Filtration of Mineral and Biological Aerosols by Natural Plant Panels

Abstract

This study investigated the potential of Tillandsia plants, which can be arranged as a soil-free living green panel, and Banksia flower spikes, which could be arranged as a non-living natural panel, to filter particulate matter (PM) and airborne microorganisms. The Tillandsia panels demonstrated superior PM filtration, achieving up to 74% efficiency for large particles (>10 μm) at air velocities of 1.0 and 1.5 m/s without increasing pressure drop substantially. Conversely, Banksia performed better at 0.5 m/s, filtering up to 53% of PM compared to Tillandsia’s 13%. Notably, both panel types demonstrated significant fungal filtration, removing over 50% of airborne spores at 1.5 m/s. These findings suggest that incorporating plant-based panels into urban environments can enhance air quality and public health especially for allergenic particles and microorganisms.

1. Introduction

Green infrastructure (GI) solutions, such as green walls and living roofs, can not only enhance the aesthetic appeal of cities but also provide numerous ecological benefits, including reduced urban temperatures and consequent energy demand for cooling buildings, reduced stormwater runoff, and increased biodiversity [1,2]. Moreover, these green spaces can contribute to the mental and physical well-being of urban residents, thus improving community health and quality of life [3].

Airborne microorganisms, including bacteria and fungi, pose significant threats to public health by causing respiratory infections and allergies [4]. Urban environments, in particular, are susceptible to increased levels of these microorganisms due to factors such as pollution, overcrowding, and inadequate ventilation in indoor spaces [5]. While traditional air filtration systems are effective, they often rely on energy-intensive technologies and may not be suitable for all settings.

Phytoremediation, the use of plants to remove pollutants from the environment, has been studied extensively for its efficacy in mitigating various forms of pollution, including air pollution [6]. Plants, with their diverse morphological and physiological characteristics, have emerged as potential candidates for natural air filters [7,8,9,10,11]. The ability of plants to remove pollutants from indoor environments, especially volatile organic compounds (VOCs) and particulate matter (PM), has been studied, yielding promising results [12,13,14,15,16]. The emergence of the COVID-19 pandemic triggered interest in investigating the ability of plants to filter pathogens such as viruses and bacteria [17]. However, their potential to filter airborne microorganisms remains relatively unexplored.

This research investigates the use of two plant typologies, Banksia flower spikes and Tillandsia plants, in improving indoor and outdoor/indoor air quality when arranged as natural filtering panels. The complex structure and large surface area of Banksia flower spikes suggest their efficacy as physical traps for airborne microorganisms, a function potentially enhanced by the antimicrobial compounds produced by some Banksia species [18]. Tillandsia plants, being soil-free epiphytes, present a distinct advantage for urban and interior integration. Their ability to absorb atmospheric nutrients and water may also play a role in capturing and retaining airborne microorganisms. Furthermore, the successful use of Tillandsia panels as GI in Melbourne, Australia, where they reflected up to 90% of light and significantly cooled a building [19], highlights their broader environmental benefits.

Such natural, plant-based panels could be used as vertical greenery systems, more specifically green walls, where their complex geometries could be used to filter air pollutants. Different to green façades, that have plant roots on the ground (climbers) or roof (cascading plants), green walls are self-sufficient systems attached to the exterior or interior of a building [20]. The plant panels investigated in this study could be strategically installed along busy urban roads or street canyons, where they could act as biofilters, intercepting PM and gaseous pollutants generated by vehicular traffic [21,22]. Their organic aesthetic could also offer a softer, greener façade alternative or addition to conventional noise barriers [23] or outdoor plantation shutters to filter sunlight into houses. The panels could also be integrated into building ventilation systems as pre-filters, potentially reducing the load on mechanical filters and improving the quality of incoming air. Moreover, their lightweight and soilless nature makes them ideal for interior green walls in office spaces or residential buildings, where they could contribute to improved indoor air quality. Research has shown that interior green walls can significantly reduce levels of VOCs and PM in indoor environments [16]. The versatility of these plant-based panels, as highlighted by this range of potential applications that range from roadside barriers to interior design elements, highlights their potential to contribute to creating healthier and more sustainable urban spaces.

By employing a controlled laboratory experiment, this study aimed to quantify the filtration efficiency of the selected plant species for particles and airborne microorganisms, namely fungi, representing a wider range of allergy-causing airborne pathogens. Understanding the filtration mechanisms and potential of these plants can inform the design of green infrastructure solutions to improve air quality and public health in urban areas and enclosed spaces.

2. Materials and Methods

2.1. Laboratory Setup and Experiment Methodology

The experimental setup, shown in Figure 1, was designed to assess the efficacy of plants arranged as plant panels in removing various aerosols. Airflow, generated by a suction centrifugal fan (1) and measured by a flowmeter (2) (Model: HP-866B/APP, Zhuhai JiDa Huapu Instrument Co., Zhuhai, China), carried aerosols into the sealed chamber (4) containing a plant holder section (3) to act as a natural filter. The cross-sectional area of the plant panel was 126 cm² and the thickness of the plant holder was 5 cm. Mineral aerosols were produced through atomization of a 5% NaCl solution by a nebulizer (5), further explained in Section 2.3. Mineral aerosol concentration and particle size distribution were monitored every 2 s using a laser particle counter (7) (Model 4705, AeroNanoTech, Moscow, Russia) used in previous research projects [10,11,24]. Biological aerosols (fungi) were produced via agar surface vibration (6), also further explained in Section 2.3, and samples were collected by bioaerosol impactors (8) onto the surface of the agar plates for subsequent incubation and colony counting.

A minimum of eight experimental runs (replicates) were performed for each plant species with different plant individuals to assess repeatability. The amount of plant material selected for each run was comparable, aiming for the same biomass weight of 123 g with an allowed variability of 10% due to biomass weight fluctuations from moisture content natural variations. Each experimental run involved operating the system for 2 min at each of the three selected velocities (0.5 m/s, 1.0 m/s, and 1.5 m/s) for the mineral aerosol experiments, and at 1.5 m/s velocity for the biological aerosol runs. These velocities were selected to represent the lower end of usual wind speeds experienced at ground level inside street canyons that—roughly—range between 0.5 m/s and 3 m/s on a typical day [25]. Aerosols were monitored and sampled from both inlet and outlet sampling points, i.e., before and after passing through the plant panel. The filtration efficiency of each run was calculated by:

$$E(\%)=\left(1-{\displaystyle \frac{C_{out}}{C_{in}}}\right)\times 100\%$$

(1)

where C_in and C_out represent the fractional aerosol concentration at the inlet and outlet, respectively [26]. The filtration efficiency of the system was determined by averaging the results of at least eight replicate runs. To account for aerosol settling and deposition on the chamber walls, control measurements were performed with an empty chamber, revealing a deposition rate of less than 5% across all particles which was deemed acceptable and therefore not used for the correction of the results.

Pressure drop was measured at the beginning of each run and used to derive the quality factor (QF) of the system, which is calculated as follows:

QF = −ln(1 − E)/Δp

(2)

where E denotes the collection efficiency and Δp is the pressure drop across the filtering material [27]. The filtration performance of fibrous filters is commonly assessed using the QF, as a higher QF offers better performance by achieving higher collection efficiency with a lower pressure drop.

2.2. Plant Selection

The plant species were selected based on three criteria: (1) their ability to be arranged as natural panels for indoor and outdoor/indoor (such as window covers) environments; (2) the durability of the panel system, including plant adaptation in the urban and indoor settings with potentially harsh environment (i.e., sun exposure and possible lack of it, rain, wind); (3) plants of small elongated anatomy (leaves, flowers) for more predictable filtration efficiencies in accordance with our previous study [11].

Banksia spinulosa (Figure 2A), commonly known as Candlestick Banksia, is a species of the Proteaceae family, widely distributed along the eastern coast of Australia and typically growing as a multi-stemmed shrub to a height of 1.5 m [28]. It is characterised by its distinctive cylindrical flower spikes (inflorescences) with individual tiny flowers. When fresh or dry, the Banksia flower spikes can be easily removed from the plant and attached side by side onto a mesh/wire structure to form a durable, non-living natural panel. The Banksia flower spikes used in this experiment were sourced from the Griffith University campus and Toohey Forest in Nathan, Queensland, Australia, to where they are indigenous and abundantly found [29]. Dried Banksia flowers can last for up to six years when stored away from direct sunlight, water, and humidity in dried flower arrangements [30]. It is estimated that the non-living panels would last at least six months in an outdoor environment subjected to the weather, and up to twelve months in indoor conditions subjected to water flushing and forced airflow, after which time the flower spikes will need to be replaced. Banksias generally have a long flowering period—e.g., B. spinulosa flowers from autumn through the winter to spring [31]—allowing high availability of flower spikes for replacement. It is important to note that Banksia cones develop through three distinct stages: bud, flower spike, and seed capsule [31]. For use on the proposed panel, the cone must be harvested specifically during its flower spike stage. Unlike living materials, the non-living nature of the Banksia flower spike panel eliminates any need for consideration regarding irrigation or sun exposure, despite eventually requiring replacement.

Tillandsia spp. are epiphytes from the Bromeliaceae family native to the Americas that are commonly known as air plants. Although they use their roots to attach themselves to branches and other plant surfaces for structural support, they are not parasitic [32]. Their unique adaptation involves absorbing all necessary nutrients and water directly from the atmosphere, a process facilitated by specialised cells called trichomes [32]. The fact that these plants do not need soil or substrate to grow is advantageous for their use as vertical green infrastructure panels offering a resilient living, soil-independent green system that requires minimum maintenance [33]. In fact, in an outdoor environment with optimum sun exposure—indirect sunlight, dappled shade, or morning sun, depending on the species [34]—and abundant rainfall, the panels would require no maintenance. This was showcased on the Eureka Tower Experiment, where an installation of Tillandsias on top of Melbourne’s Eureka Tower (295 m) withstood for over a year with no maintenance or auxiliary watering system, exposed to extreme winds, heat, cold and extended periods of drought, with its performance continuing beyond the initial reporting period [35]. When placed indoors, Tillandsias need enough watering, sufficient light, to be allowed to dry between waterings, and they benefit from liquid fertilisation [36]. The species Tillandsia paleacea used in this study (Figure 2B) was sourced from a grower on the Sunshine Coast, Australia [37].

While the above-mentioned species were used for the experiments, it is hypothesised that other species within the same genus exhibiting analogous morphological traits, e.g., Banksia robur, Banksia aemula, Tillandsia chiletensis, and Tillandsia streptocarpa, would yield comparable results. Therefore, to streamline terminology, the plants used in this study are hereafter simply referred to as ‘Banksia flower spikes’ and ‘Tillandsias’.

Within the prototype chamber, Banksia flower spikes were cut and placed adjacent to each other, while Tillandsias were arranged to match their natural outdoor growth density after one year of establishment hanging on an outdoor wall (as depicted in Figure 2). These plant arrangements in the prototype simulated their intended configuration on a full-scale plant-based panel designed for urban application.

2.3. Particle Selection and Generation/Cultivation

As previously mentioned in Section 2.1, the laboratory setup was built to measure aerosols with both mineral and biological particles.

To create an aerosol comprised with mineral particles, a 5% NaCl solution was processed through a 3-jet Collison nebulizer (BGI Inc., Waltman, MA, USA). This nebulization was powered by 6 L per minute of HEPA-filtered compressed air. The addition of approximately 2% v/v glycerol to the nebulizer was crucial in maintaining hydration of the salt aerosol particles, thereby ensuring the final particle size distribution contained particles with a diameter up to and inclusive of 10 μm. More detailed information on the particle size distribution is presented in our previously published research that used the same methodology [10]. These mineral particles were selected to represent general PM in air, which is a complex mixture of components. PM₁₀ and PM_2.5 are particles with a diameter of 10 μm or less and 2.5 μm or less, respectively. PM₁₀ are inhalable into the lungs and associated with adverse health effects, while PM_2.5 can travel deep into the lungs and enter the bloodstream, posing significant health risks [38,39].

Aspergillus niger is a species of the fungal genus Aspergillus, which is recognised as the most prevalent airborne fungal spore that is linked to the development of invasive life-threatening infection and hypersensitivity respiratory illness in humans [40]. A. niger is found widely in the environment, including soil, decaying vegetation, and indoor air. While generally considered safe, it can cause health issues in susceptible individuals [41]. For this study, A. niger was sourced from Southern Biological, Australia. A small portion of fungus was transferred into a tube containing 9 mL of sterile water. The tube was shaken and sonicated to disperse the spores evenly throughout the liquid, and 150 μL of this suspension was pipetted onto the surface of an agar plate in a 90 mm diameter plastic Petri dish. The suspension was then spread across the nutrient surface—malt extract agar (MEA)—medium (CM 59 Oxoid Ltd., Basingstoke, UK) to ensure uniform growth, and placed into an incubator set to 25 °C. A bowl of water was also placed in the incubator to maintain a relative humidity of 55–65%. All procedures and experiments were conducted within a Class II biological safety cabinet (Model BH 2000, Clyde-Apac, Brisbane, Australia) to prevent release of aerosols into the laboratory environment.

Fungal aerosols were then made airborne using our previously developed vibrating fungal spore generator that is capable of aerosolizing fungal spores directly from cultures grown in standard 9 cm Petri dishes [42]. The setup involved sealing a 10 cm diameter, 2-watt, 4-ohm speaker to the bottom of a cylindrical casing with a soft rubber gasket. A Petri dish containing mature fungal spores was placed directly on the speaker’s surface and secured with thin, flexible rubber strips. The speaker’s size was carefully selected to precisely fit a standard 9 cm Petri dish, ensuring that the dish vibrated in sync with the speaker’s membrane at different frequencies during the experiment. A Hewlett-Packard 3312A Function Generator controlled the amplitude and frequency of the speaker’s vibrations to optimise particle concentration. A fresh Petri dish with fungal culture was used for each experiment. At the air flow rate of 15 lpm through the generator, the device is capable of generating up to 1000 Colony-Forming Units (CFU) per litre of the air carrier.

3. Results and Discussion

The results obtained for the aerosol filtration efficiency of the panels, calculated by Equation (1), are shown in Figure 3. None of the systems performed well at filtering particles of 2.5 μm diameter or less (i.e., PM_2.5); Tillandsia panels subjected to the highest air velocity (1.5 m/s) was the only configuration providing filtration efficiency of around 10% for these particle sizes. Nevertheless, particles larger than 2.5 μm were filtered with at least 10% efficiency by all systems except Tillandsia at lowest air velocity. Efficiencies for particles larger than 5 μm ranged from 31% to 71% for Banksia flower spike panels and 6% to 74% for Tillandsia panels. This pattern aligns with the general expectations derived from filtration theory. As particle size increases, the inertial mechanism of aerosol removal becomes more effective. Additionally, since inertial impaction is the dominant mechanism for capturing large particles, increasing the air velocity further enhances this process, which leads to an increase in overall particle collection efficiency. This interaction between particle size, air velocity, and filtration mechanisms is consistent with the theoretical understanding of filtration systems and their performance across different particle sizes.

An important parament of the air filtration processes is the pressure drop across the filtration media at the operational gas flowrates. Figure 4 shows the pressure drop recorded for both plant systems involved in this investigation. Tillandsia panels provided considerably lower pressure drop compared with the Banksia flower spike panels; even for the highest operational air velocity of 1.5 m/s, the pressure drop across the Tillandsia panels did not exceed 20 Pa. A higher but also acceptable pressure drop of around 70 Pa was observed for the Banksia flower spike panels operated at the highest flowrate. With the abovementioned similar aerosol removal efficiency results provided by the two species, Tillandsia panels seem to be more advantageous for having lower pressure drop at a comparable filtering efficiency.

The benefits of using Tillandsia are further highlighted in Figure 5, which illustrates the QF values calculated by Equation (2) for both panel types across the entire range of air velocities. The results were determined based on the average values of the particle size intervals, and these values are displayed on the x-axis of Figure 5. The QF for the Tillandsia panel reaches values between 0.04 and 0.06 Pa⁻¹ for larger particle sizes and the highest air velocities, which is comparable to the performance of moderately efficient fibrous filters (as seen in Figure 9 of [27]). Additionally, it is important to note that, due to the very low hydrodynamic resistance of the Tillandsia panel, its thickness could be increased—potentially doubled or even tripled—without a significant rise in pressure drop (only up to 50–60 Pa). This adjustment would result in anexponential increase in efficiency while only causing a moderate increase in pressure drop.

For reference, a common polystyrene fibre filter of 1.5 mm thickness and 0.7 μm fibre diameter provides an average filtration efficiency of 99.5%, a pressure drop of 2900 Pa, and a filter quality 0.001 Pa⁻¹ with an air velocity of 0.27 m/s [26].

The last set of experiments was focused on the capability of the panels to capture biological aerosols from the ambient air. As discussed above, A. niger spores were used in the experiments. As the highest velocities yielded the best performance in mineral aerosol tests with relatively low pressure drop, the air velocity of 1.5 m/s was chosen for biological aerosol (fungi) testing. Banksia flower spikes collected A. niger spores with 71% ± 14% average efficiency and low STDEV value, outperforming Tillandsias’ 57% ± 10% removal efficiency. Given that A. niger spores range from 3 to 5.4 µm [43], these measured filtration efficiencies were better than those measured for mineral aerosols of the same size, yet the results are still within a similar range. This outperformance could be associated with the more sophisticated shape of fungal spores captured particularly well by the more complex structure of the Banksia flower spikes particularly. According to a review of the literature by M. Tomson et al. [44], urban vegetation filters air primarily through two mechanisms: dispersion and deposition; these processes are enhanced by specific micromorphological leaf traits such as the presence of grooves, ridges, leaf hairs, a higher density of leaf pores, and a greater amount of epicuticular wax on the leaf surface.

The effective performance of the plant-based panels in filtering larger particles is a good indication that pollen can be collected at high efficiencies. Wind-borne pollen ranges from 10 to approximately 100 μm in diameter, with allergy-producing pollen often cited as being around 25 µm [45,46]. Consequently, these plant panels are expected to be promising collectors of allergenic particles and microorganisms, thereby enhancing air quality and public health. This aligns with the observations by H. Li et al. [47], who documented a significant inverse correlation between the extent of urban greenness and the relative abundance of pathogenic bacteria and fungi in the air.

Natural panel systems can potentially be designed for self-sufficient operation outdoors, by using ambient wind to achieve the predicted efficiencies and rainwater for flushing and renewing the filtration system. For indoor applications where there is less airflow and no ambient precipitation, a quiet, low-power axial fan achieving a 1.5 m/s airflow, together with an irrigation system, would be required to achieve the predicted filtration efficiencies.

4. Conclusions

Banksia flower spikes and Tillandsias were both found to be promising when arranged as plant-based panels to filter PM and fungi in the air. Tillandsia panels provided the best filter QF at air velocities of 1.0 m/s and 1.5 m/s, producing the highest PM filtration efficiencies (up to 74% for large particles) with low pressure drop. However, Banksia flower spike panels performed better at lower air velocity (0.5 m/s), with a filtration capacity of up to 53% compared with 13% for Tillandsia. Effective filtration of fungi (A. niger spores) was also observed, with both plant panels capturing over 50% of the airborne spores at the tested 1.5 m/s air velocity. Plant performance in filtering fungi is likely due to a combination of factors, including their physical structure and surface properties.

The filtration efficiencies found in this study suggest that incorporating plants into urban environments in the form of passive natural panels or living green walls can contribute to improved air quality and public health by reducing the concentration of airborne particles and fungi.

The efficiency of the panels in removing small particles was found to be low, which means they would not be effective at removing airborne bacteria or, even more so, viruses. However, the system demonstrated excellent performance in removing larger allergenic particles, such as fungi and pollen. Therefore, the plant panel systems may be a good natural solution for individuals suffering from allergic reactions.