The Limits of Three-Dimensionality: Systematic Assessment of Effective Anode Macrostructure Dimensions for Mixed-Culture Electroactive Biofilms.

This study analyzes the biofilm growth and long-term current production of mixed-culture, electrochemically active biofilms (EABs) on macrostructured electrodes under low-shear-force conditions. The channel dimensions were altered systematically in the range 400 μm to 2 mm, and the channel heights were varied between 1 and 4 mm to simulate macrostructures of different scales. Electrodes with finer-structured surfaces produced higher current densities in the short term owing to their large surface area but were outperformed in the long term because the accumulation of biomass led to limitations of mass transfer into the structures. The best long-term performance was observed for electrodes with channel dimensions of 1×4 mm, which showed no significant decrease in performance in the long term. Channels with a diameter of 400 μm were overgrown by the biofilm, which led to a transition from 3 D to 2 D behavior, indicating that structures of this scale might not be suitable for long-term operation under low-shear-stress conditions.

Introduction

Microbial electrochemical technologies (METs) are based on a functional implementation of biological catalysts, mostly in the form of bacteria, into electrochemical cells.1 In primary METs, the living organisms are cultivated by using electrodes either as electron donor or acceptor,1 enabling complex electrocatalytic reactions under mild conditions.2 The microbial fuel cell (MFC) is an example of such a bioelectrochemical system (BES), in which bacteria growing on anodes generate electrical current whilst degrading the organic content in wastewater.3, 4, 5 Some of the most prominent microorganisms that find application in MFCs belong to the family of Geobacteraceae. They can be enriched easily from wastewater, sludge, or sediments by using an electrode serving as terminal electron acceptor and an appropriate carbon source (e.g., acetate). Under these conditions, the growth of mixed‐culture electroactive biofilms (EABs) is often dominated by Geobacter sulfurreducens 6, 7 or related Geobacter species.8 The performance of the resulting biofilm anode is a critical application parameter, and thus its enhancement through, for example, use of improved materials,9, 10, 11, 12 surface modifications,13, 14 and electrode structuring15, 16, 17, 18 has been the subject of a wealth of research over recent years. In general, the goal is to provide optimum growth conditions for the electroactive bacteria whilst also maximizing the ratio of anode surface area to reactor and electrode volume. The introduction of small structures (i.e., nanometer or micrometer scale) can drastically increase the electrochemically active surface area of an electrode. However, recent studies have shown that although structures that are small compared to the biofilm thickness can influence biofilm attachment and initial growth, they do not significantly enhance long‐term performance.19, 20, 21 Macrostructures with a comparatively larger size22 can, however, drastically increase the current density, which has been shown by several studies with specific 3 D macrostructure designs.15, 17, 23 Here, clogging of the 3 D structures during the biofilm growth represents a major obstacle that needs to be overcome.16, 24 Whether a structure is susceptible to clogging because of biofilm growth depends predominately on two factors: 1) the size of the electrode structural elements relative to the biofilm thickness, and 2) the presence of shear forces resulting from medium flow inside the electrode. The presence of shear forces can induce the production of exopolysaccharides (EPSs)25 as well as detachment processes at the biofilm solution interface,26, 27 which leads to the formation of more compact biofilms.25 Therefore, controlled application of shear forces can be used to control the biofilm thickness, and thus, avoid clogging even for electrodes consisting of relatively small structural elements.28 It can, however, be interesting to consider the case in which no or no significant shear forces are applied, particularly for MFCs used in wastewater treatment, for which the shear forces applied to biofilms tend to be relatively low29 and the additional energy required for sustained medium flow through an electrode would decrease the net power gain of the system. In this case, the occurrence of clogging will depend predominantly on the electrode geometry and the size of the electrode structural elements for the given biological system and growth conditions. In this context, our study aims to explore the influence of systematically altered channel dimensions of macrostructured electrodes on the growth and performance of a Geobacter‐dominated, mixed‐culture EAB. The study shows how a better understanding of EAB growth patterns under controlled conditions can help to understand the influence of biofilm growth and clogging of 3 D structures on the bioanode performance. Ultimately, this knowledge can be used to optimize the ratio of electrode surface to reactor and electrode volume for any given set of conditions.

Experimental Section

Experimental setup

The bioelectrochemical cells used in this study were composed of customized, 1 L capacity, round‐bottom glass reactors, each containing 600 mL of modified standard cultivation medium (phosphate buffer, 7.5 mL mineral and vitamin solutions)30 containing 10 mm acetate. A secondary, Geobacter‐dominated bacterial culture31 was used, which was enriched from the primary wastewater of the wastewater treatment plant Steinhof, Braunschweig (Germany). The cell was deaerated at startup and after each batch cycle by sparging with nitrogen for at least 15 min. Upon replacing the medium after each cycle, the new medium was also sparged for 15 min prior to the replacement. The cell was sealed by using butyl rubber stoppers to maintain anaerobic conditions during the cycles.

The electrodes were placed in the reactor according to Figure S1 (Supporting Information) and were connected to the potentiostat with isolated copper wires. The reference electrode (Ag/AgCl, sat. KCl, Sensortechnik Meinsberg, Germany) was placed in the center of the reactor. The counter electrode was a graphite rod (CP‐Graphitprodukte GmbH, Germany) of 1.5 cm diameter, with a surface area of over 40 cm2. It was placed approximately 3 cm from the reference electrode. Six working electrodes were placed in the reactor and positioned concentrically and at equal distances around the reference electrode with their active areas facing outward to allow for periodical photography of the electrodes without interference with the experiment. The working electrodes were made from rectangular polycrystalline graphite blocks (CP‐Graphitprodukte GmbH, Germany). The electrode structures were produced by using saw blades of different thicknesses to cut into the graphite, creating channels of different widths and heights (Table S1, Figure S2 in the Supporting Information), whereas unmachined graphite blocks served as control electrodes. Epoxy resin (R&G Faserverbundwerkstoffe GmbH, Germany) was used as insulation for the side and bottom areas of the working electrodes. A typical data set was based on two reactors with 12 electrodes (4 triplicates) in total. For a better comparability between the individual cells, the working electrodes were distributed such that every cell contained at least one electrode of each kind. A list of all experiments and electrodes used in this study is provided in Table S1 in the Supporting Information.

All anodes were poised at 0.2 V vs. Ag/AgCl, and the bioelectrocatalytic currents were measured chronoamperometrically (CA) in a multi‐anode setup with a VMP‐3 potentiostat (Bio‐Logic Science Instruments SAS, France) and an N‐Stat Box (Bio‐Logic Science Instruments SAS, France). The experiments were conducted for up to nine cycles, corresponding to a duration of up to 70 days. The currents were recorded for each electrode, with the maximum electrode current of the individual batch cycles serving as a key parameter, reflecting the catalytic biofilm performance without substrate limitation. These maximum currents can be considered as values that would be achieved in continuous‐mode experiments with a short hydraulic retention time of the substrate solution.

The temperature was kept constant at 35 °C, interrupted only briefly for medium changes between the cycles. All reactors were stirred gently with a magnetic stirrer (250 rpm) to avoid concentration gradients in the reactor. The mild stirring conditions can be considered to create only negligible shear forces at the biofilm electrodes. For systems in which biofilm growth is influenced strongly by shear forces, growth behavior might differ significantly and should therefore be investigated separately.

Optical and microscopic biofilm electrode observation and analysis

The thickness of biofilms can vary significantly both over time and location (especially for older biofilms, Figure 2 c, d). Therefore, sample measurements (e.g., by confocal laser scanning microscopy, CLSM) may have only limited significance to the question of whether clogging of electrode structures can be expected or not. Instead, an analysis of the biofilm growth behavior across the entire electrode surface is more likely to yield information that is transferable toward practical use. For this reason, both CLSM and digital photography were used as optical tools for biofilm characterization.

Figure 2

Photographic images showing bacterial growth behavior on a structured electrode with 0.8×1 mm channels: a) young biofilm (cycle 2 or ≈2 weeks), b) biofilm during maturation (cycle 4 or ≈3 weeks), c) mature biofilm (cycle 6 or ≈4–5 weeks), d) biofilm detachment (cycle 7–9 or ≈30–50 days).

CLSM analysis was conducted by using a Leica TCS SPE DM5500 confocal microscope. All images were taken as reflection images. The optical electrode characterization was performed to verify the geometry of the channels of the working electrodes (Table S1 in the Supporting Information). For this purpose, three images were taken as random samples for each electrode with a 10× air objective (Leica HC PL Fluotar 10×/0.3NA). The software Leica Map 7.0 was used for data evaluation. For biofilm characterization after termination of the bioelectrochemical experiments, the biofilm electrodes were submerged in phosphate buffer solution, and the biofilm morphology was analyzed by using a 25× water immersion objective (Leica HCX IRAPO L 25x/0.95NA). Biofilm images were processed with the software Leica Las X (Version 3.2.9702.1). High‐resolution photographic images were obtained by using a digital camera (Canon EOS 5D, 100 mm macro lens). Pictures were taken after each batch cycle, after replacement of the cultivation medium.

Results and Discussion

Long‐term biofilm growth behavior

It has been shown in several recent studies that biofilm electrode performance is directly linked to biofilm growth and/or maturation.32, 33 On the one hand, sufficient active biomass (and thus sufficient biofilm thickness) is necessary to assure high substrate turnover. On the other hand, increasing biofilm thickness may cause both electron‐32 and mass‐transport34, 35 limitations within the biofilm and may even lead to the accumulation of inactive bacteria (i.e., bacteria that do not contribute to the electrode performance) below36, 37 or on top of38, 39 the metabolically active bacteria. The values that are usually reported for the biofilm thickness of Geobacter‐dominated mixed‐culture EAB are in the range 100–200 μm.12, 24, 32, 37, 40 However, these values are generally reported without a discussion of the respective biofilm growth conditions. Biofilms growing under conditions limited by mass transfer are known to form porous structures with large surface areas.41 Additionally, biofilm density and tensile strength are known to decrease toward the biofilm–solution interface.42 In this case, detachment processes caused by shear forces applied by medium flow are likely to become relevant, in addition to cell growth, in defining the thickness of the biofilm.41 The formation of thinner and denser biofilms, in contrast, has been reported upon imposition of higher shear forces toward the biofilm.25, 43 Thus, all reports of biofilm thickness should be viewed relative to the growth conditions (substrate concentration and shear force) chosen for the respective experiments.

Biofilm growth, maturing, and aging phenomena all influence the electrocatalytic performance of a biofilm electrode, usually expressed by an initial period in increasing currents, a period of maximum performance, and a period of current decline. Thus, for analysis of biofilm growth in 3 D structures, these typical growth stages have to be taken into account. For this reason, every reactor in this study always contained a control electrode, that is, a planar, unstructured electrode serving as a reference to monitor 2 D biofilm behavior. Figure 1 shows the average maximum current densities of all control electrodes as relative values over time in relation to the highest current density measured for every electrode (a summary of all bioelectrochemical data can be found in Figures S3–S8 in the Supporting Information).

Figure 1

Average of relative maximum current densities for the control electrodes of all reactors over time (depicted as batch feeding cycles).

The current density was relatively low in the first cycle with a comparatively large standard deviation, which indicates that the biofilms were not yet fully developed at this point. Maximum current density was generally reached in the second and third cycles, after which it decreased continuously until the end of the experiment. The final current density was approximately 55 % lower than the maximum current density measured in cycle 2. This behavior is generally attributed to biofilm maturation and the limitations emerging owing to the accumulation of biomass on the electrode. The dominant limitations are assumed to be transport processes such as electron transport,32, 44 mass‐transport limitations for substrate34 and protons,45, 46, 47 as well as the accumulation of inactive cells inside the biofilm.36, 37, 38, 39 The exact contributions of the respective phenomena to the resulting anode performance and the possible influences of biofilm development and structure, however, are largely unknown. On a qualitative level, the current trends depicted in Figure 1 for the 2 D control electrodes were also observed for all structured electrodes independent of the specific structure dimensions (see discussion in the following sections).

Figure 2 documents the biofilm growth on an electrode with 0.8 mm channel diameter. Additional pictures can be found in Figures S11–S26 in the Supporting Information). Figure 2 a shows the biofilm anode at its peak performance after cycle 2, Figure 2 b at the beginning of its performance decline after cycle 4, Figure 2 c after cycle 6 when the peak current density has decreased substantially, and Figure 2 d at the end of the experiment corresponding to the final current density in Figure 1. The biofilm in Figure 2 a looks relatively thin and homogenous, whereas the biofilm in Figure 2 b seems to be thicker in some areas and shows several inhomogeneities, which become a dominant surface feature of the biofilm with increasing cycle number (Figure 2 c) together with an overall increase in biofilm thickness. Apart from an additional increase in thickness, a detachment can be observed at several locations, at which a thinner layer of biofilm becomes visible. These observations can be interpreted as follows: because the young biofilm in Figure 2 a is very homogenous, the inhomogeneities of the older biofilms can be signs of biofilm maturation. In this regard, the biofilm in Figure 2 b can be viewed as a biofilm during the maturation process and that in Figure 2 c as a biofilm that has largely matured. Biofilm dispersion and detachment could qualify as possible mechanisms leading to the observed inhomogeneities.48

Biofilm detachment denotes bacteria being transferred from the outer layer of the biofilm into the medium biofilm, whereas dispersion is described as evacuation of bacteria from the inside of the biofilm.48 Both processes are known to occur with mature biofilms48, 49, 50 and reduce the amount of biomass present on the electrode by transferring bacteria from the biofilm into the bulk medium.48, 51 This provides a good explanation for the observation that mature biofilms can provide comparatively low but stable current densities over long periods of time19, 33 as the biofilm enters a stationary phase in which growth and detachment/dispersion take place simultaneously.48, 52, 53, 54 Detachment and dispersion processes can be viewed as a reaction to starvation of parts of the biofilm owing to transport limitations.50, 51 As such, it can be assumed that the increasing appearance of inhomogeneities implies that parts of the mature biofilm are not metabolically active. Because this can affect the cohesive strength of biofilms,29, 55 it can subsequently lead to detachment of large parts of the biofilm (Figure 2 d), which has been observed especially for biofilms grown under low‐shear‐stress conditions.52, 56 It should be added here that, although optical observations of biofilm growth can provide a general description of the maturation process, more research is needed to define the underlying mechanisms and thus completely understand the maturation process of EABs.

For 3 D structured biofilm anodes, the above results mean that clogging can originate directly from the biofilm growth and also from biofilm detached from the electrode. This can increase the potential for clogging, especially for complex 3 D structures. Therefore, the relation between biofilm growth and maturation behavior and biofilm anode performance should be analyzed critically. Note that because biofilm detachment and dispersion depend largely on external factors such as medium flow, shear stress, and substrate concentration,52, 57, 58, 59 external factors will also have a significant influence on biofilm growth and maturation behavior and thus on the potential clogging of 3 D electrode structures. They should therefore always be taken into account in the electrode design process.

Bioelectrochemical performance relative to structure dimensions for partially overgrown structures

As the EAB grows in thickness (including the buildup of inactive biomass60, 61) on a 3 D structured electrode, it effectively changes the geometry of the electrode by narrowing the cross sections available for mass transfer inside the structure. Therefore, apart from the possibility of clogging, the thickness of the EAB can influence the performance of a biofilm electrode even if the 3 D structure remains accessible. Figure 3 shows side views of CLSM 3 D images taken from the channels with widths of 0.6 mm (Figure 3 a), 0.8 mm (Figure 3 b), 1 mm (Figure 3 c), and 2 mm (Figure 3 d) at the end of the experiments (i.e., after eight feeding batch cycles, corresponding to a biofilm age of more than 30 days), visualizing the biofilm growth within the channel opening. The channel openings were overgrown by the biofilm to different extents (camera pictures can be found in Figures S11–S22 in the Supporting Information). The channel of 2 mm diameter (Figure 3 d) is still wide open and easily accessible by CLSM, leaving an opening of approximately 1 mm diameter or approximately 50 % of the original diameter of the channel.

Figure 3

CLSM 3 D images of biofilms growing across a) 0.6 mm, b) 0.8 mm, c) 1 mm, and d) 2 mm channels; the scales are denoted by color bars (axial) and white scale bars (lateral).

The effective diameter of the channel of 1 mm diameter is reduced to approximately 0.4 mm, corresponding to 40 % of its original value. For the 0.8 mm channel diameter, the effective channel diameter is reduced to an even larger extent to only approximately 0.1 mm or 12 % of the original value. The 0.6 mm channel is completely overgrown by the biofilm, which could effectively block mass transfer at that spot. Because the CLSM images were always taken at the end of the cultivation experiments, at which biofilm detachment was a common phenomenon (Figure 2 d), CLSM images were taken only at electrodes for which the biofilm had not detached up to that point. Therefore, the effective channel diameters measured by CLSM should only be taken as qualitative samples considering the significant heterogeneity observed in terms of biofilm growth. This is especially relevant for the channels of 0.6 mm diameter because those were not in all cases completely overgrown before biofilm detachment occurred and should therefore be considered as partially overgrown (Figures S11–S22 in the Supporting Information). This still means that 0.8 mm constitutes the minimum diameter at which the electrode structure remained open reliably. It was also observed that the biofilm thickness was higher on the outside of the electrodes, especially at the channel openings, than on the inside of the channels, as has been observed in several studies.39, 62, 63

To evaluate the extent to which a macrostructure contributes to the electrode performance and remains accessible to unhindered current generation, we correlated the current density related to “unfolded” surface area (j unfolded), which integrates all geometric elements of the macrostructured electrodes (see Figure S2 c in the Supporting Information), to the current density of the 2 D control electrodes (j control). In the case of a fully accessible surface area, the ratio of j unfolded/j control will be 1, whereas decreasing values indicate increasing limitations in biofilm accessibility. The use of this current density ratio also eliminates contributions caused by biofilm maturation and ageing.

The results of this correlation for electrodes with different channel dimensions and during a consecutive number of batch cycles are depicted in Figure 4. The bioelectrochemical data of the individual experiments can be found in Figures S3–S8 in the Supporting Information. Figure 4 shows that the relative current densities achieved by electrodes with channels of 2 and 1 mm diameter are close to 1 and, therefore, close to those of the control electrodes. They do not decrease significantly with time, indicating that the biofilms inside the channels are not limited by mass transport into the structure despite the narrowing of the channels during biofilm growth. Electrodes with channels of 0.8 mm diameter and 1 mm height show a similar behavior. For electrodes with the channel geometries 0.8×2 mm and 0.8×4 mm and all electrodes with 0.6 mm channel diameter, a decrease in current density over time can be observed, which is probably caused by increasing mass‐transport limitations. The different behaviors amongst the electrodes with 0.8 mm diameters illustrate the role of effective channel geometries as a dominant performance‐defining factor. Here, limitations in bioelectrochemical substrate turnover caused by the narrowing of the channels owing to biofilm growth increase considerably with increasing channel height.

Figure 4

Unfolded maximum current densities for experiments 1–3 (see the Supporting Information for details) relative to the respective control electrodes over time (depicted as batch feeding cycles).

Among the electrodes with 0.8 mm channel diameter, the channel geometry of 0.8×4 mm constitutes a special case because the current density reaches a minimum in cycle 6 to increase again and reach a value similar to the other limited electrode geometries. This behavior was caused by gas bubbles accumulating inside the channels for these electrodes, especially during cycles 4, 5, and 6 (Figure S21 in the Supporting Information). If the volume of gas entrapped in the structure is sufficiently high, it can limit mass transfer inside the structure. Additionally, it reduces the electrode's active surface area because of the contact area of gas bubbles with the electrode surface. Although gas bubbles were observed on several electrodes, an influence of the retention of gas bubbles on the bioelectrochemical performance was only evident for the 0.8×4 mm channel geometry. Visual observations indicate that gas bubbles were not retained permanently in the structures but left the structures through the channel openings, as suggested by protrusions along the narrowed channel openings of some electrodes (Figure S10 in the Supporting Information). The retention of gas bubbles in 3 D bioanodes has been observed before.64 Although there are several factors influencing the entrapment of gas bubbles in 3 D structures, it can be expected in general that the susceptibility of a structure to the retention of gas bubbles will depend on the size of its structural elements and the flow conditions inside the structure.65, 66 Therefore, entrapment of gas bubbles can be expected to be of relevance for comparatively fine structures or structures that are overgrown to such an extent that stable cavities are formed in which gas bubbles can be retained. In this case, however, it is likely that mass transport inside the structure will be severely limited even without the presence of the bubbles.

To analyze the performance gain by the 3 D structures in relation to the 2 D control electrodes, we plotted the ratio j projected/j control against the relative increase in the geometric surface area, given as the ratio A unfolded/A projected (Figure 5). Here, every specific surface area ratio stands for a distinct electrode type, with the channel geometry and number of channels cut into the electrode surface determining the respective surface area. For every electrode type, the development of the relative current density during the consecutive batch cycles is depicted by symbols with increasing color saturation.

Figure 5

Projected maximum current densities for experiments 1–3 relative to the respective control electrodes plotted against the relative surface area gain provided by the electrode structure. The development of the relative current density of the individual electrode types during the consecutive batch cycles is illustrated by increasing color saturation. The solid line depicts conditions not impacted by mass‐transport limitations, whereas current densities along the dashed line are mass‐transfer limited.

Similarly to Figure 4, the electrodes can be separated into two groups on the basis of their long‐term performance: 1) all electrodes with channel diameters of 1 and 2 mm as well as the electrodes with 0.8×1 mm channels (straight lines), and 2) other electrodes with a diameter of 0.8 mm as well as all electrodes with a diameter of 0.6 mm (dashed lines). The first group can be viewed as suitable electrode geometries because their performance is not significantly affected by biofilm growth and stays relatively constant in the long term, whereas the current densities of the other electrodes drop over time as their effective geometry becomes more unfavorable (for unfolded current densities see Figure S27 in the Supporting Information). Furthermore, it is seen from Figure 5 a that 1×4 mm is the most effective channel geometry for the system in terms of performance. However, because of the significant narrowing of the channels and the additional possibility of biofilm detachment, a more open structure, that is, the 2×4 mm geometry, may be advantageous for long‐term operation.

For the present case of a low‐shear‐force environment, our results agree with the statement that millimeter‐sized electrode structures are preferable for a system using mixed‐culture EAB.15, 67 However, it should be added that those values have to be evaluated as effective structure sizes considering the expected biofilm thickness and also the inner geometry of the electrodes to ensure adequate mass transfer into the electrode structure.68

Complete overgrowth of 3 D electrode structures by EAB results in a 2 D electrode behavior

If the openings of a 3 D electrode structure become completely overgrown or clogged by biomass, the electrode will produce current densities similar to a 2 D electrode of the same footprint area. This behavior has been observed in the past for different electrode structures.16, 24, 68 A Geobacter sulfurreducens pure culture biofilm was able to overgrow a structure with pores of up to 50 μm,68 and mixed‐culture biofilms have been reported to overgrow structures with sizes of up to 200 μm.16, 24 It is to be expected that the minimum structure size will depend not only on the biological system but also on external factors such as medium flow rate and the resulting shear forces inside and outside the structure. It should, however, be relatively reproducible if those factors are kept constant.

Figure 6 shows pictures of electrodes with channel geometries of 1×1 mm and 0.4×4 mm after the second cycle (Figures 6 a, b) and after the seventh cycle (Figures 6 c, d). Whereas the channels with a diameter of 1 mm remain largely open throughout the experiment, the channels with 0.4 mm diameter are open in the early stages of the experiment (Figure 6 b) but completely overgrown by the biofilm in the long term (Figure 6 d). The impact of biofilm growth on the electrode performance can be seen in Figure 7. Electrodes with a channel geometry of 1×1 mm produce significantly higher long‐term current densities with respect to the unfolded surface area than those with a diameter of 0.4 mm (Figure 7 d). However, owing to the small area increase of this structure (Table S1 in the Supporting Information), the overall increase in current density is relatively low.

Figure 6

Photographic images showing bacterial growth behavior on differently structured electrodes: a) 1×1 mm, cycle 2; b) 0.4×4 mm, cycle 2; c) 1×1 mm, cycle 7; d) 0.4×4 mm, cycle 7.

Figure 7

Chronoamperometry curves of experiment 4 (see the Supporting Information for details): a) reactor 1, b) reactor 2; maximum current densities for experiment 4: c) reactor 1, d) reactor 2.

Electrodes with channel diameters of 0.4 mm exhibit very high current densities with respect to the projected surface area in the early stages but decline in the long term, showing much lower current densities that are comparable to the control electrodes (Figure 7 a–c). This illustrates the transition from 3 D to 2 D behavior and shows that 400 μm large openings can be overgrown by mixed‐culture EAB and are thus too small to constitute efficient 3 D structures. These results indicate that several of the architectures that have been proposed in the literature as 3 D electrode designs18, 69, 70 are probably too fine to be applied under low‐shear conditions with mixed‐culture EABs. They underline the importance of taking into consideration long‐term biofilm growth behavior and expected operation conditions for future bioanode designs.

Conclusions

In this study, mixed‐culture electrochemically active biofilms (EABs) were cultivated on structured graphite electrodes with channels of different dimensions to analyze biofilm growth behavior and bioelectrochemical performance systematically. Channels of 0.4 mm diameter were completely overgrown, resulting in a transition from 3 D to 2 D behavior in the respective electrodes during the experiment. Electrodes with a channel diameter of 0.6 mm were overgrown in some areas, with parts of the electrodes staying open until the end of the experiment. Channels with diameters of 0.8, 1, and 2 mm stayed open until the end of the experiment, with biofilm growth reducing the effective channel diameter to approximately 12, 40, and 50 % of the initial values, respectively. Whereas all electrodes with channel diameters of 1 or 2 mm as well as electrodes with a channel geometry of 0.8×1 mm provided (unfolded) current densities comparable to those of the control electrodes, the relative current densities of the other electrodes with 0.8 mm diameter channels and all electrodes with 0.6 mm diameter channels decreased over time. This phenomenon was caused by the reduction of the cross‐sectional area available for mass transport into the channels. Additionally, the formation of gas bubbles inside the electrode structure and clogging owing to biofilm detachment have been observed as possible limitations that may arise because of a suboptimal combination of electrode structure and operating conditions. These observations, indicating the influence of external factors on growth behavior and morphology on EAB biofilms, will be of great interest for future research because the effective electrode geometry of the bioanode can differ significantly from the geometry of the bare electrode material. More thorough research will be needed to understand the behavior of EAB biofilms under different operating conditions. The results also show that for low‐shear conditions, structure sizes of 400 μm or below are too fine to be available for long‐term bioelectrochemical activity and should be avoided. This study further underlines the importance of performing long‐term experiments to evaluate the suitability of electrode designs for the development of biofilm electrodes.

Conflict of interest

The authors declare no conflict of interest.

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer reviewed and may be re‐organized for online delivery, but are not copy‐edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

Supplementary

Click here for additional data file.