Enhanced extraction of natural substances using microwave energy

ENHANCED EXTRACTION OF NATURAL SUBSTANCES USING MICROWAVE ENERGY Katalin Sólyom Katalin Sólyom » Enhanced extraction of natural substances using microwave energy « ESCUELA DE INGENIERÍAS INDUSTRIALES DEPARTAMENTO DE INGENIERÍA QUÍMICA Y TECNOLOGÍA DEL MEDIO AMBIENTE TESIS DOCTORAL: Enhanced extraction of natural substances using microwave energy Presentada por Katalin Sólyom para optar al grado de Doctor por la Universidad de Valladolid Dirigida por: Prof. Rafael B. Mato Prof. María José Cocero Alonso ESCUELA DE INGENIERÍAS INDUSTRIALES DEPARTAMENTO DE INGENIERÍA QUÍMICA Y TECNOLOGÍA DEL MEDIO AMBIENTE TESIS DOCTORAL: Extracción mejorada de sustancias naturales mediante energía de microondas Presentada por Katalin Sólyom para optar al grado de Doctor por la Universidad de Valladolid Dirigida por: Prof. Rafael B. Mato Chaín Prof. María José Cocero Alonso Memoria para optar al grado de Doctor, con Mención Doctor Internacional, presentada por la Ingeniera Química: Katalin Sólyom Siendo los tutores en la Universidad de Valladolid: Prof. Dr. Rafael B. Mato y Prof.ª Dr.ª María José Cocero Alonso Y en el Karlsruher Institut für Technologie, Lebensmittelverfahrenstechnik (Karlsruhe, Alemania): Prof. Dr.-Ing. Heike P. Schuchmann Valladolid, Septiembre de 2013 UNIVERSIDAD DE VALLADOLID ESCUELA DE INGENIERÍAS INDUSTRIALES Secretaría La presente tesis doctoral queda registrada en el folio número _______ del correspondiente libro de registro número _____________________ Valladolid, a ______ de_______________ de 2013 Fdo. El encargado del registro TABLA DE CONTENIDOS Resumen (Inglés) 19 Ámbito y Objetivos (Inglés) 23 Antecedentes Teóricos 27 Capítulo 1 53 Influencia de la energía absorbida en los pre-tratamientos con microondas en la producción de biogás a partir de lodo segundario de aguas residuales s Capítulo 2 77 Intensificación de la cinética de extracción de lípidos y pigmentos a partir de microalgas aplicando microondas Capítulo 3 101 Propiedades dieléctricas del orujo de uvas: efecto de la temperatura, la humedad y el método de preparación de la muestra Capítulo 4 123 Efecto de la energía absorbida de microondas en las cinéticas de extracción de los compuestos fenólicos a partir del orujo de uva Capítulo 5 145 Extracción mejorada de polifenoles a partir del orujo de uvas después del pre-tratamiento con ultrasonidos Capítulo 6 169 Degradación térmica de polifenoles del orujo de uvas Conclusiones 193 Resumen Español 199 Agradecimientos 217 Sobre la Autora 221 Abstract Abstract 19 Nowadays, in numerous industrial extraction processes, technologies that can save the time, space, energy and solvent are welcome in order to achieve production intensification. This is of particular importance in those cases where complex natural raw materials are involved, and the biological cell wall disruption step is blamed for the slow kinetics and low throughput in the process. Among several mechanical, chemical, physical or physicochemical treatments of natural raw materials, the possibility of using microwaves has emerged. Due to the electromagnetic irradiation of the material, rapid heating up and evaporation of intracellular water can be achieved, and the resulting pressure gradient may lead to cell wall damage. Thus, the intracellular compounds of interest become more accessible after an efficient pre-treatment, and faster kinetics can be achieved in the subsequent conventional process. In this Doctoral Thesis, three different raw materials and processes were studied in order to explore the effects of microwave pre-treatment on extraction process kinetics. Firstly, microorganisms were used, such as wastewater sludge, in order to intensify biogas production by liberating any intracellular material, which thus became accessible for anaerobic bacteria in the digestion process. Due to the large cell variety present in wastewater sludge, the specific study of the cell structure becomes difficult to elucidate, although the overall process kinetics can be upgraded, just as in a conventional thermal pre-treatment. Secondly, unicellular microalgae were chosen as better candidates for this purpose. Also, lipid extraction for biodiesel production, and high added value algal pigment extraction are at the centre of attention in the field of continuously developing sustainable process engineering. The study on different microalgae and extraction techniques presented here showed that microwave pre-treatments have the potential to improve extraction kinetics, and also higher extraction yields may be obtained. Finally, the abundant viticulture in the region of Castile and Leon suggested the study of wine making by-products (grape marc) in order to obtain valuable compounds (polyphenols) with antioxidant capacity, via solid-liquid extraction. Compared to the two former raw materials, grape marc represents higher-level cell organization, where different cell wall types may have diverse responses to pre-treatments. Microwave pre-treatments were studied both in the presence and in absence of an extracting solvent. Only solvent assisted microwave irradiation led to the successful intensification of kinetics in the performed comparative study. As a competing novel technique, the use of ultrasounds was also assessed through these studies, and Abstract 20 similar results were found when compared with microwave assisted extractions. Both pretreatment techniques were able to overcome the long extraction times and large solvent stocks required in the industrial extraction process. The possible thermal degradation of these valuable compounds was considered. Besides the study of these processes and the analysis of the effect of microwave irradiation on different raw materials, special importance has been devoted to the determination of the energy absorbed by the materials and solvents during the processes. The fraction of electromagnetic microwave energy absorbed by the sample is subsequently dissipated as heat. This heat can be described using an energy balance, which takes into account the sensible heat, by the temperature increment in the system, the latent heat, by solvent evaporation, and finally the heat loss to the environment. Unfortunately, results from published literature on the aforementioned topics are hard to compare with each other due to the lack of this data. Information about absorbed energy would facilitate the obtaining of comparable results in different experimental conditions, which are indispensable for scaling up. Based on the results discussed in this dissertation, further research can be performed in order to develop an industrial scale microwave assisted extraction process of grape marc polyphenols. Furthermore, the deeper study of the microwave effect on different cell structures and the role of the solvent in the microwave assisted extraction seem to be interesting issues to follow in the future. “If you have built castles in the air, your work need not to be lost; that is where they should be. Now, put the foundations under them.” /Henry David Thoureau/ Scope and Aims Scope & Aims 23 SCOPE In industrial processes, where natural raw materials are involved, the substance of interest may have intracellular localization. In order to facilitate the accessibility of the product in question, the natural cells have first to be damaged. This primary stage of the processes is often recognised as the rate limiting step, causing slow process kinetics, and therefore leading to long operation times, low recovery efficiencies or high operation costs. Microwave irradiation can improve cell wall rupture of high moisture materials. Due to the rapid heating up and evaporation of intracellular water, the pressure gradient may lead to cell wall damage. Besides the thermal effect of microwave energy, there may be a non-thermal effect, which is highly controversial in literature: the alternating dielectric field may be able to force the polarized side chains of the cell wall macromolecules to break their hydrogen bonds, and thus alter their structure. The microwave technology is widely used for analytical purposes due to the abovementioned features, achieving faster and more economical procedures. However, the electric energy used to transfer microwaves into any sample could lead to high operation costs in industrial scale applications. For this reason, an effective and short microwave pre-treatment on the natural material may be convenient prior to the conventional processing. This additional stage would not replace the existing process, but may modify the cell structure in order to facilitate the conventional method in continuation, thus intensifying the conventional process kinetics. Theoretical background 30 It is evident that excess sludge must be reduced through the wastewater treatment process, where it is produced. Different technologies have been studied so far in order to solve this problem, and enhance the process performance. Although reduction can be achieved either in the water line of the activated sludge process, or in the final waste line, by incineration or supercritical water oxidation, a wide range of techniques are available in the sludge line to reduce the final waste stream by enhanced anaerobic digestion. Physical, chemical or biological pre-treatments can be applied to sludge prior to the digestion step, while also modified digestion processes may also be performed in order to obtain better results. The disintegration of solid particles caused by pre-treatments leads to the release of intracellular compounds and facilitates the accessibility of substrates to anaerobic bacteria, thus accelerating the digestion process and increasing the degree of degradation. Although in the late ‘80s, mechanical cell disruption methods were preferred among physical technologies, nowadays the most successful treatments involve thermal hydrolysis, pressurized systems or alternative energy input by means of ultrasounds or microwave irradiation (Chisti and Moo-Young 1986, Pérez-Elvira et al., 2006). The possible use of microwave energy as a sludge disintegration method prior to anaerobic digestion is deeply discussed in Chapter 1. The influence of absorbed microwave energy and power is evaluated in terms of organic matter solubilisation and biogas production. In addition, a comparison with thermal treatment is performed to reveal a possible non-thermal microwave effect. 1.2. Microalgae: production of biodiesel and of high added value products Microalgae are photosynthetic microorganisms that convert sunlight, water and CO2 to algal biomass. This biomass is a good candidate in various scenarios, such as biofuels, foods, feeds and high added value compounds. As renewable biofuels, methane can be produced by anaerobic digestion of algal biomass, while microalga oil is appropriate for biodiesel production, and biohydrogen can be photobiologically produced. Biodiesel consists of triglycerides, where one glycerol molecule is esterified with three fatty acid chains. Through the transesterification with an excess amount of methanol in the presence of acid or alkaline catalyst, fatty acid methyl esters and glycerol are produced. For this process Theoretical background 31 to take place at atmospheric pressure at 60 °C, the necessary time period is 90 minutes. To avoid yield loss, both the oil and the reacting methanol must be dried to avoid saponification, and in addition free fatty acid content should not surpass a maximum value. In the final step, biodiesel is separated from glycerol and methanol by repeated washing with water (Chisti, 2007). When considering the future replacement of fossil fuels, microalgae also have to be considered, as well as the amount of oil, produced from crops and animal fat at present. Microalgae might be an alternative solution due to the lower requirement in cultivation area and fast reproduction rates. Moreover, specific microalgae can provide extremely high oil yields for the purpose of biodiesel production (Huang et al., 2010). The production of algal biomass is performed either in raceway ponds or in tubular reactors, where photosynthetic growth takes place under optimal conditions of CO2, light, water and inorganic salt supply. The majority biomass produced in daylight must be withdrawn from the reactor, while the rest is consumed during the night via the respiration pathway. The processing of the harvested algae starts with a dehydration step (spray-drying, drum-drying, freeze-drying and sun-drying). After the drying procedure, cells need to be disrupted in order to release the metabolites of interest (mechanical: cell homogenizers, bead mills, ultrasounds, autoclave, spray drying; and non-mechanical: freezing, organic solvent, osmotic shock, acid-base and enzyme reactions). Although it is difficult to achieve an energy efficient method in order to liberate the intracellular lipids from microalgae, it is necessary to facilitate and upgrade the extraction method, using a low solvent amount but still achieving high recovery of oil and minor high-value products. Several organic solvents can be used in the extraction process, such as hexane, ethanol or their mixtures, to recover up to 98% of purified fatty acids. The oil obtained is appropriate for transesterification in order to be converted into biodiesel, as has been described above (Chisti, 2007, Mata et al., 2010, Scott et al., 2010). The replacement of fossil fuels has not been achieved so far by microalgae technology, since extensive research is needed to meet economical, environmental and consumer demand in the field. The process must be improved from an economic point of view at different stages. Metabolic and genetic engineering enhance algal biology to obtain high oil content strains, while advances in photobioreactor engineering can achieve more stable biomass production. Theoretical background 32 Finally, downstream processes (harvesting and oil extraction) have to be intensified in order to reach a sustainable biodiesel process (Chisti, 2007, Huang et al., 2010). Besides the lipid content of microalgal cells, other valuable intracellular compounds are of interest in different industrial sectors such as fine chemicals and bioactive products. Pigments, antioxidants, β-carotenes, polysaccharides triglycerides, fatty acids, vitamins and biomass have applications with health benefits. Their importance is recognized and has been under extended research for the last thirty to forty years. Sterols present in algae cells are used in cardiovascular disease prevention. Antioxidants and carotenoids (especially lutein) are important in the prevention and treatment of degenerative diseases, while algal polysaccharide complexes have immune-modulating properties. It is important to mention long-chain polyunsaturated fatty acids (PUFA), especially ω-3 and ω-6 types such as EPA, DHA or AA, which are pharmacologically important for dietetics and therapeutics. Those compounds are believed to have a positive effect on cardio-circulatory diseases, coronary heart disease, hypertension, cholesterol problems and in cancer treatments (Mata et al., 2010). In the light of the great importance of substances with intracellular location in microalgae, Chapter 2 was dedicated to the study of microwave energy application in order to enhance the extraction kinetics of both lipid and pigment compounds from different microalgae species. 1.3. Grape marc and antioxidant extraction processes In the red wine making process the selected grapes, arrived from the field, suffer destemming and vigorous crushing in the first stage. No excessive crushing should be performed in order to avoid the increment of non-soluble solids and seed breakage, which leads to a high release of phenolic compounds. Depending on the wine type and grape quality, the fermentation conditions, temperature ramps and the contact time between the liquid and solid phases may vary to provide the desired taste, colour and phenolic component performance to the wine. During the process, sulphur dioxide is added to inhibit the polyphenol-oxidase activity and provide more stability to the phenolic components (Zoecklein et al., 1995). The discarded solid residue after the fermentation process, the so-called grape marc, has a high content of bioactive compounds. Taking into account the whole worldwide wine production, around 9 million tonnes of residue is generated every year. Thus the demand for natural antioxidants to Theoretical background 33 replace synthetic additives in the food industry can be connected to the large amount of high added value product, recovered from the residue of the wine industry (Lafka et al., 2007). The recovery of active components is also a convenient way to achieve their removal from the grape pomace, so that it can be used as a fertilizer preventing inhibition problems (Negro et al., 2003, Pinelo et al., 2006). Grape marc is a mixture of grape seeds, skin and a minimal amount of pulp. It contains active components such as flavonoids, polyphenols, anthocyanins, proanthocyanidins, procyanidins and stilbene derivatives. These components are of interest for the food-, cosmeticsand pharmaceutical industries due to their antioxidant, antimicrobial, antiviral or anticarcinogenic features (Nassiri Asl and Hosseinzadeh 2009). Different types of polyphenols may be found in different parts of grape (Figure 2). The layers of the epidermis and hypodermis are formed by the skin, and the upper part of the pulp cells, and , contain most part of the polyphenols and aromatic compounds. Figure 2 Schematic morphology of grape tissues (adapted from Hidalgo-Togores 2006) Most of the flavonoids are found in the epidermis, and in the hypodermis cells, including anthocyanin vacuoles and aromatic compounds. Tannins are located in free vacuoles, or in polymeric form connected to the polysaccharides of the cell walls, and also immobilized in the cell wall. The majority of tannin molecules are found in the grape seeds (Hidalgo-Togores 2006). Conventional solid liquid extraction techniques have been widely used in order to gain the antioxidant components from grape marc, or from their macroscopic components, skin and seeds. Various process conditions have been studied and optimized to obtain high extraction Theoretical background 34 yields with improved extract performance in terms of antioxidant activity and product stability. Different sample preparation processes (drying, milling, degreasing), solvents and solvent mixtures (ethanol, methanol, water, acetone, ethyl acetate, and their combination), and diverse solid:liquid ratios have been studied so far. The effects of pH, extraction temperature and time have also been optimized. Several studies concluded that extraction times between 2 and 5 hours are long enough to obtain high extraction yields at temperatures below 60 ºC, in order to avoid thermal degradation. Acidic pH is beneficial for better stability of phenolics, and the ethanol – water mixture was found to be the best option from an economical point of view, and also considering health aspects (Lafka et al., 2007, Spigno and De Faveri 2007). Quantitative determination of extraction kinetics is necessary for scaling up and process design modelling. Many studies found that polyphenol extraction shows good correlation by first order kinetics. The first order rate equation (Eq. 1) can be obtained combining Fick´s second law and the steady-state model taking into account the following assumptions: » solid particles are considered as flat plates; » active compound is initially distributed homogenously in the solid; » the porous solid is considered as a pseudo-homogeneous medium; » the content of the active compound in the solid varies with time and distance; » the thermodynamic equilibrium is established at the interface. ln (Ce/(Ce-Ct)) = k·t (1) Equilibrium polyphenol concentration (Ce) is obtained at infinite extraction time (t=∞). The concentration of polyphenols (Ct) at a certain time ‘t’ is readily obtained once the overall rate constant (k) is known. Similar diffusivity coefficients can be found in literature for different grapes and grape residues, which indicates that product nature, origin and environmental conditions are mainly responsible for differences in extraction yields from different samples of the same nature (Cacace and Mazza 2003, Amendola et al., 2010). Although conventional maceration or stirred extraction under mild conditions has been widely studied, it has the disadvantage of long processing times because of slow mass transfer. Because of this reason, it requires a huge and expensive solvent stock, which must be recovered and removed from the final product. In order to improve the mass transfer and extraction kinetics of antioxidants, different sample pre-treatments and novel extraction techniques were recently studied and compared, such as supercritical fluid extraction, microwave assisted extraction, Soxhlet-extraction or ultrasound assisted extraction (Pascual- Theoretical background 35 Marti et al., 2001, Martino et al., 2006, Casazza et al., 2010, Rodriguez-Rojo et al., 2012, Peralbo-Molina et al., 2012). The positive effects of microwaves and ultrasounds on polyphenol extraction kinetics from grape marcs have been deeply studied in Chapter 4 and Chapter 5 in comparison to thermal pre-treatment and conventional stirred extraction method, according to a Spanish patent (MoroGonzález 2010). The effect of possible thermal degradation due to heat treatments is evaluated and discussed in Chapter 6. 2. EXTRACTION PRE-TREATMENTS 2.1. Microwaves application and theory Electrical volumetric heating techniques include microwave heating, conduction and induction heating, ohmic heating, and radio frequency heating. In microwave heating, energy is transferred by electromagnetic waves instead of heat flux, as happens in conventional heating processes. Intensive research on microwave started in the Second World War to develop high-definition radars. In the post-war years the scene of microwave usage extended to heating purposes, either with the appearance of domestic microwave ovens or in industrial applications. Nowadays, it is widely used in the food industry, and in the ceramics, rubber and plastic industry and has been provoking a growing interest for its use in chemical applications (Meredith, 1998). Success behind microwaves is based on the heating theory of the oscillating electromagnetic field, in a frequency range between 0.3 and 300 GHz. Heat is generated by friction of molecules inside the material, due to the reorientation of dipoles in the changing electric field. In industrial and domestic applications a frequency of 2.45 GHz is commonly used, providing 0.94 kJ/molphoton energy with a wavelength of 12.2 cm. The molecule of water, an abundant component in natural raw materials, changes its orientation 109 times per second, producing rapid heating. This effect is enhanced by the presence of ionic components and other substances with low specific heat (Schubert and Regier 2005). In addition to the obvious thermal effect of microwaves, a so-called athermal or non-thermal effect has been reported. This occurs when the alternating electric field of microwaves is able to force the polarized side chains of macromolecules to break their hydrogen bonds, and thus Theoretical background 36 alter their structure (Hong et al., 2004; Woo et al., 2000). The evidence and importance of this phenomenon in natural and organic material processing is under discussion, and was theoretically attacked by Stuerga (2008), although contradictory results have been obtained so far. Evidence of an athermal effect is usually claimed or discarded by comparing results from parallel microwave and conventional heating processes. In the case of biogas production, where microwave and conventional heating pre-treatments were compared, contradictory results were found by different authors, showing the results for microwave heating as being either worse (Climent et al., 2007; Eskicioglu et al., 2006), better (Beszédes et al., 2011), or similar (Eskicioglu et al., 2007). The electromagnetic field and the propagation of microwaves (Figure 3) can be described by Maxwell’s equations (Eqs. 2-5), in which electrical phenomena is incorporated. The equations presented here are in vector form for a sinusoidal, time-varying field with angular frequency (ω=2πf [radian/s]), and in the case of no electric charge or magnetic dipoles (Eqs. 2-3) (Metaxas and Meredith, 1988, Meredith, 1998) 𝑑𝑖𝑣 𝑫 = 0 (2) 𝑑𝑖𝑣 𝑩 = 0 (3) 𝑐𝑢𝑟𝑙 𝑬 = −𝑗𝜔𝑩 (4) 𝑐𝑢𝑟𝑙 𝑯 = 𝑱 + 𝑗𝜔𝑫 (5) Equation (4) corresponds to Faraday’s law: the time changing electric field density (E) is related to the magnetic field (B). The electric field, circulating around a contour, is determined by the rate of change in the magnetic flux through the enclosed surface. Equation (5) describes Ampere’s law connecting the magnetic field intensity (H) with current density (J) and displacement density (D). It states that the magnetic field intensity circulating around a contour is determined by the net conduction and displacement current through the enclosed surface. Theoretical background 37 Figure 3. Electromagnetic wave (adapted from Navarrete, 2010) In addition to the equations above, explaining the interaction between electric and magnetic fields, auxiliary equations are needed to describe the interaction between the material and the electromagnetic fields (Eqs. 6-8.): 𝑱 = 𝜎𝑬 (6) 𝑫 = 𝜀0𝜀′𝑬 (7) 𝑩 = 𝜇0𝜇′𝑯 (8) Where σ is the conductivity of the material. The relative permittivities of the material and of the vacuum are ε’ and ε0 respectively, describing the non-conductive interaction of the material with the electric field. Interaction with the magnetic field is described by the relative permeability of the material, μ’, and of the vacuum, μ0. In microwave heating, the electric field is considered as the prime source of energy transferred to the workload (Meredith, 1998). For this reason, the material interaction with the electric field must be studied, using different workloads, for further design of the optimal microwave process. The complex dielectric permittivity of the material is determined as follows (Metaxas and Meredith, 1988): 𝜀 = 𝜀0𝜀𝑟 (9) 𝜀𝑟= 𝜀𝑟 ′+𝜀𝑟 ′′ (10) In Equation (9), ε0 and εr refer to the free space permittivity and the complex dielectric constant, respectively. Equation (10) shows the real part of the complex dielectric constant (εr’), which corresponds to the fraction of energy stored in the material when an electric field is applied, and the imaginary part (εr”), corresponding to the fraction of energy which is converted into heat. Theoretical background 38 During the microwave process, the material undergoes physical and perhaps chemical changes (moisture content, temperature and salt concentration) which affect its interaction with the dielectric field. These factors must be taken into account when characterizing the dielectric properties of the sample (Venkatesh and Raghavan 2004). Different measurement methods, and the effect of temperature and moisture on the dielectric properties of grape marc, are presented and discussed in Chapter 3. Industrial microwave systems are mainly composed of three parts: the microwave source, the waveguide and the applicator. In the microwave source (magnetron) the electric energy is converted into microwave energy. The waveguide conveys the electromagnetic waves to the applicator, to meet the workload with the microwave irradiation. Depending on the material to be treated and on power distribution needs, different applicator types can be used for processing. The most commons are multimode applicators, as in domestic microwave ovens, single-mode applicators and near-field applicator types. In order to achieve high power absorption in the system, and to avoid back-reflection from the applicator to the source, impedance matching has to be achieved. For this end, tuning systems are used, which can be controlled during the process and follow the change in the interaction of the material and the dielectric field. Considering these aspects in the microwave design, 95% of the output power from the magnetron can be efficiently dissipated in the workload. Magnetrons may have an efficiency around 85%, relatively to the input electric power. (Meredith, 1998, Schubert and Regier 2005). In household microwave ovens and laboratory experimental devices, the impedance matching and good efficiency may not be obtained due to the varying nature of the workload. Nevertheless, thermodynamic considerations can be used as an alternative to quantify the energy absorbed in the experimental setup, since the absolute value of the absorbed energy is so important from an engineering point of view. 2.2. Ultrasounds application and theory Ultrasounds generate forces of radiative nature including acoustic streaming forces. Ultrasounds are classified into two main categories according to their frequency. High frequencies, in the MHz range, are of low energy and used for diagnostic purposes. The low frequency range, between 20-100 kHz, called power ultrasound, is used to transmit high energy densities. The latter is frequently used in extraction processes, where the generated wave Theoretical background 39 energy is observed by the solvent, dispersed molecules and particles. However, the density of the propagated energy is lower than 10-9 J/atom, which is not high enough to break chemical bonds. A cavitation phenomena occurs due to ultrasounds. The compression and stretching of the molecular spacing cause the generation and violent collapse of cavitation bubbles in the liquid. The energy thus focused in the sound field is amplified by 11 orders of magnitude, and is therefore enough to break chemical bonds, and even to induce luminescence. Stable cavitation is obtained at lower power intensities (1-3 W/cm2), when the bubble size oscillates about an equilibrium value. For intensities higher than 10 W/cm2, called transient cavitation, the violent collapse of cavitation bubbles occurs after some expansion and compression cycles (Santos et al., 2009, Zhang et al., 2011). The physicochemical effects of ultrasound radiation are due to collapsing bubbles from transient cavitation. They act as micro-reactors of hot gas phase reactions where newly produced active radicals can appear at elevated local temperature and pressure (>1000 °C, 1000 bar), in an intensive mixed gas and liquid mixture. Mechanical effects, like degradation of large molecules, microstreaming of solid surfaces, cell disruption and cell content release depend on the nature of the system (Santos et al., 2009). Ultrasound processes are mainly affected by intensity, frequency, temperature, solvent properties, the presence of dissolved gases and pressure. The intensity of the ultrasound energy is proportional to the square of the amplitude. There is a required minimum value in order to achieve cavitation in the medium, although excessively high intensities may cause degradation of the desired compounds. Additionally, an excessively high amplitude causes insufficient coupling between the transducer and the liquid, thus decreasing the amount of energy transmitted into the medium, and reducing the operating lifetime of the transducer. In the high frequency range, the cycles become too short to achieve cavitation in the medium, and so frequency is selected in the kHz range, and usually constant in the process. Increased temperature has a positive effect on the process, like in the case of extraction, but at the same time causes a less violent collapse of the cavitation bubbles, as they are filled with solvent vapour at higher temperatures. An increase in temperature also affects medium properties, like viscosity or surface tension. Low surface tension and the use of surfactants enhance the sonochemical effects. Likewise, low solvent viscosity enhances cavitation, but high resistance is showed against the generation of cavitation bubbles in high viscosity liquids. The presence of dissolved gases or the introduction of gas bubbles also promotes the generation of the Theoretical background 46 REFERENCES « Amendola, D., De Faveri, D. M., Spigno, G. (2010) Grape marc phenolics: Extraction kinetics, quality and stability of extracts. Journal of Food Engineering, 97, 384 – 392. « Appels L., Baeyens J., Degreve J., Dewil R. (2008) Principles and potential of anaerobic digestion of waste-activated sludge. Progress in Energy and Combustion Science, 34, 755781. « Beszédes, S., László Zs., Horváth, Zs. H., Szabó G., Hodúr C. (2011) Comparison of the effects of microwave irradiation with different intensities on the biodegradability of sludge from dairyand meat-industry. Bioresource Technology, 102(2), 814-821. « Cacace, J.E., Mazza, G. (2003) Mass transfer process during extraction of phenolic compounds from milled berries. Journal of Food Engineering, 59, 379–389. « Casazza, A. A., Aliakbarian, B., Mantegna, S., Cravotto, G., Perego, P. (2010) Extraction of phenolics from Vitis vinifera wastes using non-conventional techniques. Journal of Food Engineering, 100, 50 –55. « Chisti Y. (2007) Biodiesel from microalgae. Biotechnology Advances, 25, 294-306. « Chisti, Y., Moo-Young, M. (1986) Disruption of microbial cells for intracellular products. Enzyme and Microial. Technology, 8, 194-204. « Climent, M., Ferrer, I., Baeza, M., Artola, A., Vázquez F., Font, F. (2007) Effects of thermal and mechanical pretreatments of secondary sludge on biogas production under thermophilic conditions. Chemical Engineering Journal, 133, 335–342. « Eskicioglu, C., Kennedy K. J., Droste, L. R. (2006) Characterization of soluble organic matter of waste activated sludge before and after thermal pretreatment. Water Research, 40, 3725-3736. « Eskicioglu, C., Terzian, N., Kennedy K. J., Droste, L. R, Hamoda M. (2007) Athermal microwave effects for enhancing digestibility of waste activated sludge. Water Research, 41, 2457-2466. Theoretical background 47 « Hidalgo-Togores, J. (2006) La calidad del Vino desde el Viñedo. Mundi Prensa, Madrid, Spain « Hong S.M., Park J.K., Lee Y.O. (2004) Mechanisms of microwave irradiation involved in the destruction of fecal coliforms from biosolids. Water Research, 38, 1615-25. « Huang, G.H., Chen, F., Wei, D., Zhang, X.W., Chen, G. (2010) Biodiesel production by microalgal biotechnology. Applied Energy, 87, 38–46. « Lafka T.I.; Sinanoglou, V.; Lazos, E. S. (2007) On the extraction and antioxidant activity of phenolic compounds from winery wastes. Food Chemistry, 104, 1206–1214. « Martino, E., Ramaiola, I., Urbano, M., Bracco, F., Collina, S. (2006) Microwave-assisted extraction of coumarin and related compounds from Melilotus officinalis (L.) Pallas as an alternative to Soxhlet and ultrasound-assisted extraction. Journal of Chromatography A, 1125, 147 – 151. « Mata, T.M., Martins, A.A., Caetano, N.S. (2010) Microalgae for biodiesel production and other applications: A review. Renewable and Sustainable Energy Reviews, 14(1), 217-232. « Meredith, R. (1998) Engineer´s handbook of industrial microwave heating. The Insitution of Electrical Engineers. London, United Kingdom. « Metaxas A.C., Meredith R.J. (1988) Industrial Microwave Heating. Peter Peregrinus Ltd., London, United Kingdom. « Moro-González, C. (2010) Procedure of polyphenol extraction from distilled grape marc (Procedimiento de extracción de polifenoles a partir de orujo de uva procedente de destilación). 2319032 B1. Spanish National Patent. « Nassiri-Asl, M., Hosseinzadeh, H. (2009) Review of the Pharmacological Effects of Vitis vinifera (Grape) and its Bioactive Compounds. Phytotherapy Research, 23, 1197 – 1204. « Navarrete, A., (2010) Studies in the development of a continuous process for the extraction of essential oils using microwaves. PhD Thesis, University of Valladolid, Department of Chemical Engineering and Environmental Technology Theoretical background 48 « Negro, C., Tommasi, L., Miceli, A. (2003) Phenolic compounds and antioxidant activity from red grape marc extracts. Bioresource Technology, 87, 41 – 44 « O’Flaherty V., Collins G., Mahony T. (2006) The microbiology and biochemistry of anaerobic bioreactors with relevance to domestic sewage treatment. Reviews in Environmental Science and Biotechnology, 5, 39-55. « Pascual-Martí, M.C., Salvador, A., Chafer, A., Berna, A. (2001) Supercritical fluid extraction of resveratrol from grape skin of Vitis vinifera and determination by HPLC. Talanta, 54, 735 – 740. « Peralbo-Molina, A., Priego-Capote, F., Luque de Castro, M. D. (2012) Comparison of extraction methods for exploitation of grape skin residues from ethanol distillation. Talanta, 101, 292–298. « Perez-Elvira S. I., Nieto Diez P., Fdz-Polanco F. (2006) Sludge minimization technologies. Reviews in Environmental Science and Biotechnology, 5, 375 – 398. « Pinelo, M., Arnous, A., Meyer, A.S. (2006). Upgrading of grape skins: Significance of plant cell-wall structural components and extraction techniques for phenol release. Trends in Food Science & Technology, 17, 579-590 « Rodríguez-Rojo, S., Visentin, A., Maestri, D., Cocero, M. J. (2012) Assisted extraction of rosemary antioxidants with green solvents. Journal of Food Engineering, 109, 98–103 « Santos, H. M., Lodeiro, C., Capelo-Martinez, J. L. (2009) The Power of Ultrasound. In Capelo-Martinez, J.L. (Ed) Ultrasound in Chemistry: Analytical Applications (pp. 1-16 ) WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim, Germany. « Schubert, H., Regier, M. (2005) The microwave processing of foods, Woodhead Publishing. Cambridge, United Kingdom. « Scott, S.A., Davey, M.P., Dennis, J.S., Horst, I., Howe, C.J., Lea-Smith, D.J., Alison, G. (2010) Biodiesel from algae: challenges and prospects. Current Opinion in Biotechnology, 21, 277–286. Theoretical background 49 « Spigno, G., De Faveri, D.M. (2007) Antioxidants from grape stalks and marc: Influence of extraction procedure on yield, purity and antioxidant power of the extracts. Journal of Food Engineering, 78, 793 – 801 « Stuerga, D. (2008) Microwave-Material Interactions and Dielectric Properties, Key Ingredients for Mastery of Chemical Microwave Processes, in Microwaves in Organic Synthesis, Second Edition (ed A. Loupy), Wiley-VCH Verlag GmbH, Weinheim, Germany. « Venkatesh, M.S., Raghavan, G.S.V. (2004) An overview of microwave processing and dielectric properties of agri-food materials. Biosystems Engineering, 88(1), 1–18. « Vilkhu, K., Mawson, R., Simons, L., Bates, D. (2008) Applications and opportunities for ultrasound assisted extraction in the food industry — A review. Innovative Food Science and Emerging Technologies, 9, 161–169. « Woo I., Rhee I., Park H. (2000) Differential damage in bacterial cells by microwave radiation on the basis of cell wall structure. Applied Environmental Microbiology, 66, 2243–2247. « Zhang, H. Q., Barbosa-Canovas G. V., Balasubramaniam, V. M., Dunne, C. P., Farkas, D. F., Yana, T. C. (Eds) Nonthermal Processing Technologies for Food (pp. 109-134). WileyBlackwell, UK. Feng, H., Yang, W.: Ultrasound processing; Weiss, J., Gulseren, I., Kjartansson, G.: Physochemical Effects of High-Intensity Ultrasonication on Food Proteins and Carbohydrates « Zoecklein, B.W., Fugelsang, K.C., Gump, B.H., Nury, F.S. (1995) Wine Analyses and Production; Chapter 7. Chapman and Hall, New York, USA. Chapter 1. The influence of energy absorbed from microwave pre-treatment on biogas production from secondary wastewater sludge Sólyom, K. Mato, R.B., Pérez-Elvira, S.I., Cocero, M.J. Bioresource Technology 102 (2011) 10849-10854. Chapter 1 53 THE INFLUENCE OF THE ENERGY ABSORBED FROM MICROWAVE PRE-TREATMENT ON BIOGAS PRODUCTION FROM SECONDARY WASTEWATER SLUDGE Abstract In this study, microwave treatment is proposed as a way to accelerate the hydrolysis in anaerobic digestion of municipal wastewater sludge. The influence of the absorbed energy, power and athermal microwave effect on organic matter solubilization and biogas production has been studied. In addition, a novel method that considers the absorbed energy in the microwave system is proposed, in order to obtain comparable experimental results. The absorbed energy is calculated from an energy balance, considering the sum of sensible and latent heats, and heat loss from a sample surface. The highest solubilization was achieved using 0.54 kJ/ml at 1000W, where an increment of 7.1% was observed in methane production, compared to the untreated sample. Using a higher energy value (0.83 kJ/ml), methane production further increased (to 15.4%), but solubilization decreased. No power influence was found when 0.54 kJ/ml was applied at 1000, 600 and 440 W. Microwave heating was compared to conventional heating in two different experimental setups, providing similar methane yields in all cases. Chapter 1 54 1. INTRODUCTION Municipal waste water treatment plants generate large amounts of excess sludge which must be disposed of. This is a problem of growing importance, representing up to 50% of the current operating costs of a wastewater treatment plant (Appels et al., 2008). Most plants use anaerobic digestion to stabilize the organic material of this sludge and reduce its biomass. In addition, it produces biogas and improves the dewatering properties of the residual sludge (O’Flaherty et al., 2006). The anaerobic degradation of the sludge is a three-step sequence: 1) hydrolysis, 2) acidogenesis, and 3) methanogenesis, in which the former, hydrolysis, is accepted as the rate-limiting step. Sludge particles are essentially concentrated aerobic microbial cells. In order to be hydrolyzed, anaerobic bacteria must release extracellular enzymes that break down and solubilise the rigid microbial cell walls (O’Flaherty et al., 2006; Appels et al., 2008). In order to increase the biodegradability of the sludge, various mechanical, ultrasonic, chemical and thermal pretreatment methods have been proposed (Perez-Elvira et al., 2006). The pretreatment by irradiation of the sludge using microwaves has been one of the methods reported repeatedly in literature over the last decade. Microwaves can improve the rupturing of the cell wall, thus enhancing anaerobic hydrolysis in two different ways. Firstly, the thermal effect corresponds to degradation caused by temperature increase. Also, the internal heating and evaporation of the intracellular water causes an increase in internal pressure that can lead the cell wall to rupture (Lyons and Hatcher, 1972; Golmakani et al., 2008; Lucchesi et al., 2007; Uquichea et al., 2008). Secondly, the so-called athermal or non-thermal effect must be considered. This occurs when the alternating electric field of microwaves is able to force the polarized side chains of the cell wall macromolecules to break their hydrogen bonds, and thus alter their structure (Hong et al., 2004; Woo et al., 2000). Several studies have been reported investigating optimal conditions (power, treatment time and temperature) for microwave wastewater sludge treatment with the aim of solubilizing organic matter or enhancing anaerobic digestion. In some of these papers, microwave treatment is compared with other thermal or ultrasonic processes in order to assess its potential as a sludge disintegration method (Beszédes et al., 2009, Beszédes et Chapter 1 55 al., 2011, Climent et al., 2007, Eskicioglu et al., 2006, Eskicioglu et al., 2007, Guo et al., 2008 ). However, the results reported in these studies lead to conflicting results: » Effect of power on organic matter solubilization. Organic matter solubility is calculated from the COD (chemical oxygen demand) of the soluble phase. According to some authors, the use of higher power during microwave treatment leads to lower solubilization (Eskicioglu et al., 2007; Park et al., 2010; Toreci et al., 2009), while others report the opposite effect (Climent et al., 2007). » Effect of power on biogas production. Similar (Eskicioglu et al., 2007) and higher (Toreci et al., 2009) biogas production were found with increasing microwave power. » Athermal effect. The presence of an athermal effect is usually determined through the comparison of microwave heating and conventional heating processes. The biogas production reported in these studies, where microwave instead of conventional heating was used, was in some cases lower (Climent et al., 2007; Eskicioglu et al., 2006), in another case higher (Beszédes et al., 2011), and in a last one it remained unchanged (Eskicioglu et al., 2007). The use of different raw materials can lead to different quantitative results. However, the reason for these contradictory conclusions can be explained due to the difficulty in comparing results using different experimental conditions in microwave treatments. Only a fraction of the power applied in microwave heating is absorbed by the sample material. This fraction depends on factors such as dielectric properties, the size and geometry of the sample, microwave frequency and intensity, process time, and oven cavity characteristics (Swain and James, 2005; Campañone and Zaritzky, 2005; Gunasekaran and Yang, 2007; Zhu et al., 2007). In order to analyze and compare results from several experiments it is necessary to determine the fraction of energy absorbed by the sample. The triple objective of this study was, firstly, to develop an experimental procedure in which absorbed microwave energy could be calculated from an energy balance; secondly, to evaluate the influence of absorbed microwave energy and power on treated sludge; and thirdly, to reveal the presence of a possible athermal effect. The influence of absorbed Chapter 1 62 Biogas production was determined under mesophilic conditions (35ºC). Samples were inoculated with anaerobic sludge from wastewater bioreactors in 125 ml-bottles, in triplicate. Substrate/ inoculum ratio was chosen as 0.5 of the volatile solids (g/g). The evolution of headspace pressure was measured in order to quantify the biogas, and its composition was analyzed by gas chromatography (Varian CP-3800, USA) with a thermal conductivity detector (TCD), applying helium as the carrier gas. Accumulated methane volume (35ºC) was calculated per gram of added volatile solids (mL CH4/g VS). 3. RESULTS AND DISCUSSION 3.1. Energy efficiency measurements Equipment efficiency varied between 62 and 77% depending on the process variables (Table 1). Heat loss contribution was in the range of 10-16% of absorbed energy, increasing with the treatment time. Sensible heat was the main contribution (≈86.5%) to absorbed energy, when boiling temperature was not achieved. When water evaporation occurred, this percentage was divided between the sensible and the latent heat, depending on the amount of evaporated water. 3.2. Microwave energy effect on solubilization and biogas production Using a constant power (1000 W), different treatment times were used in order to apply a wide range of absorbed energy values to the sludge samples. During the treatment, the experimental setup remained opened to atmosphere, therefore the sample temperature only increased up to the boiling point, when water evaporation started at a constant temperature. Figure 2 shows the results of COD solubilization and protein concentration versus the absorbed microwave energy. At low absorbed energies, the solubilization values, as well as the protein concentrations, increased. When the absorbed energy was around 0.40.5 kJ/ml the sludge reached its boiling point, and the maximum values of solubilization and proteins were obtained. Chapter 1 63 Figure 2 (●) : COD solubilization (Eq. 6), and (♦) : protein concentration of microwave treatments at constant power (1000 W) in a wide range of absorbed energy. Beyond this value, water evaporation took place at a constant temperature, and solubilization and protein concentration showed a slightly decreasing pattern. Although the evaporated water was replaced after treatments, for high values of absorbed energy, the evaporation partially dried the sludge samples during the treatment, and therefore sludge floc agglomeration may have occurred. Part of the liberated soluble matter could have been blocked inside the agglomerates, and later rehydration using the evaporated water was not enough to solubilise them. Antoher reason for the COD solubilization decreasing could be the loss of volatiles during the water evaporation. As previously explained, experiments were performed at ambient pressure to allow for cell wall rupture by intracellular water evaporation. From these experiments, an approximate value of 0.4-0.5 kJ/ml can be established as the optimal absorbed energy for organic matter solubilization, when the sludge sample reached its boiling point. Metal content measurements (Table 4.) showed that slight liberation occurred in the case of Al, Ni, and Zn content in the soluble fraction. Fe content slightly decreased with higher energy treatment. 0 1 2 3 4 5 6 0 1 2 3 4 5 6 0.00 0.50 1.00 1.50 Protein concenctration (mg BSA/L) Solubilization Absorbed Energy (kJ/ml) Chapter 1 64 Table 4 Metal content (mg/kg) of total and soluble fractions before (Control) and after microwave treatment at 1000W for 0.7 and 1.3 min. Al Cd Cr Cu Fe Mn Ni Pb Zn Control soluble 0.564 <0.40 <0.40 <0.40 15.5 <0.40 <0.40 <0.40 0.478 0.7 min soluble 2.79 <0.40 <0.40 <0.40 14.7 <0.40 0.529 <0.40 2.11 1.3 min soluble 2.72 <0.40 <0.40 <0.40 13.5 <0.40 0.496 <0.40 1.91 Control total 701 0.083 2.14 10.1 795 5.41 2.11 3.08 66 0.7 min total 641 0.087 2.07 10.3 774 5.47 2.15 2.96 66 1.3 min total 716 0.091 2.25 10.8 836 5.71 2.25 3.24 70 Among these metals, zinc and nickel are the principal elements limiting sludge recycling to agricultural land. However, a low level of nickel would not affect the bioprocess (Alvarez et al., 2002). Guo et al., (2008) reported that microwave sludge treatment can liberate heavy metals, and that elevated Cu content inhibits the H2 production in anaerobic digestion. However, this effect was more serious in the case of ultrasonication and could not be observed in the sterilization treatment. In their experiments, 67.2 kJ microwave energy was used for an unreported quantity of sludge, thus it cannot be compared to the treatments performed in this study. In order to ensure against possible inhibition risks during the digestion process, a screening of separate metal forms by sequential extraction has been proposed in literature (Alvarez et al., 2002). After organic matter solubilization measurements, biogas production was determined for experiments 1, 2 and 3 (Table 1), and compared to one untreated sample. As shown in Figure 3, the final CH4 production was 195 ±7 ml/g added volatile solids in the case of the untreated control sample. For treated samples, production yields were 206 ±4 ml, 209 ±11 ml and 225 ±6 ml for experiments 1, 2 and 3 respectively. Although the highest biogas production was obtained in the case of the highest absorbed energy treatment, it can be observed that the time required to produce as much biogas as the control in 22 days was reduced by 20% in experiments 1 and 2, and by 50% in experiment 3. The decrease of the Chapter 1 65 residence time in digestion would allow for the use of smaller digestion tanks, so reducing the capital and operational costs of the process. Figure 3 Accumulated CH4 production (ml/g VS added) for control and treatments with 0.40, 0.54 and 0.83 kJ/ml absorbed energy at constant power (1000 W). (-●-): Control, untreated sample; (- -□- - ): 0.40 kJ/ml; (••∆••): 0.54 kJ/ml; (--◊--): 0.83 kJ/ml Although experiment 2 (0.54 kJ/ml) was the optimum for the organic matter solubilization, biogas production tests did not confirm this result, since biogas yield continued to increase with absorbed energy in microwave treatment, even beyond the boiling point. Although the soluble fraction was not positively affected over the boiling point, further structural changes in sludge flocs could have occured, having a positive effect on the anaerobic digestion. In the case of the untreated sample, the methane content in biogas composition was 45.10±0.46 % on the first day, increased up to 67.30±0.91 % on the second day and stabilized at 62.94±1.01 % during the 20 days of production. Microwave treatments did not changed the methane yield in biogas composition, indicating an undisturbed bioprocess in biogas production. In the study of Beszédes et al., (2009) microwave treatment was used in dairy sewage sludge to enhance anaerobic digestion. The 5 W/g treatments resulted in the same solubilization as 10 W/g, however 10 W/g treatment had a significantly higher biogas yield. Higher energy in closed vessels resulted in higher treatment temperature, thus, enhanced biogas generation was observed (Eskicioglu et al., 2009). Tang et al., (2010) studied the treatment energy required to obtain maximum biogas production with microwaves. In their study, the experimental setup worked up to the boiling temperature 0 50 100 150 200 250 0 10 20 30 Accumulated methane production (ml/g VS added) Digestion time (days) Chapter 1 66 of sludge (≈100ºC), as in this manuscript. The maximum organic matter solubilization was reached at the boiling point with a specific energy of 0.12 kJ/ml, four-fold lower than the energies used here (0.4-0.5 kJ/ml). The energy required to boil the sludge sample depends on their initial temperature, which could explain such a difference between the results mentioned. 3.3. The effect of microwave power on solubilization and biogas production In this set of experiments, a value close to 0.54 kJ/ml of absorbed energy was used at 1000, 600 and 440 W with 1.2; 2 and 2.7 min treatment times, respectively. The obtained COD solubilization and protein concentration results can be seen in Figure 4. Allowing for an error of 3 and 4% in solubilization data and protein concentration, respectively, no clear differences can be observed between the experiments with microwave treatment. Figure 4 COD solubilization results and protein concentrations results using 0.54 kJ/ml absorbed energy treatment with 1000, 600, 440 W and control sample. (■): Solubilization; (■): Protein concentration (mg BSA/L) Similar COD solubilization was also found by Chang et al., (2011) when waste activated sludge was treated by microwaves at 300, 450 and 600 W until sludge boiling occurred at 80ºC. Park et al., (2010) found considerably higher solubilization values, using 400 W instead of 1600 W at 60, 90 and 120ºC. The same solubilization tendency was observed using lower microwave intensity at a higher temperature (175ºC) (Toreci et al., 2009), and also at lower temperatures (50, 75 and 96ºC) (Eskicioglu et al., 2007). These results 3.62 3.35 3.27 0.05 4.90 4.59 4.56 0246 0 440W 600W 1000W Microwave power at 0.54 kJ/ml Protein (mg/l) Solubility Chapter 1 67 suggest that differences in treatment time may be responsible for better solubility results, since longer processing times are required at lower powers to reach the same final temperature. On the other hand, Climent et al., (2007) obtained higher solubilities at 800 W than at 400 W, with 13000 kJ/kg suspended solids (equivalent to 0.48 kJ/ml sludge). However, in the same study, higher solubility values were obtained at 400 W with a sample of 7800 kJ/kg suspended solids. Despite these contradictory results, the present study did not find clear differences in COD solubilization at different power levels. It is probable that the difference between the results may be caused by the alteration of the energy absorption in the experimental setup. After the analysis of the organic matter solubilization, the biogas production was measured. In Figure 5, the accumulative biogas production can be seen for raw sludge and for microwave treated samples. All treatments with 0.54 kJ/ml enhanced the methane yield (211±11 ml/g VS added), compared to the untreated sludge (195±7 ml/g VS added), but almost identical methane production curves were obtained at all power levels. This confirms the observation drawn from COD solubilization, that absorbed energy as a specific treatment condition plays a greater role than the applied power. Figure 5 Accumulated CH4 production (ml/g VS added) for power effect experiments at 0.54 kJ/ml. (-●-): Control, untreated sample; (- -□- - ): 440 W; (••∆••): 600 W; (--◊--): 1000 W Toreci et al., (2009) reported higher biogas production at higher microwave intensity; however, the solubilization was better at lower power levels. There is a lack of information about the effect on biodegradability at different microwave powers in other 0 50 100 150 200 250 0 5 10 15 20 25 Accumulated methane production (ml/g VS added) Digestion time (days) Chapter 1 68 studies, usually focusing on organic matter solubilization for this analysis (Climent et al., 2007; Eskicioglu et al., 2007; Park et al., 2010). 3.4. Microwave athermal effect In a first set of experiments (Table 2) COD solubilization results were 3.27 and 5.26 for microwave (MW1) and conventional heating (CH1) respectively. The same behavior was observed by Climent et al., (2007) for high and low temperature treatments (135ºC and 70ºC) with conventional heating versus microwave treatment. Eskicioglu et al., (2007) obtained the trend observed here, working at lower temperatures (50, 75, 96ºC). However, Beszédes et al., (2009) observed lower biological oxygen demand solubilization at 95ºC with convective heat treatments than with microwaves. It is believed that through conventional heating over a longer time period, more solubilization can be achieved, while the rapid microwave treatment does not allow time enough for organic matter liberation in the process (Eskicioglu et al., 2006). For this reason, here, a second experimental setup is proposed to compare conventional heating to the microwave treatment. Figure 6 Accumulated CH4 production (ml/g VS added) for compared experiments with conventional and microwave heating. (♦): Control 1, untreated sample; (▲): MW 1; (■): CH 1; (◊):Control 2, untreated sample; (∆): MW 2; (□): CH 2. In that set of experiments (MW2 and CH2), equal processing times and treatment temperatures were applied in order to reveal a possible athermal effect when using microwaves. COD solubilization was 22.45 and 20.07 for MW2 and CH2 respectively. 0 50 100 150 200 250 0 10 20 30 Accumulated CH4 production (ml/g VS added) Time (days) Chapter 1 69 Apart from the identical temperature and treatment times in both experiments, in the case of MW2, continuous vapor reflux was performed to avoid the agglomeration of sludge flocs due to water evaporation during treatment. These differences indicate that the comparative experiments need to be similar in as many operating conditions as possible to obtain comparable results. In the biogas production test, the same tendency was observed for both the experimental series (Figure 6). MW1 and CH1 resulted in 208±12 ml/g VSadded and 203±1 ml/g VSadded methane production yield respectively, while 195±7 ml/g VSadded methane yield was observed with untreated sludge (Control 1). A further increase in production can be seen in the second set of experiments when compared to its control (Control 2: 192±2; CH2: 212±4; MW2: 213±1), but now there is no clear difference between microwave treated and conventionally heated sludge. Higher biogas generation was reported (Eskicioglu et al., 2007) for microwave treated samples at 50, 75 and 96ºC compared to the conventionally heated sludge, although in an earlier study (Eskicioglu et al., 2006) the conventional heating at 96 ºC was more effective than microwaves in terms of biogas yield. From the results of the present study, no evidence of the athermal effect of microwaves was found, since similar biogas production was obtained using the conventional heating sludge pretreatment. Chapter 1 70 4. CONCLUSIONS » In microwave treatments experiments, the absorbed energy must be reported as a fundamental operating variable, and used when comparing operating conditions. » The highest value in COD solubilization was obtained using 0.54 kJ/ml of absorbed energy; however, the highest increment in methane production (15%) was achieved with the highest value used (0.83 kJ/ml). » Using the same absorbed energy, different power levels did not give conclusive COD solubilization results; however, there was no difference in methane production when the power lever was changed. » No evidence of the athermal effect of microwaves was found, as similar biogas production results were obtained with microwave treatment and conventional heating. Chapter 1 71 ACKNOWLEDGEMENTS The authors wish to thank the Junta de Castilla y León for their financial support (Project GR11-2008). REFERENCES « Ahn, H J, Shin, S. G.,Hwang, S., 2009. Effect of microwave irradiation on the disintegration and acidogenesis of municipal secondary sludge. Chem. Eng. J. 153. 145-150. « Alvarez, A.,Callejón Monchon M., Jimenez Sanchez, J. C., Ternero Rodriguez, M., 2002. Heavy metal extractable forms in sludge from wastewater treatment plants. Chemosphere. 47. 765-777. « Appels L., Baeyens J., Degreve J., Dewil R., 2008. Principles and potential of anaerobic digestion of waste-activated sludge. Prog. Energy Combust. Sci. 34. 755781. « Beszédes, S., László Zs., Szabó G., Hodúr C., 2009. Examination of the effect of microwave heating on the biodegradable and soluble fraction of organic matter of sludge. Journal of Engineering Annals of Faculty of Engineering Hunedoara 7(4). 8790. « Beszédes, S., László Zs., Horváth, Zs. H., Szabó G., Hodúr C., 2011. Comparison of the effects of microwave irradiation with different intensities on the biodegradability of sludge from dairyand meat-industry. Biores. Technol. 102(2). 814-821 « Campañone, L. A., Zaritzky N. E., 2005. Mathematical analisys of microwave heating process. J. Food Eng. 69. 359-368. « Chang C.J., Tyagi V. K., Lo S.L., 2011. Effects of microwave and alkali induced pretreatment on sludge solubilization and subsequent aerobic digestion. Biores. Technol. doi: 10.1016/j.biortech.2011.05.031 Chapter 2 78 1. INTRODUCTION The interest in microalgae strains from a biotechnological point of view has grown over the last decades. As photosynthetic organisms, they convert CO2, water and light into complex organic compounds of industrial interest. These include biofuels such as biogas (the consequence of the anaerobic digestion of algal biomass), photobiologically produced biohydrogen, or biodiesel, which can be obtained from algal oil. In addition microalgae contain high-added-value components such as polyunsaturated fatty acids, vitamins and pigments which play important role in food, pharmaceutical and cosmetic industries (Guedes et al., 2011, Koberg et al., 2011). When compared with the extensive land requirement and slow regeneration of crops and other plant tissues, the fast growth rate and specific accumulation or secretion of the substances to be recovered, convert the algal biomass into a potential competitor in the field (Chisti 2007). The quality and quantity of the compounds of interest can be optimized by nutritional and environmental factors, cultivation conditions and optimal harvesting at certain growth phases (Mata et al., 2010, Guedes et al., 2011). A cost effective and sustainable process for obtaining bio-products from microalgae is not yet available nowadays. To achieve this target, the different process stages are under extended research. First to be mentioned are the upstream processes: species screening or genetic modification of the strains, the cultivation conditions (open ponds or closed bioreactors) and nutrition (direct caption of industrial CO2 outlet and wastewater usage as growth medium). To continue, further challenges in the downstream scenario such as harvesting (flocculants, high throughput centrifuge) and the recovery process of the substances (dehydration for dry extraction, wet extraction, extraction condition optimization, cell disruption methods) are also under exploration in order to achieve energy efficiency and high throughput production (Mata et al., 2010, Sorguven and Özilgen 2010, Halim et al., 2011, Pfromm et al., 2011). The replacement of fossil fuels has not been achieved so far by microalgae technology, where economic, environmental and consumer issues must be altogether satisfied. Several process stages must still be improved to satisfy economic criteria. Metabolic and genetic engineering is enhancing algal biology to obtain high oil content strains, while advances in photobioreactor engineering can achieve a more stable biomass production. Finally, Chapter 2 79 downstream processes (harvesting and oil extraction) have to be intensified in order to reach a sustainable biodiesel process (Chisti, 2007, Huang et al., 2010). Different cell disruption methods can be used to enhance the extraction of valuable intracellular components from microalgae. Among various mechanical, chemical, thermal methods, and other novel methods (Lee et al., 2010, Halim et al., 2011, Wiltshire et al., 2000, Halim et al., 2012, McMillan et al., 2013, Šoštarič et al. 2012), microwave treatment has been proposed for cell wall rupture. As an oscillating electric field, microwaves cause rapid alignment and realignment of dipoles in a polar solvent, which produces rapid heating. Internal vapour generation in materials with high water content establishes a pressure gradient inside the sample (Lyons and Hatcher 1972, Navarrete et al., 2011). Such a pressure difference may cause the damage of the algal cell walls, and let the intracellular material become more accessible to the solvent. Microwaves have been applied as a pre-treatment prior to solvent extraction, thus only the biomass is involved in the irradiation stage, in the absence of the organic solvent (Lee et al., 2010, Balasubramanian et al., 2011, Biller et al., 2013). In other studies (Pasquet et al., 2011, Iqbal and Theegala, 2013) microwave assisted extraction (MAE) was performed to extract algal pigments or lipids, where the solvent and solute were both exposed together to the microwaves. In the case of the pigment extraction process, the non-polar hexane solvent did not react to the microwave irradiation, but immediately dissolved the released intracellular components (Pasquet et al., 2011). For lipid extraction, it was found that biodiesel can be used as co-solvent in MAE as an alternative to the toxic hexane (Iqbal and Theegala, 2013). Other organic solvents, like alcohols, react better to microwaves, thus allowing microwave assisted transesterification in the presence of catalysts, with methanol as a reactant (Barnard et al., 2007). By connecting these two features, a combined one-step extraction – conversion technique can also be used for biodiesel production (Patil et al., 2011, Patil et al., 2012). There are plenty of problems to be solved in order to achieve a sustainable production of biofuels or other valuable substances from microalgae. Among these, the cell disintegration step will always play an important role, since cells need to be disrupted to release intracellular lipids from microalgae. However, the achievement of an energy efficient method has proven to be difficult to overcome, but necessary to facilitate and Chapter 2 80 upgrade the extraction process, by reducing the requirement of solvent stock while still obtaining high oil and high-value products recoveries (Chisti, 2007, Mata et al., 2010). The selected technique must be adaptable to the likelihood of a variable algal feedstock. The aim of this study was to investigate the effect of microwave pre-treatments on extraction kinetics from different microalgae species using different extraction processes. The object, in this case, was that whether the algal specie, the obtained product, or the extraction technique were different, it would always be possible to demonstrate the cell damage due to the microwave irradiation. Microwave energy was applied on concentrated microalgae paste to achieve the heating up of the intracellular water. Extraction kinetics measured in Soxhlet extraction from dried Nannochloropsis gaditana and in stirred extraction from wet Secenedesmus almeriensis were studied. Gravimetric analyses, UVVis spectrophotometry, scanning electron microscopy and fatty acid composition determination by gas chromatography were performed in order to follow the changes produced by the microwave pre-treatments on the algal biomass. 2. MATERIALS AND METHODS 2.1. Microalgae Two different microalgae species were used in the experiments. Harvested and dewatered Nannochloropsis gaditana (for lipid extraction) samples were received from the Department of Vegetal Production (University School of Agricultural Engineers, Universidad Politécnica de Madrid). Initial water content was 84%. Samples were treated with microwaves within 2 days of their reception, dried and stored under vacuum at 4ºC until extraction. The other species, Scenedesmus almeriensis (principally for pigment extraction) was received from “Estación Experimental Las Palmerillas”, Almeria, Spain. Initial water content was approximately 90%. Some samples were treated and extracted on the arrival date, the others were kept in the freezer and later defrosted prior to the microwave treatment, followed by extraction. As microalgae deteriorate rapidly after harvesting, only a few experiments were performed with fresh microalgae. The rest of the raw material was frozen to avoid loss of Chapter 2 81 active compounds. To ensure that storage of samples does not alter the extraction behaviour of the raw material, some conventional extraction experiments were performed with fresh and defrosted algae, and results were compared. 2.2. Moisture content determination The moisture content of the microalgae was determined before and after microwave treatments. Samples were dried at 105°C for 24 hours. Moisture content was calculated from the wet and dry masses. 2.3. Experimental scheme Experiments were carried out in two series, referred to as “Experiment I” and “Experiment II” (Figure 1). In both experimental series, the microwave treatment was carried out on centrifuged microalgae paste in order to take advantage of the microwave energy for the heating of the intracellular water. In “Experiment I” Nannochloropsis gaditana was used for lipid extraction. The microwave pre-treatment was carried out in a domestic microwave oven. Next, samples were dried at 60 °C for further extraction. Hexane was used as the extraction solvent in a Soxhlet apparatus, where samples were taken from the boiling liquid in order to measure the kinetics of the extraction. One sample (the Control) did not suffer microwave pre-treatment, but followed the same drying and extraction process in order to compare the effect of the pre-treatment. Extracts were analysed using gravimetry and spectrophotometry. Cell surface changes in microalgae were observed using scanning electron microscopy. Chapter 2 82 Figure 1 Experimental procedures for “Experiment I” and “Experiment II”. In “Experiment II” Scenedesmus almeriensis was used for pigment extraction. The microwave pre-treatment was performed in a laboratory microwave device (CEM Discover ONE). The wet biomass was extracted with hexane in continuously stirred extraction vessels at 25ºC. In continuation, solid, aqueous and organic phases were separated by centrifugation. The organic phase was analysed by gravimetry and spectrophotometry, and fatty acid composition was determined by gas chromatography. Detailed conditions of the experimental steps are presented below. Chapter 2 83 2.3.1. Microwave treatments 2.3.1.1. Domestic microwave oven “Experiment I” was performed in a modified domestic microwave oven (Figure 2) (Panasonic NNGD-566M; oven cavity: 359 x 352 x 217 mm) with a maximum output power of 1000 W. In order to promote cell wall rupture, experiments were performed at ambient pressure. In each experiment, 60 g of wet algae paste was divided into three polycarbonate tubes (Nalgene) with a diameter of 38 mm. The tubes were connected to a general collector, allowing the produced vapour produced to be condensed as it exited the system. After exposing the microalgae to microwave radiation, evaporated and condensed water were weighed and refilled into the tubes to maintain the TS (total solid) content in the samples. Temperature was recorded before and after MW pre-treatment. Experiments were carried out an emitted energy level of at 0.5 - 1.0 - 1.5 kJ/ml to obtain maximum treatment temperatures of 65, 80 and 88 °C, respectively. Power W Time min:sec 0I 2 3 4 5 6 7 1 5 Figure 2 Domestic microwave oven used in “Experiment I”: 1 – microwave oven; 2 – tube welded onto microwave cavity; 3 – turning carousel with vapor collection (r = 160 mm); 4 – Polycarbonate tubes (85 ml, d= 38 mm); 5 – silicon tube for vapor exit; 6 – condenser Chapter 2 84 2.3.1.2. CEM Discovery laboratory equipment “Experiment II” was carried out in a laboratory microwave equipment (CEM® Discover ONE) with a maximum output power of 300 W, although only 150 W was applied. 50 g of wet Scendesmus a. was placed in a 100 ml round-bottomed flask, open to the atmosphere. The homogeneity of microwave irradiation is questionable in any volumetric device; therefore, the algal biomass was continuously stirred with a mechanical stirrer at 50 rpm. Although the device was equipped with an infrared thermometer, the temperature inside the flask was measured with a fibre optic temperature sensor both before and after the treatment. Experiments were carried out at an emitted energy level of 0.18 - 0.24 - 0.30 kJ/ml, where final temperatures of 47, 68 and 86 °C were obtained, respectively. 2.3.1.3. Determination of absorbed microwave energy The nominal emitted energy in microwave processing can be easily calculated as the product of the treatment time by the applied power. However, this emitted energy is not fully absorbed by the material. Moreover, the use of different devices or just a different geometry of the experimental setup also affects the fraction of energy which is absorbed by the sample. Since only the absorbed energy can produce an effect in the sample, it is necessary to use the absorbed instead of the emitted energy as the variable to evaluate the performance of the process. The energy absorbed by microalgae (Eabsorbed [kJ/ml]) has been estimated with an energy balance of three contributers (Eq. 1): Eabsorbed ≈ Qsensible + Qlatent + Qloss (1) The sensible heat contribution (Qsensible) was calculated from the increment in temperature before and after microwave treatment, and the latent heat contribution (Qlatent) from the weight of evaporated water. The heat loss from the vessel surface (Qloss) was evaluated through the heat transfer coefficient of the system. This coefficient was obtained by correlating the decrease in temperature over time of a hot sample placed into the microwave cavity without microwave irradiation. Detailed determination of the absorbed energy has been explained elsewhere (Sólyom et al., 2011). In the case of the laboratory equipment (Experiment II) this heat loss from the vessel surface was negligible due to the lack of free space inside the cavity. Chapter 2 85 The relation between the absorbed and emitted energy gives the efficiency of the experimental device (Table 1). Table 1 Calculation of Microwave Absorption Efficiency and Emitted and Absorbed energies from experimental final heating temperature and evaporated water, in domestic and laboratory microwave setups. Temperature Evaporated water E emitted E absrobred Efficiency °C wt% kJ/ml kJ/ml % domestic MW 65 0.17 0.5 0.22 44 80 1.83 1.0 0.41 41 88 5.83 1.5 0.78 52 CEM® Discover 47 0.5 0.180 0.151 84 68 0.4 0.240 0.234 97 86 0.0 0.300 0.297 99 2.3.2. Extraction methods 2.3.2.1. Soxhlet extraction Soxhlet extraction was performed with n-hexane in “Experiment I” to obtain the lipid fraction from Nannochloropsis g. This method was chosen because it has been frequently used as a reference technique in lipid extraction (Balasubramaniam et al., 2011, Halim et al., 2011). Microalgae provided by the microwave treatment was dried at 60ºC for 24 h prior to solvent extraction, in order to avoid immiscibility problems because of polarity difference between solvents. A Soxhlet extractor (250 ml) was used with a ratio of 60 ml hexane/gram of dry material, and 6.5 reflux/h for 4 hours. Samples were taken from the boiling flask at different extraction times, using a 1 ml syringe, in order to follow the extraction kinetics in real time. 2.3.2.2. Continuous stirred extraction Scendesmus a. was extracted directly after the microwave treatment in its wet form. Despite the very slow mass transfer between the aqueous and organic phase, this technique may be interesting on an industrial scale as the high cost of drying, prior to the extraction, can be saved. Nevertheless, the extraction process may be enhanced to achieve better contact between the two phases; in this study, it was found to be sufficient for the evaluation of the microwave pre-treatment effect. Extraction was performed at 25 °C in a 100 ml closed bottle, without previous drying. 60 ml of hexane was applied to 5 g wet Chapter 2 86 biomass, and continuously stirred using an orbital shaker at 350 rpm. Centrifuge was used for phase separation (10 000 rpm, 10 min) after the extraction, and the organic phase was further analysed. Two different sources of Scendesmus were used: 1) Fresh microalgae. It was first pretreated (0.18 - 0.24 - 0.30 kJ/ml emitted energy level) and then extracted in duplicates. Also a control sample, without microwave pre-treatment, was used for extraction. Solvent samples of 1 ml were taken from the organic phase over 24 hours, and analysed by spectrophotometry. Only the final extract was analysed by gravimetry. 2) Frozen microalgae. It was defrosted and pretreated with microwave at 0.30 kJ/ml emitted energy level. Duplicates with control (defrosted, untreated sample) were extracted over 6-12-24 hours, and all of them were analysed by gravimetry and spectrophotometry. The fatty acid composition was determined from the accumulated extract. 2.3.3. Analytical procedures 2.3.3.1. Gravimetric analysis The extracted material was quantified after solvent evaporation (Büchi Rotavapor, 40 °C, 50 rpm). The dried extract was later re-dissolved in hexane and used for spectrophotometry calibration. 2.3.3.2. Spectrophotometry In the case of “Experiment I”, the extracts were measured at the maximum absorption wavelength (498 nm) to follow colour evolution during extraction (Hitachi U-2000). In the series of “Experiment II”, a wavelength scan was performed (Shimadzu UV-Vis Spectrophotometer) between 200 and 800 nm to detect peaks of different pigments. Chapter 2 87 2.3.3.3. Environmental Scanning electron microscope (ESEM) analysis The effect caused by microwave pre-treatment on extraction kinetics gives an indirect response of the cell disruption produced by microwaves. In addition, Environmental Scanning Electron Microscopy (ESEM-Quanta 200-F) of untreated algae and algae exposed to microwaves was used to obtain a qualitative impression of this phenomenon in “Experiment I”. 2.3.3.4. Fatty acid composition – Gas chromatography Fatty acids contained in the cumulative extract from the control and pre-treated Scendesmus a. samples were analysed in Experiment II by the CSIC Instituto de la Grasa, Sevilla (Spain). Component identification was performed by monitoring retention time coincidence with the following standard components: C16:0 (palmitic acid), C16:1 (palmitoleic acid), C18:0 (stearic acid), C18:1 (oleic acid), C18:2 (linoleic acid), C18:3 (linolenic acid), C18:4, C20:0 (arachidonic acid), C20:1 (eicosenic acid), C20:2 (n-6), C20:4 (n-6), C20:5 (n-3) (EPA). 3. RESULTS AND DISCUSSION 3.1. Experiment I – Nannochloropsis gaditana 3.1.1. Extraction kinetics The extraction kinetic curves shown in Figure 3 were obtained from absorbance measurements, correlated to the gravimetric analysis. Extraction yield was expressed in terms of g extract/g dry matter. The evolution of the accumulated yield (Y) can be described by first order kinetics to quantitatively compare different extractions (Eq.2). Y= Ymax[1 – exp (-βt)] (2) In Equation (2), extraction kinetics is described by two parameters: the pre-exponential parameter (Y max), which represents the maximum yield reached at infinite extraction time, and the initial extraction rate (𝑑𝑌 𝑑𝑡|𝑡=0), which is calculated as the product of the maximum Chapter 2 94 responsible for the health benefits in the nutritional usage of microalgae oil (Balasubramanian et al., 2011). 4. CONCLUSIONS The positive effects of solvent free microwave pre-treatment favoured extraction of dry algae (SFMP). Although a positive influence was demonstrated by the spectrophotometric analyses, the slow mass transfer in liquid-liquid extraction of wet algae did not permit the proper study of pretreatment effects. ESEM pictures showed damaged cell surface and a reduction in the number of large cells, indicating the successful impact of the MW. Thermosensitive fatty acid compounds did not suffer degradation during the short high temperature SFMP. In general, pre-treatments enhanced the processes studied, however, their efficiency was found to be process and product dependent. Chapter 2 95 ACKNOWLEDGEMENTS This work was supported by the Spanish Ministry of Economy and Competitiveness for the Project CTQ2010-15475 and the project of ENE2012-33613. The authors wish to thank Raul Muñoz, Javier Pereda and Miguel Ángel Álvarez for their research contribution. Katalin Sólyom thanks the University of Valladolid for the financial support provided by the “FPI UVa” scholarship. REFERENCES « Balasubramanian, S., Allen, J.D., Kanitkar, A., Boldor, D. 2011. Oil extraction from Scenedesmus obliquus using a continuous microwave system - design, optimization, and quality characterization. Bioresource Technol. 102(3), 3396-3403. « Barnard, T.M., Leadbeater, N.E., Boucher, M.B., Stencel, L.M., Wilhite, B.A. 2007. Continuous-Flow Preparation of Biodiesel Using Microwave Heating. Energ. Fuel. 21(3), 1777-1781. « Biller, P., Friedman, C., Ross, A.B. 2013. Hydrothermal microwave processing of microalgae as a pre-treatment and extraction technique for bio-fuels and bio-products. Bioresource Technol. 136, 188-195. « Cerón, M.C., Campos, I., Sánchez, J. F., Acién, F. G., Molina, E., Fernández-Sevilla, J. M. 2008. Recovery of Lutein from Microalgae Biomass: Development of a Process for Scenedesmus almeriensis Biomass. J. Agric. Food Chem. 56, 11761–11766. « Chisti Y. 2007. Biodiesel from microalgae. Biotechnol. Adv. 25, 294-306. « Cravotto, G., Boffa, L., Mantegna, S., Perego, P., Avogadro, M., Cintas, P. 2008. Improved extraction of vegetable oils under high-intensity ultrasound and/or microwaves. Ultrason. Sonochem. 15, 898–902. « Cvetković, D., Marković, D. 2008. Stability of carotenoids toward UV-irradiation in hexane solution. J. Serb. Chem. Soc. 73(1), 15–27. « Guedes, A.C., Amaro, H.M., Malcata, F.X. 2011. Microalgae as Sources of Carotenoids. Mar. Drugs. 9, 625-644. « Halim, R., Danquah, M.K., Webley, P.A. 2012. Extraction of oil from microalgae for biodiesel production: A review. Biotechnol. Adv. 30(3), 709-732. « Halim, R., Gladman, B., Danquah, M.K., Webley, P.A. 2011. Oil extraction from microalgae for biodiesel production. Bioresource Technol. 102, 178-185. Chapter 2 96 « Iqbal, J., Theegala, C. 2013. Microwave assisted lipid extraction from microalgae using biodiesel as co-solvent. Algal Research. 2(1), 34-42. « Koberg, M., Cohen, M., Ben-Amotz, A., Gedanken, A. 2011. Bio-diesel production directly from the microalgae biomass of Nannochloropsis by microwave and ultrasound radiation. Bioresource Technol. 102, 4265-4269. « Lee, J.-Y., Yoo, C., Jun, S.-Y., Ahn, C.-Y., Oh, H.-M. 2010. Comparison of several methods for effective lipid extraction from microalgae. Bioresource Technol. 101, S75-S77. « Lichtentaler, K.H, Buschman, C. 2001. Chlorophylls and Carotenoids: Measurement and Characterization by UV-VIS Spectroscopy. Current Protocols in Food Analytical Chemistry. F4.3.1-F4.3.8 « Lyons R. W., Hatcher J.D. 1972. Drying of a porous medium with internal heat generation. Int. J. Heat Mass Transfer. 15, 897-905. « Mata, T.M., Martins, A.A., Caetano, N.S. 2010. Microalgae for biodiesel production and other applications: A review. Renew. Sust. Energ. Rev. 14(1), 217-232. « McMillan, J. R., Watson, I. A., Ali, M., Jaafar, W. 2013. Evaluation and comparison of algal cell disruption methods: Microwave, waterbath, blender, ultrasonic and laser treatment. Appl. Energ. 103, 128-134. « Minguez-Mosquera, M.I., Gandul-Rojas, B., Gallardo-Guerrero, M. L. 1992. Rapid Method of Quantification of Chlorophylls and Carotenoids in Virgin Olive Oil by High-Performance Liquid Chromatography. J. Agric. Food Chem. 40, 60-63. « Navarrete, S. Wallraf, R. B. Mato, and M. J. Cocero, 2011, Improvement of Essential Oil Steam Distillation by Microwave Pretreatment. Ind. Eng. Chem. Res. 50, 4667– 4671 « Pasquet, V., Chérouvrier, J.-R., Farhat, F., Thiéry, V., Piot, J.-M., Bérard, J.-B., Kaas, R., Serive, B., Patrice, T., Cadoret, J.-P., Picot, L. 2011. Study on the microalgal pigments extraction process: Performance of microwave assisted extraction. Process Biochem. 46(1), 59-67. « Patil, P. D., Reddy, H., Muppaneni, T., Ponnusamy, S., Cooke, P., Schuab, T., Deng, S. 2012. Microwave-mediated non-catalytic transesterification of algal biomass under supercritical ethanol conditions. J. Supercrit. Fluids. http://dx.doi.org/10.1016/j.supflu.2012.11.023 Chapter 2 97 « Patil, P.D., Gude, V.G., Mannarswamy, A., Cooke, P., Munson-McGee, S., Nirmalakhandan, N., Lammers, P., Deng, S. 2011. Optimization of microwaveassisted transesterification of dry algal biomass using response surface methodology. Bioresource Technol. 102(2), 1399-1405. « Pfromm, P. H., Amanor-Boadu, V., Nelson, R. 2011. Sustainability of algae derived biodiesel: A mass balance approach. Bioresource Technol. 102, 1185–1193. « Sólyom, K., Mato, R.B., Pérez-Elvira, S.I., Cocero, M.J. 2011. The influence of the energy absorbed from microwave pretreatment on biogas production from secondary wastewater sludge. Bioresource Technol. 102, 10849–10854. « Sorguven, E., Özilgen, M. 2010. Thermodynamic assessment of algal biodiesel utilization. Renew. Energ. 35, 1956-1966. « Šoštarič, M., Klinar, D., Bricelj, M., Golob, J., Berovič, M., Likozar, B. 2012. Growth, lipid extraction and thermal degradation of the microalga Chlorella vulgaris. New Biotechnol. 29(3), 325-331. « Wiltshire, K. H., Boersma, M., Möller, A., Buhtz, H. 2000. Extraction of pigments and fatty acids from the green alga Scenedesmus obliquus (Chlorophyceae). Aquat. Ecol. 34, 119-126. « Zang, L. Y., Sommerburg, O., Van Kuijk, F. J. G. M. 1997. Absorbance Changes of Carotenoids in Different Solvents. Free Radical Bio. Med. 23(7), 1086–1089. Chapter 3. Dielectric properties of grape marc: Effect of temperature, moisture content and sample preparation method Sólyom, K., Kraus, S., Mato, R.B., Volker, G., Schuchmann, H.P., Cocero, M.J. Journal of Food Engineering 119 (2013) 33-39. Chapter 3 101 DIELECTRIC PROPERTIES OF GRAPE MARC: EFFECT OF TEMPERATURE, MOISTURE CONTENT AND SAMPLE PREPARATION METHOD Abstract Microwave assisted extraction enhances antioxidant extraction kinetics in grape marc processing. For an effective microwave process design, the dielectric properties (DP) of the material must be known under different conditions. This study focused on the DP measurement of grape marc using a resonant cavity measuring device. Grape skin, seeds and their mixtures with three different moisture contents were studied at 28, 39 and 50 ºC. At higher moisture content DP showed increasing tendency, while temperature didn’t have a definite effect. Good correlation was achieved with the experimental results applying the complex refractive index mixing equation between 0.4‐0.6 porosity. Also, despite the complexity of natural products properties, the experimental DP values of the real mixture were in agreement with values estimated from DP properties of its individual constituents (grape skin and seeds). Milling was performed to obtain homogeneous material for the measurements, and this effect on the results was also studied. Chapter 3 102 1. INTRODUCTION There is an increasing interest in antioxidant extraction from industrial by-products such as the grape marc obtained from red wine production (Spigno and De Faveri 2007). The conventional extraction processes can be enhanced with novel techniques such as the application of microwaves, where the dipole molecules are mostly affected by the oscillation of the dielectric field. Microwave assisted extraction processes may lead to increased extraction yields in short extraction times while using a smaller amount of solvent (Proestos and Komaitis 2008). However, the efficiency of any microwave process depends on the correct design of the microwave applicator. For this reason, the interaction between the treated material and the applied electric field must be known. In order to describe this interaction, the complex dielectric permittivity (Equation (1), (2)) of the material should be known (Metaxas and Meredith 1988). 𝜀=𝜀0 𝜀𝑟 (1) 𝜀𝑟=𝜀𝑟′+𝑗𝜀𝑟" (2) In Equation (1) 𝜀0 and 𝜀𝑟 refer to the free space permittivity and the complex dielectric constant, respectively. Equation (2) shows the real part of the complex dielectric constant (𝜀𝑟′), which corresponds to the energy stored in the material when an electric field is applied, and the imaginary part (𝜀𝑟"), corresponding to the energy which is converted into heat. During the microwave process, the material undergoes physical and perhaps chemical changes (moisture content, temperature and salt concentration) which would affect its interaction with the dielectric field. These factors must be taken into account when characterizing the dielectric properties of the sample (Venkatesh and Raghavan 2004). In former studies the grapes’ dielectric constants were measured for microwave drying purposes (Dev et al. 2009, Tualsidas et al. 1995). In those measurements whole grapes or grape skin, excluding seeds, were measured in an open ended coaxial line measurement setup, requiring perfect contact between the probe and the sample. The same method was used for fresh fruit and vegetable slices, where non-monotonic dielectric behaviour was reported above 65 ºC, corresponding to the cell breakage of plant tissues (Nigmatullin and Nelson 2006). The dielectric properties of grape juice were studied at different frequencies (García et al. 2001), also with an open-ended coaxial line probe. The conductivity of the sample was also Chapter 3 103 considered, as it strongly affected the measurement results at higher frequencies. When measuring wine (García et al. 2004), it was reportedly difficult to establish perfect contact between the probe and the liquid sample because of the presence of disturbing CO2 bubbles in the fermented wine. In the afore-mentioned literature all measurements were done with seedless grape berries. The inclusion of grape seeds affects the microscopic and macroscopic content of the raw material, and so dielectric properties may change. The dielectric properties of a mixture of grape seeds and skin had apparently not been measured in former studies. When complex semi solid material is measured, open ended coaxial line probes present two main problems. On one hand, it is difficult to establish the necessary physical contact between the sample and the probe; on the other hand, the results may rely on the position and distance of the different constituent parts in the inhomogeneous material. For such samples, the resonant cavity method may be more suitable, if measurements are performed at a constant frequency (Nelson 1991), because this measurement method reflects the overall dielectric constant of the complex sample placed into the cavity. Otherwise, the global dielectric properties of a multiphase material can also be estimated from the proportions and properties of their individual constituent phases by using mixing equations. Some of the previously reported correlations are derived from Maxwell’s equations, while others are semi empirical models (Sheen et al. 2010, Simpkin 2012, Nelson 2005, Erle et al. 2000). The wide variety of equations used for different materials confirms that none of them gives the exclusive solution to the estimation of the dielectric properties of a complex mixture, and in any case, the best fit should be chosen depending on the material type, shape and complexity. Sheen et al. (2010) compared exponential and logarithmic models, estimated their theoretical error and correlated them with experimental data. They suggested that the complex refractive model and the random model were the most suitable for a polymer-ceramic mixture at a high volume fraction of the dispersed phase. These models are also widely mentioned in case of agricultural and food products (Venkatesh and Raghavan 2004). Therefore, the Complex Refractive Index Mixing model (CRIM, Equation 3), the Random model (LandauLifshitz-Looyenga, Equation 4) and a Logarithmic model (Lichtenecker-Rother , Equation 5) were used in this study to estimate the material dielectric properties of an air-particle mixture as a function of the sample porosity: √𝜀𝑚=𝜈1√𝜀1+𝜈2√𝜀2 (3) Chapter 3 110 In the case of the fresh grape skin, when converted values to 0.5 porosity are compared, both a higher dielectric constant and loss were observed for the whole skin (𝜀′ = 12.63, 𝜀” = 2.131) than for the milled product (𝜀’ = 10.83, 𝜀” = 1.778). However, the opposite effect was expected because of the higher moisture content of the milled skin. This effect was even more pronounced with the dry material, which is again in contrast to the expectations. In experiments with dry skin the porosity of the sample (0.8877 and 0.6704) was considerably higher than 0.5 due to the fragility of the dry material (Table 2). When these values were adapted to 0.5 porosity with the CRIM model, no conclusive results were obtained. This behaviour may be explained by the reported lack of suitability of the CRIM equation for high porosity mixtures. (Sheen et al. 2010) In this range, above 0.7 porosity, the same authors proposed the use of a linear approximation (Equation 8) 𝜀𝑚=𝑣1𝜀1+𝑣2𝜀2 (8) The experimental values of dry grape skin were adapted to 0.5 porosity also using the Linear model (Table 2). With this model, a slight but significant increment of dielectric constant can be observed for milled skin (𝜀′= 1.76) when compared to whole skin (𝜀′= 1.45). In the case of dielectric loss the value was almost doubled (𝜀"= 0.064 for milled, and 𝜀"= 0.027 for whole skin). Obviously, it is hard to obtain reliable results in these conditions and extrapolate them to the pure material or to a porosity of 0.5. Besides porosity, differences in sample shape could also explain the discrepancies between the results. (Jones and Friedman 2000, Di Biasio and Cametti. 2007.) Despite the non-equivalent results, it was expected that a similar tendency in dielectric property behaviour at different conditions of moisture content and temperature would be obtained through further experimentation (Section 3.3). 3.3. The effect of moisture and temperature on the milled mixture, skin and seeds A clear influence of the moisture content on the dielectric constant and dielectric loss was noted, such that higher moistures lead to higher dielectric constant and loss for all types of studied materials at any temperature (Figure 2). As all studied materials had relatively high moisture content, the observed phenomena was in accordance with the rule: the more free water is in the material, the higher the microwave absorption and energy dissipation will be (Metaxas and Meredtih 1988). Chapter 3 111 There was not such clear evidence regarding the effect of temperature. In the case of the dielectric constant, the values were seen to increase slightly with increasing temperature at every moisture content. However, a clear decrease in this property has been described in literature (Dev et al. 2009, Tulasidas et al. 1995), as happens with water. This discrepancy in behaviour may be explained by the presence of potassium metabisulfite (0.1 wt %) added to the raw material during the industrial processing of grape marc. Moreover, the increasing value of dielectric properties with temperature at high moisture content may also be explained by the complex water – oligosaccharide secondary bonds in the natural matrix. Sipahioglu and Barringer (2003) found similar behaviour for garlic samples with 57.3 % moisture content, where the increasing tendency was due to the high inulin content of the samples. In the case of the dielectric loss of the milled seeds at the highest moisture content, a slight decrease in behaviour was observed when temperature was increased, as is expected when a high amount of free water is present. This may be explained by the lower salt content in the seeds, as the whole seed is a closed structure, in which salt is only found on the outer layer of the seed. The inner part of the seeds remains intact due to the protective function guarding the possible future reproduction of the grape plant. At lower moisture content, the dielectric loss increased with temperature. Although the former interpretation of the results is hypothetical, neither the clear water-like behaviour nor the obvious salt effects were observed through the experimental data. Chapter 3 112 Figure 2 Dielectric constant (a) and dielectric loss (b) of milled grape marc (seeds and skin mixture), dielectric constant (c) and dielectric loss (d) of milled skin, dielectric constant € and dielectric loss (f) of milled seeds at ν= 0.5. Chapter 3 113 All these results were fitted with a quadratic regression model (Equation 9) to determine the complex dielectric properties at any moisture content (Mc [%]) and temperature (T [ºC]) in the measured range. The general model takes into account the linear, quadratic and crossing dependence on moisture and temperature through the corresponding coefficients for temperature (𝑘𝑇), second order temperature effect (𝑘𝑇2), joint temperature and moisture effect (𝑘𝑇𝑀𝑐), moisture (𝑘𝑀𝑐), and second order moisture effect (𝑘𝑀𝑐2), respectively. 𝜀=𝜀0+𝑘𝑇𝑇+𝑘𝑇2𝑇2+𝑘𝑇𝑀𝑐𝑇𝑀𝑐+𝑘𝑀𝑐𝑀𝑐+𝑘𝑀𝑐2𝑀𝑐2 (9) Correlation coefficients are given in Table 3 for grape skin, seeds and their mixture. Reported values are different from zero at a significance level of P<0.05. Table 3 Quadratic model coefficients (Eq. 9) Both the temperature and moisture content quadratic effects are negligible. The crossed effect of temperature and moisture is not clear either, as already reported in literature for a microwave drying process (Tualsidas et al. 1995). These results are in agreement with previous discussion. 3.4. Calculation of dielectric properties of the milled mixture from milled skin and seeds To test the established measurement procedure, the following calculations were performed, and results were compared: 1) The dielectric properties of milled grape marc (seeds and skin mixture) were measured at a porosity of 0.4-0.6. These values were extrapolated with CRIM equation to zero porosity, and are shown in Figure 3, with the reference “Experimental”. 2) The same procedure was used to estimate zero porosity values for separated grape skin and seed samples. From these values, and the known ratio of skin (58 wt%) and seeds (42 wt%) in Chapter 3 114 the grape marc, the CRIM equation was used at zero porosity to estimate the dielectric properties of the mixture by combining the contributions of seeds and skin (with the reference “Estimated” in Figure 3). The dielectric property values of grape marc at zero porosity obtained by procedures 1) and 2) are compared in Figure 3 as a function of moisture content for three different temperatures. Dielectric constant and loss both increase with increasing moisture content, as expected. In the case of lower moisture content the estimation from CRIM for seeds and skin was too great, while at higher moisture the same calculation underestimated the “experimental” values. The percentage standard deviation between ‘experimental’ and ‘estimated’ values was 17%. This level of precision was similar to that previously obtained when correlating milled gra marc as a function of porosity with the CRIM equation (Table 1), which was considered acceptable for this natural by-product material as a general purpose model. The agreement between both procedures may assure that the obtained values are close to the dielectric properties of the grape marc itself despite of the complex difficulties in the measurement method. Chapter 3 115 Figure 3 Comparison of dielectric properties of milled grape marc at zero porosity calculated from the extrapolation of grape marc experimental values (‘Experimental’), and from the estimation with the CRIM equation from seeds and skin constituents (‘Estimated’). a) T=28 ºC b) T=39 ºC c) T=50 ºC Chapter 3 116 4. CONCLUSIONS Complex dielectric properties of grape marc and its separated macroscopic constituents (grape skin and seeds) were measured as a function of temperature (28-50 ºC) and moisture content (8.6 – 65.9 wt%). In previous studies, fresh, seedless grapes were measured using the coaxial line method, which requires perfect physical contact between the probe and the material. Here, the cavity perturbation method was used for the semi solid material as an “air-particle” mixture, after a milling step for sample preparation. In order to obtain the dielectric properties of the material itself (air-free), different porosities were studied and mixing equation models were fitted to the results in order to obtain the dielectric properties at zero porosity. The complex refractive index mixing equation model (CRIM) was chosen because it provided the best fit between 0.4-0.6 porosity. The study of the influence of temperature and moisture was performed in this range, and experimental values were normalized to a porosity of 0.5 with the CRIM model. This same mixing equation was used to calculate dielectric properties of grape marc from the experimental values of their individual constituents, seeds and skin, confirming the reliability of the measurement method (percentage standard deviation ≈ 17%). In addition, correlation equations were obtained in order to calculate the dielectric properties of grape marc, seeds and skin as a function of moisture and temperature. As expected, moisture content had a significantly increasing effect on the dielectric properties for all samples. However, temperature was not shown to have a clear influence on either the dielectric constant or on the dielectric loss. This behaviour was attributed to the presence of salt in the grape marc. The so obtained results will provide useful information in the design of a microwave cavity in order to maximize the treatment efficiency of grape marc in a microwave industrial extraction process. The use of reliable experimental values is essential when performing process scaling up from laboratory scale in order to consider the interaction between the complex material and the electromagnetic field. Chapter 3 117 ACKNOWLEDGEMENTS The authors thank the Spanish Ministry of Economy and Competitiveness for projects CTQ2010-15475 and FracBioFuel ENE2012-33613 for funding. Katalin Solyom thanks the University of Valladolid for the financial support of the ”Estancias breves FPI UVa” Scholarship, and the Institute of Process Engineering in Life Sciences Section I: Food Process Engineering, Karlsruhe Institute of Technology for their technical support. REFERENCES « Dev, S.R.S., Gariépy Y., Raghavan G.S.V. (2009) Measurement of Dielectric Properties and Finite Element Simulation of Microwave Pretreatment for Convective Drying of Grapes. Progress in Electromagnetic Research Online, 5(7), 690-695. « Di Biasio, A., Cametti, C. (2007) Effect of shape on the dielectric properties of biological cell suspensions. Bioelectrochemistry, 71,149–156. « Erle, U., Regier M., Persch C., Schubert H. (2000) Dielectric Properties of Emulsions and Suspensions: Mixture Equations and Measurement Comparison Journal of the Microwave Power and Electromagnetic Energy, 35(3), 185-190. « García, A., Torres J.L., de Blas M, De Francisco A., Illanes R. (2004) Dielectric Characteristics of Grape Juice and Wine. Biosystems Engineering, 88 (3), 343-349. « García, A., Torres J.L., Prieto L., de Blas M. (2001) Dielectric properties of grape juice at 0.2 and 3 GHz Journal of Food Engineering 48, 203-211. « Jones, S.B., Friedman, S. P. (2000) Particle shape effects on the effective permittivity of anisotropic or isotropic media consisting of aligned or randomly oriented ellipsoidal particles. Water Resources Research, 36(10), 2821–2833 « Kent M., Kress-Rogers E. (1986) Microwave moisture and density measurements in particulate solids. Transactions of the Institute of Measurement and Control, 8(3), 167– 168. « Metaxas A.C., Meredith R.J. (1988) Industrial Microwave Heating. Peter Peregrinus Ltd., London, United Kingdom. Chapter 3 118 « Navarrete, A., Mato R.B., Dimitrakis G., Lester E., Robinson J.R., Cocero M.J., Kingman S. (2011) Measurement and estimation of aromatic plant dielectric properties. Application to low moisture rosemary. Industrial Crops and Products, 33, 697–703. « Nelson, S. O. (1991) Dielectric Properties of Agricultural Products. IEEE Transaction Electrical Insulation, 26(5) 845-869. « Nelson, S.O. (2005) Density-Permittivity Relationships for Powdered and Granular Materials. IEEE Transaction on Instrument Measurement, 54(5), 2033-2040. « Nigmatullin, R.R., Nelson S.O. (2006) Recognition of the ‘‘fractional’’ kinetics in complex systems: Dielectric properties of fresh fruits and vegetables from 0.01 to 1.8GHz. Signal Processing, 86, 2744-2759. « Pereira, S.M., Rivas M. A., Mosteiro L., Legido J.L., Iglesias T.P. (2002) Relative permittivity increments for the binary mixture (methanol + polyethylene glycol dimethyl ether 250) at the temperatures from 283.15K to 323.15K. Journal of Chemical Thermodynamics, 34, 1751–1759. « Proestos C., Komaitis M. (2008) Application of microwave assisted extraction to the fast extraction of plant phenolic compounds. LWT Food Science and Technology Journal, 41, 652-659. « Sheen, J., Hong Z.W., Su C.W., Chen H.C. (2010) Microwave Measurements of Dielectric Constants by Exponential And Logarithmic Mixture Equations. Progress in Electromagnetic Research, 100, 13-26. « Simpkin R. (2012) Derivation of Lichtenecker’s Logarithmic Mixture Formula from Maxwell’s Equations. IEEE Tranaction of Microwave theory and Techniques, 58(3), 545550. « Sipahioglu, O., Barringer, S.A. (2003) Dielectric properties of vegetables and fruits as a function of temperature, ash, and moisture content. Journal of Food Science, 68(1), 234239. « Spigno G., De Faveri D. M. (2007). Antioxidants from grape stalks and marc: Influence of extraction procedure on yield, purity and antioxidant power of the extracts. Journal of Food Engineering, 78, 793-801. Chapter 3 119 « Tulasidas, T.N., Raghavan G.S.V., van de Voort F., Girard R. (1995) Dielectric Properties of Grapes and Sugar Solutions at 2.45 GHz. Journal of the Microwave Power and Electromagnetic Energy, 30(2), 117-123. « Venkatesh, M.S., Raghavan, G.S.V. (2004) An overview of microwave processing and dielectric properties of agri-food materials. Biosystems Engineering, 88(1), 1–18. Chapter 4 126 although claims of higher yields using super-heated liquid extraction has been cited elsewhere (Peralbo-Molina et al., 2012). The extraction of fungicide residues from grapes for analytical purposes permitted more sensitive and faster performance with MAE than with solid-liquid extraction, or with matrix solid-phase dispersion (Lagunas-Allué et al., 2012). The kinetics of the MAE process were described by a first order model in the case of the phenolic compounds of tea (Spigno and De Faveri 2009) and cocoa leaves (Chan et al., 2013), taking into account the absorbed microwave energy. On the other hand, anthocyanin extraction from grape peel was found to follow a Gauss function (Li et al., 2012b). The aim of this study was to evaluate the extraction kinetics of phenolic compounds, and especially of anthocyanins from grape residue produced by the wine industry using two different microwave processes: SFMP (solvent-free microwave pre-treatment, prior to extraction) and MAE (microwave assisted extraction). In the case of MAE, different levels of absorbed microwave energy and power were studied in order to evaluate their influence on the extraction rate constant and the final extraction yield of total phenoland monomeric anthocyanin content. Results were compared with a patented conventional extraction process (Moro-Gonzalez 2010). For industrial considerations 1) the raw material remained untreated prior to the modified extraction processes, 2) the energy absorbed was evaluated in the microwave processing, and 3) the cost of energy in the intensified process was calculated. 2. MATERIALS AND METHODS 2.1. Raw material A mixture of red grape seed and skin (grape marc) from the region of Toro (Spain) was used in the experiments. In the wine production, the grape marc was separated from the liquid phase after the red wine. 10 kg of grape marc was distributed in 100 g packages and stored at -18 °C until the time of the experiment. Defrosting was performed at 4 °C overnight, the day before the extractions. 2.2. Extraction techniques The effect of microwave irradiation on the extraction kinetics was studied using two methods. On one hand microwave assisted extraction was applied, in which the mixture of raw material and solvent was exposed to the microwaves. On the other hand, solvent-free microwave pretreatment was used to achieve cell wall rupture in the raw material in the absence of a solvent, Chapter 4 127 and further continue with a conventional stirred extraction. The results of both techniques were compared to a conventional patented method (Moro-Gonzalez 2010), following referred to from now on as the “control” 2.2.1. Microwave assisted extraction (MAE) MAE experiments were carried out in a piece of laboratory microwave equipment (CEM® Discover) with a maximum output power of 300W. 30 g of wet grape marc was placed in a 100 ml round-bottomed flask, open to the atmosphere. The material was gently stirred with a mechanical stirrer at 50 rpm to facilitate irradiation homogeneity in the material. Although the device is equipped with an infrared thermometer, the temperature inside the flask was measured with a fibre optic temperature sensor during the treatment (FoTemp 4, OPTOcon GmbH). Microwave irradiation was performed at constant power (80, 150 and 300 W). At every power level, three irradiation times were used (Table 1). These irradiation times were chosen in order to supply similar absorbed energies at the three power levels. Before microwave irradiation, 60 ml of 1:1 volumetric mixture of ethanol (96%) and acidified water (pH=1) was added to the grape marc as the extraction solvent. After stirring for 1 minute without microwave irradiation, the first sample of 1.5 ml was taken as an initial value for the kinetic analysis, and the initial temperature was recorded. After microwave irradiation, the final temperature was recorded again, and a second sample was taken for further analyses. After irradiation, stirring continued for 20 minutes of the total extraction time, with periodic sampling from the extraction vessel. During the whole extraction, a total of five samples were gained and filtered through 0.2 µm PET syringe filters for further analyses of the total phenol content and antioxidants. Every treatment was performed in duplicate from the same batch of grape marc. Table 1 Experimental conditions in MAE: power levels and irradiation times Power (W) 80 150 300 Irradiation time (s) 80 40 20 160 80 40 240 120 60 Chapter 4 128 2.2.2. Solvent-free microwave pre-treatment (SFMP) The experimental setup was prepared as in Section 2.2., with the difference that the solvent was added after instead of before the microwave treatment. Right after irradiation, samples were cooled down in a cold water bath, and placed into a bath to continue the extraction at 40 ºC, as described in Section 2.2.3. As a preliminary test, 150W of power was applied for 55 and 80 s on grape marc with an 85% moisture content. The effect of moisture content was also studied by pressing juice out of the grape marc before the treatment, in order to reduce the moisture content from 85% to 73%. Samples from the solvent free pre-treated extraction process were compared to “control” samples, without microwave pre-treatment, as described in section 2.2.3. Experimental error was obtained from duplicate experiments. 2.2.3. Conventional solidliquid extraction The same glass vessel with mechanical stirring was used, as described in the MAE experiments (but without microwave treatment) and was placed directly into a water bath at 40°C. The same sample and solvent amounts were also used, as in the MAE experiments. 1.5 ml samples were taken for kinetic analysis after 10-20-45-90-180 min and filtered through 0.2 µm PET syringe filters for further analyses of the total phenol content and antioxidants. Extractions were performed in duplicate. 2.3. Absorbed energy calculation In microwave processes, the product of the applied power multiplied by the irradiation time determines the energy emitted by the device. However, this amount of energy is not completely absorbed by the material. The fraction of energy absorbed by the sample will depend on the dielectric and geometric characteristics of the material, as well as on the type and geometry of the microwave oven. To take advantage of the results presented here, so that they can be transferred to other devices and setups at laboratory or industrial scale, it is necessary to estimate the fraction of energy absorbed by the grape marc and the solvent mixture. This magnitude has been calculated in terms of density of theabsorbed energy (Eabsorbed [kJ/ml]) from an energy balance with three contributors (Eq. 1) (Sólyom et al., 2011): Eabsorbed = Qsensible + Qlatent + Qloss (1) Chapter 4 129 At the beginning of the irradiation period, most of the energy is spent on heating the sample, as sensible heat (Qsensible). This term can be calculated from the mass of grape marc plus the solvent mixture (msample [g]), the temperature increment in the microwave treatment (ΔT [ºC]), and the specific heat capacity of the mixture of grape marc and solvent (Cp = 3.555 [kJ/(kg*K)]) (Eq. 2). The heat capacity of the grape marc at certain moisture content was estimated as described by Tulasidas (1994). The sensible heat term: Qsensible = msample Cp ΔT (2) When the temperature reaches the boiling point of the mixture, the energy is used up on strong evaporation as latent heat (Qlatent), which can be determined as the product of the weight of the evaporated solvent and the latent heat of its vaporization. The third element in the energy balance (Qloss) takes into account the heat loss from the vessel surface into the environment during the treatment, which can be determined through the heat transfer coefficient of the system, obtained by the decrease of temperature over time of a hot sample placed into the microwave cavity without microwave irradiation (Sólyom et al., 2011). In the present case, temperatures below the boiling point were achieved and no solvent losses were observed, therefore the evaluation of the latent heat was not performed. Due to the short irradiation time, and the lack of free volume inside the cavity, the heat loss from the vessel surface was negligible. This simplification was checked experimentally. Thus, Eq. (1) can be simplified as follows: Eabsorbed ≈ Qsensible= msample Cp ΔT (3) 2.4. Moisture content analysis The moisture content of the grape marc samples was determined before the microwave treatments in the case of every new opened 100 g pack. Samples were dried at 105°C for 24 hours. From the wet and dry masses the moisture content was calculated. Analyses were performed in duplicate for every sample. 2.5. Total phenol content analysis Total phenol content (TPC) was quantified, using the Folin-Ciocalteu method, as Gallic acid equivalents per gram of dry material [mg GA/g DM] (Singleton et al. 1999). A volume of 40 μl of filtered sample was diluted with distilled water (3 ml) and mixed with 200 μl of FolinCiocalteu reagent. After 10 minutes, saturated Na2CO3 was added to the solutions, and the Chapter 4 130 samples were incubated at 40ºC for 30 min. Absorbance was measured at 765 nm (UV 2550 Shimadzu UV/VIS spectrometer) and known concentrations of diluted Gallic acid were used for calibration. 2.6. Anthocyanin content analysis Total monomeric anthocyanin pigment content was determined by the pH differential method (AOAC 2005.02). Results are expressed on the basis of cyanidin-3-glucoside [mg cyanidin-3glucoside/ g dry matter]. 2.7. Extraction kinetics model and data analysis In several occasions the extraction was found to follow kinetics of first order (Spigno and De Faveri., 2009, Chan et al., 2013), which was also suggested in the present study. 𝑌(𝑡)= 𝑌 0+ 𝑌 𝑓(1 − exp⁡(−𝛽𝑡)) (4) In Eq. (4) Y(t) refers to the extraction yield of TPC (Y TPC; [mg GA/g DM]) or anthocyanins (YAC; [mg cyanidin-3-glucoside/ g DM]) at a certain time (t; [min]). The sum of the initial yield (Y0) and the pre-exponential constant (Yf) results in the maximal extraction yield (Ymax) at infinite time. The extraction rate constant (β [min-1]) can also be expressed separately for TPC (βTPC) or Anthocyanin content (βAC). The parameters of the extraction kinetics were calculated using Statgraphics Centurion XVI Software with the Gauss-Newton non-linear regression method. Standard deviation and P= 0.05 confidence limits were also determined. 3. RESULTS AND DISCUSSION 3.1. Absorbed energy evaluation and efficiency Irradiation times were chosen with the aim of supplying the same absorbed energy at different power levels. However, when the absorbed energies were evaluated according to the sensible heat transferred to the sample, small discrepancies were observed in some cases (Table 2). The three levels of energy absorbed by the samples were 18.4±0.7, 12.3±0.5 and 6.1±0.2 kJ, with absorption energy efficiencies related to emitted energies of 90%±2%, 90%±7% and 98%±2%, respectively. The lowest energy value (6.1±0.2 kJ) was completely absorbed by the system (98%±2%). At higher energy levels (>12 kJ), the energy absorption was lower (90%), but still much more efficient than in experimental setups with domestic microwave ovens (Sólyom et Chapter 4 131 al., 2011). The absorbed energy density (Edensity [kJ/ml]) can be calculated from the determined absorbed energy and the volumes of sample and solvent. Table 2 Absorbed energy evaluation in MAE experiments Power Time T final ΔT E emitted E absorbed E density Efficiency W sec °C °C kJ kJ kJ/ml % 80 240 72 57 19.2 17.2 0.201 90% 80 160 57 41.5 12.8 12.5 0.147 98% 80 80 35 21 6.4 6.3 0.074 99% 150 120 70 55 18.0 16.6 0.194 92% 150 80 49 35.5 12.0 10.7 0.125 89% 150 40 34 19.8 6.0 6.0 0.070 100% 300 60 67 52.5 18.0 15.9 0.186 88% 300 40 48 33.5 12.0 10.1 0.118 84% 300 20 34 19 6.0 5.7 0.067 96% Most literature on antioxidant extraction with the assistance of microwave energy deals with various experimental conditions, and also makes use of different microwave oven types, sizes and geometries. On a few occasions, attention has been given to the importance of power density (W/g or ml sample) or energy density (J/g or ml sample) (Huma et al., 2011 b, Li et al., 2012 b, Hiranvarachat et al., 2013, Chan et al., 2013). Nevertheless, the method used to quantify the absorbed energy has barely been described. Energy characterization was performed in the case of antioxidant extraction from dry cocoa leaves (Chan et al., 2013) and of β-carotene extraction from carrots (Hiranvarachat et al., 2013), but no determination has been found for the extraction from grape residues so far. 3.2. Preliminary study on SFMP Pre-treatments performed at 150 W without solvent were compared in terms of irradiation time and sample moisture content. No relevant differences were observed in any of the cases, neither in total phenol content nor in extraction kinetics, between the control and SFMP experiments (Figures 1a and 1b). In the analysis of irradiation time (Figure 1a) both experiments were compared to a control extraction, made on the same day from the same sample charge (“Control 1”, “Control 2”). The tendency observed for total phenol content was similar to that of anthocyanin content (data not shown). Likewise, no difference was observed when different moisture contents were compared with the corresponding control extractions after microwave irradiation (Figure 1b). A higher yield of total phenol content was obtained in the case of grape Chapter 4 132 marc with 85% moisture. This difference is explained by the loss of active compounds after pressing, which reduces the moisture in the raw material. As in the former case, the anthocyanin content showed similar evolution to the total phenol content. These unsuccessful results led to the discarding of SFMP as a possible intensification process, although for other raw materials positive results were obtained elsewhere (Michel et al., 2011, Huma et al., 2011 a, b). Figure 1. Total phenol content evolution of extract during conventional solid liquid extraction (control) and after SFMP treatments at 150 W: a) with irradiation times of 55 and 80 s, and b) with samples with different moisture contents (73 %, 85%). 1. a) 1. b) Chapter 4 133 3.3. First order kinetic model fit of experimental data 3.3.1. MAE experiments Parameters of the first order kinetics extraction for MAE and control experiments are listed in Table 3, including standard deviations and R2 coefficients. The average coefficients of determination (R2) were 95.8 and 94.2% for TPC and anthocyanin measurements respectively, showing good correlation with the experimental data. From Y0 and Yf values, the maximum yield of TPC (Ymax TPC = Y0+Yf) was calculated and represented at different power and absorbed energy levels in Figure 2. Increasing treatment energy at every power level caused increased extraction yields, however only the highest energy level (≈ 200 J/ml) rendered the same yield as the control samples. A significant influence of power cannot be perceived from experimental results, although the increment in yield with increasing treatment energy (the slope) is more pronounced at 80 W than at 300 W. This effect is probably due to the longer irradiation time required at lower power. Table 3. First order kinetic parameters for MAE and control extraction experiments. Power E density Yo TPC σ (Yo) Yf TPC σ (Yf) β TPC σ (β) R2Yo AC σ (Yo) Yf AC σ (Yf) β AC σ (β) R2 W kJ/ml % 80 0.201 1.75 0.38 7.75 0.37 0.181 0.024 96.3 9.73E-04 2.0E-04 1.85E-03 1.9E-04 0.418 0.076 91.4 80 0.147 2.89 0.32 5.32 0.31 0.186 0.032 97.8 1.57E-03 2.7E-04 1.66E-03 2.5E-04 0.216 0.088 87.5 80 0.074 2.54 0.29 3.10 0.28 0.233 0.053 88.7 1.46E-03 8.5E-05 1.01E-03 8.6E-05 0.184 0.046 89.1 150 0.194 1.84 0.59 7.04 0.56 0.220 0.048 91.2 1.06E-03 1.9E-04 2.05E-03 1.8E-04 0.357 0.061 92.6 150 0.125 1.28 0.41 5.84 0.38 0.341 0.043 98.1 7.63E-04 1.9E-04 2.16E-03 1.8E-04 0.402 0.059 97.3 150 0.070 2.89 0.26 3.72 0.25 0.256 0.044 97.5 1.63E-03 1.6E-04 1.43E-03 1.5E-04 0.276 0.069 94.8 300 0.186 0.68 0.44 6.50 0.41 0.310 0.042 98.3 3.98E-04 1.3E-04 1.98E-03 1.2E-04 0.560 0.051 99.2 300 0.118 1.42 0.61 5.26 0.57 0.370 0.073 96.2 9.23E-04 1.9E-04 1.71E-03 1.8E-04 0.459 0.075 97.3 300 0.067 2.42 0.26 3.57 0.24 0.272 0.043 97.9 1.16E-03 6.7E-05 1.21E-03 6.2E-05 0.337 0.035 99.1 Control - 1.97 0.77 7.96 0.36 0.057 0.010 94.5 9.23E-04 1.2E-03 2.27E-03 1.0E-04 0.090 0.018 89.6 Control - 1.97 0.77 6.72 0.44 0.113 0.031 75.7 9.23E-04 1.2E-03 2.01E-03 5.0E-05 0.164 0.023 88.7 min-1 min-1 mg GA/g DM mg GA/g DM mg CG/g DM mg CG/g DM Chapter 4 134 Figure 2. First order parameters (Y max: full bullets; β: empty bullets) of TPC at different power and energy levels in MAE experiments The other key parameter in extraction kinetics is the rate constant, β. The higher the value of β, the faster the extraction kinetics. In the case of TPC values, the rate constant seems to show a maximum value versus increasing energy. The analysis of total phenols provides a global view of possible valuable compounds. These phenolic compounds are located at different cell structures within the material, with different degrees of accessibility. A possible explanation of the observed maximum in the kinetics could be based on these differences in accessibility: with higher absorbed energy, the phenolics that are more difficult to extract become available, and the maximum yield is increased, but their kinetics of extraction are lower, reducing the global kinetics of the process. The use of higher power levels provided a faster extraction, as the temperature increment in the sample is faster at higher irradiation intensities. In the case of anthocyanin content (Figure 3) maximum extraction yields did not change significantly with increasing irradiation energy or power level. Compared to control samples, final yields are similar or even lower after MAE experiments. As regards the extraction rate constant of anthocyanins, higher values were obtained at higher energy levels. Similarly to the TPC, the higher was the applied power, the higher was the rate constant in the extraction. The energy amounts studied here are in accordance with those published for cocoa leaves (Chan et al., 2013), as the extraction was insufficient below100 J/ml and equilibrium was found between 100-300 J/ml. Higher energies were not studied in order to avoid high temperatures, excessive evaporation, and insufficient precision in the determination of absorbed energy. 0.0 0.2 0.4 0.6 0.8 1.0 0.0 2.0 4.0 6.0 8.0 10.0 0.00 0.05 0.10 0.15 0.20 0.25 β(min-1) Maximum TPC yield (mg GA/ g DM) Absorbed energy density (kJ/ml) 80 W 150 W 300 W Chapter 4 135 Figure 3. First order parameters (Y max: full bullets; β: empty bullets) of anthocyanins at different power and energy levels in MAE experiments The MAE experiments presented here were not performed in the usual manner, since here the shorter microwave irradiation stage was followed by an insulated stirred extraction stage at the temperature achieved during the irradiation. Using the first order kinetic model, the extraction yield obtained right after the irradiation pre-treatment, before stirred non-irradiated extraction, can be determined. Figure 4 shows the thus calculated relative yields for TPC and anthocyanin content (AC) at different power levels, corresponding at the highest energy level (200 J/ml). In all cases, the recovered percentage of Ymax was greater when lower power levels were used. Figure 4. Relative extraction yield right after microwave irradiation in MAE experiments (Edensity≈200 J/ml) The MW irradiation at a higher power level (300 W) causes the rapid heating of localized points in the experimental system, and the so-called hot spots are formed (Chan et al., 2011). When the irradiation stops, temperature inhomogeneity among the different points of the load is more likely to appear at higher power levels. During the following minutes of extraction, while the 0.0 0.2 0.4 0.6 0.8 1.0 0.000 0.001 0.002 0.003 0.004 0.00 0.05 0.10 0.15 0.20 0.25 β(min-1) Maximum AC yield (mg CG/ g DM) Absorbed energy density (kJ/ml) 80 W 150 W 300 W 0% 20% 40% 60% 80% 100% 80 W 150 W 300 W Extracted amount after MAE pre-treatment (% Ymax) TPC AC Chapter 4 142 « Song, J., Li, D., Liu, C., & Zhang, Y. (2011) Optimized microwave-assisted extraction of total phenolics (TP) from Ipomoea batatas leaves and its antioxidant activity. Innovative Food Science and Emerging Technologies, 12. 282–287. « Spigno, G., De Faveri, D.M. (2007) Antioxidants from grape stalks and marc: Influence of extraction procedure on yield, purity and antioxidant power of the extracts. Journal of Food Engineering. 78, 793 – 801. « Spigno, G., De Faveri, D.M. (2009) Microwave-assisted extraction of tea phenols: A phenomenological study. Journal of Food Engineering 93, 210–217. « Spigno, G., Tramelli, L., & De Faveri, D.M. (2007) Effects of extraction time, temperature and solvent on concentration and antioxidant activity of grape marc phenolics. Journal of Food Engineering. 81, 200 – 208. « Tulasidas, T. N. (1994) Combined convective and microwave drying of grapes. (PhD Thesis) Department of Agricultural Engineering, McGill University, Quebec, Canadá. p.79. « Wataniyakul, P., Pavasant, P., Goto, M., & Shotipruk, A. (2012) Microwave pretreatment of defatted rice bran for enhanced recovery of total phenolic compounds extracted by subcritical water Bioresource Technology, 124. 18–22. « Wu, T.,Yan, J., Liu, R., Marcone, M.F., Aisa, H.A., & Tsao, R. (2012) Optimization of microwave-assisted extraction of phenolics from potato and its downstream waste using orthogonal array design. Food Chemistry 133. 1292–1298. « Xiao, X.H., Wang, J.X., Wang, G., Wang, J.Y., & Li, G.K. (2009) Evaluation of vacuum microwave-assisted extraction technique for the extraction of antioxidants from plant samples. Journal of Chromatography A, 1216. 8867–8873. « Yang, Z., Zhai, W. (2010) Optimization of microwave-assisted extraction of anthocyanins from purple corn (Zea mays L.) cob and identification with HPLC–MS. Innovative Food Science and Emerging Technologies, 11. 470–476. Chapter 5. Enhanced polyphenol extraction of grape marc after ultrasound pre-treatment Chapter 5 145 ENHANCED POLYPHENOL EXTRACTION OF GRAPE MARC AFTER ULTRASOUND PRE-TREATMENT Abstract In this study ultrasound (US) was applied as a pretreatment, followed by conventional extraction of polyphenols, in order to enhance extraction kinetics using a lower energy input than other ultrasound assisted extraction processes, seeking for possible implementation on an industrial scale. The antioxidant extraction from grape marc, a byproduct of wine production, is constantly gaining more interest in food, pharmaceutical and cosmetic industry. Ultrasound assisted extraction is known to enhance the slow mass transfer stage, inherent in the conventional extraction processes. The influence of absorbed energy and energy intensity were evaluated. Increasing temperature (40-55-70°C) caused strong improvement, while energy intensity (100-75-50%) showed little influence on the final extraction yield and on the initial extraction rate. Up to 70% of the final yield was obtained solely through US pretreatment. Finally, only the thermal effect of US energy was confirmed by comparison of thermal and US pretreatments, and no clear evidence of a specific ultrasound effect could be verified. Chapter 5 146 1. INTRODUCTION Worldwide annual grape production is over 65 million tons, from which more than 25 million tons of wine is produced (FAOSTAT, 2010). During the wine-making process almost the same amount of semi-solid waste is generated, called grape marc, which contains the grapes’ skin, seeds, and the wine lees after the fermentation. During the fermentation process of red wine a mild extraction of the grape skin and seed occurs, thus improving the color and polyphenol profile of the wine. The remaining residue is still rich in antioxidants, which is desirable to eliminate in order to avoid inhibition problems before using the grape pomace as fertilizers (Negro et al. 2003, Pinelo et al. 2006). The extracted bioactive compounds mentioned (flavonoids, anthocyanins and stilbene derivatives) are attractive products for the food, cosmetic or pharmaceutical industry, due to the high antioxidant, and antimicrobial, antiviral or anticarcinogenic effects (Nassiri Asl and Hosseinzadeh 2009) . The widely used and studied conventional maceration or stirred extraction under mild conditions has the disadvantage of long processing times because of slow mass transfer, and requires huge solvent amounts (alcohols and water) which must be removed from the final product (Spigno and De Faveri 2007, Spigno et al. 2007, Amendola et al. 2010, Karacabey and Mazza 2010, Lapornik et al. 2005). Different sample pretreatments and novel extraction techniques were recently studied and compared in order to improve mass transfer and extraction kinetics of antioxidants, such as supercritical fluid extraction and microwave assisted extraction, Soxhlet-extraction or ultrasound assisted extraction (Martino et al. 2006, Pascual-Marti et al. 2001, Casazza et al. 2010, Rodriguez-Rojo et al. 2012, Peralbo-Molina et al. 2012). Positive effects of ultrasound assisted extraction (UAE) in food industry were reviewed by Vilkhu et al. (2008), who reports increased extraction yields and initial extraction rates. The propagation of ultrasound (US) waves leads to cavitation phenomena. Compression and stretch of the molecular spacing cause the formation and violent collapse of cavitation bubbles in the liquid, thus increasing the mass transfer, causing surface peeling, erosion and particle breakdown. The cell disruption or enlargement of cell wall pores due to the cavitation leads to the washing out of the intracellular content. Cell plasmolysis and alteration of different cellular layers were observed on US treated grape berries, using light and transmission electric Chapter 5 147 microscopy (Fava et al 2011). In US systems the electrical energy is converted into mechanical vibrations at a constant frequency, which is then transferred to the medium in an ultrasonic bath or using an ultrasonic probe. The intensity of the US energy is proportional to the square of the amplitude where there is a required minimum in order to achieve cavitations in the medium, although excessively high intensities may cause the degradation of the desired compounds. The majority of the dissipated acoustic energy is converted into heat. The consequently increased temperature has a positive effect on the extraction but at the same time causes a less violent collapse of the cavitation bubbles as they are filled with solvent vapor at higher temperatures. For this reason, most of the UAE applications are performed in a thermostatic environment. The nature of the solvent and treatment cycles (continuous or pulsed US) are also important factors during the process (Weiss et al. 2011, Santos et al.2009). These parameters were studied and optimized for several raw materials in order to achieve the desired features in the extracted product (Palma and Barroso 2002, Wang et al. 2008, Benito-Román et al. 2013, Londoño-Londoño et al. 2010, Morelli and Prado 2012). Extraction kinetics were found to fit first order behavior (Virot et al. 2010, Pingret et al. 2012, Karabegovic et al. 2011, Topallar and Gecgel 2000.) although Naik´s model (Carcel et al. 2010, Ahmad Quasem et al. 2012) and second order kinetics (Pan et al. 2011, Qu et al. 2010) were also proposed for the modelling of extraction processes. The efficiency of the UAE process is directly connected to the intensity of the acoustic energy supplied to the sample, which is usually designated as ‘specific energy’ (power/area of sonotrode). This specific energy corresponds to the dissipated acoustic power in the liquid, and must be distinguished from the power consumption of the US device. The latter can be measured by a high precision wattmeter, while the former is experimentally determined by calorimetry in an insulated vessel (Weiss et al. 2011, Romdhane et al. 1995, Ratoarinoro et al. 1995, Kuijpers et al. 2002). This specification of the energy input is indispensable in scaling up the UAE process (Virot et al. 2010, Pingret et al. 2012). The aim of this study was to evaluate an effective low energy input ultrasound pretreatment, prior to the conventional industrial process for the extraction of polyphenols. No additional sample preparation steps (drying or mechanical particle size reduction) were applied to the raw material, because of their high operating cost in industrial scale. The extraction kinetics of polyphenols from fermented grape marc were measured after short US pretreatments, using different absorbed energy levels. The so obtained kinetic parameters were compared to those Chapter 5 148 from conventional extraction without pretreatment (Control). The operating variables in the US pretreatments were the final pretreatment temperature (40-55-70ºC) and the US amplitude (50-75-100%). Also the possible non-thermal US effects were studied by the comparison of analogous thermal and US pretreatments, following the same thermal history. 2. MATERIALS AND METHODS 2.1. Raw material Fermented grape marc was received from the Matarromera winery (Spain) in November 2011. It was a mixture of pressed, partially ruptured grape skin and whole seed, which had contributed to the color development of the red wine during the fermentation process. For industrial reasons the samples had 1 g/kg grape marc NaHSO3 content. Samples were stored in a freezer at -20ºC and defrosted at 5ºC overnight before the pretreatments and extraction processes. The moisture content of the grape marc was determined in an oven at 105ºC after 24 hours drying. The original moisture content was around 60% for the milled skin and seed mixture. 2.2. Conventional extraction A continuous stirred extraction was performed for 3 hours in a closed round-bottomed flask containing 30 g wet grape marc in 150 ml solvent of 50 v/v% ethanol and acidified water at pH=1 (Moro Gonzalez 2010). Temperature (40ºC) was maintained using a water bath. During the extraction 0.5 ml samples were taken at 5, 15, 30, 60, 120, 180 min to study the extraction kinetics of polyphenols. The samples that did not suffer any treatment prior to the extraction will subsequently be referred to as “control”. Two parallel extraction setups were used for every treatment condition. 2.3. Ultrasound treatment and absorbed energy determination Ultrasound was performed before the conventional extraction procedure as a pretreatment stage. A probe type ultrasound system was applied (UP400S; Hielscher GmbH, Germany) with constant frequency (24 kHz), a titanium sonotrode (D=22mm), maximum power of 400W and maximum amplitude of 100μm. Pretreatments were carried out (30 g of wet grape marc and 150 ml solvent) in a double-walled glass vessel providing thermo insulation (air) which allowed the medium to heat up from room Chapter 5 149 temperature. The vessel was covered with aluminum foil to diminish the evaporation of the solvent, although the evaporated amount was measured by weight and replaced after the pretreatment. Additional magnetic stirring was carried out in order to avoid the settlement of the grape marc. After the pretreatment, the solution was rapidly cooled down to 40ºC and the extraction continued for 3 hours in the conventional stirred system, as it was formerly explained. The first sample for the kinetic study was taken just after the pretreatment, before the start of the conventional extraction. The effect of ultrasound intensity was studied with different amplitudes (50-75-100%) and at different final temperatures (40-55-70ºC). The pretreatment time was set to reach the specified final temperature. Every pretreatment was carried out in parallel and the duplicate of the middle point (55ºC, 75%) was repeated 3 times. The consumed power (Pconsumed) of the system is distributed between the inefficiency in the generator and the transducer, and the dissipated acoustic power in the liquid. Only the latter represents the efficient (absorbed) energy in the treatment, since the others depend on the type of equipment (Weiss et al. 2011). The energy absorbed (Eabsorbed [kJ]) by the grape marc and the solvent mixture is finally transformed into heat. So, the absorbed energy may be calculated using an energy balance taking into account the temperature increment of the system (Qsensible), the solvent evaporation (Qlatent) and the heat loss (Qloss) through the vessel wall (Sólyom et al., 2011) (Eq. 1): Eabsorbed ≈ Qsensible + Qlatent + Qloss (1) Sensible heat (Qsensible [kJ]) is calculated from the mass of the grape marc and the solvent mixture (msample [g]), the temperature difference (ΔT [ºC]) before and after the ultrasound treatment, and the specific heat capacity of the grape marc and solvent mixture (Cp = 3.374 [kJ/(kg•K)] ). (Eq. 2) The heat capacity of the grape marc at certain moisture content was estimated as it was described by Tulasidas (1994). Qsensible = msample Cp ΔT (2) Latent heat of vaporization (Qlatent [kJ]) is calculated from the amount of evaporated solvent (mvap), measured by weight loss after microwave treatment, and the heat of vaporization of the solvent (ΔHvap= 540 [kJ/kg]) (Eq. 3). Chapter 5 150 Qlatent = mvap ΔHvap (3) The heat loss from the vessel surface (Qloss) was evaluated through the global heat transfer coefficient of the system, obtained from the decrease of temperature (0.36 ºC/min) over time of a hot sample after finishing the ultrasound treatment. To estimate the total energy consumed by the ultrasound equipment, power consumption (Pconsumed) was recorded during the pretreatments. The ratio of energy absorbed by the sample to energy consumed by the ultrasound equipment was referred to as energy efficiency, and was calculated as follows (Eq.4): 𝐸𝑓𝑓𝑖𝑐𝑖𝑒𝑛𝑐𝑦 (%)= 𝐸𝑎𝑏𝑠𝑜𝑟𝑏𝑒𝑑 ∫𝑃𝑐𝑜𝑛𝑠𝑢𝑚𝑒𝑑 𝑑𝑡 (4) 2.4. Thermal treatment The thermal history (temperature versus time) of an ultrasound pretreatment at 100% amplitude and 70ºC was repeated in a conventional thermal treatment process using an electric heater, and it was followed by a conventional extraction stage. The comparison of both pretreatments allowed for the investigation of a probable non-thermal effect of ultrasounds. 2.5. Total phenol content analysis Total phenol content (TPC) was chosen to quantify polyphenols extraction kinetics. TPC is an indicative parameter when obtaining valuable compounds from grape marc, but does not provide relevant information on antioxidant activity or the content of any specific component of interest. However, its ease of analytical measurement and its representative behavior in global polyphenol extraction makes of TPC a suitable election to follow the extraction kinetics in this process. Ghafoor et al (2009) found that up to 60ºC during 30 min TPC, antioxidant activity and anthocyanin content change similarly in an UAE process of grape seeds. TPC was quantified, using the Folin-Ciocalteu method, as Gallic acid equivalents per gram of dry material [mg GA/g DM]. (Singleton et al. 1999) A volume of 40 μl of filtered sample (0.2 μm, nylon membrane) was diluted with distilled water (3 ml) and mixed with 200μl of FolinCiocalteu reagent. After 10 minutes saturated Na2CO3 was added to the solutions, and the samples were incubated at 40ºC for 30 min. Absorbance was measured at 765 nm (UV 2550 Chapter 5 151 Shimadzu UV/VIS spectrometer) and known concentrations of diluted Gallic acid were used for calibration. 2.6. Extraction kinetics and activation energy The extraction yield obtained from TPC measurements was plotted versus extraction time and first order kinetics was fitted to the experimental points (Eq.5) TPC=A·(1-exp(-βt)) (5) The extraction process is characterized by 1) the final extraction yield, A [mg GA/g DM], and 2) the initial extraction rate, Aβ [mg GA/g DM min], where β is the time constant [min-1]. The fitting of the experimental data was achieved by absolute error minimization, using the Solver tool in MS Excel. The dependence of the extraction rate constant on temperature can be described by the Arrhenius equation (Eq. 6): β=β0·exp(-Ea/RT) (6) , where β0 is the pre-exponential temperature independent factor [min-1], R is the universal gas constant (8.314 J/mol K), T is the absolute temperature [K] and Ea is the activation energy of extraction [J/mol]. These parameters, Ea and β0, were obtained from the slope and interception in the linear fitting of ln(β) versus 1/T, respectively. 2.7. Data Analysis Full factorial experimental design (32) was used to study the ultrasound effect in extraction experiments. Two-way ANOVA analysis was performed where results were considered to be significantly different at 95.0 % confidence level. Dependent variables were the final extraction yield (mg TPC/g DM) and the initial extraction rate. Power level (50-75-100%) and Absorbed energies corresponding to temperatures (40-55-70 ºC) were the two factors, each on 3 levels.