Study of the Scale Formation Mechanism on Gold ... - ACS Publications

Jan 14, 2013 - Centre Recherche et Développement Arvida, Rio Tinto Alcan, Jonquière, Québec G7S 4K8, Canada. Langmuir , 2013, 29 (5), pp 1395–139...
0 downloads 0 Views 283KB Size
Article pubs.acs.org/Langmuir

Study of the Scale Formation Mechanism on Gold Modified with an Alkanethiol Monolayer Maxime Clément,*,† Émilie St-Jean,† Nicolas-Alexandre Bouchard,‡ and Hugues Ménard† †

Département de Chimie, Université de Sherbrooke, Sherbrooke, Québec J1K 2R1, Canada Centre Recherche et Développement Arvida, Rio Tinto Alcan, Jonquière, Québec G7S 4K8, Canada



ABSTRACT: Scaling is a problem in many industrial processes. To control and minimize it, it is important to understand the dynamics of the scale formation. In this paper, the scale formation was examined on two kinds of gold surfaces. One was a pure metallic gold surface, and the other was a gold surface modified with an alkanethiol self-assembled monolayer. A series of surface characterization experiments were performed to ensure a good understanding of the gold−thiol bond stability in a caustic solution.



INTRODUCTION In many industrial processes, scale formation represents a major problem. For example, in the Bayer process used to recover aluminum oxide from bauxite, the whole circuit operates from saturated to supersaturated concentration for many of its components, such as aluminate, silicate, and oxalate. Scale formation occurs in areas of the process where aluminate and silicate have high supersaturation, and it has a major impact on alumina production costs. It decreases the efficiency of heat exchangers,1 reduces the flow in the lines, increases the maintenance costs, and lowers the plant productivity. This explains why numerous researchers have been challenged by scaling problems over the last few decades.2−6 These researchers explained scale formation phenomenon using different techniques or hypotheses. Our group investigated some mechanisms to explain why the gibbsite scale is present on metallic surfaces. Breault and Bouchard7 demonstrated that, in a dynamic system, the passivating layer on the surface of mild steel or copper is responsible for the scale adherence to the metallic surface and that scaling can be delayed by a cathodic protection. Brisach et al.8 worked in a static system and suggested an apparent global nucleation mechanism in solution, which combines cementation and sudden particle showering. The objective of our group’s study was to improve the understanding of the aluminum hydroxide scale formation in the Bayer process. Therefore, we studied its scaling rate on various metals, such as mild steel, copper, gold, platinum, and palladium.9 Noble metals, such as gold, platinum, or palladium, do not form surface oxides or hydroxides as easily as iron in Bayer conditions, according to their Pourbaix10 diagrams. However, our previous results showed significant scale formation on these metals, which suggests that the scale formation is not influenced by the chemical surface © 2013 American Chemical Society

composition. To confirm this hypothesis, we decided to use the experiment performed on the gold surface as a base model to understand the details of the metal/solution interactions. To achieve this purpose, we measured the scaling time on two types of gold surfaces: the first one is a pure metallic gold surface, and the second one is a gold surface modified with an alkanethiol self-assembled monolayer.11−14 The non-modified gold surface is then considered as our reference, and it represents a model material because it is immune in our working conditions.10 The techniques used to investigate the dynamics of the scaling process consist mainly of gravimetric methods that follow the mass gain of a sample material during a certain period of time. By plotting the mass gain as a function of time, we can then determine the scaling time of gibbsite onto the sample material in the specific conditions of the experiment.15,16



EXPERIMENTAL SECTION

Nucleation Time Measurements. The scaling time is measured using different types of balances coupled with automated acquisition systems that follow the mass gain as a function of time. The first type of balance is an interfaced analytical balance used as previously described by St-Jean et al.15 The second type of balance is a quartz crystal microbalance (QCM) and is used as previously described by Brisach et al.8 In this case, the acquisition system collects frequency data as a function of time. Diminutions in resonance frequency are attributed to changes in mass because of aluminum trihydroxide nucleation and crystallization deposition onto the inner gold electrode. Received: July 7, 2012 Revised: September 6, 2012 Published: January 14, 2013 1395

dx.doi.org/10.1021/la302743s | Langmuir 2013, 29, 1395−1399

Langmuir

Article

Figure 1. Scale formation study over time on a QCM. The relation between the frequency variation (Δf) and mass gain (Δm) is given by the Sauerbrey17 relation (eq 1). Δf = − Cf Δm

was used on all samples, and the binding energy scale was calibrated to Au 4f (83.96 eV).



(1)

RESULTS AND DISCUSSION The scaling times on the different gold surfaces were measured using two types of balances: QCMs and interfaced analytical balances. Figure 1 shows the mass gain profile obtained as a function of time on the QCM with the pure gold surface and with the modified gold surface. For each kind of surface, two series of experiments were performed to estimate the method reproducibility. The results obtained on the QCM show a small increase of the nucleation time. However, this increase is negligible in comparison to the long nucleation time on pure gold. Figure 2 shows the mass gain profile obtained as a function of time on the interfaced analytical balances. This method has been used in many of our group’s studies,15,16 and its reproducibility has already been tested. The results obtained on the interfaced analytical balance show that there is no difference in the nucleation time between the gold sample and the modified gold sample.

The correlation coefficient Cf can be estimated by means of silver or copper electrodeposition on the working electrode and by quantifying Δm from Faraday’s law. Theoretically, Cf = 5.4 × 10−9 g cm−2 Hz−1 at 10 MHz. In our case, no correlation between Δf and Δm has been performed. Synthetic Bayer Liquor. The supersaturated sodium aluminate solution was prepared using 150 g of NaOH pellets (Fisher Scientific), 40 g of Na2CO3 (Fisher Scientific), and 180 g of gibbsite [Al(OH)3] supplied by Rio Tinto Alcan and used as received. The total volume was 1 L (distilled water), and the solution was prepared in a pressure reactor (Parr 4843) at 150 °C. The solution was then cooled to 65 °C to give an alumina/caustic ratio of 0.62. Formation of Alkanethiol Self-Assembled Monolayer on Gold. It has long been known that thiols can be used effectively to form a self-assembled monolayer on the surface of gold.11−14 For nucleation time measurements, gold foils (0.1 mm thick, 99.95%, 25 × 25 mm, Alfa Aesar) and gold-covered quartz crystal resonators (ICM) are treated with a 1-octanethiol (98.5+%, Aldrich) solution in hexane (Fisher Scientific). For the surface characterization experiments, gold foils are treated with a 4-methylthiophenol (Oakwood Products, Inc.) solution in cyclohexane (Anachemia). In all cases, an excess of thiol is used to make sure that the gold surface is covered by a complete monolayer. These surface modifications are carried out by dipping the gold surface into the alkanethiol solution and mixing for 2 h to ensure better organization of alkyl chains at the surface of the metal. The following steps consist of cleaning with hexane or cyclohexane and drying the samples. All chemicals were used as received. Caustic Treatment. The gold sample treatment was performed in a 5 N NaOH solution prepared from 10 N NaOH (Fisher Scientific). Micro-Raman Analysis. Phonon spectra with 0.5 cm−1 resolution were recorded using a micro-Raman spectrometer in backscattering configurations with the 632.8 nm He−Ne laser line and a liquidnitrogen-cooled charge-coupled device (CCD) detector.18,19 The vibration frequency was calibrated with the plasma ray at 180 cm−1. X-ray Photoelectron Spectroscopy (XPS) Analysis. The XPS spectra were recorded in UHV (