Sound-Assisted Fluidization for Temperature Swing Adsorption and Calcium Looping: A Review

Fine/ultra-fine cohesive powders find application in different industrial and chemical sectors. For example, they are considered in the framework of the Carbon Capture and Storage (CCS), for the reduction of the carbon dioxide emissions to the atmosphere, and in the framework of the thermochemical energy storage (TCES) in concentrated solar power (CSP) plants. Therefore, developing of technologies able to handle/process big amounts of these materials is of great importance. In this context, the sound-assisted fluidized bed reactor (SAFB) designed and set-up in Naples represents a useful device to study the behavior of cohesive powders also in the framework of low and high temperature chemical processes, such as CO2 adsorption and Ca-looping. The present manuscript reviews the main results obtained so far using the SAFB. More specifically, the role played by the acoustic perturbation and its effect on the fluid dynamics of the system and on the performances/outcomes of the specific chemical processes are pointed out.


Introduction
Fine and ultra-fine powders, i.e., characterized by particle size lower than 30 µm, have been characterized by increased research interest in recent years. In fact, due to their exceptional properties, mainly derived from their characteristic dimensions and large surface area per unit mass (which means that they are able to offer larger contact/reaction efficiency), they can be applied in a variety of industrial sectors, such as in the production of foods, plastics, cosmetics, catalysts, and metallurgic components [1].
For instance, nanostructured catalysts, characterized by a hierarchical porosity, have been synthesized in the framework of different chemical processes, such as combustion and gasification and fuel and polymer production [2]. Likewise, fine/ultrafine particles have also found application in the Carbon Capture and Storage (CCS) sector, for the reduction of the CO 2 emissions to the atmosphere [3,4], and in the thermochemical energy storage (TCES) in concentrated solar power (CSP) plants [5][6][7][8].
In regards to the CCS, different separation techniques are available to separate the CO 2 from the combustion flue gases: absorption, adsorption, cryogenics separation, membranes, etc. [9][10][11]. Among these, amine-based absorption, even though being the most mature technological option [12,13], is characterized by several shortcomings, when applied for post-combustion capture: huge energy demand for the regeneration of the sorbent, thermal/chemical degradation of the amines, amine losses due to evaporation and corrosion problems [12]. Therefore, adsorption has been proposed as one of the most attractive and promising alternatives, being characterized by relatively low energy consumption for the sorbent regeneration, high selectivity, no liquid waste streams, and quite a wide range of possible operating temperatures [14][15][16][17]. In particular, regarding the critical issue of the sorbent regeneration [18][19][20][21], different strategies can be adopted, acting either on the pressure (PSA-pressure swing adsorption) or on the temperature (TSA-temperature swing adsorption) of the system to induce the desorption of the CO 2 molecules [22][23][24][25]. In this framework, TSA combined with an indirect heating process (for example by means of heat exchanger tubes), is considered to be very promising for post-combustion operations [21,24,26]. However, the selection of the sorbent is crucial for adsorption to become a competitive solution [27][28][29][30]. Indeed, a good sorbent should be relatively convenient from the economic point of view, but, at the same time, it should be able to provide good performances at low CO 2 pressure (<0.2 atm [23]): high equilibrium adsorption capacity, fast kinetics of adsorption/desorption, stability to cyclic operations and tolerance to water vapor and other possible impurities in the exhaust stream [31]. In the framework of the need to produce specific adsorbent materials, fine/ultrafine particles have attracted growing interest in research [32]. Indeed, they can quite easily serve as the substrate for the production of ad-hoc synthesized sorbents, i.e., characterized by remarkable affinity towards the molecules of CO 2 [27][28][29]. As a matter of fact, due to their distinctive properties, size, and nature their surface chemistry and porous structure can be easily adjusted/functionalized at the molecular level, i.e., by introducing chemical groups/ligands with basic functionalities (e.g., carbonates, amino groups, etc.) able to tailor their adsorption behavior towards the CO 2 molecules [27][28][29][33][34][35][36]. Besides that, it must be considered that most of the sorbents commercially available are produced in a fine powdered form [27][28][29].
Likewise, fine/ultrafine powders have also been proposed as a possible solution to overcome the drawbacks of the calcium looping (CaL) process, which can be carried out for both CCS [37] and TCES-CSP applications [7,8], according to the following reaction scheme: In the former case, CaO/CaCO 3 is used as sorbent material to capture/release CO 2 by means of carbonation/calcination cycles [37]. In the latter case, CaO is used as an energy intensive material for the storage of energy in a chemical form by means of the reversible carbonation/calcination reaction [7,8]. In regards to the carbonation reaction, it must be underlined that it takes place in two successive stages: (i) a fast initial stage controlled by the reaction kinetics and by the transport of the CO 2 molecules from the bulk phase to the particle surface and (ii) a slow final stage controlled by the diffusion of the CO 2 molecules through the carbonate layer, which covers the remaining free surface of the CaO particles [38]. In light of this, it is clear that a poor gas-solid contact efficiency can strongly hinder the carbonation reaction [37]. Moreover, regardless of the specific application, i.e., either CCS or TCES, the main shortcoming of CaL is represented by the loss of CaO activity, i.e., the progressive drop of the carbonation conversion over repeated cycles due to sintering and pore-plugging [39,40]. Therefore, thanks to the extensive research effort devoted to finding solutions to this issue in the last years, different options are available to limit the natural CaO deactivation, such as: thermal activation [41,42], hydration [43], mechanical crushing [44,45], and the synthesis of Ca-based sorbents able to provide superior reactivity [8,44,46]. In this framework, the utilization of fine CaO particles has been recently proposed as a promising alternative solution [7,47]. In fact, it has been demonstrated that the sintering and pore-plugging phenomena can be remarkably hindered if CaO small particles are employed, thus improving the obtainable multicyclic performances [7,47]. Besides this, it has been reported that the use of fine CaO particles, thanks to their intrinsic high surface to volume ratio, makes it possible to enhance the gas-solid contact efficiency with respect to coarser CaO particles [37,48]. Likewise, the use, in small concentrations, of inert nano-sized particles as additives to fine cohesive limestone powders has been also proved to be a viable option to improve the flowability and reactivity performances of the system [49,50].
Therefore, in light of all the possible applications reviewed above, it is evident that it is necessary to develop new technologies to process these particles in big amounts. In this framework, several techniques can be used to handle and disperse fine/ultrafine powders in a continuous manner. Among these, fluidization is one of the most effective technological options. Indeed, fluidized beds can rely on different advantages, such as: high gas-solid contact area and efficiency, large heat/mass transfer coefficients, relatively easy control of the bed temperature uniformity, high particle flowability, flexibility in terms of types of powders to be processed, and suitability for large-scale applications [51].
However, regardless of these benefits of fluidized beds in handling powdered materials, it must be considered that fine/ultrafine particles, i.e., belonging to Geldart's C group, cannot be fluidized under ordinary fluidization conditions due to their intrinsic cohesive nature [1,2,52]. Indeed, these particles are characterized by strong interparticle forces (IPFs), causing the natural formation of large aggregates, due to the particles tendency to stick to one another giving rise to larger structures as large as hundreds of microns [1,2,53]. This agglomeration tendency causes, in turn, channeling/plugging phenomena, thus hindering the possibility to obtain a proper fluidization quality under ordinary conditions [2,[53][54][55][56].
Therefore, fine/ultrafine powders can be fluidized only with the use of some kind of externally assisted fluidization devices, namely it is necessary to introduce some kind of external energy in order to overcome the strong inter-particle forces. Among the different alternatives (magnetic [57,58], electric fields [59,60], mechanical vibrations [61,62], etc.), sound-assisted fluidization has been proposed as one of the most effective solutions [2,55,63]. In particular, it has been reported that agglomeration, channeling and/or slugging phenomena can be efficiently hindered if a suitable acoustic field, i.e., capable of overcoming the IPFs, is applied, thus assuring the possibility to achieve a proper fluidization regime [19,37,64,65].
In this context, the sound-assisted fluidized bed reactor (SAFB) designed and set-up in Naples by our research group has proven itself as a consistent device for the fundamental and phenomenological study of the fluidization behavior of the fine/ultrafine cohesive powders at atmospheric pressure [2,54,55,66,67]. More recently, the SAFB has also been upgraded to be used for both low-and high-temperature chemical processes, such as CO 2 adsorption and Ca-looping, studied in the framework of both CCS and TCES-CSP applications [3][4][5][6]15,[19][20][21]36,37,64,65,68,69].
In this work, the main results recently obtained by means of the SAFB setup are reviewed, particularly focusing on how the application of the acoustic perturbation reflects not only on the fluid dynamics of the system, but also on the performances/outcomes of the specific chemical processes carried out therein. Figure 1 reports the scheme of the SAFB designed and set-up in Naples by our research group and used in all the works reviewed in the present manuscript. It is based on the initial setup originally conceived by Chirone et al. [70,71] to study the fundamental behavior of cohesive powders at atmospheric pressure. The SAFB mainly consists of five parts: (i) the fluidization/reaction section; (ii) the acoustic section; (iii) the control section; (iv) the analysis section; and (v) acquisition section.

SAFB
The fluidization section is made of a fluidization column in Plexiglas, Pyrex, or quartz, depending on the operating temperature. The column is provided with a gas distributor, consisting of a quartz porous plate. The section situated below the gas distributor is a wind-box. In particular, the proper gas distribution below the quartz plate is assured by a quartz rings filling. A pressure probe (linked to pressure transducers) located just above the gas distributor makes it possible to directly measure the pressure drops across the bed, i.e., without the contribution of the distributor plate.
The acoustic section consists of a sound-generation/propagation system. In particular, the acoustic field is generated by a digital signal generator, which is capable of generating an electrical sine wave of specified frequency (f). Then, the signal, amplified in a power audio amplifier rated up to 400 W, is sent to an 8 W woofer loudspeaker and finally introduced in the fluidization column through a specifically designed sound wave guide, which is linked to the upper part of the fluidization column so that the sound wave propagates from the top to the bottom of the column. In particular, the geometry of this sound wave guide and, in turn, the reciprocal position of the loudspeaker and the fluidization column, have been properly conceived in order to protect the loudspeaker from any possible elutriated materials and/or from the heat emitted from the bed in the case of high temperature tests. The response of the bed to the acoustic perturbation is monitored through a microphone located in the upper part of the fluidization column and linked to an oscilloscope. This makes it possible to monitor and adjust the intensity (SPL) of the acoustic field according to the signal received from the bed through the microphone. Considering that the intensity of the acoustic fields generated in our research activity is relatively high (>110 dB) [3,54], the experimental set up envisages a noise insulation system based on the principles of the Helmholtz resonator [72][73][74]. The resonator simply consists of a cavity connected to the system of interest through one or several short and narrow tubes (necks). The resonator has a characteristic resonance frequency (to which corresponds the maximum oscillation), which is strictly dependent on its geometry, i.e., the volume of the cavity and the length of the neck [72][73][74]. Therefore, by properly selecting these geometrical parameters, it has been possible to make the resonance frequency characteristic of the resonator fall inside the frequency domain of interest, thus effectively dampening the noise [72][73][74]. The control section consists of all the devices used to set, measure, and control the operating variables of the experiments, i.e., the inlet gas flow rates and temperature. A set of mass flow meters and controllers (Brooks) with different full-scale values (ranging from 1 to 300 NL h −1 ) is used to set and control the inlet gas flow rate. Likewise, a mass flowmeter (Brooks) located at the exit of the column is used to measure the flow rate of outlet stream. The gas feed is prepared using separate high purity cylinders (99.999%) containing the gaseous species (N 2 and CO 2 ) required for the experiments. In regards to the temperature, it is set, measured, and controlled by a PID controller equipped with a type K thermocouple (positioned at the center of the bed), which is linked to an ad-hoc designed electric heating jacket (Tyco Thermal Controls GmbH) specifically designed with an open window in order to visually assess the fluidization quality.
The analysis system is made of a continuous online analyzer (ABB, AO2020) with an infrared (URAS 14) detector (with a 0-100 CO 2 vol.% full range) able to measure the outlet CO 2 concentration continuously. In particular, a part of outlet stream is sampled by means of a sample gas feed unit (ABB, SCC-F) and sent to the analyzer, whereas the residual part is sent to the stack. The system is also equipped with a Peltier cooler (ABB, SCC-C) for cooling the sample and separating and removing any possible condensate. Besides that, the analysis system is also protected from any possible elutriated particles by a system of filters placed downstream from the exit of the fluidization column. With reference to the elutriation phenomena, it should be considered that under the application of acoustic fields, the elutriation of fine particles from the fluidized bed is remarkably reduced, thus limiting possible related issues such as clogging of valves [2].
Finally, in regards to the data acquisition system, the pressure drops (∆P), temperature (T), gas flowrate (Q), and concentration (C) signals are continuously recorded and elaborated on a PC, at a sampling rate of 1 Hz, via LabView software.

Materials
All the materials tested in the SAFB are reported, with their main properties, in Table 1. Clearly, except for the silica sand, belonging to the Geldart's A group, all the powders are characterized by dimensions lower than 30 µm, i.e., they all belong to the Geldart's C group. However, they can be categorized in two main groups on the basis of their primary particle size: nano-sized and micro-sized powders. ρ p = particle density; BET SSA = BET specific surface area; V p = pore volume; m = mass of powder loaded in the SAFB. * Tests performed in the Seville experimental apparatus.
In particular, the nano-powders are: Al 2 O 3 , Fe 2 O 3 , ZrO 2 , and CuO, all of them produced by Sigma-Aldrich. Whereas, in regards to the micro-powders, the two commercial activated carbons are an activated carbon DARCO FGD (Norit), AC Norit, and an activated charcoal powder (Sigma Aldrich, Milan, Italy), AC Sigma. The metal organic framework (MOF), HKUST-1, has been synthesized according to the procedure reported in [68]. The natural limestone, CaCO 3 , comes from Belchite quarries (Spain) and has been supplied by OMYA. Finally, the synthetic CaO is produced by Sigma-Aldrich.

Fluidization Tests
All the fluidization experiments have been carried out in the SAFB at atmospheric pressure and at temperatures ranging from 25 to 850 • C under ordinary ad sound-assisted operating conditions. The fluidizing gas is N 2 in order to avoid air moisture that could intensify the powder cohesiveness. In regards to the sound-assisted tests, acoustic fields with different sound frequency (f: 20-300 Hz) and intensity (SPL: 120-150 dB) have been applied in order to study the effect of the acoustic parameters on the fluidization behavior.
Each fluidization experiment consists in measuring and recording the pressure drops as a function of the superficial gas velocity (u), i.e., by both decreasing (DOWN) and increasing (UP) u. All of the results shown in this work have been obtained from the DOWN curves. Then, the main fluidization parameters are evaluated by working out to the obtained experimental data, i.e., the minimum fluidization velocity (u mf ) and the average aggregate size (d agg ).
In regards to u mf , the well-known graphical procedure is applied to the experimental pressure drop curve: u mf is evaluated by crossing the line that fits the data for flow through a packed bed and the horizontal line that fits the data for the fully fluidized bed [75].
In regards to d agg , the correlation proposed by Wen and Yu [75] has been applied using the experimental values of u mf . In particular, an apparent density, smaller than the primary particle density, has been considered for the aggregates [54].

CO 2 Adsorption Tests
Dynamic adsorption/desorption experiments have been carried out in the SAFB under atmospheric pressure and different adsorption/desorption temperatures (T ads , T des ) ranging from 25 to 150 • C and either with or without the assistance of the sound perturbation. In particular, relatively low values of CO 2 partial pressures (P CO 2 : 0.05-0.20 atm) have been tested in order to simulate the composition of a typical post-combustion flue gas. The feeding stream in both the adsorption and desorption step is set so that superficial gas velocity is sufficiently bigger than the u mf of the tested sorbent, thus assuring that the whole bed is fully fluidized.
In particular, cyclic tests of CO 2 capture and recovery have been performed according to the TSA approach. Therefore, each cycle consists of an adsorption step, in which the CO 2 is captured on the solid sorbent, and a subsequent desorption step, in which the sorbent is regenerated by increasing the temperature of the system and the CO 2 , previously captured, is released.
Adsorption: Before the adsorption test, a preliminary step is always performed: the sorbent is firstly dried/cleaned by flowing N 2 at 150 • C and atmospheric pressure for 60 min (this is necessary to remove any trace of water, which would end up reducing the CO 2 adsorption capacity). Then, it is pre-conditioned by fluxing N 2 through the system until the system reaches the desired temperature. After this preliminary stage, the adsorption step is started, in which the feed to the column is switched to a CO 2 /N 2 gas mixture of specified composition (C 0 ). Thanks to the continuous monitoring of the CO 2 concentration in the outlet stream, the breakthrough (BT) curves, consisting in the plot of C/C 0 versus time, are obtained. Then, the BT curves are elaborated in order to calculate some important adsorption parameters: • q e -the amount of CO 2 adsorbed per unit mass of adsorbent at the thermodynamic equilibrium. In particular, this is evaluated from the integration of the BT curve according to the following mass balance: where m is the sorbent mass, F CO 2 ,in and F CO 2 ,out are the inlet and outlet CO 2 molar flowrate, respectively, and t s is the saturation time; • t b -the BT time, namely, the time in which the outlet CO 2 concentration is the 5% of the feed concentration. Technically, in real industrial adsorption processes, it is the time at which the adsorber is taken off-line for regenerating the sorbent, which means that larger values of t b correspond to the more effective capture capacities; • ∆τ-a time parameter providing information on the kinetics of the adsorption process. The smaller ∆τ is, the faster the adsorption will be. In particular, it is representative of the slope of the linear part of the BT curve and it is evaluated as t 95 -t b (with t 95 being the time for which the outlet CO 2 concentration reach 95% of the concentration in the feed), which means that higher values of ∆τ represents steeper BT curves; • ψ-the fraction of bed used at BT, namely the percentage of the total captured CO 2 (q e ) which is adsorbed at the BT time, t b .
Desorption: In regards to the desorption step, different regeneration modes have been used: • non-isothermal purge-in this mode of operation, the heating happens contextually to the N 2 purging step. In particular, once the adsorption step is concluded, the column is heated up (20 • C/min) to the selected T des and, at the same time, the feed (Q IN ) is switched from the CO 2 /N 2 mixture, used during the previous adsorption step, to pure N 2 (Q p ) to purge the bed. • isothermal purge-in this mode of operation, the purging step happens isothermally.
In particular, once the adsorption step is concluded, the column is isolated so that all the desorbed CO 2 is kept inside the column. To do so, the inlet and outlet of the column are closed by means of two-way valves and, in the case of sound-assisted adsorption tests, the acoustic field is switched off and the system is heated up to the selected desorption temperature. Then, once the set T des is reached, the purging step is started, namely, the acoustic field is switched on, in the case of sound-assisted tests, the feed to the column is opened fluxing N 2 and the column exit is un-sealed. In this way, the CO 2 already desorbed from the adsorbent, which has been trapped inside the isolated column, and the CO 2 still adsorbed on the sorbent (i.e., the CO 2 that can be released only by reducing the CO 2 partial pressure) exits from the column diluted in the N 2 stream.
In analogy to the adsorption step, thanks to the continuous monitoring of the CO 2 concentration in the outlet stream, the desorption peaks are elaborated in order to calculate some important desorption parameters: • R-the CO 2 recovery level, defined as the percentage of the previously captured CO 2 which is recovered during the desorption step; • t d -the time needed for CO 2 desorption at a fixed value of R; • C m -CO 2 purity, expressed as the average CO 2 concentration in the desorbed stream.

Ca-Looping Tests
CaL tests have been performed for both CCS and TCES applications either with or without the assistance of the acoustic perturbation. However, only the tests for TCES have been performed in the SAFB in Naples, whereas those for CCS have been performed in the experimental apparatus set up in Seville in collaboration with the University of Seville. Detailed information on this experimental apparatus can be found in [37].
The experimental procedures used for both the CCS and TCES are similar, they only differ for the operating temperatures of the carbonation and calcination steps. In fact, the optimal operating conditions to perform the CaL process are strictly dependent on the particular application. In the case of CCS applications: the calcination, whose conditions are dictated by the need of extracting from the calciner high-concentration CO 2 stream to be compressed and sequestered, is performed at high temperature (∼950 • C) under high CO 2 partial pressure. The carbonation, whose conditions are dictated by the fixed concentration of the combustion flue gases, is performed under low CO 2 partial pressure (∼0.15 bar) at the lowest possible temperature (∼650 • C) able to favor the reaction thermodynamically [5]. In the case of TCES applications: since there is no more CO 2 capture/storage issue, relatively low temperatures (~750 • C) and CO 2 partial pressures (i.e., by using a gas easily separable from CO 2 , such as superheated steam or helium) are employed to perform the calcination. On the contrary, aiming at achieving high global efficiency for energy storage and electricity generation, high CO 2 partial pressures and temperatures (around or above 800 • C) are employed to perform the carbonation [5].
In both cases, CCS and TCES tests, the sorbent is subjected to a sound-assisted (150 dB-120 Hz) pre-treatment step, in which it is pre-calcined in N 2 at 900 • C in order to obtain a completely de-carbonated sorbent for the following carbonation step.
CCS: After the pretreatment, the sorbent is subjected to a carbonation step, performed at 650 • C feeding a mixture of 15% CO 2 /85% N 2 v/v until the saturation of the sorbent. Then, a calcination step is performed at 900 • C in N 2 atmosphere to fully de-carbonate the sorbent. TCES: The operating conditions have been carefully chosen in order to simulate a CaL-CSP integration scheme in which the calcination is carried out at low temperature feeding helium and the carbonator is integrated with a closed CO 2 Brayton cycle [76]. In particular, after the pre-treatment, the sorbent is subjected to a carbonation step carried out at 850 • C using a mixture of 70% CO 2 /30% N 2 v/v until the sorbent is saturated. Then, a calcination step is performed at 750 • C in N 2 atmosphere until the sorbent is completely de-carbonated.
Also for these tests, the outlet CO 2 concentration profiles are elaborated in order to calculate the carbonation conversion, X.

Results
In the following paragraphs, the main results obtained by means of the SAFB setup are presented and reviewed, pointing out how the application of the acoustic perturbation affects the fluid dynamics of the system and, in turn, also the performances/outcomes of the specific chemical processes carried out therein.

Fluid Dynamics of Sound-Assisted Fluidized Beds
Studying the behavior of different fine/ultrafine cohesive particles, it has been shown that, regardless of the specific particle size, the application of an acoustic perturbation, with proper f and SPL, makes it possible to either achieve an appropriate fluidization regime or improve their fluidization quality [2,55], in terms of ideal-like pressure drop curves and affecting the values of all the most important fluidization parameters (lower values of u mf and d agg ). In this framework, Figures 2 and 3 report, as examples, the fluidization and expansion curves obtained for different nano-sized and micro-sized powder, respectively, under ordinary and sound-assisted conditions.  Clearly, from the analysis of Figures 2a and 3a, it is evident that when the powders are fluidized without any external excitation, the fluidization quality is relatively poor, as confirmed by the dimensional pressure drops being unstable and lower than 1, which means that a fraction of the bed remains not fluidized. On the contrary, when the acoustic perturbation is applied, the fluidization curves become more regular, with the whole bed fluidized (∆P/∆P 0 always reaching 1) and higher values of expansion ratio (Figures 2b and 3b). These experimental evidences are due to the fact that the application of a proper sound excitation is able to overcome the strong interparticle forces, thus hindering the agglomeration, channeling, and/or slugging phenomena. More specifically, the application of the acoustic perturbation generates both inertia and viscous forces that induce an oscillatory motion of gas molecules and solid particle/aggregates. In particular, depending on the dimension of the fluidizing aggregates (i.e., cluster and sub-clusters), the entity of this motion can be more or less intense [53]. This results in a relative motion between particle aggregates of different dimensions, which, in turn, induces a dynamic break-up mechanism of the bigger clusters into smaller and more easily fluidizable sub-clusters, as theoretically explained in [54].
The dynamics of the break-up mechanism has been also clearly highlighted investigating the mixing of different nano-sized powders (F 2 O 3 and Al 2 O 3 ). In particular, combining the color tracing method with SEM/EDS morphological/chemical analysis of samples taken from the bed at different times, it has been demonstrated that the acoustic perturbation promotes a dynamic evolution of the fluidizing aggregates. In fact, the fluidizing aggregates, initially formed by only one of the two powders, are not static but dynamic structures, as confirmed by the fact that their chemical composition varies with time during the sound-assisted mixing test. Due to the aggregates continuously breaking-up and re-aggregating, there is an actual exchange of material at the micron scale which creates hybrid aggregates containing both of the two types of particles. More specifically, while the bed appears macroscopically well mixed after just few minutes, the local mixing inside the aggregates takes longer times. These considerations are clearly evidenced in Figure 4a,b reporting, respectively, the chemical composition of the hybrids aggregates and the time evolution of the mixing degree M, where M(t) is the ratio between the number of aggregates having a chemical composition different from the theoretical value of less than 10% and the total number of aggregates analyzed at the time t [66].

Effect of Sound Intensity and Frequency
Aiming at highlighting the effect of the acoustic parameters, i.e., SPL and f, on the fluidization behavior of fine/ultrafine particles, tests have been performed in the SAFB applying acoustic fields of different SPL (120-150 dB) and f . In this framework, Figure 5a,b report the effect of SPL and f on the fluidization behavior, in terms of u mf , of different nano-sized and micro-sized particles, respectively. Clearly, in regards to the effect of SPL, it has been found that increasing intensities of the acoustic perturbation lead to a general enhancement of the fluidization quality ( Figure 5). Indeed, the strength of the acoustic field is proportional to SPL [2,77] and, as a consequence, a minimum acoustic energy (SPL > 120 dB) is required to start the break-up mechanism, which, in turn, promotes the fluidization process. Besides that, an increase of SPL results in enhancing the fluidization quality, as clearly confirmed by u mf decreasing with increasing SPL ( Figure 5). Indeed, larger values of SPLs are representative of larger amount of external energy introduced inside the system, which means that the break-up mechanism is more effective, i.e., the larger fluidizing aggregates are disrupted in smaller and more fluidizable structures. However, it must be also considered that there is some experimental evidence that at very high values of SPL, this positive effect may be reversed because of the collision between particles and/or agglomerates becoming more and more probable, thus leading to an increase of the aggregate size [78].
On the contrary, in regards to the effect of sound frequency on the fluidization quality, it has been found that there is a non-monotonic correlation, indeed, it can be found an optimum range of frequency (80-120 Hz) in the proximity of to the natural frequency of the particle bed. This is clearly shown in Figure 5, where the existence of the optimum range of frequency, providing the lowest values of u mf (i.e., more efficient break-up mechanism), is evident for all the tested nano-sized and micro-sized powders. In regards to the explanation of this experimental evidence, it must be considered that for too small values of sound frequency, there is no relative motion between smaller and larger aggregates and, as a consequence, there is no aggregate break-up [2,53,54]. On the contrary, for too high, values of sound frequency, the propagation of the acoustic waves though the particulate bed cannot take place properly [2,53,54]. In fact, there is a proportional correlation between the sound absorption coefficient and the square of sound frequency [2,53,54]. Therefore, when the sound frequency is too high, the upper part of the bed absorbs most of the sound energy, whereas, the bottom of the bed is only interested by a weakened sound energy, thus meaning that the large aggregates located there are not efficiently disrupted [2,53,54].

Effect of Temperature
It is generally accepted that the influence of the temperature on the fluidizing gas characteristics (density and viscosity) can be used to explain how the fluidization behavior is affected by the temperature only if the fluidization is controlled by the hydrodynamic forces (HDFs) [2,53,54,[79][80][81]. On the contrary, other aspects must be taken into account if the IPFs are either comparable with or predominate over the HDFs, as typical in the fluidization of fine/ultrafine particles [53,54]. Indeed, the temperature affects as the gaseous phase, in terms of density and viscosity variations, as the solid phase, in terms of intensification of IPFs [53,54]. In this context, several authors studied the effect of the temperature on the fluidization behavior of different types of powders, reporting an increase of the IPFs with increasing values of the temperatures [2,53,54,79,80]. It has been also found that, in spite of the fluidization of coarser powders being generally controlled by HDFs and not affected by IPFS, non-cohesive A powders can switch to C-type fluidization behavior when the temperature is increased, as a consequence of the intensified IPFs outbalancing the HDFs [81].
The effect of the temperature on the fluidization behavior of A and C powders has been also investigated in the SAFB. According to this general consensus of the works available in the literature, it has been found that increasing temperatures negatively affect the fluidization quality due to the intensification of IPFs. Indeed, regardless of the powders being either A or C type, the fluidization parameters, u mf and d agg , are increased when the temperature is increased, as clearly shown in Figure 6. This can be explained considering that, as consequence of the intensified IPFs, the achievement of a fluidization regime becomes more and more difficult due to a less effective break-up mechanism. Indeed, with the intensification of IPFs the cohesive character of the powder is enhanced, i.e., with increasing temperatures, more and more particles adhere to one another to form larger aggregates (higher values of d agg ) (Figure 6b) which lead to larger values of u mf (Figure 6a). In particular, in regards to the C type powders, it is clear that the efficiency of the acoustic perturbation decreases as the temperature is increased. In fact, as clearly inferable from Figure 6b, the fluidizing aggregate size, which is always larger than the powder average size (Ashes = 8 µm, CaO = 4 µm- Table 1), increases more and more as the temperature is increased, becoming 13-15 times larger than the Sauter diameter around 800 • C.
Likewise, also in the case of the A type powder, increasing temperatures lead to a reduced efficiency of the acoustic field in counteracting the effect of IPFs. In fact, even though the fluidizing parameter increase whether or not the acoustic field is applied, the acoustic field keeps its effectiveness up to 200 • C. In fact, the fluidizing aggregates are broken down to the Sauter diameter (60 µm- Table 1) only for temperatures smaller than 200 • C, as confirmation of an effective break-up mechanism. On the contrary, when the temperature is increased beyond 200 • C, the effectiveness of the acoustic field decrease until becoming negligible at 600 • C, where no difference is observed between the ordinary and sound-assisted tests (d agg is about two times the D sauter of the powder) (Figure 6b).

Temperature Swing Adsorption (TSA)
Since fine/ultrafine powders have great potential to be used for the CO 2 adsorption process, the SAFB has been proposed in a TSA configuration (TSA-SAFB) as possible technology for the handling and processing of these materials in large quantities. In particular, an intense research activity has been performed to highlight and demonstrate that fine/ultrafine powdered sorbents, regardless of their specific chemical nature, can be directly used in the TSA-SAFB, i.e., they can be used as free-flowing powders, without being previously treated. In particular, there is no need to perform a previous pelletization step (which, on the contrary, would be necessary to face the huge pressure drops of fixed bed reactors of fine particles), which ends up in decreasing the sorbent surface area and, consequently, its adsorption capacity for CO 2 , especially for some kinds of novel cutting-edge sorbent materials [82]. Besides this crucial benefit, the TSA-SAFB has been also proved to significantly improve the CO 2 adsorption and desorption performances of fine/ultrafine powdered sorbents with respect to their use in other standard reactor configurations (e.g., ordinary fluidized beds and fixed beds).

Adsorption
In regards to the adsorption step, Figure 7 reports the BT curves obtained for two adsorbents belonging to two different class of adsorbent materials, i.e., activated carbons and MOFs, under ordinary and sound-assisted fluidization conditions. Table 2 reports the adsorption parameters calculated from the elaboration of the BT curves.  The analysis of Figure 7 and Table 2 shows that the application of the acoustic perturbation remarkably improves the CO 2 adsorption performances of both the two sorbents. Indeed, all the adsorption parameters of AC Norit and HKUST-1 are enhanced in the sound-assisted test with respect to the ordinary test: (i) the equilibrium adsorption capacity, q e , is increased of about 20% and 45%; (ii) the BT time, t b , is increased of about 5 times; and (iii) the fraction of bed utilized at t b , ψ, is increased of about 4 and 3 times for AC Norit and HKUST-1, respectively. Besides this, it is also evident that the adsorption kinetics is also enhanced when the system is assisted by the acoustic excitation, as clearly evidenced by the BT curves becoming steeper and shorter. Indeed, the saturation time, t s , is roughly reduced to a half and ∆τ is reduced of about 60%, thus meaning that the CO 2 adsorption is speeded up when the acoustic perturbation is applied.
These experimental evidences can be explained considering that without the assistance of the acoustic perturbation, i.e., under ordinary conditions, the fluidization quality is extremely poor and characterized by strong instability. This is clearly evidenced by the fluidization curve obtained for both the sorbents (Figure 3a). Due to cohesive nature of the sorbents, the gas-solid contact efficiency is dramatically hindered and, therefore, most of the reactive gas (CO 2 ) bypasses the bed through the formation of channels rather than actually permeating the bed. Likewise, from the point of view of the solid phase, a great amount of the sorbent does not actually take part in the adsorption process, which means that a great amount of adsorbent surface is not exploited, being completely precluded to the gaseous phase. Then, once the most readily available adsorption sites (i.e., those located on the surface of the channels) are quickly saturated, the outlet CO 2 concentration increases, due to the above-mentioned by-pass phenomenon, with a sudden increase of the slope of the BT curve. Then, after this sharp rise, the slope of the BT curve abruptly decreases approaching the saturation with a very slow tail of the BT curve, which accounts for most of the adsorption process. In particular, the saturation of the bed occurs very slowly due to the fact that only a small fraction of the solid and gaseous reactants takes part in the adsorption process.
On the contrary, when the acoustic perturbation is applied to the system, the BT curves show a more regular shape, i.e., with no sudden initial increase and slow tail, thanks to the enhanced of the fluidization quality and, in turn, of the gas-solid contact efficiency. More specifically, the general improvement of the CO 2 adsorption performances observed during the sound-assisted tests is due to the characteristic dynamic break-up and re-aggregation mechanism of the fluidizing aggregates, as distinctive feature of the soundassisted fluidization process. In fact, this mechanism leads to the continuous renewing of the sorbent particles exposed to the fluid phase and, in turn, of the active sites available to the CO 2 adsorption process.
This strong connection between the CO 2 adsorption process and the sound-assisted fluidization process has been clearly evidenced by performing an OFF-ON test, in which the acoustic field is switched on only at a time t = t * , i.e., soon after the above-discussed change of slope characteristic of the ordinary tests, (Figure 7). It is evident that when the acoustic perturbation is switched on, at t = t*, the CO 2 concentration profile suddenly decreases and, after few minutes, restart the rising with a sound-assisted-like shape. This experimental evidence is the confirmation that the acoustic perturbation makes it possible to better exploit the sorbent surface. In fact, the sorbent surface, that is precluded to the gaseous phase while the sound is switched off in the first part of the test, suddenly becomes available, as soon as the acoustic field is switched on. As a consequence, the CO 2 concentration abruptly drops down, due to the renewed sorbent adsorption capacity.
The superiority of the SAFB reactor configuration to perform the CO 2 adsorption process on fine/ultrafine sorbents have been also proved by testing the same sorbent (AC Norit) in a fixed bed reactor. In particular, the fixed bed experiments have been performed with the sorbent in both powdered and pelletized (pellet size: 180-400 µm) form. The comparison of the obtained BT curves is reported in Figure 8. It is evident, in regards to the fixed bed tests, that the pelletization improves the performances with respect to the free-flowing powders. This is due to the fact of the strong cohesive nature of the fine sorbent, the gas flows through the bed through channels rather than permeating it. However, whether pelletized or not, when used in a standard fixed bed reactor, the sorbent provides remarkably worse CO 2 adsorption performances than when used in the SAFB. In particular, in the SAFB, its adsorption capacity can be increased of about 76% with respect to the fixed bed operation with the pellets [4]. This is most likely due to diffusive resistances inside the pellets, hindering the full exploitation of the sorbent surface. The effect of some important operating variables on the CO 2 adsorption process, i.e., CO 2 partial pressure, P CO 2 , and temperature, T ads , has also been investigated in the SAFB (Figure 9). It is clear, from the analysis of Figure 9, that the pressure and the temperature have an effect on both the CO 2 adsorption thermodynamics and kinetics.
In regards to the CO 2 adsorption thermodynamics, it is positively affected by increasing values of P CO 2 , whereas, it is negatively affected by increasing values of T ads . Indeed, higher values of P CO 2 lead to higher values of q e , due to the fact that the CO 2 adsorption process is thermodynamically controlled by P CO 2 [17]. On the contrary, q e decreases with increasing T ads , due to the adsorption being an exothermic process [11].
In regards to the CO 2 adsorption kinetics, it is positively affected by increasing both P CO 2 and T ads . In fact, higher values of P CO 2 lead to lower values of ∆t, i.e., faster adsorption processes, due to the fact that longer times are needed by the CO 2 concentration to go through the bed [11,16]. Likewise, larger values of T ads also result in smaller values of ∆t, due to the enhancement of the mass-transfer coefficients at higher temperatures [11,16].

Desorption
In regards to the desorption step, Figure 10a reports the desorption peaks and temperature profiles inside the bed obtained for the AC Norit under ordinary and sound-assisted fluidization conditions, using the non-isothermal purging mode described in the experimental section.
The analysis of Figure 10 evidences that, under both ordinary and sound-assisted fluidization conditions, the CO 2 desorption profile shows the typical shape of a peak followed by a final tail, which is indicative of the CO 2 desorption occurring slowly in the final stages due to the reduction of the driving force. Also, it is important to underline that, regardless of the fact that the process is carried out under sound-assisted or ordinary conditions, the sorbent is always fully regenerated, i.e., the desorbed CO 2 is always the same as the CO 2 previously adsorbed. However, even though the shapes of the ordinary and sound-assisted profile are similar, it is evident, in analogy to the results obtained for adsorption step, that the desorption profile is smoother and more regular under sound-assisted fluidization conditions. When no acoustic field is applied, on the contrary, a clear instability can be evidenced, which is caused by the fact that the establishment of preferential channels through the bed prevents a smooth desorption of the CO 2 . Besides this, the application of the sound also results in a faster heating rate, as clearly evidenced by the temperature profiles reported in Figure 10a, thanks to the larger heat transfer coefficients obtainable when the bed is fluidized, namely under sound-assisted conditions, with respect to those obtainable when the fluidization quality is poor, i.e., under ordinary conditions. The beneficial effect derived by the application of the sound is also highlighted by evaluating some important desorption parameters, namely the recovery level, R, the desorption time at a fixed R, t d , and the average CO 2 concentration C m . Clearly, as consequence of the peak-shaped desorption profile, whether the acoustic field is applied or not, higher t d values result in higher regeneration efficiency, i.e., larger values of R. However, at the same time, larger values of t d lead to lower values of C m , due to the increasing dilution effect caused by prolonged N 2 purging. This is evidently shown in Figure 10b, where it is clear that, when t d is increased, R monotonically increases and C m monotonically decreases. It is also evident that, at each fixed value of t d , both R and C m are always larger in the sound-assisted test than in the ordinary test.
In analogy to the study performed on the CO 2 adsorption step, the effect of some important operating variables on the CO 2 desorption process, i.e., desorption temperature (T des ) and N 2 purge flowrate (Q p ), has been also investigated in the SAFB, using the isothermal purging mode described in Section 2.3. The results obtained are shown in Figures 11 and 12. First of all, it is important to highlight that, regardless of T des and Q p , the sorbent can be fully regenerated (R = 100%), which agrees with the literature indicating the activated carbons as a physisorbent, i.e., the CO 2 is adsorbed by the formation of weak physical bonds rather than strong chemical bonds on the sorbent surface [15,18].  With reference to the effect of T des , the analysis of Figure 11a highlights that the CO 2 desorption process is fastened by increasing values of T des , as clearly denoted by the CO 2 desorption peaks becoming higher and narrower as T des is increased. This experimental evidence can be explained considering that at higher values of T des : (i) the thermodynamics equilibrium is shifted towards the desorption, namely the adsorption process is more and more penalized, meaning that the CO 2 are more easily found in the desorbed rather than in the adsorbed state and (ii) the CO 2 and N 2 molecular diffusivities are increased. Then, based on the elaboration of the experimental desorption peaks, it has been highlighted in Figure 11b that: (i) the time required to achieve a certain value of R monotonically decreases with increasing T des and (ii) at a fixed value of R, the concentration of the recovered CO 2 stream can be remarkably improved increasing T des . Moreover, it is important to underline that, regardless of the N 2 purge flow rates, values of T des larger than 70 • C are enough to concentrate the captured CO 2 ; indeed, as clearly shown in Figure 11b, C m is larger than C 0 (10 vol.%).
With reference to the effect of Q p , the analysis of Figure 12a evidences that the CO 2 desorption process is fastened by increasing values of Q p , as evidently shown by the CO 2 desorption peaks becoming narrower as Q p is increased. However, differently from what observed with increasing T des , Q p has no effect on the maximum of the desorption peak. As observed for T des , increasing values of Q p also positively affect the desorption rate. Then, the analysis of Figure 11b shows that: (i) the time required to achieve a certain value of R monotonically decreases with increasing Q p and (ii) the concentration of the recovered CO 2 stream achievable at a fixed R is only slightly affected by Q p .

Calcium Looping (CaL)
Since fine/ultrafine powders have been indicated as viable solution to overcome the limitation of the CaL, in the framework of both CCS and TCES application, the SAFB has been proposed in a CaL configuration (CaL-SAFB) as possible technology for the handling and processing of these materials. In particular, Figure 13a,b reports the results, in terms of time evolution of the outlet CO 2 concentration, obtained in the SAFB, for TCES applications, and in the Seville experimental apparatus, for CCS applications, respectively. It is clear that when the acoustic field is applied the carbonation performances are remarkably enhanced, regardless of the specific operating conditions of CCS and TCES. Indeed, the reaction degree, namely the quantity of CO 2 and CaO reacted, is proportional to the area between the curve and the horizontal line corresponding to the inlet CO 2 concentration. Clearly, under sound-assisted fluidization conditions, this area is bigger than under ordinary conditions, meaning that more CO 2 /CaO has reacted. More specifically, whether the acoustic perturbation is applied or not, the variation of the slope of the CO 2 concentration profile clearly evidences the existence of the two stages of the carbonation reaction, namely a fast stage under kinetic control followed by a slow diffusion controlled one. However, the time at which the transition between these two stages occurs is strongly dependent on whether the sound is applied or not. In particular, under sound-assisted fluidization conditions the transition occurs much later than under ordinary conditions. Indeed, under ordinary conditions, the CaO is carbonated mainly in the slow stage under diffusion control, due to the relatively fast switch from the previous fast stage of the reaction. Under sound-assisted conditions, on the contrary, the contribution of the fast kinetically controlled stage to the global carbonation conversion is remarkably enhanced, thus resulting in a general speed-up of the process.
These experimental evidences are consistent with the observed enhancement of the fluidization quality under sound-assisted fluidization conditions ( Figure 3). Indeed, as also clearly shown in Figure 14a, the fluidization quality is particularly poor under ordinary conditions, i.e., characterized by severe channeling. As a consequence, when no sound perturbation is applied, the gas-solid contact efficiency is dramatically low: a large portion of both the gaseous, i.e., CO 2 , and solid, i.e., CaO, hardly takes part in the carbonation reaction due to gas by-passing, though channels, and solid agglomeration, which reduces the availability of free CaO surface. It is also evident from Figure 13 that the effect of the acoustic perturbation is stronger in the initial fast stage of the reaction, when the reaction is mainly due to the reaction of CO 2 with the CaO surface and the sound can improve the gassolid contact efficiency. On the contrary, the effect of the sound becomes less evident in the slow diffusion controlled stage, when the reaction is no longer controlled by the availability of free CaO surface and starts to be controlled by the diffusion of the CO 2 molecules through the carbonate layer. Indeed, in this stage, the improved fluidization quality and gas-solid contact efficiency yielded by the acoustic perturbation have a negligible effect on the reaction. It is also evident, from the results reported in Figure 14b, that the application of the acoustic field makes it possible to increase the residual carbonation conversion and reduce the rate of the natural deactivation of CaO due to sintering phenomena.
According to the literature [7,83,84], the progressive decrease of surface due to sintering leads to a dramatic loss of CaO multicyclic during cyclic operations. In particular, when performed under sound-assisted conditions, the fine CaO sorbent shows a residual conversion as high as X20 = 0.55 (X at the 20th cycle), which is remarkably higher than the value obtained with coarser limestone particles (X20 = 0.41 [7,47]) and also with CaO stabilized with sintering inhibitors (X20 = 0.46 [7,8]) at TCES-CSP operating conditions. This experimental evidence confirms that the employment of fine CaO particles can remarkably increase the carbonation performances, due to the hindrance of both sintering and pore-plugging phenomena. Besides this, it is also noteworthy that the results obtained using fine CaO particles in the SAFB are remarkably better than those reported for fine natural limestone tested in thermobalances [7,47]. This further confirms the capability of the SAFB configuration of getting the best performances out of fine CaO particles.
In regards to the calcination step, few calcination tests have been performed in the Seville experimental apparatus [37]. It was found that, even though the calcination of CaCO 3 takes place at an already quick rate at temperatures as high as 900 • C under ordinary fluidization conditions, the application of sound makes it possible to accelerate it further.

Conclusions
The present manuscript presents the sound-assisted fluidized bed reactor (SAFB), designed and set-up in Naples. In particular, the main results obtained using the SAFB are reviewed, highlighting how the acoustic affects the fluid dynamics of the system and the performances/outcomes of the specific chemical processes performed therein.
The fluidization behavior of different nano-sized and micro-sized powders has been investigated. The application of an acoustic field, characterized by proper f and SPL, makes it possible to either achieve an appropriate fluidization regime or enhance the fluidization quality of the powders, regardless of the specific particle size of the tested materials. In particular, while under ordinary conditions the fluidization quality is particularly poor and heterogeneous, the bed is completely fluidized, with ideal-like fluidization curves and higher values of expansion ratio, when the acoustic perturbation is applied. The effect of the acoustic parameters, i.e., SPL and f, and of the temperature has been highlighted:

•
Higher SPLs lead to a general enhancement of the fluidization quality (i.e., lower values of u mf and d agg ), whereas, an optimum range of frequency (providing minimum values of u mf and d agg ) can be found in the proximity of to the natural frequency of the powder.

•
Increasing temperatures negatively affect the fluidization quality due to the intensification of IPFs. Regardless of the powders being A or C type, the fluidization parameters, u mf and d agg , are increased when the temperature is increased.
The SAFB has been used in a TSA configuration (TSA-SAFB), demonstrating that fine/ultrafine powdered sorbents, regardless of their specific chemical nature, can be directly used as free-flowing powders in the TSA-SAFB also with a remarkable improvement of their adsorption/desorption performances with respect to their use in other standard reactor configurations (e.g., ordinary fluidized beds and fixed beds). The use of a proper acoustic field remarkably enhances both CO 2 adsorption performances, in terms of higher values of equilibrium adsorption capacity (q e ), BT time (t b ) and fraction of bed utilized at t b (ψ), and the CO 2 desorption performances, in terms of higher values of CO 2 recovery (R) and CO 2 purity (C m ), and shorter desorption times (t d ). The effect of some important operating variables on the adsorption/desorption performances has been pointed out: • Increasing values of the adsorption temperature (T ads ) negatively affects the CO 2 adsorption thermodynamics, i.e., higher values of T ads lead to lower q e . On the contrary, increasing values T ads positively affect the CO 2 adsorption kinetics, i.e., higher values of T ads lead to lower values of ∆t.

•
Increasing values of the CO 2 partial pressure (P CO 2 ) positively affects the CO 2 adsorption thermodynamics, i.e., higher values of P CO 2 lead to higher q e . Likewise, increasing values P CO 2 positively affect the CO 2 adsorption kinetics, i.e., higher values of P CO 2 lead to lower values of ∆t.

•
Increasing values of the desorption temperature (T des ) positively affects the CO 2 desorption kinetics, i.e., for a fixed R, higher values of T des lead to shorter t d and higher C m .

•
Increasing values of Q p also positively affects the desorption rate, i.e., for a fixed R, higher values of Q p lead to shorter t d and higher C m .
The SAFB has been proposed in a CaL configuration, CaL-SAFB, as a possible technology for the handling and processing of fine/ultrafine CaO/CaCO 3 particles aiming at overcoming the shortcomings/limits of the standard CaL, in the framework of both CCS and TCES application. The results obtained shows that, when the acoustic field is applied, the carbonation performances are remarkably enhanced, regardless of the specific operating conditions of CCS and TCES. In particular:

•
The beneficial effect of the acoustic perturbation is much stronger in the early fast stage of the carbonation, when the reaction is mainly governed by the reaction of CO 2 with the CaO surface and the sound can improve the gas-solid contact efficiency.

•
The acoustic perturbation also leads to the increase of the residual carbonation conversion and to the reduction of the rate of the natural deactivation of CaO due to sintering phenomena.

•
When performed under sound-assisted conditions, the fine CaO particles can provide a residual conversion remarkably higher that the values reported for coarser limestone and even for CaO stabilized with sintering inhibitors.
Author Contributions: Conceptualization, F.R. and P.A.; methodology, F.R. and P.A.; investigation, F.R. and P.A.; writing-original draft preparation, F.R.; writing-review and editing, F.R. and P.A. All authors have read and agreed to the published version of the manuscript.