BCCMS  /  Groups  /  CMS  /  Research  /  Metal-oxid surf.
English  /  Deutsch

Computational studies on epoxy adhesion at the surface of native Alumina

Introduction

During the past two decades aluminum has constantly gained importance in technical applications, with its use spreading from aerospace to automotive applications and now covering nearly every area of industrial and consumer appliances. At the same time adhesive technology has achieved advances which enable adhesive bonding of metal parts to nowadays supplement or even replace traditional metal joining technologies such as welding, bolting or riveting. Correlated to this is development of metal-resin or metal-resin-fiber compound materials, which have recently advanced to the point of technical application. The demands for lighter and at the same time stronger materials, in order to build more fuel-efficient, i.e. lighter, vehicles and aircraft without sacrificing structural strength leads to a strong interest in the development and improvement of fiber reinforced aluminum-polymer hybrid materials and adhesive bonding technology for aluminum. For both of these technologies, the aluminum-polymer adhesion is of crucial importance.

Since, outside high vacuum environments, aluminium instantly develops a surface oxide layer, the problem of organic adhesion on aluminum translates into the problem of organic adhesion on native aluminum oxide. The improvement of adhesion technology requires an understanding of the underlying chemical processes of the bonding of organic adhesives to the alumina surface. In this study we aim at finding a suitable methodology for gaining insight into the initial bonding of adhesive molecules as well as the bonding competition between different organic species at the native Al2O3 -surface. To gain a better understanding of these adhesion phenomena, we work to model the binding behaviour of a simple epoxy adhesive system at the surface of alumunim oxide.

Problem Details and recent activities

As a first step towards understanding the adhesion of organic molecules, we investigate the reaction products and -paths of the initial adsorption reactions of the components of our model adhesive at the hydroxylated surface of γ-Al2O3. These components are diglycidylesterbisphenol-A (DGEBA) as the resin, diethyltriamine (DETA) as the hardener and 3-aminopropylmethoxysilane (sold as: Dynasilan AMEO, here referred to as AMEO) as an adhesion promoter component. We calculate the structures and energies of the isolated organic molecules and the adsorbed molecules on the surface model using the self-consistent charge density-functional based tight binding (SCC-DFTB) method, to obtain the reaction energies of the adsorption reactions. Figure 1 shows the calculated products structures of the AMEO adsorption reaction at two different surface sites. Our surface model contains 374 atoms in a 2D-periodic supercell, together with the organic compounds, this leads to about 400 atoms for the image geometries.

Figure 1: Structures of AMEO adsorbed at two different sites of the alumina surface model. (Al: grey, O: red, C: black, N: blue, Si: yellow, the surface model is truncated at the bottom).

We then employ the nudged elastic band (NEB) methodi to calculate the total energies of a chain of intermediate structures, called images, between the educt and product geometries. These images sample the minimum energy path of the given reaction and allow to asses the reaction barriers and kinetics. Figure 2 shows the MEP-s obtained for the adsorption of AMEO. Figures 2 (b) and (c) show the two most important transition states along the adsorption path. We sample between 7 and 11 images along each reaction path.

Figure 2: Interpolated minimum energy paths of the AMEO adsorption (left) and selected image geometries along the adsorption path of AMEO at site 2 (right) (Al: grey, O: red, C: black, N: blue, Si: yellow, +-symbols show calculated image total energies, the red line serves to guide the eye.)

From this data we aim to construct a model of the competition between the organic compounds for the adsorption sites available at the surface.

Additional analyses of the reaction paths allow to asses the ranges of electronic and mechanical disturbance of the surface, induced by the adsorption reactions. Figure 3 shows the root of mean square displacements (RMSD) and maximum charge differences of the atoms during the adsorption of AMEO at site 2.

Figure 3: (left) Movement of the atoms during the adsorption of AMEO at site 2, color coded by RMSD over the entire reaction path. (right) Mulliken charge difference ΔQ=Qmax-Qmin of each atom for the reaction of AMEO at site 2 in units of electrons.

These results will be used to help us in developing a multi-scale coupling method for modeling reactions at organic/inorganic hybrid interfaces, by providing an insight into the influence ranges across the model.

Example adsorption path

This video shows an interpolation between the NEB calculated images along the MEP of the condensatin of a trucated, ring-opened DGEBA molecule at the hydroxylated native Al2O3 surface.