Bio-inspired tubular hierarchical porous materials with selective liquids absorption

ABSTRACT Cacti have an excellent ability to capture and store water in the air because of their unique hierarchical porous structure. Inspired by this, we proposed a hierarchical emulsion-tube-lattice (ETL) material, comprised of the lattice structure, tubular struts, and porous walls, with superior oil–water separation and storage capabilities. A one-step coaxial direct ink writing process was developed, with a silicone emulsion serving as the typical ink. After that, the oil absorption performances of the sample were evaluated. The results showed that the ETL material possesses higher oil absorption and retention capacity, and could allow the water of the water–oil mixture to pass through while absorbing the oil. Besides, the sample could reuse after squeezing out the oil and keep the soft mechanical properties. In summary, with the bio-inspiring design, the sample showed promising application in water–oil separation and storage. Furthermore, the method could be easily extended to other materials.


Introduction
Water pollution has been one of the significant sources of environmental pollution these days.For wastewater resource cleaning and recycling, the demand for materials with efficient and low-cost water-oil separation and liquid storage performance is increasing.Porous materials are generally used for oil-water separation applications (Zhang and Seeger 2011), such as fabric (Cortese et al. 2014), sponge (Zhang et al. 2013;Jiang et al. 2013), and also aerogel (Zhang et al. 2022;Zhang et al. 2021;Zhang et al. 2021), due to their lightweight properties (Zeiger et al. 2017;Chen et al. 2019) and high absorption capabilities (Khosravi and Azizian 2016).
Some natural living organisms (Figure 1a), due to their unique lattice structure, provide different special characteristics such as water harvesting and absorption properties or high mechanical properties like strength and toughness, which can help to provide novel ideas for fabricating hierarchical water-oil separation materials (Li et al. 2019;Wegst et al. 2015;Xu et al. 2021).Figure 1a shows some natural living organisms with the properties mentioned above.The stems of the cactus (Zhang et al. 2017;Ju et al. 2014;Kim et al. 2017) have multiple branches, increasing the contact area with the air.There are also abundant cells in the stem used for storing water.Additionally, the surface of the cactus has evolved spines with the function of directing water droplets.The surface of a butterfly wing (Han et al. 2016) has laminar structures, which allow for the spreading of fog droplets.As a result of that, the fog droplets will evaporate rapidly.Therefore, the butterfly wings can be anti-fog.A human bone cross-section (Liu, Luo, and Wang 2016) has a hierarchical structure, giving it high strength and toughness.Inspired by an insect's compound eye (Huang et al. 2019), microwave absorption materials could be prepared with a hierarchical honeycomb structure to obtain broad bandwidth and hydrophobic properties.
Inspired by nature materials, we proposed a water-oil separation material with a three-level structure, each with a different function (Figure 1b).The first level is the Marco lattice structure.It will increase the contact area with liquids.The second level is the tubular struts This is used to store the separated liquid.The third level is the porous wall, which selectively allows liquid to pass through.
Herein, different processes were used to form different scales of structures.First, the macro lattice structure was fabricated using the Direct Ink Writing (DIW) technology (Lewis 2006).DIW uses an extrusionbased 3D printing technology that offers significant advantages in preparing customised structures.During the process, materials were prepared into a paste and extruded through a syringe.A two-dimensional pattern was formed according to the movement track of the syringe controlled by a computer.Then different patterns were printed layer-by-layer by moving vertically so that the pattern formed a three-dimensional part.The second tubular structure can be fabricated by designing a special core-shell nozzle (Rao et al. 2005).In the DIW process, the profile of extruded material was shaped and limited by the configuration of the nozzle.It was reported (Zhang et al. 2019) that the core-shell structure could be fabricated using a DIW printer with a coaxial nozzle.When printing, the coaxial nozzle will be loaded with two different materials, one with carbon nanotubes as the inner layer and the other with silk fibroin as the shell.Finally, the micro-level porous wall was fabricated by introducing emulsion ink (Li et al. 2022) into the coaxial DIW method.The emulsion is a stable mixture of two or more immiscible liquids.Depending on the type of the liquids, the emulsion can be divided into an oil-inwater emulsion or water-in-oil emulsion.After posttreatment, the dispersed phase of the emulsion is removed, and the continuous phase forms a porous structure (Liu and Zhai 2022).(Kim et al. 2017) under the terms of CC-BY license; ii.The surface of butterfly wings.Reproduced with permission from Ref. (Han et al. 2016).
Copyright 2016, American Chemical Society; iii. the structure of bone.Reproduced with permission from Ref. (Liu, Luo, and Wang 2016).Copyright 2016, John Wiley and Sons; iv. the compound eyes of insects.Reproduced with permission from Ref. (Huang et al. 2019).Copyright 2019, John Wiley and Sons.(b) Fabrication of porous samples with hierarchical structure via coaxial nozzle direct ink writing process (the scale bars for lattice structure and tubular struts are 500 μm and the scale bar for porous wall is 100 μm).
This paper proposed a hierarchical emulsion-tubelattice (ETL) material with superior oil-water separation and storage capabilities.The coaxial DIW, with printing parameters optimised, is developed as a one-step method to print commercially available silicone elastomers.The printed structures were then observed, the effects of structure on oil absorption performance were studied, and the compressive and rebound properties were measured.Finally, the feasibility of the material used for water-oil separation was tested.The method used in this study could easily be extended for materials.

Preparation of the ink
Two types of ink were prepared for the coaxial DIW method.The water-in-oil emulsion ink used silicone elastomer as a representative material and the support ink.Two types of silicone elastomers were used for preparing the oil phase in the emulsion ink, Sylgard 184 (Dow Corning, USA) and SE 1700 (Dow Corning, USA).First, the Sylgard 184 was prepared in a 10:1 base to curing agent ratio.The SE 1700 (mixed by 10:1 base to curing) was added into Sylgard 184 with a mass ratio of 1:1 for adjusting the rheology.Span 80 (Sigma-Aldrich, USA), a surfactant, was added to the mixture at a mass ratio of 2 wt.% of the deionised water (DI water).The final emulsion was prepared by dropping DI water into the mixture with stirring.The proportion of the oil phase and water phase are listed in Table S1.The emulsion was further homogenised and degassed by a planetary mixer (SK-300SII, Kakuhunter, Japan).Afterwards, the emulsion was labelled and loaded into a syringe carefully for subsequent use.
The support ink was prepared by adding Pluronic F-127 (Sigma-Aldrich, USA) powder into DI water with stirring and heating at 60 °C.After the powder was swollen entirely, the solution was kept in a refrigerator at −4 °C.Red pigments (Taobao, China) were added to the solution to distinguish it from the emulsion ink.

Fabrication of samples
Three types of samples were fabricated to characterise the effect of structure based on absorption performances.The emulsion-tube-lattice (ETL) samples with a three-level structure, as seen in Figure 1b, were fabricated with a coaxial nozzle (Nanoapparatus Co., Ltd., China).The outer layer of the nozzle was loaded with emulsion ink, and the inner layer was loaded with support ink.The emulsion and support ink can be extruded under different pressures during printing.
The pressures used during the print for each syringe can be adjusted individually; this allows the flow rate for each material to be roughly the same.A set of parallel lines were printed and measured to determine the appropriate printing speed.The rest of the steps were the same as the single-nozzle DIW process (Hinton et al. 2016) using a commercial DIW machine (Allevi, USA).The critical printing parameters are listed in Table S2.Once the samples were printed, it was immersed in DI water and cured at a temperature of 80 °C for 6 h.Before the support ink was removed by ultrasound, the sample was first cut slightly to expose the support ink and immersed in DI water.Two different samples were fabricated as the control groups.One was the lattice samples with solid tubular struts (labelled as 'STL').The STL was fabricated via the same process by replacing the emulsion ink with pure silicone ink.Another one was the lattice samples without tubular struts (labelled as 'SL').The SL samples were fabricated via the single-nozzle DIW process (Hinton et al. 2016).
To improve the hydrophobicity of the sample, 3 ml of hexadecyltrimethoxysilane (Sigma-Aldrich, USA) and 5 ml of tetraethyl orthodox (Sigma-Aldrich, USA) were dissolved into 25 ml ethanol (Sigma-Aldrich, USA).After curing, the samples were immersed in the solution for 30 mins.The samples were then treated under an alkaline atmosphere for 24 h and heated at 120 °C for 30 mins.

Characterisation of the properties
The viscosities and moduli of the inks were measured using the rheometer (HAAKE MARS rheometer, Thermo Scientific, USA) at room temperature with two plates of 35 mm and a gap of 0.15 mm.The digital microscope (VHX-6000, Keyence, Japan) was used to measure the dimension of the printed sample.The microstructure of samples was observed by a field emission scanning electron microscopy (FE-SEM, S-4300, Hitachi, Japan) after being coated with gold for 120 s at 20 A. The compression test was carried out on SL, SLT, and ETL samples (10 × 10 × 10 mm) using a universal testing machine (Instron 5848 Microtester, USA), and the loading rate was 50% strain min −1 .The sample density was calculated by the ratio of mass to volume.The porosity was determined according to the following equation (1): Where the ρ l is the density of the lattice sample, and the ρ s is the density of the solid sample.The water contact angle of samples was measured by a standard contact angle measuring instrument (SDC-80, Sindin Precision Instrument Co., Ltd., China).The oil absorption capacity (Zhang et al. 2021) was calculated with the equation ( 2): Where C represents the absorption capacity (g/g), m 0 is the samples' mass (g) before absorption.The samples were immersed in liquids until equilibrium was reached, and mass (g) was measured and recorded as m e .The sample was squeezed to drain the oil and weighed.The cyclic oil absorption capacity was tested by repeat oil absorption and oil drain.The ability of the sample to retain the absorbed oil was characterised by the oil retention capacity (Zhang et al. 2021) as follows: Where R is the oil retention capacity (%).m 0 and m e are the mass (g) of the sample before and after oil absorption, respectively.m r is the mass (g) of the samples placed on a mesh for 24 h.The liquids used in these experiments include DI water, ethanol (Sigma-Aldrich, USA), simethicone (Sigma-Aldrich, USA), sunflower oil (purchase from local distributor), chloroform, dimethylformamide (DMF), and pump oil (purchase from local distributor).

Fabrication of samples
Fabricating the hollow tubular struts limits the acceptable rheological properties of the ink-most importantly viscosity and yield stress (Shao et al. 2021).The ideal ink for the DIW process should have low viscosity during extrusion with high yield stress to keep its shape.Due to the weak rheological properties of silicone elastomer ink, the hollow structure could not be printed directly.Therefore, the supported ink was added to support the structures.During printing, the emulsion ink was extruded from the outer needle and wrapped around the support ink extruded from the inner needle.The Pluronic F127 solution was also commonly used as the support ink, which flowed like a liquid at low temperature (−4 °C) and was gelatinous at room temperature (Bohorquez et al. 1999).The viscosities and moduli of the support ink are shown in Figure S1a and b.As a result of the satisfactory rheology of the support ink, the inner cavity of the coaxial strut could be formed with good circularity.
In addition, the emulsion ink must have good rheological properties to retain its printed shape and allow curing Figure 2a shows the curves of log viscosity versus log shear rate of inks with different water content.It can be seen that the viscosity of ink decreases as the shear rate increases.Furthermore, greater water content increases the viscosity of the ink.At a shear rate of 40 s −1 , the viscosity of the ink with 70 wt.%water content was less than that of the ink with 50 wt.%water content.The reason was that the water droplets in emulsion were independent of each other due to the presence of surfactants.In this way, the water droplets were like ceramic particles in ceramic ink (Lewis 2000).However, the stable water phase in ink was destroyed when the shear rate was high, resulting in a decrease in viscosity.This indicates that the emulsion ink with 70 wt.%water content is not stable enough.
Figure 2b shows the curves of the storage modulus (G') and loss modulus (G") of different inks.When G' is larger than G", the ink shows an in-deformable flow behaviour like a solid.Conversely, the ink exhibits a liquid-like flow behaviour when G" is larger than G' (Lewis 2000).Hence, the intersection of the two curves is determined as the yield point of the ink.Higher yield stress means the ink does not flow well and can retain its shape after extrusion.Conversely, higher yield stress also requires higher pressure during extrusion.Figure 2b shows that the yield stress of mixed inks with 30 wt.% water content reaches 110 Pa.Correspondingly, as the water content increased to 50 and 70 wt.%, the yield stress of the samples raised to 520 Pa and 940 Pa, respectively.This indicates that in terms of the ability to maintain shape, both 50 and 70 wt.%water content inks were more suitable.
The porous structure was obtained by removing the dispersed phase from the emulsion after the continuous phase was cured.The microstructures of samples with a water content of 30, 50, and 70 wt.%were observed (Figure 2c-e).From each sample, it can be seen that the pores showed a relatively complete spherical shape and the size of the pores increased substantially with the increasing water content (Figure 2c3, d3, and  e3).When the water content was 30 wt.%, the pores were mostly closed with an average size of about 0.8 μm.When the water content was at 50 and 70 wt.%, the pores became interconnected through-holes with sizes of about 2 μm.However, the structure of the sample with 70 wt.%water content was damaged, due to high water content it caused the pores to enlarge and weakened the struts within and the internal void showed an incomplete form.This was because, upon curing and drying, the dispersed phase was evaporated, but the shape was preserved (Riesco et al. 2019).Due to surface tension, the dispersed phase was spherically stabilised in the continuous phase before curing.As the water content increased, the pores in the samples were changed from closed to open pores.When the content of water was too high, the structure of the oil phase was then destroyed.Altogether, considering the rheology of inks and internal porous morphology of samples, the emulsion ink with 50 wt.%water content was selected for the following experiments.
To determine to proper printing speed of the coaxial nozzle, a set of lines was extruded with the speed of 0.5, 1, 1.5, 2, 2.5, and 3 mm/s.Figure 3a shows that when the extrusion speed was determined, the increase in the printing speed would result in a decrease in the crosssection area of the printed strut.The lowest printing speed (0.5 mm/s) led to the largest width of the extruded strut (Figure S2).The overfilled material and the limited layer height meant that the cross-section of the strut was formed into a rectangular shape.However, when the printing speed exceeded 2 mm/s, the strut could not be printed continuously.The dimensions of the struts with printing speeds 1, 1.5, and 2 mm/ s are shown on the right of Figure 3a.It can be seen that the strut with 1 mm/s printing speed showed a flat top surface like the strut with 0.5 mm/s printing speed.The strut with 1.5 mm/s printing speed possessed a better shape with a width and height was 0.587 ± 0.008 mm and about 1216.53 mm, respectively.This was because, with the increase in printing speed, the width and height of the strut decreased gradually.This was due to the high coupling between the extrusion and printing speed (Li et al. 2019).Since the ink was incompressible, the volume of the extruded ink was equal to that of the ink that stayed on the substrate.When the extrusion speed increased, the extruded volume also increased, which increased the strut's cross-section.Similarly, an increase in print speed led to a decrease in crosssection for the same volume.Based on these results, the printing speed of 1.5 mm/s was selected.
The SL, STL, and ETL samples were obtained from optimised process parameters.SL and STL samples' structure are shown in Figure S3a and b.The SL sample showed a lattice structure with solid struts.The STL sample showed a similar lattice structure with tubular struts but without the porous wall.The structure of the ETL sample is shown in Figure 1b.The samples contained three main levels of structure.First, the macro lattice structure was fabricated with the DIW method.The size of the extruded struts was studied, allowing the gaps between the struts to be adjusted by customising the samples' structure.Second, the tubular structure was obtained by removing the inner supported ink.Lastly, the microporous wall is formed by the emulsion ink.Owing to the hierarchical structure, the ETL sample possessed the highest porosity (Figure 3b), which reached 0.80 ± 0.06, and the lowest density, 0.19 ± 0.01 g/cm 3 .Some pores could be observed on the inside (Figure 3d) and outside (Figure 3e) surface of the tubular strut.This indicated that the inner cavity could connect well to the external environment through micro-pores.The morphology of tubular struts inside surface and outside surface were different.This was because during curing, the inner surface was in contact with the support ink and the outer surface was exposed to hot DI water.The temperature and water content of the support ink (room temperature, 70 wt.%water) and DI water (80°C, 100 wt.% water) were different, thus leading to differences in the structure of the surface after curing.

The oil absorption performances
The oil absorption performances of samples are highly correlated with the structure (Shin et al. 2019).As seen in Figure 4a, the stem of the cactus possesses many spines, which can direct the flow of droplets condensed in the air to the stem.The stalks, on the other hand, have a large number of thin-walled tissue cells that can store large amounts of water inside.As a result of the particular hierarchical structure, the cactus has an excellent ability to capture and store water in the air (Shin et al.  ).Inspired by the structure of the cactus, the tubular struts with porous walls were fabricated for absorbing and storing oil liquid (as shown in Figure 4a).Excluding the effects of the lattice, which is the gaps between the struts, the block samples of tubular struts were printed.As seen in Figure 4b, the volume of the oil droplets decreased immediately after they touched the surface.Subsequently, the oil droplets were gradually absorbed into the hollow tubular structure.This indicates that the oil was absorbed into the sample through the wall of the tube.In this way, the oil droplets could be stored.
To characterise the potential of the samples as oilwater separation materials, the oil absorption capacity, oil absorption rate, and absorption capacities of samples for different liquids were tested.Figure 4c shows the results of the oil absorption and retention capacity of SL, STL, and ETL samples.It can be observed that the ETL sample possessed the highest oil absorption capacity, and SL sample was the lowest.The oil retention capacity of the samples showed the same trend.This suggested that the macro-tubes in ETL samples could contribute to liquid storage and improve absorption and retention capacities.The tested oil absorption rate is shown in Figure 4d.The oil absorption rate of all samples reached its equilibrium value in about 30s.During the rising oil absorption capacity, the ETL sample had greater capacity than the STL sample, while the SL sample had the lowest capacity.By comparing the test results between STL and SL samples, it could be found that the tubular strut helped to improve the oil absorption rate.Moreover, the porous walls could increase the oil passage into the tubular strut, further increasing oil absorption speed (Khosravi and Azizian 2016).In addition, the ETL samples also could absorb different liquids (Figure 4e).This demonstrated the universality of the tubular strut to absorb liquids.
Due to the high lipophilicity of the samples, oil could easily adhere to these surfaces, because the contact angle (101.11± 0.50 °C) of the sample to water was relatively high.Therefore, the ETL samples could allow water in the water/oil mixture to pass through the gaps of the lattice while adsorbing the oil, as seen in Figure 5a.Knowing this characteristic, ETL samples can be used as a material for water-oil separation.As seen in Figure 5b, the water could flow into a beaker when decanting the water-oil mixture in the centrifuge tube.At the same time, the oil was absorbed into the ETL sample.The ETL sample could also absorb the oil in the water-oil mixture by dipping, thus achieving water-oil separation (Figure 5c).

The repeatability of samples
Generally, soft materials could possess relatively lower modulus and high elasticity than rigid materials (Chen and Cao 2018; Chen et al. 2019).When silicone elastomers are used as water-oil separation materials, the oil can be discharged from the material by squeezing it.This allows the water-oil separation materials to be reused.The compression test characterised the compressive properties of the SL, STL, and ETL samples (Figure 6a).The result showed that the SL samples possessed the highest modulus (9.43 ± 0.35 kPa), followed by STL samples (1.17 ± 0.73 kPa) and ETL samples (0.88 ± 0.09 kPa).The compressive strain of the ETL samples exceeded 90%.Additionally, the ETL samples could recover after 20%, 40%, 60% and 80% strains (Figure 6b).These results suggested that the sample with tubular struts had higher ultimate elongation and became softer than SL samples, which was beneficial for compressing and transporting.The cyclic compression curves under the compressive strain of 80% are shown in Figure 6c along with the maximum stress and retention modulus of every cycle.The results showed that, after the first compression, the maximum stress was reduced to 89.10%, and the modulus was reduced to 42.77%.After further compression, the maximum stress and modulus continued to decrease.After ten compressions, the maximum stress was only 71.03% at the beginning, and the modulus was reduced to 38.08%.Overall, the compression test showed that this layered structure design could reduce the sample's modulus, and the sample's mechanical properties tend to be stable after cyclic use.Moreover, the oil stored in the sample could be squeezed out, restoring the ability to absorb oil. Figure 6d shows the testing performance of the cyclic absorption capacity and indicates that the sample's ability to absorb oil did not decrease significantly after 20 cycles.

Conclusion
This paper developed the coaxial DIW process to fabricate bioinspired ETL material consisting of a three-level structure.First was the macro lattice structure, which could increase the contact area of the liquid.The second was the tubular structure that served to store the liquid.The last was the microporous wall with selective oil absorption capacity that could absorb the oil from outside into the tubular structure.
The silicone elastomer was used as a typical soft material.The structure of the samples was observed.Additionally, the oil absorption performances and mechanical properties, were tested.The results of microstructure observation suggested that the pores inside the samples were interconnected and allowed air to pass through.ETL samples could absorb various liquids.Benefiting from the tubular struts, the ETL samples have a satisfactory oil absorption capacity and oil absorption rate.Moreover, with the hierarchical lattice design, the ETL samples could absorb the oil phase from the water-oil mixture as it was filtered.Therefore, the samples could be used for oil-water separation.The compression test showed that the compression ratio of samples was improved with the tubular strut, but the modulus grew worse.Despite the decreasing modulus, the ETL samples could withstand multiple cycles of compression.Therefore, the oil in the sample could be extruded for reuse.The results showed that after 20 cycles of oil absorption and drainage, there was no significant decrease in the oil absorption capacity of the ETL samples.Overall, the ETL material possessed an excellent oil absorption capacity and could selectively filter the oil in water/oil mixtures.More importantly, the coaxial DIW process can easily be extended for fabricating other soft and rigid materials.

Disclosure statement
No potential conflict of interest was reported by the author(s).

Notes on contributors
Mr. Guanjin Li is a Ph.D. candidate at ZhaiGroup@NUS, from the Department of Mechanical Engineering, National University of Singapore.He studies bio-inspired hierarchical structure materials.
Ms. Xinyu Dong is a Ph.D. candidate at ZhaiGroup@NUS, from the Department of Mechanical Engineering, National University of Singapore.She studies the freeze casting of hierarchical structure materials.
Mr. Ambrose Raphael Chuan is a Ph.D. candidate at Zhai-Group@NUS and a research engineer from the Department of Mechanical Engineering, National University of Singapore.He studies the 3D printing of flexible electronics.
Dr. Beng Wah Chua is a research scientist at the Singapore Institute of Manufacturing Technology (SIMTech), A*STAR.He leads the advanced forming and surface technologies for SIMTech.
Dr. Li Tao is a research scientist at the Singapore Institute of Manufacturing Technology (SIMTech), A*STAR.He leads the advanced powder processing technologies for SIMTech.
Dr. Wei Zhai is an assistant professor at the Department of Mechanical Engineering, National University of Singapore.She leads the ZhaiGroup@NUS, developing nature-inspired advanced materials engineering via multiscale and multimaterial manufacturing technologies.

Figure 1 .
Figure 1.(a) Examples of hierarchical structures in nature: i. the structure of cactus.Reproduced with permission from Ref.(Kim et al. 2017) under the terms of CC-BY license; ii.The surface of butterfly wings.Reproduced with permission from Ref.(Han et al. 2016).Copyright 2016, American Chemical Society; iii. the structure of bone.Reproduced with permission from Ref.(Liu, Luo, and Wang 2016).Copyright 2016, John Wiley and Sons; iv. the compound eyes of insects.Reproduced with permission from Ref.(Huang et al. 2019).Copyright 2019, John Wiley and Sons.(b) Fabrication of porous samples with hierarchical structure via coaxial nozzle direct ink writing process (the scale bars for lattice structure and tubular struts are 500 μm and the scale bar for porous wall is 100 μm).

Figure 3 .
Figure 3. (a) Left was the optical image of a group of printed lines with different printing speeds and right were the dimensions measured by the confocal microscope.(b) the density and porosity of solid, SL, STL, and ETL samples.The (c) cross-section of the tubular strut and the (d-e) were the outside and inside surfaces of the tubular strut, respectively.

Figure 4 .
Figure 4.The (a) Schematic diagram of the cactus oil absorption and storage (left) and the inspired sample (right).(b) the oil absorption and storage behaviour of the sample with tubular struts.The results of (c) oil absorption and retention capacity of SL, STL, and ETL samples, (d) the oil absorption rate of SL, STL, and ETL samples, and (e) the different liquids absorption capacity of ETL sample. 2019

Figure 5 .
Figure 5. (a) The schematic diagram of selective filtration of oil/water mixture, (b) filtration of oil/water mixture through ETL sample, and (c) Selective oil absorption from water-oil mixtures.

Figure 6 .
Figure 6.(a)-(c) The results of the compression test.(a) the curves of compressive stress vs. compressive strain of SL, STL, and ETL samples, (b) rebound curves of ETL samples after 20%, 40%, 60%, and 80% strain, (c) curves of ten compression cycles at 80% strain of ETL samples and inner was the retention of maximum stress and modulus of ETL samples during the 10 cycles at 80% strain.(d) The result of the cyclic oil absorption test.