Introduction

Natural fibers attract a lot of attention in research and industry, due to their biodegradability, biocompatibility, versatility and ubiquity (Seki et al. 2022). Apart from the general cellulosic fibers extracted from various plants utilized as textiles, artwork, thermoplastic reinforcement composites (Sanjay et al. 2019), or other value-added products (Nascimento et al. 2021; Rajeshkumar et al. 2021; Elseify et al. 2022; Sanjay et al. 2022), cellulosic foams with three-dimensional hierarchical structures have been increasingly appreciated by researchers recently. For example, cellulosic sponges/aerogels were widely applied in oil/water separation (Thilagavathi et al. 2018; Datiles et al. 2021; Chen et al. 2023a), nanogenerators (Zhang et al. 2021; Ram et al. 2021), dye adsorption (Fauziyah et al. 2019; Xia et al. 2020a), thermal insulation (Sen et al. 2022), packaging (Miranda-Valdez et al. 2023), biomedical applications (Zampieri et al. 2006; Seddiqi et al. 2021), and as carbon source/template for electrochemical devices (Sun et al. 2019). Based on recent reviews (Budtova 2019; Rahmanian et al. 2021; Zaman et al. 2020), cellulosic foams can be categorized into two groups: synthetic cellulose aerogels and natural cellulose foams. The solubility and processability of raw natural cellulose derived from biomass resources pose inherent limitations in the production of synthetic cellulose aerogels. Additional treatments are often required prior to the freeze-/supercritical-drying of the cellulose composites (Vinod et al. 2020). These treatments include extraction processes (e.g. high-frequency ultrasonic crushing, mechanical stirring, high-pressure homogenization, TEMPO oxidation, acid hydrolysis, or bacterial secretion) (Manian et al. 2021; Mokhena and John 2020), or regenerated processes in various solvents (e.g. ionic liquids, N-methylmorpholine-N-oxide (NMMO), LiCl/DMAc systems, or alkali/urea aqueous solutions). Consequently, substantial amounts of energy and chemicals are required to process raw natural cellulose into synthetic cellulose aerogel. In terms of cost and environmental impact, it is therefore more attractive to consider directly utilizing natural cellulose foams. Until now, sources of reported cellulosic foams have been limited to the plant components such as phloem (hemp), xylem (balsa wood), vascular vessels (sugarcane bagasse), fruits (luffa cylindrica) and seeds (kapok or poplar catkins) (Pennells et al. 2020). Additionally, foam structures derived from these sources are primarily tubular in nature (e.g., hemp, cotton, kapok, poplar catkins, and coir) or honeycomb-like, as observed in balsa wood or cork. However, there are limited reports on naturally occurring cellulose foams that possess regularly interconnected networks (Bismarck et al. 2002; Chen et al. 2013; Cruz et al. 2013; Shen et al. 2013; Zhang et al. 2020a). Therefore, there is an urgent and compelling need to explore additional sources of natural cellulose foams that are cost-effective and exhibit high-performance. It is also crucial to discover novel foam structures (Li et al. 2021). Hence, the study of plant tissue with unique structures, such as aerenchyma, holds significant potential and garners considerable interest.

Aerenchyma tissue is considered to be the most striking morphological feature within plants. It is commonly found in aquatic and wetland species, assisting with their floating and internal air circulation (Blossfeld et al. 2011). This unique foamy tissue, which serves to retain air, is predominantly composed of aerenchymatic channels that possess large intercellular spaces, contributing to the high porosity of the aerenchyma. The arrangement of aerenchymatic channels exhibits distinct patterns in various plant species, including the radial arrangement in Ceratophyllum demersum, the perpendicular arrangement in Myriophyllum aquaticum, the honeycomb arrangement in water hyacinth, and the stellate network in Juncus effusus L. (Justin and Armstrong 1987). These aquatic plants show a high reproductive ability and possess sophisticated aerenchyma tissue, which can potentially be applied for water purification (Wolverton and Mcdonald 1979; Keskinkan et al. 2004; Xia et al. 2020a; Guo et al. 2020). On the other hand, they are also assessed as hard-to-handle weeds—controlling their quantities costs millions of dollars annually (Wolverton and Mcdonald 1979; Kaczmarek-derda et al. 2014). Thus, utilizing aerenchyma tissue as a natural cellulose foam resource offers a practical and valuable solution that addresses three interconnected problems. Firstly, it enables the utilization of abundant prolific wetland/aquatic pests. Secondly, it contributes to the promotion of biodiversity by exploring novel natural foam sources. Lastly, it leads to the discovery of natural foam structures. However, turning these weeds into an added-value product requires an in-depth study of their characteristics. Here, the aerenchyma tissue of Juncus effusus L. with the extraordinary stellate network was selected and characterized for its further potential foam applications such as oil adsorption, water purification, nanogenerator (Chen et al. 2023b) or foam structural templates.

Juncus effusus L. is a wetland perennial monocot, distributed widely throughout warm temperature zones in Eurasia and the American continents as one of the dominant species in multiple altitude ranges (Tweed and Woodhead 1946). Due to its high seed production and faster regrowth rate outcompeting neighboring species, Juncus effusus L. has the annual production of ~ 10 kg ash-free dry mass (AFDM) per m2 (Ervin and Wetzel 2002). Consequently, it is considered as a problematic weed in pastures and meadows as it impairs the forage yield and quality (Lazenby 1955; Kaczmarek-derda et al. 2014). According to the literature, the epidermis of their stem was usually utilized in weave while their pith was historically applied as wicks in train-oil lamps across north-western Europe and China (Kaczmarek-derda et al. 2018), or as traditional Chinese medicine (Noguchi et al. 2017). Recently, it was utilized in solar-driven steam generators (Zhang et al. 2020b) and soluble pollutant removals (Xia et al. 2020a). Nevertheless, comprehensive characterization studies of the pith aerenchyma foam in Juncus effusus L. (JEF) and the separation of its structural stellate unit have not yet been reported. This work aims to assess the potential usage of Juncus effusus L. pith aerenchyma as 3-dimensional flexible cellulosic foams, and to provide a new structure template for lightweight highly porous materials.

The structural, physico-chemical, mechanical and thermal characteristics of delignified stellate units of JEF and raw JEF were comprehensively analyzed. The properties of JEF were evaluated via single fiber tensile tests, cyclic compression tests, N2 adsorption–desorption experiments, liquid immersion tests, water contact angle measurements (WCA), chemical analysis, Scanning Electron Microscope (SEM) observations, X-ray diffraction (XRD), Fourier transform infrared spectroscopy (FTIR) and Thermogravimetric analysis (TGA). The properties were compared with other natural lignocellulosic foams.

Materials and methods

Fibers obtainment

Juncus effusus L. stems were collected from the Sprikkenburg riverside (53° 14′ N, 6° 31′ E) as shown in Fig. 1a within the university complex in Groningen. The stems were cut to 50 cm in length and their outer green epidermis was subsequently peeled off manually with a tweezer (Fig. 1c). The obtained JE aerenchyma foam (JEF in Fig. 1c) was kept in an air oven at 105 ℃ overnight to remove the moisture for further characterizations.

Fig. 1
figure 1

a Juncus effuses L., b old (left) and young (right) JE stems, c JEF extracted from young JE stems, d schematic illustration of JEF unit’s separation process

Characterization methods

Morphological characterization

The collected JE stems were divided into young and old groups (Fig. 1b) according to the color of the stem and the size of central lacunas. The morphological characteristics of these two kinds of stems were investigated by scanning electron microscopy (SEM) via a FEI Nova NanoSEM 650 with an accelerating voltage of 15 kV. Samples were coated with 10 nm Au in a Cressington Sputter Coater 208HR before imaging. The JEF diameter (D) and JEF units’ parameters in Table 2 were calculated from SEM images using the software ImageJ 1.53a. A total of 5 samples were measured in 20 different spots. The JEF stellate units’ cellulose crystals were studied by polarized optical microscopy (POM). A Zeiss Axiophot polarizing microscope equipped with a Sony DICC-500 camera for image acquisition was used and the images of bright field and POM were all recorded by Zeiss KS3000 software.

Density and porosity measurements

The solid density (\({\rho }_{s}\)) was obtained by liquid immersion test using a pycnometer (Blaubrand, Gay-Lussac) with toluene at 20 °C (Siva et al. 2020). The \({\rho }_{s}\) (g/cm3) of JEF was calculated via Eq. (1) (Lee et al. 1983):

$$\rho_s = \frac{{\left( {m_2 - m_1 } \right)}}{{\left( {m_3 - m_1 } \right)\left( {m_4 - m_2 } \right)}}\rho_t$$
(1)

where, \({\rho }_{t}\) is the density of toluene at 20 °C (0.867 g/cm3), \({m}_{1}\) represents the mass of the empty pycnometer (g), \({m}_{2}\) is the mass of pycnometer with JEF (g), \({m}_{3}\) is the mass of pycnometer filled with toluene (g) and \({m}_{4}\) is the mass of a pycnometer with JEF and toluene (g).

The bulk density (\({\rho }_{b}\)) of JEF (g/cm3) was determined via Eq. (2):

$$\rho_b = \frac{m}{V}$$
(2)

where \(m\) is the mass of JEF (g) and \(V\) is the calculated volume of JEF while regarding them as the cylindrical shape.

With the obtained values of \({\rho }_{s}\) and \({\rho }_{b}\), the porosity P of JEF could be estimated via Eq. (3):

$$P = \left( {1 - \frac{\rho_b }{{\rho_s }}} \right) \times 100$$
(3)

The obtained density and porosity results are listed in Table 1.

Table 1 Comparison of the physico-chemical properties of JEF with other reported natural foams and fibers

Chemical composition analysis

The carbohydrate composition analysis was conducted according to the method provided by the National Renewable Energy Laboratory (NREL/TP-510-42618) (Sluiter et al. 2012). The wax content was quantified referring to the standard protocol (Conrad 1944). The obtained data were compared with other reported natural foams and fibers in Table 1 (Anjos et al. 2008; Bismarck et al. 2002; Borrega et al. 2015; Chaiarrekij et al. 2011; Chen et al. 2013; Cruz et al. 2013; Luz et al. 2017; Dhakal et al. 2007; Driemeier et al. 2011; Gowthaman et al. 2018; Guimarães et al. 2009; Guna et al. 2017; Prakash et al. 2018; Hsieh et al. 1996; Leite and Pereira 2017; Likon et al. 2013; Lim and Huang 2007; Mahmood et al. 2005; Meng et al. 2021; Narayanasamy et al. 2020; Pereira 2015; Saeed and Iqbal 2013; Shen et al. 2013; Singh et al. 2013; Xie et al. 2020; Zhang et al. 2020a, 2017; Zheng et al. 2016).

Delignification treatment

The JEF units were separated via a previously reported delignification method (Sun et al. 2020). As shown in Fig. 1d, g JEF was submerged into a mixture of 50 mL of acetic acid (≥ 99.7%, Sigma-Aldrich) and 50 mL of hydrogen peroxide (30%, Sigma-Aldrich) at 85 °C overnight. Subsequently, the separated JEF units were dialyzed in deionized water with a dialysis membrane (MWCO 1kD, Spectrum Laboratories, Inc.) for 1 week to remove the chemicals. Finally, the JEF units were naturally dried at ambient temperature.

Fourier transform infrared spectroscopic analysis

Attenuated Total Reflection-Fourier Transform Infrared (ATR-FTIR) spectra were recorded on a Bruker VERTEX 70 spectrometer equipped with a Platinum-ATR diamond single reflection unit in the range of 4000–500 cm−1. Each spectrum was the average of 16 scans collected with a resolution of 4 cm−1. Atmospheric compensation and baseline correction of all collected spectra were applied via OPUS spectroscopy software (Bruker Optics v7.0).

Thermogravimetric analysis

Thermogravimetric analysis (TGA) and DTG were performed using a TA-instruments Discovery TGA 5500. Samples of approximately 5 mg were heated in a nitrogen environment from room temperature to 800 ℃ (10 ℃/min). Results were collected and analyzed via the software TA instruments Trios v4.3.1.39215.

X-ray diffraction analysis

X-ray diffraction (XRD) spectra were acquired at room temperature using a Bruker D8 Advance diffractometer (Cu Kα radiation, λ = 0.1542 nm) with a 10.5° divergent slit and a 10.5° anti-scattering slit. The pattern was recorded for 2θ ranging from 10° to 80° in steps of 0.01° and a counting time of 1 s per step, with an intensity of 40 mA and a voltage of 40 kV. For all XRD measurements, approximately 30 mg of JEF samples or isolated JEF units were compressed into a pellet under a dry state using a stress of 12.75 MPa. The environmental background data was also collected under the same measurement conditions with the blank sample holder, which was then subtracted from the data of all experimental samples (Hermans and Weidinger 1948; French 2020). Subsequently, the obtained experimental XRD spectra were fitted with the cellulose Iβ pattern (French 2014) in the software MAUD (Nascimento et al. 2021; Saville et al. 2021). Experimental data of cotton (Kruidvat cosmetic cotton puff) was collected under the same measurement conditions for comparing the crystallinity of the samples.

The crystallinity index (CrI) was calculated according to the empirical Eq. (4) (Segal et al. 1959).

$$CrI = \frac{{I_{200} - I_{am} }}{{I_{200} }} \times 100$$
(4)

where \({I}_{200}\) represents the maximum intensity at 2θ of 22° referring to the cellulose crystalline content, and \({I}_{am}\) is the minimum intensity at 2θ of 18.5° referring to the amorphous non-crystalline content of cellulose.

The crystallite size (CS) was calculated by using Scherrer’s equation as Eq. (5). (Siva et al. 2020)

$$CS = \frac{K\lambda }{{\beta \cos \theta }}$$
(5)

where K is the Scherrer’s constant (0.9), λ is the wavelength of the X-ray source (0.1542 nm), β is the peak’s full-width half maximum (FWHM), and θ is the corresponding Bragg angle.

BET surface area analysis

The specific surface area and pore size were detected with a nitrogen adsorption–desorption experiment at 77 K using a Micromeritics ASAP 2420 V2.05 (V2.05 J) porosimeter. Prior to analysis, ~ 1 g JEF samples was degassed at 105 °C for 12 h under vacuum.

Water contact angle test

The surface wettability of JEF was evaluated by water contact angle (WCA) measurements, using a Dataphysics OCA 25 system. Milli-Q water droplets of 2 μL were deposited directly on the samples automatically, where a camera recorded the pictures and the contact angles were calculated using the Young–Laplace (YL) equation. The WCA was obtained with three tests per sample.

Tensile and compressibility test

The tensile tests of JEF were conducted using an Instron 5565 0.1kN Series IX tensile tester with a 100 N load cell. Prior to the test, the ends of samples were wrapped with tape to prevent the damage of jaws’ clamping. Tests were conducted at various gauge lengths of 10 mm, 20 mm, 30 mm, 40 mm and 50 mm with 10 samples for each gauge length.

The compressive properties of JEF were evaluated with a cyclic compressive method by an Instron 5565 0.1kN Series IX tensile tester. Prior to the measurements, the samples were cut and fitted into a cylindrical Teflon tube (φ 15 mm, h 30 mm). Supported by the tube, JEF were placed between the testing plates and compressed with a speed of 1 mm/min to 30% and 60% of their original thicknesses. All tensile and compressive measurements were conducted at room temperature (20 °C) and at a relative humidity of 60%, and all results were collected via the software Merlin V5.50.00.

Results and discussion

Morphologies of JE stems

The morphologies of Juncus effusus L. (JE) stems with two representative aging degrees were studied through its epidermis and aerenchyma by scanning electron microscopy (SEM)—see Fig. 2. Since JE is a wetland monocotyledon, its epidermis of stem works as primary photosynthetic carbon-fixing organs instead of the absent leaves (Stebbins and Khush 1961; Hinchliff and Roalson 2009). As a result, its epidermis contains not only the strong sclerenchymatous fibers (Fig. 2c) for mechanical support (Dou et al. 2019; Lo et al. 2004), but also the vascular bundles (Fig. 2a, b) for water, salts and organics transportation within the plant (Eltahir and AbuEReish 2010). It can clearly be seen from Fig. 2a that the vascular bundles are embedded in parenchyma cells, with sclerenchymatous fibers surrounding the outermost part of the stems. As the maturity of the JE stems increased, the size of scalariform vessels within the vascular bundles was enlarged by approximately 70% (27–46 μm), which led to more voids.

Fig. 2
figure 2

SEM images of transverse sections of a a young and b an old JE stem; SEM images of longitudinal sections of c young and d old JE stems. The green rectangles in the upper images are shown in higher magnification in their corresponding bottom images. Black and white scale bars here represent 500 μm and 20 μm, respectively

In addition, due to the special aquatic habitat of Juncus effusus L., their aerenchyma tissue (auriferous parenchyma) was well developed to ensure a continuous oxygen supply in the oxygen-deficient high humidity soils (Blossfeld et al. 2011). Aerenchyma is regarded as a modified parenchyma, which can always be found in the soft parts of aquatic plants such as lotus or water hyacinth (Guna et al. 2017). As for Juncus effusus L., apart from the epidermis, its stem is filled with aerenchymatous cells that have a visual resemblance to cotton wool (Fig. 2c). This spongy culm exhibits extensive inter-connected multicellular structures, which contain numerous small regularly shaped air chambers along both transverse and longitudinal directions for gas diffusion and buoyancy support. Obviously, with the increase of the growth time, the programmed death of aerenchymatous cells results in a large central lacuna in the JE stems (Fig. 2b). This structure of thickened vascular bundle fibers and central voids from old JE stems also showed great consistency with previously reported JE stem fibers (Maache et al. 2017).

According to the observation of JE stems with two representative aging degrees, the spongy JE aerenchyma foams (JEF) from young plump stems were selected for the further studies below.

Morphologies of JEF

JEF of young JE plump stems were characterized by SEM and optical microscopy. It can be observed that JEF is filled with large symmetrical air spaces (Fig. 3a), which make them appear snow-white color (Fig. 1c). This is comparable to aerenchyma tissue from water hyacinth (Salas-Ruiz et al. 2019). The three-dimensional JEF structure is assembled by large layers of similar cross-linked structures, and each layer is formed by regularly interconnecting the stellate unit (six-rayed stars), which can be seen in Fig. 3f from different directions. The coupling of different layers is achieved via the stellate units’ random vertically aligned rays. This hierarchically assembled structure also showed a great similarity with the reported extremely low-density super-compressible graphene foam (Qiu et al. 2017), which might be a hint at its potential as an elastomer in the future. Meanwhile the linkage at the end of each unit can be clearly distinguished from both SEM (Fig. 3d) and bright field optical images (Fig. 3e). Surprisingly, this stellate unit of JEF could be separated while reserving its original structure after the delignification process (Fig. 3c), which showed a great difference with reported shrunken foam structures after alkaline treatment (Xia et al. 2020b). As lignin molecules cross-link different plant poly-saccharides and cellulose, the cell wall’s mechanical strength of JEF is enhanced by this process (Chabannes et al. 2001). In contrast to the structure illustrated in Fig. 3b, the stellate unit cell wall showed less stiffness and was inclined to overlap after the delignification process (Fig. 3c), which could be advantageous for further modifications or various lightweight applications.

Fig. 3
figure 3

SEM images of a a JEF network, b a JEF stellate unit and d the binding parts of the units; Optical microscopy images of c delignified JEF stellate units under polarized light and e the JEF structure under bright field; f schematic structures of JEF interconnected network

Interestingly, JEF contained a similar wool-like open-cell foam structure consisting of hollow cylindrical channels as compared with the seed fibers from kapok, C. gigantea and poplar catkins (Likon et al. 2013; Zheng et al. 2016). By comparison, JEF stellate units had the narrowest outer diameter OD (7.91 ± 1.13 μm) and the shortest channel length L (76.66 ± 8.77 μm) among all reported hollow fibers. A more detailed comparison of morphological characteristics of JEF and other typical reported hollow tubular fibers was summarized in Tables 1 and 2. As shown in Table 2, the JEF units contained a similar channel wall thickness T (0.86 ± 0.18 μm) as that of C. gigantea fibers (0.72 μm), meanwhile it also shared a similar OD with poplar catkins fibers (8.7 ± 5.7 μm). In contrast, the OD (16.5 ± 2.4 μm) and T (2 ± 2.4 μm) of kapok fibers was twice and three times that of JEF units, respectively. On the basis of these hollow channels’ morphological differences, the resulting varied cavity ratios further differentiated the bulk densities and porosities of the materials as shown in Table 1. By Comparison, JEF possessed the lowest bulk density (0.017 g/cm3) and the highest porosity (98.08%) among all listed natural lignocellulosic open-cell foams, namely balsa wood (0.06 ~ 0.38 g/cm3 and 85.60%) (Borrega et al. 2015), kapok (0.384 g/cm3 and 91%) (Likon et al. 2013; Thilagavathi et al. 2018), C. gigantea (94%) (Thilagavathi et al. 2018), poplar catkins (0.36 g/cm3 and 89%) (Likon et al. 2013; Saeed and Iqbal 2013), and luffa cylindrica (0.025 ~ 0.065 g/cm3 and 79 ~ 93%) (Shen et al. 2013; Saeed and Iqbal 2013). Furthermore, the large amount of air chambers kept within these stacked or interconnected hollow fibers could prevent the entry of liquids with high surface tension (e.g. water), which resulted in their hydrophobic properties. Specifically, poplar catkins fibers showed the highest water contact angle (150°) among all listed natural fibers, and JEF displayed the second highest hydrophobicity (WCA of 147°). The hydrophobic hollow micro-channels of JEF also cause its strong capillary effect for non-polar solvents, which explains why ancient Asians and Europeans used them as wicks for oil lamps (Kaczmarek-derda et al. 2018; Noguchi et al. 2017). Moreover, as density, porosity, hydrophobicity and environmental friendliness play vital roles in oil-absorbing applications, the above-mentioned materials have been reported to exhibit excellent performance as oil absorbents (Datiles et al. 2021; Likon et al. 2013; Liu et al. 2021; Thilagavathi et al. 2018; Yang et al. 2020; Zhang et al. 2017; Zheng et al. 2016). In this case, based on JEF’s low density, high porosity and excellent hydrophobicity, it could also be regarded as potential material for adsorption applications or as a lightweight 3D structural template (Manimaran et al. 2018).

Table 2 Comparison of morphological parameters of JEF units with other typical hollow fibers (Likon et al. 2013; Zheng et al. 2016)

Physico-chemical properties of JEF

Chemical composition analysis

The chemical composition of natural fibers plays an important role in their mechanical and morphological properties, as well as their thermal stability. Therefore, the chemical composition of JEF was investigated. The obtained results were compared with other reported natural foams and fibers in Table 1. JE contained a similar cellulose content (39.67%) as balsa wood (42 ~ 45%) (Zhang et al. 2020a), poplar catkins (41 ~ 44%) (Likon et al. 2013) and coconut shell coir (32 ~ 43%) (Fiore et al. 2014), while a higher content of cellulose in wet plant water hyacinth aerenchymas can be observed (25.64%) (Guna et al. 2017). The hemicellulose and lignin content of JEF was 35.97% and 20.18%, respectively. This chemical composition ratio demonstrated high similarity to that of Napier grass fibers (cellulose 47.12%, hemicellulose 31.27%, lignin 21.63%) (Kommula et al. 2015). The content of hemicellulose is also similar to that of water hyacinth (31.81%). These considerably high levels of hemicellulose or lignin content can be commonly found in soft grass (Kommula et al. 2015), which necessitates both softness and flexural rigidity to withstand various environmental forces. Hemicellulose, with relatively low mechanical strength, provides the desired softness, while lignin, known for its high mechanical strength, imparts the necessary rigidity (Mohanty et al. 2000; Manimaran et al. 2018). A higher lignin content (20.18%) also helps the plant against biological attack (Manimaran et al. 2018). In addition, the wax content (2.52%) acts as an adhesive between JEF and epidermis in the stem (Jebadurai et al. 2019). At the same time, it serves to protect the stem in aquatic environments due to its hydrophobic nature.

Although the raw JEF foam comprises a mixture of cellulose, hemicellulose, lignin, and wax, each of these components plays a crucial role in the properties of the foam. Specifically, hemicellulose and lignin contribute to its structural integrity, as evidenced by the isolated stellate foam structural units obtained after the delignification process. The presence of wax in JEF also contributes to its high hydrophobicity, thereby expanding its potential applications. Furthermore, the hemicellulose-rich JEF holds great promise as a hemicellulose resource for diverse applications such as hydrogels, thermoplastics, coatings, drug carriers, and additives in food or papermaking (Cunha and Gandini 2010; Fredon et al. 2002; Gabrielii et al. 2000; Jain et al. 2001; Kommula et al. 2015). More importantly, the direct utilization of raw JEF foam allows us to leverage cellulose foam in its natural form, while capitalizing on its inherent excellent foam characteristics (as shown in Table 1). By adopting this approach, we can avoid additional energy/hazardous chemicals consumption associated with other processes to achieve the true environmental sustainability (from raw materials to manufacturing and final products) (Li et al. 2021). This approach in our study was rooted in the fundamental motivations of utilizing cellulose, with a strong emphasis on energy conservation, sustainability, the abundance of sources, versatility, and cost-effectiveness. These principles were pivotal in driving our efforts to accelerate the transition from the fossil fuel era to a sustainable future.

ATR-FTIR

FTIR spectra of the raw and delignified JEF units are shown in Fig. 4. The small band centered at 667 cm−1 is ascribed to the C–OH out-of-plane bending from cellulose (Pereira et al. 2011; Jaouadi et al. 2009). The band observed at 897 cm−1 corresponds to the presence of β-glycosidic bonds between the monosaccharides (Fiore et al. 2014), which may also indicate the existence of saline content in JEF, according to literature (Khan and Drzal 2004; Sreenivasan et al. 2011). The highest band of 1032 cm−1 refers to the alkoxy C–O bond bending vibration and O–H stretching of polysaccharides in cellulose (Prasad et al. 2010; Saravanakumar et al. 2013; Maepaac et al. 2015). The band at 1159 cm−1 represents the presence of strong and broad C–OH stretching of cellulose in JEF (Binoj et al. 2016). In addition, the band at 1315 cm−1 is attributed to the CH2 asymmetric stretch vibration in crystallized cellulose I (Colom et al. 2003), and the band at 1421 cm−1 is assigned to the CH2 bending vibration at –OCH3, representing both crystalline cellulose I and amorphous cellulose (Huang et al 2012). The detection of the band at 1602 cm−1 represents the C=C aromatic stretching with strong conjugated C–C bonds in the lignin content (Saravanakumar et al. 2013), while it was also reported to indicate the moisture from natural fibers (Karbowiak et al. 2011). Meanwhile, water trapped in lignin or cellulose can be observed at 1629 cm−1, which was usually reported to be present between 1620 cm−1 and 1640 cm−1 (Nascimento et al. 2021; Boukir et al. 2019). Similarly, the band at 1513 cm−1 also refers to the presence of lignin, which can be concluded from the vibrations of C=C groups (Sugiyama et al. 1991). The main bands associated with hemicellulose were expressed at 1242 cm−1, 1371 cm−1 and 1736 cm−1. Specifically, bands at 1242 cm−1 corresponds to the syringyl nuclei and C–O stretching in hemicellulose and lignin (Nascimento et al. 2021; Boukir et al. 2019), 1371 cm−1 is assigned to the CH3 symmetric bending vibration in hemicelluloses (H3C–(C = O)–O–) and cellulose (Boukir et al. 2019), and 1736 cm−1 is correlated to the C=O bonds stretching in acetoxy groups (H3C–(C=O)–O–) of hemicellulose (Porras et al. 2015; Nascimento et al. 2021; Boukir et al. 2019). The two small bands at 2854 cm−1 and 2920 cm−1 are ascribed to the CH2 symmetrical and asymmetrical stretching in methyl/methylene groups, respectively (Zghari et al. 2017; Nascimento et al. 2021; Boukir et al. 2019). These two bands (2854 cm−1 and 2920 cm−1) were also reported to represent the wax content (C≡C stretching) in natural fibers (Moshi et al. 2020). The last strong and broad band at 3331 cm−1 represents the hydroxyl and carboxylic O–H stretching vibrations in cellulose I (Åkerholm et al. 2004; Jayaramudu et al. 2010; Reddy and Yang 2005; Sugiyama et al. 1991). The observed band centers of natural fiber can vary around ± 16 cm−1 from their position for different studies (Zhbankov et al. 2002).

Fig. 4
figure 4

FTIR of JEF and JEF units

After the delignification treatment, the intensity of bands representing hemicellulose at 1242 cm−1, 1371 cm−1, and 1736 cm−1 were significantly reduced, while bands referring to lignin at the wave numbers of 1513 cm−1 and 1602 cm−1 completely disappeared. This clearly indicated the reduction of hemicellulose content and the successful removal of lignin.

Additionally, the degree of crystallinity of different samples was compared by the Lateral order index (LOI) value based on the ratio of the absorbance at 1428 and 893 cm−1 (Eq. 6) (Liu et al. 2004).

$$LOI = \frac{{a_{1428 \,{\text{cm}}^{ - 1} } }}{{a_{893 \;{\text{cm}}^{ - 1} } }}$$
(6)

Here the absorption bands at 1428 and 893 cm−1 are assigned to crystalline and amorphous structures of fibers (Korte 2006). The obtained LOI value was 0.32 and 0.52 for JEF and JEF units, respectively. By comparison, these values were lower than that of althaea fiber (0.79), linden (0.96), ferula fibers (0.70) and Hierochloe Odarata fiber (1.01) (Dalmis et al. 2020). However, it is important to note that the increased LOI of JEF units indicated a higher degree of crystallinity (Korte 2006), which is also consistent with the relatively high crystallinity of cellulose from stellate JEF units observed under polarized light (Fig. 3c).

X-ray diffraction analysis

The crystal structures of raw JEF samples and JEF units were evaluated with X-ray diffraction as shown in Fig. 5a. It can be seen that both diffractograms exhibited three obvious peaks around 16°, 22° and 35°, which corresponded to the crystalline planes of (1–10)/(110), (200) and (004), respectively (French 2014; Ling et al. 2019). Specifically, the peak observed around 16° was associated with the cellulose Iβ crystal pattern, which contained overlapping reflections of plane (1–10) and (110) (French 2014; Xia et al. 2020b; Nascimento et al. 2021). After being refined in the software MAUD with the cellulose Iβ crystal pattern (Ling et al. 2019; Saville et al. 2021), these two crystallographic peaks can be distinguished without overlapping in the Rietveld fitted data of JEF units (Fig. 5a). In addition, the absence of “shoulder” peaks nearby the peak (200) can be an indicator for the preferred crystallites orientation of JEF and JEF units cellulosic samples (French 2014, 2020; Ling et al. 2019).

Fig. 5
figure 5

a X-ray spectra of JEF, JEF units and the corresponding fitting with Rietveld refinement software MAUD; b X-ray pattern of cotton

The crystallinity index (CrI) of the JEF and delignified JEF were calculated according to the Segal empirical equation as shown in Eq. (4) (Segal et al. 1959), and the \({I}_{am}\) and \({I}_{200}\) was the intensity at 2θ of 18.5° and 22°, respectively. The calculated CrI value of raw JEF and JEF units was 42.5% and 65.1%, respectively. The higher crystallinity observed in JEF units (65.1%) compared to raw JEF indicated the effective partial removal of the amorphous hemicellulose, wax and lignin through the delignification treatment (Chen et al. 2023a). This finding also aligned with the LOI values and FTIR results discussed above (Hashim et al. 2014). Meanwhile the diffractograms peaks between JEF and JEF units shared great similarity, suggesting that the cellulose Iβ crystalline structure of JEF was well-preserved in JEF units during the delignification process (Chen et al. 2023a). As the higher cellulose crystallinity is often recognized as a contributing factor in improving important mechanical characteristics of fibers, including Young's modulus, tensile strength, and hardness (Sanjay et al. 2019), the influence of crystallinity on the stiffness of natural fibers cannot be overstated (Benítez and Walther 2017). As a result, the CrI of cotton was also measured and calculated for comparison here (Fig. 5b) since it is difficult and inaccurate to compare with the reported results obtained under different conditions (Ling et al. 2019; French 2020; Nascimento et al. 2021). Here the CrI of cotton (85.1%) was 2 and 1.3 times that of JEF and JEF units, respectively. The crystallinity of the cellulose fibers obtained from various biomass sources and chemical treatments could vary significantly, ranging from 40 to 91% (Jonoobi et al. 2015). Therefore, the crystallinity of JEF (42.5%) increased substantially from a lower end to a relatively higher end of this range after the deligninfication treatment (65.1%). Besides, the crystallite size (CS) was calculated via Eq. (5) (Scherrer’s equation) with the highest intensity peak at 22°. The CS value of raw JEF increased from 2.0 to 2.7 nm (JEF units) after the delignification treatment, rendering its high potential as promising reinforcements in polymeric matrix composites. Moreover, it has been reported that residual amounts of hemicellulose and lignin can potentially enhance the characteristics of resulting composites and even introduce new functionalities (Imani et al. 2019; Song et al. 2018). This suggests that the utilization of delignified JEF units can offer a more cost-competitive option, as they require reduced or less intensive processing, even though it may lower the material’s purity. It is important to note that the performance of cellulose materials is not solely dependent on high purity; instead, it varies based on the specific application (Li et al. 2021). This also highlights the significance of considering application-specific requirements when evaluating the suitability of cellulose materials.

Thermogravimetric analysis

To prevent the degradation of the foams in any potential applications, thermogravimetric analysis was conducted. As shown in Fig. 6a, the thermal degradation process of JEF was presented in four major stages. The first stage occurred in the range of 20–100 °C with a weight loss of 5%, which was regarded as the removal of moisture and wax (Belouadah et al. 2015). Subsequently, when the temperature rose from 160 to 310 °C, the degradation of hemicellulose, partial lignin and the glycosidic links of cellulose occurred as the second and third JEF weight loss (24%) stages with the maximum changing rate at 218.2 °C and 285.2 °C, respectively (Sgriccia and Hawley 2007; Seki et al. 2013; Azwa et al. 2013). Similar thermal degradation stages of hemicellulose (200 ~ 260 °C) and the glycosidic links of cellulose (240 ~ 350 °C) were reported elsewhere (Mohan et al. 2006). Further, the fourth DTG peak at 340.5 °C (weight loss 56%) indicated the direct thermal degradation of cellulose I and α-cellulose (Seki et al. 2013). A similar weight loss was also detected in various natural lignocellulosic fibers such as Lygeumspartum (338.7 °C), hemp (308.2 °C), Cissusquadrangularis root (328.9 °C) and napier grass (383 °C) (Dhakal et al. 2007; Kommula et al. 2015; Indran and Raj 2015; Belouadah et al. 2015). A residue percentage of 15% was produced at the ending temperature of 800 °C. By comparison, the sample of JEF units had one major weight drop maximizing at 336 °C, which was attributed to the degradation of α-cellulose (Fig. 6b), proving the partial removal of the amorphous hemicellulose and lignin. In addition, previous research has reported that cellulose fibers with lower crystallinity and smaller crystallite size tend to undergo accelerated degradation and exhibit reduced thermal stability (Poletto et al. 2012). Accordingly, the high thermal stability observed in JEF units can be explained by their higher crystallinity and larger crystallite size. In conclusion, thermogravimetric analysis suggests that the raw and delignified JEF were thermally stable below 160 °C and 220 °C, respectively. Therefore, it would be possible to modify the fibers with for instance silane coupling agents for different applications when the reaction was set under the corresponding temperatures (Binoj et al. 2016).

Fig. 6
figure 6

TG and DTG curve of a raw JEF and b JEF units

Tensile and compressive behavior of JEF

Table 3 shows the tensile strength and Young's modulus values of a single JEF with various gauge lengths (10 ~ 50 mm). The average tensile strength of JEF (15.2 ~ 46.6 kPa) were significantly lower than that of extracted fibers from Juncus effusus L. stem (113 ± 36 MPa) (Maache et al. 2017), which could be attributed to the vascular vessels and ordered strong sclerenchymatous fibers within the epidermis of the stem (Fig. 2c). Furthermore, the tensile strength of JEF was also lower than most of the natural fibers (≥ 20 MPa) (Indran and Raj 2015), and this low value could be explained by two aspects: chemical composition and microstructures. In terms of chemical composition, the varying contents of cellulose, hemicellulose and lignin in natural fibers have a great impact on their mechanical strength. Cellulose, as the strongest reinforcement material in natural fibers, has the most significant effect on the mechanical properties, followed by lignin (Sreenivasan et al. 2012). In contrast to this, the high hemicellulose content of JEF explains its lower mechanical strength (Sreenivasan et al. 2012). Additionally, from the structural perspective, natural fibers with larger porosity or increased cross-sectional area/diameter typically have lower tensile strength (Fidelis et al. 2013; Wen et al. 2004). This explains the relatively low tensile strength results of JEF as well. Meanwhile, JEF tensile strength values showed an obvious decreasing trend with increased gauge length, and this variability (typical in natural fibers) can be attributed to the different distributions of defects within JEF that are caused by the extraction process (Fidelis et al. 2013).

Table 3 Mechanical properties of JEF

Compressibility is a vital property for flexible foams. Here, raw JEF exhibited good compressiblity (Fig. 7), which is rare among natural lignocellulosic foams with high porosity (98.08%) (Qiu et al. 2017). As shown in the insets of Fig. 7a, b, JEF is almost fully reversible to its original volume after the first cycle in both two moderate strain uniaxial compressions (ε = 30%). The volume deformation ratio increased gradually along with the continuing compression cycle, then reached a final value of 12% and 7.5% under the force against the transverse section (direction T) and longitudinal section (direction L), respectively. When the strain was set to a larger value of 60%, JEF still sustained deformations by recovering most of its volume elastically (60% of direction T and 70% of direction L), and revolt the structural fatigue after 10 cycles. This capacity of reversible structural recovery is owed to the highly ordered three-dimensional interconnected configuration of JEF, which could prevent the breaking of linkages between the stellate units, while having the space for densifying at the same time (Gui et al. 2010). Both plateaus at lower ε and gradually rising slope at high ε (Fig. 7a, b) can be found in all hysteresis curves, indicating a substantial energy dissipation as a result of the friction between the air or the Teflon mold and the sponge skeletons. In addition, the higher deformation ratios from compression tests along the T direction at the same strain were observed (insets of Fig. 7a, b) indicating less resilient or supporting substances of JEF in these directions, which again corresponds to the stacked layer by layer parallel structure as explained above (Fig. 3f). Interestingly, JEF displayed better structural integrity with the compression force from direction L in accordance with the direction of common environmental stresses (e.g., wind, or animal activities) (Gardiner et al. 2016).

Fig. 7
figure 7

Stress-strains curves of JEF a at a maximum ε of 30% along the height of JEF (direction H). Inset, volume deformation ratios initiated via compression for 10 cycles at ε of 30% and 60%, respectively; b at a maximum ε of 60% along the diameter of JEF (direction D). Inset, volume deformation ratios developed via 10 cycles’ compression tests at ε of 30% and 60%, respectively; c the fully-recovered volume of JEF pellet in water

To further investigate the compressibility of the hierarchical structure, JEF was compressed to form a 100 μm thick pellet with 12.75 MPa stress in a dry state (Fig. 7c), in which its original volume shrunk by 99% during the densification. Surprisingly, this densified pellet would instantaneously swell to larger sizes while submerging it in water, and recover most of its thickness within 6 s. This behavior confirmed the interconnected three-dimensionally highly porous structure of JEF. This structure prevents the formation of strong van der Waals interactions even at a densified state, while allowing the absorbed water to refill the open pores very fast and to push back the lignocellulosic cell walls away to their original configurations (Qiu et al. 2017; Wen et al. 2004). Therefore, the structure of JEF holds a great potential as a template in 3D flexible materials’ applications.

Conclusion

In this work, the structural, physico-chemical, mechanical and thermal properties of stellate units of JEF and raw JEF were investigated in-depth. By comparing with other natural cellulosic foams, JEF displayed a distinctive hierarchical interconnected structure consisting of unique hollow stellate units. This unique structure was identified as the key primary factor contributing to JEF’s remarkable characteristics, such as its ultra-lightweight nature (0.017 g/cm3), high porosity (98.08%) and impressive compressibility. The exceptional properties exhibited by JEF offer the possibility for utilizing abundant and problematic aquatic or wetland weeds as alternative sources for versatile 3-dimensional cellulosic foams in a true environmentally sustainable way (from raw materials to manufacturing and final products). By introducing the overlooked aerenchyma plant tissue as a lignocellulosic fiber source and structural template, we have opened up possibilities for various potential applications, including but not limited to oil adsorption, water purification, nanogenerators, and foam structural templates.