particle surface is taken apart and brought in contact with the solution (Fig. 4), thus forming a unit area of particle-solution interface. This process is the opposite of particle coagulation, so that it can be written

DG^el and DG^VW are the double layer and Van der Waals contributions to the energy change when two identical, plane surfaces are brought to contact from infinity; these are the interaction energies between identical particles, which form the basis of DLVO and related theories of colloid particle interactions. Their dependence with the solvent composition can be obtained from the respective equations as follows.

The Van der Waals interactions

The Van der Waals interaction energy between two plane parallel surfaces at a distance D apart is given by²⁹:

so that DG^VW is the energy difference between infinity and a minimum contact distance D₀, which can be taken as 0.165 nm²⁹, that is

In Eqs. 5 and 6 A, known as the Hamaker constant, is approximately given by

where e_P and e_S are, respectively, the particle and solvent static dielectric constants, and e_P(in) and e_S(in) are approximately given by³⁰:

where n is the refractive index and n_e the main electronic absorption frequency. In Eq. 7 the integration is carried out in the optical frequency range, so the lower limit is usually taken as²⁹n₁» 4 x 10¹³ s^-1. Eq. 7 includes the zero frequency interaction, with the Debye and Keesom contributions (first term) as well as the London dispersion energy (second term). n_e for water and ethanol can be taken²⁹ as 3.0 10¹⁵ s^-1 and for hematite, which absorbs in the visible range, n_e» 5.510¹⁴ s^-1.

From Eqs. 7 and 8 it is clear that the Van der Waals interaction depends on the static dielectric constants and the refractive indices of both media, particle and solvent. For hematite, e_{P = 12} and n_P = 3.05³¹. For water-ethanol mixtures, using the available data³², the dielectric constant can be approximately expressed, in the range 0 £ x_Et£ 0.4, as ep = 78.5 - 59.5(x_Et)^1/2. On the other hand, as there is very little difference between the refractive indices of water (1.333) and ethanol (1.361), the n_S values are linearly interpolated.

Figure 5 shows the resulting DG^VW as a function of ethanol mole fraction. A slight increase with x_Et is observed.

The electrostatic interactions

The subject of colloidal particle electrostatic interactions has been widely studied^17-20,23. Hogg et al.¹⁹ gave the following approximate analytical expression, valid for low surface potentials, of the free energy change for two plane parallel double layers which are brought from infinity to a distance D apart:

where Y₀ is the surface potential, e_o the vacuum permitivity and k is the inverse Debye-Hückel length, given for a 1-1 electrolyte by

c being the electrolyte concentration, and F, R and T have their usual meaning.

The surface potential Y₀ is a quantity relatively difficult to obtain. Instead, the so called z potential is customarily obtained from electrokinetic measurements. z is the potential value at the slipping plane in the electrolyte surrounding the particle. Heisleitner et al.²⁴ have measured the z potential of hematite in water-ethanol mixtures, and gave an estimation of the distance from the outer Helmholtz plane to the slipping plane in the order of 1.2 nm. From these data, approximate values for Y₀ can be obtained. With these values in turn, DG^el at D = D₀, gives a dependence with x_Et as shown in Fig. 6.

The other contributions to g_PS, DG^chem and DG^im are not known, and initially they will be considered negligible, as is usually done^17-20,27.

The dependence of DG_adh on solvent composition

From Eqs. 1, 3, 4 and the above considerations, it can be written that

Using data from Ref. 28 for g_MS, the adhesion free energy dependence on x_Et is depicted in Fig. 7. It is observed that as the ethanol mole fraction is increased, DG_adh increases reaching a plateau at x_Et» 0.25. This is in qualitative agreement with the experimental findings. Mainly, the increase in the first part of the curve approximately matches the experimental decrease of N_d in this range. The decrease of DG_adh at higher x_Et (and the corresponding increase in number of particles) can be attributed to a higher instability of the particles in a medium of low dielectric constant.

Although not generally considered, there should be a contribution of DG^chem to this effect. Because hematite is an oxide, the presence of hydrogen bonding between water and the particles should be expected. As the ethanol (which has less capability to form hydrogen bonds than water) content increases, the number of such bonds should diminish, and so diminishing the stability of the particles in suspension, thus favoring the deposition.

On the other hand, the above calculations do not explain the observed crossing of the N_dvs. x_Et curves at different potentials. This is a difficult point to explain, because the electrode potential is assumed to affect g_MP and g_MS but not g_PS. The dependence of g_MS with the ethanol concentration is monotonic and cannot lead to the behavior of Fig. 2. Although Ockrent²⁸ used a different electrolyte than ours, it is unlikely that this difference could affect strongly the results. A possible explanation is that g_MP could be solvent dependent, in addition to potential dependent. This might be due to a small amount of solvent being retained upon particle adhesion.

As it is mentioned above, the metal - particle electrostatic interactions do not appear to have significant influence in the observed behaviour, because the hematite particles have a positive charge at the pH studied here, whereas the metal has in one case negative charge and positive charge in the other case. The differences observed do not seem to be related to such interactions. In fact, the calculations show that g_MS is the dominant term in DG'_adh, that is, the energy gained in diminishing the metal-solution interfacial area seems to overcome other contributions.

Acknowledgments

Grateful acknowledgement is made to the Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET) de la República Argentina for support of this work. E.M.A. and F.V.M. are also indebted to the Grupo de Electroquímica and to INQUIMAE for partial support of this work.

• 1.Marshall, J.; Kitchener, J. J. Colloid Interf. Sci. 1966, 22, 342.
• 2.Hull, M.; Kitchener, J. Trans. Far. Soc. 1969, 65, 3093.
• 3.Usui, S.; Yamasaki, T. J. Colloid Interf. Sci 1969, 29, 629.
• 4.Dabros, T.; van de Ven, T.G.M. Colloid Surf. A: Physic. Eng. Aspects 1993, 75, 95.
• 5.Sjollema, J.; Buscher, H.J. J. Colloid Interf. Sci. 1989, 132, 382.
• 6.Ryde, N.; Kihira, H.; Matijevic, E. J. Colloid Interf. Sci 1992, 151, 421.
• 7.Urrutia, G.; Pasaggio, S.; Maroto, A.; Blesa, M.A. Nucl. Sci. Eng. 1983, 84, 120.
• 8.Kallay, N.; Barouch, E.; Matijevic, E. Adv. Colloid Interf. Sci. 1987, 27, 1.
• 9.Kuo, R.J.; Matijevic, E. J. Colloid Interf. Sci. 1980, 78, 407.
• 10.Caselli, M.; Lippolis, G.; Gierst, L. J. Electroanal. Chem. 1972, 38, 451.
• 11.Cohen, H.; A. Soffer Y.; Oren, J. Colloid Interf. Sci. 1987, 120, 272.
• 12.Alince, B. Colloids Surf. 1989, 34, 39.
• 13.Andrade, E.M.; Molina, F.V.; Posadas, D. J. Colloid Interf. Sci. 1994, 165, 450.
• 14.Andrade, E.M.; Molina, F.V., Gordillo, G.J.; Posadas, D. J. Colloid Interf. Sci. 1994, 165, 459.
• 15.Andrade, E.M.; Gordillo, G.J.; Molina, F.V.; Posadas, D. J. Colloid Interf. Sci. 1995, 173, 231.
• 16.Andrade, E.M.; Molina, F.V.; Posadas, D. J. Colloid Interf. Sci. 1995, 176, 450.
• 17.Verwey, E.; Overbeek, J.T. In Theory of the Stability of Lyophobic Colloids; Elsevier, Amsterdam, 1948.
• 18.Derjaguin, B. Disc. Faraday Soc. 1954, 18, 85.
• 19.Hogg, R.; Healy, T.; Fuerstenau, D. Trans. Far. Soc 1966, 62, 1638.
• 20.Usui, S.In Progress in Surface and Membrane Science; Danielli, J. F. Ed.; 5, 223, 1972.
• 21.Adamczyk, Z.; Siwek, B.; Zembala, M. Colloid Surf. 1992, 62, 119.
• 22.Adamczyk, Z.; Siwek, B., Zembala, M.; Belouschek, P. Adv. Colloid Interf. Sci. 1994, 48, 151.
• 23.Adamczyk, Z.; Warszynski, P. Adv. Colloid Interf. Sci. 1996, 63, 41.
• 24.Hesleitner, P.; Kallay, N.; Matijevic, E. Langmuir 1991, 7, 178.
• 25.Regazzoni, A.E. Ph. D. Thesis, Buenos Aires, 1984.
• 26.Israelachvili, J.N. In Intermolecular and Surface Forces; Academic Press, London, 1985, ch. 14.
• 27.Bowling, R.A. In Particles and Surfaces; Mittal, K.L. Ed.; Plenum Press, New York, 1988.
• 28.Ockrent, C. J. Phys. Chem. 1931, 35, 3354.
• 29.Israelachvili, J.N. In Intermolecular and Surface Forces; Academic Press, London, 1985, ch. 11.
• 30.Mahanty, J.; Ninham, B.W. In Dispersion Forces; Academic Press, New York, 1976.
• 31.CRC Handbook of Chemistry and Physics; Lide, D.R. Ed.; CRC Press, Florida 1995.
• 32.International Critical Tables; Washburn. ,E. Ed.; McGraw-Hill, New York, 1929 Vol. IV.
• 33.Delahay, P. In Double Layer and Electrode Kinetics; Interscience, New York, 1965.

Publication Dates

• Publication in this collection
  19 Nov 2010
• Date of issue
  1997

History

• Received
  25 July 1996