The rapid development of the biofuels production technologies from differ­ent biomass types has given rise to the biomass and food markets as it was aforementioned. Besides the use of residual biomass, research and in par­ticular biotechnology has moved forward into seeking alternative biomass production technologies that will offer higher yields per hectare as well as lipids and carbohydrates, which are not part of the human and animal food — chain, avoiding competition between food/feed and energy crops. Targeted research efforts have offered a promising solution by the selection of unicel­lular microorganisms for the production of biofuels [48][49]. Micro-algae are photosynthetic microorganisms that can produce lipids, proteins and car­bohydrates in large amounts over short periods of time.

Micro-algae are currently considered a prominent source of fatty acids, which offers large yields per hectare with various fatty acid foot-prints from each strain. In fact, there are certain strains that offer fatty acids of in­creased saturation (small content of unsaturated fatty acids) and of smaller carbon-chain length such as Dunaliellasalina, Chlorella minutissima, Spi — rulina maxima, Synechococcus sp. [50] etc. Another advantage of algal oils is that their fatty acid content can be directed to small carbon-chain mol­ecules either genetically or by manipulating the aquaculture conditions such as light source and intensity [51], nitrogen starvation period [52], nutrients and CO2 feeding profiles [53].

Micro-algae and their products formulated the so called 3rd generation biofuels, as they incorporate various characteristics, which render them superior over other biofuels and biomass types. Micro-algae can also be produced in sea water [54] or even waste water, while they are biodegrad­able and relatively harmless during an eventual spill. Furthermore, their yield per hectare can reach 3785-5678lit, which is 20-700 higher over the conventional energy crops yield (soy, rape and palm). The lipids contained in most micro-algal oils have a similar synthesis with that of soy-bean oil, while they also contain some poly-saturated fatty acids with four double bonds. As a result catalytic hydrotreating of micro-algal oil is the most promising technology for converting it into biofuels.

R. BINDIG, S. BUTT, and I. HARTMANN

6.1 INTRODUCTION

The use of biomass or bioenergy can be traced back to the beginning of hu­man civilization when people started to burn wood for heating and cook­ing purposes. Ironically, after so many years have gone by, wood still re­mains the largest biomass resource in the world [1]. However, one major difference which has occurred over this period of time is the introduction of the concept “modern biomass” which states the usage of traditional bio­mass resources in highly efficient systems. This concept has been put into practice with more conviction and determination during the last decade, particularly in Europe, due to ever rising CO2 levels in our environment. By now, it is an established fact that about 10-30% of total energy demand for hot water supply and domestic heating in European countries like, Aus­tria, Germany, Sweden and Finland is provided by small scale biomass

Emission Abatement at Small-Scale Biomass Combustion Unit with High-Temperature Catalysts. © Bindig R, Butt S, and Hartmann I. Journal of Thermodynamics & Catalysis 4,125 (2013), doi: 10.4172/2157-7544.1000125. Licensed under Creative Commons Attribution License, http://creative- commons. org/licenses/by/4.0/.

combustion systems [2]. Moreover, it has been also concluded that despite the vast spread of technologically advanced small scale combustion devices in European countries (like countries mentioned above) during the recent years, still the old biomass combustion systems (stoves and boilers) occupy more consumers [3]. These conventional systems which are based on natu­ral draft play a pivotal role in contributing to the high emission levels of par­ticulate matter (PM), carbon monoxide (CO), organic gaseous compounds (OGC) and polycyclic aromatic hydrocarbons (PAH). These facts and fig­ures have triggered an enormous understanding and awareness among the researchers as well as local population concerning harmful pollutants emit­ted by residential biomass combustion systems. For this reason inefficient small scale biomass combustion systems have been heavily criticized and demanded to be replaced by new efficient technologies.

Speaking of older and newer technologies, it has to be mentioned here that two types of technologies exist concerning small scale bio­mass combustion systems. The old biomass combustion systems are based on “up-burn” which is in a process of being rapidly replaced by “down-firing” systems (new technologies). As mentioned above, these older systems are a main source of PM1 (particles with diameter less than 1 pm) in European countries. It has been also concluded that such particles serve as a purpose of “support” onto which carbonaceous par­ticles (organic compounds and soot) are deposited which are primarily responsible for the adverse health effects [4]. So in order to counter such an undesired release of pollutants, particularly from small scale biomass systems, a concept has been conceived according to which “down-firing” technology will be implemented in specially designed wood log stove in combination with catalytic treatment in order to abate harmful emis­sions to minimum possible values. It is noteworthy to mention here that the abatement of emissions through catalytic treatment from small scale biomass combustion systems has not been studied or implemented on a wide scale. So this novel concept of integrating catalytic components in different parts of the stove i. e. grate, walls of combustion chamber and the base will open more channels and schemes in order to accomplish the acceptable emission levels coming out of biomass combustion systems particularly, those used for residential purposes.

FIGURE 2: Time-dependent behavior of temperature during the reference experiment.

In the past, the process of catalysis has been strongly linked to chemi­cal and refinery industries. However, recently the catalytic converters have been deployed and installed in automobiles, biomass fired boilers and power generation facilities in order to promote the environmentally friendly usage of technological devices. It has been estimated that the mar­ket of catalysis around the world worth around US$9 billion, out of which, one third is occupied by the environmental catalysis. So building on this ever growing trend of environmental catalysis, this article gives a further insight into the integration of catalytic components in a downdraft wood log stove to foresee the feasibility of this novel approach to resolve the problem of high emissions (e. g. carbon monoxide, volatile organic com­pounds, dust particles etc.) at small-scale furnaces for solid biomass.

The rational design and optimisation of palladium selox catalysts require a microscopic understanding of the active catalytic species responsible for alcohol and oxygen activation, and the associated reaction pathway to the aldehyde/ketone products and any competing processes. A key char­acteristic of palladium is its ability to perform selox chemistry at tem­peratures between 60 and 160 °C and with ambient oxygen pressure [39, 142] via the widely accepted oxidative dehydrogenation route illustrated in Scheme 3 [39, 67]. Whether O-H or C-H scission of the a-carbon is the first chemical step remains a matter of debate, since the only fundamental studies over well-defined Pd(111) surfaces to date employed temperature — programmed XPS [143] and metastable de-excitation spectroscopy (MDS) [144] with temporal resolutions on the second ^ minute timescale, over which loss of both hydrogens appears coincident. However, temperature — programmed mass spectrometric [145] and vibrational [146] studies of unsaturated C1-C3 alcohols implicate O-H cleavage and attendant alkoxy formation over Pd single crystal surfaces as the first reaction step [142, 147]. It is generally held that the resultant hydrogen adatoms react with dissociatively absorbed oxygen to form water, which immediately desorbs at ambient temperature thereby shifting the equilibrium to carbonyl for­mation [39, 67]. Temperature-programmed XPS studies of crotyl alcohol adsorbed over clean Pd(111) [143] prove that oxidative dehydrogenation to crotonaldehyde occurs at temperatures as low as -60 °C (Fig. 11), with alcohol dehydration to butane only a minor pathway. These ultra-high vac­uum measurements also revealed that reactively formed crotonaldehyde undergoes a competing decarbonylation reaction over metallic palladium above 0°C yielding strongly bound CO and propylidene which may act as site-blockers poisoning subsequent catalytic selox cycles, coincident with evolution of propene into the gas phase. Unexpectedly, pre-adsorbed atomic oxygen switched-off undesired decarbonylation chemistry, pro­moting facile crotonaldehyde desorption.

Structural information on lignins could be obtained by wet-chemical and spectroscopic methods using the approach for analysis of wood given in Figure 17. Here, 5 mL 20% AcBr in pure acetic acid is added to ca. 1-10 mg of wood followed by the addition of 0.1 mL perchloric acid (70%) and keeping the mixture for 3 hours at 50 °C, with subsequent neutralization with NaOH, dilution and UV-Vis analysis at 280 nm.

Direct analysis of lignin in wood can be performed by selective / spe­cific degradation followed by GC analysis. Among degradation methods acidolysis, thioacidolysis, permanganate oxidation and pyrolysis can be mentioned. Acidolysis (Figure 18) cleaves predominantly p-O-4-ether bonds by acid hydrolysis and gives many degradation products with a rather low yield of ca. 60%.

Thioacidolysis (Figure 19) gives selective cleavage of p-O-4-ether bonds and results in less complex mixtures than acidolysis and also gives higher yields (> 80%) being able to quantify units with p-O-4-ether bonds and a free hydroxyl. This reaction is performed in dioxane-ethanethiol with boron trifluoride etherate. The degradation products are silylated prior to analysis by GC.

Oxidation by permanganate (KMnO4 — NaIO4 at 82 °C for 6 hours) in acidic solution of ethylated (at pH = 11 and 25 °C with diazoethane for 25 hours) samples with further oxidation by alkaline peroxide for 10 min at 50 °C and final methylation results in samples which can be analyzed by GC (Figure 7.20). This method gives information only on aromatic units with a free hydroxyl, comprising about 10% of lignin in wood and ca. 70% of lignin after kraft cooking.

Another method for the analysis of lignin is pyrolysis combined with GC-MS (Py GC-MS) allowing even a simultaneous determination of lig­nin and carbohydrates. Py GC-MS can be combined with advanced che — mometric methods such as principal component analysis to enable a more complete identification of various lignin fragments. In summary, it can be stated that because of the heterogeneity of lignin there is no universal degradation method giving all desired information on the lignin structure, however, by combination of several methods the structure of lignin can be described fairly well.

FIGURE 22: Analytical procedures for wood extractives

Nitrogen sorption measurements were performed to examine the poros­ity of the different materials obtained by the reaction pathway shown in Figure 1 (EP, BEP, EP-(CH2)3-SH and EP-(CH2)3-SO3H). The nitrogen adsorption and desorption isotherms are shown in Figure 2. The type IV isotherms with the condensation step at relative pressures between 0.55 and 0.75 and the H1 hysteresis of the solids clearly indicate that the ma­terials are mesoporous and possess cylindrical pores with a narrow pore size distribution.

A summary of the properties of these materials is shown in Table 1. The materials exhibit high specific surface areas (SBET) ranging from 850

to 523 m2 g-1 and large total pore volumes around 0.84 mL g-1. The SBET decreases when the material is functionalized due to the decoration of the pore walls with the bromine and later on with the propylthiol functionality but also due to the overall weight gain of the functionalized materials. The pore diameter of all the materials lies in the range of 6 to 5 nm. Only a mi­nor shift to smaller pore diameters and a slight broadening of the pore size distribution is observed (Figure 2 and Table 1). The structural characteris­tics of the commercially available resin Amberlyst-15 are also presented in Table 1 for comparison. This ethenylbenzenesulfonic acid polymer is a strong acid ion exchange resin with unordered macropores. The material is also prone to swelling.

The XRD patterns of the materials in Figure 3 reveal three well-resolved signals originating from the low angle (100) and second-order (110) and (200) reflections. This evidently indicates that the materials possess a 2D — hexagonal ordered structure and thus retain their P6mm space group order­ing throughout the syntheses. Only a slight broadening can be observed at the patterns of sample BEP, EP-(CH2)3-SH and EP-(CH2)3-SO3H.

It is quite remarkable that all the materials discussed in this study, show outstanding structural stability. The materials retain porosity and ordering after three consecutive reactions as can be seen from the nitrogen sorption and XRD data. These results also confirm the reported stability of Periodic Mesoporous Organosilicas [20,23].

Table 2 presents an overview of the chemical characterization of the solids after the different synthetic procedures. The bromination of the eth- ene bridge is a very straightforward reaction and approximately 25%-30% of the double bonds are brominated as the remaining fraction of double bonds are buried inside the walls and are unavailable for further reaction [45]. The subsequent substitution with the Grignard reagent results in thiol functionalities. The amount of thiol functionalities is determined using a silver titration [44,46]. After oxidation of the thiol groups using sulfuric acid and a thorough washing step, a total amount of 0.60 mmol H+ per gram of material has been observed. This also includes the intrinsic acid­ity of the PMO material originating from the surface silanols (—0.15 mmol g-1), as we described earlier [49]. It is clear that the conversion of the thiol containing PMO into the sulfonic acid containing-material has occurred via the oxidation process. This is also confirmed by Raman spectroscopy by the appearance of two signals in the region between 1160 and 1190 cm-1 (See figure S1 in supplementary information). Also, the thiol titration after oxidation showed a zero concentration of remaining thiol groups. Amberlyst-15 exhibits a high acidity of 4.7 mmol H+ g-1.

Catalytic hydroprocessing of liquid biomass is a technology currently un­der developed and there is a lot of room for optimization. For example there are not many commercial catalysts specifically designed and devel­oped for such applications, while conventional commercial catalysts, em­ployed for catalytic hydroprocessing of refinery streams, are used instead. Common hydrotreating catalysts employed contain active metals on alu­mina substrate with increased surface area. The most known commercial catalysts employ Cobalt and Molybdenum (CoMo) or Nikel and Molybde­num (NiMo) in alumina substrate (Al2O3) as shown in Figure 4.

Hydrotreating catalysts are dual action catalytic material, triggering both hydrogenation and cracking/isomerization reactions. On one hand hydrogenation takes place on the active metals (Mo, Ni, Co, Pd, Pt) which catalyze the feedstock molecules rendering them more active when subject to cracking and heteroatom removal, while limiting coke formation on the catalyst. Furthermore hydrogenation supports cracking by forming an active olefinic intermediate molecule via dehydrogenation. On the other hand both cracking and isomerization reactions take place in acidic envi­ronment such as amorphous oxides (SiO2 — Al2O3) or crystalline zeolites (mainly z-zeolites) or mixtures of zeolites with amorphous oxides.

During the first contact of the feedstock molecules with the catalyst, a temperature increase is likely to develop due to the exothermic reactions that occur. However, during the continuous utilization of the catalyst and coke deposition, the catalyst activity eventually reduces from 1/3 to 1/2 of its initial one. The catalyst deactivation rate mainly depends on tempera­ture and hydrogen partial pressure. Increased temperatures accelerate cata­lyst deactivation while high hydrogen partial pressure tends to mitigate catalyst deactivation rate. Most of the catalyst activity can be recovered by catalyst regeneration.

The selection of a suitable hydroprocessing catalyst is a critical step defining the hydroprocessing product yield and quality as well as the operating cycle time of the process in petroleum industry [5]. However the hydrotreating catalyst selection for biomass applications is particu­larly crucial and challenging for two reasons: a) catalyst activity varies significantly, as commercial catalysts are designed for different feed­stocks, i. e. feedstocks with high sulfur concentration, heavy feedstocks (containing large molecules), feedstocks with high oxygen concentration etc, and b) there are currently no commercial hydroprocessing catalysts available for lipid feedstocks and other intermediate products of biomass conversion processes (e. g. pyrolysis biooil), and thus commercial hy­drotreating catalysts need to be explored and evaluated as different cat­alyst have different yields (Figure 5) and different degradation rate [8].

Nevertheless, significant efforts have been directed towards developing special hydrotreating catalysts for converting/upgrading liquid biomass to biofuels [9-12].

The catalytic transformation of sugars to liquid hydrocarbon fuels is a complicated process that ideally should combine deep oxygen removal and adjustment of the molecular weight using a small number of reactors and with minimal utilization of fossil fuel-based external hydrogen. This goal can be achieved by (i) using multifunctional catalysts able to carry out different reactions in the same reactor [92] and (ii) utilizing a fraction of the sugar feedstock as a source of in situ hydrogen through aqueous — phase reforming reactions. [73]

Both approaches are combined by Kunkes et al. in a recent process that transforms aqueous solutions of sugars and sugaralcohols into liquid hydrocarbon fuels in a two-step cascade process [116] (Fig. 8). Firstly, aqueous sugars and polyols (typically glucose and sorbitol) are converted into a mixture of monofunctional compounds (e. g., acids, alcohols, ke­tones and heterocycles) in the C4-C6 range, which are stored in an organic phase that spontaneously separates from water. This step is carried out at temperatures near 500 K over a Pt-Re/C catalyst, which achieves deep de­oxygenation (up to 80% of the oxygen in the initial feedstock is removed) by means of C-O hydrogenolysis reactions. Importantly, the hydrogen required to accomplish the C-O cleavage step is internally supplied by aqueous-phase reforming (involving C-C cleavage and WGS reactions) of a fraction of the feed (Fig. 8). The Pt-Re/C material allows production of hydrogen and removal of oxygen in a single reactor. Unlike bio-oils produced by pyrolysis (Section 4.2), the organic stream of monofunctional compounds produced by sugar processing over Pt-Re/C is completely free of water and has a well-defined composition that is controlled by the feed­stock type (e. g., sugars or polyols) and the reaction conditions. [117]

The retention of functionality in the organic intermediates is key to control reactivity and to allow subsequent C-C coupling upgrading strate­gies. This approach has been demonstrated to be conceptually adequate to process sugars into fuels, [3] and important biomass derivatives such as lactic acid (3-hydroxypropanoic acid) [89,118] and levulinic acid [119] have been upgraded following this strategy. Each group of compounds (e. g., alcohols, ketones, acids) in the monofunctional stream can be up­graded to targeted hydrocarbons through different C-C coupling reactions (e. g., oligomerization, aldol-condensation and ketonization). For example, the organic stream enriched in alcohols by hydrogenation of ketones can be processed over an acidic H-ZSM5 zeolite at atmospheric pressure to yield 40% of C6+ aromatic gasoline components. Ketones can be upgraded to larger hydrocarbon compounds (C8-C12) with low extents of branching by means of aldol-condensation reactions over bifunctional Cu/ Mg10Al7Ox catalysts. However, carboxylic acids present in the organic stream caused deactivation of the basic sites responsible for aldol-condensation, and ap­proaches based on upstream removal of acids by ketonization (similar to those proposed herein for bio-oils upgrading, Section 4.2) and subsequent aldolcondensation have been successfully developed. [120-122] Ketoni — zation acquires special relevance when the organic stream is rich in car­boxylic acids, as is the case when the feed is glucose.

Among the alkaline and/or alkaline-earth oxides, various lithium, sodium, potassium, calcium and magnesium ceramics have been proposed for CO2 capture through adsorption and chemisorption processes [1-20]. These materials can be classified into two large groups: dense and porous ceram­ics. Dense ceramics mainly trap CO2 chemically: the CO2 is chemisorbed. Among these ceramics, CaO is the most studied one. It presents very in­teresting sorption capacities at high temperatures (T > 600 °C). In addition to this material, alkaline ceramic oxides have been considered as possible captors, mostly lithium and sodium based ceramics (Li5AlO4 and Na2ZrO3, for example). In these cases, one of the most interesting properties is re­lated to the wide temperature range in which some of these ceramics trap CO2 (between 150 and 800 °C), as well as their high CO2 capture capacity.

In these ceramics, the CO2 capture occurs chemically, through a che­misorption process. At a micrometric scale, a general reaction mechanism has been proposed, where the following steps have been established: Ini­tially, CO2 reacts at the surface of the particles, producing the respective alkaline or alkaline-earth carbonate and in some cases different secondary phases. Some examples are:

Li5AlO4 + 2CO2 ^ 2Li2CO3 + LiAlO2 (1)

Na2ZrO3 + CO2 ^ Na2CO3 + ZrO2 (2)

CaO + CO2 ^ CaCO3 (3)

The above reactions show that surface products can be composed of carbonates, but as well they can contain metal oxides or other alkaline/al­kaline-earth ceramics. The presence of these secondary phases can modify (improve or reduce) the diffusion processes described below [1].

Once the external carbonate shell is formed, different diffusion mecha­nisms have to be activated in order to continue the CO2 chemisorption, through the particle bulk. Some of the diffusion processes correspond to the CO2 diffusion through the mesoporous external carbonate shell, and some others such as the intercrystalline and grain boundary diffusion pro­cesses [1, 18, 21].

Figure 1 shows the theoretical CO2 chemisorption capacities (mmol of CO2 per gram of ceramic) for the most studied alkaline and alkaline-earth ceramics. As it can be seen, metal oxides (Li2O, MgO and CaO) are among the materials with the best CO2 capture capacities. Nevertheless, Li2O and MgO have not been really considered as possible options due to reactivity and kinetics factors, respectively. On the contrary, CaO is one of the most promising alkaline-earth based materials, with possible real industrial ap­plications. Other interesting materials are ceramics with lithium or sodium phases, which present better thermal stabilities and volume variations than CaO. In addition, the sodium phases may present another advantage if the CO2 capture is produced in the presence of steam. Under these condi­tions the sodium phases may produce sodium bicarbonate (NaHCO3) as the carbonated phase, which is twice the amount of CO2 could be trapped in comparison to the Na2CO3 product.

Other ceramics containing alkaline-earth metals are the layered double hydroxides (LDH) or hydrotalcite-like compounds (HTLc). LDHs, also called anionic clays due to their layered structure and structural resem­blance to a kind of naturally-occurring clay mineral. These materials are a family of anionic clays that have received much attention in the past decades because of their numerous applications in many different fields, such as antacids, PVC additives, flame retardants and more recently for drug delivery systems and as solid sorbents of gaseous pollutants [22-24]. The LDH structure is based on positively charged brucite like [Mg(OH)2] layers that consist of divalent cations surrounded octahedrally by hydrox­ide ions. These octahedral units form infinite layers by edge sharing [25]. Due to the fact that certain fraction of the divalent cations can be substi­tuted by trivalent cations at the centers of octahedral sites, an excess of positive charge is promoted. The excess of positive charge in the main lay­ers of LDHs is compensated by the intercalation of anions in the hydrated interlayer space, to form the three-dimensional structure. These materials have relatively weak bonds between the interlayer and the sheet, so they exhibit excellent ability to capture organic or inorganic anions. The ma­terials are easy to synthesize by several methods such as co-precipitation, rehydration-reconstruction, ion exchange, hydrothermal, urea hydrolysis and sol gel, although not always as a pure phase [26].

The LDH materials are represented by the general formula: [M1-xn MxnI(OH)2]x+[Am-]x/mmH2O where Mn and MnI are divalent (Mg2+, Ni2+, Zn2+, Cu2+, etc.) and trivalent cations (Al3+, Fe3+, Cr3+, etc.), respec­tively, and Am — is a charge compensating anion such as CO32-, SO42-, NO3-, Cl-, OH-, where x is equal to the molar ratio of [MIn/(Mn + MIn)]. Its value is commonly between 0.2 and 0.33, i. e., the Mn/MnI molar ratio is in the range of 4 — 2 [25], but this is not a limitation ratio and it depends on the Mn and MnI composition [27-29].

Among various CO2 mesoporous adsorbents, LDH-base materials have been identified as suitable materials for CO2 sorption at moderate tempera­tures (T < 400 °C) [30-46] due to their properties such as large surface area, high anion exchange capacity (2-3 meq/g) and good thermal stability [37-40]. The LDH materials themselves do not possess any basic sites. For that reason, it is preferred to use their derived mixed oxides, formed by the thermal decomposition of LDH, which do exhibit interesting basic properties. Thermal decomposition of the material occurs in three stages, first at temperatures lower than 200 °C, at which the dehydration of su­perficial and interlayer water molecules takes place on the material. Then the second decomposition stage takes place in the range of 300-400 °C, at which the structure collapses due to a partial dehydroxylation process, typically associated with both the decomposition of Al-OH and the Mg — OH hydroxides. During dehydroxylation, changes occur in the structure. A portion of the trivalent cations of the brucite like layers migrates to the in­terlaminar region, allowing the preservation of the laminar characteristics of the material [41]. Finally, the total decomposition of the material occurs at temperatures higher than 400 °C, when the decarbonation process is completed [42].

Once the temperature reaches about 400 °C, LDH forms a three-dimen­sional network of compact oxygen with a disordered distribution of cat­ions in the interstices, where the cations M+3 are tetrahedrally coordinated (interlayer region) and M+2 are octahedrally coordinated. The compres­sive-expansion stresses associated with the formation of the amorphous three-dimensional networks and their connection to the octahedral layer increases the surface area and pore volume, which can help improve the storage capacity properties, for example for gas sorption related applica­tions, besides decreasing the ability of the Mg+2 cation to favor physisorp — tion instead of chemisorption [30, 42]. For instance, the thermal evolution of the Mg/Al-CO3 LDH structure is considered to be crucial in determin­ing the CO2 adsorption capacity, so there are several studies about this issue [42-44].

Reddy et al. [43] studied the effect of the calcination temperature on the adsorptive capacity of the Mg/Al-CO3 LDH. They found out that the best properties were obtained at calcination temperature of 400 °C, which they attributed to the obtaining of a combination of surface area and the availability of the active basic sites. Actually, at this temperature the mate­rial is still amorphous, which allows having a relatively high surface area. Therefore, there is a high number of exposed basic sites, allowing the re­versible CO2 adsorption according to the following reaction:

Mg-O + CO2 ^ Mg-O… CO2 (ad) (4)

However, if the LDH is calcined under 500 °C, the material is able to transform back to the original LDH structure when it is exposed to a carbonate solution or another anionic containing solution. Finally, if the sample is heated to temperatures above 500 °C, the structural changes become irreversible because of the spinel phase formation [37].

As mentioned, the mixed oxides derived from the LDH calcination possess some interesting characteristics such as high specific surface area, excess of positive charge that needs to be compensated, basic sites and thermal stability at elevated temperatures (200 — 400 °C). Besides these aspects, it is important to consider the advantage of acid-base interactions on the CO2 sorption applications, where acidic CO2 molecules interact with the basic sites on the derived oxide. These characteristics make the

LDH-materials acceptable CO2 captors [43, 45]. However, the CO2 ad­sorption capacity of this material is low compared with other ceramic sor­bents; reaching mean values smaller than 0.1 mmol/g [46]. Nevertheless, many studies suggest that the adsorption capacity of LDH materials can be improved by modifying a factor set such as: composition, improvement of the material’s basicity and contaminant gas stream composition [30-32, 36, 41-45, 47-59].

As previously mentioned, Reddy et al. [43] studied the influence of the calcination temperature of LDHs on their CO2 capture properties. The Mg3/Al1-CO3 material was calcined at different temperatures ranging from 200 to 600 °C. The results showed that when the calcination temperatures are under 400 °C, LDH is considered to be dehydrated and materials still keep the layered structure intact, wherein the CO32- ions are occupying the basic sites. The obtained samples calcined at 400 °C have the maximum BET surface area of 167 m2/g compared with samples calcined at lower temperatures. Moreover, during the calcination of the LDH at higher tem­peratures (T > 500 °C), most of the CO32- decompose to release some basic sites for CO2 adsorption. However, the final amount of basic sites decreas­es with the subsequent crystallization of the MgO and spinel (MgAl2O4). Hence, LDH materials obtained at 400 °C have the highest surface area and the maximum quantities of active basic sites exposed. Because of these characteristics, they achieved a total sorption capacity of 0.5 mmol/g [43]. The same researchers observed that 88% of the captured gas can be desorbed and during the material regeneration 98% of the original weight is gained. This is another important property of LDH materials in high temperature CO2 separation applications as described later..

As mentioned, the thermal evolution of the layered structure has a great influence on the CO2 capture. The loss of superficial interlayer water occurs at 200 °C. Then at temperatures between 300 and 400 °C the layer decomposition begins, resulting in an amorphous 3D network with the highest surface area [30], so the adsorption temperature improves the CO2 capture in the order of 400 > 300 > 20 >200 °C [41-42, 47, 52]

Several researchers have investigated a set of different factors to im­prove the CO2 sorption capacity. Yong et al. [47, 48] studied the factors which influence the CO2 capture in LDH materials, such as aluminum content, water content and heat treatment temperature. Regarding the M/

Al-CO3 LDHs (M = Mg, Ni, Co, Cu or Zn), the best CO2 sorption capacity was obtained for the Mg/Al materials degassed at 400 °C and with adsorp­tion conditions of 25 °C. In general, the sorption capacity follows the trend Ni > Mg > Co > Cu = Zn. However, when the degassed temperature is in­creased, the trend is modified to Mg (400 °C) > Co (300°C) > Ni (350°C) > Cu (300°C) >Zn (200°C). These results show that Mg/Al-CO3 is the best composition at the degassing temperature of 400 °C [47]. At this tempera­ture, the material consists of an amorphous phase with optimal properties for use as CO2 captor [42]. Also, the influence of Al+3 has been studied as a trivalent cation at 25 and 300 °C adsorption temperatures, by Yong [41] and Yamamoto [49] respectively. Both samples were degassed at 300 °C and the results showed that the CO2 capture is influenced by the adsorption temperature. At a temperature of 25 °C, the maximum adsorption was 0.41 mmol/g with an Mg/Al ratio equal to 1.5 (x = 0.375) [41] and at 300 °C the amount of CO2 adsorbed was 1.5 mmol/g for a cation ratio of 1.66 (x = 0.4) [49]. The differences between the two capacities can be attributed to the Al content differences. The Al incorporation in the structure has two functions: 1) to increase the charge density on the brucite-like sheet; and 2) to reduce the interlaminar distance and the number of sites with high resistance to CO2 adsorption [48].

On the other hand, Qian et al. [50] studied the effect of the charge compensation anions (A — = CO3-2, NO3-1, Cl-, SO4-2 and HCO3-1) on the structural properties and CO2 adsorption capacity of Mg/Al-A — (molar ra­tio equal to 3). Despite all of the prepared LDH materials showed the typical XRD patterns of LDH materials, slight structural and microstruc­tural differences were observed. In fact, the interlayer distance changed by varying the interlayer anions due to their difference in sizes and carried charges. These differences affect the morphology and the BET surface area of both fresh and heat-treated LDH materials. Additionally, thermal treatments were performed in order to optimize the adsorption capacity of these materials. The optimal temperature treatment was established for each Mg/Al-A — based on the surface area of each calcined LDH. Then the CO2 adsorption capacities of calcined LDH were tested at 200 °C. Mg3/ Al1-CO3 showed the highest CO2 adsorption capacity (0.53 mmol/g). This value was much higher than those obtained for calcined Mg3/Al1-NO3 > Mg3/Al1-HCO3, Mg3Al1-Cl, and Mg3/Al1-SO4 ( ~ 0.1 mmol/g). The results indicated that BET surface area of calcined LDHs seems be the main pa­rameter that determines the CO2 adsorption capacity because the Mg-O active basic site [43, 45].

It has been demonstrated that the quasi-amorphous phase obtained by the thermal treatment of LDH at the lowest possible temperature has the highest CO2 capture capacity. This finding is in line with the fact that high calcination temperature can decrease the number of active Mg-O sites due to the formation of crystal MgO [51].

Yong [41] and Yamamoto [49] investigated the influence of the sev­eral types of anions. The results suggested that the amounts CO2 capture decrease as a function of the anion size, which promotes a larger inter­layer spacing and the higher charge: Fe(CN)64- (1.5 mmol/g) > CO32- (0.5 mmol/g) > NO3- (0.4 mmol/g) > OH — (0.4-0.25 mmol/g). The reason is that Fe(CN)64- and CO32-, because they have more void space in the interlayer due size, and are able to accommodate higher CO2 quantities. Calcined layered double hydroxide derivatives have shown great potential for high temperature CO2 separation from flue gases. However, the presence of SOx and H2O from flue gases can strongly affect CO2 adsorption capac­ity and regeneration of hydrotalcite-like compounds. Flue gases emitted from power stations contain considerable amounts of water in the form of steam. The percentage of water found in the flue gas emitted from dif­ferent sources varies between 7 and 22%, with the emissions from brown coal combustion having the highest water content [45]. For many other gas adsorption sorbents, steam generally has a negative effect on the adsorp­tion performance because of competition for basic sites between CO2 and H2O. However, the presence of water or steam seems to be favorable for the adsorption capacity onto LDH [31,43,53,54]. This fact is the result of the increasing potential energy that is able to further activate basic sites, possibly by maintaining the hydroxyl concentration of the surface mate­rial and/or preventing site poisoning through carbonate or coke deposition [31]. An example of the above was reported by Yong et al. [47], who found that water or steam can increase the adsorption capacity of CO2 by about 25%, from 0.4 mmol/g to 0.5 mmol/g.

Ding et al. [31] studied CO2 adsorption at higher temperatures (480 °C) under conditions for steam reforming of methane. They found an ad­sorption capacity of 0.58 mmol/g, which was independent of water vapor content in the feed. In turn, Reddy et al. [45] investigated calcined LDHs’ sorption performance influenced by CO2 wet-gas streams. LDH samples were calcined at 400 °C [43] before measuring CO2 sorption at 200 °C. The gas streams used were CO2, CO2 + H2O, flue gas (14% CO2, 4% O2 and 82 % N2) +12% H2O.

For a pure CO2 dry sorption, the maximum weight gain was 2.72% (~0.61 mmol/g) after 60 min, whereas the wet adsorption increased the weight of the calcined LDHs to 4.81%, showing an additional 2.09%, where He and He + H2O were used to remove the H2O water capture. The results showed that the helium has virtually no significant sorption affin­ity for the material, whereas the water-sorption profile of it clearly indi­cates a water weight gain of 1.67%, i. e., the gain was 0.1mmol/g due to steam presence, showing that water has a positive effect, shifting the CO2 sorption by 0.42% as compared to dry CO2 sorption. Also, these results revealed that in all cases about 70% sorption occurs during the first 5 min and reaches equilibrium after around 30 min.

To determine the influence of CO2, Reddy et al. [43] tested a sample in both, wet and dry CO2 stream conditions. The experiments showed that the same quantity of CO2 can be trapped for the solid sorbent after two hours. The presence of water in the stream only affects the kinetics of the process. This result is in agreement with that reported by Ding et al. [31]. On the other hand, the results of the material tested suggest that the fact the CO2 capture from flue gas was higher than in a pure stream of CO2 might have been because the polluted gas was diluted in the stream. The presence of the water does not enhance de CO2 capture; the maximum CO2 adsorbed was 0.9 mmol/g. The differences between Reddy et al. results and the pre­viously mentioned studies can be caused by the use of uncalcined LDHs, which already contain an — OH network.

To apply these materials in industrial processes, it is important to know the times during which each sorbent material can be used. Tests of the cy- clability in LDH materials disclose that as function of the temperature the CO2 capture time can vary. This can be attributed to CO2 chemisorbed dur­ing each cycle [54] and/or to the formation of spinel-based aluminas, such as y-Al2O3 (at temperatures higher than 400 °C). Hibino et al. [52] found that the carbonate content, acting as charge-compensating anion, continu­ously decreases in subsequent calcination-rehydration cycles. Reddy et al.

tested LDH materials during six CO2 adsorption (200 °C)-desorption (300 °C) cycles. The average amount gained was 0.58 mmol/g, whereas 75% of this value is desorbed, reaching desorption equilibrium after the third cycle. This can be attributed to the stabilization of the material phase and basic sites during the temperature swing.

Hufton et al. [54] studied a LDH material during several cycles in dry and wet CO2 flows. As previously discussed, the presence of steam in the flow gas improves the CO2 adsorption. However, after 10 adsorption cy­cles, the capture decreased 45%. The same behavior was observed in the dry gas flow. However, the final capture was similar to the wet gas stream, in agreement with Reddy et al. [43].

Recent studies have demonstrated that K-impregnated LDH or K-im- pregnated mixed oxides have a better CO2 capture capacity due to the addition of K alkaline-earth element that improves the chemical affinity between the acidic CO2 and alkaline surface of the sorbent material [32, 36, 55-56]. Additionally, it has been proposed that K-impregnation re­duces the CO2 diffusion resistance in the material. [57]. Hufton et al. [58] showed that the K-impregnation increases the CO2 capture, but there is an optimal quantity of K to reach the maximum capture. Qiang et al. [50] tested an Mg3/Al1-CO3 (pH = 10) impregnated with 20 wt.% K2CO3. The CO2 adsorption capacity was increased between 0.81 and 0.85 mmol/g in the temperature range of 300-350 °C. This adsorption capacity is adequate for application in water gas shift reactions (WGS).

Lee et al. [59] tested the behavior of three commercial LDHs impreg­nated with K (K2CO3/LDH ratio between 0 and 1). Three Mg/Al-CO3 LDH with different contents of magnesium were used. Results indicated that the sorption capacity of the LDH is improved by about 10 times with the optimal K2CO3 additions. Additionally, it was observed that impregnation is not the only factor that influences the adsorption but the composition too. The best value was obtained when the content of divalent cation was reduced and therefore, the material had a composition with the maximum trivalent cation content. The CO2 adsorption capacity improved from 0.1mmol/g to 0.95mmol/g with K2CO3/LDH weight ratio equal to 0.35 at 400 °C. After determining the optimal alkaline source/LDH ratio, a set of samples was evaluated as a function of the temperature and the results showed a maximum of 1.35 mmol/g, at 50 °C. In the impregnated materi­als, CO2 chemisorption can occur and the sorbed CO2 can be further stored as metal carbonate forms.

Other alkaline elements can be used to improve the sorption capac­ity of materials. Martunus et al. [46] studied the impregnation of LDH with Na and K. The LDH samples were thermally treated at 450 °C for 5 min then calcined samples were re-crystallized in K2CO3 and Na2CO3 (1 M) solutions. The re-crystallized materials were tested as CO2 captors and the capture was maximum with LDH-Na (0.688 mmol/g) > LDH — K (0.575 mmol/g) at 350 °C after five cycles. Finally, the re-crystallized material with the highest capture was calcined at 650 °C for 4 h and re­crystallized with a solution containing the appropriate quantities of K and Na to achieve alkaline metal loading up to 20%. When the sample was Impregnated with additional K and Na at 18.4% and 1.6%, respectively, the adsorption capacity rose from 0.688 to 1.21 mmol/g. This capacity increase was achievable despite the relatively low BET surface area, equal to 124 m2/g.

Other alkaline elements such as cesium have been studied as reinforce­ment. Oliveira et al. [55] tested commercial Mg1/Al1-CO3 and Mg6Al1-CO3 impregnated with K and Cs carbonates. The materials were evaluated in the presence of steam (26% v/v water content) gas at different temperatures (306, 403 and 510 °C) at 0.4 bar of CO2 partial pressure (total pressure 2 bar). The LDH with the highest sorption capacity was Mg1/Al1-CO3-K with 0.76 mmol/g at 403 °C. Among the Cs impregnated samples, the Mg6Al1-CO3-Cs presented the highest capacity with 0.41 mmol/g, while the commercial LDH samples presented CO2 sorption capacities around 0.1 mmol/g.

The results suggest the existence of a sorption mechanism combining physical adsorption and chemical reaction. First the maximum physical adsorption is reached, then the chemisorption begins, but there is an op­timal temperature. If the temperature is too low, the chemisorption does not happen, but with higher temperatures the loss of porosity impedes the contact of CO2 molecules with active basic sites promoted by the alkaline element addition.

These results suggest there is an optimum amount of K2CO3 to impreg­nate the LDH that achieves a balance between the increase in the basicity of the sorbent material and its reduction of surface area, associated with

CO2 capture capacity. The influence of potassium is currently not clear and the relevant research is still ongoing. Finally, CO2 adsorption capacity on the synthesized 20 wt.% K2CO3/Mg3/Al1-CO3 (pH = 10) probably could be further increased in the presence of steam.

The effectiveness of catalytic hydroprocessing was also explored for co­processing of lipid feedstocks with petroleum fractions as catalytic hydro­processing units are available in almost all refineries. The first co-process­ing study involved experiments of catalytic hydrotreating of sunflower oil mixtures with heavy petroleum fractions aiming to produce high quality diesel [55]. The experiments were conducted in a continuous fixed-bed reactor over a wide range of temperatures 300-450°C employing a typi­cal NiMo/Al2O3 hydrotreating catalyst. The study was focused on the hy­drogenation of double C-C bonds and the subsequent paraffin formation via the three different reactions routes: decarbonylation, decarboxylation

and deoxygenation. Furthermore the large carbon-chain paraffins can also undergo isomerization and cracking leading to the formation of smaller paraffins. This study concluded that the selectivity of products on decar­boxylation and decarbonylation is increasing as the temperature and veg­etable oil content in the feedstock increase [55].

In a similar study catalytic hydrocracking over sunflower oil and heavy vacuum gas oil mixtures was investigated [56]. The experiments were conducted in a continuous-flow hydroprocessing pilot-plant over a range of temperatures (350-390°) and pressures (70-140bar). Three different hy­drocracking catalysts were compared under the same conditions and four different feedstocks were employed, incorporating for 10% and 30%v/v of lipid bio-based feedstock and considering non-pretreated and pretreated sunflower oil as a bio-based feedstock. The results indicated that a prior mild hydrogenation step of sunflower oil is necessary before hydrocrack­ing. Furthermore, conversion was increased with increasing sunflower oil ratio in the feedstock and increasing temperature, while the later decreased diesel selectivity.

The effect of the process parameters and the vegetable oil content of the feedstocks on the yield, physical properties, chemical properties and application properties during co-hydrotreating of sunflower oil and gas-oil mixtures utilizing a typical NiMo/Al2O3 hydrotreating catalyst was also studied [57]. The experimental results of this study indicated that catalytic co-hydrogenation of gas oil containing sunflower oil in different percent­ages allowed both vegetable oil conversion reactions (saturation, deoxy­genation) and the gas oil quality improvement reactions (hetero atom re­moval, aromatic reduction). The optimal operating conditions (360-380°C, P=80 bar, LHSV=1.0h-1, H2/oil=600 Nm3/m3 and 15% sunflower oil con­tent of feed) resulted in a final diesel product with favorable properties (e. g. less than 10 wppm sulfur, ~20% aromatics) but poor cold flow prop­erties (CFPP=3°C). The study also indicated that for sunflower content in the feedstock higher than 15% reduced the desulfurization efficiency. Furthermore, the authors also concluded that the presence of sunflower oil in the feedstock has augmented the normal and iso-paraffins content of the final product and as a result has increased the cetane number but degraded the cold flow properties, indicating that an isomerization step is required as an additional step.

The issue of catalyst development suitable for co-hydrotreating and co-hydrocracking of gas-oil and vegetable oil mixtures was recently ad­dressed [10], as there are no commercial hydroprocessing catalysts avail­able for lipid feedstocks. New sulfided Ni-W/SiO2-Al2O3 and sulfided Ni-Mo/Al2O3 catalysts were tested for hydrocracking and hydrotreating of gas — oil and vegetable oil mixtures respectively. The results indicated that the hydrocracking catalyst was more selective for the kerosene hy­drocarbons (140-250°C), while the less acidic hydrotreating catalyst was more selective for the diesel hydrocarbons (250-380°C). The study ad­ditionally showed that the deoxygenation reactions are more favored over the hydrotreating catalyst, while the decarboxylation and decarbonylation reactions are favored over the hydrocracking catalyst.

6.2.1 SETUP AND DESCRIPTION OF THE TESTED EQUIPMENT ALONG WITH MEASURING TECHNIQUES

A test bench has been developed (as shown in the Figure 1) in order to ex­amine the emissions from a prototype downdraft stove. The test bench is designed in such a way that it can facilitate the analysis of dust composition.

For the determination of flue gas and combustion chamber tempera­ture profiles, the thermocouples of Type K (manufactured by the company “Newport Electronics GmbH”) have been used. For this purpose, a set of various thermocouples has been inserted into the grate, in the middle of upper and lower combustion chambers as well as in the walls of the lower combustion chamber. Moreover, the pressure conditions were recorded with the help of pressure sensors, inserted into the combustion chambers (upper and lower) as well as into the exhaust pipe. The measurement of static and dynamic pressures in the flue gas has been done with the aid of Prandtl tube produced by the company “Testo AG”. The continuous transmission and data recording of Prandtl tube and pressure nozzles in the combustion chamber is carried out through data logging module provid­ed by the company “Ahlborn”. The data of the thermocouples have been recorded through a data logger of the company “National Instruments” along with the help of the software “Labview”.

The emissions coming out of the stove are measured by means of a gas analyzer which consists of a Fourier Transform Infrared Spectrometer (FTIR, Manufacturer: Calcmet), a Flame Ionization Detector (FID, Manu­facturer: Mess- & Analysentechnik GmbH, Typ: thermo-FID ES) and a paramagnetic oxygen analyzer (Manufacturer: M&C, Type: PMA 100). The infrared spectrum of FTIR can measure simultaneously organic as well as inorganic components. At the moment, about 44 different compo­nents can be recorded through FTIR.

The Volatile Organic Compounds (VOC) can be recorded by means of both FID and FTIR measuring devices. In case of VOC, the concentrations ranging under 50 mg/m3 (at standard conditions i. e. =0°C, 1 atm) can be considered from the FID measuring device. On the other hand, the values above 50 mg/m3 (at standard conditions i. e. =0°C, 1 atm) can be assumed from the FTIR measuring device. Following parameters can be measured simultaneously:

1. Oxygen O2 (paramagnetic analyzer)

2. Carbon dioxide (FTIR)

3. Moisture in the flue gas i. e. H2O (FTIR)

4. Carbon monoxide CO (FTIR)

5. Volatile organic compounds (VOC) as organic carbon (Org.-C) (FTIR and FID)

6. Nitrogen oxide as nitrogen dioxide equivalent (NO2equi) (FTIR)

7. Sulphur dioxide SO2 (FTIR)

8. Methane CH4 (FTIR)

9. Organic compounds like, alkanes, alkenes, aromatics, aldehydes as well as ketones (FTIR)

10. Flue gas temperature, gas velocity and draft conditions.

The recording of the above mentioned parameters took place on con­tinuous time basis except for the dust measurement. During the evaluation of the data, the average values of the pollutants were calculated for each dust sampling cycle whereas each cycle lasts for 30 minutes. With the aid of a chimney fan, a constant negative pressure of 12 Pa has been main­tained in the chimney stack in order to achieve a fuel thermal output from 8 to 9 kW.

The gravimetric analysis of total amount of dust was done in accor­dance with VDI guidelines 2066-1, according to which a partial volume flow must be taken in isokinetic manner out of the main flue gas stream. In this process, the accompanied particles can be deposited on the already weighed plane filter. Since the filter housing is located outside the flue gas pipe, this sampling procedure is termed as “outstack process”. The filter system was heated up with a heating jacket in order to prevent the falling down of temperature under saturation temperature of the flue gas. The temperature of the filter was maintained at 70°C so that the semi-volatile hydrocarbons could also be deposited on the filter. After the experiment, the deposited dust amount was determined gravimetrically and then can be specified by taking into consideration the measured partial volume and oxygen concentration. The plane filter was made of micro-glass fibers hav­ing a diameter of 45 mm.