Introduction

China has experienced severe air pollution problems in recent years due to rapid industrialization and urbanization and increased energy consumption (Zhang et al. 2014). The Chinese economy still mainly depends on fossil fuels. In 2011, 69% of total energy consumption was linked to coal combustion, with 18% related to oil and 4% to natural gas use. As a result, the ambient air has been profoundly affected, especially in urban areas (Chung and Kim 2008). In 2012, only 5 cities out of 367 in China met the air quality standards recommended by the WHO. The primary air pollutant in Chinese cities is suspended particulate matters (PMs), specifically PM2.5 and PM10 (PM10 denotes fine inhalable particulate matters with diameters that are generally 10 μm and smaller; PM2.5 denotes fine inhalable particulate matters with diameters that are generally 2.5 μm and smaller) (Gao et al. 2016). The PM2.5 concentrations in most urban areas have exceeded the acceptable national standards and are often worse in winter in North China due to increased fuel combustion for heating (Zhang and Crooks 2012; Gu et al. 2014). Shao et al. (2006) illustrated that over three-quarters of the population in urban areas is exposed to air quality levels that do not meet the Chinese National Air Quality Standard, and the primary contaminant in the ambient air is PM. Rohde and Muller (2015) found that high air pollutant levels are widespread throughout the northern and central parts of China, not only in the major cities and geologic basins but also in small cities and plain areas. They further noted that in the north-eastern corridor, which extends from Beijing to Shanghai, the air pollution is extremely severe.

In addition to PM, the concentrations of the other air pollutants, such as nitrogen oxide and carbon monoxide, have increased in recent decades. Data from the Chinese Ministry of Environmental Protection (2015) indicated SO2 emissions had increased to 25.5 Mt/year in 2005, 27% higher than the emission level in 2000. Additionally, NOx emissions increased by over 150% between 2005 and 2015, due to the installation of new power plants and increases in the vehicle population.

The 13th Five-Year Energy Development Plan of China claimed that over the next 10 years, coal will remain the dominant energy source in China. Therefore, in the future, China will face more severe air pollution problems (NDRC and NEA 2016).

Since it is impossible to eliminate ambient air pollution in a short time, people have begun to implement air pollution mitigation measures inside their buildings. A market study showed that the sales volume of household air purification systems has dramatically increased since 2013 (Zhu 2013). In 2011 and 2012, 1.12 million and 1.26 million air purification systems were sold in China, respectively (Zhao 2014). Then, an extreme increase occurred in 2013, when the severe air pollution issue became well publicized and the total number of air purification systems sold reached 2.4 million. In 2017, over 7.69 million air purification systems were sold in China, three times the number in 2013. Moreover, a report from AskCI Consulting (2017) indicated that air purification systems are more popular in autumn and winter than in summer and spring, which is consistent with the air pollution trends in China.

However, some studies have verified that these air purification systems cannot completely clean the air to a level that meets the Chinese standards and the WHO indoor air quality standards (Li et al. 2015). Moreover, these products consume electricity and thus indirectly increase building energy consumption, which further contributes to ambient air pollution. In addition, cost is another barrier to installing air purification systems widely. For example, in December 2015, the Beijing government announced that they cannot install air purifier systems in all public primary school buildings due to financial restrictions (Wang 2016). In Hangzhou, the Education Department of the Hangzhou government rejected a proposal to install air purifiers in primary school classrooms due to a financial shortage (Wang 2016).

In view of the drawbacks of air purification systems, alternatively, passive methods, such as the use of plants, have gained increased recognition. Many studies have tested the efficiency of plants in removing various air contaminants (Darlington et al. 2001; Lohr 2010; Wood et al. 2003). Plants play a significant role in regulating climate changes. They have a considerable impact on the global climate and are highly related to the local weather conditions. Plants can lower temperatures, reduce energy use, and improve air quality mainly through two processes: evapotranspiration and photosynthesis. Evapotranspiration involves the absorption of water from the soil, ocean, or waterbodies and the subsequent release of water as vapor into the air through the stomata, which increases air humidity and decrease air temperature. This process is part of the water cycle and influences the thermal comfort level of the air, especially in dry climate regions. Photosynthesis is a process used by plants to convert light energy into chemical energy and simultaneously release oxygen. The atmospheric oxygen levels are maintained at a steady level by photosynthesis, and the energy necessary for life on Earth is provided (Bryant and Frigaard 2006).

The atmospheric CO2 concentration has rapidly increased from 280 ppm to more than 380 ppm since the industrial revolution; moreover, this concentration is predicted to reach 700 to 1000 ppm by the end of the century (Team 2018). However, Berkeley Labs (2016) found that the CO2 levels in the atmosphere have stabilized in recent years because plants are removing more carbon from the air than in the past. The Conversation (2014) stated that land plants absorb approximately 25% of the carbon emissions produced by human activities. Milcu et al. (2012) indicated that plants can remove much more CO2 from the atmosphere than was previously believed. By absorbing CO2, plants emit oxygen. Usman et al. (2014) stated that plants play a significant role in human life by maintaining atmosphere oxygen levels and that the oxygen production level of some plants is greater than the CO2 absorption level. In addition to stabilizing the oxygen level and reducing the atmospheric CO2 level, plants also affect indoor CO2 levels. Tarran et al. (2007) indicated that among 55 city offices, rooms with three or more potted plants had 10% lower CO2 concentrations in an air-conditioned building and 25% lower concentrations in a non-air-conditioned building. Pegas et al. (2012) conducted an empirical study by installing six plants in a classroom of 52.5 m2 in Aveiro, Portugal. The results indicated that this classroom had a lower CO2 concentration by 50% compared to other classrooms without installing plants.

Plants can also directly clean the atmosphere by intercepting PMs in three ways (Beckett et al. 2000; Freer-Smith et al. 2004). The first way is retention. Plants reduce the airflow velocity and PMs settle on the surfaces of leaves. The quantity that settles on the surfaces of leaves is based on the airflow velocity and PM concentration in the ambient air. However, PMs adsorbed by retention can return to the air easily if the airflow speed increases and deteriorate the stable surrounding environment. The second way is attachment based on the rough leaf epidermis and cuticle characteristics. PMs are fixed to the leaf surfaces by attachment, which is relatively stable, and it is difficult for the wind to blow away the PM. The third way for plants to adsorb PMs is adhesion by the leaf exudates. Adhesion is the most stable way of capturing PMs and fixing them to the surfaces of leaves. Chen et al. (2014) claimed that some garden plants have the ability to remove PMs from the air and significantly improve the urban environmental quality. At the local scale, the inner portions of forest patches in urban areas exhibited significantly reduced concentrations of PMs compared to those near forest edges (Cavanagh et al. 2009). McDonald et al. (2007) claimed that in the West Midlands area of the UK, 110 tons of PM10 was removed per year by increasing the total tree cover from 3.7 to 16.5%. In Glasgow, the PM10 concentration was reduced by 2%, accounting for almost 4 tons per year, by increasing the tree cover from 3.6 to 8%. Nowak et al. (2013) stated that trees remove 4.7 to 64.5 tons of PM2.5 annually in Syracuse and Atlanta. In the Greater London area, almost 852–2121 tons of PM10 is annually removed by the urban tree canopy (Tallis et al. 2011). In the USA, urban trees remove approximately 214,900 tons of PM10 per year (Nowak et al. 2006). Trees can also indirectly reduce PM concentrations. For example, trees can alter the air temperature and influence building energy consumption (e.g., cooling the air temperature to reduce cooling energy consumption, shading buildings, and reducing solar heat gains), as well as subsequently affect the emissions from power stations (Nowak and Crane 2002; Beckett et al. 1998). Thus, plants both directly and indirectly play an essential role in the cleaning of the ambient air.

In addition to removing the PMs, plants also make a significant contribution to reducing chemical air pollutants, such as VOCs, ozone, NOx, and SOx. Tarran et al. (2007) performed laboratory studies and found that VOC load was significantly reduced by plants. Papinchak et al. (2009) stated that chambers containing snake plants, spider plants, and golden pothos had higher ozone depletion rates than chambers without plants. Takahashi et al. (2005) found that plants have a high capacity for the uptake of nitrogen dioxide.

Therefore, based on the benefits noted above, indoor planting is an important design factor that can improve energy savings, indoor air quality, and work performance. Table 1 lists some indoor plants that can reduce indoor air pollution.

Table 1 Plants and target air pollutants (NASA 1989)
Full size table

Identification of problems

As previously discussed, plants contribute to removing indoor air pollutants, such as PMs, VOCs, and CO2 (Pettit et al. 2017). Many studies, especially those conducted in chambers, have shown that plants can clean air. Orwell et al. (2006) indicated that plants have the capability to remove contaminants from the air based on a laboratory study in a small and sealed chamber. Wood et al. (2003) investigated the effectiveness of plants in removing benzene and hexane in a small sealed space and found that the concentrations of benzene and hexane were reduced by 80% and 70%, respectively.

However, many studies have claimed that the laboratory results are not adequate evidence for concluding that the use of plants indoors can result in significant reductions in indoor air pollutant levels. Dingle et al. (2000) performed a field study in office buildings in Perth, Australia, to investigate the effectiveness of plants in removing formaldehyde. The results showed “no change in the formaldehyde concentrations with the addition of 5 or 10 plants in the room and only an 11% reduction in formaldehyde concentrations with 20 plants in the room.” Wood et al. (2006) conducted a similar field study in Sydney, Australia. The study found that VOC concentrations in rooms with plants were randomly higher or lower than those in rooms without plants.

Therefore, conflicting conclusions were provided by different studies. On one hand, the reductions in pollutant concentrations by plants in chamber studies suggest that plants are capable of removing air contaminants. On the other hand, field studies have indicated that plants are not efficient in cleaning air in practical settings. Two reasons are listed below.

  • Firstly, most chamber studies were performed with many plants in a small space. Chambers can create a relatively stable environment for the plants. To observe the same effect in a real space, several times as many houseplants as required in the chamber may need to be included in the field study. Installing one or two plants for a large space will not make a significant difference in terms of air quality.

  • The second reason for the differences between the laboratory and field studies is that air pollutants in real buildings are constantly affected by human activities. There may be not enough time for plants to completely absorb these pollutants. Therefore, laboratory studies and field measurements are very different cases.

Therefore, the key issue that should be considered in the use of plants to clean indoor air is the number of plants and the stability of indoor environments.

Research hypothesis

Plants play a significant role in building design and have been widely used in many ways, including green facades, garden atriums, vertical forests, and planted terraces. For example, the University of Guelph-Humber established a four-storey plant wall in their central atrium in early 2000s to recycle more inside air and reduce the need for heating and cooling. The plant wall was made up of over 1000 plants from 100 different species. In addition, it was fully integrated into the building’s ventilation system and processes 1133 m3 of air very minute, removing air pollutants, and decreasing the temperature (Guelphhumber n.d.). Another example is the Junglefy Breathing Wall developed by the company Junglefy Pty Ltd. and the Plants and Environmental Quality Research Group at the University of Technology Sydney. It is an active, modular green wall system that pulls the air through the leaves of plants by an electric axial impeller. Report published in 2016 indicates that a Breathing Wall of 10 m2 in size can produce 1623.4 m3 particle-free air per hour on average and balance-out CO2 emissions from 4 to 5 occupants during normal working hours (Irga, et al. 2016).

Based on the literature review, this research proposed an integrated ventilation system designed for buildings (Fig. 1). The concept of the greenhouse with plants originated from laboratory studies involving small sealed chambers with many plants. Therefore, the sunspace in this system acts as an enlarged air storage chamber. The plants were placed inside the sunspace rather than scattering them randomly inside the buildings. Air flows into the greenhouse at a slow speed to ensure the stable inside environment. Therefore, the use of plants eliminated many factors that affect the plant cleaning efficiency, such as wind and occupant activities. Moreover, the sunspace provided a stable indoor environment for plants to remove air pollutants.

Fig. 1
figure 1

Initial idea for plants’ utilization in building design

Full size image

Although the greenhouse originated from chamber studies, a significant difference also should be emphasized. In chamber studies, normally, air pollutants were only introduced once, and the air was then monitored to observe the reductions in pollutant concentrations. In the greenhouse, air constantly flows into the greenhouse at a very slow speed. Thus, the cleaning efficiencies of plants in these two situations were also different.

To identify the potentialities of the ventilation system presented in Fig. 1 and also determine the capabilities of plants to remove air pollutants in a dynamic process (air constantly flows into the sunspace), an experimental work was conducted based on the financial and material resources currently available to this research.

Experimental setup

The experimental study established a greenhouse with a steel framework and polyethylene films to represent the sunspace designed in the ventilation system. Theoretically, the inside environment of the greenhouse was more stable than the ambient environment. The width, length, and height of the greenhouse were 1.4 m, 14 m, and 2 m, respectively. The primary air pollutants in China are currently PM2.5 and PM10 (Peng et al. 2017); thus, the experimental work was mainly used to investigate the changing trends of PM concentrations. Moreover, the indoor PM concentrations were measured at four monitoring points (see Fig. 1 in the Supplementary Materials).

The greenhouse was located in Tai’an city. The city is part of the cold climate zone in China, which is bitterly cold in winter and hot in summer, indicating a large temperature variation. The coldest average monthly temperature was between − 10 and 0 °C in winter, while in summer, the highest air temperature usually was above 30 °C. It also has heavy ambient air pollution. Based on the data from the local environmental protection department, the daily highest ambient PM2.5 and PM10 concentrations in 2019 were 195 μg/m3 and 262 μg/m3 on average, respectively (Environment Protection Department of Shandong Provincial Government 2019).

Based on the available financial resources, 20 potted plants (labeled from A, B, C…, to T) bought from the local market were placed inside the greenhouse (see Fig. 2 in the Supplementary Materials). The plants used were Epipremnum aureum because it is effective in removing PMs from the air (Zhou et al. 2009). The 20 plants were put in clay containers full of grow potting mix. Moreover, it is common in the plant market in the case study area of China and its number of leaves can be easily counted. The selected potted plants had similar heights, crown diameters, and ages.

Many methods, such as the graticulation method and regression equations, can be used to estimate the surface area of leaves. Some devices, such as portable leaf area meters, can also directly measure the surface area of leaves (Xu et al. 2002). In this study, due to resource limitations, only scanners and drawing software, including Photoshop and AutoCAD, were used to quantify the leaf surface area. The scanners scanned a single leaf and imported the scanned picture into Photoshop and AutoCAD. Then, AutoCAD can directly provide the surface areas of leaves (see Fig.3 in the Supplementary Materials).

To estimate the surface areas of leaves on each plant, the number of leaves should be determined. Table 2 summarizes the number of leaves on each plant. The maximum number of leaves was 160 for plant H. Leaves falling and growing are two natural processes that occur simultaneously. Within a specified period, it is assumed that the number of leaves and the leaf surface area is constant.

Table 2 The quantity of leaves on selected plants
Full size table

According to the empirical principles of the sampling survey, for a subject with a population less than 1000, the smallest number of viable samples should be at least 30% of the total population (Shao 2012). Therefore, for the selected 20 plants, 50 leaves from each plant were collected as samples. The samples of leaves were collected randomly from the bottom, middle, and top of each potted plant. Then, the overall leaf surface area was estimated based on these samples. Table 3 summarizes the average leaf surface area data for each plant. The average surface area for one plant was approximately 1.16 m2.

Table 3 Detail surface areas calculations of leaves from the selected plants (A-T) (m2)
Full size table

To ensure that the air could pass through as many plants as possible, it flowed into the greenhouse from inlet A1 and through the outlet A2 (see Fig. 1 in the Supplementary Materials). A fan with a rated power of 50 W was used to generate an air flow rate of 225 m3 per hour. Moreover, the diameter of the fan’s outlet was 100 mm. Therefore, the supplied air flow speed at the outlet of the fan was 8 m/s. A connection between the fan and the pipes was created to generate different airflow speeds (see Fig. 4 in the Supplementary Materials). For example, when the fan was connected to a pipe with a diameter of 400 mm, an airflow speed of 0.5 m/s could be generated. The box in the middle acted as a connector between the pipes and the fan. The diameters of selected pipes were 250 mm, 300 mm, and 400 mm, and each pipe corresponded to one air flow velocity (Table 4). In addition, polyvinyl chloride (PVC) pipes with different diameters were used for the measurements (see Fig. 5 in the Supplementary Materials).

Table 4 The diameters of the selected pipes and the corresponding airflow speed
Full size table

Totally, five portable hand-held detectors of BRAMC-SMART-126 models (see Fig. 6 in the Supplementary Materials), which have been certified by the Department of Environment Protection of Central Government of China, were used to measure the PM concentrations at the inside monitoring points of the sunspace and the outside PM concentrations, simultaneously. The detector is a high-precision dust sensor capable of detecting 0.5-μm particulate matters. It provides great airflow intake and generate consistent and accurate results.

To identify the influences of the plants on the inside PM concentrations, measurements of different scenarios were conducted (Table 5). Totally, 8 scenarios were created, and the duration of each measurement was 24 h. The plants were cleaned in advance before each measurement. For example, the plants’ leaves were washed 1 day before the measurements.

Table 5 A summary of the measurement
Full size table

Measurements results and discussion

Inside and outside PM concentrations

Table 6 indicates the indoor and outdoor PM concentrations and the ratios of indoor PM concentrations to outdoor PM concentrations (I/O ratios) of scenarios. The indoor PM concentrations were estimated based on the measurement results from the 4 monitoring points (see Fig. 1 in the Supplementary Materials). Figures 2, 3, 4, and 5 show the changing tendency of PM concentrations in both indoor and outdoor environments in scenarios 1, 2, 6, and 7. It was noted that in all scenarios, the indoor PM concentrations were all lower than the outdoor PM concentrations. Thus, in an enclosed space, the stable inside environment led to a lower inside PM concentration, as more PMs were retained, attached, or adhered on the leaf surfaces of the plants in the greenhouse.

Table 6 Measurement results (24-h average)
Full size table
Fig. 2
figure 2

Measurement results of scenario 1 (without air supply, and plants were placed inside greenhouse)

Full size image
Fig. 3
figure 3

Measurement results of scenario 7 (air supply at 0.5 m/s, and plants were not inside)

Full size image
Fig. 4
figure 4

Measurement results of scenario 2 (no air supply, plants were not inside)

Full size image
Fig. 5
figure 5

Measurement results of scenario 6 (air supply at 0.9 m/s, plants were placed inside)

Full size image

In scenarios 1 and 7, there were no plants inside the sunspace. Therefore, the ambient PM concentrations were the sole factor that affected the inside PM levels. Moreover, it was also noted that the changing tendency for both the inside and outside PM concentrations was more consistent than those situations in scenarios 2 and 6. A similarly changing tendency also can be observed in scenarios 3 and 5.

However, at some points, the inside PM concentrations were closed to the outside PM concentrations. For example, in scenario 1, there was a rapid decline of the outside PM concentrations between the time 11:00–13:00, and then, it quick reversed before the inside PM concentrations could respond. Therefore, the inside PM concentrations reached a similar level as the outside PM concentrations. In scenario 7, the rapid decline of the ambient PM levels led a significant decrease of the inside PM levels of the sunspace.

In scenarios 2 and 6, plants were moved into the greenhouse. It was noted that the inside PM concentrations were more stable than outside PM levels because plants provided more inner surfaces for PMs to sunk. For example, in Figs. 4 and 5, the outside PM concentrations fluctuated greatly during 9:00–14:00, while the inside PM concentrations were maintained at a stable level. A similarly pattern also can be found in scenarios 4 and 8.

In general, in an empty space, the ambient PM concentration was the only influential factor that affected the indoor PM concentrations. This finding was also proved by the field studies conducted by Peng et al. (2017), which stated that, in unoccupied space, the inside PM concentrations are normally lower than the outside PM concentrations. In addition, plants inside the greenhouse provided more inner surfaces for PMs to be retained, attached, or adhered. Thus, the inside PM levels were more stable than that case of greenhouse without plants.

I/O ratios

Figures 6 and 7 describe the I/O ratios of different scenarios. Notably, the greenhouse with plants inside has lower I/O ratios than the sunspace without plants inside. I/O ratios of scenarios 1, 3, 5, and 7 were between 0.64 and 0.98. In scenarios 2, 4, 6, and 8, the values ranged from 0.23 to 0.66.

Fig. 6
figure 6

Performance of I/O ratios of PM2.5 in different scenarios

Full size image
Fig. 7
figure 7

Performance of I/O ratios of PM10 in different scenarios

Full size image

The I/O ratios of scenario 1 were much higher than those in scenario 2. The only difference between scenario 1 and scenario 2 was the presence of plants. Scenario 2 revealed that the plants can remove approximately 38% of PMs from the infiltration based on scenario 1. Thus, there is no doubt that the plants increased the surface areas of the greenhouse and more of the PMs originated from the outside were retained, attached, or adhered on the surfaces of leaves.

Figure 8 shows the comparison between scenarios 1, 3, 5, and 7. The airflow speed is the single variable in this comparison. It was noted that in scenario 1, the greenhouse without plants has approximately 77% of the PMs coming from the outside by infiltration on average. In scenarios 3, 5, and 7, the I/O ratios were slightly higher than that in scenario 1, since air comes inside not only by infiltration but also through the supply of the fans at speeds of 1.3 m/s, 0.9 m/s, and 0.5 m/s, respectively.

Fig. 8
figure 8

A comparison of I/O ratios of PMs between scenarios 1, 3, 5, and 7

Full size image

Scenarios 4, 6, and 8 suggested that when the air flows at a relatively slow speed, a greenhouse with plants still has a lower PM concentration than those of the ambient air. Scenario 6 performed similar I/O ratios to those of scenario 8. Scenario 4 has higher values than scenarios 6 and 8 on average.

Comparison between scenarios 3, 5, and 7 and 4, 6, and 8 suggested that plants can significantly reduce the inside PM concentrations although they have the same airflow speeds. For example, in scenarios 3 and 4 (airflow speed was 1.3 m/s), plants approximately removed 28% of the inside PMs based on the level of scenario 3. In scenarios 5 and 6, 7, and 8, this number was 30% on average.

Supported by findings from the field study conducted by Peng et al. (2017), a consistent conclusion was raised based on the analysis of I/O ratios in this research: an enclosed space with plants tends to have a lower PM concentration than the ambient environment. In addition, plants and slow airflow speed have a positive effect on the reduction of indoor PM concentration.

The designed experiment contains two variables: plants and airflow speed. Comparisons between scenarios revealed the significance of the plants in removing PM and the effects of the airflow speed on reducing PM concentrations. These comparisons have directly answered the research question raised earlier in the “Identification of problems” section. The key issue that should be considered in the use of plants to clean air is the number of plants and the stability of the indoor environments.

Discussion

Taking scenario 1 as a baseline, the removal efficiencies of plants in scenarios 2, 4, 6, and 8 were summarized in Table 7. Scenario 2 has the highest removal efficiency, as the values for PM2.5 and PM10 were 0.36 and 0.39, respectively. Scenario 6 has similar removal efficiency to those of scenario 8. Scenario 4 has lower removal efficiency than scenarios 6 and 8, on average.

Table 7 Removal efficiency of plants in different scenarios
Full size table

Compared with an earlier study from Irga et al. (2017), it was noted that only scenario 2 has removal efficiencies of PMs within the range determined by Irga et al. (2017). In contrast, scenarios 4, 6, and 8 have lower removal efficiencies (see Fig. 7 in the Supplementary Materials). The differences between these two studies may be explained as follows.

  • Firstly, the study from Irga et al. (2017) was carried out in a chamber which is relatively small compared to the greenhouse used in this work (see Fig. 8 in the Supplementary Materials).

  • Secondly, the airflow rate in this work was much higher than the value used in the study from Irga et al. (2017) (62.5 L/s vs 11.25 L/s).

However, due to the limited financial resources, the greenhouse was built with a steel framework and polyethylene films. Moreover, the connection of the polyethylene films of the sunspace only relies on the transparent adhesive tape (see Fig. 9 in the Supplementary Materials), which might lead to a low level of airtightness. Therefore, the greenhouse might have a high infiltration rate allowing more outdoor PMs penetrations, resulting in high I\O ratios and low removal efficiencies, even though plants can remove parts of them. Moreover, a high infiltration rate also leads to a high indoor pollution load, which might accelerate the process of adsorption saturation of plants since their adsorption capacity is certain. Supported by the findings of Peng et al. (2017), if materials such as glass were used to build the greenhouse, the airtightness levels would be significantly improved. Hence, the removal efficiencies can be further improved.

Conclusions

This research provided information about an experiment designed to evaluate the capabilities of plants to remove PMs. The measuring results have generated the following conclusions.

  • An enclosed greenhouse has a lower PM concentration than the outside environment as the ventilation only relies on infiltration.

  • Plants can further reduce the indoor PM concentrations, since they increase the surface area of the greenhouse. Therefore, more PMs coming from outside can sink on the surfaces of leaves.

  • Airflow speed has adverse effects on the efficiency of plants’ PM removal. A relatively slow air supply speed is beneficial to create a stable indoor environment and to increase plants’ efficiency in reducing PM concentrations.

This research focused on plants’ capabilities to remove PMs (PM2.5 and PM10). The changes in the concentrations of other air pollutants inside the sunspace were excluded from this research. For example, plants could reduce the CO2 concentrations and increase the oxygen level by photosynthesis. Other air pollutants, such as VOCs, NOx, and SOx, could also be removed from the air. Future work could focus on the potential cleaning performance of the plants regarding CO2 and other air pollutants. Due to the limited financial resources, the sunspace in the experiment was built with a steel framework and polyethylene films. The airtightness could be significantly improved if glass was used and the reduction of PM inside the sunspace could be different.