Flask culture is one of the most widely used methods in bioprocess research [1-5]. Prior to the full-scale reactor-level study, flask cultures have been utilized mainly in small-scale experiments that require the screening for many variables. Such experiments include isolation of microbial strains, design of medium and optimization of culture conditions. There are various kinds of flasks depending on size, and shape of the baffle [3]. Among them, baffled flasks, which are generally known to have high gas-liquid mass transfer characteristics by generating strong turbulence [4], have been widely used in most of aerobic cultures except for microorganisms sensitive to shear force.

Recently, gases such as methane (CH₄), carbon monoxide (CO) and carbon dioxide (CO₂) have attracted much attention as useful carbon substrates in microbial fermentations for the production of biochemical and biofuels [6-10]. Many researchers have conducted studies to screen microorganisms that can assimilate those gases and to establish the culture conditions. Since the solubility of CH₄ and CO gases in aqueous solution is very low, the performance of the CH_4- or CO-dependent gas fermentation is directly linked to gas-liquid mass transfer characteristics [8]. Therefore, it is essential to develop efficient flasks with high gas-liquid mass transfer rate even in smallscale experiments. This is likely to assist in the development of strains for gas fermentation, investigating physiological characteristics of developed strains, and optimizing culture conditions.

In this study, with the consideration of turbulence generation, various types of baffled flasks were designed and fabricated using 3D printer. Afterwards, the experiment for CO₂ absorption into the alkaline water was conducted in the prepared flasks to evaluate which types of flasks are suitable for efficient gas-liquid mass transfer. Finally, the effect of mass transfer of gas on microbial culture was investigated from the cultivation of Escherichia coli cells in flasks.

2.1. Design and fabrication of flasks

Ten kinds of baffled flasks with different shapes were designed and classified into following categories: 1) three conventional flasks with or without baffles, 2) two kinds of latticed-baffled flasks, 3) three kinds of curved-baffled flasks, and 4) two kinds of flasks with swirl type baffles (Fig. 1).

Figure 1. Flask types used in the experiments.

All flasks (600 mL volume) have same specification except the shape of baffles as shown in Fig. 2 and were designed by 3D modeling software (Solidworks 2017, Dassault systems).

Figure 2. Flask specification in this study.

The designed flasks were fabricated using 3D printer (LUGO M Printer, Formerspham Co., Korea), which includes a 0.40 mm single outlet and a heating bed. For 3-D printing, a transparent PLA (polylactic acid) filament (1.75 mm) was used as a flask material, and slicer program (ideaMaker, Raise3D, CA, USA) was used. After printing, the fabricated flasks were dried in an oven at 70°C for 15 min, and was later subjected to a test for liquid leak for 7 days. The results of the leak tests indicated that, no leakage occurred at all for 7 days in flasks fabricated under following printing conditions: layer height of 0.8 mm, 2 shell, first layer height of 0.25 mm, printing speed of 50 mm/s, a rectilinear patterned infill of 100% density, and a raft of 5 mm offset without support. Those leak-free flasks were used in CO2 absorption and microbial cultivation experiments.

2.2. Measurement of CO₂ transfer rate

In this study, titration experiments using CO₂ gas were conducted to evaluate the gas-liquid mass transfer performance of the flasks [11]. Fifty milliliters of the aqueous solution with a pH of 11 was added into the fabricated flasks and then 100 μL of phenolphthalein solution was fed into the flasks using a pipette. After confirming that the color of the solution in flask changed to purple, 20 ml of CO₂ was injected into the flasks stirred at 150 rpm in an orbital shaker (SHK-30, Jung Biotech Co., Korea). The CO₂ transfer rate was determined by measuring the time from the start of the injection until the pH of the solution in flask became 8.2 at which the color of solution became colorless.

2.3. Cultivation of Escherichia coli in flasks

Microbial flask culture experiments were carried out using Escherichia coli W3110 in LB medium which contained 10 g/L of tryptone, 5 g/L of yeast extract, and 10 g/L of NaCl. For experiments, four flasks selected from the CO₂ transfer performance experiments were used and as a control experiment, E. coli was also cultivated in conventional non-baffled Erlenmeyer flask. Flask cultures were conducted at 200 rpm and 37°C in a shaking incubator (VS-8480SF, Vision Scientific Co., Korea) for 6 h. The cell concentration was determined by measuring optical density at 600 nm (OD₆₀₀) using a UV-VIS spectrophotometer (Jenway 7300 Spectrophotometer, Cole-Parmer, UK), and the acetate concentration was measured by HPLC (HPLC 1500, Waters, MA, USA).

3.1. Gas transfer performance of designed flasks

In this study, three types of baffled flasks were made of plastic material (PLA) using 3D printer (Fig. 1). The latticed-baffled flasks have the lattice-shaped baffles at the bottom. The curved-baffled flasks consist of wavy baffles at the bottom, allowing the fluid to flow smoothly after touching the baffle. The swirled-baffled flasks were fabricated to incorporate both latticed- and curved-baffle characteristics. Considering the possible effect on the experiment results due to the difference in base material, PLA flask of the same specification and shape as those of ready-made glass Erlenmeyer flask was fabricated and used as a control (Fig. 1(A)). The results from the CO₂ titration experiment indicated that the performance of gas transfer depends on the types of baffle (Fig. 3). From the comparison of the time when the pH of the solution dropped from 11.0 to 8.2 by CO₂ dissolution, it was clear that a longer time was needed for CO₂ to be completely dissolved in flasks without baffles. On the other hand, the latticed-baffled flasks were observed to have the best CO₂ transfer rate, while the curved-baffled flasks showed slightly lower gas transfer performance than the conventional bottom-baffled flasks. In the case of latticed-baffled flasks, despite of a small difference, the soccer-ball, which consisted of pentagon and hexagon, showed better gas transfer performance than the honeycomb, which consisted of hexagon only. The swirled-baffled flasks were observed to have moderate gas transfer performance between the curved- and latticed-baffled flasks. The difference seen in the results is explained by the degree of turbulence formation depending on the type of baffle.

Figure 3. Comparison of CO2 dissolution rate according to baffle type (A: Unbaffled, B: 3 Bottom, C: 4 Bottom, D: Soccer ball, E: Honeycomb, F: 3 Wave, G: 6 Wave, H: 6 Wave with center ball, I: 6 Vortex clockwise, and J: 6 Vortex counter-clockwise).

3.2. Batch cultivation of E. coli using designed flasks

Batch cultures of E. coli were performed for 6 hours each in 4 types of PLA flasks fabricated using 3D printer (non-baffled, soccer-ball-baffled, 4-bottom-baffled and 6-vortex-counterclockwisebaffled flask), which were selected from the CO₂ absorption experiment. To investigate the effect of plastic material on the cell growth, E. coli was cultured in the ready-made glass flask without baffle of the same specification and shape as the nonbaffled PLA flask. As a result, no significant difference in cell concentration (OD₆₀₀) and specific growth rate (μ) was observed (OD₆₀₀ = 4.29 ± 0.36 and μ = 0.59/hr for glass flask, whereas OD₆₀₀ = 4.19 ± 0.55 and μ = 0.59/hr for PLA flask), and thus it was confirmed that flask culture performance did not depend on the material if the flask had the same structure as in this study. When comparing the E. coli cultures in PLA flasks, it was found that the higher the gas transfer rate, the higher cell concentration and the lower acetic acid concentration (Table 1). The soccer-ball-baffled flask designed in this study was measured to have a final cell concentration of 30.4% and 6.1% higher than the conventional non-baffled flask and bottom-baffled flask, respectively. This implies that the presence of a sharper and more baffles at the bottom of the flask promotes the formation of more turbulence, thus leading to the improvement of mass transfer and consequently higher cell concentration. It was also observed that the higher the cell concentration in the culture, the lower the acetic acid concentration. This meant that, the oxygen was sufficiently supplied to the culture broth and the acetic acid was well oxidized through the TCA cycle.

Table 1 Cell growth and acetate production during 6h batch cultivation of E. coli in designed flasks

Flasks	OD600	Acetate (g/L)	Specific growth rate (1/hr)	Acetate production rate (g/L/hr)
Unbaffled	4.19 ± 0.55	0.37 ± 0.04	0.59	0.06
4 bottom	5.27 ± 0.26	0.24 ± 0.01	0.62	0.04
6 vortex counter-clockwise	5.10 ± 0.08	0.24 ± 0.01	0.61	0.04
Soccer ball	5.73 ± 0.12	0.09 ± 0.02	0.63	0.02

Fig. 4 shows the relationship between mass transfer performance and cell concentration. The reciprocal of CO₂ dissolution time was shown to have a linear relationship with the cell concentration. Considering that kLa is the value of transferred gas amount divided by time, it was reconfirmed that the higher mass transfer coefficient, the greater the cell growth.

Figure 4. Relationship between CO₂ dissolution time and final cell concentration.

This study demonstrates the possibility of easily developing a tailor-made small-scale microbial culture system with decent gas-liquid mass transfer rate by applying 3D-printing technology. By the 3D printer, we could easily control the shape and number of baffles located at the bottom of the flasks, and develop the new baffled flasks with higher gas-liquid mass transfer rate than the conventional non-baffled and bottombaffled flasks. As a result, the highest cell growth was obtained in the new latticed-baffled flask. Since the continuous development of computational fluid dynamics and 3D-printing technology is expected, new flasks with desired properties such as better gas-liquid mass transfer performance could be tailored for different purposes.

In this study, we tried to develop a tailed-made flask culture system suitable for aerobic fermentation using 3D-printing technology. For this, various baffled flasks were fabricated from PLA material using a 3D printer, and the mass transfer rate of CO₂ in each flask was evaluated. As a result, it was confirmed that the difference in CO₂ mass transfer rate occurred due to the difference in the formation of turbulence depending on the number and shape of the baffles. In addition, in order to investigate the effect of transfer efficiency of gaseous substrate on the culture characteristics, aerobic batch culture of E. coli was conducted in baffled and non-baffled PLA flasks. As a result, the highest cell concentration and the lowest acetic acid concentration were observed in the soccer-ball-baffled flask which showed the highest CO₂ mass transfer rate, whereas the opposite result was observed in the non-baffled PLA flask with the lowest CO₂ mass transfer rate. It implies that the difference of O2 supply rate to the culture broth strongly affects the oxidation of acetic acid through the TCA cycle, leading to the difference in culture characteristics. This study shows the potential of 3D-printing technology to develop an efficient, customized small scale-scale gas fermentation culture systems.

This research was supported by C1 Gas Refinery Program through the National Research Foundation of Korea (NRF) funded by the ministry of Science, ICT & Future Planning (NRF-2015 M3D3A1A01064926), the Sogang University Research Grant of 2022 (202219012.01), and the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) (2020R1F1A1068541).