Enhancing Biolm Formation and Microalgae Growth by Preparing Cellulose Substrate With Rough Surface

A series of cellulose lms with rough surface were prepared by employing a simple solution casting method and using the waterproof abrasive papers with different grits as the substrate. The cellulose lms possessed a rough surface with the maximum height difference (S z ) of 128-217 μm, a macroporous structure with a high porosity of 84.7%-90.5%, and a negative potential between -40.00 and -54.15 mV. Furthermore, the cellulose lms exhibited excellent microalgae adhesion properties. After 18 days of attached Chlorella sp. cultivation experiments, the average productivities of C-A-120 lms (C-A-X, X means the mesh number of the substrate) reached 20.80 g m -2 d -1 ), which is 2.69 times than that of the cellulose lm with a smooth surface. The result indicates that the cellulose lms with a rough surface and high water adsorption ratio have a huge potential in serving as the substrate of the attached microalgae cultivation to promote microalgae cells growth and biolm formation.


Introduction
Microalgae have been regarded as one of the most promising sources for biofuel (Benemann 2013;Maeda et al. 2018) and other value-added products (Borowitzka 2013;Matos et al. 2017) due to their fast growth rates and high value-added bioproducts production. However, it is still a great challenge to achieve a large-scale commercialized production of microalgae, because one of the key challenges is the high cost of microalgae harvesting and dewatering in conventional suspended culture systems (Bharathiraja 2015;Pierobon 2018). Recently, attached cultivation based on microalgae bio lm has received widespread attention because of the low water demand, the ease of biomass harvesting and its higher concentration (Wang et al. 2017;Gross et al. 2015). For this mode, the cells are xed to the solid substrate surfaces, which made them separate easily from the culturing medium and signi cantly reduced the cost of microalgae harvesting and separation.
During the attached cultivation process, the formation of microalgae bio lm depends on various factors including microalgae species, substrate materials, culture conditions, hydrodynamic conditions, et al. (Zhuang et al. 2018) The substrate material is one of the most important factors in the development of an attached culture system, because it in uences the initial adhesion and the subsequent growth of the microalgae bio lm (Schnurr and Allen, 2015). Up to now, various materials used in the microalgae attachment have been selected (Rosli et al. 2020). Cellulose material has been regarded as one of the most potential materials due to its high hydrophilicity, non-biotoxicity, low price, and porosity (Zhang et al. 2017). Several kinds of cellulose materials, such as cellulose acetate/nitrate membrane (Ji et al. 2014), newspaper (Naumann et al. 2013), printing paper (Bernstein et al. 2014) and cotton rope (Christenson and Sims 2012;Bernstein et al. 2014), have been evaluated. However, the reported cellulose-based substrate materials are directly obtained from the market, and their structures have not been controlled and designed. Thus, in order to obtain high microalgae productivity and understand the relationship between the structure of cellulose materials and the microalgae growth, it is urgently needed to develop new cellulose materials with novel and de ned structures to serve as the substrate of microalgae cells in attached microalgae cultivation.
In this work, by changing the roughness of the substrate (glass, waterproof abrasive paper with different roughness), a series of cellulose lms with different surface roughness were prepared by a solution casting method in ionic liquids. Subsequently, they were used as the supported materials in attached microalgae cultivation (Fig. 1). The structure and properties of cellulose lms were characterized by Fourier transform infrared (FTIR) spectra, zeta potential, scanning electron microscopy (SEM), mercury porosimetry and water absorption experiment. The effect of surface roughness on the microalgae adhesion of cellulose lms was investigated by attached Chlorella sp. cultivation in 24-well tissue culture plate.

Materials
The strain Chlorella sp. FACHB-1514 was obtained from Institute of Hydrobiology, Chinese Academy of Sciences, which was grown and maintained in sterilized BG 11 medium at 25 ± 2 o C under a light intensity of 100 µmol m − 2 s − 1 in HZQ-QG light incubator (HDL Apparatus) (Yan et al. 2016). The culture with optical density (OD 680 ) about 0.3 was used as the seed for the subsequent attached cultivation experiments. The cellulose (cotton pulp) with a degree of polymerization (DP) of 620 and cellulose content of bigger than 98% was supplied by Hubei Chemical Company Limited (Xiangfan, China), which was dried at 80 o C under vacuum for 6 h prior to use. The ionic liquid, 1-allyl-3-methylimidazolium chloride (AmimCl), was obtained from Institute of Chemistry, Chinese Academy of Sciences. The water content in the resultant ionic liquid was less than 0.3% as measured by Carl-Fisher method. All the inorganic salts and organic solvents were analytical reagents and used without further puri cation.

Preparation of cellulose lms
The cellulose lms with different surface roughness were prepared in AmimCl using dissolution and regeneration method ( Fig. 1) by changing the surface roughness of the substrate. Glass plate and waterproof abrasive papers with different micrometer-sized morphology (40 mesh, 80 mesh, 120 mesh and 300 mesh) were used as the substrate. A typical preparation procedure used is as follows: 4.0 g of cellulose was dispersed into 96.0 g of AmimCl in a ask, and then the mixture was stirred at 80 o C for 1.5 h to ensure a complete dissolution of cellulose. The solution was degassed in vacuum, then cast onto the abrasive paper (40 mesh) and immediately coagulated in 60 o C deionized water to make regenerated cellulose lms. To remove residual ionic liquid AmimCl in the cellulose lms, they were further washed with distilled water at least three times until no Cl − ions were detectable by the AgNO 3 test. After solvent exchange with tert-butanol and further freeze drying, the cellulose lm (C-A-40) was obtained and kept in a desiccator prior to characterization.

Microalgae bio lm cultivation
The microalgae adhesion properties of the cellulose lms were explored by attached Chlorella sp. cultivation experiments as shown in Fig. 1. A typical cultivation procedure used is as follows: the cellulose lms were rstly cut into pieces with the diameter of 1.5 cm, and then were immersed into a beaker with BG 11 media. Then, the cellulose lms were placed in the wells of a sterile and transparent 24-well tissue culture plate. The algal culture solutions of 200 µL were slowly added on the cellulose lm surface, then the plate was kept in HZQ-QG light incubator (HDL Apparatus) under a light intensity of 100 µmol m − 2 s − 1 and at a temperature of 25 ± 2 o C. The sterilized BG 11 medium of 500 µL was added into each well by slowly dripping from the side of the culture-well every 24 h. The uorescence intensity of attached microalgae cells on the cellulose surface was measured every 3 days. The cellulose lm was taken out and rinsed thrice with water to remove planktonic microalgae cells, then the absorbance was measured at 570 nm using a multi well plate reader (In nite M200, TECAN, Switzerland). Moreover, the surface changes of the cellulose lms during the attached microalgae cultivation experiment were observed using photographs and light microscope.
At the end of the attached cultivation experiments (with a total duration of 18 days), the attached microalgae on the surface of cellulose lm were quanti ed by measuring the cell dry weight. The cellulose lm was washed ultrasonically three times with deionized water to remove the attached microalgae, then all the washed solution was collected and centrifugated at 8000 rpm to separate the microalgae. The microalgae were dried by freeze drying and then weighed using an analytical scale (XS105DU, Mettler Toledo, Switzerland).
The attached microalgae productivity (g m − 2 d − 1 ) was calculated as the Eq. (1): 1 Where W 0 and W 1 are the dry weight of the inoculated and harvested microalgae biomass, respectively. A is the surface area of the cellulose lm, which is equal to 1.77×10 − 4 m 2 . t is the days of the attached cultivation.

Measurements
The surface morphology of cellulose lms was observed on a Leica DMI4000B inverted uorescence microscope. The cross section of cellulose lms was observed on a JSM-6700F JEOL scanning electron microscope (SEM) at an accelerating voltage of 10 kV. The specimens were coated with platinum before observation. The pore size and its distribution of cellulose lms were measured using an AutoPore IV (Micromeritics, USA) instrument.
The surface zeta potential of cellulose lms was measured with SurPass zeta potential analyzer (Anton Paar, Austria) at room temperature using BG11 medium (pH = 7.5-7.8) as the electrolyte, and the potential was measured three times to get the average potential for each sample.
The ATR-FTIR spectra of pulp cellulose and regenerated cellulose lms were measured with Nicolet™ IS5 FTIR spectrometer. The spectrum was recorded at room temperature with a resolution of 4 cm − 1 and 24 scans per sample.
The water absorption ratio (W A ) of cellulose lms was quanti ed by measuring the dry weight (W D ) and wet weight (W W ) of cellulose lms. The cellulose lm was dried at 80°C under reduced pressure until constant weight and then was cooled down to room temperature. The sample was taken out and weighed immediately. The weighted cellulose lm was submerged in water at 25°C for 48 h, then the water on the surface was removed by lter papers and the cellulose lm was weighed. The water absorption ratio (W A ) was calculated using the equations (2) as follows: 2 The arithmetic mean height (Sa) and the maximum height of the cellulose lm surface (Sz) were measured with a confocal laser microscope with a program (OLS4000, Olympus, Massachusetts, USA), which allowed the measurement of surface roughness in a linear manner and in speci c areas. The images were captured with a 10 magni cation and 0.2 µm registration accuracy. The central area of the cellulose lms (500 µm 2 ) was selected to analyze the Sa and Sz of the cellulose lms that was expressed as a numerical value (µm). Three equidistant measurements were performed for each specimen in three different areas.

The structure and properties of cellulose lms
In order to characterize the morphology of the cellulose lms, they were lyophilized to maintain their microstructure. The surface morphology of the prepared cellulose lm was studied by visually and optical microscope. The photographs of cellulose lms are shown in Fig. 2a. It can be seen that the cellulose lms are white and opaque sheets. The photographs also show that the C-G lm exhibits smooth surface, while the C-A lms exhibit coarse surfaces with irregular shape and a change in the morphology occurs for different C-A lms due to the surface roughness of the substrate during the regeneration process. To gain more insights into the surface morphology of those cellulose lms, optical microscope observations were further carried out. As shown in Fig. 2b, it can be observed that a large number of concaves cover the surface of the C-A lms with diameter of dozens to hundreds µm, and the diameter of the concaves for different C-A lms decreases with the increase of the mesh of the substrate, while there is no obvious concave on the surface of C-G lms.
Since the microalgae adhesion capacity is highly dependent on the surface properties and structure of the cellulose substrate such as surface hydrophilicity/hydrophobicity, surface charge and surface roughness/micropattern. An attempt was made to measure the water contact angle of the cellulose lms, but failed due to the porous structure and water adsorption behavior of the cellulose lms. The zeta potential of cellulose lms was investigated in BG11 medium at pH 7.5-7.8, which is illustrated in Table. W The zeta potential of all the cellulose lms exhibited negative values ranging from − 40.00 ± 3.31 mV to -54.15 ± 3.84 mV probably due to the full deprotonation of OH in the cellulose skeleton, which is similar to that of cellulose reported in previous research (Qian et al. 2019). The surface roughness of the cellulose lms was measured using confocal laser-scanning microscopy (CLSM), which reports the vertical deviation from the actual surface using the arithmetic mean height (Sa) and the maximum height of the pro le (Sz). Sa provides statistically stable and more accurate measurement with optical pro lometry. However, it has limitations to differentiate between peaks and valleys on the surface. Sz is more sensitive to peaks and valleys than that of Sa, thus both Sa and Sz were measured. As shown in Table 1, the C-A-120 owned the roughest surface (Sa = 26.17, Sz = 217). This is followed by C-A-80, C-A-40, C-A-300, the smoothest surface of C-G being 15.68 µm (Sa) and 108 µm (Sz). The C-A lms show a signi cant difference between Sa and Sq due to the difference of the substrate, which indicates that it is feasible to design the surface roughness of cellulose lm by changing the roughness of the substrate. All the cellulose lms were further utilized as the substrate of attached microalgae cultivation and the effect of the surface roughness on the microalgae adhesion was investigated. The IR spectra of the cellulose pulp and regenerated C-A-40 lm are shown in Fig. 3a. The spectra of C-A-40 is similar to the native cotton pulp, indicating that no chemical reaction occurs during the preparation of the cellulose lm. Similar results were also reported in the other reports about the cellulose dissolution and regeneration process in ionic liquids (Zhang et al. 2005). A broad vibration band around 3284 cm − 1 is assigned to the O-H vibrations of cotton pulp, while the -OH stretching vibration band in the C-A-40 shifts to a higher frequency (3337 cm − 1 ) and becomes sharper and narrower as a result of splitting hydrogen bonds to some extent during the dissolution and regeneration process (De Silva et al. 2015). A new shoulder round 899 cm − 1 is observed only in the C-A-40 but not in the cotton pulp, which is assigned to the C-O stretching vibration in the amorphous region (Zhang et al. 2005). The absorption band at 1426 cm − 1 is assigned to the CH 2 scissoring motion for the cotton pulp, while it weakens and shifts to a lower wavenumber at 1421 cm − 1 for C-A-40 due to the destruction of the intramolecular hydrogen bond.
The porous structure of the cellulose lms was determined by SEM and mercury porosimeter measurements. The cross-section SEM images of cellulose lms are shown in Fig. 2c. The porosity, total intrusion volume, average pore diameter and pore size distribution are shown in Table 1 and Fig. 3b. All the cellulose lms possessed high porosities of more than 84.68%, the average pore diameter is in range of 185.52-303.51 nm, which is bene cial to water adsorption. For these cellulose lms, the porous structure is formed primarily in the regeneration step. The gelation of the cellulose solution proceeds after immersing the cellulose solution into 60 o C deionized water, so the phase separation occurs very quickly due to the high driving force caused by the large concentration gradient and high diffusion coe cient of the ionic liquids in the bath of hot water, resulting in large pores in the cellulose lms (Mi et al. 2016).
Obviously, there is no relationship between the porosity and average pore diameter and the substrate used in the preparation of cellulose lms. The pore volume distributions of the cellulose lms are shown in Fig. 3b, the results suggested that the pore size distribution of all the cellulose lms is similar and covers the size range between 100 nm and 3 µm, which is less than the size of microalgae and makes the microalgae adhere the surface but not the inner of the cellulose lms in attached microalgae cultivation.
The water adsorption ability is of great importance for cellulose substrate in attached microalgae cultivation, which decreases liquid evaporation and maintains moist environment during attached microalgae cultivation. The water adsorption ratio of cellulose lms was measured and shown in Table 1.
It is found that all the cellulose lms display high water-uptake properties with higher than 174% due to their water a nity and porous structure (Jiang et al. 2019). The C-A-120 shows the highest water adsorption ratio (419%), which enables them to absorb a great deal of microalgae cultivation liquid and bene ts for the growth of microalgae. The rough surface and good water adsorption ability of cellulose lms make them potential in serving as the substrate in the attached microalgae cultivation.

Microalgae adhesion and growth properties on cellulose lms
The surface changes of cellulose lms during the attached microalgae cultivation were observed visually by light microscope. The photographs of cellulose lms at different time intervals are shown in Fig. 4a. The C-A-120 becomes green at 3th day, while C-G shows green color until 12th day. Moreover, the surface of C-A-120 is much greener and more area is covered with green colour than those of C-G gels until the end of the cultivation. Similar results are also observed by light microscope as shown in Fig. 4b. After 6 days of cultivation, the surface of C-A-120 has been covered completely by microalgae, while the surface of C-G is covered with microalgae partially even after 18 days of cultivation. The growth of microalgae cells on the surface of cellulose lms is also conducted by measuring the photosynthetic activity of the cells as an indication of the cellular growth at different time intervals during attached microalgae cultivation. As shown in Fig. 5a, the uorescence intensity on the surface of all C-A lms is higher than that of C-G lms. For example, the uorescence intensity of the C-A-120 is 1.78 times than that of C-G after the 18 days cultivation, indicating that more microalgae cells adhere on the surface of the C-A-120.
The areal productivity of Chlorella sp. on the surface of cellulose lms were measured and shown in Fig. 5b. After 18 days of cultivation, the area productivity of Chlorella sp. bio lm in C-A-40, C-A-80, C-A-120, C-A-300 is 16.28 ± 2.97, 10.44 ± 1.17, 20.8 ± 3.76 and 9.04 ± 2.24 g m − 2 d − 1 , respectively, which are 110.6%, 35.06%, 169% and 16.95% higher than 7.73 ± 3.99 g m − 2 d − 1 attained in the C-G. The C-A-120 shows the highest area productivity, illustrating a positive correlation between microalgae productivity and the surface roughness of cellulose lms. The cellulose lms with rougher surface have a relatively larger surface area and deeper grooves, so a bigger microalgae bio lm productivity is obtained (Gross et al. 2016;Huang et al. 2018). Up to now, numerous studies about substrate materials have been witnessed that the substrate materials with an appropriate rough/textured surface provides a "shelter" for the attached microalgae and makes the sloughing of the attached microalgae signi cantly reduce (Zhang et al. 2020;Guo et al.2019). Moreover, the C-A-120 gains the maximum productivity of 20.8 ± 3.76 g m − 2 d − 1 , which is also much higher than that of most cellulose-based substrates in previous reports ( Fig. 5c) (Berner et al. 2015), such as 10.92 g m − 2 d − 1 for pine sawdust (Zhang et al. 2017

Conclusions
A series of cellulose lms have been prepared from an AmimCl solution of cellulose by solution casting using waterproof abrasive paper as the substrate. The cellulose lms possess rough surface and macroporous structure together with good water absorption properties. Moreover, the cellulose lms exhibit good microalgae adhesion properties. After 18 days of cultivation, the maximum productivity of Chlorella sp. bio lm in C-A-40, C-A-80, C-A-120, C-A-300 reaches 16.28 ± 2.97, 10.44 ± 1.17, 20.8 ± 3.76 and 9.04 ± 2.24 g m -2 d -1 , respectively, which are 110.6%, 35.06%, 169% and 16.95% higher than 7.73 ± 3.99 g m −2 d -1 attained on the C-G with a smooth surface. These cellulose lms with a rough surface and high water adsorption ratio can be considered as the potential candidate for substrate materials in attached microalgae cultivation. Figure 1 Schematic illustration of the fabrication process of the cellulose lms and their application as the substrate materials in attached microalgae cultivation