Hydrogen detection/concentration

A prestressed roller bearing is used to detect the presence of hydrogen sulfide, but more specifically it is used to test for hydrogen embrittlement tendency of the drilling fluid. When introduced to the environment, the bearing has sufficient residual stresses to cause failure if sufficient hydrogen sulfide concentration is present. 

The value of the ratio [InB]/[InA] (i.e. [Basic form]/[Acidic form]) can be determined by a visual colour comparison or, more accurately, by a spectrophotometric method. Both forms of the indicator are present at any hydrogen-ion concentration. It must be realised, however, that the human eye has a limited ability to detect either of two colours when one of them predominates. Experience shows that the solution will appear to have the acid colour, i.e. of InA, when the ratio of [InA] to [InB] is above approximately 10, and the alkaline colour, i.e. of InB, when the ratio of [InB] to [InA] is above approximately 10. Thus only the acid colour will be visible when [InA]/[InB]> 10 the corresponding limit of pH given by equation (5) is ... 

Luminol (3-aminophthalhydrazide) reacts with hydrogen peroxide in the presence of a metal catalyst in basic solution to produce luminescence. This reaction is extremely sensitive and may be used to detect concentrations of parts per 1012 of metals separated by liquid chromatography [199-202]. 

The interference to the hydrogen detection of C-I-S structures caused by varying amounts of water vapor is also summarized in Table II. As seen in that table, high concentrations of H2O vapor lower the sensitivity of Pd/SiOx/Si diodes whereas water vapor, in general, lowers the sensitivity of Pd/TiOx/Si diodes at room temperature. 

The effect of elevated operating temperatures on the interference to hydrogen detection arising from O2 for Pd/SiOx/Si diodes is also seen in Table III as a function of H2 ppm levels. As may be noted from this table, the reduced sensitivity to hydrogen, especially at low concentrations, caused by the interference from oxygen is not significant at elevated device operating temperatures i.e., temperature effects dominate. 

In a limited number of experiments, in which six different hydrogen peroxide concentrations were examined with four different pH values, the validity of the procedure used was checked by comparing the results obtained with the sensor electrode to those obtained by means of titration. The results obtained with the sensor electrode have a maximum divergence of 3% compared with the concentrations obtained by titration. The final aspect of the amperometrical detection method that was examined at laboratory scale is the stability in time of a calibrated sensor electrode. 

An initial series of measurements concerned the accurate determination of the dilution factor. As is represented in Fig.5.16, the absolute position of the tubes of the process solution and of the sodium hydroxide solution over the rollers of the pump is different. One tube is situated higher on the roller than the other, which can lead to a small difference in flow rate. By determining the hydrogen peroxide concentration by means of titration of a sample taken from the bath and a sample taken at the exit of the detection cell (diluted solution), the dilution factor can be determined. Table 5.1 presents the results of 23 measurements of the hydrogen peroxide concentration, performed at a liquid flow rate of 11/h. It is concluded that the dilution factor equals 1.9773+0.0014, which in fact is very close to 2. 

In a second series of experiments, the dead time is examined by means of the sensor signal at a liquid flow of 11/h. It is clear that when modifying, for example, the hydrogen peroxide concentration in the bath, this change will only be observed in the detection cell after a period of time. The length and the diameter of the tubes and the actual internal volume of the detection cell play an important part here. After changing the hydrogen... 

This means that one should bear in mind that the values of the process parameters detected at the sensor correspond to a situation which occurred in the bath of the process 30 s earlier. This is extremely important if one wishes to control the hydrogen peroxide concentration in the considered process. Though the dead time can be shortened by increasing the flow rate, the problem arises then that higher flow rates generate more waste-water and turbulent behaviour in the detection cell which affects the sensor signal. Similarly, the dead time of the same sensor expanded with the FIA system was verified with different flow rates (Fig.5.18). With a flow rate of, for example, 21/h, the dead time amounts to ca. 23 s. 

Another important element that should be taken into account is that, in certain processes, relatively low hydrogen peroxide concentrations are used. This means that the sensor s detection limit can be a crucial parameter for these applications. As a criterion, one speaks of useful measurements when the sensor signal is three times higher than the noise observed in the absence of hydrogen peroxide. As appears from the data shown in Fig. 4.9, the sensor signal increases with increasing pH. Hence, it is interesting to... 

Hence another strategy was followed, where the calibration is done by means of a solution from the detection cell. With this solution, the sensor signal, as well as the concentration of hydrogen peroxide through titration, was obtained. In order to be able to calculate subsequently the original concentration in the bath where the process occurs, the dilution factor should be taken into account, and for the control of the hydrogen peroxide concentration one should bear in mind a 30-s delay. In this way, a possible difference in process-bath composition is eliminated. The equation for the calculation of the hydrogen peroxide concentration in the process bath is ... 

At a laboratory scale, some preliminary experiments were performed to determine hydrogen peroxide concentrations in solutions with pH < 10.5. To this end, hydrogen peroxide solutions from a paper pulp process (pH ca. 8) were used, which were obtained from a French company. The results are displayed in Fig. 5.19. Here, mainly relatively low hydrogen peroxide concentrations were examined, with output signals that were never higher than 35pA. The pH in the detection cell was ca. 12.5, and the temperature fluctuated around 318K. 

---