Phytotoxicological effects of phytosynthesized nanoparticles: A systematic review and meta-analysis

Abstract Nanomaterials (NMs) have become one of the most attractive new materials because of their excellent performance and application in the fields of chemistry, agriculture and medicine. However, high prices and potential pollution also limit widespread application of NMs. To prepare cheaper and widely applicable NMs, biological materials, especially plant materials, are being tested for their feasibility as the host of carbon NMs (CNMs) or the carrier of metal NMs (MNMs). This review analyzes the performance and practical applications of current plant-based NMs (PB-NMs) by meta-analysis. The results show that the particle size of PB-NMs is significantly smaller than traditional NMs. Plants improve the production, and hence potential application of NMs, resulting in not only economic advantages but also environmental suitability. At the same time, CNMs below 20 nm have significant protection and promotory effects on plants compared with larger size NMs (above 20 nm) NMs also show contrasting effects on different parts and growth stages of plants. Based on these results, plants can be used as raw materials to produce efficient, environmentally friendly and low toxicity NMs. The applicability of new nanomaterials, including their environmental effects, physicochemical properties and economic advantages, can also be preliminarily confirmed through meta-analysis. Graphical Abstract Abstract Figure Difference between PB-NMs and non PB-NMs in properties and phytotoxicity


Introduction
Nanomaterials (NMs) are minute structures with unique properties, measured at the nanoscale.Among the various types, metal nanomaterials (MNMs), and carbon nanomaterials (CNMs) are the two most studied (Saleh, 2020).Examples of MNMs, include Ag NMs and TiO 2 NMs while multi-walled carbon nanotubes (MWCNT) and graphene oxide (GO) are examples of CNMs.Properties of NMs are significantly different from those of macroscopic materials, such as differences in magnetism, catalytic ability, mechanical properties and sensing (Abid et al., 2022, Hauser & Nowack, 2021).Therefore, NMs have been broadly applied in optics and electrics (Kumar et al., 2015, Narayanan et al., 2019), while the application of NMs in medicine and environmental protection fields has been gradually increasing in recent years (Su & Kang, 2020, Ren et al., 2022).
However, current NMs still have many problems.On the one hand, production scale is difficult to satisfy the market demand, resulting in high prices (Singh et al., 2018).On the other hand, production processes consume a lot of energy and possibly generate secondary pollution (Lin et al., 2021, Letchumanan et al., 2021).Hence, these ultimately cast doubt on the practicality of NMs.Hence, there is a need to research the economical and practical applications of NMs.NMs tend to enter ecological cycles due to the long-term wear (Candotto Carniel et al., 2020).Then, NMs gradually accumulate in organisms and affect genetic processes (Liu et al., 2020).This accumulation poses a potential threat to microorganisms, plants, and animals alike (Carley et al., 2020, Ge et al., 2018).Additionally, exploring the effects of NMs on organisms is crucial for the future development and application.
Plants are becoming a good source material for economical and environmentally friendly NMs, which serve as not only the host to produce CNMs, but also the main carrier of MNMs to enclose or attach metal elements (Vijayan et al., 2018, Wang et al., 2022).In terms of economic impact, some plants such as wheat from China (Chen et al., 2019a) or bamboo from India (Tade & Patil, 2020) have a high yield and a low price, and even the waste products can be used, such as sugar cane bagasse (Du et al., 2014).In terms of practicality, these plant-based nanomaterials (PB-NMs) have exhibited excellent properties in the environmental and biomedical fields.For instance, PB-NMs exhibit higher photocatalytic degradation rates (Prasad et al., 2019, Elumalai et al., 2015) and prove to be more suitable as anticancer agents compared to traditional NMs (Yang et al., 2021).This superiority may be attributed to their special functional groups and the mechanical strength derived from the fibrous plant structure (Li et al., 2021).In terms of environmental impact, there is currently no evidence that PB-NMs can be reabsorbed by organisms, but some traditional NMs have been shown to be taken up by different organisms (Ge et al., 2018(Ge et al., , A et al., 2019)).Thus, it is important to conduct systematic sorting and analysis for the production, stability, practicality and biotoxicity of PB-NMs compared with traditional non PB-NMs.
The potential threat of NMs to plants is concerning due to the function of plants as producers in the ecosystem and the increase in broad applications of engineered nanomaterials (ENMs) for sustainable agriculture (Murali et al., 2022, Cai et al., 2023).The inhibitory aspects of NMs on plants is diverse, including germination rate (Lee et al., 2018), biomass (Lahiani et al., 2018), photosynthesis (Bai et al., 2021), enzymes (Cota-Ruiz et al., 2020) and gene expression (Jackson et al., 2022, Chen et al., 2019b).On the contrary, some NMs show a promoter effect on plant growth from biomass to stress resistance (Adeel et al., 2021, Gohari et al., 2021) and have even been used in delivery of DNA and proteins (Demirer et al., 2019).In addition, NMs also have equally contradictory effects on rhizosphere microorganisms and indirectly affect the plant's environment (Ge et al., 2018, Wang et al., 2014).Therefore, it is urgent to conduct systematic sorting and analysis of these contradictory results to determine the tendency of NMs to affect plants.
Meta-analysis is the examination of data from several independent studies on the same theme to determine overall trends (Wang et al., 2020a).It has been successfully used as an analysis method in the medical field (Dang et al., 2022).Through meta-analysis, the research results of different samples can be comprehensively calculated to improve statistical analyses and reduce data inconsistency (Huangfu & Atkinson, 2020).This may lead to improvements in estimating promotion or inhibition, and the discovery of new research questions.Furthermore, answers to a scientific question under the influence of a single variable can be visually presented in a more accurate manner than traditional tabulated summaries (Althea A. Archmiller et al., 2015, Hauser & Nowack, 2019).However, meta-analysis has not been used much in studying the application and impact of NMs on the environment, especially non-human beings (Book & Backhaus, 2022).Therefore, this review uses meta-analysis as the main analysis tool to more intuitively reveal the differences in stability, properties and phytotoxicity between PB-NMs and traditional NMs.
This review mainly used meta-analysis to compare and systematically analyze the differences between PB-NMs and traditional NMs.The main goals of the review are to explore and discuss the following: (1) to investigate the advantage of PB-NMs compared with traditional NMs; (2) to reveal the phytosociological effects of NMs and predict the phytosociological effects of PB-NMs; (3) to discuss the feasibility and limitations of meta-analysis for NMs and phytotoxicology.

Literature search and eligibility criteria
The "Web of Science" platform was the main literature search platform used in this study.In addition, the results from Google Scholar and Nature platforms were also utilized and compared to make up for shortcomings by using the software Publish or Perish 8.6.The keywords used were "plant-based nanomaterials" OR "plants and nanomaterials" when searching for studies on PB-NMs.When searching for studies on the effects of nanomaterials on plants, the keywords used were, "nanomaterials" OR "nanoparticles" AND "plants".In the search results, conference papers, newspapers and review articles were excluded.
The literature was divided into two categories: "PB-NMs" and "Phytotoxicity of NMs".Data were collated from the literature based on the following: (1) Literature describing complete and detailed accounts of the experimental work, for example, purchase of NMs, sampling of plants, cultivation and determination of phytotoxicity; (2) Experimental groups of "PB-NMs" where plants were used to prepare NMs.Experimental groups of "Phytotoxicity of NMs" that must use plants for the research, and did not mix with microorganisms, soil insects and human factors.
Experiments that used at least triplicate samples; (3) Literature providing mean, median or mode with standard deviations, including absolute or relative data (Dang et al., 2022, Zhang et al., 2022).The whole research and collation process is shown in the Supplementary Information (SI) Figure S1.In the literature, each type of NM is recorded as a sample (N).There are 59-235 samples from 171 references that are used in different analyses (Figure S1).In the process of sample extraction, the mean and standard deviations were the first to be extracted.For samples with standard error, the standard deviation can be obtained by formula conversion (Lee et al., 2015).For samples that do not give the relevant difference, the standard deviation can be estimated by the formula (Zhang et al., 2020a).The number of samples used for each time of meta-analysis is displayed in the Results section.The formulas used are shown below.
1.The standard deviation (SD) was calculated from the standard error (SE).Applicable when both numerical value and SE are provided.
SD means standard deviation; SE means standard error; N means number of samples.
2. The standard deviation (SD) was calculated from the mean coefficient of variation (m r ).This is applicable when only numerical values are provided, but no SE or SD is provided.
SD i means standard deviation of literature without standard deviation; SD r means each known standard deviation; M r means each known data (mean, median or mode); m r means mean coefficient of variation; M i means known data (mean, median or mode) of required literature.

Meta analysis of data
Meta-analysis is used on data for systematic analysis and visualization.The detailed information required for analysis includes the name of the author, published year, data of both experimental and control groups which is a process of quantitative analysis.However, there are many indicators that cannot be quantified in two directions on which this study focuses.For example, the effects of nanomaterials on plant cells and genes are usually stated in the literature as "promoting effect" or "inhibitory effect".Therefore, the meta-analysis in this paper adopts quantitative and qualitative analysis methods according to different indicators (Dang et al., 2022).Before the analysis was carried out, Akechi Information Criterion (AIC) supported by software Review Manager 5.4 was used to determine whether the analysis should use a random or fixed effects model (Huang et al., 2019).Ultimately, the random-effects model was selected because the AIC value of this model was small, and it is more commonly used in the environmental and materials research fields (Zhang et al., 2022).

Quantitative meta-analysis of data
The main targets of quantitative analysis are the particle size of NMs and the growth indicators of plants.
1.The difference in particle size between PB-NMs and traditional non-PB-NMs.The single variable is "plants".The characterization index is "NMs particle size" (quantitative index).Particle size stands out as a typical and frequently cited quantitative index, offering insights into the characteristics of NMs (Hu et al., 2020).Therefore, particle size is used in this meta-analysis to characterize the differences between PB-NMs and non-PB-NMs.2. The effects of NMs on plant growth.The single variable is the "use of NMs".The characterization indicators are "plant growth indices" (quantitative indicator).
Plant growth indicators, including root length, plant height, dry/fresh weight and biomass, are generally quantitatively given in the form of absolute or relative values in the literature.Therefore, these growth toxicity indicators are used for quantitative analysis.

Qualitative meta-analysis of data
Meta analysis requires the use of specific data, thus qualitative indicators need to be quantified, which is a problem faced by both the scientific and social scientific communities but there are some solutions (Sijbesma & Postma, 2008, Frericks, 2021).This review focuses on two main goals, which are the practicalities of PB-NMs (such as economy and accessibility) and NMs effects on plants.Therefore, two scientific questions are raised around a single variable in these areas: i. Effect of plant addition on NMs.The single variable is "plants".Localization, non-food/non-medicinal and consumption are three important indicators that researchers focus on when discussing the practicality of PB-NMs.Based on this, this meta-analysis constructed a simple scoring system as shown in Table S1.
ii.The effects of NMs on plant growth.The single variable is the "use of NMs".
The five most reported aspects of NM toxicity on plants are used in this meta-analysis.These include growth, function, cells, genes and others (Wang et al., 2020a, Sun et al., 2020).A scoring system was constructed as shown in Table S2.

Subgroup meta-analysis and bias correction of data
Sub-analysis in meta-analysis can divide samples into different groups to explore the characteristics and effects of the same or similar NMs, reducing masking of individual phenomena caused by large amounts of data.Comparisons between groups can also highlight differences between them (Chen & Hoek, 2020).Therefore, PB-NMs were divided into six groups according to two categories (MNMs and CNMs) and three particle size ranges (1-10, 10-100, 100-1000 nm) as shown in Figure 1.Traditional NMs were divided into two groups according to two categories (MNMs and CNMs) in Figures 2 and 3.The sample size of each group should be >10 to meet the needs of bias analysis.
The test and visualization of bias is achieved by a funnel plot.Specifically, data from the literature directly affects the accuracy of the results of the whole analysis (Zhang et al., 2022).Therefore, a second manual verification can be performed for the authenticity of the literature to determine whether it still needs to be included, and ultimately improve the accuracy of the results.
Visualization and analysis of data was carried out using R 4.2.1 and Adobe Illustrator 2020 (AI).

Characteristics of PB-MNMs and PB-CNMs
Figure 4a and b show the percentage of studies that have used that particular shape of particle and potential application of PB-NMs, respectively.Figure 4a illustrates shape types of PB-NMs Figure 1.Quantitative analysis of differences between PB-nms and common non-PB-nms by sub analysis according to both type and size of nms. the blue horizontal line represents the confidence interval of the nms and corresponds to the 95%-Cl on the right.horizontal lines completely on the left or completely on the right side of the vertical line (invalid line) represent that the experimental group is significantly smaller or larger than the control group, otherwise there is no significance.the red diamond represents the overall data of the group, and the significance judgment is the same as the horizontal line.the dotted line represents the position of the mean of the whole data, so that it is easy to compare with each group of data.
to reflect their structural differences.Fibril, layer and spherical are the most studied structures of PB-CNMs, accounting for 63.70% of all samples in total.The proportion of spherical PB-MNMs (39.02%) studied is much higher than the remaining shapes.Shape also represents differences in particle size, surface area, and physicochemical properties.The cross-sectional diameter of PB-NMs such as "fibril" is only 3.6-8.0nm, but the length can reach 352-624 nm."Spherical" refers to regular or nearly regular spherical particles with a diameter of 5-50 nm. Figure 4b shows that biomedicine is the main potential application field of PB-CNMs and PB-MNMs, accounting for 55% and 57%, respectively, followed by the environmental protection field.CNMs and MNMs have potential applications in food engineering and the chemical industry, respectively.

Characteristics of PB-NMs and traditional NMs
Figure 1 shows whether there are similarities between PB-NMs and non-PB-NMs.The particle size of NMs is used as a quantitative indicator for meta-analysis, and sample data from 52 studies were used.The detailed analysis can be viewed in Figure S1-2.The analyzed data were integrated as Figure 1, covering six groups.Particle size data of non-PB-NMs do not come from literature that provides PB-NMs size, because the latter rarely makes comparisons between these two NMs.Therefore, particle size data of some traditional NMs are compared to PB-NMs (these data can be viewed in Supporting Information).Figure 1 shows that the particle size of PB-NMs is generally smaller than that of common non-PB-NMs with a significant difference (p < 0.01).66.66% of PB-NMs groups' center points (mean of group) are on the left side and 22.22% of points are on the right side.The rest are on the central axis, which means the particle size of these PB-NMs is similar to traditional NMs.However, subgroup analysis shows that only three groups have the same significant difference as the total results, including MNMs-(10,100], CNMs-(10,100] and CNMs-[0,10].The MNMs-(100,1000) group has only one type of NMs (Au), and the data are quite different (248.61nm) to the traditional NMs.In addition, some blue lines reflecting the results of each NMs are striking.Lines of black carbon (BC), reduced GO (rGO) in the CNMs-(100, 1000) group and nanofilms in the CNMs-(10,100) group are obviously skewed to the left, but that of BC in the CNMs-(10,100) group is skewed to the right.Figure S2 shows the respective conditions of the literature used in the meta-analysis process of Figure 1. Figure S2(a) shows that there is further literature about MNMs with similar data trends, while there are fewer on CNMs with differences in data trends.This corresponds to Figure 1a and also is the reason that data of CNMs presents a greater bias than that of MNMs in Figure S4a and b. Figure S2b shows that some literature has large errors, especially in the size-(100,1000) group, while Figure S4b-d illustrates that the analysis results of Figure 1 are accurate but biased.
Figure 5 presents a meta-analysis of 45 raw materials for PB-NMs (including secondary plant products such as bagasse) provided by 45 research papers, indicating that the practicability of PB-NMs is better than that of non-PB-NMs with a significant difference (p = 0.03).Table S1 shows the scoring system and an example of this meta-analysis.Figure S5 shows that the meta-analysis of Figure 5 is accurate and free of bias, but there is still small heterogeneity, which means 1 point is outside the confidence interval.

Effects of NMs on plants
Figure 2 shows the effects of different types or size of NMs on plant growth indicators and counts the data of 702 samples from 27 studies between plants and nanomaterials.Figure 2a indicates that plant growth indicators of the control group are significantly higher than those of the NMs group (p < 0.01).The same can be said for the MNMs group which also shows a significant difference (p = 0.01) compared to the control group.100% of MNMs groups center points (i.e., mean of the group) are on the left.However, the red diamond is located on the right side of the vertical axis in the CNMs group, indicating that growth indicators of the control group are significantly lower than that of the NMs group (p < 0.01).40% of CNMs groups center points are on the left side and the rest are on the right side.Furthermore, there are two correspondences that provide a large number of samples, namely Cu-alfalfa (180) and , which can lead to a certain bias in the overall results.Ag and few-layer graphene (FLG) have the most obvious inhibitory effects on plant growth (rice) among MNMs and CNMs, respectively.The perfect funnel plot shown in Figure S6(a) and (b) illustrates that the meta-analysis process of Figure 2a is accurate, unbiased and non-heterogeneous.
Figure 2b shows that there are significant differences in the effects of different NM particle sizes on the plant growth index (p < 0.01).When the particle size is less than 20 nm, CNMs significantly promote plant growth (p < 0.01).When the particle size is 20-50 nm, CNMs promote plant growth but not significantly (p = 0.81).In the two intervals of particle size ≥50 nm, CNMs inhibited plant growth but not significantly (p = 0.81 and 0.17, respectively).MNMs of all particle sizes inhibited plant growth, but not significantly.The bias analysis of Figure S6(c) shows that the data of Figure 2b has a small bias, but the overall credibility is high.
Figure 3 is the result of further integration of analyzed data, while the original analysis results can be viewed in Figure S3.Table S2 shows the scoring system and two examples of this meta-analysis.111 samples are used in this analysis, indicating that there is no significant difference between the NMs group and control group (p = 1.00).However, there are differences in the data of the two groups in the subgroup analysis.For MNMs groups, the Control group is better than the NMs group, while the CNMs group has the opposite results, but neither have a significant difference (p = 0.83 and 1.0, respectively).The blue lines of the MNMs group are generally longer than those of the CNMs group, indicating that the data of the same MNMs have a larger standard deviation than CNMs. Figure S3 shows that the data from each piece of each literature has a similar confidence interval range whether they be CNMs or MNMs. Figure S7 shows that the data sample size used in this analysis is small and the variance is large.However, the data are all within the confidence interval so there is no bias or heterogeneity.

Production stability and practicality of PB-NMs
As a new NMs production method, the preparation of PB-NMs is still in its preliminary stages.Table 1 shows the common methods for the production of PB-NMs.In the material selection process, different parts of the plant can be used to prepare NMs; from seed husks (Tkachenko et al., 2020) to flowers (Piruthiviraj et al., 2015) to the whole plant (Haj Bloukh et al., 2021) as shown in Table 1.This obviously depends mainly on the species and economic value of the plant.The large green flowers of Brassica oleracea (broccoli) occupy the main volume of the plant (Piruthiviraj et al., 2015), while the seed husks of Oryza sativa (rice) are numerous and lack economic value (Tkachenko et al., 2020).Table 1 also records the state of the plant materials from these examples before preparation, including dryness and moisture, but has no significant relationship with the preparation method.In the pre-processing procedures for plants, most research chooses to grind plants in a 20-100 mesh in order to obtain the extracts (Wang et al., 2022, Nethravathi et al., 2015, Lee et al., 2016).In the reaction process, the reduction reaction or heat treatment is the most important step.The reduction reaction after mixing of plant extracts and acids containing metal elements has become a key step in the preparation of most PB-MNMs, such as Ag and Au (Singh et al., 2018, Yang et al., 2021, Lee et al., 2016).Heating is usually used to promote decomposition and mixing of materials to improve efficiency, such as thermal decomposition for SiO 2 (Tkachenko et al., 2020) and thermal hydrolysis for carbon quantum dots (CQDs) (Du et al., 2014).Some PB-NMs need specialized preparation methods.In the preparation of ZnO, reduction and heating methods are combined (Anbuvannan et al., 2015, Khan et al., 2019).In the preparation of gliadin − phospholipid hybrid nanoparticles (GPH), antisolvent precipitation as a dedicated method is used (Chen et al., 2019a).Therefore, the preparation process of PB-NMs already has some well-used routes and is suitable for plants with different characteristics.
Plants are the most important reason for the differences in production, characteristics and application between PB-NMs and traditional NMs.In the production process, plant substances, such as saponins, phenols, and flavonoids are abundant, and promote reduction or thermal decomposition as reducing agents or stabilizing agents (Figure 6a) (Elumalai et al., 2015, Haj Bloukh et al., 2021).This can replace additional and dangerous catalysts, solvents or templates (Prasad et al., 2019).Therefore, plants reduce the cost and risk in the PB-NMs production process.
The characteristics of NMs also can be changed and controlled by plants.For chemical properties, plants can increase strength or quantity of key chemical bonds and functional groups in PB-NMs (Lv et al., 2021), such as Si-O-Si of SiO 2 NMs (Tkachenko et al., 2020) and 4 types of amino functional groups of Au NMs (Guo et al., 2016).Plants can also add new functional groups to NMs, such as surface hydrophilic groups and phospholipid groups, which give NMs the potential to be used for the preparation of hydrogels and increase the resistance to extreme pH and salt (Chen et al., 2019a, Kummala et al., 2020).
For physical properties, the morphology of PB-NMs is related to plant additives, so it can be changed by simply controlling the reaction kinetics and changing reaction parameters (Zohora et al., 2017).Different types of PB-NMs can be obtained by the selection of plant characteristics and species.On one hand, the morphology of PB-NMs can be easily changed by the adjustment of the quantity of plants.Therefore, PB-NMs can obtain uniform shapes and significantly smaller particle sizes than traditional NMs (Figures 1 and 4a), which can increase the nano-effects and reaction stability of nanomaterials (Kummala et al., 2020, Matinise et al., 2017).Smaller particle size means PB-NMs are more easily absorbed, utilized and digested by plants and rhizosphere microorganisms than traditional NMs (Jiang et al., 2020).Figure 2b also shows that the small size of CNMs can significantly improve the growth of plants.This advantage can be combined with the properties of plant components to reduce phytotoxicity, allowing PB-NMs to have the  potential to have broader applications as shown in Figure 4b and Table 1 (Kummala et al., 2020).The controllable particle size and uniform shapes of NMs also can simplify the chemical reaction process (Zhang et al., 2019).On the other hand, PB-NMs can obtain special characteristics with the selection of different plants.On the other hand, brittle plant components tend to increase porosity, and flexible plant components tend to increase strength.For example, the pores of PB-NMs prepared from shells of cottonseed or walnut are large and continuous with high specific surface area to increase adsorption capacity (Jiang et al., 2020, Cheng et al., 2017).However, the pores of PB-NMs prepared from cattails or wood pulp are small and compact with high structural strength to face the demand of extreme conditions (Hauser & Nowack, 2019, Guo et al., 2016, Jiang et al., 2020).These characteristics also can be unique.Walnuts endow CQDs with zigzag and armchair edges (Cheng et al., 2017), but this is not seen in CQDs prepared from other plants, such as Tamarindus indica L. (tamarind) and bamboo (Tade & Patil, 2020, Bano et al., 2018).These different structures specifically improve the adsorption or catalytic degradation ability of PB-NMs to pollutants (Jiang et al., 2020, Prasad et al., 2018).Meanwhile, components also can be attached to the surface to form a special chemical coating to achieve functionalization, including antioxidant and antibacterial functions (Jiménez Pérez et al., 2017).Therefore, plants increase controllability of NM production.Plants also change the physicochemical properties of NMs through different species and concentration.Plants improve the application potential of NMs.Table 1 demonstrates the promising application feasibility of PB-NMs.The efficient degradation of pollutants is an important application of PB-NMs.For example, ZnO can catalyze the photodegradation of toxic dyes (Elumalai et al., 2015, Varadavenkatesan et al., 2019).CQDs can be used in the adsorption of Hg due to their porous structure (Bano et al., 2018).The application potential of PB-NMs in the medical field is reflected in its higher antibacterial properties than traditional NMs (Haj Bloukh et al., 2021).The potential feasibility of rGO in treating cancer continues to expand the application scope of PB-NMs (Yang et al., 2021).In the fields of materials production and agriculture, PB-NMs can serve as good catalysts or additives to promote the synthesis of other NMs (Khan et al., 2019) or crop growth (Ehsan et al., 2022), respectively.The ability of PB-NMs to be used as raw materials for chromatography represents the potential of PB-NMs in the development of new instruments (Tkachenko et al., 2020).
Figure 5 visually evaluates the economy and practicality of PB-NMs.Firstly, locally grown and high-yielding plants as an excellent raw material can give PB-NMs an economic advantage.This not only includes the economics of the production process (Ehsan et al., 2022), but also drives the cultivation and systematic utilization of these plants, especially those that are used as cheap raw materials or ornamental plants, such as bamboo (Tade & Patil, 2020) and watercress (Gordiichuk et al., 2021).Secondly, environmentally friendly features of PB-NMs include no secondary pollution as some waste materials can be changed to PB-NMs, such as bagasse (Du et al., 2014), coconut shells (Narayanan et al., 2019) and walnut shells (Wang et al., 2022).Moreover, the production and application processes are generally less secondary pollution and sustainable (Lv et al., 2021).Collection and reuse of plants can promote local industrial transformation, employment and environmental protection (Ramakrishnan et al., 2020).However, plants as a raw material also have some disadvantages.The drawback of edible plants is also in the analysis of Figure 5.The current world-wide food shortage is destined to give a higher priority for plants such as wheat (Chen et al., 2019b) to be used as a food source and hinder the large-scale production of PB-NMs.Meanwhile, high prices of some plants also mean they can be unsuitable for the production of PB-NMs, such as ginseng (Jiménez Pérez et al., 2017) and strawberries (A et al., 2019).Therefore, the selection of plants should consider the actual market demand to achieve a multi-faceted balance, rather than blindly pursuing the high performance of NMs.There are no doubt that new and well-established production technology makes the application of PB-NMs feasible.

Contradictory effects of nanomaterials on plants
NMs have begun to be widely used in the chemical industry, agriculture, and environmental protection fields.This is due to their unparalleled advantages, including physicochemical properties, environmental stability, catalysts, and pollutant degradation compared with biochar, metal semiconductor materials and compost (Elumalai et al., 2015, Zohora et al., 2017, Ehsan et al., 2022).During this process, NMs enter the environment and transport between air, soil, water, and finally reach plants.Due to differences in species, availability and physicochemical properties, NMs may have promotory or inhibitory effects on plants (Lahiani et al., 2018, Cota-Ruiz et al., 2020), showing the dual nature of NMs in relation to many plant indicators (Figure 6b) (Agathokleous et al., 2019, Pandey et al., 2019).
Plant growth indicators are the most intuitive data.MNMs generally show inhibition of plants growth in Figure 2, even at low concentration such as 0.006 mg Ag/L (Wu et al., 2020).But CNMs tend to show more promotory effects than MNMs (Lee et al., 2018), especially at low concentrations and low size (Chen et al., 2021).Figure 2b and Table 2 shows the effects of NMs in 0-20 nm (small size) and NMs in 20-100 (large size) on plants.Small CNMs can significantly (p < 0.01) promote plants growth as shown in Figure 2b, but MNMs show inhibition.At the same concentration, small size CNMs can increase the growth of Catharanthus L. by 42.58% and cotton by 27.93% compared to large particle size CNMs (Pandey et al., 2019).The growth of Amaranth is increased by 95.00% because small-sized CNMs can promote growth, while large-sized CNMs inhibit growth.MNMs can increase the growth of rice by 17.00% (Bai et al., 2021).Small size NMs can also promote plant germination rate, elongation and photosynthesis, even in long-term process (Lahiani et al., 2018).Small particle size NMs can also improve stress resistance and nitrogen fixation.This is due to the improvement of the rhizosphere environment by NMs, including promoting root growth, nutrient flow and rhizosphere microbial growth (Pandey et al., 2019, Wang et al., 2020b).Therefore, low NMs particle size with low phytotoxicity can promote plant growth and stress resistance.PB-NMs exhibit a significantly lower particle size than traditional NMs (Figure 1), so PB-NMs have the potential to become excellent low-phytotoxic materials.
This inconsistency may be due to the toxicity of metal elements (Velicogna et al. 2016).Figure 2 shows a contradictory relationship between plants and NMs, especially the most inhibitory effect being caused by Ag. Figure 2 also shows that even the same NMs have opposite effects for different plants, including MWCNTs, GO and ZnO.This may be related to specific plants, for example rice is a kind of plant usually affected by both MNMs and CNMs, while this also may be caused by different sources of NMs (Adeel et al., 2021, Jordan et al., 2020).
The large differences in indicators show that the effects of NMs on the inside of plants are special and unfixed.Figure 3 quantifies these effects of NMs on plants and demonstrates that they may be neutralizing.However, toxic and promotory effects are not counteracted.Specifically, these effects are reflected in different parts of the same plant as shown in Figure 6b.On one hand, NMs promote plant uptake of nutrients (such as C, N, S) (Ge et al., 2018) and resist salt stress (Gordienko et al., 2019).NMs also reduce the effects of some viruses by increasing plant resistance (Wang et al., 2014) or direct virus suppression (Adeel et al., 2021).On the other hand, NMs accumulate in plants and transport from roots to leaves, sometimes affecting plant photosynthesis (Bai et al., 2021, Sun et al., 2020) and generally causing antioxidant responses (Cota-Ruiz et al., 2020, Guha et al., 2018), even some NMs directly affect cell integrity and mitosis, such as SWCNTs and C 60 (Adeel et al., 2021, González-Grandío et al., 2021).Inhibition and promotion can also occur in the same plant by the same NMs, but not at the same time (Wang et al., 2022).This indicates that the effects of NMs on plants may also be due to demand or resistance in different growth stages (Agathokleous et al., 2019).However, all NMs induce antioxidant responses in plants with the increase of ROS or antioxidase.This shows that NMs disrupt the normal life activities of plants, whether they actually promote or inhibit plant growth (Guha et al., 2018).In addition, NMs generally alter microbial community structure (Ge et al., 2018, Zhang et al., 2020b), microbial activity (Chen et al., 2024) and gene relative abundance (Chen et al., 2019b) in the rhizosphere, and whether this has a potential impact on plant growth or not still deserves attention.The surface functional groups of traditional NMs are one of the possible phytotoxic causes, especially common carboxyl and polyethyleneimine (PEI) (González-Grandío et al., 2021, McGehee et al., 2017).Therefore, properties of NMs and plants remain factors that cannot be ignored when NMs are used in the environment.

Potential relationships and cycles between PB-NMs and plants
The effects of traditional NMs on plants are complex but traceable.Meanwhile, the recycling of traditional NMs is still energy-intensive and has low recovery rates (Hansen et al., 2022).However, PB-NMs, as a new type of material with the potential to replace traditional NMs for large-scale application in agriculture and industry as shown in Table 1 (Chen et al., 2019a), have not yet been studied in terms of their entry into the ecological cycle and their effects on plants.Some traditional CNMs can already provide carbon to plants (Ge et al., 2018).Figure 6a shows the effect of plants on the functional groups of PB-NMs and surface modification, on the contrary, Figure 6b shows the effect of metal element/carbon structure and functional groups of traditional NMs on plants.Obviously, this combines with the results of Figures 1 and 5 can show that the addition of plants did change the composition, structure and functional groups of PB-NMs.Thus, a reasonable guess is that PB-NMs are more easily absorbed and utilized by plants than traditional NMs due to small particle size by Figures 1 and 2 with Table 2, further promoting plant growth and minimizing phytotoxicity.In addition, there is also a complex relationship between NMs and the rhizosphere system of plants (Wang et al., 2020a), which is a possible direction to explore the impact of PNs on plants.
Both plants and NMs have been used for pollutants uptake and degradation (Ren et al., 2022, Tang et al., 2019).Plants can promote PB-NMs have better absorption and degradation function compared with traditional NMs as shown in Figure 6a, which means they have potential to replace expensive traditional NMs and inefficient biochar for pollutants treatment (Zand et al., 2020, Gonçalves et al., 2018).Hence, a recycle and reuse system composed of "Plants + PB-NMs" may be able to realize the secondary and efficient utilization of plants.In detail, plants can be produced as PB-NMs and reused after pollution adsorption.This may be able to minimize the changes in volume, physical and chemical properties brought about by the recycling of NMs, so as to be more in line with the needs of circular economy (Hansen et al., 2022).

Limitation and further applications of meta-analysis
In this review, meta-analysis can quantify and clearly demonstrate the practicability of PB-NMs and phytotoxicity of NMs.Further sub-group analysis shows individual effects of different types of NMs.Summary data, grouped data and sample data can all be visualized in one figure (Figures 1, 2, 3, and 5 and Figures S2 and S3) (Wang et al., 2020a).The error and distribution of data from different literatures can also be revealed through the bias analysis of meta-analysis (Figures S4-S7).Therefore, instead of vague generalizations, precise data can appear in this review to answer scientific questions.This also reduces reading, classification, comparison and comprehension difficulties associated with common cumbersome tables.
However, there are still limitations in these meta-analyses.The main limitation of this meta-analysis is the selection of samples.NMs samples that have been analyzed are limited to simple types, such as single Ag NMs and GO.Because many complex combined NMs are difficult to classify, they have not been included, such as Ag/ZnO NMs (Ehsan et al., 2022).These types of NMs usually contain more than one carbon and metal element with multiple structures, exhibiting properties opposite to raw NMs (Arumugam et al., 2021).Therefore, these meta-analysis results can only include the effects of common simple NMs and cannot represent novel complex NMs.Table 3 also reflects the similar situation that samples lack generalizability or representativeness.
Another limitation is the lack of standards and tools for the meta-analysis of plants.This review constructs some simple but limited scoring systems (Figures 3 and 5 and Table S1-2).This simple treatment may not accurately reflect all characteristics and toxicity of NMs.For example, the risk of damage to plant DNA by NMs may be greater than the root length indicator, but the two effects have the same weight in the scoring system.
Thus, there are uncertainties when the results of this meta-analysis are extrapolated to the typical levels of PB-NMs application and NMs phytotoxicity due to the increase of new NMs and the simplicity of the scoring system (Chen et al., 2024).
Recently, standardization and systematization of data and tools are often given as expectations in these related fields (Book & Backhaus, 2022, Huang et al., 2019).Therefore, this review expects scientific scoring criteria, models and analyzing tools, such as machine learning (Dang et al., 2022), in non-medical fields, such as environmental protection, biomaterials, and chemical industry.This can improve the in-depth, systematic, intuitive and scientific understanding of research progress.

Conclusion
This review used a scientific analysis method called meta-analysis to discuss and go toward answering two scientific issues, which are "Do PB-NMs have advantages over traditional NMs in terms of stability and practicability?" and "Do NMs have promoting or inhibiting effects on plants?".
PB-NMs show significantly smaller particle sizes than traditional NMs with a small bias, this characteristic makes PN-NMs less phytotoxic.Cheap plants as raw materials make PB-NMs more suitable for large-scale applications because they can reduce the production cost of NMs, promote the development of local planting industry, reduce secondary pollution and finally achieve a circular economy.By affecting particle size, surface and structure, plant components can also improve catalytic efficiency, physicochemical properties and pollution remediation ability of PB-NMs.
The effects of NMs on plants are both promotory and inhibitory.The surface functional groups and metal elements of traditional NMs are phytotoxic, which can affect plants at all time points from germination to fruiting and all levels from physiology to genes.However, some NMs, especially small particle size CNMs, can promote plant growth and the ability to resist external stressors.This size range is as same as common PB-NMs size range.Therefore, these results show the advantages and potentiality of PB-NMs, the complexity of the relationship between NMs and plants, and the feasibility of meta-analysis for researchers in the field of environment and materials.We expect to see more studies exploring differences between PB-NMs and traditional NMs, such as the catalytic function in pollution degradation.We also expect to see the research of relationships between PB-NMs and plants, such as the phytotoxicity of PB-NMs.PB-NMs may have lower phytotoxicity to plants, which is more suitable for large-scale and high-concentration applications in agriculture and industry.Furthermore, it is anticipated that this review will serve as inspiration and promotion for the application of meta-analysis in the fields of environment and materials.

Figure 2 .
Figure 2. Quantitative analysis of plant growth indexes affected by different (a) types of nms; (b) sizes of nms.

Figure 3 .
Figure 3. Quality analysis of nanomaterial toxicity effects on plants.

Figure 4 .
Figure 4. differences between PB-mnms and PB-Cnms in (a) shape of particles studied and (b) potential applications of these particles.

Figure 5 .
Figure 5. Quality analysis of PB-nms and non-PB-nms in practicality.
is the mixture of ag and Zno.2 assistance means PB-nms can become raw materials for chromatography and catalysts in material synthesis.

Figure 6 .
Figure 6.(a) enhancement effect of plants on PB-nms; (b) effects of traditional nms on plants.

Table 1 .
examples of PB-nms plant materials and applications.

Table 2 .
examples about promotion or protection on plants by small size nms.

Table 3 .
examples of meta-analysis limitation in fields similar to this review.8 heavy metals include Cd, Cr, hg, Pb, as, Cu, Zn and ni. 1