Co-Immobilization of Clostridium carboxidivorans and Clostridium kluyveri in a Synthetic Dual-Layer Biofilm for Syngas Conversion

Abstract

Syngas fermentation in combination with chain elongation offers great promise for sustainable medium-chain fatty acid production. While immobilization has proven effective for stabilizing monocultures of C. kluyveri for chain elongation, its applicability to co-cultures involving C. carboxidivorans for simultaneous syngas fermentation remains unexplored. This study investigates the physiological compatibility of C. carboxidivorans with agar-based hydrogel immobilization and its co-cultivation potential with C. kluyveri in a synthetic dual-layer biofilm reactor. First, we conducted autotrophic batch fermentations using suspended and immobilized cells, proving metabolic activity similar for both. Applying different sulfur feeding rates, experiments showed best ethanol formation with C. carboxidivorans at increased sulfur feeding, enabling better conditions for chain elongation with C. kluyveri. In the synthetic dual-layer biofilm reactor, with the C. carboxidivorans biofilm in contact with the CO-containing gas phase above the C. kluyveri biofilm, the formation of 1-butyrate and 1-hexanoate was observed with product formation rates of 0.46 g L⁻¹ d⁻¹ 1-butyrate, and 0.91 g L⁻¹ d⁻¹ 1-hexanoate, respectively. The formation rate of 1-hexanoate in the dual-layer biofilm reactor was approximately 7.6 times higher than that reported with suspended cells in a stirred tank bioreactor. Spatial analysis revealed species-specific migration behavior and confirmed that C. carboxidivorans reduced local CO concentrations, improving the environment for C. kluyveri.

1. Introduction

The increasing concentration of greenhouse gases in the atmosphere, particularly carbon dioxide (CO₂), is widely recognized as a major driver of anthropogenic climate change. According to the Intergovernmental Panel on Climate Change (IPCC), global CO₂ emissions must be drastically reduced to limit global warming to below 1.5 °C and avoid severe environmental, economic, and societal impacts [1]. This has led to an urgent demand for carbon-neutral and carbon-negative technologies capable of utilizing CO₂ or CO-rich waste gases as raw materials in industrial production.

Among the strategies being explored, biological gas fermentation has emerged as a promising route for converting synthesis gas (syngas)—a mixture typically composed of CO, CO₂, and H₂—into value-added chemicals and fuels under mild reaction conditions [2,3]. Unlike thermo-catalytic processes, gas fermentation operates at ambient pressure and temperature, and allows for integration with microbial metabolic engineering strategies [4]. Moreover, syngas can be derived from a variety of carbon-rich waste streams, including industrial off-gases, municipal solid waste, and agricultural residues, enabling a circular bioeconomy approach [5].

A key advantage of gas fermentation is its capacity to convert C1 gases into predominantly short-chain organic products, such as acetate and ethanol. But only small amounts of medium-chain organic acids and alcohols are formed. This product profile is of particular relevance to the chemical industry. For example, acetate, ethanol, 1-butyrate, 1-hexanoate, 1-butanol, and 1-hexanol are platform chemicals that can serve as precursors for biofuels, plasticizers, solvents, and specialty chemicals [6,7]. These compounds also show potential as antimicrobial agents or feed additives, broadening their applicability across multiple industrial sectors [8]. The production of such molecules from waste-derived syngas not only adds economic value but also reduces reliance on fossil resources and mitigates CO₂ emissions from both ends of the value chain. In the context of the European Green Deal and the United Nations Sustainable Development Goals, the development of microbial bioprocesses for syngas valorization directly addresses targets related to climate action, sustainable industry, and responsible consumption and production [9,10]. Thus, advancing the biotechnological conversion of syngas into functionalized C4 and C6 molecules contributes to the broader goal of transforming today’s linear petrochemical industry into a resilient and bio-based circular economy.

The biological fixation of CO and CO₂ into organic molecules is primarily enabled by autotrophic acetogenic bacteria via the Wood–Ljungdahl pathway, a highly energy-efficient form of anaerobic carbon fixation [11]. Among these, Clostridium carboxidivorans has emerged as a particularly promising species for syngas fermentation due to its ability to grow autotrophically with CO, CO₂, or H₂ and convert these gases into a range of reduced carbon products [12,13]. In contrast to many other acetogens, C. carboxidivorans not only produces acetate and ethanol as major fermentation products but also forms 1-butanol and even 1-hexanol in small but detectable amounts [13,14].

The product spectrum of C. carboxidivorans reflects the classical biphasic fermentation behavior of many Clostridia: in the acidogenic phase, it consists of acetate and 1-butyrate accumulate, while in the solventogenic phase, these acids are reduced to ethanol, 1-butanol, and longer-chain alcohols—potentially as a cellular strategy to counteract acidification and maintain pH homeostasis [15].

However, while C. carboxidivorans can produce C4 and C6 products from syngas, the yields and selectivities are typically low under standard conditions [16]. To improve both carbon chain length and product titer, metabolic cooperation with other microbes—specifically chain-elongating species—has been investigated.

One such partner is Clostridium kluyveri, a strict anaerobe capable of performing reverse β-oxidation. This is a pathway that elongates carbon chains by the successive addition of two-carbon units. In this pathway, ethanol is oxidized to acetyl-CoA, which then condenses with acetate-derived acetyl-CoA to the production of 1-butyrate and 1-hexanoate, respectively [17,18,19]. Notably, more than 25% of the cellular carbon in C. kluyveri can originate from CO₂, underlining its mixotrophic potential during chain elongation [20].

The combination of C. carboxidivorans and C. kluyveri in a synthetic co-culture exploits their complementary metabolic capabilities: the acetogen produces acetate and ethanol from syngas, which the chain elongator then uses to generate C4/C6 fatty acids, forming a cooperative bioconversion route [21,22].

A review of recent studies reveals a broad spectrum of experimental setups and performances. In anaerobic shake flasks (ASF), Diender et al. (2016) demonstrated that C. acetoethanogenum and C. kluyveri can co-produce 1-butyrate (2.3 g L⁻¹), 1-hexanoate (0.7 g L⁻¹), 1-butanol (0.7 g L⁻¹), and 1-hexanol (0.4 g L⁻¹) from syngas mixtures [23]. Using a more controlled continuous stirred-tank reactor (CSTR) with inline product extraction, Richter et al. (2016) achieved significantly higher product concentrations with a co-culture of C. ljungdahlii and C. kluyveri, reaching up to 4.7 g L⁻¹ 1-hexanol and 0.78 g L⁻¹ 1-octanol using a mixture of 60% CO, 35% H₂, and 5% CO₂ as synthesis gas [24].

Focusing specifically on syngas-based chain elongation, recent studies using C. carboxidivorans and C. kluyveri have shown promising productivities. In a batch stirred-tank reactor fed with 80% CO and 20% CO₂, Bäumler et al. (2022) reported combined concentrations of 1.2 g L⁻¹ 1-butyrate/1-butanol and 0.6 g L⁻¹ 1-hexanoate/1-hexanol [21]. Fernández-Blanco et al. (2022) investigated a co-culture of the acetate producing C. aceticum and the chain-elongating C. kluyveri in a stirred-tank reactor supplemented with ethanol, achieving up to 8.2 g L⁻¹ 1-hexanoate and 7.0 g L⁻¹ 1-butyrate with 30% CO and 5% CO₂ in the gas phase [25]. Most recently, Bäumler et al. (2023) employed a cascade of two continuously operated stirred-tank bioreactors with a synthetic co-culture of C. carboxidivorans and C. kluyveri. At pH 6.0 and 100 mbar CO in the first reactor, ethanol and acetate were produced by C. carboxidivorans and elongated by C. kluyveri. In the second reactor (pH 5.0, 800 mbar CO), the organic acids were reduced to alcohols, yielding steady-state concentrations of 3.78 g L⁻¹ ethanol, 0.97 ± 0.07 g L⁻¹ 1-butanol, 0.16 ± 0.01 g L⁻¹ 1-hexanol, and 0.77 ± 0.03 g L⁻¹ 1-octanoate. The CO partial pressure was reduced in the first reactor to enable the growth of C. kluyveri [15].

Despite the advances delivered by this kind of co-culture, the fundamental challenge remains: the dual role of CO. CO is an essential electron donor and carbon source for acetogens, and thus must be supplied in sufficiently high concentrations to drive autotrophic growth and ethanol formation by C. carboxidivorans [13,15]. At the same time, however, CO exerts an inhibitory effect on C. kluyveri, the chain-elongating partner. While C. kluyveri does not require CO for its metabolism, it is sensitive to elevated CO concentrations, which interfere specifically with cellular growth and viability, although chain elongation activity itself may remain functional [23]. The mechanism of inhibition is thought to involve disruption of the energy metabolism. CO may interfere with membrane-bound redox systems such as the Rnf complex, compromise the proton motive force, and disturb redox homeostasis essential for growth [19,26]. This poses a fundamental dilemma for co-cultivation. The acetogen requires a high partial pressure of CO for sufficient productivity, while the chain elongator performs best in its absence or at low concentrations. Attempts to balance these conflicting requirements with suspended cells are inherently limited, as both organisms are exposed to the same environment.

A promising solution to this challenge lies in the use of biofilm-based reactors, which offer spatial structuring and microenvironment control. In biofilm reactors, microorganisms can be immobilized within defined regions of a matrix, enabling localized metabolite gradients and partial separation from toxic gases like CO. For instance, if C. carboxidivorans is localized on the gas-exposed side of the biofilm, it can serve as a protective barrier, reducing CO diffusion into deeper layers where C. kluyveri resides. This physical separation mimics microbial community structures found in natural anaerobic consortia, where spatial stratification enables stable co-existence under otherwise inhibitory conditions [27,28]. In addition, biofilm reactors allow for high initial inoculation densities—especially of C. kluyveri—which compensates for its inhibited growth under CO exposure.

In addition to suspended cultivation strategies, the immobilization of microorganisms has been studied as a powerful tool to enhance the stability, productivity, and resilience of microbial bioprocesses [29]. Similar to co-cultures combining syngas-utilizing acetogens with chain-elongating bacteria, immobilization techniques have been applied with a wide range of anaerobic microorganisms. This was in order to improve biocatalytic performance under stress conditions and enable long-term operation in continuous or semi-continuous processes.

Karube et al. (1982) immobilized C. butyricum in an acetylcellulose filter containing agar and conducted repeated batch fermentations using glucose as the substrate [30]. This membrane reactor configuration with immobilized cells significantly enhanced hydrogen yield in batch processes, as approximately 45% (w/w) of the supplied glucose was converted into H₂. After several reuse cycles, the immobilized cells even achieved stoichiometric hydrogen yields from glucose, i.e., yields approaching the theoretical maximum. The addition of peptone and riboflavin further improved H₂ productivity and reduced undesirable by-products such as lactate. This early study demonstrated that immobilization in a solid matrix can increase cell density and stabilize biohydrogen production [30].

Riegler et al. (2019) employed C. aceticum in a retrofitted stirred-tank bioreactor operated either as a packed-bed or trickle-bed biofilm reactor for continuous CO₂/H₂ fermentation at pH 8. The trickle-bed configuration enabled superior gas–liquid mass transfer and achieved acetate space-time yields of up to 14 mmol L⁻¹ h⁻¹ with hydrogen conversion rates reaching 96%. Biofilm formation occurred on sintered ceramic carriers without artificial immobilization, resulting in robust cell retention under continuous operation. However, the naturally forming biofilms developed slowly and exhibited low overall biomass concentrations, highlighting a major limitation of natural biofilm formation for rapid start-up of biofilm reactors and urging for a new approach like synthetic biofilms [31].

A lab-scale alternative for natural biofilm formation was shown by Zhang et al. (2019). They applied wheat straw to soak media with varying bacterial concentrations of C. kluyveri and observed increased the production of 1-hexanoate from ethanol and acetate and greater tolerance to inhibitory levels of ammonium (up to 5 g L⁻¹ NH₄⁺) [32]. The straw with the immobilized cells could be reused in multiple batch cycles in the anaerobic flasks, and adaptation over cycles reduced lag phases from 72 to 30 h [32].

In a synthetic biofilm reactor described by Herzog et al. (2025), C. kluyveri cells were embedded in 4 mm planar agar hydrogel. Ethanol and acetate supplementation, as well as product removal, were achieved via the recirculating liquid phase drawn from the bottom of the synthetic biofilm, while the gas phase—containing defined amounts of CO₂—was supplied from above [33]. They demonstrated production of up to 10.1 g L⁻¹ 1-hexanoate and 2.7 g L⁻¹ 1-butyrate in batch mode with yeast extract and even proved production without yeast extract was possible. The synthetic biofilm configuration showed space-time yields of up to 4.95 g L⁻¹ h⁻¹ 1-hexanoate and 1.33 g L⁻¹ h⁻¹ 1-butyrate (related to the biofilm volume), highlighting the effectiveness of synthetic biofilms with C. kluyveri for chain elongation [33].

The combination of immobilization and co-cultivation—specifically between syngas-utilizing acetogens and chain-elongating organisms—has not yet been studied in depth. Most current research focuses either on suspended co-cultures (Bäumler et al., 2023; Diender et al., 2016) or on the immobilization of single species under isolated conditions (Zhang et al. 2019) [15,23,32]. However, the synergistic potential of spatially structured, synthetically immobilized co-cultures remains largely untapped.

This study aims to address this gap by establishing a synthetic biofilm reactor in which C. carboxidivorans and C. kluyveri are co-immobilized within a hydrogel matrix, enabling controlled spatial organization. The upper layer of the synthetic biofilm contains CO-converting C. carboxidivorans in direct contact with a syngas atmosphere (CO/CO₂), facilitating the autotrophic production of ethanol and acetate. Positioned beneath, a second hydrogel layer harbors C. kluyveri, which utilizes these intermediates for chain elongation to produce valuable medium-chain carboxylates. Both microbial populations are in contact with a recirculating aqueous phase at the base of the reactor, ensuring nutrient supply and the continuous removal of fermentation products. Through this approach, we seek to combine the metabolic advantages of both strains with the process stability conferred by immobilization, while also mitigating the limitations caused by CO toxicity for C. kluyveri. The spatial structuring of the biofilm creates localized microenvironments that shield the CO-sensitive C. kluyveri and allow for sustained metabolic activity, even under conditions that would otherwise inhibit growth. As growth is not essential in synthetic biofilms, this setup also enables the potential of inoculating with high cell densities.

To this end, the present study aims to first investigate the suitability of C. carboxidivorans for immobilization in agar-based hydrogels using anaerobic flask experiments. Based on these findings, optimal operating conditions for C. carboxidivorans will be established in a biofilm reactor to enable efficient ethanol and acetate production. Finally, a dual-layer biofilm co-culture with C. kluyveri will be implemented and characterized under these defined conditions, aiming to demonstrate the feasibility of a spatially structured, synthetic biofilm for syngas-based chain elongation.

2. Materials and Methods

2.1. Organisms and Culture Conditions

The bacterial strains employed in this study were C. carboxidivorans DSM 15243 and Clostridium kluyveri DSM 555, sourced from the German Collection of Microorganisms and Cell Cultures (DSMZ, Braunschweig, Germany). All experimental work was carried out using a modified Hurst medium (for composition see Tables S1–S9 in the supporting information), originally developed by Hurst and Lewis (2010) [34]. The formulation was adapted according to the procedure described by Schneider et al. (2021) [22], incorporating 0.4 g L⁻¹ L-cysteine-HCl as a reducing agent. As part of this study, an alternative trace element solution (TESnew) for an enhanced alcohol production was employed based on the formulation described by Shen et al. (2017) [35] (see SI). Whenever this alternative composition instead of the standard solution (TESstandard) was used, it is explicitly stated in the respective sections. The pH was adjusted to pH 6.0 prior to use. Medium preparation was carried out under strictly anaerobic conditions following established protocols [36,37]. This modified Hurst medium was consistently used for all mono-cultivations of C. carboxidivorans in anaerobic flasks and in the biofilm reactor, as well as for co-cultivations of C. carboxidivorans and C. kluyveri.

2.2. Precultures

Strains acquired from DSMZ were initially propagated in the modified Hurst medium. Following cultivation, glycerol was added as a cryoprotectant, and the cultures were preserved at −80 °C as glycerol stocks until further use. For the initiation of precultures, 2.5 mL (5 mL) of frozen stock for 100 mL (700 mL) medium was thawed and aseptically transferred into nitrogen-flushed anaerobic flasks (1.0 bar total pressure) via sterile syringes (BD Discardit II, Becton Dickinson, Franklin Lakes, NJ, USA) and needles (Sterican 0.9 × 70 mm, B. Braun, Melsungen, Germany) through a butyl rubber septum (Glasgerätebau Ochs, Bovenden, Germany). Heterotrophic cultivation of C. carboxidivorans and C. kluyveri was performed in 0.5 L (1.0 L) anaerobic flasks containing 100 mL (700 mL) of the modified Hurst medium. The medium was supplemented with 5.0 g L⁻¹ glucose for C. carboxidivorans, and with 10 g L⁻¹ potassium acetate and 20 mL L⁻¹ ethanol for C. kluyveri. For C. kluyveri, an additional 2.5 g L⁻¹ sodium bicarbonate (Na₂HCO₃) was included to provide a source of inorganic carbon in the form of bicarbonate/CO₂ [20,38]. Both strains received 0.4 g L⁻¹ L-cysteine-HCl to maintain reducing conditions. Cultures were incubated at 37 °C under agitation (100 rpm, 2.5 cm orbital diameter) using a shaking incubator (Wisecube WIS-20R, witeg Labortechnik GmbH, Wertheim, Germany) for 24 h in the case of C. carboxidivorans, and 120 h for C. kluyveri. Cells were harvested during exponential growth via centrifugation at 3620×g for 10 min (Rotina 50 RS, Hettich GmbH, Tuttlingen, Germany), followed by resuspension in anaerobic phosphate-buffered saline (PBS, pH 7.4) prior to suspended or immobilized inoculation.

2.3. Synthetic Biofilm Fabrication

To prepare the hydrogel matrix in order to fabricate the synthetic biofilm, agar-agar was dissolved at a concentration of 18 g L⁻¹ in the modified Hurst medium used for cultivations and sterilized by autoclaving at 121 °C for 20 min. Following sterilization, the hydrogel solution was cooled to ambient temperature under a nitrogen atmosphere to prevent oxygen diffusion. Upon reaching 45 °C, 0.4 g L⁻¹ of L-cysteine-HCl was added to the hydrogel. Afterward, the hydrogel was inoculated with the pre-prepared cell suspension in PBS. Defined aliquots of the inoculated hydrogel were subsequently transferred into a sterile mold (as previously described by Herzog et al., 2025 [33]) or an anaerobic flask using a positive displacement pipette (Transferpettor Digital, 2000–10,000 µL; BRAND GMBH + CO KG, Wertheim, Germany). The filled mold or flasks were then allowed to solidify at room temperature under a nitrogen atmosphere, resulting in the formation of a synthetic biofilm with thicknesses of 4 mm, due to the defined aliquot volumes in context with the geometry of the mold. As for co-cultivation experiments with the biofilm reactor, the procedure was repeated and the second biofilm layer was poured on top of the solidified first layer, forming a synthetic double layered biofilm with individual thicknesses of each layer of 4 mm.

2.4. Parallelized Anaerobic Flask Experiments

Batch experiments in 500 mL anaerobic flasks were conducted, either employing synthetic biofilms immobilized at the bottom or as suspension cultures. For experiments utilizing immobilized C. carboxidivorans cells, synthetic biofilms were overlaid with 100 mL of modified Hurst medium containing 0.4 g L⁻¹ of L-cysteine-HCl. The flasks were then sealed gas-tight under a nitrogen atmosphere and subsequently pressurized to 2.0 bar with a 1:4 (v/v) mixture of CO₂ and CO. For suspension culture experiments, 100 mL of modified Hurst medium was added to 500 mL anaerobic flasks, which were then similarly sealed and pressurized with the same gas mixture and total pressure. All flasks were equilibrated to the target inoculation temperature using thermostated water baths or the environmental temperature for experiments with inoculation at room temperature. Subsequently, each flask within the suspended cell approaches was supplemented via sterile syringe and needle with L-cysteine-HCl to a final concentration of 0.4 g L⁻¹, followed by inoculation with the prepared cell suspension in PBS. Immediately afterward, the cultures were incubated at 37 °C in a shaking incubator (Wisecube WIS-20R, Witeg Labortechnik GmbH, Wertheim, Germany) set to 100 rpm with a 2.5 cm orbital shaking diameter for the duration of the experiments. Samples for high-performance liquid chromatography (HPLC), optical density measurement (OD₆₀₀) or cell dry weight (CDW), and pH determination were aseptically withdrawn using sterile, single-use syringes (BD Discardit II; Becton Dickinson, Franklin Lakes, NJ, USA) and needles (Sterican 0.9 × 70 mm; B. Braun, Melsungen, Germany) through the butyl rubber stoppers and subsequently processed.

2.5. Synthetic Biofilm Reactor Experiments

A complete and detailed explanation of the novel synthetic biofilm reactor setup can be found in previously published work [33]. For all biofilm reactor experiments in this work, the biofilm reactor chamber, previously sterilized by autoclaving at 121 °C for 20 min, was transferred into an anaerobic glovebox and opened under a nitrogen atmosphere. The previously prepared solidified synthetic biofilm (Section 2.3) was carefully extracted from the mold using a custom-fabricated tool [33] and placed into the open chamber. The reactor was then hermetically sealed to ensure gas tightness and removed from the anaerobic environment. Subsequently, 600 mL of modified Hurst medium—excluding L-cysteine-HCl—was introduced into the reactor via a septum using a peristaltic pump. L-cysteine-HCl was then added to a final concentration of 0.4 g L⁻¹ through the same septum. During the entire duration of the batch experiments, the reactor headspace was continuously sparged with a 1:4 (v/v) mixture of CO and CO₂ at a flow rate of 1 L h⁻¹ and a total pressure of 1.0 bar. The pH of the medium was maintained at pH 6.0 by the automated addition of either 2 M NaOH or 3 M H₂SO₄. Temperature was maintained at 37 °C throughout the processes. Additionally, a continuous Na₂S feed was implemented throughout the entire duration of the processes to ensure a constant sulfur supply. The feed rate was varied between experiments and is explicitly stated for each individual experiment. Samples for OD measurement or CDW determination and HPLC quantification were aseptically taken through a septum using sterile, single-use syringes (BD Discardit II; Becton Dickinson, Franklin Lakes, NJ, USA) and needles (Sterican 0.9 × 70 mm; B. Braun, Melsungen, Germany).

2.6. Analytical Methods

2.6.1. High-Performance Liquid Chromatography (HPLC) and Optical Density (OD)

Quantification of organic acids and alcohols was carried out using high-performance liquid chromatography (HPLC, 1100 Series; Agilent Technologies, Santa Clara, CA, USA) equipped with a refractive index (RI) detector and an Aminex HPX-87H ion-exchange column (Bio-Rad, Munich, Germany). Chromatographic separation was achieved using 5 mM H₂SO₄ as the mobile phase at a constant flow rate of 0.6 mL min⁻¹. The column temperature was maintained at 60 °C. Prior to injection, all samples were filtered through 0.2 µm regenerated cellulose membrane filters (Chromafil RC20/15 MS; Macherey-Nagel GmbH & Co. KG, Düren, Germany). Time-resolved product formation rates were estimated by fitting the data to a biphasic sigmoidal model using non-linear regression. CDW concentration was estimated by measuring the optical density at 600 nm (OD₆₀₀) using a UV-Vis spectrophotometer (Genesys 10S UV-Vis; Thermo Scientific, Neuss, Germany). CDW concentrations were calculated using a previously established linear correlation factor (0.48 ± 0.03 g L⁻¹) for C. carboxidivorans mono cultures [21]. To distinguish between the two cell types, fluorescence in situ hybridization (FISH) was applied as described in the following.

2.6.2. Flow Cytometry (FC) and Fluorescence Microscopy

Cell counts were determined individually using a flow cytometer (Cytoflex; Beckman Coulter, Brea, CA, USA) equipped with blue and red lasers operating at excitation wavelengths of 488 nm and 640 nm, respectively. The acquisition rate was set to 10,000 events per second, and data were analyzed using CytExpert (Version V3.12.0) software (Beckman Coulter). Side scatter (SSC) and forward scatter (FSC) parameters were recorded using the 488 nm laser in conjunction with a 488 nm band-pass filter. C. kluyveri cells were detected via fluorescein isothiocyanate (FITC)-labeled probes, with excitation at 488 nm and fluorescence emission measured using a 525/32 nm band-pass filter (FITC filter set). C. carboxidivorans cells were detected via cyanine 5 (Cy5)-labeled probes, excited at 640 nm, with fluorescence emission measured using a 660/20 nm band-pass filter (APC-Cy7 filter set). An acquisition threshold was applied in the SSC channel to minimize background noise and exclude interference from PBS components. Cell populations were quantified by the gating strategy predefined in the software. OD₆₀₀ for each strain was calculated based on strain-specific linear correlation factors previously determined from cell count data. These correlations were derived by analyzing manually mixed cultures of known OD₆₀₀, each comprising different proportions of C. kluyveri and C. carboxidivorans, summed to a total OD₆₀₀ of 0.5. After confirmation of the values, the conversion factors were obtained from Bäumler et al. (2021) [21], with values of 5.1 ± 0.03 × 10⁻⁵ (cell counts per 10,000 events) for C. kluyveri and 5.3 ± 0.1 × 10⁻⁵ (cell counts per 10,000 events) for C. carboxidivorans. Unlabeled samples served as negative controls. The CDW concentrations of both strains were subsequently calculated using linear correlation factors between OD₆₀₀ and CDW as described above.

Image acquisition of gel slices was performed using a lightning confocal laser scanning microscope (SP8; Leica Microsystems, Wetzlar, Germany) operated via LAS X software (version 3.5.7.23225). Imaging was conducted using a 10×/0.45 air objective and an HC PL APO 40×/1.10 water immersion objective. Fluorophore excitation was achieved using diode lasers at wavelengths of 405 nm, 488 nm, 552 nm, and 638 nm. The system was operated in photon integration mode with a photomultiplier tube (PMT) detector to optimize signal sensitivity. Fluorescence emission was recorded across defined spectral windows: 493–548 nm, 557–643 nm, and 646–700 nm, enabling the selective detection of specific fluorophores such as Cy3, FITC, and Cy5. A pinhole diameter corresponding to 1.00 Airy units was employed to enhance the axial resolution. Image acquisition settings included a frame accumulation and line averaging value of 1, with unidirectional scanning at a speed of 100 Hz.

3. Results and Discussion

3.1. Inoculation Temperature

Previous studies have demonstrated that C. kluyveri can be successfully immobilized in agar hydrogel, maintaining both metabolic activity and product formation under anaerobic conditions [33]. This strategy has proven effective in stabilizing chain elongation processes. Before advancing toward biofilm-based co-culture applications, it is essential to evaluate whether C. carboxidivorans—similar to C. kluyveri—is physiologically compatible with immobilization in agar-based hydrogel matrices. A critical prerequisite is that the organism tolerates short-term exposure to the temperatures required for gel liquefaction and solidification, which typically involve initial heating steps above ambient temperature.

To this end, the effect of inoculation temperature on C. carboxidivorans was investigated in suspended batch fermentations in anaerobic flasks under autotrophic conditions. Inoculations were carried out at 25 °C, 45 °C, and 70 °C. Following inoculation, cultures were transferred to a shaking incubator at 37 °C, which was reached within the cultures within 10–30 min depending on the initial temperature. A synthetic gas phase composed of CO₂ and CO in a 1:4 ratio at 2.0 bar served as the substrate, while product and biomass formation were monitored over a 7-day period (Figure 1).

Across all conditions, C. carboxidivorans exhibited growth and metabolic activity, but the lag phase duration varied depending on the inoculation temperature. At 25 °C and 45 °C, cell growth and product formation initiated immediately after process start. In contrast, inoculation at 70 °C resulted in a pronounced delay, with observable metabolic activity occurring only after 24 h. The extended lag phase at 70 °C coincided with a delayed pH drop, and slower product formation of acetate, ethanol, and 1-butyrate, and 1-butanol.

Despite these differences in process dynamics, end-point values after 144 h were largely comparable across all test conditions, while the process termination was attributed to the pH drop, which reached values outside the adequate range of metabolic activity of C. carboxidivorans. Biomass concentrations stabilized at approximately ~0.55 g L⁻¹, and the pH levelled off at ~pH 4.8. Acetate and ethanol accumulated to ~3.0 g L⁻¹, and ~0.2 g L⁻¹, respectively, while final concentrations of 1-butyrate and 1-butanol reached ~0.3 g L⁻¹ and ~0.04 g L⁻¹. These results suggest that C. carboxidivorans is capable of recovering from transient thermal exposure up to 70 °C, although metabolic reactivation is considerably delayed.

The data indicate that higher inoculation temperatures slow down process initiation without impairing the overall yield, pointing to a temporary metabolic suppression rather than loss of viability. Such resilience is particularly relevant for bioprocesses involving thermally activated carrier materials, which may expose cells to elevated temperatures during immobilization. Given that comparable process performance was observed for inoculation temperatures ≤ 45 °C without notable delays, future immobilization strategies should aim to remain below this threshold to ensure rapid process onset and avoid thermally induced lag phases.

3.2. Suspended Versus Immobilized Cells

The performance comparison of suspended versus immobilized C. kluyveri has been carried out before [33]. To assess the performance of C. carboxidivorans under immobilized conditions, batch fermentations in anaerobic flasks were conducted using either a synthetic agar hydrogel biofilm or freely suspended cells. Both configurations were inoculated with the same initial biomass concentration (CDW₀ = 0.13 g L⁻¹, referred to the liquid phase volume). For the biofilm setup, C. carboxidivorans was embedded in 1.8% (w/v) agar hydrogel at 45 °C. The immobilized biomass was subsequently overlaid with 100 mL of fermentation medium, whereas the other flasks were directly inoculated with suspended cells in 100 mL medium. All flasks were incubated under strictly anaerobic conditions at 37 °C and 100 rpm, with a gas mixture of CO₂ and CO (1:4) at 2.0 bar absolute pressure. The batch process was monitored over 7 days, and the results are summarized in Figure 2.

Throughout the fermentation, both culture types showed similar metabolic behavior with slightly slower reaction rates for the immobilized cells. As products were formed, the pH declined steadily from approximately pH 6.0 to pH 4.7 in both experiments. In contrast to the experiments with C. kluyveri [33], biomass accumulation in the suspension was observed in the flasks with freely suspended C. carboxidivorans and with immobilized cells. In addition to this, a shift in final product concentrations from shorter- to longer-chain carbon products can be observed with immobilized cells. Final concentrations for acetate, ethanol, and 1-butyrate were 3.19 ± 0.60 g L⁻¹, 0.58 ± 0.05 g L⁻¹, and 0.22 ± 0.03 g L⁻¹ with the suspended, and 2.41 ± 0.49 g L⁻¹, 0.34 ± 0.05 g L⁻¹, and 0.36 ± 0.04 g L⁻¹ with the immobilized cells. Moreover, 1-butanol concentrations in both cultures were below 0.1 g L⁻¹.

The slower reaction rates observed between the two batch processes are likely attributable to the physiological effects of immobilization. It is conceivable that some loss of cell viability occurred during the embedding of C. carboxidivorans in the molten agar, which may have reduced the effective number of metabolically active cells at the start of the process. In addition, the spatial confinement of cells within the hydrogel may have introduced slight diffusion limitations or local variations in pH, which could have influenced activity and growth dynamics.

The presence of suspended cells observed in the immobilized setup suggests that C. carboxidivorans, unlike C. kluyveri in anaerobic flasks, is capable of moving within the agar hydrogel matrix and even moving out of it. However, the fact that the concentration of cells in suspension is significantly lower than in the experiment only with suspended cells, while product formation remains similarly high, indicates that the immobilized cells contributed substantially to overall product formation, which demonstrates metabolic activity of C. carboxidivorans within the agar hydrogel. While ethanol production remained unchanged, acetate formation was reduced within the immobilized approach. Consequently, the pH decreased more slowly, allowing the cells to remain in the acidogenic phase for a more extended period and thus continue growing until the end of the cultivation. The onset of 1-butyrate production was also delayed and extended over a longer duration, which appears to be linked to the prolonged cell growth. These findings suggest that the pH trajectory is a key factor influencing the overall metabolic dynamics.

Nonetheless, the final yields of key fermentation products were comparable, indicating that C. carboxidivorans remains robust and catalytically active even when immobilized in agar hydrogel. The higher amount of the longer-chain products can even be counted as an improvement, since the longer=chain products are more valuable. These findings support the feasibility of using synthetic biofilms for autotrophic processes involving this strain and pave the way for co-immobilization strategies with C. kluyveri in structured biofilm reactors.

3.3. Increasing Ethanol Production

The ethanol concentrations achieved in these batch experiments with C. carboxidivorans are relatively low. However, for a co-culture with C. kluyveri to function effectively, it is crucial that the substrates acetate and ethanol are produced in sufficiently high amounts. Moreover, C. kluyveri requires a supra-stoichiometric ratio of ethanol-to-acetate [39]. Therefore, increasing the ethanol production of C. carboxidivorans is a necessary first step. Several strategies are available for this purpose, including continuous sulfur feeding, variation of the trace element composition, and increasing the initial cell density [15,35,40]. In the following, the impact of continuous sulfur supply was investigated using the biofilm reactor previously described by Herzog et al. (2025) [33].

The continuously gassed biofilm reactor was operated at 37 °C and pH 6.0 in the aqueous phase with a constant gas flow of CO (800 mbar) and CO₂ (200 mbar). C. carboxidivorans was immobilized in a planar agar-based hydrogel with an initial biomass concentration of 0.88 g L⁻¹ related to the biofilm volume. Two distinct Na₂S feeding rates were applied, 0.01 mmol S L⁻¹ h⁻¹ and 0.08 mmol S L⁻¹ h⁻¹, to assess their effect on growth and product formation. The experiments were performed as single batch runs for each condition. The results are shown in Figure 3.

In the experiment with the higher sulfur feeding rate, a markedly increased acetate production was observed immediately after process initiation. Ethanol production began to rise steeper than in the other experiment after approximately 72 h but subsequently reached significantly higher levels as well. From 144 h onward, 1-butyrate formation also exceeded that of the experiment with the lower sulfur feed. Final product concentrations were substantially higher under high sulfur conditions, with acetate reaching 0.85 g L⁻¹ compared to 0.54 g L⁻¹, ethanol at 0.53 g L⁻¹ versus 0.06 g L⁻¹, and 1-butyrate at 0.06 g L⁻¹ compared to 0.01 g L⁻¹. Moreover, 1-hexanoate was formed only in trace amounts in both batch experiments. OD₆₀₀ in the suspension showed similar behavior in both setups, rising slightly after the start of the process and reaching a maximum of approximately 0.32 at 50 h before gradually declining.

The observed enhancement of product formation under high sulfur feeding conditions highlights the critical role of sulfur in the autotrophic metabolism of C. carboxidivorans.

While 1-hexanoate remained low in both setups, the increased 1-butyrate formation under sulfur-rich conditions indicates that chain elongation by C. carboxidivorans benefits from higher precursor availability. The similar optical density profiles across both conditions suggest that the metabolic activity rather than biomass formation was affected by sulfur availability. These findings emphasize the importance of an increased sulfur feeding in biofilm-based syngas fermentation with C. carboxidivorans to maximize product concentrations.

To quantitatively estimate the contribution of suspended cells to overall product formation in the biofilm reactor with C. carboxidivorans, both an optimistic and a pessimistic scenario are evaluated—following the methodological framework previously introduced for C. kluyveri [33].

In the optimistic scenario, it is assumed that biomass growth within the biofilm proceeds analogously to suspended cultures and occurs in parallel throughout the process. This assumption is supported by the observation that the total product formation in the immobilized C. carboxidivorans experiments was comparable to that of suspended cells at similar initial cell densities (Figure 2). Based on the observed growth in these experiments, in which the CDW concentration increased from 0.13 g L⁻¹ to a maximum of 0.86 g L⁻¹, corresponding to a 6.6-fold increase, the maximum CDW concentration achievable can be estimated applying this multiplication factor. With this, a maximum biofilm-associated biomass concentration of approximately 5.8 g L⁻¹ is estimated, corresponding to 0.31 g L⁻¹ related to the liquid volume. Assuming a similar growth pattern in the biofilm as exhibited in suspension, this biomass concentration would be reached after 50 h. The observed suspended CDW concentration, which peaked at 0.15 g L⁻¹, would therefore represent approximately 33% of the total biomass, suggesting a contribution of 33% to product formation, if CO is solved in the aqueous phase at the same concentration as in the biofilm. In the pessimistic scenario, no cell growth occurs within the hydrogel matrix. Under this assumption, considering the peak of the suspended biomass concentration of 0.15 g L⁻¹, the suspended fraction would represent about 76% of the active biomass. However, both scenarios likely overestimate the influence of suspended cells, as the CO supply is reduced to the liquid phase after passing a biofilm layer of 4 mm thickness.

3.4. Synthetic Dual-Layer Biofilm Co-Culture

The synthetic biofilm was inoculated with high cell densities of both microorganisms, corresponding to 5.82 g L⁻¹ C. carboxidivorans and 3.78 g L⁻¹ C. kluyveri in each layer of the dual-layer biofilm. The reactor was operated at 37 °C with a continuous gas flow of syngas (CO:CO₂ = 4:1; total gas flow = 1 L h⁻¹), and the pH was maintained at pH 6.0 using automated dosing of 2 M H₂SO₄ or 3 M NaOH. The medium contained the trace element solution TESnew, and elemental sulfur was supplied continuously at a rate of 0.08 mmol S L⁻¹ h⁻¹. The experiment was carried out in duplicate, and the results are shown in Figure 4.

It can be observed that the OD₆₀₀ of the dual-layer biofilm reactor increased slowly but steadily in suspension over the first ~120 h, followed by a plateau phase at approximately 0.35. Regarding the product dynamics, a distinct increase in acetate concentration was observed during the first 120 h, followed by a steady decline. Ethanol concentration peaked after approximately 24 h and then gradually decreased. After 192 h, 1-butyrate reached a concentration of 0.11 ± 0.02 g L⁻¹, and 1-hexanoate reached 0.05 ± 0.03 g L⁻¹.

The slow but steady increase in OD₆₀₀ in the medium during the initial 120 h suggests gradual cell release from the biofilm into the surrounding liquid phase. This phenomenon is consistent with previous observations made with C. kluyveri or C. carboxidivorans in the same biofilm reactor, indicating that shear forces or natural biofilm dynamics contribute to the detachment of the microbial cells.

The product profiles indicate a parallel metabolic interplay rather than a strictly sequential one. During syngas fermentation by C. carboxidivorans in a stirred tank bioreactor, both acetate and ethanol were produced in roughly equal amounts at a CO partial pressure of 800 mbar [21]. However, since C. kluyveri requires a higher ethanol-to-acetate ratio of approximately 2.5 g g⁻¹ [39] for efficient chain elongation, ethanol was consumed more rapidly than acetate. As a result, ethanol concentrations increased only during the initial phase until ~24 h. Despite the presence of both substrates for chain elongation, the concentrations of 1-butyrate and 1-hexanoate showed only a brief initial increase, followed by stagnation rather than continuous accumulation. This suggests that chain elongation occurred early but was subsequently limited—likely due to insufficient ethanol availability. Notably, a renewed increase in 1-butyrate concentration was observed only after acetate levels began to decrease, indicating that a shift in substrate availability re-enabled chain elongation activity within the co-culture.

In comparison to the results obtained with the C. carboxidivorans mono-biofilm, shown in Figure 4 as light grey inverted triangles, a shift in product formation from short-chain carbon products (acetate and ethanol) to longer-chain products (1-butyrate and 1-hexanoate) was observed with the dual-layer biofilm. In the co-culture, the formation of these longer-chain products begins immediately after the start of the process, whereas with the mono-layer biofilm of C. carboxidivorans, a substantial increase was only observed after 144 h. This demonstrates that chain elongation by C. kluyveri is functional and shows the greater potential of the co-culture for efficient production of longer-chain products.

To gain insights into the spatial distribution and relative abundance of the two species in the biofilm and in suspension, fluorescence in situ hybridization (FISH), fluorescence microscopy, and flow cytometry analyses were performed (see Figure 5), allowing for further evaluation of the distribution of suspended and immobilized cells. The concentrations of C. carboxidivorans and C. kluyveri in the liquid phase at the end of the autotrophic batch process, as determined by FISH and flow cytometry, are shown in blue. In parallel, the initial concentrations of each microorganism immobilized in the synthetic biofilm—normalized to the total liquid volume of 600 mL—are depicted in grey. The data reveal that by the end of the process, the suspended cell population was almost evenly distributed between C. carboxidivorans and C. kluyveri. Notably, the initial biomass concentrations within the biofilm were significantly higher than those observed finally in suspension. It was approximately 6.9-fold higher for C. carboxidivorans and 4.5-fold higher for C. kluyveri. Interestingly, the presence of C. carboxidivorans, which was initially confined to the upper layer of the biofilm, was also detected in the liquid phase, indicating migration through the C. kluyveri biofilm during the process. To gain a deeper understanding of this phenomenon, post-process biofilm samples were analyzed using FISH and fluorescence microscopy, following the protocol previously published [41].

Microscopic analysis of the post-process biofilm revealed a distinct spatial distribution of the two microbial populations. In the upper biofilm region, only red-labeled cells corresponding to C. carboxidivorans were detected, whereas in the lower region, both green-labeled C. kluyveri and a minority of C. carboxidivorans cells were present. This observation strongly suggests that C. carboxidivorans was capable of migrating through the agar hydrogel matrix during the course of the process, while C. kluyveri remained confined to its initial location within the lower biofilm layer. A possible explanation for this differential mobility lies in the physiological and structural differences between the two organisms. C. carboxidivorans is known to exhibit a more motile phenotype under autotrophic conditions and has been reported to possess peritrichous flagella [12], potentially enabling limited movement even within viscous or porous environments such as agar hydrogels. In contrast, C. kluyveri lacks a driving force to migrate to the upper regions of the biofilm, as the elevated CO concentration present there exerts an inhibitory effect on its activity, which likely explains its stationary behavior within the biofilm.

To assess the effect of the upper C. carboxidivorans layer on CO partial pressure reduction, thereby potentially creating a more favorable environment for C. kluyveri in the lower region of the biofilm, we tried to estimate the CO consumption in the biofilm with immobilized C. carboxidivorans. Linear fits of the acetate production curve over the first 50 h measured with a mono-layer biofilm of C. carboxidivorans (see Figure 3) lead to CO consumption rates of r_CO = 2.15 × 10⁻⁴ mol m⁻³ s⁻¹ based on the assumption that 4 mol of CO are consumed per mol of acetate produced [42]. This CO consumption rate by the C. carboxidivorans biofilm is most probably underestimated, because the initial cell density was 6.6 times higher in the dual-layer biofilm reactor.

To model CO diffusion, the CO diffusion coefficient in water at 37 °C was taken as 2.3 × 10⁻⁵ cm² s⁻¹ [43]. According to findings from glucose diffusion studies, the diffusion coefficient in agar hydrogel is approximately 5% lower than in water [44], resulting in an effective CO diffusion coefficient in the agar hydrogel of D_COgel = 2.19 × 10⁻⁹ m² s⁻¹.

With the assumption that the agarose-based hydrogel exhibits thermodynamic equilibrium at the gas–hydrogel interface comparable to water, the equilibrium concentration of CO at the gas–hydrogel interface ${\mathrm{c}}_{\mathrm{C}\mathrm{O}}^{\mathrm{*}}$ was calculated using Henry’s law (Equation (1)), based on a partial pressure ${\mathrm{p}}_{\mathrm{C}\mathrm{O}}$ of 800 mbar CO, and the Henry’s constant ${\mathrm{k}}_{\mathrm{H}}^{\mathrm{C}\mathrm{O}}$ at 37 °C (${\mathrm{k}}_{\mathrm{H}}^{\mathrm{C}\mathrm{O}}$ = 8.2 × 10⁻⁶ m³ Pa⁻¹ mol⁻¹). This results in an equilibrium concentration of CO at the gas–hydrogel interface of ${\mathrm{c}}_{\mathrm{C}\mathrm{O}}^{\mathrm{*}}$ = 0.656 mol m⁻³.

$${\mathrm{c}}_{\mathrm{C}\mathrm{O}}^{\mathrm{*}}={\mathrm{p}}_{\mathrm{C}\mathrm{O}}·{\mathrm{k}}_{\mathrm{H}}^{\mathrm{C}\mathrm{O}}$$

(1)

Assuming steady-state diffusion, the CO concentration profile in the planar hydrogel was estimated using Fick’s second law. The equation is shown below, with ${\mathrm{c}}_{\mathrm{C}\mathrm{O}}\left(\mathrm{z}\right)$ being the concentration of CO depending on the biofilm depth $\mathrm{z}$, ${\mathrm{r}}_{\mathrm{C}\mathrm{O}}$ the CO consumption rate, and ${\mathrm{z}}_{\mathrm{m}\mathrm{a}\mathrm{x}}$ the depth into the biofilm where CO concentration approaches 0 mol m⁻³:

$${\mathrm{c}}_{\mathrm{C}\mathrm{O}}\left(\mathrm{z}\right)=-{\displaystyle \frac{{\mathrm{r}}_{\mathrm{C}\mathrm{O}}}{2{\mathrm{D}}_{\mathrm{C}\mathrm{O}\mathrm{g}\mathrm{e}\mathrm{l}}}}{·\mathrm{z}}^2+\left({\displaystyle \frac{{\mathrm{r}}_{\mathrm{C}\mathrm{O}}{·\mathrm{z}}_{\mathrm{m}\mathrm{a}\mathrm{x}}}{2{\mathrm{D}}_{\mathrm{C}\mathrm{O}\mathrm{g}\mathrm{e}\mathrm{l}}}}-{\displaystyle \frac{{\mathrm{c}}_{\mathrm{C}\mathrm{O}}^{\mathrm{*}}}{{\mathrm{z}}_{\mathrm{m}\mathrm{a}\mathrm{x}}}}\right)·\mathrm{z}+{\mathrm{c}}_{\mathrm{C}\mathrm{O}}^{\mathrm{*}}$$

(2)

Assuming that the active biomass is homogeneously distributed in the hydrogel and has the same specific activity everywhere, the penetration depth at which the CO concentration theoretically reaches zero was thus estimated to be approximately 3.66 mm. Given the 4 mm thickness of the C. carboxidivorans biofilm layer, this indicates a significant reduction of CO before reaching the underlying C. kluyveri layer. This estimation shows that C. carboxidivorans in the upper biofilm is able to establish a favorable CO reduction for the co-cultivated C. kluyveri immobilized in the lower biofilm. In addition, it is most probable that the CO concentration in the liquid phase is significantly reduced compared to the upper biofilm. Thus, the C. carboxidivorans cells in suspension will make a significantly smaller contribution to product formation than was previously estimated on the basis of the biomass concentration.

A comparison of the final concentrations of 1-butyrate and 1-hexanoate with data from the literature with suspended cells reveals that, under comparable process conditions, higher end concentrations have been achieved in stirred-tank reactors [21]. However, it is important to note that the reaction volumes in those studies were significantly larger. For example, in the study by Bäumler et al. (2021), the reaction volume was 1 L, whereas the active biofilm volume in the biofilm reactor used here was only 32 mL. Therefore, product formation rates normalized to the active volume provide a more meaningful basis for comparison. Bäumler et al. (2021) reported maximum product formation rates of approximately 0.51 g L⁻¹ d⁻¹ 1-butyrate and 0.12 g L⁻¹ d⁻¹ 1-hexanoate [21]. In contrast, the corresponding product formation rates observed in the present study were 0.46 g L⁻¹ d⁻¹ 1-butyrate, and 0.91 g L⁻¹ d⁻¹ 1-hexanoate, respectively. It should be noted that 1-butyrate is both produced and consumed simultaneously, which likely accounts for the slightly lower rate in the dual-biofilm reactor. The formation rate of 1-hexanoate in the biofilm reactor was approximately 7.6 times higher than that reported with suspended cells in a stirred tank bioreactor [21], proving the efficiency of the biofilm-based process.

4. Conclusions

This study demonstrates that C. carboxidivorans is physiologically compatible with agar-based hydrogel immobilization and maintains robust autotrophic metabolism. While higher inoculation temperatures temporarily delayed metabolic activation, final product yields were unaffected, confirming thermal resilience during immobilization procedures. Comparative batch fermentations with suspended and immobilized cells revealed slightly slower kinetics in the latter but similar endpoint concentrations, indicating that the immobilized cells remain metabolically active. Crucially, the application of continuous sulfur feeding significantly enhanced the ethanol production of C. carboxidivorans, an essential prerequisite for syntrophic interaction with C. kluyveri. In the dual-layer biofilm reactor with the synthetic C. carboxidivorans biofilm in contact with the CO-containing gas phase, the stable formation of 1-butyrate and 1-hexanoate confirmed successful chain elongation. Spatially resolved microscopy and quantitative analysis indicated that C. carboxidivorans actively migrated within the hydrogel matrix and reduced local CO concentrations, thereby enabling C. kluyveri to operate in a less inhibitory CO environment.

Together, these findings demonstrate the potential of a synthetic biofilm for syngas-based chain elongation. However, the current laboratory-scale biofilm reactor setup is not scalable and exhibits an unfavorable biofilm-to-liquid volume ratio, which limits product concentrations and volumetric productivities in the liquid phase. To render this approach scalable and industrially relevant, novel bioreactor designs must be considered. One promising strategy involves, e.g., membrane contactors, such as hollow fiber modules with adaptable fiber wall thickness and porosity, where hydrogel-embedded C. carboxidivorans could be immobilized within the pores of syngas-perfused fibers, while C. kluyveri forms an outer biofilm layer. This configuration could significantly reduce the external liquid volume and improve volumetric performance. Alternatively, spherical dual-layer hydrogel beads could be used in a bubble column or gas-lift bioreactor, with C. kluyveri encapsulated in the core and C. carboxidivorans coating the surface. This spatial separation would prevent CO exposure to C. kluyveri, while enabling, in addition, efficient conversion of the medium-chain organic acids to longer-chain alcohols by C. carboxidivorans.

Nevertheless, this study has some limitations, including the laboratory-scale setup, the absence of long-term stability tests, and the lack of experiments with real syngas streams containing impurities. Also, an optimization in sulfur feeding is lacking so far. Addressing these aspects in future work will be essential to further evaluate the robustness and industrial applicability of the synthetic biofilm approach.