summarizes the ash content on a dry basis (d.b.), the ultimate analyses on a dry ash free basis (d.a.f.), and the preparation yields of the different carbon catalysts obtained in this work. For the sake of comparison, the ash content and the ultimate analyses of the starting material has been also included. In this sense, the black liquors in Kraft pulp production plants, from which carbons are obtained, are alkaline solutions. The evaporation of these black liquors yields solid lignin with a high content of sodium compounds. The ash content of the bio-chars is around 30%, which is approximately twice the value of the lignin raw material, L (16.9%). This increase is due to the loss of volatile matter during the carbonization, and it is in agreement with the yield of the process (around 50%), suggesting that the inorganic matter is not significantly volatilized in this stage. After the carbonization process of L at the different temperatures, the carbon and the sulfur contents significantly increased, meanwhile, the amount of oxygen and hydrogen considerably decreased, due to the devolatilization and aromatic condensation occurring during the carbonization process.
The washing of the L900 char removed most of its inorganic matter, obtaining a final ash content of 4%. However, this process was not totally effective, probably due to the narrower, porous structure of this char. This treatment also reduced the oxygen and the sulfur amounts, and consequently increased the carbon content in the obtained carbon, L900W.
Gasification of L900 and L900W chars to obtain L900G and L900WG activated carbons, respectively, produced an increase of the ash content (L900G: 44%; L900WG: 8%), as a consequence of the carbonaceous matter loss (burn-off (B.O.) = 44% and 60%, respectively). This loss was also accompanied by an increase of the oxygen content due to formation of new carbon-oxygen surface groups which were created by chemical reactions between the activation agent (CO2) and the carbon surface.
2.1.1. Textural Properties
shows the N2
adsorption-desorption isotherms at −196 °C of the different carbon catalysts prepared. The catalyst obtained by the direct carbonization at 500 °C presents low N2
adsorption in the entire relative pressure range of the isotherm, typical of a nonporous solid. The increase of the carbonization temperature produced a slight development of a porous structure, as is shown by the increased amount of adsorbed N2
a), due to the activation produced by sodium oxide when the carbonization temperature was increased. The isotherms of the non-washed catalysts obtained at 900 °C before and after the gasification step (Figure 1
b) are very similar, due to the presence of inorganic matter that hinders the development of porosity during the gasification step, probably by hampering the complete devolatilization [23
]. In contrast, the washing treatment releases to a great extent the porosity of the sample L900W due to the removal of the inorganic matter, mainly composed of Na and S. The gasification of this sample produces high porosity in L900WG, showing a type IV isotherm with an increase of the adsorbed volume in the whole relative pressure range, which corresponds to a large amount of porosity with micropores and mesopores. The presence in this sample of H4 type hysteresis loops is associated to slit-shaped pores.
c,d also shows the micropore size distribution of the different carbon catalysts. Most of the carbon materials have micropores within an effective pore width in the range of 4–15 Å. The increase of the carbonization temperature (Figure 1
c) produces a slight increase of the micropore size as a consequence of a higher devolatilization. In this sense, the samples treated at 900 °C present a multimodal pore size distribution. After the washing of the L900 carbon, an important widening of micropore size is revealed, showing a maximum around 6 Å. The following gasification of this sample (L900WG) produces a further broadening of the porosity, presenting another important maximum around 80 Å.
summarizes the values of the textural properties that characterize the porous texture calculated from N2
adsorption-desorption and CO2
adsorption isotherms of the different carbon catalysts. The catalysts obtained directly by carbonization at the different temperatures show significantly higher values of specific surface area from the CO2
) than that of apparent surface area (ABET
), indicative of a very narrow porous structure, with mean pore size values under 7 Å [24
]. After the washing treatment, the sample L900W has high values of micropore volume obtained from the N2
), mesopore volume (Vmes
), and micropore volume derived from the CO2
) than those of L900, revealing that the removal of the inorganic matter (from 31 to 4%) removes to a great extent the porosity. The similarities between ABET
values, obtained for L900W carbon, are typical of microporous solids.
Gasification of L900 does not produce any apparent development of the porosity. This fact probably can be associated to the high ash content (44%) of the sample L900G, which may block access to the pores. In contrast, gasification of L900W generated a notable development of the porosity, obtaining an ABET as high as 850 m2 g−1. In this case, an ABET higher than the value of ADR indicates the presence of a wide porous structure. Furthermore, Vmes is higher than Vt, indicating a high contribution to mesoporosity.
2.1.3. Surface Chemistry
X-ray photoelectron spectroscopy (XPS) analyses were carried out in order to evaluate the surface element distribution and the surface chemical composition of the carbon catalysts. The main components of the carbon catalysts are carbon, oxygen, sulfur, and sodium. It is important to highlight the low carbon content of these carbon catalysts (L500, L700, and L900), these values are considerably lower than those obtained by other lignocellulosic chars [23
], indicating that the main part of the inorganic matter is present in the outer surface. The weight ratio O/C determined by XPS in these three carbon catalysts is approximately 3, and the ash content is around 30%. If it is assumed that most of the inorganic matter is composed of sodium, a weight ratio Na/CXPS
of approximately 2 can be observed. These values suggest that sodium is mainly present in the form of sodium salt (carbonates, sulfates, and sulfides). After the washing treatment, more of the ashes are removed and, consequently, the carbon content significantly increases in detriment of the oxygen, sulfur, and sodium contents.
On the other hand, it was expected that the carbon content of L900G would decrease with respect to that of L900 as a consequence of the carbon loss by the gasification. However, the gasification process increased the carbon amount and decreased the sulfur and sodium percentages. This could be due, in part, to the carbonation of the sodium ions present on the surface during the gasification process. In this sense, some volatilization of the sulfur compounds can take place during gasification with CO2, as suggested by the decrease of the sulfur content close to the formation of sodium clusters that reduce the relative concentration of this atom in the external surface, so increasing the carbon content. Gasification of the carbon L900W produces, as expected, an increase of the sulfur and sodium contents and a reduction of the carbon amount caused by the carbon gasification.
shows the C1s, O1s S2p, and Na1s spectra of the different carbon catalysts. C1s spectrum of the samples carbonized at the different temperatures shows two broad peaks associated to the presence of different chemical states. In this sense, it shows the presence of C-C (284.5 eV), C-OH (285.5 eV), C-O-C (286.3 eV); C=O (288 eV), and CO32−
bonds (290 eV). The contribution of carbonates is very important in these samples, being even higher than those of the other species. The C1s spectrum of the washed carbons (Figure 2
e) shows an important reduction of the carbonates peak at the same time as an increase of the peak associated to carbon in the C-C bond. O1s spectrum shows a wide and similar band around 531.5 eV in all the carbon catalysts, this band can be deconvoluted into several peaks in the range of 530.4–530.8, 532.4–533.1, and 534.8–535.6 eV, which have been ascribed to C=O groups, C-OH and/or C-O-C groups, and chemisorbed oxygen and/or water, respectively. Furthermore, an important band around 531.1 eV can be also obtained due to the oxygen atoms in the carbonate ions of sodium carbonate [26
]. These samples also present a broad peak found around 535.5 eV, which corresponds to the sodium Auger peak (Na KLL). This last peak and the one found at 531.1 eV are considerably decreased in the washed samples.
The sulfur spectrum presents a doublet corresponding to S2p3/2 and S2p1/2. The S2p spectrums of the samples L500 and L700 show two contributions at 167.6 and 169.4 eV that could be associated to sulfite and sulfate groups, respectively. The carbonization at the highest temperature produces the appearance of another peak at 162.1 eV, related to the formation of sulfide (thiolate) groups. The washing treatment reduces considerably the presence of sulfur, only leaving some sulfur in the form of sulfate groups. Only the sample L900WG has a peak at 163.4 eV, associated to disulfide or sulfide and unbound thiol. The Na1s spectrum for the carbon catalysts obtained at different carbonization temperatures can be assigned to sodium carbonate, sodium sulfate, and sodium sulfide, with the more significant contribution associated to sodium carbonate. These species almost disappear with the washing treatment.
displays the sodium dispersion derived from SEM equipped with energy-dispersive X-ray analyses (SEM-EDX) analyses of the different carbon catalysts obtained at 900 °C. As can be seen, sodium is well distributed in the sample carbonized at 900 °C. After the washing treatment, its distribution is very homogeneous, in spite of the important reduction of its content. The formation of sodium clusters can be noticed in the sample L900G, in agreement with the decrease of Na content detected by X-ray photoelectron spectroscopy (XPS) analyses, however, those clusters are not observed in the carbon L900WG. This difference can be associated to the lower Na content of this sample. Furthermore, Na is a well-known catalyst of the gasification reactions, and its high amount in the sample L900G can produce local increases of the temperature, favoring the sintering of clusters in that case.
The surface chemistry of the carbon catalysts was also analyzed by the temperature-programmed desorption (TPD) technique, which is used to characterize the carbon-oxygen groups of the samples. Carbon materials release different oxygen surface groups whose nature and quantity depend on the starting material and the activation treatment [27
]. The acidic groups (carboxylic, lactonic) evolve as CO2
, whereas the non-acidic (carbonyl, ether, quinone) and phenol groups produce CO. Anhydride evolves as both CO and CO2
]. Figure 5
a–d shows, respectively, the CO2
and CO evolved during the TPD experiments performed on the different carbon catalysts. Decomposition of surface oxygen complexes such as CO and CO2
is quite high in all the carbon catalysts. In general, the evolution of CO is higher than that of CO2
. The chars obtained at the different temperatures (Figure 5
a,c) show high amounts of very stable carbon-oxygen groups, which decompose at temperatures higher than 600 °C, where decomposition of both inorganic compounds and carbonyl and quinone groups takes place. The increase of the carbonization temperature produces an increase of the CO and CO2
evolution. In the case of CO, the last peak associated to more stable carbon-oxygen complexes remains very similar after carbonization at the different temperatures, however, the peak around 700 °C decreases in the carbon catalyst prepared at 900 °C, probably as a consequence of the higher temperature of the treatment. On the other hand, CO2
profiles increase with the carbonization temperature. The evolution at low temperatures is almost negligible, and most of the evolution is centered on 700 °C, which probably can be associated to the decomposition of sodium carbonates [29
]. However, the different chars have very similar ash and Na content (determined by XPS analyses), so their differences must be associated to the increase of the sulfur content and its possible reduction from sulfate to sulfide groups. This reduction would produce the oxidation of the carbon to carbonate surface in the form of groups, which decompose as CO2
. Nevertheless, another feasible explanation could be the formation of sodium atom clusters in the external surface, decreasing the relative amount of sodium, as revealed by XPS [30
b,d collects the CO and CO2
evolutions of the carbon catalysts obtained at 900 °C. When L900 is washed, an important reduction of the CO and CO2
evolution is noticed, related to the considerable removal of the inorganic matter. With regard to the gasified samples, gasification of carbon samples with high ash content produces the increase of the amount of oxygen groups that evolved as CO2
during the TPD, as occurs in the L900WG catalyst. However, the L900G sample shows a similar CO profile, and a CO2
evolution substantially decreased. Specifically, the peak around 700 °C is significantly reduced. This decrease may be due to decomposition of alkaline metal carbonates, which can take place at temperatures below the melting point (820 °C) when heating carbonate-carbon mixtures. In this case, the temperature of the gasification treatment was 750 °C, so decomposition of the groups of L900 was probably produced during its gasification [31
]. Table 3
also summarizes the amounts of CO and CO2
evolved with temperature during the TPD experiment for the different carbon catalysts. As can be seen, the catalyst L900 presents the highest CO2
evolution; meanwhile the sample L900G evolves the highest amount of CO.
2.1.4. Catalytic Activity
Alcohol molecules, mainly 2-propanol, are widely used for testing the acidic-basic character of heterogeneous catalysts [20
]. In this sense, alcohol dehydrogenation products (aldehydes and ketones) are preferentially formed on basic catalysts, while dehydration products (olefins and ethers) are favored when acidic sites are present. Figure 6
shows the evolution of the steady-state conversions of 2-propanol and the selectivity to acetone and propylene with the temperature for all the carbon catalysts (Partial alcohol pressure: PoALC
= 0.0185 atm; space time W/FoALC
= 0.073 g s μmol−1
). As expected, 2-propanol conversion increases with the reaction temperature. The increase of the carbonization temperature produces an enhancement of the steady-state conversions, showing the catalyst carbonized at 900 °C as the highest steady-state conversions of the three chars. On the other hand, the series of the carbon catalysts prepared at 900 °C (Figure 6
b) follows the sequence L900WG > L900 > L900W > L900G. The most active catalysts correspond to the carbon with the highest development of porosity and a high dispersion of Na, meanwhile the catalyst L900G presents quite a low apparent surface area and a poor dispersion of sodium, as evidenced the SEM-EDX analyses.
The main product is acetone in most of the carbon catalysts, obtained as 2-propanol dehydrogenation product. However, the catalyst L500 and to a lesser extent L900W also yield an important amount of propylene, indicating the presence of both acid and basic surface sites. Furthermore, most of samples carbonized at 900 °C, showing selectivity to acetone higher than 95% at reaction temperatures higher than 350 °C.
In this sense, the hydrogen atom on the hydroxyl group of alcohols is acidic in the presence of basic catalysts and/or basic sites. So, the interaction between a basic site (for example, quinone groups present in these catalysts, especially in L900WG) and an alcohol molecule causes the abstraction of the proton from the -OH group, forming a surface alkoxide intermediate. The following abstraction of a Hα or Hβ from the corresponding alkoxide intermediate leads to ketone or alkene formation, respectively; the latter is less favored.
Decomposition of ethanol and methanol was also evaluated for the carbon catalyst more active in the 2-propanol decomposition of L900WG. Figure 7
shows the evolution of the steady-state conversions of ethanol (a) and methanol (b). Furthermore, the selectivity to aldehydes (filled marks) and to ethylene and dimethyl ether (hollow marks) was also included as a function of the reaction temperature in a helium flow (PoALC
= 0.0185 atm; W/FoALC
= 0.073 g s μmol−1
). Although the conversions observed for these catalysts were not as high as some reported in the literature [32
], selectivity of ethanol to acetaldehyde was higher than 80% in the whole range of conversions studied, values considerably larger than those obtained by those authors. In addition, our catalyst was obtained from a renewable source by a cheap and easy synthesis method. The stability of this catalyst for the ethanol conversion was also evaluated at two different temperatures: 420 and 450 °C (Figure 7
c). The catalyst showed the same activity along the experiment, with the same selectivity toward acetaldehyde.
With regard to methanol, conversions were comparable to those reported in the literature [35
]. The selectivity to formaldehyde increased gradually with temperature and reached a maximum (approximately 80%) at the highest temperature studied, which is near the commercial operation temperature. At the lowest temperatures (<475 °C), methanol was mainly dehydrated to dimethyl ether, making the yield of olefins (propylene or butenes) very low. The possible deactivation of the catalyst was also analyzed for 6 h at 450 and 540 °C. The conversion remained the same during this reaction time.
In general, the selectivity of ethanol and methanol to acetaldehyde and formaldehyde, respectively, were lower than those observed for 2-propanol to acetone. Specifically, the formation of alkene was higher when ethanol was used versus 2-propanol, due to the chemical similarities of Hα and Hβ to ethanol. In the case of methanol, there is a competition between the nucleophilic substitution to yield dimethyl ether (DME) and the abstraction of Hα, which is slower as it is less favored than in higher alcohols.