Production of Glucose by Starch and Cellulose Acid Hydrolysis and its Use as a Fuel in Low-Temperature Direct-Mode Fuel Cells

The use of glucose, which is produced from the acid hydrolysis of starch and cellulose, is studied as a fuel in a low-temperature direct-mode fuel cell (LTDMFC) with an alkaline electrolyte. Glucose is regarded as being as good a fuel as bioethanol, because both the fuels give 2 electrons per molecule in the fuel cell without carbonisation problems. However, glucose can be produced with fewer processing stages from starch and cellulose than can bioethanol. In the LTDMFC the fuel and the electrolyte are mixed with each other and the fuel cell is equipped only with metal catalysts. Cellulose as a fuel is of great importance because the fuel for the energy production is not taken from food production. A description of an acid hydrolysis method for starch and cellulose is presented. Values for glucose concentrations in each hydrolysate are analysed by means of a chromatographic method. Each glucose hydrolysate was made alkaline by adding of potassium hydroxide before feed in the fuel cell. Polarisation curves were measured, and they were found to produce lower current density values when compared to earlier tests with pure glucose. The Coulombic efficiency of pure glucose electrochemical oxidation in LTDMFC, which was calculated from a ratio of detected current capacity (As) to the maximum current capacity with the release of two electrons per molecule, was also found to be very low. Concerning the hydrolysates, the glucose concentrations were found to have values that were too low when compared to the earlier tests with pure glucose in a concentration of 1 M. The further development demands for the system under consideration are indicated. The concentration of glucose in the hydrolysate is essential to achieve high enough current density values in the LTDMFC.


Introduction
The ideal fuel for low-temperature fuel cells in power generation systems would be a directly used fuel in a liquid phase, which allows the fuel to be stored and used more easily than gaseous hydrogen does [1,2]. There are fuel cells which operate with solidphase electrolyte and tissue cultures, which contain a biological electro-catalyst either in the form of enzymes or of living microorganisms, together with proton-selective membranes for the in situ reforming of bioorganic materials [3][4][5]. However, fuels such as glucose or other carbohydrates can easily be utilised by mixing them directly in the electrolyte without the use of any auxiliary equipment or additional microbial cultures. Many studies have been conducted in which metallic catalysts were used in electrolytes with different pH values [1][2][3][4][5][6][7][8][9][10][11].
As far as glucose is concerned, the oxidation reactions take place with a yield of only two electrons per one molecule from 24 available electrons [10]. This yield is equal to aqueous bioethanol, when it is used as a fuel without carbonisation problems in the fuel cells [6,7]. In the earlier studies regarding direct bioorganic oxidations, the anodic catalyst materials were different Pt alloys [4][5][6][7][8][9][10][11].

Bioorganic material as a fuel source
In this study, a low-temperature direct-mode fuel cell (LTDMFC) in which a fuel and an aqueous electrolyte are mixed with each others is tested. The fuel is glucose [C 6 The average cellulose content in different wood species (spruce, pine and birch) varies between values of 33 and 51% [15]. It is possible to reach up to a 90% yield of glucose, depending on the wood species, the concentration of acid, and the process temperatures [14]. However, highly concentrated acids and the process temperatures cause difficult corrosion problems in the process equipment. The principal flow sheet for the acid hydrolysation of polysaccharides is shown in Figure 1.
From Figure 1 it can be seen that the undissolved solids from the hydrolysis stage can be fed to the combustion stage. When the heat production of the combustion stage is taken into account, the heat balance for glucose production is positive, whereas the heat balance for ethanol production is negative because of the heat-consuming concentration [16]. In the development of the direct-mode bioorganic fuel cell, the target is to increase the ionic oxidation reactions in the liquid phase. This would make possible a high electronic yield and electric power production from one bioorganic molecule. In this study, an anode catalyst of alloyed platinum with palladium (Pt-Pd) and a combined cathode catalyst of cobalt porphyrin complex (CoTPP) and of spinel (MnCo 2 O 4 ) are tested in the test fuel cell with an alkaline electrolyte. It was reported earlier that several electro catalysts can not work well in near-neutral-state electrolytes (at pH 7.4) because of low electric conductivity values [17].

Experimental
Test equipment. The test equipment included an LTDMFC (developed by Hydrocell), a recirculation pump (capacity 100 ml / min), an air fan (i.e. a PC cooling fan) and a recirculation tank for the fuel electrolyte solution. The system also included a heating system (water bath) for the solution, but it was not used in these tests. The test fuel cell system is shown in Figure 2. The test fuel cell consisted of a two-cylinder structure. The inner cylinder was an anode electrode (Pt-Pd catalyst) and the outer one was a cathode electrode (combined MnCo 2 O 4 and CoTPP catalysts). The anode catalyst structure was wetted on both sides of its surfaces. The distance between the anode and cathode was 1 mm. The catalyst loadings of the anode electrode were 2.46 mgcm -2 for both the Pt and Pd in the bimetallic combination (in a weight ratio of 1 to 1) with concentrations of 10% on carbon. The catalyst loadings of the combined cathode electrode contained a catalyst loading of 3.15 mgcm -2 of CoTPP with a concentration of 18 % on carbon and of 17.5 mgcm -2 of MnCo 2 O 4 . The anode electrode consisted of a single-layer structure without diffusion material. The cathode electrode was of the double-layer type and contained both the active (catalyst) and the diffusion layers. The electrodes were sintered   Figure. 1. Principal flowsheet of the acid hydrolysation process applied from Ref. [14].  [13]. Because of the larger volumes (10 dm 3 ) the time of hydrolysis was extended to 80-120 min compared to the 40 minutes used in the Ref. [13]. White cellulose paper was weighed and dissolved in 72% H 2 SO 4 for one hour at room temperature (RT) to bring the cellulose into solution, followed by diluting the solution to 4% H 2 SO 4 and then autoclaving the solution at 121 o C for one hour to depolymerise the cellulose to glucose. The glucose content was determined by using an Aminex HPX-87P column (Bio-Rad) at 70 °C with distilled water as the mobile phase at an elution rate of 0.6 ml min-1. All components were analysed with a refractive index (RI) detector. Test routine in fuel cell equipment. The test routine in all the tests was as follows:

Results
Acid hydrolysis. The hydrolysis of starch reached a yield of 1g of glucose per g of starch, when the weighted maize starch concentration was kept below 50 gdm -3 according to Ref. [13]. When the maize starch concentration was increased to 200 gdm -3 the glucose yield decreased to 0.24 g glucose per g of starch. The colouring effect was also very

THERMEC 2009
strong. In order to avoid the strong colouring effect and the formation of inhibitors the starch hydrolysate, which contained 24.8 gdm -3 of glucose, was concentrated by evaporation. When 200 g of cellulose was hydrolysed according to Ref. [13], the glucose yield was 1.32 gdm -3 and the colouring effect was very strong. In addition, the performance of the hydrolysate in the fuel cell was insufficient (the data are not shown here). Therefore, the cellulose hydrolysis was carried out by strong-dilute acid hydrolysis followed by autoclavation. This method resulted in 0.29 g glucose/g cellulose from 50 g of weighed cellulose paper (see Table 1). In all cases only a small part of the glucose was used in the fuel. The cellulose hydrolysate had a 58% glucose utilisation rate. The fuel cell tests. In Figure 3 the measured polarisation curves with three glucose concentrations from starch acid hydrolysis, and with one glucose concentration from cellulose acid hydrolysis, are shown. The measured OCV values related to Figure 3 in respect to the concentrations of the fuels at room temperature (RT) are shown in Table 2. The average measured OCV generation rates (AMGR OCV) in each test are also presented. Glucose with a concentration of 80 gdm -3 was also used, but the determination of the polarisation curve did not produce a higher current density than 0.4 mAcm -2 although the OCV value was 0.678 V. From the results shown in Fig. 3 and Table 2 it can be seen that the glucose content 24.8 gdm -3 (0.14 moldm -3 ) produced the best results in the test direct-mode fuel cell. The higher concentrated glucose values indicated that the glucose has been decomposed in the concentrating. Other compounds (e.g. dissolved sulphate) also disturb the operation of the electro catalysts.  However it was clearly indicated that the hydrolysation and the following concentrating methods have to be much improved to produce higher glucose concentrations in the hydrolysates. It was reported early on that pure glucose in a concentration of 1 M provides current density values from 5 (at RT) to 8 mAcm -2 (at 51 o C) with a voltage of 0.5 V [1,2]. Thus, the detected OCV and current density values with a lower glucose concentration in hydrolysates did not produce values equal to those reached earlier with pure glucose [1,2]. The current density values for pure glucose at RT and at 51 o C were found to be essentially higher than the findings in earlier reports [3][4][5], in which biofuel cells with starch or glucose as a fuel were provided with either a microbial culture and/or a proton-selective membrane. According to the results in this presentation, the development of the acid hydrolysis of starch and cellulose should be developed so as to be able to concentrate glucose in hydrolysates to minimum values of 1 M. However, in the concentrating step the temperature must not exceed 40 o C in order to prevent the decomposition of the glucose as described in Ref. [19]. CE of the fuel cell. The calculated maximum value of the Coulumbic efficiency (CE) with the pure glucose as a fuel according to Equation 2 is shown in Table 3. Table 3. The calculated value for the maximum CE value (n max =2e-) with glucose as a fuel.  The current output was repeated three times as shown in Figures 4 and 5. The fuel cell voltage was dropped from the OCV value down to 0.5 V. The generation of OCV was then followed after each discharging. The calculated CE values for each cycle and in the entire test are shown in Table 4. It is shown in Table 4 that the CE values are very low. One reason for the low CE values is that the voltage drop in each tests stopped at a value of 0.5 V and unusable glucose remained that was usable for the electrochemical oxidation. Therefore it is also assumed that in the electro-chemical oxidising of the glucose molecule the maximum yield of the electrons was two. It can be seen from Table 4 that the value of CE decreased by 38.5 % in the second cycle from the value in the first cycle, and by 20.8 % in the third cycle from the value in the second cycle. According to the tests in the fuel cell, the development of the direct-mode fuel cell also includes selecting and searching for more effective catalyst materials.

THERMEC 2009
Summary Starch and cellulose are interesting candidates as fuel sources for fuel cells, because they need fewer processing steps compared to other biofuels, such as bio-ethanol, before being supplied to the fuel cell. In this study glucose was produced by acid hydrolysis from starch and cellulose. The glucose concentrations in the hydrolysates were found to be too low to be able to produce equal power densities (from 2.5 to 4 mWcm -2 ) as were earlier reported with pure glucose of a 1 M concentration. Thus the concentration of glucose has to be higher in the hydrolysates. The research should also concentrate on the development of more effective and active catalyst materials for the oxidation of glucose, because both the current density and the Coulombic efficiency (CE) values were found to be low in the test fuel cell. The low CE values indicate that the maximum yield of transferred electrons in the electro-chemical oxidation of the glucose molecule was two. However, starch and cellulose are interesting fuel sources for the low-temperature fuel cells. Cellulose is an especially interesting candidate for the production of the electric power, because it is not taken from the raw materials used in food production.