rus
Issue #5/2018
B.Gribov, К.Zinoviev, О.Kalashnik, N.Gerasimenko, D.Smirnov, V.Sukhanov, L.Sukhanova, V.Chetverikov
Obtaining hydrogen for fuel cells using finely dispersed silicon
Obtaining hydrogen for fuel cells using finely dispersed silicon
The development of hydrogen energy causes the growth of interest in the creation of chemical hydrogen generators. The paper considers the prospect of using fine-dispersed silicon in such generators. In particular, the waste products of the production of high-purity single-crystal and polycrystalline silicon, as well as metallurgical silicon, can be suitable and cheap materials for the obtaining of hydrogen when interacting with a weak-alkaline KOH solution. The study of the change in the rate and the thermal effect of the reaction as a function of its time, the parameters of the silicon powder, and the ratio of the initial components of the solution made it possible to establish that for an identical particle size of the powder, the specific rate of hydrogen generation depends little on the method of obtaining silicon, its crystal structure and the impurity content. The main factor influencing the intensity of the chemical reaction is the particle size of the powder. Based on the research, recommendations for the selection of powders of polycrystalline and monocrystalline silicon for practical use in autonomous chemical hydrogen generators are proposed.
Теги: chemical hydrogen generator hydrogen energy monocrystalline and polycrystalline silicon водородная энергетика монокристаллический и поликристаллический кремний химический генератор водорода
Recently, in connection with the development of hydrogen energy, there is growing interest in materials that can be used in chemical hydrogen generators of various types and purposes. Portable chemical hydrogen generators complete with small fuel cells and a battery are able to function as alternative power sources for electronic devices in the field. The operation of such generators is based on the thermal decomposition of hydrides or the reaction of metals and hydrides with water. The production of hydrogen in specified volumes is also possible with the interaction of finely dispersed silicon with water. In a number of studies, this interaction was investigated to find the most active form of a silicon powder [1–4].
It was shown in [4] that particles of nano-crystalline silicon with a size of about 10 nm react actively with water containing KOH forming hydrogen and silicic acid. At the same time, the process speed exceeds the values for known systems that generate hydrogen (for example, hydride systems). However, the practical use of such nanoparticles for hydrogen generation is problematic because of their high cost. In addition, too active interaction of reagents can make it difficult to control the process.
The purpose of this work is to search and study silicon materials that are suitable for practical use in portable chemical hydrogen generators. A very suitable and cheap material can be wastes from the production of high-purity single-crystal and polycrystalline silicon, and also metallurgical silicon – polycrystalline silicon of low purity [5, 6].
EXPERIMENT CONDITIONS
Metallurgical silicon is produced in Russia and abroad on an industrial scale and has a relatively low cost. The powder of metallurgical silicon of Кр0 and Кр1 is used as a raw material for the production of silanes and high-purity silicon.
In this study, the main object of investigation was a powder of KR0 polycrystalline metallurgical silicon with a particle size of 40 μm. Other silicon powders, in particular powders obtained from high purity silicon wastes, were also investigated to expand the range of materials suitable for use. The characteristics of all the powders are given in Table 1.
When the experiment was carried out, the portions of the powder were weighed with an accuracy of 0.05 g and placed into a glass flask, into which an aqueous solution of KOH with a volume of 75 ml was then poured. The flask was quickly closed with a stopper with a gas outlet tube, the end of which was fed to a measuring cylinder filled with water. The volume of hydrogen generated by the chemical reaction in the flask was determined by the volume of water displaced from the graduated cylinder. The error in determining the volume was ± 5 ml. The temperature of the solution in the flask was measured by a laboratory thermometer with an accuracy of 0.5°C.
The intensity of interaction of silicon powder with water in an alkaline medium and, accordingly, the nature of hydrogen generation directly depend on the mass of the powder and the concentration of KOH in the solution (at its constant volume). The mass of the powder was 0.4–1.2 g, and the KOH concentration was in the range of 0.25–4%, which made it possible to work with small volumes of hydrogen and low-aggressive solutions.
INVESTIGATION OF KR0 POWDER
Fig.1 and Fig.2 show the dependences of the hydrogen generation volume on the reaction time for solutions with different concentrations of the initial components. It follows from Fig.1 that at a KOH concentration of 0.25% the reaction practically does not go, but even at a KOH concentration of 0.5%, the interaction takes place, although its intensity is very low. Increasing the concentration of KOH rapidly increases the intensity of the reaction, so the upper limit was limited to 2% KOH.
Analysis of the curves in Fig.2 shows that for all values of the mass of silicon (mSi), the highest intensity of the reaction is observed at the initial stage. If we arbitrarily limit the initial stage to 5 minutes (300 s), then during this time, an average of 52% of the theoretically possible volume of hydrogen will be allocated for all the indicated values of mSi. Further, the intensity of the reaction gradually decreases due to the consumption of the reagent (Si): in the second stage (5–10 minutes), 26–28% are released, and at the third (10–15 min.) – less than 10% of the theoretically value. The subsequent stages are no longer of interest because of the small volumes of hydrogen, although the reaction can last more than 30 minutes, and the total generation can reach 90–94% of the theoretical value.
Of greatest interest is the first stage of the reaction (0–5 min.), where all the dependences are close in shape to the straight lines drawn from the origin, and their inclination to the abscissa axis characterizes the hydrogen generation rate (the volume of H2 liberated per unit time). With an increase in the initial mass of silicon, the volume of hydrogen release increases very rapidly and is directly proportional to mSi. Table 2 shows the rates of the H2 generation rate in the first stage.
As a characteristic of a particular type of silicon powder, the term "specific rate of hydrogen generation" is used in the literature on this topic, that is, the speed per unit mass (one gram of silicon) [4]. Table 2 shows the values of the specific velocity calculated for the values of mSi shown in Fig.2. These values are very close: the arithmetic mean value is 166 ml with a spread of ± 3.5%.
If we increase the sample weight of the powder mSi by more than 1.2 g, then the H2 generation rate will also increase proportionally to a certain limit, which can be considered as the maximum rate for a given solution volume (75 ml). The fact is that the reaction of silicon with water in an alkaline medium is exothermic and quickly leads to a significant heating of the solution. To estimate the influence of the mass of silicon on the magnitude of the thermal effect, measurements of the temperature of the solution prior to the start of the reaction and in its course were made. The results of the measurements are given in Table 3.
On the basis of the data obtained, it can be concluded that weighed portions of Si powder more than 2 g are not desirable, since they can bring the temperature of the solution to the boiling point. But even with a sample of Si of more than 1.7 grams, rapid solution heating and gas evolution lead to the fact that part of the powder is carried to the surface, and the droplets of the solution are sprayed with the powder particles. Thus, it is expedient to limit the maximum sample of silicon to a value of 1.5–1.6 g.
For practical use, not less than 10–15 liters of hydrogen is required, therefore, increased amounts of silicon powder should be used. To avoid overheating of the reaction mass, it is possible to increase the volume of the alkaline solution. Water in this process serves not only as a source of hydrogen, but also as a reaction medium. According to the stoichiometry, no more than 2.6 ml of H2O per 1 g of Si is required [4], and a multiple excess of water is required to dissolve the reaction product (silicic acid) and to cool the solution. The conducted experiments indicate the possibility of increasing the volume of the solution several times, for example, up to 200 ml. An increase in volume by an order of magnitude or more requires additional investigation, since problems associated with a large volume of hydrogen and an increase in the pressure in the reaction vessel may arise.
ALTERNATIVE POWDERS
In the present work, we did not confine ourselves to the study of one type of silicon powder, but also tested alternative variants. Among them, powders obtained from monocrystalline silicon waste (residues after cutting out a branded single crystal from an ingot) and intermediate or by-products of pyrolysis of monosilane [7, 8] (such a powder is usually unsuitable for the production of silicon ingots due to rapid oxidation in air). Both options can be considered as a way of recycling waste from the production of high-purity silicon.
As the third alternative, a powder of metallurgical silicon with a particle size of less than 10 μm was tested. It is known that grinding the material to particles of no more than 10 μm in size can create an activation effect in some chemical processes [9]. To determine the effect of this factor on the intensity of hydrogen generation, we compared powders obtained from the same material – metallurgical silicon, but differing in particle size.
These powders were tested according to the procedure described above with constant parameters: powder mass is 0.5 g; concentration of KOH is 2%; volume of solution is 75 ml. The results of the tests, as well as comparative literature data, are given in Table 1, where the main characteristics of the powders and the corresponding values of the specific generation rate of hydrogen are indicated.
Analysis of the data in Table 1 shows that the powder of single-crystal silicon with a particle size of 40 μm in intensity of hydrogen generation is close to a polycrystalline silicon powder of the same size. The experimental results are in good agreement with the literature data.
The polycrystalline silicon powder with smaller particles is characterized by a higher intensity of hydrogen generation, but less than one might expect based on the value of its specific surface area. The same applies to the silicon powder obtained by pyrolysis. Apparently, this effect is associated with the behavior of a very fine silicon powder in solution at a high rate of H2 generation, that is, with the removal of powder particles to the surface and their aggregation. Another reason for the discrepancy between the actual generation rate of H2 and the expected (from the point of view of the degree of particle size reduction) can be the oxidation of particles in air.
ANALYSIS OF RESULTS
According to the results of tests of all silicon powders, it can be concluded that for the same particle size the specific hydrogen generation rate depends little on the method of obtaining Si, the crystal structure and the impurity content. The particle size is the main factor influencing the intensity of the chemical reaction, so its decrease allows to increase the rate of hydrogen generation accordingly. However, it should be noted that the generation of hydrogen in the "generator-fuel cell-battery" system has the ultimate goal of charging the battery, and this process, as a rule, requires a long time. Therefore, for practical use, the long and stable operation of a chemical generator is much more important than the rate of hydrogen generation. In addition, the use of fine powders is economically unprofitable, since their cost is inversely proportional to the particle size.
For these reasons, the use of silicon powders with a particle size of less than 40 μm in chemical hydrogen generators is impractical. In this case, powders with a particle size of more than 40 μm are required for significant mSi values.
Powdered silicon is a safe, neutral and cheap material for generating hydrogen, favorably differing from hydrides, which are dangerous, decomposing in air and expensive materials. It is supposed that it is possible to use silicon containing a significant amount of impurities, since practically all of them remain in solution during the reaction of silicon with water.
On the basis of the work done, powders of polycrystalline and monocrystalline silicon with a size of 40 μm or more can be recommended for practical use in autonomous chemical hydrogen generators. Such powders ensure the production of H2 in a volume of more than 1.2 liters per gram of silicon in a short time (about 10 minutes). Based on this, it is possible to calculate the amount of powder needed to produce a given volume of hydrogen over a certain period of time.
The mass of the silicon powder and its size have a significant effect on the generation rate of H2, so it is advisable to use powders with a particle size of more than 40 μm for large silicon loads. On the other hand, autonomous chemical generators now allow very precise control over the hydrogen production reaction and its supply to consumers. As the latter, we consider small-size fuel cells that provide recharging of batteries or direct power to mobile phones and other electronic devices in the field.
CONCLUSION
The processes of hydrogen production by the reaction of powdered silicon with water in a slightly alkaline medium are studied. The dependences of the hydrogen generation volume on the reaction time, the parameters of the silicon powder and the ratio of the initial components of the solution are determined. The chemical activity of the silicon powder was estimated from the specific hydrogen generation rate and the thermal reaction effect. The main limitations associated with the maximum mass of silicon in solution and the particle size of the powder are revealed. The prospects of practical use of the investigated silicon powders in chemical hydrogen generators are shown. ■
It was shown in [4] that particles of nano-crystalline silicon with a size of about 10 nm react actively with water containing KOH forming hydrogen and silicic acid. At the same time, the process speed exceeds the values for known systems that generate hydrogen (for example, hydride systems). However, the practical use of such nanoparticles for hydrogen generation is problematic because of their high cost. In addition, too active interaction of reagents can make it difficult to control the process.
The purpose of this work is to search and study silicon materials that are suitable for practical use in portable chemical hydrogen generators. A very suitable and cheap material can be wastes from the production of high-purity single-crystal and polycrystalline silicon, and also metallurgical silicon – polycrystalline silicon of low purity [5, 6].
EXPERIMENT CONDITIONS
Metallurgical silicon is produced in Russia and abroad on an industrial scale and has a relatively low cost. The powder of metallurgical silicon of Кр0 and Кр1 is used as a raw material for the production of silanes and high-purity silicon.
In this study, the main object of investigation was a powder of KR0 polycrystalline metallurgical silicon with a particle size of 40 μm. Other silicon powders, in particular powders obtained from high purity silicon wastes, were also investigated to expand the range of materials suitable for use. The characteristics of all the powders are given in Table 1.
When the experiment was carried out, the portions of the powder were weighed with an accuracy of 0.05 g and placed into a glass flask, into which an aqueous solution of KOH with a volume of 75 ml was then poured. The flask was quickly closed with a stopper with a gas outlet tube, the end of which was fed to a measuring cylinder filled with water. The volume of hydrogen generated by the chemical reaction in the flask was determined by the volume of water displaced from the graduated cylinder. The error in determining the volume was ± 5 ml. The temperature of the solution in the flask was measured by a laboratory thermometer with an accuracy of 0.5°C.
The intensity of interaction of silicon powder with water in an alkaline medium and, accordingly, the nature of hydrogen generation directly depend on the mass of the powder and the concentration of KOH in the solution (at its constant volume). The mass of the powder was 0.4–1.2 g, and the KOH concentration was in the range of 0.25–4%, which made it possible to work with small volumes of hydrogen and low-aggressive solutions.
INVESTIGATION OF KR0 POWDER
Fig.1 and Fig.2 show the dependences of the hydrogen generation volume on the reaction time for solutions with different concentrations of the initial components. It follows from Fig.1 that at a KOH concentration of 0.25% the reaction practically does not go, but even at a KOH concentration of 0.5%, the interaction takes place, although its intensity is very low. Increasing the concentration of KOH rapidly increases the intensity of the reaction, so the upper limit was limited to 2% KOH.
Analysis of the curves in Fig.2 shows that for all values of the mass of silicon (mSi), the highest intensity of the reaction is observed at the initial stage. If we arbitrarily limit the initial stage to 5 minutes (300 s), then during this time, an average of 52% of the theoretically possible volume of hydrogen will be allocated for all the indicated values of mSi. Further, the intensity of the reaction gradually decreases due to the consumption of the reagent (Si): in the second stage (5–10 minutes), 26–28% are released, and at the third (10–15 min.) – less than 10% of the theoretically value. The subsequent stages are no longer of interest because of the small volumes of hydrogen, although the reaction can last more than 30 minutes, and the total generation can reach 90–94% of the theoretical value.
Of greatest interest is the first stage of the reaction (0–5 min.), where all the dependences are close in shape to the straight lines drawn from the origin, and their inclination to the abscissa axis characterizes the hydrogen generation rate (the volume of H2 liberated per unit time). With an increase in the initial mass of silicon, the volume of hydrogen release increases very rapidly and is directly proportional to mSi. Table 2 shows the rates of the H2 generation rate in the first stage.
As a characteristic of a particular type of silicon powder, the term "specific rate of hydrogen generation" is used in the literature on this topic, that is, the speed per unit mass (one gram of silicon) [4]. Table 2 shows the values of the specific velocity calculated for the values of mSi shown in Fig.2. These values are very close: the arithmetic mean value is 166 ml with a spread of ± 3.5%.
If we increase the sample weight of the powder mSi by more than 1.2 g, then the H2 generation rate will also increase proportionally to a certain limit, which can be considered as the maximum rate for a given solution volume (75 ml). The fact is that the reaction of silicon with water in an alkaline medium is exothermic and quickly leads to a significant heating of the solution. To estimate the influence of the mass of silicon on the magnitude of the thermal effect, measurements of the temperature of the solution prior to the start of the reaction and in its course were made. The results of the measurements are given in Table 3.
On the basis of the data obtained, it can be concluded that weighed portions of Si powder more than 2 g are not desirable, since they can bring the temperature of the solution to the boiling point. But even with a sample of Si of more than 1.7 grams, rapid solution heating and gas evolution lead to the fact that part of the powder is carried to the surface, and the droplets of the solution are sprayed with the powder particles. Thus, it is expedient to limit the maximum sample of silicon to a value of 1.5–1.6 g.
For practical use, not less than 10–15 liters of hydrogen is required, therefore, increased amounts of silicon powder should be used. To avoid overheating of the reaction mass, it is possible to increase the volume of the alkaline solution. Water in this process serves not only as a source of hydrogen, but also as a reaction medium. According to the stoichiometry, no more than 2.6 ml of H2O per 1 g of Si is required [4], and a multiple excess of water is required to dissolve the reaction product (silicic acid) and to cool the solution. The conducted experiments indicate the possibility of increasing the volume of the solution several times, for example, up to 200 ml. An increase in volume by an order of magnitude or more requires additional investigation, since problems associated with a large volume of hydrogen and an increase in the pressure in the reaction vessel may arise.
ALTERNATIVE POWDERS
In the present work, we did not confine ourselves to the study of one type of silicon powder, but also tested alternative variants. Among them, powders obtained from monocrystalline silicon waste (residues after cutting out a branded single crystal from an ingot) and intermediate or by-products of pyrolysis of monosilane [7, 8] (such a powder is usually unsuitable for the production of silicon ingots due to rapid oxidation in air). Both options can be considered as a way of recycling waste from the production of high-purity silicon.
As the third alternative, a powder of metallurgical silicon with a particle size of less than 10 μm was tested. It is known that grinding the material to particles of no more than 10 μm in size can create an activation effect in some chemical processes [9]. To determine the effect of this factor on the intensity of hydrogen generation, we compared powders obtained from the same material – metallurgical silicon, but differing in particle size.
These powders were tested according to the procedure described above with constant parameters: powder mass is 0.5 g; concentration of KOH is 2%; volume of solution is 75 ml. The results of the tests, as well as comparative literature data, are given in Table 1, where the main characteristics of the powders and the corresponding values of the specific generation rate of hydrogen are indicated.
Analysis of the data in Table 1 shows that the powder of single-crystal silicon with a particle size of 40 μm in intensity of hydrogen generation is close to a polycrystalline silicon powder of the same size. The experimental results are in good agreement with the literature data.
The polycrystalline silicon powder with smaller particles is characterized by a higher intensity of hydrogen generation, but less than one might expect based on the value of its specific surface area. The same applies to the silicon powder obtained by pyrolysis. Apparently, this effect is associated with the behavior of a very fine silicon powder in solution at a high rate of H2 generation, that is, with the removal of powder particles to the surface and their aggregation. Another reason for the discrepancy between the actual generation rate of H2 and the expected (from the point of view of the degree of particle size reduction) can be the oxidation of particles in air.
ANALYSIS OF RESULTS
According to the results of tests of all silicon powders, it can be concluded that for the same particle size the specific hydrogen generation rate depends little on the method of obtaining Si, the crystal structure and the impurity content. The particle size is the main factor influencing the intensity of the chemical reaction, so its decrease allows to increase the rate of hydrogen generation accordingly. However, it should be noted that the generation of hydrogen in the "generator-fuel cell-battery" system has the ultimate goal of charging the battery, and this process, as a rule, requires a long time. Therefore, for practical use, the long and stable operation of a chemical generator is much more important than the rate of hydrogen generation. In addition, the use of fine powders is economically unprofitable, since their cost is inversely proportional to the particle size.
For these reasons, the use of silicon powders with a particle size of less than 40 μm in chemical hydrogen generators is impractical. In this case, powders with a particle size of more than 40 μm are required for significant mSi values.
Powdered silicon is a safe, neutral and cheap material for generating hydrogen, favorably differing from hydrides, which are dangerous, decomposing in air and expensive materials. It is supposed that it is possible to use silicon containing a significant amount of impurities, since practically all of them remain in solution during the reaction of silicon with water.
On the basis of the work done, powders of polycrystalline and monocrystalline silicon with a size of 40 μm or more can be recommended for practical use in autonomous chemical hydrogen generators. Such powders ensure the production of H2 in a volume of more than 1.2 liters per gram of silicon in a short time (about 10 minutes). Based on this, it is possible to calculate the amount of powder needed to produce a given volume of hydrogen over a certain period of time.
The mass of the silicon powder and its size have a significant effect on the generation rate of H2, so it is advisable to use powders with a particle size of more than 40 μm for large silicon loads. On the other hand, autonomous chemical generators now allow very precise control over the hydrogen production reaction and its supply to consumers. As the latter, we consider small-size fuel cells that provide recharging of batteries or direct power to mobile phones and other electronic devices in the field.
CONCLUSION
The processes of hydrogen production by the reaction of powdered silicon with water in a slightly alkaline medium are studied. The dependences of the hydrogen generation volume on the reaction time, the parameters of the silicon powder and the ratio of the initial components of the solution are determined. The chemical activity of the silicon powder was estimated from the specific hydrogen generation rate and the thermal reaction effect. The main limitations associated with the maximum mass of silicon in solution and the particle size of the powder are revealed. The prospects of practical use of the investigated silicon powders in chemical hydrogen generators are shown. ■
Readers feedback
