Do D D, Do H D
Department of Chemical Engineering, University of Queensland, St. Lucia, Qld 4072, Australia.
J Phys Chem B. 2005 Oct 20;109(41):19288-95. doi: 10.1021/jp052448y.
In this paper, we evaluate the performance of the 1- and 5-site models of methane on the description of adsorption on graphite surfaces and in graphitic slit pores. These models have been known to perform well in the description of the fluid-phase behavior and vapor-liquid equilibria. Their performance in adsorption is evaluated in this work for nonporous graphitized thermal carbon black, and simulation results are compared with the experimental data of Avgul and Kiselev (Chemistry and Physics of Carbon; Dekker: New York, 1970; Vol. 6, p 1). On this nonporous surface, it is found that these models perform as well on isotherms at various temperatures as they do on the experimental isosteric heat for adsorption on a graphite surface. They are then tested for their performance in predicting the adsorption isotherms in graphitic slit pores, in which we would like to explore the effect of confinement on the molecule packing. Pore widths of 10 and 20 A are chosen in this investigation, and we also study the effects of temperature by choosing 90.7, 113, and 273 K. The first two are for subcritical conditions, with 90.7 K being the triple point of methane and 113 K being its boiling point. The last temperature is chosen to represent the supercritical condition so that we can investigate the performance of these models at extremely high pressures. We have found that for the case of slit pores investigated in this paper, although the two models yield comparable pore densities (provided the accessible pore width is used in the calculation of pore density), the number of particles predicted by the 1-site model is always greater than that predicted by the 5-site model, regardless of whether temperature is subcritical or supercritical. This is due to the packing effect in the confined space such that a methane molecule modeled as a spherical particle in the 1-site model would pack better than the fused five-sphere model in the case of the 5-site model. Because the 5-site model better describes the liquid- and solid-phase behavior, we would argue that the packing density in small pores is better described with a more detailed 5-site model, and care should be exercised when using the 1-site model to study adsorption in small pores.
在本文中，我们评估了甲烷的单位点和五位点模型在描述石墨表面及石墨狭缝孔吸附方面的性能。已知这些模型在描述流体相行为和汽液平衡方面表现良好。本文针对无孔石墨化热炭黑评估了它们在吸附方面的性能，并将模拟结果与Avgul和Kiselev的实验数据（《碳的化学与物理》；Dekker出版社：纽约，1970年；第6卷，第1页）进行了比较。在这个无孔表面上，发现这些模型在不同温度下的等温线上的表现与它们在石墨表面吸附的实验等量吸附热上的表现一样好。然后测试了它们在预测石墨狭缝孔吸附等温线方面的性能，在其中我们想要探究限制对分子堆积的影响。本研究中选择了10 Å和20 Å的孔径，并且我们还通过选择90.7 K、113 K和273 K来研究温度的影响。前两个温度代表亚临界条件下，90.7 K是甲烷的三相点，113 K是其沸点。选择最后一个温度代表超临界条件，以便我们能够研究这些模型在极高压力下的性能。我们发现，对于本文所研究的狭缝孔情况，尽管这两个模型产生了相当的孔密度（前提是在孔密度计算中使用可及孔径），但无论温度是亚临界还是超临界，单位点模型预测的粒子数总是大于五位点模型预测的粒子数。这是由于受限空间中的堆积效应，使得在单位点模型中被建模为球形粒子的甲烷分子比五位点模型中的融合五球模型堆积得更好。因为五位点模型能更好地描述液相和固相行为，我们认为在小孔中用更详细的五位点模型能更好地描述堆积密度，并且在使用单位点模型研究小孔吸附时应谨慎。
