Template Synthesis of Mesoporous Carbon Materials for Electrochemical Capacitors

The effect of AlOOH endo-template on porous structure, electrical conductivity, and electrochemical properties of lactose-derived carbon materials is investigated in the article. It is found that a carbon material with a specific surface of 1707 m2/g and a total pore volume of 1.546 cm3/g can be obtained when the mass ratio of C : AlOOH is 1 : 1. Electrochemical capacitors, formed on the base of the synthesized samples, have specific capacity of a wide range (66–170 F/g), at discharge current 10 mA. The developed mesoporous structure of carbon materials synthesized via the template method allows charge/discharge electrochemical capacitors at currents up to 200 mA, providing the value of specific capacity 121 F/g.

As a rule, CMs have microporous structure, when using the methods of thermal or thermochemical modification. Activation thermochemical processes, which lead to an increase in the pore volume in a carbon matrix, are due to the reactive interaction of H 2 O or CO 2 molecules with carbon atoms at temperatures of 800-1100°C [24,25].
Carbon structures with the required ratio of microand mesoporous are synthesized using reagents KOH, K 2 CO 3 or ZnCl 2 [26][27][28][29][30]. When heated composite mixtures consisting of a carbon precursor and indicated compounds, K 2 O and ZnO particles are formed, which form mesopores of 2-6 nm in size after removing from the carbon matrix. H 2 O, CO 2 , and HCl molecules, released due to thermodissociation of reagents, are involved in the oxidation of carbon atoms and contribute to the growth of pore volume in the material.
To obtain spatially ordered porous CMs the methods of exo-and endotemplating are used. The first method involves thermolytic decomposition of a carbonaceous precursor in the volume of pores of inorganic matrices-zeolites [31][32][33], silica molecular sieves or alumina membranes [34][35][36][37]. The voids of exo-templates as a result of organic matter carbonization are carbon skeletons, and free volume as pores and channels is formed after dissolution and removal of the matrix. The endo-template method consists in filling the carbon precursor volume by nanoparticles of an inorganic oxide material (for example, SiO 2 or Al 2 O 3 ) and washing of these particles from CMs by an aqueous solution of KOH, fluoride or chloride acid after carbonation of the precursor [38][39][40].
The mesoporous aluminum hydroxide AlOOH, formed during thermal decomposition of aluminum nitrate nonahydrate Al(NO 3 ) 3 • 9H 2 O [41], is one of the endo-templates that can be used to correct the porous structure of CMs. Therefore, the purpose of this work is to obtain CMs using AlOOH endo-template and to find out its effect on the porous structure, electrical conductivity of CMs, and electrochemical properties of ECs with electrodes based on them.
EXPERIMENTAL Carbon was obtained from the crystalline D-lactose (C 12 H 22 O 11 ), as a precursor, at 400°C for 30 min in an air atmosphere. Subsequent activation of carbon was carried out at 800°C for 30 min in an argon atmosphere. This material is denoted as sample 1.
Sample 2 (used for comparison) was formed by mixing the saturated solution of the activating agent (KOH) and D-lactose in a ratio of 1 : 3 and then heating the mixture to a liquid homogeneous consistency. After evaporation (100-175°C) and caramelization (220-230°C), the composition mixture was carbonized at 350°C for 1 h. After that, the mixture of carbon with the reagent was heated in an argon atmosphere at 800°C for 30 min.
D-Lactose was mixed with aluminum nitrate nonahydrate Al(NO 3 ) 3 • 9H 2 O in ratios of 3 : 1 (sample 3) and 1 : 1 (sample 4) for realisation of an endo-template synthesis of CMs. After thermolytic decomposition of disaccharide precursor containing aluminum nitrate molecule, a composite material of C-AlOOH is formed. The composite material was impregnated by KOH solution, dried, and calcined at 800°C for 40 min. KAlO 2 formed in the CM volume was washed by water.
The structural and adsorption characteristics of CMs were analyzed on the base of low-temperature (77 K) nitrogen adsorption-desorption isotherms recorded using a Quantachrome Autosorb Nova 2200e adsorption analyzer. Before measurements, the samples were heated at 180°C for 24 h. The specific surface area (S BET ) of the samples was calculated via the Brunauer-Emmett-Teller (BET) method [35]. The specific surface (S micro ) and the volume (V micro ) of micropore was calculated via t-method [36], the surface (S meso ) and the volume (V meso ) of mesoporus were calculated as the difference between the total surface (volume) of pores and micropores. The total pore volume (V p ) was evaluated from the nitrogen adsorption at p/p 0 ≈ 0.99, where p and p 0 denote the equilibrium and saturation pressure of nitrogen at 77 K, respectively. PSD were calculated via the nonlocal density functional theory [37] using an equilibrium model with slit-shaped pores (samples 1 and 2) and slitshaped/cylindrical ones (samples 3 and 4).
To explore the conductive characteristics of CMs, a condenser system was used that consisted of two copper electrodes between which the test samples were located. Impedance hodographs Z'' = f(Z'), where Z' and Z'' are real and imaginary parts of the system complex resistance, respectively, were carried out using a Metrohm Autolab FRA-2 (Frequency Response Analyzer) at frequencies of 10 -2 -10 5 Hz and a voltage amplitude of 10 mV. Taking into account the geometric parameters of the samples, the specific values of resistances, electrical conductivity and frequency dependences of the electrical parameters were calculated according to the equation: and , А and d-electrode surface area and sample thickness, respectively. The complex specific conductivity was determined from the relation: (2) where , , , and complete specific conductivity was calculated as in [45]: The mixture of a CM and a conductive addition (graphite KS-15, Lonza) in a mass ratio of 75 : 25 was used for the manufacture of electrodes of symmetrical ECs. The electrodes were impregnated with an electrolyte (30% KOH aqueous solution), separated by nonwoven polypropylene, placed in a two-electrode cell, and sealed.
Methods of galvanostatic cycling and cyclic voltammetry were used for the study of electrochemical properties of ECs with electrodes on the base of experimental samples of CMs. Measurements were carried out using a Metrohm Autolab with GPES (General Purpose Electrochemical System) software.
Galvanostatic measurements were carried out in a voltage range of 0-1 V, the charge/discharge current of EC varied in a range of 10-200 mA. The specific capacity of the electrode material was calculated according to the formula: (4) where І-charge/discharge current, t-discharge time, U max -maximal voltage, ΔU-voltage drop when discharge circle is switched on, m-CM mass.
The internal resistance of ECs was determined via voltage drop after ten charge/discharge cycles according to equation: (5) Cyclic voltamperograms of ECs were recorded in the voltage range of 0-1 V; the scan rate was 1, 5, 10, and 20 mV/s.

RESULTS AND DISCUSSION
Experimental dependences of changes in volume V of adsorbed/desorbed nitrogen from the value of its relative pressure р/р 0 at constant temperature ( Fig. 1) provide the necessary information for calculating the structural and morphological characteristics of CMs.
An important feature of isotherms for samples 1 and 2 is that the limit value of the adsorption value is recorded when the pressure of saturated vapour reaches p 0 . With these features, adsorption dependence is characteristic for monomolecular adsorption ' ' of substances by microporous sorbents. According to the IUPAC classification, they belong to type I of adsorption isotherms [39]. Those isotherms are characterized by an open loop of hysteresis due to the presence of long and narrow pores with narrow necks, the size of which is close to the size of nitrogen molecules [39]. In particular, the isotherm hysteresis for sample 1 is recorded in the entire range of relative pressure р/р 0 and for sample 2-within 0.1-1.0 р/р 0 .
Isotherms characterize the processes of polymolecular adsorption in micro-and mesoporous materials of organic origin for samples 3 and 4 [40]. The shape of the hysteresis loop belongs to the type H4 [39], which is associated with capillary condensation in mesopores.
As follows from the calculation data of parameters of CMs porous structure (Table 1), non-activated carbon (sample 1) has a moderate specific surface of 499 m 2 /g and a small pore volume of 0.222 cm 3 /g. As a result of activation thermochemical processes when using KOH, the specific surface area and the total pore volume of the material increase by 2.1 times (sample 2). There is also a slight increase in the mesopores portion from 9 to 18%.
Significant changes in the porous structure are observed as a result of the use of AlOOH endo-template. In particular, although there is no increase in the specific surface for sample 3, the pore volume increases almost thrice, and there is a significant development of mesopores. The specific surface area and the pore volume increase by 3.4 and 7 times, respectively, and the portion of mesopores is 85% when a ratio of C : AlOOH is 1 : 1 (sample 4).
Changes in the porous structure of CMs as a result of thermochemical activation and endo-templating are confirmed by the PSD curves (Fig. 2).
The non-activated carbon is microporous, i.e. most of pores are 1.17 nm in size (Fig. 2, curve 1). Sample 2, in general, is also microporous, with peaks close to 1.3 and 1.5 nm. The material contains an insignificant amount of mesopores (18%) with a maximum close to 2.2 nm (Fig. 2, curve 2). Sample 3 is characterized by an intense maximum near 1 nm, which corresponds to micropores, and a rather wide distribution of mesopores in a range from 2 to 25 nm (Fig. 2, curve 3). The most noticeable changes in PSD relative to the initial material are observed for sample 4 (Fig. 2, curve 4). The endo-templating of lactosebased precursor by aluminum hydroxide creates preconditions for the formation of micro-mesoporous structure, in which the maximum in the distribution of micropores is close to 1.3 nm, and of the mesopores-5 nm. Less intense maxima in mesopores distribution are observed at 2.4, 3.1, and 3.8 nm. Comparing the results of low-temperature porometry with the study of thermal decomposition of aluminum nitrate nonahydrate Al(NO 3 ) 3 • 9H 2 O [34], the differences in the PSDs of the materials become clear. A composite material C-AlOOH, the volume of the carbon matrix in which is filled by boehmite globules of 3-5 nm in diameter, is formed during the thermolytic decomposition of a disaccharide precursor containing a molecule of aluminum nitrate. Removing boehmite from the CM volume by leaching allows to obtain mesoporous carbon with a large pore volume.
A high value of electrical conductance is one of the requirements that a material must possess to be used as    (Fig. 3). As can be seen from the diagrams, there is an inductive component in the high-frequency region of spectrum for all samples. It is related to the percolation mechanism of the passing of electric charge carriers through a mixture of conducting and non-conducting particles of a material, where the pores are non-conducting particles. Thus, the carbon structure of the electrode material becomes an analogue of inductance due to a complex trajectory of the charge carrier movement. Only a real component of the resistance is changed when the frequency of the electric potential decreases. The value of the imaginary resistance first decreases but then remains unchanged.
Using Z' і Z'' values and formulas (1)-(3), real σ', imaginary σ'', and complete σ* specific conductivities of CMs were calculated and their dependence on the frequency was found. Since the values of the real components of electrical conductivity are greater by 2-3 orders of magnitude than an imaginary one at frequencies below 10 3 Hz, the contribution of the latter to the complete conductivity of CMs can be neglected. Under these circumstances, the behaviour of the total resistance will be determined, mainly, by the frequency dependence of the real component of electri-cal conductivity. Representation of the frequency dependences of electrical conductivity in semilogarithmic coordinates (the plot of σ* vs. log f) allows to determine the specific conductivity as the value of σ* at f → 0 (Fig. 4).
The calculations have shown that high-temperature activation of the initial carbon material by KOH increases its specific conductivity by 1.34 times (Table 2). Such result of a KOH effect on the electrical conductivity of CMs is due to the intercalation of K + ions into the interlayer of carbon microcrystallites, which contributes to their structural rearrangement at high temperatures in the graphite-like state. When using the AlOOH template, the specific conductivity of the material is reduced due to an increase of the mesoporous number, which acts as a barrier for passing the electric charge curriers through the sample.
To find out a possibility of using CMs obtained by the endo-template method, as electrode of ECs, galvanostatic studies were performed. Specific capacity of electrode materials and internal resistance of ECs were calculated by formulas (4) and (5), respectively, from dependencies U = f(t) (Fig. 5).
Analyzing the dependences shown in Fig. 6a, one can conclude that sample 2 has the largest specific capacity (176-157 F/g) at discharge currents of 10-100 mA. Specific capacity of the initial material (sample 1) at these discharge currents is smaller (160-137 F/g) due to its smaller specific surface. The use of endo-template AlOOH leads to a decrease in specific capacity of ECs formed on the base of sample 3 and 4. For sample 3, this is due to a two-fold decrease of the specific surface of micropores (relative to sample 1), which was involved in the formation of a double electric layer (DEL) and provided a capacity of EC (the value of the material specific surface is almost the same). Smaller values of specific capacitance of EC based on sample 4 (147-132 F/g) can be explained by a significant  decrease in the micropore surface relative to the total surface of the material (S micro /S BET is 91% and 27% for samples 1 and 4, respectively). The redistribution in the ratio of micro-and mesopores, caused both by thermochemical activation and endo-templating, changes the internal resistance of ECs (Fig. 6b). There is a certain correlation between the values of the specific conductivity of CMs (Table 2) and internal resistance of ECs.
One of the criteria for choosing the maximum charge/discharge current during cycling of ECs is a decrease in the voltage drop ΔU no more than 20% of the maximum voltage (1 V), i.e. 0.2 V when the discharge circle is switched on [41]. For ECs, formed on the basis of samples 1 and 2, indicated decreasing in voltage is observed already at discharge current of 100 mA. In contrast, ECs on the base of samples 3 and 4,  can withstand cycling at currents of 110-170 mA and 110-200 mA, respectively, providing a specific capacity of 54-52 F/g and 131-121 F/g. The use of endotemplate during material obtaining leads to an increase in the mesoporous number (Table 1), which, as transport pores, provides rapid access of electrolyte ions to the material surface at large charge/discharge currents. According to cyclic voltammetry voltamperograms, recorded at lower scan rate (1 mV/s), have a symmetrical shape similar to the rectangle one (Fig. 7a). This form of voltamperograms shows that the accumulation of charge in the EC occurs due to the formation of DEL at the interface between the surface of the electrode material and the electrolyte, and the secondary oxidation-reduction reactions (socalled Faraday processes) are absent.
The rectangular shape of the voltamperograms is disturbed at higher scan rate (20 mV/s, Fig. 7b) due to the increase in the resistance of the electrochemical system as a result of local decrease in the electrolyte ions concentration in the micropores of the electrode material-so-called effect of "starvation" of the electrolyte [42]. The most noticeable changes are observed for samples 1 and 2, which are dominated by micropores. For samples 3 and 4, dominated by mesopores, the deviations from the rectangular form are less significant, which confirms the results of galvanostatic studies on the possibility of charge/discharge of ECs based on these materials with large currents.

CONCLUSIONS
The method of getting synthetic CMs using lactose (as a precursor) and mesoporous aluminum hydroxide AlOOH (as a template), which can be obtained during the thermal decomposition of aluminum nitrate nonahydrate Al(NO 3 ) 3 • 9H 2 O, is proposed.
The specific surface area and the pore volume of the synthesized CM at a ratio of C : AlOOH = 1 : 1 increase by 3.4 and 7 times, respectively, in relation to the initial material, the mesoporous portion being 85%. Increasing mesopores results in reducing the specific conductivity of the material two-fold.
Mesoporous carbon-based EC has a slightly lower specific capacity (147 F/g at 10 mA discharge current) in contrast to that on lactose-derived carbon (160 F/g at the same discharge current), but due to developed mesoporous structure allows discharging of EC by currents up to 200 mA.