The effect of diurnal light cycles on biohydrogen production in a thermosiphon photobioreactor

Bacterial species and culturing

In this study, Rhodopseudomonas palustris (NCIMB 1774) was used as model organism. Cells were precultured in Van Niels medium—a fast-growing medium containing (per 1 L): 1 g K2HPO4, 0.5 g MgSO4, 10 g yeast extract and the balance deionised water (du Toit and Pott 2021). After autoclaving (121 °C, 20 min), 10 mL of sterile glycerol (4 M) was added to the medium aseptically (du Toit and Pott 2021). Bacterial cells were suspended in the medium and grown anaerobically in 500 mL Schott bottles under argon atmosphere. Temperature was maintained at 35 °C (± 0.2 °C) and light intensity was calibrated to 200 W m−2 (± 20 W m−2) in the wavelength range of 500–1100 nm (tungsten filament incandescent light, Eurolux©, South Africa) (Bosman et al. 2022a) using a handheld spectrophotometer (RGB Photonics, Qmini VIS–NIR). Culturing time was approximately five days to allow cells to reach the mid-logarithmic phase.

All growth and hydrogen production experiments were conducted using a Rhodospirillaceae medium containing (per 1 L): 0.6 g K2HPO4, 1.7 g KH2PO4, 0.02 g MgSO4·7H2O, 0.005 g CaCl2·2H2O, 0.4 g NaCl, 0.3 g Na2S2O3, 0.0005 g ferric citrate, 0.0002 g para-aminobenzoic acid, and 1 mL of trace element solution containing (per 1 L): 70 mg ZnCl2, 100 mg MnCl2·4H2O, 60 mg H3BO3, 200 mg CoCl2·6H2O, 20 mg CuCl2·2H2O, 20 mg NiCl2·6H2O, and 40 mg NaMoO4·2H2O (Pott et al. 2014). The medium was autoclaved (121 °C, 20 min) and the pH measured at 7.2. Also aseptically added to the medium after autoclaving was a vitamin solution containing (per 1 L): 1.2 g thiamine HCl and 0.01 g cyanocobalamin (filter-sterilized), and lastly 10 mL of 5 M sterile glycerol (final concentration of 50 mM) and 5 mL of sterile glutamic acid (final concentration of 10 mM) were added to the medium aseptically (Pott et al. 2014).

Experimental setups

For the hydrogen production and growth experiments, two different photobioreactor setups were implemented. A test-tube reactor was used to investigate the effect of light cycles on the growth and hydrogen productivity of the bacterial species R. palustris at constant temperatures, while a TPBR setup was used to determine the effect of light cycles on the hydrogen productivity and internal working of the TPBR itself at varying temperatures.

Test-tube photobioreactor

The test-tube reactor system consisted of four glass test-tubes (72 mL working volume) mounted inside a temperature-controlled water bath (35 °C ± 0.2 °C). The test-tubes were illuminated from one side using halogen floodlights (Eurolux©, South Africa, FS13, 150 W) and the light intensity inside the test-tubes was calibrated using a handheld spectrophotometer (RGB Photonics, Qmini VIS–NIR). The tubes were fitted with gastight lids consisting of two stainless steel sample ports—one for liquid samples (biomass and glycerol concentration) and one for gas measurements. The gas port was connected to an inverted measuring cylinder, immersed in a water bath, using low hydrogen-permeability tubing (Saint Gobain, South Africa, Tygon E3606). A one-way valve was also used to ensure no reverse flow into the reactor itself. The water-displacement technique was used to monitor the volume of evolved gas.

Thermosiphon photobioreactor

The TPBR is a tubular loop (inner diameter 24 mm) reactor made of glass and can be divided into three sections—an illuminated riser section (600 mm length), an insulated downcomer section (450 mm length) and a cooling section surrounded by a cooling water jacket (442 mL volume) (Bosman et al. 2022a). The entire reactor was insulated and shaded from light, except for the illuminated riser section which was illuminated by halogen floodlights (Eurolux©, South Africa, FS13, 150 W). A water chiller (model F25, Julabo GmbH, Germany) was used to continuously circulate cooling water through the jacket at a fixed flow rate of 0.5 L min−1. The reactor had a reflector in the centre of the loop behind the riser section, facing the riser and the halogen floodlights—this was to ensure light from both sides of the riser section for enhanced light distribution. The reactor was fitted with three temperature sensors (3-wire PT100, 3 mm diameter, stainless steel sheath) that were connected to a data logger—the three sensors were positioned at the top and bottom of the illuminated riser section, and one directly below the cooling section of the TPBR. The reactor was fitted with a GL45 polypropylene lid, and samples were taken from the sampling ports. The volume of evolved gas was quantified using the water-displacement method, similar to that described for the test-tube reactors.

Illumination system

In order to simulate outdoor conditions, more specifically the diurnal light cycle, outdoor light intensities were measured. Light intensities within the spectral range of 360–1120 nm were measured using a pyranometer (Campbell Scientific, South Africa, CS300, Apogee Instruments Inc.) connected to a data logger (Campbell Scientific, South Africa, CR300). The light intensities were measured over a period of 24 h (Stellenbosch, South Africa, April 2022) and the mean of triplicate measurements were used. The time-dependent light intensity data were coded into an Arduino® microcontroller light simulator system. The system comprised a microcontroller board (Micro Robotics©, South Africa, Arduino® UNO R3 Original) connected to four vertically oriented halogen floodlights (Eurolux©, South Africa, FS13, 150 W), a 4-channel AC dimmer module (Rocket Controller, China, 3.3 V/5 V logic, AC 50/60 Hz) and a real-time clock module (Micro Robotics©, South Africa, DS3231, 3.3 V/5 V) to maintain accurate time-keeping in the event of a power outage. This allowed automated control of the emitted light intensities, simulating an indoor light cycle mimicking that measured outside (Fig. 1).

Fig. 1figure 1

Measured and simulated 24-h diurnal light cycle, and continuous light intensity (500 W m−2) over time

Experimental procedure

For the growth and hydrogen production experiments, precultured R. palustris cells were added to Rhodospirillaceae medium (with glycerol as carbon source) to obtain a starting biomass concentration of approximately 0.1 g L−1 in all reactors. Thereafter, the cultures were aseptically added to an autoclaved (121 °C, 20 min) photobioreactor, i.e. either the test-tube or TPBR setup, depending on the experiments. Reactors were sparged with filtered-sterilised (Midisart®, 2000 PTFE filter, diameter of 50 mm, pore size of 0.2 µm) argon gas (purity of > 99.9%) for approximately 10 min to ensure an anaerobic atmosphere inside the reactors.

Experimental runs using the test-tube reactors were initialized by switching on the light sources and the water baths (35 °C). In the TPBR, experimental runs were initialized by switching on the light sources and the cooling system to induce circulation in the reactor. Sampling (cell growth and glycerol concentration) was done approximately every 10, 14 or 24 h, and the cumulative volume of evolved gas was also noted at these timestamps. All test-tube experiments were done in quadruple, while the TPBR runs were conducted in triplicate to allow for the determination of statistical significance and standard deviation—that is, each individual experimental run was repeated four and three times, respectively, at identical conditions.

Analytical methods

Cell dry weight (CDW) was determined using a CDW vs optical density (OD) standard curve. A UV/Vis-spectrophotometer (Model AE-S60-4U) was used to measure OD, and OD was correlated to CDW as follows: CDW = 0.7126 × OD660 nm—0.007 (Van Niels medium), R2 = 0.9981; CDW = 0.6391 × OD660 nm + 0.0619 (Rhodospirillaceae medium), R2 = 0.9996 (Bosman et al. 2022a). Glycerol concentration was measured using high-performance liquid chromatography (Dionex UltiMate 3000 HPLC). Filtered samples (FilterBio® Nylon Syringe Filter, 13 mm diameter, 0.22 µm pore size) were injected into the HPLC column (Bio-Rad Laboratories Ltd., Johannesburg, South Africa, HPX-87H column, 250 × 7.8 mm with guard cartridge and ERC Refracto Max520 RI detector) operating at 65 °C and using a 0.005 M H2SO4 solution as mobile phase (0.6 mL min−1). Evolved gas samples were analysed using a gas chromatograph (Global Analyser Solutions Compact Gas GC) with a thermal conductivity detector (110 °C), packed columns (Rt-QBond 3 m × 0.32 mm and Molsieve 5A 3 m × 0.533 mm) and argon as carrier gas (45 kPa, 50 µL injections with a split of 5 mL min−1). The GC oven and filament temperatures were specified as 65 and 210 °C respectively, and the reference flow rate was 1 mL min−1.

Calculations

The effect of light protocol was investigated based on the rate of hydrogen production, glycerol consumption and hydrogen yield. The rate of hydrogen production was calculated using the molar volume of hydrogen produced (Δn) at the time of interest (t) together with the culture/photobioreactor volume (V), as denoted in Eq. 1. The molar volume of hydrogen was determined at NTP using the analysed composition of the evolved gas, i.e. ~ 92% (± 1.9%) hydrogen and the balance CO2. Glycerol consumption was evaluated using the molar concentration of glycerol in the system at a specific time (nt), and the glycerol concentration initially in the system (no), as shown in Eq. 2.

$$ }_2 \,}\,\left( }\,}^ \,}^ } \right) \, = \frac}_2 \,measured} }} $$

(1)

$$ }\,\left( \% \right) = \frac - n_ }} }} \times 100 $$

(2)

Hydrogen yield was determined as the molar ratio of hydrogen produced to the theoretical maximum of glycerol that can be consumed according to the stoichiometric conversion of glycerol to hydrogen: C3H8O3 + 3H2O → 3CO2 + 7H2 (Eq. 3).

$$ }_2 \,}\,\left( \% \right) \, = \frac}_2 \,measured} }} }} \times 100. $$

(3)

留言 (0)

沒有登入
gif