Quantifying negative air ion release from urban tree seedlings in Beijing using a stripping curve approach

During photosynthesis, plants release O₂ at rates ranging from 0.5 to 2.0 g·m⁻² leaf area·h^-1 in temperate tree species¹, this O₂ can interact with atmospheric electrons to form negative air ions (NAIs). This process effectively enhances the negative air ion concentration (NAIC) in the atmosphere. Previous studies(Krueger, 1976; Hawkins, 1981) demonstrated that NAIs can absorb particulate matter and other pollutants, purifying the atmosphere^2,3. Additionally, they have been found effective in treating human diseases and maintaining overall health. Currently, the content and distribution of NAIs are considered important indicators of a region’s ecological environment and air cleanliness^5,6,7,8. The formation pathways of NAIs are diverse and influenced by a wide range of environmental factors, including climatic conditions, forest coverage, atmospheric flow, particle adsorption, and natural geographic conditions^9,10,11,12 . Due to these external environmental factors, NAIs in the air exhibit strong self-characterization and variability, with rapid generation and extinction rates, making them difficult to capture. Understanding the dynamic characteristics of NAI released by the plants at any given moment remains a problem to be solved.

The release mechanism of green plants, as the main source of NAI, is very complex, showing different characteristics at different spatial and temporal scales and being affected by a variety of factors^13,14,15,16. Existing studies have mainly focused on continuous monitoring of NAIC in natural environments such as urban forests, green spaces, fountains, and plant communities^{8,14,17,18,19,20,35}. Yang et al.²¹ studied the characteristics of the dynamic changes in NAIC and PM2.5 concentration in nine landscape recreational forests with different community structures, including arboreal-shrub-grass, arboreal-grass, and shrub-grass, in Xishan National Forest Park, Beijing. The daily dynamics showed that NAIC changed continuously throughout the day, with peaks appearing at different times²², and that most plants released higher NAIC during the daytime compared to nighttime^23,24. In terms of seasonal dynamics, NAIC was generally highest in summer and autumn and lowest in winter²⁵.

As mentioned above, most studies are based on field observation, where natural environmental conditions change rapidly, and numerous uncertain factors interfere, making it challenging to effectively determine the exact role of NAI released by the plants. In the present study, the stripping curve of NAI released by the plants was constructed for the first time by setting up a semi-enclosed open-top chamber (OTC) that effectively isolates external environmental conditions. This method allowed for the characterization of the continuous dynamic changes in NAI release, achieving the purpose of stripping the self-releasing effect from plants. This study is conducive to maximizing the potential of NAI released by common seedling species and will provide guidance for landscape area planning, seedling species selection, and configuration. We hypothesized that:(H1) Coniferous seedlings would exhibit significantly higher SR-NAI than broadleaf seedlings due to greater leaf tip density facilitating corona discharge;(H2) Daily NAI release would follow a bimodal pattern, peaking during maximal photosynthetic activity (10:00–12:00 and 14:00);(H3) Seasonal NAI release would be highest during peak growth (July–August) when photosynthetic capacity and transpiration rates are maximal. It is critical to note that these hypotheses were tested on 3-year-old potted seedlings, which differ substantially from mature trees in canopy structure, leaf area index, and atmospheric coupling. Therefore, scaling up these results to urban forest planning requires caution and field validation.

Overview of the study area and selection of seedling species

The experimental site was located in the experimental base of the Institute of Forestry and Fruit Tree Research, Beijing Academy of Agricultural and Forestry Sciences (latitude 39° 58′ 01″ N, longitude 116° 13′ 02″ E). This site is situated at the foot of Fragrant Hill in Beijing, adjacent to Xishan Mountain, and features a warm-temperate continental monsoon climate. The area experiences four distinct seasons, with high temperatures and heavy rainfall in summer, cold and dry winters, short springs and Autumn, and extremely uneven precipitation. The annual average temperature is 16.2 ℃, and the annual precipitation is about 484 mm²⁴.

The study selected two evergreen conifer trees in Beijing: Pinus tabuliformis and Pinus bungeana, along with six deciduous broadleaf seedling species: Acer truncatum, Sophora japonica, Quercus variabilis, Koelreuteria paniculata, Robinia pseudoacacia, and Populus tomentosa. These species are key recommendations for Beijing’s plains afforestation project. All seedlings were purchased from Wenzé Nursery (汶泽苗圃), Tai’an, Shandong Province, China. All selected seedling species were well-grown, had a consistent shape, and showed no signs of pests or diseases among 3-year-old seedlings. The seedlings were planted in a pot measuring 25 cm in height, with an upper caliber of 23.5 cm and a lower caliber of 19 cm. The potting substrate consisted of a potting substrate of local loam:sand:peat (3:1:1) with pH 6.8 and organic matter content of 12% (weight-based, determined by loss-on-ignition method following Nelson & Sommers, 1996). After cultivation in the test base, the seedlings were allowed to adapt for 14 days before the start of the experiment under ambient temperature (22 ± 2 °C), relative humidity (60 ± 5% RH), and regular watering conditions.

As shown in Table 1, an overview of experimental seedling species is provided. Plant height was measured in centimeters (cm) using a digital caliper (Mitutoyo Absolute, resolution 0.01 mm) from the root collar to the apical bud. Diameter at breast height (DBH) is inappropriate for seedlings; collar diameter (mm) was measured at 5 cm above soil level. All morphological measurements were recorded in millimeters (mm) for precision. Species characteristics are summarized below.

Research methodology

The experimental design consisted of two treatment groups: EXP (experimental group with plants) and CTRL (control group without plants). As illustrated in Fig. 2, the stripping curve principle calculates SR-NAI (plant self-released NAI) as the difference: SR-NAI = EXP—CTRL at each time interval, isolating the biogenic NAI signal by removing background atmospheric ionization (Fig. 1).

Test setup

The semi-closed device, known as the open-top chamber (OTC), allows for a certain degree of isolation between the plants inside the chamber and the external environment, effectively reducing the influence of external factors. The upper section is sealed with a vented glass lid (0.5 cm diameter perforations, 10 cm spacing) to minimize external ion ingress while permitting gas exchange for seedling respiration. The OTC has a height of about 2.5 m, with its main structure composed of stainless-steel tubing and colorless transparent glass. An electrically insulating PTFE film is applied to the inside of the chamber walls to maintain a relatively stable testing environment with minimal external interference. Each side of the OTC features a test hand entrance with a diameter of 0.2 m, positioned 0.5 m from the ground, facilitating maintenance and instrument measurement during the experiment. The experimental setup follows established protocols for controlled plant NAI studies, with chamber design adapted from previous open-top chamber research. CO₂ concentration was monitored continuously using an infrared gas analyzer (LI-COR LI-830, Lincoln, NE, USA) at 30-min intervals. Internal temperature and humidity were actively synchronized with external ambient conditions using a split-type constant temperature and humidity air conditioning unit (Midea KFR-72GW/BP3DN8Y-YH200(1)), achieving control precision of ± 2 °C and ± 5% RH as verified by HOBO UX100-003 temperature loggers and Onset HOBO MX2301 humidity sensors logging at 5-min intervals. The technical specifications of the OTC are detailed in Table 2.

Test methods

The experiment was conducted during the growing season, from May to October. Potted seedlings that had been slowed down in the early stage were selected. Six seedlings of each species were placed together in a single OTC, representing a density of 1.07 seedlings·m⁻³ (based on OTC internal volume of 5.625 m³). NAI release was measured per OTC (i.e., cumulative flux from all six seedlings), then divided by six to obtain per seedling values reported in Tables 2 and 3. An equal number of pots filled with soil was placed in the plant-free control group. To minimize experimental error, the outer surfaces of the pots in both the experimental and control groups were wrapped in PTFE.

Prior to measurements, dust was removed from leaf surfaces using a deionized water sprayer, followed by air-drying for 2 h. Leaves were gently wiped with a lint-free moist towel, and all pots were wrapped in PTFE film. Three independent replicates (OTC chambers) were used per species, with six seedlings per replicate, totaling 18 seedlings per species tested across the growing season. The plant-free control group contained identical PTFE-wrapped pots filled with substrate only. Each seedling species batch was observed continuously for 72 h (three full diurnal cycles), which has been shown sufficient to capture representative NAI patterns in preliminary studies. Twenty-four-hour NAIC data were collected daily from 8:00 to 20:00 the following day. During this time, the temperature and humidity settings inside the OTC were kept at the same levels as those outside. After each monitoring period, the chamber was purged with filtered air for 24 h to prevent carryover effects before introducing the next species batch.

NAI monitoring and data processing

The monitoring of NAIs using negative air ion instruments, along with data processing, evaluation methods, and other technical requirements, follows the Technical Specification for Observation of Negative (Oxygen) Ion Concentration in Air (LY/T 2586–2016) issued by the State Forestry Administration (SFA) in 2016. Inside each OTC, one set of the negative air ion instruments automatic monitoring system (model: AQ3600) is installed, equipped with an internal GPRS remote transmission module that transmits data directly to a cellphone server. The instrument measures the concentration of small-sized atmospheric negative ions with ion mobility ≥ 0.7 ion·cm⁻³·v⁻¹·s⁻¹, with a measurement range of 0 to 5.0 × 10⁶ ion·cm^-3 and an error of ≤ 5% in ion concentration measurement. The collection interval was set to 5 min per measurement. The height of the NAI measurement was adjusted according to the average growth height of the different seedling species being studied. After completing the monitoring of each seedling species, the room was left empty for 24 h before the next batch of seedling species was introduced, reducing any influence on the results of subsequent tests. The NAI measurement heights of the plant-free control group were synchronized with those of the planted experimental group.

All NAI data were processed following the single reference standard: Technical Specification for Observation of Negative (Oxygen) Ion Concentration in Air (LY/T 2586–2016). Data screening procedures included: (1) removal of instrument failure outliers, (2) elimination of consecutive identical values (> 3 repeats), (3) exclusion of values differing by > 3 × from adjacent measurements, and (4) linear interpolation for gaps < 30 min.²⁷.

Construction of the stripping curve

The OTC effectively mitigates the uncertainty and interference from various factors present in natural ecosystems. By comparing the experimental group with plants to the control group without plants, the process of increasing NAI concentration due to the stimulation can be revealed, allowing for the precise identification of the plants’ role in releasing NAI. Continuous monitoring of each plant’s NAIC through the experimental and control treatments provided the dynamic characteristics of NAIC released by the plants. The change curves of NAIC in the experimental group (EXP) and the control group (CTRL) were plotted, as shown in Fig. 2. These curves illustrate the change of NAIC released by the plants, resulting in the stripping curve (SR-NAI). Figure 2 was generated using OriginPro 2021(OriginLab Corp., Northampton, MA, USA). The stripping curve (GSR-NAI) was calculated by point-by-point subtraction: GSR-NAI(t) = GEXP(t)–GCTRL(t), where t represents each 5-min measurement interval. Statistical software used for all analyses was spss27 (IBM SPSS Statistics v27.0, IBM Corp., Armonk, NY, USA).

Characteristics of daily changes in NAI released by different seedling species

As shown in Fig. 3, the daily stripping curves of NAIC released by each seedling species demonstrate that the daily changes follow the pattern of EXP > SR-NAI > CTRL. All seedling species released significantly more NAIC compared to the control group. Self-released NAI (SR-NAI) was 1.79 to 3.66 times higher than that produced by atmospheric ionization in the control group (CTRL). SR-NAI could reach 74.88% to 90.24% of the levels observed in the experimental group with plants in a simulated natural environment, indicating that plant self-released NAI has a strong effect.

The NAIC curve of the CTRL exhibited a “single-peak” pattern, fluctuating between 32.65 and 367.36 ion·cm⁻³, with the peak occurring at 12:00 (317.02 ± 34.81 ion·cm⁻³). The NAIC curves of the EXP showed a “double-peak” pattern, with Robinia pseudoacacia and Populus tomentosa peaking at 10:00 and 12:00, releasing the strongest NAIC at 12:00. The other species showed peaks at 10:00 and 14:00, with the highest NAIC at 10:00. All seedling species released the lowest NAIC at 13:00.

The SR-NAI stripping curves derived from EXP and CTRL curves for each species, which represent the NAIC changes released by the trees themselves, follow the same trend as the EXP changes, displaying a “double-peak” pattern. The peaks appeared at the same time. The release of NAIC from each seedling species was as follows: Pinus tabuliformis (535.71 ± 90.26 ion·cm⁻³) > Pinus bungeana > Acer truncatum > Sophora japonica > Quercus variabilis > Koelreuteria paniculata > Robinia pseudoacacia > Populus tomentosa (261.92 ± 48.77 ion·cm⁻³). During the daytime, the highest range of NAIC release was 1364.33–1892.33 ion·cm⁻³ for coniferous trees and 626.50–1047.33 ion·cm⁻³ for broadleaf trees. The lowest NAIC release occurred in the late afternoon, ranging from 472.83 to 493.67 ion·cm⁻³ for coniferous trees and from 376.00 to 390.67 ion·cm⁻³ for broadleaf trees.

As shown in Fig. 4 and Tables 3 and 4, both diurnal and nocturnal NAIC released by coniferous trees was significantly higher than that by broadleaf trees (p < 0.05), being 1.48 times higher on average. Pinus tabuliformis exhibited the strongest NAIC release, while Populus tomentosa showed the weakest. Analysis of variance (ANOVA) showed that all seedling species released significantly higher NAIC than CTRL, with the following multiples: 3.43 times (Pinus tabuliformis), 2.96 times (Pinus bungeana), 2.46 times (Acer truncatum), 2.36 times (Sophora japonica), 2.27 times (Quercus variabilis), 2.11 times (Koelreuteria paniculata), 1.84 times (Robinia pseudoacacia), and 1.67 times (Populus tomentosa). Among the broadleaf trees, Sophora japonica and Acer truncatum had significantly higher NAIC than other broadleaf trees (p < 0.05), with release capacities being 31.23% and 34.50% higher, respectively, compared to Populus tomentosa, which had the lowest NAIC. After the mean multiple comparison test, there was no significant difference in NAIC between Quercus variabilis and Robinia pseudoacacia, and between Populus tomentosa and Koelreuteria paniculata. All seedling species released significantly more NAIC during the day than at night. The day/night NAIC release ratios were as follows: Pinus tabuliformis (2.02) > Populus tomentosa (1.98) > Sophora japonica (1.91) > Quercus variabilis (1.88) > Pinus bungeana (1.82) = Acer truncatum (1.82) > Robinia pseudoacacia (1.78) > Koelreuteria paniculata (1.75). Pinus tabuliformis, showed the greatest difference in diurnal NAIC release, while Koelreuteria paniculata showed the least.

Characteristics of monthly variation in NAI released by different seedling species

As shown in Figs. 5 and 6, the monthly stripping curves of NAI released by different seedling species during the growing season indicate that the order of release is EXP > SR-NAI > CTRL. The NAIC released by the plants themselves was about 1.95 to 2.78 times higher than the NAI produced by atmospheric ionization in each month of the growing season. Additionally, the NAIC released by the plants was highest in July, followed by August, June, September, May, and October.

During the growing season, the NAIC ranged from 117.16 to 174.20 ion·cm⁻³ for CTRL group, 331.17 to 664.73 ion·cm⁻³ for broadleaf trees (EXP group), and 530.60 to 799.39 ion·cm⁻³ for coniferous trees (EXP group). This data was used to construct the monthly stripping curves of NAIC release from the seedling species themselves (SR-NAI). The highest NAIC release occurred in July, with values ranging from 539.40 to 625.19 ion·cm⁻³ for coniferous trees and 334.77 to 490.53 ion·cm⁻³ for broadleaf trees. Conversely, the lowest release was in October, with ranges of 449.64 to 484.80 ion·cm⁻³ for coniferous trees and 228.32 to 343.04 ion·cm⁻³ for broadleaf trees. This indicates that the self-release of NAIC was stronger during the peak growth period for the seedling species.

All seedling species released NAI during the growing season, with coniferous trees showing higher concentration levels than broadleaf trees. The NAIC release for conifers ranged from 413.44 to 625.19 ion·cm⁻³, while broadleaves released NAIC between 214.01 and 490.53 ion·cm^-3, making the coniferous release approximately 1.27 to 1.93 times higher. Acer truncatum released the highest NAIC among broadleaf trees, with a monthly average of 399.92 ± 63.11 ion·cm⁻³, while Populus tomentosa released the lowest NAIC at 261.92 ± 43.83 ion·cm⁻³. Among coniferous trees, Pinus tabuliformis released the highest NAIC, with a monthly average of 535.71 ± 71.65 ion·cm⁻³, while Pinus bungeana released the lowest at 482.29 ± 46.46 ion·cm⁻³. This indicates that coniferous trees can induce greater NAI production through tip discharge, and different species exhibit varying capacities for NAI release.

Characteristics of seasonal variation in NAI released by different seedling species

As shown in Fig. 7, the seasonal stripping curves of NAIC released by different seedling species followed the pattern EXP > SR-NAI > CTRL, with all exhibiting a “single-peak” type. The NAIC released by the plants was highest in summer and lowest in spring. By stripping the NAIC change curves from the EXP and CTRL groups, the seasonal stripping curve for each species was obtained. In summer, the release of NAIC was the highest, with coniferous trees between fluctuating between 528.58 and 610.68 ion·cm⁻³ and broadleaf trees 309.69 and 463.72 ion·cm⁻³. Conversely, spring exhibited the lower NAIC release, ranging from 442.79 to 478.10 ion·cm⁻³ for coniferous trees and from 228.77 to 353.64 ion·cm⁻³ for broadleaf trees. This indicates that the release of NAIC by the plants is significantly influenced by meteorological factors such as temperature and humidity.

The NAIC of each seedling species in different seasons is shown in Table 5, revealing that conifers consistently release higher NAIC than broadleaves. The NAIC release from broadleaf trees ranges from 228.77 to 610.68 ion·cm⁻³, while conifers range from 442.79 to 610.68 ion·cm⁻³.

The ranking of seedling species in terms of NAIC release is as follows: Pinus tabuliformis > Pinus bungeana > Acer truncatum > Sophora japonica > Quercus variabilis > Koelreuteria paniculata > Robinia pseudoacacia > Populus tomentosa. Among coniferous trees, Pinus tabuliformis released the highest NAIC (610.68 ± 20.52 ion·cm⁻³), while among broadleaf trees, Acer truncatum ranked highest (463.72 ± 37.92 ion·cm⁻³). This suggests that branch and leaf tip discharges in coniferous trees are significant pathways for promoting NAI release. The seasonal variation patterns across species are visually compared in Fig. 8.

The consistently higher NAIC release from coniferous species compared to broadleaf species can be attributed to their morphological and physiological traits. For instance, coniferous seedlings exhibit greater needle tip density, which enhances corona discharge potential¹¹. Furthermore, higher stomatal conductance in species such as Pinus tabuliformis facilitates greater O₂ and volatile organic compound fluxes, thereby promoting NAI formation. These findings align with previous studies^25,26 and support the role of tip discharge as a key mechanism for NAI generation in conifers.

Inter-species differences in NAIC released by different seedling species

Different plant species exhibit varying capacities for NAIC release due to differences in morphological and physiological traits. Our findings confirm that coniferous trees release significantly more NAIC than broadleaf trees, which is consistent with previous reports^25,26. This study investigated the NAIC release characteristics of different seedling species within an OTC environment. The findings revealed significant differences in NAIC among eight common seedling species^29,30. This is because plants are the main source of NAI. The tip discharge of conifers prompts air ionization, and photosynthesis accelerates the plants’ ability to release NAI. The number of leaf tips in coniferous trees is positively correlated with NAIC^29,30,31, and electrons are released into the atmosphere using the leaf tips as a medium, significantly increasing the negative ion content in the air.

Significant differences in the release of NAIC by different seedling species on a daily, monthly, and seasonal basis

The observed bimodal diurnal pattern, with peaks at 10:00–12:00 and 14:00, aligns with periods of high photosynthetic activity and stomatal opening³. The midday trough at 13:00 likely results from stomatal closure under high vapor pressure deficit. These findings are consistent with Huang²⁸ who reported similar daily trends in broadleaf species under controlled conditions. The morning peak (10:00) suggests this as an optimal time for outdoor activities to benefit from elevated NAIC. NAIC was the lowest in the afternoon, which aligns with Huang²⁸ findings on the daily trend of NAI effect on the release of three native broadleaf trees under confinement conditions. This is because the release of NAIs is related to each seedling species’ own photosynthesis, with morning and evening being less affected by photosynthetically active radiation and less stomatal opening.

The amount of NAIC released by plants is also affected by meteorological factors such as temperature and humidity. Ma et al.³³ found that negative ion concentrations were significantly higher in the autumn than in the spring at the Xining study site, which is consistent with the results of this study. This study found that each seedling species released the highest NAIC in July and the lowest in October. This was due to the decrease in temperature in the autumn, which caused plants to enter the leaf drop stage, reducing the activity of enzymes involved in photosynthesis. In contrast, plants were in their peak growth period in July and August due to higher temperatures and longer daylight hours, resulting in significantly higher NAIC in these months³⁴.

Physiological mechanisms driving NAI release patterns: The bimodal pattern (10:00, 14:00) aligns with photosynthetic photon flux density (PPFD) peaks, but the midday trough (13:00) likely reflects stomatal closure due to high vapor pressure deficit (VPD > 2.0 kPa) Species-specific lag times (Robinia, Populus) may be attributed to differences in stomatal response kinetics³.The July maximum corresponds to peak leaf area index and maximum Rubisco activity¹, whereas the October decline reflects senescence-induced chlorophyll loss and reduced transpiration-driven ion exchange³⁶.

Significant differences in the diurnal and nocturnal release of NAIC among seedling species

Environmental factors, stand characteristics, and anthropogenic factors all influence the release of NAI from plants³⁶. Temperature and solar radiation are the main meteorological factors affecting NAI release. This study found that the NAIC release from each seedling species was significantly higher during the day than at night, aligning with Wang et al.²¹. This is mainly due to solar radiation. Seedling species with high 24-h mean values and elevated daytime and nighttime values did not necessarily exhibit the highest diurnal release differences. The observed day/night differences in NAIC release, with all species releasing significantly more NAIC during daytime, underscore the role of solar radiation and photosynthesis in NAI generation^21,34. Interestingly, species with the highest overall NAIC did not necessarily exhibit the largest day/night ratios, suggesting that nocturnal mechanisms (e.g., residual tip discharge) also contribute³². The diurnal-to-nocturnal ratio varied from 1.75 (Koelreuteria paniculata) to 2.02 (Pinus tabuliformis), reflecting species-specific photosynthetic characteristics. Based on these findings, we recommend prioritizing coniferous species for maximizing NAIC, and among broadleaf species, Acer truncatum and Sophora japonica are preferable choices. These variations are primarily related to the plant’s photosynthetic physiological properties³⁶ and also reflect the effect of light on NAI release from plants. When prioritizing tree planting, it is recommended to choose coniferous trees. To improve ornamental effects, a mix of needle and broadleaf trees is advisable, with Acer truncatum preferred among broadleaf trees, followed by Sophora japonica.

In this study, the stripping curve was established for the first time, effectively excluding interference from other factors to isolate the NAI released by the plant itself. By setting up NAIC change curves for the experimental group with plants (EXP) and the control group without plants (CTRL), the study was able to strip the stripping curve of self-released NAI by the plant (SR-NAI), characterizing the dynamics of NAIC release from plants. The study found the following: (1) The daily, monthly, and seasonal changes in the capacity of each curve for each seedling species were as follows: EXP > SR-NAI > CTRL. NAIC release was significantly higher in coniferous trees than in broadleaf trees, ranked as follows: Pinus tabuliformis > Pinus bungeana > Acer truncatum > Sophora japonica > Quercus variabilis > Koelreuteria paniculata > Robinia pseudoacacia > Populus tomentosa; (2) The peaks of NAIC released by each seedling species occurred around 10:00–12:00 and 14:00, with the lowest NAIC released at 13:00; (3) Each seedling species released the highest NAIC in July, the lowest in October, the highest in summer, and the lowest in spring. Several limitations should be acknowledged. First, seedling-scale extrapolation to mature urban trees is uncertain due to 10–50 × differences in biomass and canopy coupling with the atmosphere. Second, the PTFE lining, while reducing external interference, may amplify internal electric fields, potentially overestimating tip discharge relative to open-field conditions. Third, we did not measure VOC emissions , which are known NAI precursors. Future work should scale up to 10-year-old saplings in open-air lysimeters and integrate leaf-level gas exchange and electrostatic potential measurements to validate these findings.