Introduction

Three-dimensional (3D) bioprinting can facilitate the flexible design of 3D structures of biomaterials by printing bioinks, enabling the design of new biomaterials such as meat analogues, artificial organs, pharmaceuticals, and cosmetics1,2,3. 3D bioprinters are equipped with extrusion, inkjet, stereolithography, and lasers, which are selected based on the printing speed, structural controllability, textural properties of the ink, and temperature control during printing. Most food-grade bioinks are prepared by the extrusion and stacking of highly viscous pastes. Bioinks include agarose, alginic acid, chitosan, collagen, gelatin, graphene, hyaluronic acid, and hydroxyapatite, all of which allow mild temperature conditions for gel formation and printing4,5,6. After printing, the bioink is further stabilised by crosslinking with ionic solutions, heating, or ultraviolet (UV) radiation.

Particularly in food applications, 3D bioprinters can enhance the flexibility of food design and reproducibility of recipes1,2,7,8,9,10,11. The appearance of foods associated with dysphagia is vital for appetite stimulation. Consequently, gel-based foods are essential for improving the quality of life in patients with dysphagia12,13. Hydrogels are dispersion systems consisting of proteins and polysaccharides that are cross-linked to form a three-dimensional network structure14,15,16. Hydrogels are widely used in foods to treat dysphagia, as owing to their cohesive nature, they do not remain in the pharynx and pass smoothly. For instance, agar and κ-carrageenan gels enhance texture, regulate flavour release, and encapsulate bioactive compounds17. In particular, emulsion gels are intriguing for dysphagia diets because of their texture and aromatic release behaviour, simultaneously offering easier swallowing and a better flavour13,18,19,20. In addition, emulsions can encapsulate both water- and oil-soluble functional ingredients, enhancing the nutritional value for individuals with dysphagia12. Hence, the 3D printing of emulsion gels can improve the appearance and nutritional factors of dysphagia diets21,22.

Dielectric heating using radiofrequency (RF) and microwaves (MWs) promotes thermosetting gel formation. RF generally refers to frequencies between 3 and 300 MHz, whereas MWs are between 300 MHz and 30 GHz. They are widely used in the heating, sterilisation, thawing, and processing of food in industry as well as in household microwave ovens23,24. Dielectric heating occurs because of the conversion of electromagnetic wave energy into thermal energy inside dielectric materials. The electric field of microwaves induces molecular motion through orientation polarisation in dipolar molecules, such as water and alcohol, and the vibration of ions, which greatly depends on the dielectric properties of the material and the applied frequency25. The dipoles are more affected in the MW range than in the RF range. In contrast, ion vibrations occur in the RF range. Dielectric heating has been used to enhance the gelation of hydrogels, such as those composed of polysaccharides, bovine serum albumin, ovalbumin, egg white, and wheat gluten proteins, by promoting their aggregation and cross-linking26,27,28,29,30,31,32. Solid-state microwave generators are expected to exhibit higher energy efficiencies and more uniform microwave heating than classical magnetron-based kitchen microwaves.

This study demonstrates the facile formation of emulsion gels by RF and MW heating and their application to 3D printing. The effects of dielectric heating at different frequencies on preparing a model emulsion gel consisting of egg white protein, xanthan gum, MgCl2, and canola oil were analysed. In addition, a facile RF/MW 3D bioprinter was developed to perform 3D printing of emulsion gels by equipping a LEGO 3D bioprinter with Mindstorms EV334. RF and MW were used to instantaneously solidify the emulsion gel ink for 3D printing and to control its textural properties by adjusting the frequency, power, and flow rates.

Experimental

Materials

The emulsion gel ink was prepared using egg white albumin (Nacalai Tesque, Inc.), xanthan gum (Tokyo Chemical Industry Co., Ltd.), and MgCl2 (Nacalai Tesque, Inc.). Canola oil (Nissin Oillio Co. Ltd.) was used as the dispersed phase. Polyoxyethylene sorbitan monooleate (Tween 80, Wako Pure Chemical Industries, Ltd.) was used as emulsifiers.

Preparation of emulsion gel

Premix membrane emulsification was used to prepare monodispersed emulsions. Membrane emulsification using a porous glass membrane enables a better control of the dispersed particle size of the emulsion for the preparation of monodisperse emulsions35,36. Premix membrane emulsification was performed using a Shirasu porous glass (SPG) membrane. The continuous phase consisted of an aqueous solution of egg white protein whereas the dispersed phase consisted of canola oil in a ratio of 3:2 (continuous phase : dispersed phase). A crude emulsion was obtained by using a homogeniser (AHG-160D; As One Co., Ltd.). The emulsion was then refined by extruding the dispersion using membrane emulsification through an SPG membrane (10.0 µm, SPG Techno Co., Ltd.) using a high-flow plunger pump (NP-GXL1000, Nippon Precision Science Co., Ltd.).

The emulsion gel ink was prepared by the addition of egg protein (7.5 wt%) to xanthan gum (0.5 wt%) and Tween 80 (1 wt%) solution. Egg protein and xanthan gum were used as gelling agents to enhance the emulsion stability. Then, MgCl2 (2 wt%) was added to the mixed solution (Figs. S1, S3, S4). This solution was transferred to quartz test tubes and heated for 10 min using conventional heating (CH), RF, or MW irradiation to produce emulsion gels (Fig. 1A).

Fig. 1
figure 1

(A) Schematic of emulsion gel ink preparation and (BD) RF and MW heating devices: (B) 200 MHz, (C) 915 MHz, and (D) 2.45 GHz.

RF heating was performed using an RF generator (MR-0.2G-100-AC, Ryowa Electronics) equipped with a wideband semiconductor amplifier (ZHL-100W-GAN+, Mini-Circuits, frequency range of 20–500 MHz, max. output 100 W) and a parallel plate applicator (Fuji Electric Industrial Co., Ltd. Fig. 1B)37. For MW heating at 915 MHz, a TM010 mode cavity resonator and semiconductor microwave oscillator (MR-0.9G-300, Ryowa Electronics) were used (Fig. 1C). A TM110 mode ellipsoidal cavity resonator (Chronix, Inc.)38 and a semiconductor microwave oscillator (MWPS-2450050-01, Plasma Applications) were used for MW at 2.45 GHz heating (Fig. 1D). The temperature during heating was measured using an optical fibre thermometer (FL-2000, Anritsu Meter Co., Ltd., Fig. S2). A water bath was used for CH. The temperature was increased to the target temperature of 70 °C (200 MHz; 0–40 W, 2.45 GHz; 0–7 W). The temperature was maintained at a constant value by adjusting the power.

Characterisation of the emulsion gel

Compression tests (two-bite tests) were performed twice on the emulsion gel using a texture analyzer RE-3305S (Yamaden Co., Ltd.) to measure the hardness (N/m2), adhesiveness (J/m3), and cohesiveness (–) with a sample thickness of 10 mm and plunger diameter of 3.0 mm. The strain and crosshead speed were set to 70% and 10.0 mm/s, respectively. Hardness represents the maximum force during the first compression. The adhesion value was obtained using the force required to pull the plunger from the gel. Cohesiveness indicates the ease with which a material holds together. It is given by the ratio of the energy required for the first and second compressions of the emulsion gel by the plunger.

The water holding capacity was obtained by the ratio weight after centrifuging the emulsion gel (5 mL) at 10,000 g and 4 °C for 10 min and that of the whole emulsion gel before centrifugation using the following Eq. (1).

$$Water\;holding\;capacity {\text{\% }} = \frac{Weight\;of\;precipitate}{{Weight\;of\;whole\;emulsion\;gel}} \times 100$$
(1)

The surface morphology of the emulsion gel was characterised via cryoscanning electron microscopy (cryo-SEM; SU3500, JEOL) using a cold stage. The nanoscale mesh structure of the gel was evaluated using dynamic light scattering (DLS; Partica LA-960V2, HORIBA, Ltd.) by dispersing the emulsion gel in deionised water. The gel mesh size was calculated using the Smoluchowski equation:39 The denaturation of proteins in the emulsion gel was evaluated using differential scanning calorimetry (DSC, X-DSC700, Hitachi High-Tech Corporation) using a sealed aluminium pan with a temperature program of 2 °C/min within the range of 30–100 °C under N2 flow (50 mL/min). Thermograms were obtained using an empty pan as standard.

The microwave absorption properties of the emulsion gel components were measured as complex permittivities using the coaxial probe method with N1500A and P9371A network analysers (Keysight Technologies, CA, USA). The dielectric tangent (tan δ) of the samples was calculated from the real (ε′) and imaginary (ε″) parts of the complex permittivity (Fig. S5).

$${\text{tan }}\updelta =\upvarepsilon^{\prime \prime} /\upvarepsilon^{\prime}$$
(2)

Development of MW 3D bioprinter

A 3D bioprinter using LEGO Mindstorms EV3 was combined with the MW and RF devices (Figs. 2 and S6). The 3D bioprinter deposits the emulsion gel ink in a 2D pattern by moving the X/Y stage at a speed proportional to the ink flow rate. The 3D structure was formed sequentially stacking the 2D patterns using a Z-stage. The stage and nozzle were programmed using LEGO Mindstorms Education EV3 software (Fig. S7). The emulsion gel ink was extruded using a syringe pump through a silicon tube with an internal diameter of φ2 mm. MW at 2.45 GHz (TM010 mode, MR-2.45G-100, Ryowa Electronics, Fig. 2A)40,41 or RF of 200 MHz (MR-0.2G-100-AC, Ryowa Electronics, Fig. 2B) were applied to the ink to form an emulsion gel as the ink flowed inside the silicon tube before being printed onto a glass dish placed on the stage. The flow rate of the syringe pump and the RF/MW power were varied to control the temperature of the ink and optimise its gelation and printing conditions. Namely, the flow rate of the syringe pump was set to 100, 200, 500, or 1000 µL/min, the microwave output was set to 8–12 W, and the stage speed of the 3D bioprinter was set to 0.5 cm/s. The temperature at the outlet of the RF/MW applicator was measured by thermography (Xi400, Optris GmbH & Co. KG).

Fig. 2
figure 2

Schematics of 3D bioprinter using (A) MW at 2.45 GHz and (B) RF at 200 MHz.

Results and discussion

Dielectric heating controlled the textural property of emulsion gel

The gelation behaviour of the emulsion gel ink was initially evaluated using batch RF and MW heating to understand the effects of electromagnetic waves on the formation of the emulsion gel. Figure 3 shows the results of the textural properties and water-holding capacity of the emulsion gel obtained at 70 °C using RF at 200 MHz, MW at 915 MHz, or MW at 2.45 GHz. The textural properties of emulsion gels were evaluated according to the criteria for dysphagia foods in Japan, within a range of stages I (2.5 × 103–1 × 104 N/m2), II (1 × 103–1.5 × 104 N/m2), and III (3 × 102–2 × 104 N/m2) depending on their hardness, to provide optimum dysphagia food for individuals (Fig. 3A, Table S1). The hardness of the emulsion gel increased with decreasing frequency; in particular, RF at 200 MHz was effective in improving the gel hardness four-fold compared with CH. The hardness of the emulsion gel obtained using RF (200 MHz) satisfied Stage III, whereas those obtained using MW (915 MHz and 2.45 GHz) and CH satisfied Stage I. The water-holding capacity of the emulsion gel showed a behaviour similar to that of hardness (Fig. 3B). The water-holding capacity improved with decreasing frequency, and using RF at 200 MHz, it reached 1.4 times that of CH. In contrast, adhesiveness increased with increasing frequency. The adhesiveness at 2.45 GHz was almost twice that at 200 MHz (Fig. 3C). However, there was no significant difference in cohesiveness between the CH, MW, and RF groups (Fig. 3D). These results show that RF and MW irradiation change the textural properties of the emulsion gel depending on the applied frequency. The hardness and water-holding capacity of the emulsion gel improved, particularly at 200 MHz, while 2.45 GHz enhanced the adhesiveness. Thus, the texturall properties of the emulsion gel can be controlled by varying the incident microwave frequency. In addition all the prepared emulsion gels satisfied the standard criteria for dysphagia diets in Japan (Table S1).

Fig. 3
figure 3

Textural properties and water holding capacity of the emulsion gel prepared by CH, RF at 200 MHz, MW at 915 MHz, and MW at 2.45 GHz. (A) Hardness, (B) water-holding capacity, (C) adhesion, and (D) cohesiveness. Error bars indicated mean value ± standard deviation (n = 3).

Subsequently, the microstructures of the emulsion gels formed via RF and MW irradiation were characterised. Figure 4A shows the cryo-SEM images of the emulsion gels. The average fibre thickness reported as the meane of 20 points was measured from the image as 40 ± 8.51 µm (CH), 133 ± 6.67 µm (RF at 200 MHz), 75 ± 4.85 µm (MW at 915 MHz), and 67 ± 4.06 µm (MW at 2.45 GHz), respectively. These results revealed that RF at 200 MHz formed thicker fibres than CH, MW at 915 MHz, MW or at 2.45 GHz. Therefore, the thicker fibres formed by RF at 200 MHz contributed to the improvement in the strength and water-holding capacity of the emulsion gel. Figure 4B shows the nanometre-scale mesh size of the gel network analysed using DLS. The gel mesh size created by the RF at 200 MHz was approximately 2.5 nm. The mesh size gradually increased with frequency. This indicates a dense gel network structure was formed using RF irradiation at 200 MHz, whereas MW at 2.45 GHz produced a more porous structure. This also contributes to the observed trend of increased water-holding capacity with decreasing frequency. The reduced mesh size at 200 MHz results in water being retained within the network structure in smaller cavities, affording an increased surface-area-to-volume ratio of the aqueous cavities. This greater contribution of surface forces retains water with the gel mesh under centrifugation.

Fig. 4
figure 4

Characterization of gel emulsion prepared by CH, RF at 200 MHz, MW at 915 MHz and MW at 2.45 GHz. (A) Cryo-SEM images, (B) gel mesh size, (C) DSC curves, (D) dielectric property of constituents of emulsion gel. Error bars indicate mean value ± standard deviation (n = 3).

The degree of egg protein denaturation in the emulsion gel was analysed using DSC (Fig. 4C). Before heating, the ungelled emulsion gel ink exhibits two endothermic peaks. This was attributed to ovotransferrin and ovalbumin in egg white protein42,43. The peak at 61 °C is due to the denaturation of ovotransferrin, and that at 75–80 °C is due to the denaturation of ovalbumin. Heating with RF at 200 MHz and MW at 915 MHz resulted in no endothermic peaks in the DSC, indicating that ovotransferrin and ovalbumin were denatured. On the other hand, the samples heated using CH and MW at 2.45 GHz exhibited endothermic peaks above 70 °C, indicating the presence of ovalbumin, which was not completely denatured. This is because the heating temperature (70 °C) was below the thermal denaturation temperature of ovalbumin. The un-gelled emulsion gel ink required 0.51 mJ/mg for denaturation of ovalbumin, while that for CH and MW at 2.45 GHz were 0.30 mJ/mg and 0.02 mJ/mg, respectively. This indicated that partial denaturation of ovalbumin occurred at a MW of 2.45 GHz. These results suggest that heating by RF at 200 MHz and MW at 915 MHz promotes the formation of a denser gel network structure owing to enhanced protein denaturation and aggregation. This is consistent with the microstructures of the emulsion gels observed by cryo-SEM and DLS. RF heating at 200 MHz promoted protein aggregation at temperatures lower than the denaturation temperature of ovalbumin and formed a strong gel. However, MW heating at 2.45 GHz denatured ovotransferrin more selectively, resulting in a highly adhesive porous gel with larger gel mesh microstructures.

Complex permittivity measurements were performed to analyse the mechanism by which the textural properties of the emulsion gel varied depending on frequency. Figure 4D and S5 show tan δ and complex dielectric permittivity of the components of emulsion gel ink. For pure water, tan δ increased as the frequency increased. This is attributed to the dielectric relaxation of water. On the other hand, the values of tan δ for the emulsion gel ink, MgCl2 aqueous solution, and egg protein were higher at lower frequencies than that for water. This is attributed to the conduction of ionic species. We could not obtain an emulsion gel without MgCl2 even with RF heating (Fig. S4); ionic MgCl2 was highly selectively affected by 200 MHz RF, promoting the endothermic thermal denaturation of ovalbumin and ovotransferrin. Aggregation of egg proteins was enhanced by RF irradiation with MgCl2, which improved the strength and water-holding capacity of the emulsion gel.

MW and RF 3D bioprinting of emulsion gel

A 3D bioprinting technique was developed that harnessed RF and MW to control emulsion gel formation. A LEGO-based bioprinter equipped with a nozzle on the Z-stage was used to eject the emulsion gel ink onto the XY stage34. The emulsion gel ink was flowed using a syringe pump and heated by passing it through a cavity resonator-type microwave device at 2.45 GHz, TM010 mode (Fig. 2A), or a parallel plate-type RF device (Fig. 2B). Figure 5A shows photographs and thermograms of the emulsion gel ink flowing through a silicon tube placed in the MW and RF applicators. By monitoring the outlet temperature using thermography, we confirmed that the emulsion gel ink was efficiently heated while passing it through the MW and RF devices. The effect of the emulsion gel ink flow rate was evaluated using MW at 2.45 GHz and 10 W (Fig. 5B). At 1000 µL/min, the temperature did not reach 61 °C, at which ovotransferrin in the emulsion could start to coagulate. In contrast, much lower flow rates caused overheating and the emulsion gel ink shrank inside the tube. Therefore, the optimal flow rate was set at 500 µL/min, allowing the gel to reach the target temperature for ovotransferrin denaturation without overheating. Figure 5C shows the outlet temperatures and photographs of the printed emulsion gels at different MW power levels. This temperature was higher than the denaturation temperature of ovotransferrin. The MW heating at 11 W or less resulted in good printability. 3D printing was achieved at an RF frequency of 200 MHz. Figure 5D shows the outlet temperature and a photograph of the emulsion gel printed at different powers. Emulsion gels with good printability were obtained at 11 W or less; however, the ink flow became intermittent at 12 W. Figure 5E shows the textural properties of the emulsion gels were produced by the MW and RF 3D bioprinters. The printed emulsion gels were within the approval stages for dysphagia diets. It was found that the heating by RF at 200 MHz improved the hardness compared to MW at 2.45 GHz (Fig. 5E). Notably, the textural properties of 3D printed emulsion gels were consistent with the gel formation behaviour observed in the batch process (Fig. 3A). Hence, the control of the textural properties of emulsion gels prepared using MW and RF 3D bioprinters was demonstrated, with consistent textural properties obtained by a batch process. The 3D printed emulsion gels also met the requirements for textural properties of dysphagia diet (Table S1).

Fig. 5
figure 5

MW and RF 3D bioprinting of emulsion gel. (A) Photographs and thermograms by 2.45 GHz (MW) and 200 MHz (RF) heating. (B) Temperature of emulsion gel by different flow rates during MW 3D bioprinting (MW power; 10 W). Appearance and temperature of emulsion gel at various power levels after (C) MW and (D) RF 3D bioprinting (Flow rate; 500 μL/min). (E) Hardness of 3D printed emulsion gel by MW and RF heating at different power levels (Flow rate; 500 μL/min). Error bars indicate mean value ± standard deviation (n = 3).

This study demonstrates that RF and MW irradiation at different frequencies can effectively control the textural properties of emulsion gels, composed of egg white protein, xanthan gum, MgCl2, Tween 80, and canola oil. It was found that RF at 200 MHz effectively improved the strength and water retention of the gel. This was attributed to the promotion of ovalbumin and ovotransferrin aggregation, and the formation of thick fibrous aggregates. In contrast, MW at 2.45 GHz was more selective for denaturing ovotransferrin, forming a highly adhesive gel. A complex permittivity test revealed that RF was efficiently adsorbed by MgCl2, promoting endothermic protein denaturation and the development of a network structure of protein fibres. In addition, a 3D bioprinter equipped with an RF and MW was developed, using which 3D printing of emulsion gels was achieved that met the approval criteria for dysphagia diets. Furthermore, the emulsion gel prepared using the RF/MW 3D bioprinter exhibited textural properties that were equivalent to those prepared using batch processes. These results indicate that the properties of the emulsion gel can be controlled by changing the irradiation frequency, power, and flow rate of the gel ink. In addition, the 3D printed emulsion gel satisfied the standard criteria for dysphagia diets. Our results will lead to the design of textural properties and 3D structures that meet the criteria for a dysphagia diet. Nutrients and flavours can be added to enhance the functionality of 3D-printed foods in dysphagia diets. This process can be applied to artificial meat, cultured meat, and candies.