Effect of nanobubbles on powder morphology in the spray drying process

This study investigated the influence of gas-injected nanobubbles on the morphology of particles during spray drying under various experimental conditions. The nanoparticle tracking system was used to measure the generation, size, and concentration of nanobubbles. Experiments were conducted at different temperatures (160–260°C) and feed rates (0.2–0.26 g/s) to examine the impact of nanobubbles on spray drying and present diverse results. The DI water with generated nanobub-bles had a particle concentration of 1.8 × 10 8 particles/ ml and a mean particle size of 242.6 nm, which was approximately 3.31 × 10 7 particles/ml higher untreated DI water. The maltodextrin solution containing nano-bubbles also showed a significant increase in particle generation, with a concentration of 1.62 × 10 9 particles/ ml. The viscosity of the maltodextrin solution containing nanobubbles decreased by approximately 18%, from 9.3 mPa•s to 7.5 mPa•s. Overall, the size of the generated particles was similar regardless of nanobubble treatment, but there was a tendency for particle size to increase under specific temperature (260°C) and feed flow rate (0.32 g/s) conditions. Furthermore, it was observed that the hausner ratio significantly varied with increasing temperature and feed flow rate, and these results were explained through SEM images. These findings confirm that the gas nanobubbles mixed in the feed can exert diverse effects on the spray drying system and powder characteristics depending on the operating conditions. This study suggests that nanobubbles can contribute to a more efficient process in spray drying and can influence the morphological characteristics of particles depending on the spray drying conditions.


INTRODUCTION
Drying is one of the oldest and most common unit operations of various food and chemical processes, which removes the moisture or solvent by evaporating, sublimating, or freezing the raw material using hot air, gas, steam, or electromagnetic waves to form dried products (Ekechukwu, 1995, Jain et al., 2012, Bhandari, 2015).Drying enables volume reduction and long-term storage of the desired raw material, making it an essential process for storing and transporting intermediate goods and final products in all industrial fields, such as chemicals, food, textiles, medicine, paper, wood, electronics, metal, and waste.Although various drying processes and dryers are available, spray drying has widely been used to produce powders from liquids in which solids and solvents are mixed.In this method, a powder is processed or manufactured by spraying a solid-containing liquid in the form of fine water droplets in a high-temperature drying gas atmosphere heated by various heat sources.
Maltodextrin is a widely used common material in the dairy and food industry.It is cost-effective, highly soluble, and exhibits low moisture absorption, making it suitable for application as a filler or carrier in the food and dairy industry (Turk-Gul et al., 2023).Moreover, maltodextrin is known to enhance particle size, bulk density, and moisture absorption properties (Erbay and Koca, 2015).Additionally, due to its encapsulation characteristics, maltodextrin significantly influences the particle attributes in spray drying of solutions containing maltodextrin (Minj and Anand, 2022).Maltodextrin plays a crucial role in improving the solubility and stability of dairy and food products, thus enhancing the quality of powdered products.Therefore, investigating the spray drying characteristics of maltodextrin in the dairy and food industry holds significant importance.
Hence, the spray drying process is widespread owing to the mass production of powder products.Moreover, the usage of this process steadily increases because of the increase in the application of powder products in food, pharmaceuticals, fine chemical materials, and the market.For the powder product manufactured using the spray drying process to have competitiveness,

Effect of nanobubbles on powder morphology in the spray drying process
Sang Hyun Oh,12 Sung Il Kim, 2 Younghwan Joo, 2 and Hyung Hee Cho1* the size and morphology of the spray-dried powder are essential.In addition, the quality of the powder is evaluated in terms of its reconstruction behavior and flowability (Takeiti et al., 2010, Fu et al., 2012).
The quality of the powder is mainly determined by the physical properties of the powder product, in particular its shape and size, and the usability may vary depending on the powder quality.Therefore, controlling the shape of the powder particles is critical.Several factors, such as dried material form, atomization, and operating conditions, affect the quality of the powders prepared through spray drying (Lewandowski et al., 2019).Among them, the inlet and outlet temperatures, flow rate, droplet shape by the nozzle, and properties of the raw material liquid are the main variables of spray drying that affect the particle shape.Accordingly, many studies have been conducted on the effect of these major variables on the shape of the particles (Vicente et al., 2013, Munoz-Ibanez et al., 2016, Both et al., 2020, de Souza Lima et al., 2020, Teo et al., 2021).If a liquid substance is evaporated and a solid substance is obtained from a solution in which a liquid and a solid substance are mixed, the concentration of the solid and the characteristics of the solvent have a significant influence on the morphology and encapsulation of the final powder (Walton and Mumford, 1999, Ahmed et al., 2010, Shepard et al., 2020).
Foam spray drying refers to the general method of drying liquid feed materials by incorporating gas before, during or after the drying process.It can be classified into 2 categories based on the manner and timing of gas injection: gas admixing and gas absorption methods (Hassenklöver andEggers, 2008, David Pour et al., 2022).
Various studies have been conducted on this effect; in particular, foam spray drying has been proposed.Here, a liquid mixture is applied to the spray drying process as a mixing gas (Bell et al., 1963).Foam spray drying has several advantages, such as supplying and drying liquid substances as high solid concentrates and reducing the energy cost of production.Many studies have reported that using the direct gas injection method in foam spray drying allows particle size control, increases the surface area, eliminates harmful substances, and increases the positive effects of drying efficiency (Hanrahan et al., 1962, Crosby and Weyl, 1977, Frey and King, 1986a, b).In particular, Frey and King (1986b) conducted a comparative study between gas admixing and gas absorption methods, highlighting the characteristics of volatile retentions.They reported a significant decrease in n-propyl acetate retention in the gas-desorption foam method compared with the nonfoamed method.This was attributed to the nucleation of bubbles within the droplets, which facilitated the escape of volatile compounds (Frey andKing, 1986a, Zbicinski andRabaeva, 2009).Zbicinski and Rabaeva (2009) performed various studies on foam spray drying using the direct gas injection method, including the basic characteristics of spray drying using direct gas injection and encapsulation-related studies (Zbicinski and Rabaeva, 2009, Lewandowski et al., 2019, Lewandowski et al., 2020).They analyzed a range of operating conditions to produce low, degraded powder during foam spray drying and presented results on the effect of the amount of gas injected.It was concluded that controlling the operating conditions for various drying materials and conditions is essential.However, directly injecting gas can affect the quality of the dried powder depending on the flow rate, pressure, and type of gas injected.In addition, when foam spray drying is applied to an existing device, various design changes are required with respect to the device's cost and pressure.
The gas dissolution method is an alternative to the direct injection method.Here, the gas is dissolved in a liquid material and then supplied to the spray drying process (Deotale et al., 2020).Moreover dissolving inert gas in liquid materials can also have positive effects, such as reducing the viscosity of liquid materials, which can improve productivity (Hassenklöver and Eggers, 2008).
Recently, the application of nanobubbles in spray drying has been proposed as a promising approach.It allows for efficient dissolution of gases like CO 2 and N 2 in liquids, necessitating the storage of gas-containing liquids at various pressures, ranging from atmospheric pressure to high-pressure conditions (up to 30 MPa).Notably, the impact of added gas becomes more pronounced at higher pressures.Foam spray drying, which utilizes gas dissolution, is widely employed for processing powdered products such as coffee (Hassenklöver and Eggers, 2008) and whole milk (McSweeney et al., 2021), owing to its numerous advantages.
In the study conducted by Deotale et al. (2020), nanobubbles were employed during the dissolution of spray-freeze-dried instant coffee, with a specific focus on investigating the effect of nanobubbles on the stability of the resulting foam.The researchers reported a significant improvement in the stability of the coffee foam when nanobubbles were introduced.Recent studies have shown that the utilization of micro and nano bubbles in spray drying milk protein can enhance productivity and improve the rehydration behavior of the resulting powder (Babu andAmamcharla, 2022, Babu et al., 2022).
It is crucial to differentiate between foam spray drying, which involves incorporating a substantial amount of gas into the medium, and the use of nanobubbles.Foam spray drying creates foam by incorporating a Oh et al.: Effect of nanobubbles… significant volume of gas, thereby increasing the effective volume of the feed.In contrast, the utilization of nanobubbles does not alter the volume of the medium significantly.Nanobubbles, being tiny gas bubbles, primarily enhance the stability of the system without substantially affecting its volume.
However, the currently proposed methods have several limitations.As mentioned above, processes such as optimizing additional equipment and operating conditions are required to implement foam spray drying with the direct gas injection method.In addition, to change the conventional spray drying methods, a high-pressure feed tank must be installed to maintain the liquid substance mixed with the inert gas and as the inert gas cannot be dissolved sufficiently at room temperature and pressure.Moreover, even when the liquid material is changed to foam, the time for maintaining the foam is short, making it difficult to maintain the stability of the liquid material in a constant foam form.
Thus, the present study investigated implementing foam spray drying to supplement the limitations of existing foam spray drying methods.We proposed a method of mixing foaming gas into nanobubbles with a diameter of 500 nm or less (Nirmalkar et al., 2018), which were then added to the liquid for drying.Compared with conventional bubbles, the main characteristic of nanobubbles is a decrease in their rising rate in water with decreasing bubble size because of the principle of buoyancy.Moreover, compared with that of normal bubbles, the buoyancy of nanobubbles is small; however, the surface of the nanobubbles is strong, preventing them from collapsing easily.Furthermore, according to previous studies, nanobubbles are stable in water for a long time as they move slower than the ascent rate of Brownian motion (Azevedo et al., 2016).To implement gas dissolution spray drying, nanobubbles of inert gas were added to the liquid material as a pretreatment concept.After drying, the effect of the nanobubbles on the morphology of the powder was studied.We also compared the present study with previous studies on foam spray drying and discussed the advantages and limitations of gas dissolution spray drying using nanobubbles.

Materials
Maltodextrin (DE 18.6,Qinhuangdao Lihua Starch Co. Ltd.,China) is widely used as a skin-foaming material.The feed solution concentration used in the experiment was 25% by mass, and the experimental studies were conducted using a similarly prepared basic feed solution.For nanobubbles mixing conditions, experi-mental studies were performed by adding nanobubbles to this feed solution.
The feed solution was prepared by mixing the maltodextrin solution with a stirrer at 7000 rpm for 2 h while maintaining a similar sample state and sample preparation conditions for various experiments.Then it was refrigerated for 12 h and stored at room temperature (about 25°C) for 3 h.The feed solution was also supplied at room temperature during the experiment.

Mixing of feed solution and nanobubbles
In this study, a nanobubble generation device (UFB-R201, EnH Co. Ltd., South Korea) was employed to simultaneously supply liquid substances and gas, in contrast to the conventional method of mixing gas and liquid substances.This device enabled the creation of a liquid solution containing mixed nanobubbles.The flowing fluid and the injected gas were passed through the nanobubble-generating membrane, and the injected gas was mixed with the fluid in the form of nanobubbles.The nanobubble-generating fluid was stored in a storage tank (Figure 2).
Recently, Gallo et al. (2022) conducted a study in which a gaseous supersaturated drug aqueous solution was placed in a pressurized container and then treated with spray drying.They used air, N 2 , and CO 2 microbubbles and found that the presence of microbubbles in the aqueous solution influenced the shape of the particles produced by the spray drying process.
In line with these findings, in this study, nitrogen was utilized as a reasonable mixing gas candidate because approximately 79% of the general air is composed of nitrogen.The nitrogen supplied by the nanobubble generator was set at a flow rate of 0.02 /min, with a supply pressure of 0.1 MPa.
After preparing the aforementioned maltodextrin feed solution (25 wt.%), approximately 3 L of the prepared sample was introduced into the nanobubble generator for mixing with nanobubbles.Microbubbles generally disappear within a few hours after their formation (Meegoda et al., 2018).Therefore, to isolate the effects of nanobubbles, the feed mixed with nanobubbles generated by the nanobubble generator was stored for more than 24 h to remove microbubbles before being utilized in the study.Samples were collected after a single cycle.The diameter, number, and particle size distribution of the nanobubbles within the feed solution were measured using a nanoparticle tracking analysis Oh et al.: Effect of nanobubbles… instrument (NanoSight NS300, Malvern Panalytical, UK), which has also been employed in previous similar studies (Ferraro et al., 2020, Zhou et al., 2021).The same measurement device and method used in previous studies involving the addition of nanobubbles to protein powder solutions were utilized (Babu and Amamcharla, 2022, Babu et al., 2022, Babu and Amamcharla, 2023).Initially, the nanobubble generation device was assessed using DI water to confirm the generation of nanobubbles.Subsequently, measurements were conducted on the maltodextrin solution with and without nanobubble treatment using the aforementioned measurement device, providing the respective values.
Additionally, viscosity and density measurements were carried out to evaluate changes in feed properties and the incorporation of nanobubbles.The density was calculated by measuring the volume and mass using a specific gravity bottle and a digital precision balance (XS802S, Mettler Toledo, US), respectively.The apparent viscosity was determined using a viscometer (DV-1 Viscometer, Ametek, Inc., USA) and a constant temperature circulation water tank (Stereo-Vis, Hwashin Tech Co., Ltd., South Korea).All experiments were performed in 3 times for repetition, and the reported values are presented as mean values.

Experimental facility
A single-stage lab scale spray dryer was manufactured for this study.A 2-fluid nozzle (XA-SR400, Hanmi Nozzle Co. Ltd., South Korea) was used to spray the feed solution.A spray dryer was designed in which the spray fluid and hot air were in contact in the forward direction (Co-Current).The diameter and length of the drying chamber were 388 and 800 mm, respectively.The evaporation capacity was up to 1200 g/h, and hot air was supplied up to 200°C.A peristaltic tubing pump (YZ1515x, Longer Precision Pump Co. Ltd., UK) was used to supply the feed solutions.Thermocouples were installed at the inlet and outlet to measure the temperature of the air being introduced and discharged from the spray dryer.In addition, visible windows were installed to examine the evaporation and particularization inside the spray drying device.The spray drying chamber was insulated to minimize the heat loss of the spray drying system.The dried particles were collected using a cyclone, while the uncollected particles and exhaust gas were discharged by installing an additional fan.A flowmeter (KTR-550, KOMETER Co. Ltd., South Korea) was installed at the inlet and functioned as the entrance for the hot air.Figure 1 shows a detailed schematic illustration of the spray drying test equipment and its performance.

Operating conditions of experiments
To investigate the influence of gas nanobubbles injected into a liquid feed on the morphology of spraydried particles, we conducted spray drying experiments under 3 different high-temperature air conditions (160, 200, and 260°C).The position of the thermocouple for temperature measurement is depicted in Figure 1, and the recorded values of the exit temperature for each experimental condition are presented in Table 2.
As mentioned in the research objective, experiments were performed with and without inert gas nanobubbles to observe their effect on the particle shape.The flow rate of the feed solution was conducted for 3 cases (0.2 g/s, 0.26 g/s, 0.32 g/s).Additionally, the number of nanobubbles in the feed solution mixed with nanobubbles is presented as the number per unit volume.

Particle size of powder samples
The particle size and size distribution were measured using a laser-light diffraction unit (Mastersizer 3000, Malvern Instruments Ltd., UK).All sample measurements were performed in triplicates, and the average value was used.Accordingly, various particle diameters for D 10 , D 50 , D 90 , D [4,3] , and D [3,2] and the particle size distribution for the volume (%) were presented and analyzed.

Density
The apparent bulk density of the powder was obtained by measuring the mass and volume of the powder.After carefully filling the 100-mL measuring cylinder with the powder, the weight was measured.The bulk density was calculated by dividing the mass by the volume and expressed in g/mL.
The tapped bulk density of the powder was obtained using a mechanically automated tapping machine (Autotap, Quantachrome Instruments, USA).All tapped densities were measured and presented after tapping more than 1,000 times.
The Hausner ratio is defined as the ratio of the tapped bulk density to the apparent bulk density.

Moisture contents
The moisture content of the dried powder was measured using a moisture content meter (FD-720, Kett, Japan).The moisture content was measured at 105°C for 4-6 g of the dried powder and presented in wet basis (w.b.) moisture contents.

Scanning electron microscopy
An analysis of the morphological and microstructural characteristics of the powder was conducted using scanning electron microscopy (SEM).A high-resolution scanning electron microscope (S-4800, Hitachi High-Technologies Corporation, Japan) was used for the analysis.The images were collected at 5.00 kV and presented at various magnifications.

Statistical analysis
All experiments were conducted in 3 times, and the results were presented as the mean ± standard deviation.ANOVA (One-way ANOVA; Tukey's HSD) was performed using OriginPro 2019b (OriginLab Corporation, USA).All levels of significance were set to P < 0.05.

Characteristics of nanobubble in water and feed solution
To measure the presence of nanobubble generation using a nanobubble generator, we evaluated it using NanoSight NS300.To determine the occurrence of nanobubble generation, we compared the number of particles measured with and without nanobubble treatment using DI water and maltodextrin solution.In the experiment with DI water, without nanobubble treatment, the mean particle size was measured as 169.7 nm with a mode of 85.0 nm and a standard deviation (SD) of 79.1 nm.The particle size distribution was as follows: D 10 : 91.6 nm; D 50 : 145.0 nm; D 90 : 316.2 nm.With nanobubble treatment, the mean particle size of DI water was measured as 242.6 nm with a mode of 182.2 nm and an SD of 171.3 nm.The particle size distribution was D 10 : 73.8 nm; D 50 : 200.6 nm; D 90 : 405.9 nm.Significant differences were observed in the particle counts between the 2 solutions.The particle count in DI water without nanobubble treatment was measured as 3.31 × 10 7 particles per milliliter, whereas the nanobubble-treated DI water showed a particle count of 1.8 × 10 8 particles per milliliter.While there is still uncertainty about whether all the particles measured as nanobubbles are in a gaseous state, multiple previous studies have reported the generation of nanobubbles using the same measurement method (Azevedo et al., 2016, Qiu et al., 2017).
The same experiment was performed with maltodextrin solution.The maltodextrin solution without nanobubble treatment exhibited a mean particle size of 144.2nm, a mode of 37.4nm, and a standard deviation (SD) of 248.3nm.The particle size distribution was D 10 : 33.5nm; D 50 : 40.6nm;D 90 : 606.1nm.In contrast, the maltodextrin solution treated with nanobubbles showed a mean particle size of 120.2nm, a mode of 37.0nm, and an SD of 195.1nm.The particle size distribution was D 10 : 33.1nm; D 50 : 38.2nm; D 90 : 387.7nm.Following nanobubble treatment, the maltodextrin solution exhibited a particle count of 3.37 × 10 9 particles per milliliter, whereas the particle count in the maltodextrin solution without nanobubble treatment was measured at 1.65 × 10 9 particles per milliliter.The overall decrease in particle size and increase in particle count after nanobubble treatment align with previous studies and demonstrate a similar trend (Babu and Amamcharla, 2022).Babu et al. (2022) reported changes in physical properties, such as the viscosity, of the nanobubble-added liquids.Accordingly, the viscosity and density of the maltodextrin feed solution with and without nanobubbles were measured in this study.No significant difference was observed in the density of the 25% maltodextrin feed solution with and without nanobubbles.However, the measured viscosity was approximately 18% lower in the feed solution mixed with nanobubbles.The obtained result demonstrated a trend similar to that observed in the previously mentioned previous studies by Babu et al. (2022).Another recent study also confirms these results.Babu and Amamcharla (2022) conducted a study using micro-and nanobubbles for spray drying of milk protein.They reported that, as a result of measuring the viscosity by adding nanobubbles to the milk protein solution, there was a marked decrease in viscosity compared with the milk protein solution without nanobubbles (Babu and Amamcharla, 2023).
We cannot be certain that all particles observed in the measurement results are solely attributed to the presence of nanobubbles in the solution.This has also been mentioned in previous studies (Alheshibri et al., 2021).However, as mentioned above, the effect of viscosity reduction observed in the dissolved gas or liquid containing nanobubbles, and the significant increase in particle count in the maltodextrin solution, support the expectation that the gas was mixed in the form of nanobubbles in the liquid substance in this study.

Moisture contents of the powders
Table 3 presents the final moisture content observed in all experimental cases conducted in this study.The moisture content ranged from 3% to 7.57% (w.b.).It was observed that higher temperatures resulted in lower water content and higher flow rates.The sample used in this study had a limiting moisture content of approximately 3%.
The addition of nanobubbles showed more significant results in terms of moisture content.In all cases, the inclusion of nanobubbles led to lower moisture content.Furthermore, the samples containing nanobubbles exhibited an average moisture content that was approximately 15% lower compared with the samples without nanobubbles, with a maximum reduction of approximately 33%.
In this study, it was observed that the viscosity of the liquid decreased with the addition of nanobubbles.The viscosity of a liquid is determined by the interactions between its molecules or particles.Generally, lower viscosity corresponds to higher heat transfer efficiency (Ramamurthy and Krishnan, 2022).Therefore, the decrease in viscosity associated with the addition of nanobubbles can be considered as a contributing factor to the improved heat transfer.
Moreover, previous research has reported that adding gas to the liquid feed in spray drying can enhance the drying performance.The inclusion of gas has been shown to facilitate faster drying (Zbicinski and Rabaeva, 2009).
Therefore, it can be concluded that the addition of nanobubbles to the liquid phase of the spray drying process lowers the viscosity, which is generally associated with improved heat transfer.This, in turn, leads to more efficient drying.Additionally, previous studies have reported that adding gas to the liquid phase in spray drying can expedite the drying process.As shown in the results of the particle size analysis, small and large peaks are observed in a particle area of 0.5-1 µm and ≥10 µm, respectively.This feature Oh et al.: Effect of nanobubbles… may occur in spray drying using a 2-fluid nozzle (Koç and Kaymak-Ertekin, 2014).The particle size was measured based on volume.However, more particles may exist in the particle area range of the small peak (particle area of 0.5-1 µm), assuming that the particle size is measured based on the particles count number.Because the range considered in this study was based on volume, phenomena occurring in a range of 1 µm or less are not mentioned separately.

Particle size of the powders
The volume-weighted average particle size increased in all cases when the nanobubbles were added.In particular, the maximum average particle size of the samples without nanobubbles was approximately 14.5 µm at a flow rate of 0.2 g/s, while that of the samples with nanobubbles was approximately 18.7 µm.Additionally, the maximum average particle diameter under the supply condition of 0.26 g/s was observed at 16.4 µm for samples without nanobubbles.In contrast, the maximum average particle diameter was 21.2 µm for samples with nanobubbles.Finally, the peak of the average particle size was 18.7 µm for samples without nanobubbles under a supply condition of 0.32 g/s, while that of the samples with nanobubbles was 24.1 µm.As reported in a previous study, these observations were attributed to the expansion of nitrogen nanobubbles in the liquid droplet by the high ambient temperature immediately after being sprayed from the spray nozzle (McSweeney et al., 2021).
The variation in particle size in this study can be attributed to 2 factors: a decrease in viscosity and the influence of nanobubbles mixed in the solution.As mentioned previously, the viscosity decreases when the solution is mixed with nanobubbles.In the context of spray-drying, the viscosity of the liquid material is closely associated with droplet size.It is widely known that as viscosity decreases, smaller droplets are (Premi andSharma, 2017, Gallo et al., 2022).However, in studies where certain gases were mixed with liquid materials, different phenomena were observed.For instance, in previous research involving the dissolution of CO 2 gas in liquid solutions, a decrease in viscosity was observed, yet it resulted in larger droplet sizes due to gas expansion effects (Hassenklöver and Eggers, 2008).
In this study, particle size measurements presented the volume weighted diameter D [4,3] , as shown in Figure 4; the various particle diameters for D 10 , D 50 , D 90 , and D [3,2] are illustrated in Tables 4, 5, and 6.The particle size demonstrates significant variations depending on the presence or absence of nanobubble mixing.Under the conditions of 0.20 g/s and 0.26 g/s, the particle sizes remained similar regardless of the nanobubble mixing, while in some cases, a decrease in particle size was observed when nanobubbles were mixed.This phenomenon can be attributed to the aforementioned influence of viscosity reduction.
However, under high temperature conditions (260°C), an increase in particle size was observed when nanobubbles were included.This phenomenon can be explained by a similar trend observed in studies utilizing gas mixing.In the case of a liquid mixed with nanobubbles, physical gas bubbles are incorporated into the liquid and supplied to the spray-drying process in this state.During spray drying, the liquid mixed with bubbles undergoes evaporation at elevated atmospheric temperatures.At this moment, various previous studies in different fields have reported that micro and nanobubbles can grow and expand, leading to an increase in the size of the liquid droplets (Hou et al., 2015, Wang et al., 2019).Furthermore, recent research findings suggest that nanobubbles can influence the shape of particles in the food industry (Javed et al., 2023) In this study, the addition of nanobubbles resulted in a viscosity reduction of approximately 18%, which is relatively low compared with the significant viscosity reductions of up to 65% reported in previous studies (Babu andAmamcharla, 2022, Babu et al., 2022) and approximately 30% reported in studies utilizing CO 2 dissolution (Hassenklöver and Eggers, 2008).Therefore, while the particle size of the dried particles in this study is influenced by viscosity and liquid droplet size, the impact of bubbles contained within the liquid cannot be ignored.This suggests that there is a complex interplay of factors including viscosity, feed flow rate, temperature, and other related parameters.There are limited reported cases on this study, indicating the need for further research in the future.In all cases, an increase in feed flow rate leads to larger particle sizes.Previous studies have reported that under constant drying temperature conditions, an increase in the feed solution flow rate results in an increase in average particle size, which aligns with the observations in this study (Koç and Kaymak-Ertekin, 2014).
For the maltodextrin solution, a well-known encapsulation material, spray drying typically produces hollow particles.Therefore, it is difficult to determine whether nanobubbles directly contribute to the size of the formed particles.In this study, when gas was mixed in the form of nanobubbles and used in the spray drying process, it was observed that viscosity, as well as the presence of bubbles, could have a combined effect on the particle size, and we mention the possibility of such contributions.Although a direct comparison between our study and previous research regarding the influence of nanobubbles on particle size cannot be made, it can be inferred that a similar trend is expected in our study since the liquid supplied contained mixed nanobubbles.This result was observed because, in contrast to conventional methods, the method of dissolving and mixing nanobubbles proposed in this study required less gas, and the gas was injected as fine bubbles.The expansion of the bubbles is related to the temperature and occurs rapidly in high-temperature conditions.Additionally, particles with a large diameter appeared because of the bubble expansion effect at high temperature conditions.
In the present study, no research was conducted to change the density of the gas bubbles in the liquid related to the performance of the nanobubble-generating apparatus.In previous studies, only the amount of injected gas was mentioned, and the quantitative values for the rate of gas conversion into bubbles and the effect on drying were not discussed.However, the present study quantitatively analyzed the nanobubbles introduced into the feed solution and investigated the effect of quantitatively introduced nanobubbles on spray drying and powder shape.
When comparing the effect of temperature and flow rate, the increase in temperature was found to have a greater effect on the average diameter of the particles than the increase in flow rate.This is because the expansion and temperature of the nanobubbles are closely related.

Density of the powder particles
The material used in this study was an aqueous solution of maltodextrin.Maltodextrin is a representative encapsulation material consisting of hollow particles  commonly used in drying.The powder density is closely related to the average diameter of the particles.However, the apparent and tap densities tend to be different, which is an important indicator for confirming the characteristics of the produced powder.Thus, apparent and tap densities were measured in this study (Table 7).
Apparent and tap densities highly depend on particle size and distribution.In particular, the apparent density is an index that efficiently reflects the properties related to the size of the powder particles.This is because the apparent density is determined by the volume accumulated by gravity without an external force.The apparent density decreased as the temperature increased at all flow rates of the feed solution because the size of the powder particles increased with increasing temperature.Additionally, the decrease in the apparent density of the feed solution with a high flow rate was greater than that of the feed solution with a low flow rate.This is because the particles are better formed in a feed solution with a high flow rate than in a solution with a low flow rate.Furthermore, the characteristics of the shape were more accurate.In particular, the particle size evidently increased at 260°C compared with 160°C and 200°C, while the apparent density decreased with the same tendency.
The average diameter of the particles was larger when nanobubbles were added under all conditions.This trend was also reflected in the density; the apparent density was smaller when nanobubbles were added compared with that when nanobubbles were not added.In the feed solutions with 0.2 and 0.26 g/s flow rates, 160°C and 200°C exhibited only a small difference in density irrespective of the presence or absence of nanobubbles.However, the effect of particle expansion due to nanobubbles was significant at 260°C, resulting in a large difference in density.In the feed solution with a flow rate of 0.32 g/s, the apparent density of the nanobubbles at all temperature ranges was lower than that of the feed solution without nanobubbles.This observation was attributed to the higher absolute  number of nanobubbles in the feed solution with a flow rate of 0.32 g/s.
When compared under constant temperature, in the absence of nanobubbles, the apparent density increased with an increase in the flow rate.The obtained results are similar to those reported in a previous study wherein the density increased with an increasing flow rate (Babu and Amamcharla, 2022).For all cases, the feed solution with nanobubbles demonstrated lower apparent and tap densities than the feed solution without nanobubbles; this phenomenon was more evident at 260°C.This is because when drying is performed at lower temperatures (160 and 200°C), the drying process is completed before the particles can sufficiently expand.
A similar phenomenon was also observed for the tap density.The density decreased as the temperature increased, while the tap density increased as the flow rate increased.However, the difference in density according to the presence or absence of nanobubbles was lower than that observed in the apparent density.This is because, in the tap density, the particles are more densely stacked by an external force; hence, the small particles are either mixed with the hollow particles or densely located between the large particles.
The difference between the tap and apparent densities in dependence on the presence or absence of nanobubbles was negligible at 160°C and 200°C.This can be explained by the smaller difference in tap density at 160°C and 200°C due to the lower effect of the nanobubbles.However, the difference between the tap and apparent densities became larger at 260°C.This is because particles with a large average diameter are easily generated under high-temperature conditions, and the uniformity of the particle size is lowered, resulting in a larger difference between the apparent and tap densities.
At constant temperature, the tap density increased with an increasing flow rate of the feed solution.However, this phenomenon is insignificant compared with the trend in apparent density owing to the nature of the tap density measurements and properties of maltodextrin particles.When maltodextrin is dried, it produces a powder with hollow particles consisting only of an outer shell.Hence, depending on the drying conditions, the expansion of the nanobubbles can burst the particles, and cracking is more significant when larger particles are formed.Therefore, the difference between the tap and apparent densities becomes larger as the flow rate increases.

Hausner ratio
The Hausner ratio, which is determined by the apparent and tap densities, explains the fluidity of the particles.The Hausner ratio for all experimental conditions is shown in Figure 5.
The particle flowability was improved when nanobubbles were added to the feed solution with a low flow rate (0.2 g/s).However, the particle flowability decreased as the flow rate of the feed solution and drying temperature increased; at a flow rate of 0.32 g/s, the Hausner ratio increased from 1.3 to 1.9.
Moreover, in most conditions, the Hausner ratio increased with temperature.This result indicates that the expansion effect of the nanobubbles in the feed solution increases as the feed solution flow rate and spray drying temperature increase, which is in accordance with previous studies (Lewandowski et al., 2019, Mc-Sweeney et al., 2021).In addition, particle flowability can be controlled by controlling the injected amount of nanobubbles, drying temperature conditions, and concentration of the aqueous solution.

SEM imaging
The obtained results were further investigated through SEM, as shown in Figure 6.The particles did not have a perfectly spherical shape under the low-temperature drying conditions of 160°C and 200°C, indicating that droplets were contracted during the drying process.The effect of the nanobubbles was relatively low, which affected the particle size and the apparent and tap densities.
In contrast, the shape of the generated particles was close to a spherical shape under the high-temperature drying condition of 260°C.It should be noted that a particle bursting phenomenon may occasionally occur during the 260°C drying condition.As shown in the Figure 6, several particles ruptured, resulting in a large difference between the apparent and tap densities observed under the 260°C drying condition.
A similar phenomenon was also reported in a previous studies (Lewandowski et al., 2019, Babu andAmamcharla, 2023).Compared with the present study, a much higher particle destruction rate was demonstrated in the previous study (Lewandowski et al., 2019).This is because, in the present study, the amount of inert gas added as nanobubbles is significantly lower than that reported in the previous study, wherein the gas was directly injected.However, recent SEM images of powder particles treated with nanobubbles have also reported the presence of spherical-shaped particles with complex internal structures.This finding closely resembles our own results (Babu and Amamcharla, 2023).Moreover, in this study, a smaller amount of gas was quantitatively injected for mixing nanobubbles with droplets compared with the conventional method of direct gas injection.Therefore, spray drying using nanobubbles can control the size of the particles and minimize their destruction.

CONCLUSION
In this study, we conducted a spray drying investigation by incorporating nitrogen as nanobubbles into a maltodextrin solution.Various case studies were performed to investigate the influence of the injected nanobubbles on the particle morphology during spray drying.The nanobubbles exhibited a significant impact on the shape and density of the particles.Through this study, we have confirmed that gas injected in the form of nanobubbles can directly affect the morphology in spray drying while also contributing to increased production rates through the influence of viscosity reduction and enhanced energy efficiency in gas-assisted spray drying.However, a comprehensive analysis of the effects of nanobubbles in spray drying necessitates further diverse research in the future.Nonetheless, our study proposes a gas-assisted method utilizing nanobubbles, which presents the potential to employ conventional gas utilization techniques in spray drying technology cost-effectively.
Figure 1.Schematic illustration of the lab-scale spray drying facility.
Figure 2. Schematic of nanobubble generator and nanobubble image, and the measurement result sample.

Figure 3
Figure 3 illustrates the particle size distribution according to the drying temperature in each feed solution at different flow rates.As shown in the results of the particle size analysis, small and large peaks are observed in a particle area of 0.5-1 µm and ≥10 µm, respectively.This feature
Oh et al.: Effect of nanobubbles…

Figure 5 .
Figure 5. Hausner ratio of maltodextrin powder with (solid symbol) and without (empty symbol) nanobubble treatment as a function of flowrate and temperature.

Table 2 .
Oh et al.: Effect of nanobubbles… Experimental conditions

Table 1 .
Properties of maltodextrin solution with and without nanobubbles Oh et al.: Effect of nanobubbles…

Table 3 .
Moisture content of the particles as flowrate, temperature NB X (without nanobubble), NB O (with nanobubble).

Table 4 .
Particle size distribution parameters at 160°C

Table 5 .
Particle size distribution parameters at 200°C

Table 6 .
Particle size distribution parameters at 260°C

Table 7 .
Apparent and tap densities of the particles as flowrate, temperature NB X (without nanobubble), NB O (with nanobubble).a-c Values within columns not sharing common superscripts differ significantly (P < 0.05).