- 9:45 AM

Hydrogen Production by Sulfur-Iodine Thermochemical Cycle: Experimental Tests on Hydriodic Section Concerning Iodine Purification System

Raffaele Liberatore, Antonio Ceroli, Claudio Felici, Michela Lanchi, Salvatore Sau, Annarita Spadoni, and Pietro Tarquini. Ter Soltermsvil, ENEA, Via Anguillarese, 301, Rome, I-00123, Italy

The Sulfur-Iodine thermochemical cycle is one of the most promising routes currently under development for sustainable hydrogen production from water. It is based on three main reactions carried out at three different temperature levels (low, medium and high temperature):

- Bunsen reaction, exothermic at (80÷120) °C:

2H2O + I2 + SO2 --> H2SO4 + 2HI

- Hydriodic acid decomposition, endothermic, at (300÷500) °C:

2HI --> I2 + H2

- Sulfuric acid decomposition, endothermic, at (800÷900) °C:

H2SO4 --> H2O + SO3 --> H2O + SO2 + ˝ O2

In the Bunsen section, water, iodine, and SO2 at a temperature between 80 °C and 120 °C, produce two liquid immiscible phases: the superior one containing a mixture of H2SO4 and water, and the other including HI, I2 and water.

In the Sulfuric acid decomposition section the Bunsen reaction superior phase, mainly composed of a H2SO4-H2O mixture, is concentrated from (50÷57) %w/w to (90÷98) %w/w and decomposed in water, oxygen and SO2 which is returned in the Bunsen section.

The Bunsen reaction inferior phase, instead, is sent to the hydriodic acid decomposition section where HI is decomposed in hydrogen and iodine. HI, residual H2SO4, and a very high content of iodine (~ 82 %w/w) compose this phase, since the Bunsen reaction is conducted with an excess of iodine to promote the separation of the two phases. A preventive SO2 and iodine removal by an intermediate purification-separation step is necessary, in particular considering the very low decomposition reaction conversion rate (22 % at 480 °C) that requires a recirculation of these components.

Concerning hydriodic section, the most critical phase is represented by the separation-purification step of HI. Several techniques have been proposed on this purpose and their technical and economical sustainability have been demonstrated (Allen et al., Process Safety and Environmental Protection, 83 (B4) 343-350).

In the present work, in view of a forthcoming realization of an integrated loop plant at a laboratory scale, classic distillation process for I2 separation has been considered. This technique is a conventional way to separate HI and water from I2, in view of the high difference of the boiling temperature among the three elements (at room pressure: -35°C for HI, 100°C for H2O, 184 °C for I2).

In the Bunsen inferior phase, the ratio HI/(HI+H2O) generally approaches to 57 %w/w. For iodine free mixtures, this ratio corresponds to the azeotrope, characterized by a maximum temperature boiling point, corresponding to 127 °C (1 bar). Our tests and an Engels's work (H.Engels, K.F. Knoche, International Journal of Hydrogen Energy, 11, pp. 703-707,1986) showed that isotherms of different HI-H2O-I2 mixtures containing fixed I2 contents, have the minimum vapor pressure, defined as pseudo-azeotrope points, just near at this particular ratio between HI and HI+H2O.

In these conditions, it is plausible, by a distillation column, obtaining this value in the entire equipment, where the only iodine composition will vary from very low percentages, at the top, to 99 % w/w, at the bottom.

Nevertheless, opportunely dosing the reagents and in particular temperature conditions, it could be possible obtaining, by Bunsen reaction, HI-I2-H2O mixtures with HI content superior than the pseudo-azeotropic point, so that at the top of the column we could have only HI (although in low amount). On the contrary, if the amount of water loaded in the Bunsen reactor is too much, we could obtain HI-I2-H2O mixtures with a HI content inferior than the pseudo-azeotrope. In this case, the distillation column produces water at the top.

Since the purification-separation equipment must resist at a very aggressive environment and the column boiler need a great energy amount (see: Kasahara et al., “Journal of chemical engineering of Japan, vol.36, n°7, pp.887-889, 2003; Goldstein et al., “International Journal of Hydrogen Energy, 30 (2005) 619-626), this section affects for a high share the global process cost and efficiency. Therefore, a scrupulous design is suitable. In order to scale and optimize the distillation column, experimental HI-I2-H2O liquid-vapor equilibrium data for a wide concentration domain of the liquid phase, have been investigated.

These mixtures had I2 concentration in a range included from 0.2 %w/w to 90 %w/w and HI from 4 %w/w to 67 %w/w in the liquid phase.

The liquid phase composition versus temperature measurements demonstrated that temperatures show a different behaviour depending on the HI and I2 concentration.

If HI content is higher than pseudo-azeotrope: the I2 increase produces a temperature rising, on the contrary, the HI augment turns out a temperature lowering, because its concentration goes away from the binary azeotrope HI-H2O.

HI augmentation, instead, in mixtures with HI minor than pseudo-azeotrope, rises the boiling temperature. Concerning iodine, experimental evidence showed a boiling temperature increase for mixtures with HI content near pseudo-azeotrope, but if this content is substantially lower (i.e.: 25 %w/w or 45 %w/w in iodine free basis) a light temperature reduction has been observed, at least for medium-low iodine contents.

In these last cases, some problems occur if we need to condense the vapor because it has a high content of water and small quantity of iodine (HI is very low). Since the iodine solubility in water is high only when in the mixture is being a remarkable HI content (C. F. Powell, I. E. Campbell, Journal of the American Chemical Society , Vol.: 69, pp.1227-1228, 1947), iodine precipitates and piles up on the condenser walls, causing its gradual obstruction.

In every case, iodine high content mixture samplings need an insulated and thermostatic sample system in order to avoid the sample cooling and the consequent iodine solidification.

At the end of the liquid-vapor equilibrium characterization (at room pressure), the present study has been focused on distillation preliminary tests of the mixtures coming out from Bunsen reaction inferior phase. These tests have been executed, in batch, by two kind of Pyrex equipment: one Vigreux column (Height = 600 mm, diameter = 32 mm) and one packed column (Height = 600 mm, diameter = 32mm) with Rasching rings (diameter = 7 mm). Other tests, using, as feed, inferior solutions, directly coming from Bunsen reactor, have been performed in continuous by a packed column (with Rasching rings) bigger than the others (Height = 1000 mm, diameter = 50 mm) and equipped by pumps and an external condenser.

The condenser showed obstruction problems, concerning mixtures with HI content substantially lower than the pseudo-azeotrope, due to the cited precipitation of the iodine. In these last cases, indeed, at the top of the column, water essentially went out, HI being absent. The Bunsen inferior phase, generally, includes small amounts of H2SO4 (Sakurai et al., Int. Journal of Hydrogen Energy, 25 (2000) 605-611) so, if they entered the column, go out as distillate together water, before the HI-H2O mixture. Therefore, it is evident the need to control the feed composition be appropriate and to insert a purification equipment in order to eliminate the H2SO4 and to avoid the water excess.

However, these tests confirmed the possibility to separate iodine from Bunsen inferior phase and the results are showed in this work.