Elsevier

Fuel

Volume 87, Issue 2, February 2008, Pages 222-231
Fuel

The use of fractal analysis in the textural characterization of coals

https://doi.org/10.1016/j.fuel.2007.04.020Get rights and content

Abstract

The materials studied were four bituminous coals as well as the corresponding coal samples oxidized in air at 543 K for different periods of time. The method proposed by Friesen and Mikula, the procedure of Neimark and the methodology of Zhang and Li were successfully employed to analyze mercury porosimetry data from a fractal perspective. Fractal dimensions as well as fractal profiles are sensitive to oxidation treatment, being useful to follow the changes undergone by the coal samples. The evolution of the fractal dimension of coals with oxidation is determined by a balance between two main mechanisms of pore development: the oxidation of the pore surface itself that tends to lower the fractal dimension and the access of mercury to previously non-accessible regions that tends to increase this dimension.

Introduction

The manufacture of active carbons aims to develop a carbon-based substance with suitable textural properties. When the objective is to obtain a granular active carbon from a bituminous coal it is necessary to carry out a series of steps in order to obtain the desired material [1], [2]. The first operation consists in an adequate series of grinding and sieving stages that allows the obtention of the appropriate particle size. Then, an oxidation step reduces the plastic properties of the coal. Otherwise, the char obtained would not be suitable for obtaining a good active carbon. The following step, the carbonization of the coal, produces a char with lower volatile matter content and a more developed porous network. The final step, the activation of the char, is carried out by reaction with an oxidant that leads to an active carbon with a suitable pore structure. Returning to the air oxidation of coals, it is highly important to note the following: air oxidation not only reduces and can even destroy any remainders of plastic properties, but also modifies the pore network of the coal. The pore development of coals during air oxidation is of prime importance as it is well known that the textural properties of chars and active carbons depend – among other factors – on the textural properties of the precursor coals. Each precursor leaves its fingerprint in the properties of the materials obtained from it [1], [3], [4], [5], [6].

According to the above mentioned facts, it can be deduced the need to know as much as possible about the textural features of active carbons and those of their precursors. For characterizing the texture of active carbons as well as of other porous solids, several “classical” techniques are employed. Mercury and helium picnometries, mercury porosimetry, gas adsorption isotherms or SEM, are among the most well-known. These techniques allow us the estimation of useful textural parameters such as specific pore volume, porosity, pore size distribution or specific surface area.

For some years, there has been a new approach towards the study of textural properties of materials based on the fractal theory. This fractal theory appeared formally in the mid-seventies [7] and relies on the concept of self-similarity. An object is self-similar if its appearance does not depend on the scale at which it is observed. A familiar consequence of this fact is, for example, the reference to the actual size that is found in any geological report where photographs of rocks or geological formations are shown. The determination of the pore volumes, densities or specific surface areas of a solid may be variable depending on the technique and the data processing employed and they do not necessarily give information about the pore surface itself. These facts should not be seen as a pitfall of the methods, but as a consequence of the different foundations and mechanisms of measurement. Fractal determinations may have, in some sense, the same drawbacks as classical methods, since results also depend on the original data being processed and the theoretical approach being applied. Despite this fact, the fractal dimension of the surface of a porous solid does not depend, theoretically, on the size of the pores or the “amount of surface” and is an intrinsic characteristic of the surface itself. Methods have been developed for obtaining fractal information regarding porous solids by processing data from different techniques such as gas adsorption, micrographs or SAXS, among others. These methods have been extensively applied to carbonaceous materials [8], [9], [10], [11], [12].

In the next section, some procedures for the fractal analysis of mercury porosimetry data are presented. These methods have been successfully employed for the textural analysis of coals, chars and active carbons [4], [5], [13] together with the use of classical textural determinations. In the present work, the previous analysis of series of an oxidized coal are compared with the results from the fractal study of three other oxidized coals in order to establish the general validity of the previous conclusions. The main findings are that, despite the differences observed between different coals, the fractal dimension is a reliable parameter in order to follow textural evolution during the oxidation step. Moreover, fractal analysis complements classical approaches, enhancing the understanding of the textural changes undergone by coals during the oxidation process.

Section snippets

Methods for fractal analysis of mercury porosimetry data

Mercury porosimetry enables the obtaining of pore size distributions for macropores and wide mesopores and is based on the relationship between the intrusion pressure of mercury and the dimension of the smallest pore that can be reached under this pressure. This relationship between the average radius of curvature of the meniscus, R, and the intrusion pressure, P, is given by the Laplace equationR=2σcosθPwhere σ is the surface tension and θ the contact angle of the mercury. Data from

Experimental

The coals employed in this study were two high-volatile A bituminous coals (coals A and B), a low-volatile bituminous coal (coal C) and a medium-volatile bituminous coal (coal D). The particle size utilized in this study was the fraction 1.0–3.0 mm. The yield in this fraction was maximized by means of successive crushing and sieving operations. Coal samples were oxidized in an oven with forced air circulation at a temperature of 543 K for 1, 2, 3, 4 and 14 days. Characterization of the raw coals

Results and discussion

For the fractal analysis of the pore surface of the materials, the three above-mentioned methods were employed: the procedure of Friesen and Mikula, the approach proposed by Zhang and coworkers and the methodology of Neimark.

The application of the methods of Friesen, Zhang and Neimark to the series of coals B, C and D enables the calculation of average fractal dimensions for the pore size range considered. It has already been reported for coal A that a decrease in the fractal dimension with

Conclusions

The oxidation of coals induces textural changes that are also reflected in the evolution of the fractal dimensions of the corresponding materials. The evidence compiled and discussed in the present paper can be summarized in the following conclusions:

  • 1.

    The methods proposed by Friesen, Neimark and Zhang and Li are valid approaches for obtaining values of the average fractal dimensions for each material corresponding to series of oxidized coals.

  • 2.

    Previous findings stating that the average fractal

Cited by (173)

View all citing articles on Scopus
View full text