Elsevier

Applied Surface Science

Volume 324, 1 January 2015, Pages 594-604
Applied Surface Science

Quantitative morphological and compositional evaluation of laboratory prepared aluminoborosilicate glass surfaces

https://doi.org/10.1016/j.apsusc.2014.10.132Get rights and content

Highlights

  • Aluminoborosilicate glass surfaces were prepared through both melting and polishing/etching and the surface composition and morphology were quantified as a function of processing method.

  • Glass surface morphology was quantified using PSD analysis, followed by both fractal and ABC model fitting, resulting in a comprehensive description of the spatial distribution of roughness.

  • All melt surfaces showed a depletion in Na, Ca, and B with respect to the bulk composition. Polished/etched surfaces showed a depletion in Na, B, and Al with respect to the bulk composition.

  • It was found that increasing heat treatment temperature of melt surfaces lead to a decrease in equivalent roughness and an increased spatial homogeneity of roughness while etching of polished ISG glass surfaces decreases the roughness and spatial distribution homogeneity of roughness.

Abstract

Surface finishing techniques including polishing, etching and heat treatment can modify the topography and the surface chemical composition of glasses. It is widely acknowledged that atomic force microscopy (AFM) can be used to quantify the morphology of surfaces, providing various parameters including average, peak-to-valley, and apparent root-mean-square roughness. Furthermore advanced power spectral density (PSD) analysis of AFM-derived surface profiles offers quantification of the spatial homogeneity of roughness values along different wavelengths, resulting in parameters including equivalent RMS, Hurst exponent, and fractal dimension. Outermost surface (∼8 nm) chemical composition can be quantitatively measured by X-ray photoelectron spectroscopy. In this paper, we first developed a series of surface finishing methods for an aluminoborosilicate glass system by polishing, etching or heat treatment. The chemical composition and environment of prepared glass surfaces were quantified by XPS and topographical analysis was carried out by fractal and k-correlation model fitting of PSD profiles derived via AFM. The chemical environment of elements, as determined via XPS, present on the prepared surfaces are similar to those within the pristine bulk glass. The compositional evolution of polished and melt surfaces are discussed in context of corrosion phenomena associated with the grinding, polishing, and etching of surfaces and the thermal heat treatment utilized for processing, respectively. Good correlation between surface finishing methods, chemical composition and topographical parameters were observed. More importantly, extensive discussions on topographical parameters including equivalent RMS, Hurst exponent, and fractal dimension are presented as a function of processing method.

Introduction

The surface quality has great impact on the performance properties of many novel and emerging technological applications for glasses including those for displays, biomedical applications, microelectronics and advanced optical lithography systems [1], [2]. In some cases the glass surface morphology and overall roughness has become the threshold for the continued development of novel electronics and portable devices [3], [4]. The surface morphology, chemical composition and homogeneity of glass manufacturing products can influence a wide variety of performance related properties including the mechanical strength and chemical durability [5], [6], [7], [8]. Furthermore, smooth, homogeneous and compositionally reproducible glass surfaces are required for mechanistic investigations associated with glass corrosion, chemical tempering, and thin film coating on glasses [9]. For example studies on the dissolution behavior of silicate glasses suggested surface morphology and composition can greatly affect the dissolution or alteration rate of these glasses [10], [11], [12].

A variety of processing methods can be utilized to engineer glass surfaces including mechanical force, chemical and mechanical polishing, and thermal treatment [9], [13], [14]. Each will affect the glass morphology and composition. For example, the annealing of some glass compositions can deplete or enrich the surface in metallic ions due to evaporation and segregation [15]. In the case of chemo-mechanical polishing, chemical reactions between the glass surface and the polishing media polishing can alter the surface composition relative to the bulk glass composition [16], [17], [18], [19]. These changes in surface morphology and chemistry have been shown to effect mechanical, chemical, and aesthetic properties [5], [6], [7], [8]. There are a variety of characterization tools available for the surface compositional and morphological analysis of glass surfaces, each with their own capabilities and limitations. Of particular interest here is X-ray photoelectron spectroscopy (XPS) for quantification of chemical composition and chemical environment and atomic force microscopy (AFM) for surface topographic imaging and quantitative morphological analysis.

XPS is routinely used to determine both the chemical composition and local chemical environment of elements associated with glass surfaces [20], [21]. In particular, under standard operating conditions, as those used in this work, the XPS probes to a depth of approximately 5–10 nm and therefore is a useful probe of the uppermost surface composition and chemistry. Through the empirical derivation of relative sensitivity factors and high resolution peak area analysis absolute quantification of the glass surface composition is achievable [22]. Furthermore through analysis of peak position and shape the chemical environment of the glass forming elements can also be analyzed [23], [24].

AFM is one of the more common techniques in the qualitative and quantitative analysis of glass surface morphology, given its flexibility, relatively low cost, high lateral resolution, and high sensitivity to topographic features on the angstrom to micron scale [25], [26]. Quantitative roughness analysis measured by AFM is often represented by simple statistical parameters, such as average roughness, apparent root-mean-squared roughness (RMS), or peak-to-valley roughness. In these cases, 250,000 data points (512 × 512 pixels per image for example) are expressed by a single number. However, apparent RMS values are problematic since two drastically different topographies can have the same RMS value [27], [28]. This is due to the fact that apparent RMS values are only sensitive to z-axis (vertical) height deviation, not x, y-axis (horizontal) structures. Therefore, such measurements are greatly dependent on the homogeneity of the surface scanned and can be quite problematic, in some cases leading to the difficulty in understanding surface roughness and its spatial distribution homogeneity.

Rather than using the simple above mentioned statistical parameters, surfaces can also be represented by power spectral density (PSD) functions over different spatial frequency regions. PSD is advantageous as it allows the comparison of roughness data measured at different spatial frequencies, offering a convenient representation of the spatial distribution and homogeneity of roughness [25]. This is realized through a 2D fast Fourier transformation algorithm allowing correlation of the z-axis height deviation with the x-, y-axis location data in real and reciprocal space. Furthermore, from the PSD profiles a series of spatially sensitive quantitative roughness parameters can be derived including the fractal dimension, Hurst exponent, correlation length, and equivalent RMS roughness [26].

In this paper we provide a systematic method to prepare smooth glass melt and polished surfaces with surface compositions similar to that of the bulk. Glass surface composition and morphology were quantified using complimentary surface sensitive characterization tools including XPS and AFM. In particular, both vertical and spatial distribution of roughness was investigated using advanced PSD analysis and for the first time, we report spatially sensitive roughness parameters of a variety of glass surfaces as a function of processing.

Section snippets

Glass melting and bulk glass composition analysis

The aluminoborosilicate glass used in this study, referred to as international simple glass (ISG), is a reference waste glass composition developed and utilized by a 6-nation collaborative effort in examining nuclear wasted glass corrosion [29]. The ISG glass used in this study was commercially melted by Mo-Sci Corporation, with melting procedures documented in detail elsewhere and briefly described here [30]. The ISG glasses were batched to yield a nominal weight% composition of 56.2% SiO2,

Bulk glass composition

Table 1 shows the bulk ISG glass elemental composition measured from ICP-AES digestion. The atomic concentration of Na, Ca, Zr, Al, B, Si and O in the bulk ISG glass is 8.0%, 1.6%, 0.4%, 2.4%, 9.4%, 18.0% and 60.1% respectively. Minor deviations, below 0.4%, compared to the ISG bulk glass composition (Na = 7.6%, Ca = 1.7%, Zr = 0.5%, Al = 2.3%, B = 9.6%, Si = 18.0% and O = 60.1%) as reported by Savannah River National Laboratory [30], were observed.

Thermal analysis of bulk ISG glass

Fig. 1 shows the dilatometry/thermal expansion curve and the

Conclusions

In this paper, ISG glass surfaces were prepared through both melting and polishing/etching and the surface composition and morphology were quantified as a function of processing method. In particular, for the first time glass surface morphology as a function of processing method was quantified using PSD analysis, followed by both fractal and ABC model fitting, resulting in a comprehensive description of the spatial distribution of roughness.

All melt surfaces showed a depletion in Na, Ca, and B

Acknowledgements

Funding for this work was provided by the Department of Energy Nuclear Engineering University Program (DOE-NEUP) program under contract number DE-AC0705-ID14517. The authors would like to thank Vince Bojan of The Pennsylvania State University Materials Characterization Laboratory for XPS analyses. We acknowledge Melanie Saffer of The Pennsylvania State University Laboratory for Isotopes and Metals in the Environment for bulk ICP analysis.

References (52)

  • R. Bruckner et al.

    XPS Measurements and Structural Aspects of Silicate and Phosphate-Glasses

    J. Non-Cryst. Solids

    (1980)
  • M. Senthilkumar et al.

    Characterization of microroughness parameters in gadolinium oxide thin films: A study based on extended power spectral density analyses

    Appl. Surf. Sci.

    (2005)
  • H.W. Guo et al.

    Preparation of high strength foam glass-ceramics from waste cathode ray tube

    Mater. Lett.

    (2010)
  • P.K. Ojha et al.

    Investigation of crystallization kinetic of SrO-La2O3-Al2O3-B2O3-SiO2 glass and its suitability for SOFC sealant

    Intl. J. Hydrogen Ener.

    (2011)
  • Y. Saito

    Mechanism of the etching rate change of aluminosilicate glass in HF acid with micro-indentation

    Appl. Surf. Sci.

    (2008)
  • L.S. Du et al.

    Network connectivity in aluminoborosilicate glasses: A high-resolution B-11, Al-27 and O-17 NMR study

    J. Non-Cryst. Solids

    (2005)
  • K.A. Murphy et al.

    Solid-state NMR examination of alteration layers on nuclear waste glasses

    J. Non-Cryst. Solids

    (2013)
  • L. Wong et al.

    The effect of HF/NH4F etching on the morphology of surface fractures on fused silica

    J. Non-Cryst. Solids

    (2009)
  • A.M. Hu et al.

    Effect of nucleation temperatures and time on crystallization behavior and properties of Li2O-Al2O3-SiO2 glasses

    Mater. Chem. Phys.

    (2006)
  • N. Khan et al.

    The origin of the exothermic peak in the thermal decomposition of basic magnesium carbonate

    Thermochimica Acta

    (2001)
  • S.B. Shim et al.

    Wetting and surface tension of bismate glass melt

    Thermochimica Acta

    (2009)
  • P.K. Gupta et al.

    Nanoscale roughness of oxide glass surfaces

    J. Non-Cryst. Solids

    (2000)
  • P.L. Bocko et al.

    Surface Chemistry and Microstructure of Flat-Panel Display Substrates

    SID 91 Dig.

    (1991)
  • S.K. Tang et al.

    Investigation of borophosphosilicate glass roughness and planarization with the atomic force microscope technique

    Thin Solid Films

    (1990)
  • M. Depas et al.

    Critical processes for ultrathin gate oxide integrity: Physics and Chemistry of SiO2 and the Si-SiO2 Interface

    Electrochem. Soc.

    (1996)
  • M. Kimura et al.

    Si/SiO2, Interface States and Neutral Oxide Traps Induced by Surface Microroughness

    J. Appl. Phys.

    (1995)
  • Cited by (20)

    • Spheroidization of borosilicate glass powder by RF induction coupled plasma

      2021, Ceramics International
      Citation Excerpt :

      The International Simple Glass (ISG) has been defined as a fundamental glass standard to enhancing knowledge of more complex industrial glasses used for nuclear waste confinement, and ISG is used to develop mechanistic understanding of the alteration behavior and kinetics [6]. To date, some studies have been performed with ISG [7–12]. Tricelpiece et al. [13] have developed new insights into transition by creating a relatively novel glass form for nuclear glass research: fibers.

    • Effect of ZnAl<inf>2</inf>O<inf>4</inf> crystallization on ion-exchange properties in aluminosilicate glass

      2021, Journal of Alloys and Compounds
      Citation Excerpt :

      These combined effects cause the shift and broadening of O1s (Fig. 6a). Si2p peaks of these specimens are located at ∼102.5 eV (Fig. 6b), comparable with those previously reported [56,57]. This observation shows that Si maintains their chemical states during the formation of ZnAl2O4 NCs.

    • Study of fractal dimension and power spectral density analysis of superconductor/ferromagnetic bilayer

      2018, Nuclear Instruments and Methods in Physics Research, Section B: Beam Interactions with Materials and Atoms
    View all citing articles on Scopus
    View full text