Elsevier

Journal of Membrane Science

Volumes 411–412, 1 September 2012, Pages 99-108
Journal of Membrane Science

Effect of temperature on electro-electrodialysis of HI–I2–H2O mixture using ion exchange membranes

https://doi.org/10.1016/j.memsci.2012.04.019Get rights and content

Abstract

In the thermochemical water-splitting iodine–sulfur process for hydrogen production, an electromembrane process (electro-electrodialysis) was applied to increase the HI molality of a HI–I2–H2O mixture. In order to investigate the temperature dependence of the membrane performance for increasing the HI concentration, a theoretical formula for the performance indexes (proton transport number, initial cell voltage, and ratio of water permeation to proton permeation in the membrane) was derived on the basis of the Nernst–Planck equation and electrophoresis theory. The obtained formula could reproduce the experimental values of the grafted membranes and Nafion 212 in the temperature range 313–373 K and was validated. The temperature dependence of the performance indexes was then clarified as follows: (1) The initial cell voltage decreased with temperature upon increasing the diffusion coefficient of H+ and (2) the proton transport number decreased with temperature because the diffusion coefficient of I increased more markedly than that of H+. For the ratio of water permeation to proton permeation in the membrane, we find the optimum regression parameters for achieving low water permeation, and controlling the ion exchange capacity might help in optimizing the membrane structure so that water permeation can be minimized.

Highlights

► We proposed a permeation model for HI concentration using the EED system. ► We investigated the temperature dependence of the membrane performance indexes for the EED operation. ► The model was validated by reproducing results of the EED performance test at 313–373 K for various membranes.

Introduction

Hydrogen production by thermochemical water-splitting methods involving the use of high-temperature solar heat or high-temperature nuclear heat has been studied for developing next-generation hydrogen energy systems. The iodine–sulfur (IS) process is one of the most promising thermochemical methods because of its potential for large-scale CO2-free hydrogen production [1]. The basic chemical reactions are as follows:I2 + SO2 + 2H2O  2HI + H2SO42HI  H2 + I2H2SO4  H2O + SO2 + 1/2O2Reaction (1), known as the Bunsen reaction, produces hydriodic acid (HI) and sulfuric acid from iodine (I2), sulfur dioxide, and water. Both the produced acids are then decomposed endothermically at elevated temperatures into hydrogen and oxygen as shown in reactions (2), (3), respectively.

The overall process efficiency is a major factor in the IS process. An important step for achieving this high efficiency is to increase the concentration of HI in the HI–I2–H2O mixture (hereafter referred to as HIx solution) supplied to the Bunsen reactor [2]. To increase the HI/H2O molar ratio, i.e., HI molality, of the catholyte, the authors have previously conducted experiments involving the electro-membrane process called electro-electrodialysis (EED) [3], [4]; the experiments involved redox reactions of iodine-iodide ions at electrodes and selective proton permeation through a cation exchange membrane (Fig. 1). Onuki et al. [5] confirmed the concentration of HI in the HIx solution experimentally by using commercially available proton-exchange membranes such as Nafion. However, they and other researchers found the EED performance to be quite low because of the requirements of high electric power [4], [5], [6], [7]. Hence, the authors used polymer electrolyte membranes prepared by radiation-induced graft polymerization and cross-linking methods, which have enabled their chemical, electrochemical, and mechanical properties to be controlled [8], [9], [10]. Some of the membranes required lower electric power for the EED operation than Nafion did, probably because of the lower cell resistance [3]. This result led us to expect that our method would make it possible to prepare a suitable membrane for increasing the HI concentration.

In the next step, the objective was to design a membrane for maximizing the EED efficiency. For EED, HIx solution was taken from the bottom of the Bunsen reactor as well as from the recycle line of an HI distillation column maintained at a temperature near the boiling point of the HIx solution (typically 473 K) [11]. Imai et al. [12] achieved the desired high HI concentration in the Bunsen reactor at high temperatures up to 363–373 K. The subsequent EED operation should preferably be performed at such a high temperature because it is then energetically favored and no preheating devices are required.

Hong et al. [6] investigated the effects of the operation temperature on Nafion's performance in EED. The proton conductivity increased with the temperature, while the proton and water selectivity of Nafion decreased. This result suggests that high temperature would not always lead to satisfactory EED performance. Thus, the thermodynamics in the EED process would be more complex than expected and have to be examined more closely in terms of the membrane performance and its temperature dependence.

In this study, we propose an EED model for studying the temperature dependence on the basis of Nernst–Planck equation and electrophoresis theory and demonstrate its validity by comparing the obtained results with experimental results for some grafted membranes and Nafion 212. Here, on the basis of regression parameters, we discuss the detailed thermodynamics involved in EED.

Section snippets

Model

The electric power required for the EED operation (P) can be considered to comprehensively indicate the membrane performance because a lower P value contributes to an increase in the total process efficiency [11]. We consider the initial stage of the EED operation, which is the condition that there should be little concentration difference between anolyte and catholyte. The range of application was described in Ref. [3]; the difference between anolyte and catholyte of HI molality was within ca.

Chemicals and membranes

Hydriodic acid (HI, 55–58 wt%, Kanto Chemical Co., Inc.) and iodine (I2, >99.8 wt%, Kanto Chemical Co., Inc.) were used to prepare the HIx solution. Sodium thiosulfate solution (0.1 M Na2S2O3, Kanto Chemical Co., Inc.), sodium hydroxide solution (0.1 M NaOH, Kanto Chemical Co., Inc.), and potassium iodide powder (KI, >99.7%, Kanto Chemical Co., Inc.) were used to analyze the composition of the HIx solution by titration; an automatic potentiometric titrator (COM-2500, Hiranuma Sangyo Co., Ltd.) was

Regression analysis for performance indexes

As mentioned above, the comprehensive membrane performance was represented by the P value, which is dependent on t+, E, and β in the EED process. In our previous EED experiments [3], the t+ value was estimated to be 0.59–0.84 for the grafted membranes, and 0.91–0.95 for Nafion 212. Similarly, the E and β values were estimated to be 0.06–0.17 V and 1.4–3.0 for the grafted membranes, and 0.11–0.25 V and 1.6–2.4 for Nafion 212, respectively. We consider our proposed model for each of these

Conclusions

We proposed a permeation model for the EED system; the model is based on the Nernst–Planck equation and electrophoresis theory. We then investigated the temperature dependence of the membrane performance indexes for the EED operation, i.e., the proton transport number t+, initial cell voltage E and ratio of water permeation to proton permeation in the membrane β. Their temperature dependence was expressed as follows:t+=1+BexpΔUIHRT1E=t+idYTexpΔUH+RTβ=GT3/2expUβRTThe model was validated by

Acknowledgment

The authors are grateful to Mr. K. Ikenoya for performing the EED concentration experiments.

References (27)

  • H. Ohshima et al.

    Improvement on Hogg–Healy–Fuerstenau formulas for the interaction of dissimilar double layer

    J. Colloid Interface Sci.

    (1982)
  • S. Sawada et al.

    Water transport properties of crosslinked-PTFE based electrolyte membranes

    Solid State Ionics

    (2008)
  • S. Kubo et al.

    Flow sheet evaluations for the thermochemical water-splitting iodine–sulfur process (I)

  • Cited by (0)

    View full text