Abstract
A central challenge in achieving sustainable urban transformation is reducing the environmental impact of existing buildings, many of which were not designed to meet today’s sustainability standards. While integrating green infrastructure into new construction is increasingly feasible, retrofitting older structures presents significant technical and spatial constraints. We address this challenge by developing a proof-of-concept for integrating modular flat-panel photobioreactors into existing urban buildings. Photobioreactors cultivate photoautotrophic microorganisms that fix atmospheric through photosynthesis, producing oxygen and biomass. We constructed a window-integrated prototype ( cm) and installed it directly into a standard window frame at the Biology Building, RWTH Aachen University, Germany, without major structural modifications. Window integration offers distinct advantages over facade-mounted systems for retrofit applications which include elimination of structural reinforcements and reduced visual impact on building envelope, as existing structural openings can be used, and stable operating temperatures due to the integration with climate-controlled building interiors. Two sequential cultivation trials using Synechocystis sp. PCC 6803 (5-day window-installed and 20-day stand-alone tests) demonstrated biological activity while revealing critical technical barriers that must be addressed for continuous operation. This proof-of-concept successfully demonstrates the feasibility of window-integrated PBRs for building retrofit applications and establishes a research pathway for system optimization toward practical deployment.
1 Introduction
As global urbanization accelerates, cities have become central actors in the climate crisis. Although they occupy only 3% of Earth’s land, urban centers are responsible for approximately 75% of carbon emissions and 60%–80% of global energy consumption (). With urban populations projected to increase from 55% to 68% by 2050 (), the mitigation of the Urban Heat Island effect (UHI) is a growing priority (; ; ; ) in urban planning. This disproportionate environmental impact, combined with poor air quality that affects nine out of ten urban residents (), requires urgent transitions toward sustainable urban development, particularly through the retrofitting of existing built environments. Globally, buildings account for roughly one-third of final energy consumption and a similar share of energy-related emissions, with most of this demand arising from the existing building stock rather than new construction (; ). The built environment is therefore increasingly recognized as a priority domain for decarbonization interventions, with the European Union and national governments implementing mandatory energy performance standards for existing building stock (). Green infrastructure strategies have emerged as promising interventions, yet retrofitting older structures often requires extensive structural modifications that are economically unfeasible in many cases (). In this context, nature-based and bio-integrated solutions embedded within the building envelope have attracted growing research interest as complementary strategies to conventional energy retrofits (). Bio-integrated building systems, particularly photobioreactors (PBRs) cultivate photoautotrophic microorganisms that fix atmospheric , produce oxygen, and generate biomass through photosynthesis, offering a dual benefit of environmental remediation through carbon capture and resource generation through biomass production. Unlike passive green infrastructure strategies, such as planted facades, green roofs, or carbon-sequestering construction materials, PBRs enable active and quantitatively controllable biological fixation under defined conditions, with regulatory control of culture composition, nutrient cycling, and contamination management ().
The potential of building-integrated PBRs has progressed from theoretical modelling toward experimental validation (; ; ; ; ). Beyond carbon fixation, PBR-integrated envelopes reduce energy consumption through shading and thermal buffering (), improve air quality , and produce harvestable bioenergy (). demonstrate through experimentally validated thermal modelling that a microalgae biofacade almost entirely removes a building’s cooling demand, with reductions in PBR heating demand of 92%–100% depending on location, confirming that thermal integration benefits are quantifiable and climatically robust. Nevertheless, recent systematic reviews identify critical knowledge gaps with respect to long-term operational reliability, contamination management, and integration into existing structural openings, and note that most systems to date have been installed on new construction rather than retrofitted into existing building stock (; ).
Cyanobacteria such as Synechocystis sp. PCC 6803 are particularly suitable for architectural applications due to their rapid biomass production, their ability to grow under variable light conditions and their tolerance to environmental fluctuations typical of urban settings (; ). Several pioneering projects have demonstrated architectural integration of PBRs, most notably the BIQ Das Algenhaus in Hamburg, Germany, which incorporated algae cultivation panels into building facades (). However, these facade-mounted systems face significant adoption barriers: reviews highlight high construction and maintenance costs . They require custom mounting systems and structural reinforcement to support additional weight, demand complex maintenance protocols for outdoor exposure, add substantial visual bulk to building exteriors, and raise concerns about aesthetic integration in historic or architecturally sensitive contexts. These requirements have limited PBRs deployment primarily to high-profile demonstration projects rather than widespread retrofit applications ().
Windows represent an alternative integration opportunity that has received limited research attention in the scientific literature. Unlike solid facades requiring modification to host biological systems, windows are inherently transitional architectural elements: they mediate between interior and exterior environments, occupy standardized structural openings, and can be replaced during normal building lifecycle maintenance without major construction intervention. Modern multi-pane window assemblies already create interior cavities for thermal insulation, spaces that could potentially accommodate biological cultivation systems. This approach offers several potential advantages for retrofit applications: utilization of existing structural openings minimizes construction impact and cost, access to climate-controlled interior environments may reduce temperature fluctuation stress on cultures, interior viewing provides direct occupant connection to biological processes, standard window dimensions enable modular deployment across diverse building types, and integration during scheduled window replacement cycles may reduce incremental costs compared to facade installations.
We present a proof-of-concept for integrating modular PBRs directly into standard window frames as a retrofit strategy for existing buildings. A prototype ( cm) was constructed and installed into a window opening at the Biology Building, RWTH Aachen University, Germany. The specific aims of this study are: (1) to demonstrate basic feasibility of cultivating Synechocystis sp. PCC 6803 under window-specific environmental conditions; (2) to provide a transparent assessment of performance limits, failure modes, and integration challenges that must be addressed for future building-scale applications; and (3) to identify the specific technical obstacles that must be overcome for this approach to achieve practical viability in building retrofit applications. The system is not intended as a fully optimized or deployment-ready design; rather, it aims to provide a transparent assessment of performance limits, failure modes, and integration challenges that must be addressed for future building-scale applications.
2 Methods
2.1 Window-integrated photobioreactor design
2.1.1 Flat-panel photobioreactor construction
The prototype PBR (also referred to as reactor) was designed to fit within a window opening of 124 cm 85.2 cm, at the RWTH Aachen University Biology building (Figure 1). No structural modification to the window frame or surrounding masonry was required for prototype installation. The reactor consists of a custom acrylic frame creating a flat-panel cultivation chamber with 25 mm internal depth. Frame construction utilized four corner pieces and two spacer components, assembled using 6 mm stainless steel pins (Figures 2a,b) and acrylic-compatible adhesive (single-component, UV-resistant, transparent cure). The rear panel comprises 4 mm tempered glass providing structural rigidity and thermal insulation. The front panel ( cm) is polymethyl methacrylate (PMMA, 3 mm thickness) selected for optical clarity (>92% visible light transmission), impact resistance, and compatibility with aqueous cultivation systems. Eight precision-machined access ports (70 mm diameter, positioned 55 mm from top/bottom edges and 66/134 mm from side edges) accommodate glass plugs with compression seals, enabling aseptic medium addition and sampling without system disassembly (Figures 2d-f). The front panel is secured via stainless steel clamps with silicone gaskets ensuring watertight seal while allowing panel removal for maintenance. Total reactor volume is approximately 26 L (calculated from internal dimensions 115 cm 75 cm 2.5 cm (Figure 2c), accounting for frame volume displacement). Working volume was initially limited to 8–10 L (filling to mid-height of frame) to assess system integrity before full-scale operation.
FIGURE 1
FIGURE 2
2.1.2 Sensor integration and monitoring
Real-time cultivation monitoring was achieved through an integrated sensor suite connected to an external sampling loop (Figure 1A). A submersible pump (4W, flow rate 300 L/h) circulated culture through modified acrylic tubing housing a Hamilton Arc pH electrode (range 0–14, accuracy 0.02 pH units, response time <30 s) and Hamilton VisiFerm dissolved oxygen (DO) 120 sensor (range 0%–200% saturation, accuracy 2%, response time <90 s). Optical density (OD) was measured using a modified Chi.Bio reactor unit (measurement wavelength 650 nm) with flow-through configuration, enabling continuous OD monitoring without the Chi.Bio’s standard temperature control or agitation systems.
Culture circulation was facilitated by pump-driven flow from bottom to top of the reactor chamber, supplemented by fine-bubble aeration through two diffusers positioned at the lower corners (air flow rate 0.5–1 L/min, compressed air filtered through 0.22 m PTFE membrane). This circulation pattern ensured: (1) homogeneous nutrient and gas distribution, (2) prevention of cell sedimentation, (3) representative sampling for sensor measurements, and (4) effective light penetration through maintained culture suspension.
All sensors were connected to a data acquisition laptop (Figure 1A) running manufacturer-specific software (Hamilton ArcAir for electrochemical sensors, Chi.Bio web interface for OD). Data logging frequency was 2 min for OD, 5 min for pH and DO.
2.1.3 Instalation site
The installation window was located on the ground floor of the Biology Building (Sammelbau 1, RWTH Aachen University, Aachen, Germany; 50.78°N, 6.06°E). The window faces south-east, providing exposure to direct solar irradiance primarily during morning hours. The building’s HVAC system maintains interior ambient temperatures of 20 °C–24 °C year-round, providing passive thermal regulation of the culture. Aachen is located in a temperate oceanic climate zone (Köppen: Cfb), characterized by moderate solar irradiance and frequent overcast conditions, representative of Central European urban building stock and therefore relevant to the broader applicability of the proof-of-concept.
The Biology Building (Sammelbau 1, RWTH Aachen University, Germany) was selected as the installation site based on three criteria: institutional accessibility enabling safe installation and rapid intervention in case of system failure; window dimensions ( cm) representative of standard European institutional and residential window formats, lending the proof-of-concept broader applicability; and the availability of a climate-controlled interior environment providing passive thermal regulation without additional infrastructure.
2.2 Organism selection and preliminary screening
Four cyanobacterial strains were evaluated for suitability in window-integrated PBR applications: Synechocystis sp. PCC 6803 (DSM 101314), a unicellular freshwater cyanobacterium, characterized by oxygenic photosynthesis and mixotrophic growth capabilities (); Nostoc punctiforme (DSM 101384), a filamentous nitrogen-fixing species capable of heterocyst formation (); Cyanothece sp. ATCC 51142, a unicellular diazotrophic cyanobacterium with temporal separation of photosynthesis and nitrogen fixation (); and Arthrospira platensis (Institut Pasteur), a filamentous alkaliphilic species known for high biomass productivity ().
Preliminary screening was conducted using controlled mini-bioreactor systems (Chi.Bio platform, described below) to assess growth performance under standardized conditions. Synechocystis sp. PCC 6803 was selected for prototype testing based on three criteria: (1) unicellular morphology ensuring uniform suspension and preventing sedimentation issues observed with filamentous strains, (2) superior growth rate compared to other tested species under fluctuating light conditions typical of window environments, and (3) extensive characterization in the literature providing baseline data for comparison ().
2.3 Cultivation media
All culture were maintained in blue-green Medium (BG-11), a standard inorganic-defined medium for cyanobacterial cultivation (). The medium composition includes macronutrients such as nitrate, phosphate, and magnesium as well as trace elements such as iron, manganese, and zinc. Medium pH was maintained at . All media were prepared with ultrapure water and filter-steralized or autoclaved prior to use. During the course of the study, the composition of the BG-11 medium was modified, adjusted, or supplemented to achieve better growth for specific strains such as Synechocystis sp. PCC 6803 and Arthrospira platensis ().
2.4 Preliminary growth parameter optimization
Growth optimization experiments were performed using the Chi.Bio mini-bioreactor platform (), a modular system enabling real-time optical density monitoring and automated culture dilution. Each Chi.Bio reactor (20 mL working volume) was equipped with a 650 nm laser for optical density measurement at 2-min intervals, a 6500K white LED array for illumination (adjustable between light intensities: L0.1 is about 325 mol photons , L0.2 is 560 mol photons , and L0.3 is about 850 photons m-2s-1), 280 nm UV LED, and automated medium exchange capability via peristaltic pump. Temperature was controlled via external heating block.
Within the experimental design we tested temperature (28 °C, 30 °C, 32 °C, 34 °C, 36 °C) and light intensity (325, 560, 850 mol photons , designated L0.1, L0.2, L0.3 respectively) effects on Synechocystis sp. PCC 6803 growth. Reactors were inoculated to initial = 0.05–0.1 and maintained under 16:8 h light:dark cycles. The system’s turbidostat function maintained cultures at = 1.5 through automated dilution with fresh medium, enabling calculation of specific growth rates (, ) from dilution frequency. Each condition was tested in duplicate with experimental duration of 5–7 days. Negative controls (sterile medium, no inoculation) and positive controls (fixed optimal conditions: 30 °C, L0.1) were included in each run. Growth rates were determined from exponential phase OD measurements using semi-log plots.
Statistical analysis was performed using two-way ANOVA with Tukey’s multiple comparisons test (GraphPad Prism 9). Significance levels were set at p < 0.05 (*), p < 0.01 (**), p < 0.001 (***), and p < 0.0001 (****).
2.4.1 Sterilization protocol
Due to the large size and acrylic construction of the prototype, conventional sterilization methods (autoclaving, UV irradiation) were impractical. Chemical disinfection was performed using 3% (v/v) hydrogen peroxide as follows: (1) reactor surfaces were manually cleaned with lint-free wipes; (2) the assembled system was filled with 7 L of 3% and circulated for 15 min with aeration to ensure contact with all interior surfaces; (3) was drained and the system was flushed twice with 7 L sterile ultrapure water (15 min circulation per flush); (4) final filling with sterile BG-11 medium; (5) sensor calibration and baseline measurements prior to inoculation. Ethanol (70% v/v) was avoided due to potential for acrylic stress-cracking and swelling with repeated exposure.
2.4.2 Prototype operation
Two sequential cultivation experiments were conducted with the same prototype. In Experiment 1, the prototype was installed directly within the window frame of the Biology Building at RWTH Aachen University (Figure 3), receiving natural daylight without supplemental LED illumination. Inoculation was performed with 100 mL of actively growing Synechocystis sp. PCC 6803 preculture (, grown in 250 mL Erlenmeyer flasks at 30 °C, 80 rpm orbital shaking, 50–60 mol photons ) continuous illumination for 7 days). This experiment ran for approximately 120 h before structural instability (cracking of acrylic frame components and leaking at adhesive joints) necessitated removal from the window frame.
FIGURE 3
In Experiment 2, the prototype was removed from the window frame, placed on a stand adjacent to the same window, and supplemented with an LED panel (white light, 50mol photons ) average irradiance measured at reactor surface) positioned 30 cm from the front panel. Due to leakage during Experiment 1, working volume was limited to approximately one-third of total reactor capacity. Ambient room temperature (20 °C–24 °C) provided thermal regulation through the building’s HVAC system. Initial inoculation was performed with 100 mL preculture . Due to contamination and low cell density (described in Section 3.3), a second inoculation was performed at 300 h with 200 mL of fresh preculture . Experiment 2 ran in batch mode without medium exchange for a total of 474 h (20 days) with continuous monitoring of OD, pH, and dissolved oxygen.
3 Results
3.1 Preliminary strain screening
Four cyanobacterial strains were evaluated under standardized conditions (30 °C, 325 mol photons , 16:8 h light:dark cycles) in Chi.Bio mini-bioreactors to assess growth performance (Supplementary Figures S1A–C). Synechocystis sp. PCC 6803 exhibited a lag phase of approximately 30 h followed by consistent linear growth during light periods, reaching = 1.5 after 187 h (Supplementary Figure S1A). Nostoc punctiforme showed an extended lag phase of approximately 90 h but achieved comparable final cell density to Synechocystis after 187 h (Supplementary Figure S1B. Cyanothece sp. ATCC 51142 demonstrated the longest lag phase (>100 h) and reached only low cell densities by the end of the cultivation period (Supplementary Figure S1C). Arthrospira platensis failed to grow under BG-11 medium conditions and was excluded from further analysis.
Based on these results, Synechocystis sp. PCC 6803 was selected for prototype testing due to: (1) shortest lag phase and most consistent growth kinetics, (2) unicellular morphology preventing sedimentation observed with filamentous strains (Nostoc punctiforme aggregated significantly under low agitation), and (3) superior performance under the fluctuating light conditions anticipated in window-integrated applications.
3.2 Growth parameter optimization in controlled bioreactors
Temperature and light intensity effects on Synechocystis sp. PCC 6803 growth were systematically evaluated across nine conditions (Supplementary Figure S2; Figure 4). Temperature demonstrated stronger influence on specific growth rate than light intensity across all tested conditions. At low light intensity (325 mol photons ), optimal growth occurred at 32 °C ( = 0.0506 0.012 ), significantly higher than all other temperatures at the same light level ( 0.001, two-way ANOVA with Tukey’s test). At moderate light intensity (560 mol photons ), growth rates were comparable at 30 °C ( = 0.0423 0.011 ) and 32 °C ( = 0.0385 0.009 ) with no significant difference ( 0.05). At high light intensity (850 mol photons ), 32 °C again produced optimal growth ( = 0.0473 0.014 ).
FIGURE 4
Growth rates declined at temperatures both above and below the 30 °C–32 °C optimum, demonstrating a clear thermal optimum window. Light intensity showed secondary influence: within the optimal temperature range (30 °C–32 °C), increasing light from 325 to 850 mol photons produced no significant growth rate improvement, suggesting saturation of photosynthetic capacity under these conditions.
Representative growth curves showed characteristic turbidostat behavior with linear OD increase during light periods followed by automated dilution at = 1.5 (Supplementary Figure S2). Internal reactor temperature tracked LED state, confirming consistent light:dark cycling. Growth rates were calculated from exponential phase slopes across multiple dilution cycles per condition ( = 10–54 individual measurements per condition depending on dilution frequency).
3.3 Prototype cultivation experiments
3.3.1 Experiment 1: Window-installed prototype (120 h)
The window-integrated photobioreactor consisted of a flat-panel reactor embedded into a standard window frame ( cm), constructed from transparent acrylic (PMMA) plates and sealed using adhesive joints. The system was operated as a closed-loop cultivation unit with external pumping, gas exchange, and sensor integration (Chi.Bio platform) to enable monitoring of optical density, dissolved oxygen, and pH under real environmental conditions. The Chi.Bio system enabled automated monitoring; however, several system crashes resulted in partial data loss. The flat-panel geometry ensured light penetration from both sides, while the window installation exposed the culture to natural diurnal fluctuations in light intensity and temperature. Following inoculation (initial 0.1), the window-installed prototype exhibited brief exponential growth between 15 and 20 h, reaching 0.4 before transitioning to stationary phase (Supplementary Figure S3A). Growth rate during this exponential phase ( 0.03 , estimated from OD slope) was substantially lower than optimal rates observed in Chi.Bio experiments ( = 0.05 ), suggesting suboptimal conditions. Temperature peaks in the medium correlated with daily solar radiation maxima, confirming that the window-installed configuration exposed the culture to direct solar heating (reaching up to 35 °C). Dissolved oxygen concentration dropped sharply from 84% to 62% saturation during the initial exponential phase (0–30 h), indicating active cellular respiration (Figure 5). As growth plateaued, DO recovered to 75%–80% saturation. Subsequent periodic DO fluctuations correlated with intermittent pump operation affecting sensor readings rather than biological activity. pH remained stable at 8.4 0.1 throughout the initial phase. Microscopic examination after 100 h revealed contamination: rod-shaped bacterial cells were observed alongside spherical Synechocystis cells, explaining the early growth plateau and low final OD. During extended operation, mechanical limitations of the prototype became apparent. Cracks developed in the acrylic frame, and leakage occurred at adhesive sealing interfaces, leading to gradual medium loss. These effects ultimately limited the operational stability of the system under window-installed conditions. Experiment 1 was therefore terminated at approximately 120 h, as continued operation in the current configuration was no longer feasible.
FIGURE 5
3.3.2 Experiment 2: Stand-alone prototype adjacent to window (474 h)
Following the structural failure of Experiment 1, the prototype was removed from the window frame and repositioned on a stand directly adjacent to the same window, with supplemental LED illumination (50 mol photons ). Working volume was limited to approximately one-third of total capacity due to residual leakage. Following initial inoculation, OD increased briefly before plateauing, with contamination again evident. A second inoculation was performed at 300 h with 200 mL preculture (2 initial inoculum volume). OD increased immediately following reinoculation (Supplementary Figure S3B), followed by a 50-h lag phase, then sustained linear growth through 474 h. Final culture showed characteristic green coloration indicating Synechocystis dominance. Growth rate post-reinoculation ( 0.02 estimated from linear phase) remained below Chi.Bio optimization values, consistent with suboptimal temperature and lower light availability. Following reinoculation, pH increased to 8.9 0.2 and exhibited greater fluctuation (Supplementary Figure S3D), consistent with increased photosynthetic consumption. DO patterns remained stable with pump-related fluctuations continuing throughout (Supplementary Figure S3C). Carbon sequestration was estimated indirectly from biomass formation using a literature-based conversion factor for Synechocystis of approximately 0.15–0.16 g dry cell weight (DCW) per unit optical density. An increase from one to two in the PBR filled with 8 L therefore corresponds to the formation of approximately 1.2 g biomass. Assuming a biomass carbon content of 5̃0% (w/w), this corresponds to the fixation of approximately 2.2–2.4 g . Given the observed growth rate ( 0.02 ), corresponding to a doubling time of 3̃5 h, this translates into an average CO fixation rate of approximately 1.5–1.6 g per day for the entire reactor, or 0̃.19–0.20 g .
4 Discussion
4.1 Proof-of-concept demonstration: Successes and limitations
The present study moves window-integrated PBRs from conceptual design toward empirical validation. Sustained cultivation of Synechocystis sp. PCC 6803 for 474 h under real building conditions confirms that cyanobacterial growth is compatible with the thermal and optical boundary conditions of interior-facing window systems. At the same time, the experiments clearly delineate the technical constraints that currently limit performance. Three factors: suboptimal growth conditions, microbial contamination, and system robustness; define the critical development axes for translation beyond prototype scale.
4.1.1 Growth performance: Gap between controlled and field conditions
Both prototype experiments achieved growth rates substantially below optimized Chi.Bio conditions ( at 32 °C). Experiment 1 (window-installed) additionally experienced uncontrolled temperature peaks up to 35 °C due to direct solar gain, which likely contributed to growth limitation despite the advantage of natural daylight. Experiment 2 (stand-alone) operated under more stable temperature conditions (20 °C–24 °C from building HVAC) but at very low supplemental light (50 molphotons ), well below the photosynthetic saturation threshold identified in Chi.Bio optimization. The performance gap across both experiments likely reflects: suboptimal temerature and light intensity versus optimized conditions achieved in Chi.Bio, reduced mixing efficiency in the large flat-panel geometry, and contamination effects.
The low light intensity is particularly significant: our optimization experiments demonstrated that growth saturates above 325 mol photons , but the prototype operated well below this threshold during cloudy periods. Future window-integrated designs must either: (a) select for low-light-adapted species, (b) incorporate supplemental lighting sufficient to reach photosynthetic saturation, or (c) accept lower productivity as a trade-off for architectural integration. The optimal solution likely involves species selection, as high supplemental lighting would undermine energy efficiency benefits.
Temperature control represents a more tractable challenge. The prototype operated at 20 °C–24 °C, which our optimization data shows is 30%–40% below optimal growth rate potential. However, utilizing building interior temperatures (typically 20 °C–22 °C) is a key advantage of window integration compared to outdoor facade systems experiencing greater thermal fluctuation. Rather than heating the culture (energy-intensive), future work should focus on selecting or adapting strains with shifted thermal optima matching building interior conditions. Carbon sequestration in the present study was estimated indirectly based on biomass formation and therefore represents a first-order approximation of system performance. While this approach enables an initial assessment under realistic operating conditions, a rigorous quantification of net carbon sequestration will require direct gas-phase CO measurements, comprehensive carbon mass balances, and evaluation under defined boundary conditions in future studies.
4.1.2 Contamination: The primary technical barrier
Bacterial contamination emerged as a major obstacle to sustained cultivation. The applied sterilization protocol (3% HO, 15 min circulation) reduced initial microbial load but proved insufficient to ensure axenic conditions in the large-scale acrylic reactor system. This limitation reflects both the increased surface area and structural complexity of the prototype, which likely hinder complete sterilant contact and facilitate microbial persistence in dead volumes and connection interfaces. Microscopic examination confirmed the presence of heterotrophic bacterial contaminants, which temporarily dominated biomass formation during the initial phase (0–100 h) before nutrient limitation curtailed their proliferation. Subsequent growth of Synechocystis was only achieved following high-density reinoculation (200 mL at OD650 = 1.5–1.8), indicating competitive interactions and suggesting that culture establishment requires a sufficiently high initial biomass to overcome contaminant pressure. These observations highlight that contamination not only affects culture purity but also fundamentally alters apparent growth dynamics and system performance. The precise contamination source could not be unambiguously identified but likely includes: (i) incomplete sterilization of internal reactor surfaces, (ii) introduction via gas exchange despite 0.22 m filtration, and/or (iii) contamination of the inoculum or auxiliary components (e.g., tubing, pumps). In addition, pump-driven flow and intermittent pressure fluctuations may have facilitated the transport of microorganisms across partially sealed interfaces. Future system designs should therefore implement more robust sterilization and contamination control strategies, including: (i) extended or repeated HO exposure (>30 min), (ii) integration of UV treatment for accessible surfaces, (iii) sterile filtration of all gas and liquid streams, and (iv) minimization of dead volumes and non-sterile junctions through improved reactor and tubing design. The use of selective antibiotics during initial cultivation phases may further support culture establishment, although this approach conflicts with the long-term goal of chemical-free operation and was therefore not pursued in this study. Alternatively, the results suggest that maintaining strictly axenic conditions in large-scale, building-integrated systems may be inherently challenging. In this context, a shift toward controlled mixed microbial systems may represent a more realistic operational paradigm. If contamination cannot be fully prevented, future research should focus on: (i) selecting cyanobacterial strains with strong competitive fitness, (ii) operating under conditions favoring phototrophs (e.g., elevated pH > 9), and (iii) designing stable microbial consortia optimized for robustness and functional performance rather than monoculture purity.
4.1.3 System reliability and monitoring
Sensor integration functioned adequately, but Chi.Bio monitoring system instability (two crashes in 20 days) indicates insufficient robustness for long-term architectural applications. Buildings require maintenance-free operation over months to years, not days. These limitations highlight the need for redundant monitoring architectures, including parallel sensor systems, data buffering, and improved correction algorithms for pump-induced signal fluctuations. Future systems must incorporate: (1) redundant monitoring with automatic fail-over, (2) remote alert systems for contamination or growth failure, and (3) automated medium exchange to enable continuous operation.
The external sampling loop design successfully enabled real-time pH, DO, and OD monitoring without comngng reactor sterility—a key achievement for future scaling. However, pump-induced DO fluctuations suggest circulation requires optimization to maintain stable sensor readings. Despite these limitations, the available data were sufficient to capture the main system dynamics.
4.2 Window integration versus facade mounting: comparative assessment
Our window-integrated approach differs fundamentally from facade-mounted PBR systems exemplified by the BIQ House project in Hamburg (). The BIQ House integrated 129 algae-filled glass panels (2.5 m 0.7 m each) into the building’s southwest facade, achieving biomass productivity and demonstrating architectural integration feasibility. However, BIQ House faced significant challenges: aluminum frame corrosion from high pH culture (necessitating replacement with glass fiber-reinforced plastic), glass panel fracture during maintenance, and substantial technical complexity requiring dedicated building systems infrastructure (). Compared to conventional façade-mounted PBR systems, the window-integrated design offers a higher level of architectural integration but currently operates at lower areal productivity due to geometric and light distribution constraints. Window integration offers multiple potential advantages: (1) utilization of existing structural openings eliminates need for facade reinforcement, (2) access to climate-controlled interior environments reduces thermal stress, (3) retrofit installation during normal window replacement cycles minimizes disruption and cost, and (4) interior visibility may enhance occupant connection to biological processes (though systematic assessment of aesthetic response is needed). Conversely, window integration accepts lower light availability and constrains reactor geometry to standard window dimensions. In contrast to alternative building-integrated carbon mitigation strategies such as green facades or carbon-sequestering materials, the presented system enables active and controllable biological carbon fixation, albeit at the cost of increased system complexity.
BIQ House also provided valuable lessons on occupant acceptance: resident surveys revealed mixed responses to algae facades, with some reporting noise disturbance from pumps while others appreciated the innovation or found the sound calming over time (). This highlights the necessity of incorporating human factors research into bio-integrated architecture. Our window prototype operated quietly due to low pump flow rates, but systematic occupant acceptance assessment was beyond this study’s scope. The impact of the window-integrated PBR on building performance parameters such as daylight transmission, thermal insulation, and acoustic behavior was not quantitatively assessed in this study. However, based on the optical density of the culture and reactor geometry, significant effects on light transmission and thermal behavior can be expected. Future work should therefore include systematic measurements of building performance indicators to evaluate trade-offs between biological functionality and indoor environmental quality. The key insight from comparing approaches is that facade and window integration likely serve complementary roles: facades offer high light availability and maximum surface area for high-productivity applications, while window integration suits retrofit contexts where structural modification is constrained and moderate productivity is acceptable. A detailed techno-economic and life-cycle comparison is beyond the scope of this proof-of-concept study but represents a critical next step for evaluating the practical relevance of such systems. The impact of the window-integrated PBR on building performance parameters such as daylight transmission, thermal insulation, and acoustic behavior was not quantitatively assessed in this study. However, based on the optical density of the culture and reactor geometry, significant effects on light transmission and thermal behavior can be expected. Future work should therefore include systematic measurements of building performance indicators to evaluate trade-offs between biological functionality and indoor environmental quality.
4.3 Future research priorities
Future work should move the system from a proof-of-concept toward a robust, deployable technology by addressing biological, engineering, and architectural constraints in parallel. Contamination control remains a primary bottleneck and requires a systematic comparison of physicochemical sterilization strategies, including extended exposure, UV treatment, and operation at elevated pH, alongside a strategic shift from strict monocultures toward controlled consortia where appropriate. In parallel, species and strain selection must be aligned with the environmental boundary conditions imposed by building integration, prioritizing cyanobacteria with lower thermal optima, improved low-light performance, and tolerance to alkaline conditions, while explicitly evaluating morphological trade-offs in flat-panel reactors. From an engineering perspective, improved hydrodynamic design informed by computational fluid dynamics is required to eliminate dead zones, optimize gas–liquid exchange, and enable automated, continuous operation with fail-safe monitoring and remote alerts. Quantitative performance assessment is equally critical: direct measurements of capture rates, biomass productivity, and thermal insulation effects must replace qualitative observations, and these data should feed into a full life-cycle assessment to evaluate environmental and economic viability. Finally, successful architectural integration will depend not only on structural and optical optimization of frames and glazing, but also on modular system design and systematic evaluation of occupant acceptance, ensuring scalability across diverse building types without compromising daylight access or aesthetics.
5 Conclusion
This study demonstrates that flat-panel photobioreactors can be integrated into standard window frames of an existing building and operated under ambient conditions without major structural modification, confirming the basic biological and architectural feasibility of this retrofit concept. Sustained cultivation of Synechocystis sp. PCC 6803 for 474 h under real building conditions establishes this feasibility empirically, while the performance gap relative to controlled conditions, recurrent bacterial contamination, and prototype structural failure define the three technical development axes that must be addressed before window-integrated PBRs can progress beyond proof-of-concept scale. Future work should prioritize robust contamination control strategies, selection of strains with thermal optima and low-light tolerance matched to building interior conditions, and direct quantification of capture rates and thermal performance under defined boundary conditions.
Statements
Data availability statement
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.
Author contributions
FR: Methodology, Validation, Visualization, Writing – original draft, Writing – review and editing, Data curation, Formal Analysis, Investigation. TN: Visualization, Writing – original draft, Writing – review and editing. LF: Conceptualization, Funding acquisition, Methodology, Project administration, Resources, Supervision, Writing – review and editing. LL: Conceptualization, Funding acquisition, Methodology, Project administration, Resources, Supervision, Writing – original draft, Writing – review and editing. ABM: Conceptualization, Funding acquisition, Methodology, Project administration, Resources, Supervision, Validation, Visualization, Writing – original draft, Writing – review and editing.
Funding
The author(s) declared that financial support was received for this work and/or its publication. This work was conducted as part of the LIVEWIN (Living Window) project at RWTH Aachen University, Germany, funded from the Sustainability Fund.
Acknowledgments
We thank Prof. Dr. Lars Blank for initial fruitful discussions on green algae facades and the integration of cyanobacterial systems into flat-bed reactors.
Conflict of interest
The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Generative AI statement
The author(s) declared that generative AI was used in the creation of this manuscript. Generative AI tools (specifically Claude by Anthropic and ChatGPT by OpenAI]) were used during the preparation of this manuscript for grammar and language editing, paraphrasing for clarity, and to assist in assembling introductory background literature on retrofitting. All AI-generated suggestions were reviewed, curated, and verified by the authors. The authors take full responsibility for the accuracy, integrity, and originality of the content presented. AI was not used in data collection, analysis, interpretation, or the formulation of conclusions.
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
Supplementary material
The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fbuil.2026.1818560/full#supplementary-material
References
1
AdamczykM.LasekJ.SkawińskaA. (2016). CO2 biofixation and growth kinetics of Chlorella vulgaris and Nannochloropsis gaditana. Appl. Biochem. Biotechnol.179, 1248–1261. 10.1007/s12010-016-2062-3
2
AroraR.SudhakarK.RanaR. S. (2024). Photobioreactors for building integration: a overview of designs and architectural potential. Heliyon10, e35168. 10.1016/j.heliyon.2024.e35168
3
ChewK. W.KhooK. S.FooH. T.ChiaS. R.WalvekarR.LimS. S. (2021). Algae utilization and its role in the development of green cities. Chemosphere268, 129322. 10.1016/j.chemosphere.2020.129322
4
DahiyaD.LaishramB. (2024). Life cycle energy analysis of buildings: a systematic review. Build. Environ.252, 111160. 10.1016/j.buildenv.2024.111160
5
DixonT.EamesM. (2013). Scaling up: the challenges of urban retrofit. Build. Res. & Inf.41, 499–503. 10.1080/09613218.2013.812432
6
ElrayiesG. M. (2018). Microalgae: prospects for greener future buildings. Renew. Sustain. Energy Rev.81, 1175–1191. 10.1016/j.rser.2017.08.032
7
GentschevaG.NikolovaK.PanayotovaV.PeychevaK.MakedonskiL.SlavovP.et al (2023). Application of Arthrospira platensis for medicinal purposes and the food industry: a review of the literature. Life13, 845. 10.3390/life13030845
8
GirardF.ToublancC.AndresY.DechandolE.PruvostJ. (2023). System modeling of the thermal behavior of a building equipped with facade-integrated photobioreactors: validation and comparative analysis. Energy Build.292, 113147. 10.1016/j.enbuild.2023.113147
9
InamA.OncelS. S. (2022). Photobioreactors as potential tools for environmentally friendly and sustainable buildings. Int. J. Environ. Sci. Technol.19, 2353–2364. 10.1007/s13762-021-03281-7
10
Institute for Housing and Urban Development Studies (2024). SDG 11 – sustainable cities and communities. Available online at: https://www.ihs.nl/en/about/ihs-and-sustainable-development-goals (Accessed January 21, 2026).
11
KernerM.GebkenT.SundarraoI.HindersinS.SaussD. (2019). Development of a control system to cover the demand for heat in a building with algae production in a bioenergy façade. Energy Build.184, 65–71. 10.1016/j.enbuild.2018.11.030
12
KimK. H.ParrowM. W.Kheirkhah SangdehP. (2025). Microalgae-integrated building enclosures: a nature-based solution for carbon sequestration. Front. Built Environ.11, 11–2025. 10.3389/fbuil.2025.1574582
13
McCarthyM. P.BestM. J.BettsR. A. (2010). Climate change in cities due to global warming and urban effects. Geophys. Res. Lett.37, L09705. 10.1029/2010gl042845
14
MeeksJ. C.ElhaiJ.ThielT.PottsM.LarimerF.LamerdinJ.et al (2001). An overview of the genome of Nostoc punctiforme, a multicellular, symbiotic cyanobacterium. Photosynth. Res.70, 85–106. 10.1023/a:1013840025518
15
PengS.PiaoS.CiaisP.FriedlingsteinP.OttleC.BréonF.-M.et al (2011). Surface urban heat island across 419 global big cities. Environ. Sci. & Technol.46, 696–703. 10.1021/es2030438
16
PruvostJ.Le GouicB.LepineO.LegrandJ.Le BorgneF. (2016). Microalgae culture in building-integrated photobioreactors: biomass production modelling and energetic analysis. Chem. Eng. J.284, 850–861. 10.1016/j.cej.2015.08.118
17
RippkaR.StanierR. Y.DeruellesJ.HerdmanM.WaterburyJ. B. (1979). Generic assignments, strain histories and properties of pure cultures of Cyanobacteria. Microbiology111, 1—–61. 10.1099/00221287-111-1-1
18
SahaR.VerseputA. T.BerlaB. M.MuellerT. J.PakrasiH. B.MaranasC. D. (2012). Reconstruction and comparison of the metabolic potential of Cyanobacteria Cyanothece sp. atcc 51142 and Synechocystis sp. pcc 6803. PLoS ONE7, e48285. 10.1371/journal.pone.0048285
19
ShahamatH. (2025). Advancing bio-integrated building systems: a systematic review of architectural, structural and mechanical–electrical challenges in photobioreactor façades. Smart Sustain. Built Environ., 1–32doi. 10.1108/SASBE-03-2025-0156
20
SteelH.HabgoodR.KellyC. L.PapachristodoulouA. (2020). In situ characterisation and manipulation of biological systems with chi.bio. PLOS Biol.18, e3000794. 10.1371/journal.pbio.3000794
21
TalaeiM.PrietoA. (2024). A review on performance of sustainable microalgae photobioreactor façades technology: exploring challenges and advantages. Archit. Sci. Rev.67, 387–414. 10.1080/00038628.2024.2305889
22
United Nations, Department of Economic and Social Affairs, Population Division (2025). World urbanization prospects 2025: summary of results. Available online at: https://desapublications.un.org/publications/world-urbanization-prospects-2025-summary-results (Accessed January 21, 2026).
23
VajdiS.AslaniA. (2023). Design and techno-economic analysis of direct co2 capturing with integrated photobioreactors as a building façade. Sustain. Energy Technol. Assessments56, 103068. 10.1016/j.seta.2023.103068
24
VujovicS.HaddadB.KarakyH.SebaibiN.BoutouilM. (2021). Urban heat island: causes, consequences, and mitigation measures with emphasis on reflective and permeable pavements. CivilEng2, 459–484. 10.3390/civileng2020026
25
WardK.LaufS.KleinschmitB.EndlicherW. (2016). Heat waves and urban heat islands in Europe: a review of relevant drivers. Sci. Total Environ.569 (570), 527–539. 10.1016/j.scitotenv.2016.06.119
26
WelshE. A.LibertonM.StöckelJ.LohT.ElvitigalaT.WangC.et al (2008). The genome of cyanothece 51142, a unicellular diazotrophic cyanobacterium important in the marine nitrogen cycle. Proc. Natl. Acad. Sci.105, 15094–15099. 10.1073/pnas.0805418105
27
[Dataset] World Health Organization (2024). Ambient (outdoor) air pollution: fact sheet. Available online at: https://www.who.int/news-room/fact-sheets/detail/ambient-(outdoor)-air-quality-and-health (Accessed January 21, 2026).
28
WurmJ.PauliM.KernerM.HindersinS.KrümmelS.BrecknerI.et al (2015). Monitoring Fassadenkonstruktion aus Photobioreaktoren am Pilotprojekt BIQ auf der IBA 2013 in Hamburg: Abschlussbericht (Kurzfassung). Tech. rep., Bundesinstitut für Bau-, Stadt- Raumforsch. (BBR), Bonn. Available online at: https://www.irbnet.de/daten/kbf/kbf_d_F_2996.pdf.
29
YamanY.İremD.KöktürkG.EzanM. A.TokuçA. (2025). Advancing building energy simulations: integrating photobioreactor façades for enhanced energy efficiency and comfort. Energy Build.342, 115860. 10.1016/j.enbuild.2025.115860
30
Zarco-SotoF. J.Zarco-SotoI. M.Sharif AliS. S.Zarco-PeriñánP. J. (2025). Energy consumption in buildings: a compilation of current studies. Energy Rep.13, 1293–1307. 10.1016/j.egyr.2024.12.069
31
ZavrelT.OcenasovaP.CervenyJ. (2017). Phenotypic characterization of synechocystis sp. pcc 6803 substrains reveals differences in sensitivity to abiotic stress. PLOS ONE12, e0189130. 10.1371/journal.pone.0189130
Summary
Keywords
bio-integrated architecture, building retrofit, carbon sequestration, cyanobacteria, flat-panel photobioreactor, sustainable building systems, Synechocystis, urban sustainability
Citation
Radzio F, Nies T, Fürtauer L, Lauterbach L and Matuszyńska AB (2026) Window-integrated photobioreactors: a proof-of-concept for green building retrofitting. Front. Built Environ. 12:1818560. doi: 10.3389/fbuil.2026.1818560
Received
26 February 2026
Revised
19 March 2026
Accepted
25 March 2026
Published
19 May 2026
Volume
12 - 2026
Edited by
Giovanni Ciampi, Università degli Studi della Campania Luigi Vanvitelli, Italy
Reviewed by
Vitta Ibrahim, Pyramids Higher Institute for Engineering and Technology, Egypt
Tao Luo, China Academy of Building Research, China
Updates
Copyright
© 2026 Radzio, Nies, Fürtauer, Lauterbach and Matuszyńska.
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: Lars Lauterbach, lars.lauterbach@iamb.rwth-aachen.de; Anna Barbara Matuszyńska, anna.matuszynska@cpbl.rwth-aachen.de
Disclaimer
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.