Investigating the anti-bioadhesion properties of short, medium chain length, and amphiphilic polyhydroxyalkanoate films

Abstract Silicone materials are widely used in fouling release coatings, but developing eco-friendly protection via biosourced coatings, such as polyhydroxyalcanoates (PHA) presents a major challenge. Anti-bioadhesion properties of medium chain length PHA and short chain length PHA films are studied and compared with a reference Polydimethylsiloxane coating. The results highlight the best capability of the soft and low-roughness PHA-mcl films to resist bacteria or diatoms adsorption as compared to neat PDMS and PHBHV coatings. These parameters are insufficient to explain all the results and other properties related to PHA crystallinity are discussed. Moreover, the addition of a low amount of PEG copolymers within the coatings, to create amphiphilic coatings, boosts their anti-adhesive properties. This work reveals the importance of the physical or chemical ambiguity of surfaces in their anti-adhesive effectiveness and highlights the potential of PHA-mcl film to resist the primary adhesion of microorganisms.


Introduction
Proliferation of marine living organisms on surfaces in contact with water, called biofouling, is an inevitable phenomenon with major consequences both from an economic and an environmental point of view.To prevent problems associated with fouling, toxic molecules embedded in coatings have been historically used.However, the proven environmental impact of these antifouling paints directs research towards more environmentally friendly solutions that do not release any biocides.Thus, today, antifouling research aims at developing green, sustainable, and widely applicable technologies.Fouling release coatings (FRC), based on hydrophobic and low surface energy properties are materials that prevent the adhesion of microorganisms.From the 1970s, Baier (1972Baier ( , 1984) ) and Baier and Meyer (1992) reported that the minimum adhesion of bacteria on hydrophobic polymer surfaces was obtained for materials with critical surface tension between 20 and 30 mN m −1 , corresponding to silicone and fluoropolymers.Silicones became a reference polymer for antifouling applications, such as PDMS, one of the most studied and pave the way for fouling-release coatings.These coatings are effective in limiting the adhesion of microorganisms.It was further demonstrated that PDMS coatings are notoriously poor at preventing the adsorption of diatomaceous slimes (Callow and Callow 2002), which is the primary critical step in the formation of any biofilm.It has been shown that silicone polymer films were also effective in facilitating the desorption or detachment of microorganisms under fluid flow, similar to hydrodynamic forces exerted on marine immerged systems (Valentin et al. 2019).Both of these steps (adsorption and detachment under shear stress) are important for testing the effectiveness of FRC coatings but the film properties involved in each phenomenon are not necessarily based on the same physico-chemical principles and more importantly, are not yet clearly understood.
While low surface energy is an important requirement for FRC coatings, other parameters have also been identified as being important in optimizing the release performance of materials for biofouling.Among them, surface topography and mechanical properties have been the subject of intense research.In a recent review on the effect of the surface structure on bacterial adhesion, Yang et al. investigated more than 15 works from the literature and plotted the adhesion of cells as a function of the roughness parameter (Ra or RMS) for different kinds of substrate, either organic or inorganic (Yang et al. 2022).
Their curve shows a general positive correlation between the number of cells adsorbed on the surface and its roughness above a critical Ra value, estimated to be about 6 nm, while a negative correlation was pointed out below Ra.The observed positive correlation can be explained by the possible deformation of bacteria on rough surfaces and the consequent increase of the contact area and adhesion force between both objects.On the other hand, when the surface is smooth (<6 nm), the small variation in height of the surface does not disturb the microorganisms nor generate stress, which can be accompanied by an enhancement of the cellular metabolic activities as illustrated by a bacterial size increase or production of EPS (Mitik-Dineva et al. 2008;Mitik-Dineva et al. 2009;Yang et al. 2022).
The effect of mechanical properties is also documented in the literature, but more than the two other parameters, its effect is still a matter of debate with contradictory results proposed in the literature.This could be explained by the difficulty in de-correlating the different parameters as illustrated by several works, in which the authors also modified the topography and/or the chemistry in addition to the mechanical properties.Most of these studies relate to systems based on more or less cross-linked PDMS (Song and Ren 2014;Valentin et al. 2019;Arias et al. 2020).A change of the reticulation rate modifies the modulus or more broadly the mechanical properties but not only.The number of free chains and the number of chain ends that can interact with organisms are also modified.
The effectiveness of FRC coatings can be greatly increased by incorporating in low proportion hydrophilic additives like polyethylene glycol (PEG), thus creating what is designed as 'advanced FRC coatings'.There is still no clear understanding of the reaction mechanism involved in the enhanced properties of these coatings.However, it is often proposed that the hydrophilic segments of PEG allow the structuring of a dense layer of water thus creating a steric barrier limiting the adhesion of organisms (Ngo and Grunlan 2017).Another hypothesis has also been suggested and correlates with the possible brush-like conformation of PEG polymers at the surface of the coating in contact with water (Yeh et al. 2012).This brush conformation may limit the adsorption of microorganisms via strong entropic effects of the polymer chains.
Although relatively effective and devoid of toxicity, these PDMS based systems remain expensive, with poor mechanical properties, non-repairable, and insufficiently effective against diatoms.The present work aims to develop anti-bioadhesion films based on biobased neat materials, more precisely bacterial polymers named polyhydroxyalkanoates (Miu et al. 2022).Produced and stored by many microorganisms, they constitute carbon reserves.PHAs are non-toxic, biodegradable and biocompatible.PHAs are polymers made up of monomeric units of hydroxy fatty acids linked together by an ester bond.Each monomer has a side chain whose number of carbon atoms varies according to the producing bacterial strain and/or the nature of the carbon source used for the production.Three groups of PHA can be differentiated according to the number of carbon atoms presents within the monomeric side chain: short chain length (PHA-scl) (3-5 carbon atoms), medium chain length (PHA-mcl) (6-14 carbon atoms) and long chain length (PHA-lcl) (more than 14 carbon atoms).It has been shown that the length of the lateral chain within the PHA structure plays a key role in the interchain organization and interactions (Bugnicourt et al. 2014).Interchain interactions have an impact on the rate and kinetics of crystallization, and the resulting semi-crystalline structure of the biopolymer film is correlated with its overall mechanical and/or surface properties.For example, Volant et al. (2021) observed that the semi-crystalline behavior of the poly3-(hydroxybutyrate-co-hydroxyvalerate) (PHBHV) and poly(3hydroxybutyrate-co-hydroxyhexanoate) (PHBHHX) microbeads were positively correlated with increasing surface topographies and mechanical properties (indentation modulus).The versatility of structures, obtained by modulating the monomer chain length in the polyhydroxyalkanoate, makes it possible to investigate various ranges of surface properties while maintaining significant hydrophobicity.One of the most widely used and studied PHA is the copolymer PHBHV, a short-chain length PHA but other PHAs, still little studied in the form of films, have interesting characteristics with a view to produce fouling-release films (Behera et al. 2022).A previous study has already explored the potentiality of using PHBHV as a matrix to design fouling-release systems.However, even when doped with PEG copolymers, the amphiphilic PHBHV anti-adhesive properties were still less efficient than conventional PDMS/PDMS-PEG films.Its high crystallinity content and a subsequent high roughness surface property in addition to mechanical properties higher than those of elastomers, were pointed out as responsible for a non-optimized antifouling properties (Guennec et al. 2021).A previous work by Mauclaire et al. (2010) has explored the impact of the purity of PHA-mcl films on their antibiofilm properties.These authors observed that the biofilm formation by gram-negative and gram-positive bacteria was lower or similar to the one observed in polystyrene control.
The present study aims to compare the anti-adhesive efficiency of two purified polyhydroxyalkanoates films, formulated from PHA with different lateral chain length monomer, and to determine the physicochemical parameters that play a critical role in limiting bioadhesion.The two investigated biopolymers are respectively the PHBHV (PHA-scl) and a PHA-mcl which is composed of 3-hydroxyhexanoate (HHx), 3hydroxyoctanoate (HO), 3-hydroxydecanoate (HD) and 3-hydroxydodecanoate (HDD) monomers.Adhesion studies on these materials remain focused on the preliminary stage of the colonization process, including the adhesion of some marine microorganisms, such as bacteria and diatoms.Microalgae were chosen because of their persistence on PDMS-based coatings and both microorganisms are highly representative of the pre-colonization species observed on seawater immerged surfaces after few hours.
In this study, the following questions are addressed: (1) is it possible to reach the same antiadhesive performance as PDMS with a biodegradable polymer, such as PHA-mcl? and (2) is the addition of an amphiphilic additive a benefit for the two PHA films in regards of the anti-adhesive properties of the PDMS/PEG reference amphiphilic coating?

Materials
Poly[(3-hydroxybutyrate)-co-(3-hydroxyvalerate)] (PHBHV) containing 30% of hydroxyvalerate and 70% of hydroxybutyrate was purchased from Pacific Biotech (Tahiti, French Polynesia).PHA-mcl was kindly provided by Ifremer.The constitutive monomer units of PHA-mcl are mainly hydroxyoctanoate (48.8%) and hydroxydecanoate (36%).Its Tg is lower than that of PHBHV with a value of −45 � C, as is its enthalpy of fusion, thus showing a crystallinity that is clearly lower than expected.The composition and characteristics of the polymers are given in Table 1.The synthesis of the PHBHHx-b-PEG copolymer was previously described in Guennec et al. (2021).For this purpose, the Poly[(3-hydroxybutyrate)-co-(3-hydroxyhexanoate)]) (PHBHHx) with 20% of HHx was supplied by CHEN Georges, Tshingua University, China.
Polydimethylsiloxane (PDMS) was obtained from Momentive Performance Materials (Waterford, NY, USA) as the trademark RTV-615.A commercial PDMS-PEG copolymer, traded under the name CoatOsil 7602, was obtained from Momentive Performance Materials (Waterford, NY, USA).The graft copolymer (PDMS/PEG) has a molar mass of about 3,000 g mol −1 with PEG pendant side chain, as revealed by carbon and proton NMR (Gillet et al. 2018).

Coating preparation
PHA (10 wt.%) was dissolved for 2 h under reflux (65 � C) in dichloromethane.The polymer solution was then continuously stirred until its return to ambient temperature and 500 mL of the solution was deposited on a clean glass surface (glass microscope slide of 18 � 18 � 1 mm).A polystyrene Petri dish cover was placed on top of each glass slide during the drying process that was set at room temperature for 2 days for the PHA coatings.The final coating thickness was estimated to be �100 mm as measured by an Elcometer 456 (UK).The PDMS was prepared, as specified by the supplier, by mixing the PDMS base and the curing agent (in a weight ratio of 10:1, respectively) in dichloromethane (50 wt %) for 2 h.The resulting solution was deposited, as above, on a glass slide to obtain a film with a typical thickness of around 100 mm.The PDMS coating was dried under ambient temperature for at least 7 days.In a second step, the advanced amphiphilic coatings were prepared by mixing the polymer solution (10% w/w), under the same conditions as above, with the desired PEG copolymer (0.1% w/w for PHAs and 0.01% w/w for PDMS).The final solutions were stirred overnight before being deposited onto glass slides to form the amphiphilic coating.Similar thicknesses were measured for these films.Six coatings were formulated in this study: two PDMS coatings (PDMS and PDMS/PDMS-PEG), two PHA-scl coatings (PHBHV and PHBHV/PHBHHx-PEG), and two PHA-mcl coatings (PHA-mcl and PHA-mcl/PHBHHx-PEG).For the PEG doped PHA samples, PHBHHx-PEG was added at a final concentration of 0.1 wt.% in the polymer solution whereas for PEG doped PDMS coating, the lowest concentration of 0.01 wt.% of PDMS-PEG copolymer (CoatOsil) was added to the PDMS solution.The lowest mass of PEG copolymer added in the advanced PDMS film was optimized and set at this value to avoid any defect on the surface of the coating (cracks for example).To sum up, three coatings were used: glass as commonly used in lab, and on neat or PEG-doped PDMS coatings that were compared with the different PHA coatings (scl and mcl).The coatings composition and names used in the study are presented in Table 2.In the following results and discussion sections, the following terms coatings or films are used equivalently.

Microorganisms culture and adhesion tests
Bacillus (4J6) sp. was isolated from the surface of the glass cover immersed in the Atlantic Ocean (Gulf of Morbihan, France, 47 � 34 0 37 00 N-2 � 44 0 54 00 W) for 6 h at 1 m depth (Grasland et al. 2003).The marine bacterium was cultivated on Zobell agar medium (4 g/L tryptone, 1 g/L yeast extract, and 30 g/L sea salt) and then, incubated for 24 h at 20 � C with shaking at 125 rpm.The cultures were allowed to produce a stock of several microtubes which were cryo-conserved in 20% glycerol at 80 � C for all experiments.The diatom Phaeodactylum tricornutum (AC590) was obtained from the Algo-Bank (Biological Resource Center of the University of Caen Normandie, France).The diatom Navicula perminuta CCAP1050/15 was obtained from the culture collection of algae and protozoa of the Scottish Marine Institute.Diatoms were grown in a sterile artificial seawater (SASW) culture medium with 2% Guillard's F/2 Marine Enrichment Basal Salt Mixture (stored at 4 � C before use), streptomycin (500 mg L −1 ), penicillin G (1 g L −1 ) and chloramphenicol (100 mg L −1 ).Diatom suspensions were maintained under controlled illumination of 110 mmol photons m 2 s −1 using a white fluorescent lamp at 20 � C with light/dark periods (18/6 h, respectively) in a H� elios 600 phytotron (Cryotec, Saint-G� ely-du-Fesc, France).After reaching the stationary phase (15 days), they were left in the dark for 40 h for synchronization.
Coated glass microscope slides were sterilized by UV radiation (at 365 nm) for 30 min, to ensure a non-contamination of the investigated coating before adhesion tests (Portier et al. 2021).Contact angle measurements of the sterilized polymer films reveal that the wetting surface properties were not modified by UV treatment.For the bacterium Bacillus, coated surfaces were incubated in 5 mL of a given suspension in a solution of SASW (10 8 cells mL −1 ) for 2 h at room temperature under static conditions whereas, Influence of the addition of an amphiphilic copolymer in PHBHV film PHA-mcl/PHBHHx-PEG (1%) Influence of the addition of an amphiphilic copolymer in PHA-mcl film for diatoms Phaeodactylum and Navicula, coated surfaces were incubated in 5 mL of a given suspension in a solution of SASW (10 5 cells mL −1 ) for 72 h in a phytotron.Adsorption times were set at 2 h for the bacterium and 72 h for the diatoms, respectively (Guennec et al. 2021).After these adsorption times, the coatings were removed and rinsed three times for 5 s in SASW solution to remove any non-adsorbed microorganisms.The remaining adsorbed bacterial cells on the coatings were then stained with 5 mM Syto 9 green (k excitation ¼ 485 nm and k emission ¼ 498-550 nm) for 20 min in the dark before the observation with the confocal microscope.The diatom cells were imaged via the autofluorescence of chlorophyll a (k excitation ¼ 633 nm and k emission ¼ 650-700 nm).Images of the adsorbed organisms were captured with a confocal microscope CLSM 710 from Zeiss (Zeiss, Germany).The percentage coverage was evaluated using Image J software (NIH, USA).The results reported are an average of a total of 30 independent measurements.Statistical analysis of the percent coverage data was performed with GraphPad Prism 8.2.1.Error bars represent the standard deviation (SD) between the fluorescence measurements of 40 randomly selected regions on five microscope slides (8 measurements per slide).

Contact angle measurements
Contact angles (h) were measured by a Digidrop (GBX, UK) at room temperature (23 � C, RH 50%).The droplet volume was fixed at 3 mL and the analysis time was fixed at 60 ms.Three liquids were used for the contact angle measurements including ultrapure deionized Water (MilliQplus), formamide, and diiodomethane (TCI, purity � 98%).The surface energy was evaluated using the Owens-Wendt method, in which the surface energy is the sum of the hydrogen, or polar, surface energy (c s P ) and the dispersive van der Waals surface energy (c s d ).
Surface energy was then determined by solving graphically the Owens-Wendt Equation ( 2) Where h is the contact angle of the liquid on the surface, then c L , c L P , and c L d are the surface tension of the liquid, the polar component, and the dispersive component of the liquid surface tension, respectively.The surface energy of the coatings was then obtained by plotting the graph ð1þcoshÞc L 2 ffi ffi ffiffi ), from which a linear regression was realized to extract the slope corresponding to the dispersive component of the coating surface energy, and the origin coordinates corresponding to the polar component of the coating surface energy.
Captive bubble method was used to evaluate the hydrophilic/hydrophobic properties of the coatings immersed in MilliQ water.Measurements were taken at room temperature with a contact angle meter Digidrop GBX equipped with a homemade cell.The polymers films were immersed in deionized water in a transparent homemade glass cell for a period of 20 min before measurements.Five air bubbles of 3 lL were dropped at various locations on each coating.

Atomic force microscopy
AFM images of the polymer films were obtained with a Multimode 8 AFM instrument equipped with a NanoScope V controller (Bruker, Santa Barbara, CA, USA), operated in the Peak-Force Quantitative Nanomechanical mode (PFQNM).In PFQNM, the piezo was vertically oscillating at a frequency of 2 KHz, with an amplitude set between 70 and 150 nm.As the piezo scanned the samples in X and Y directions beneath the tip (with a scan velocity < 2 mm s −1 ), a force-distance curve was recorded at each pixel with a constant peak force setpoint.The peak force set-points were typically set to 500 pN for the PHBHV sample and to 50 pN for the PHA-mcl and PDMS surfaces.Images were recorded under ambient conditions (at 23 � C and RH of 50%), using Scanasyst-Air tips (Bruker, Billerica, MA, USA) for the PDMS and PHA-mcl films and RTESP-525 tips (Bruker, USA) for the PHBHV films.A precise calibration of the spring constant of the cantilevers was obtained from the thermal tuning method for the Scanasyst tip or using the Sader method for the RTESP-525 tip.Their deflection sensitivities were determined from the linear part of an average of 10 force-distance curves obtained by ramping the tips onto a Sapphire surface.Their tip radius was calibrated using the relative method based on references samples (PFQNM SPM kit-12 M, Bruker), a PDMS standard of 2.5 MPa for the Scanasyst tip, and a polystyrene film of 2.7 GPa for the RTESP-525 tip, respectively.Indentation modulus was calculated according to the Derjaguin-Muller-Toporov (DMT) model (Derjaguin et al. 1975) (i.e.Hertz model including the effect of adhesion without any change in the contact area).
All AFM data analysis and image processing were made with NanoScope software version 1.8.To ensure a good reproducibility in the sample preparation and surface properties of the polymer films (topographical and nanomechanical), for each sample, at least three different coated surfaces were prepared and observed, and on each surface of a sample, a minimum of three different areas were analyzed for each sample.

Bacterial adhesion tests
First adhesion tests were carried out with bacteria since they are representing one of the primary colonizers of immerged surfaces.Their adsorption allows further colonization by higher organisms, such as barnacle larvae or algae spores (Dobretsov and Rittschof 2020).However, the presence of bacteria and their molecular signals are not essential for the establishment of larvae of higher organisms.Indeed, the linear or chronological succession of the fouling model is not representative of the complex fouling dynamics in nature (Callow and Callow 2002;Huang and Hadfield 2003;Rajitha et al. 2020).
The marine bacterial strain Bacillus 4J6 sp. has a length of 2.05 mm and is particularly adherent thanks, in particular, to the production of numerous exopolysaccharides (EPS) (Grasland et al. 2003).Adhesion results for the coatings are presented in Figure 1.Results obtained for the glass slide and the coatings without additives (PDMS, PHBHV, and PHA-mcl) in the presence of the marine bacterium Bacillus 4J6 sp.revealed: (i) a strong adhesion on the glass reference surface (14.9% coverage) thus validating experimental conditions to study the adhesion of bacterial cells; (ii) a low adhesion on the second reference surface, namely PDMS.A low coverage rate (0.7%) was measured and agreed with the literature data regarding the efficiency of this type of hydrophobic coatings; and (iii) both PHBHV and PHA-mcl coatings revealed anti-bioadhesion activity with respectively 0.6 and 0.4% coverage.PHBHV coating was therefore as effective as PDMS and PHA-mcl presented very interesting properties with the lowest bacterial colonization.
The addition of amphiphilic copolymers, which were expected to improve the anti-adhesion properties of the coatings, was really effective.A 50% decrease in the amount of adhered bacterial cells was observed for the PHBHV films containing the PHBHHx-PEG copolymer.The same trend was observed for PHA-mcl.For PDMS coatings, the result was even more important, with a significant decrease (85%, p < 0.0001) in the amount of bacterial cells adhered to the PDMS/PDMS-PEG coatings as compared to the neat PDMS coatings.
This first study showed that the PHA-mcl film is more repellent to the bacterium Bacillus 4J6 sp.than the PDMS polymer film.When coatings are boosted with the addition of PEG copolymers, the efficiency of both PHA-mcl and PDMS amphiphilic films is enhanced and the two films show similar trends and become the most effective against the adhesion of bacteria.

Diatom adhesion tests
The second adhesion tests were realized with two diatoms, P. tricornutum and N. perminuta.Diatoms are photosynthetic microalgae that live in salt or freshwater and generally have a silica-rich cell wall.In biofouling processes, diatoms are early and efficient colonizers producing extracellular polymeric surface (EPS) (Laviale et al. 2019).The P. tricornutum cell is between 1 and 20 mm in size.It is a widespread diatom, found both in coastal areas and inland (Rushforth and Brock 1991).In the context of the study, this diatom is an interesting species since the ovoid form has an actively mobile benthic lifestyle, known to secrete EPS and thereby adheres to substrates (De Martino et al. 2011;Willis et al. 2013).Navicula perminuta cell is 8-20 mm long and 3-4 mm wide.Little described this diatom is still one of the most used microorganisms in research work related to adhesion and/or biofilm formation.Indeed, studies have shown that N. perminuta adheres more strongly on hydrophobic surfaces, such as elastomers than on hydrophilic surfaces (Holland et al. 2004;Wanka et al. 2021).Adhesion tests on both diatoms are represented in Figure 2 (P.tricornutum) and Figure 3 (N.perminuta).
For P. tricornutum, the glass surface was highly colonized, with 3.1% coverage on average.Significant adhesion was also observed on the two hydrophobic coatings, namely PHBHV and PDMS, with coverage rates >2%.On the opposite, PHA-mcl retained good efficiency with only 0.8% recovery (i.e. a reduction of 74% compared to glass).Once again, the neat matrix PHA-mcl showed remarkable anti-adhesive efficiency against this type of cell.For both PHA coatings, the addition of PHBHHX-PEG amphiphilic copolymer allowed a reduction in adhesion of around 70-75% compared to the neat polymer coatings.For PDMS, the reduction was greater and reached 96%.It is important to note that PDMS/PDMS-PEG and PHA-mcl/ PHBHHx-PEG coatings exhibited similar performance.
For N. perminuta, the glass surface was highly colonized with an average of 4.1% coverage.On hydrophobic neat polymer films, the adhesion of N. perminuta was reduced: (i) about 50% of reduction in PHBHV films as compared to glass, (ii) a 77% reduction in PDMS, and (iii) a 90% reduction in PHA-mcl coating which revealed an exceptional efficiency.For the PHBHV coatings, the addition of amphiphilic copolymer significantly reduced the adhesion of the diatom cells (48% as compared to neat PHBHV film).On PDMS coatings, the reduction reached 87% as compared with neat PDMS, thus reaching a similar low organism coverage as that obtained with the marine bacterium.For PHA-mcl, the addition of the PEG derivative allowed an 82% reduction in colonization compared to its neat matrix, with similar performance as the amphiphilic PMDS coating reference.
The results revealed that the three native coatings have a higher affinity for P. tricornutum than for N. perminuta.This difference could be attributed to their adhesion mechanism and to a secretion of exopolysaccharides over time.Indeed, the adhesion time used here was 72 h.It is possible that the secretion of proadhesive molecules in N. perminuta could be delayed over time, which could account for the difference in surface coverage between the two diatoms.When the PHA-mcl and PDMS matrix are modified with PEG, the two amphiphilic coatings are revealing the same performances for the two diatoms, with % coverage as low as those obtained for the bacteria.
The results obtained are partly at odds with certain works present in the literature showing increased adhesion of diatoms on hydrophobic surfaces (Holland et al. 2004;Statz et al. 2006;Wanka et al. 2021).This discrepancy could be explained by different experimental conditions: cellular concentration of the diatoms inoculum, static or dynamic adhesion conditions, temperature, composition of the surrounding medium, and composition of commercial PDMS coatings (presence of additives).

Surface energy of the films
The hydrophobicity of the coatings was investigated by water contact angle (WCA) measurements.It was observed that the most hydrophobic polymer coating is the PHA-mcl film with a WCA of 106 � , then followed by the PDMS (WCA of 100 � ) and by the PHBHV (WCA of 89 � ) These measurements are in good agreement with other works from the literature data (Insomphun et al. 2017;Gillet et al. 2018).
An advanced chemical characterization of surface materials is commonly achieved by wetting properties through the surface energy parameter.This parameter provides information on the chemical interactions a surface can have with another object, including the dispersive or non-dispersive interactions.Surface energy can be estimated by analyzing the wetting properties of a surface with different liquids which can measured by contact angle measurements using the sessile drop method.Figure 4 gives the surface energy components estimated from the Owens-Wendt theory for the three neat polymer films.The results show that the PHBHV film had the highest surface energy with a value of 34.8 mN m −1 whereas both PDMS and PHA-mcl films have a surface energy of about 20.5 mN m −1 .Differences in surface energy between the two PHA films were well correlated with other works from the literature, that have shown that the presence of a longer alkyl chain in the PHA monomer structure is responsible for a surface energy decrease (Shishatskaya et al. 2019).With the exception of the PDMS surface, the comparison of the surface energy of PHA films with data from the literature is difficult and must be cautious.Indeed, PHA film surface energy is highly correlated with the composition of the polymer (length of lateral alkyl chain) and with the film preparation, that itself governs the polymer structuration, such as the crystallisation content and possible polymer ageing.Nonetheless, close values of surface energy, between 36 and 40 mN m −1 could be found for PHBHV films (Menzyanova et al. 2019;Shishatskaya et al. 2019).The surface energy of PHA-mcl film has not been reported in the literature to the best of our knowledge, but Mauclaire et al. (2010) reported a water contact angle on polyhydroxyoctonaote (PHO film) of 108 � , in good agreement with our value (106 � ).As a reminder, the octanoate chain is the main monomer in PHA-mcl copolymer.From the Owens-Wendt theory, it can also be observed that the polar contribution of the surface energy, for the three polymer films was small (<1 mN m −1 ), underlying the importance of the hydrophobic character of the films in limiting the adhesion of microorganisms.
When the polymer films are doped with their corresponding PEG copolymer, the first attempt was to measure the water contact angles using the sessile drop method, however, no variation of the contact angle could be observed under the operated experimental conditions (time of measurement fixed at 60 ms).Upon immersion of the doped coatings, the PEG macromolecules are likely to organize at the surface of the coatings and modify their interfacial energy (Hawkins et al. 2014).To demonstrate the role of the addition of PEG into the native polymer film (amphiphilic coatings), more relevant conditions could be obtained by measuring the hydrophobicity by the captive bubble method.Figure 4b shows that when PEG is added, even in the low amount of 1%, an increase in the angle is observed, indicating an increase in the hydrophilic nature of the doped films due to the presence of PEG chains on the surface of the coatings.After 20 min, no further variation in the contact angle could be measured on the different coatings, up to 3 days of immersion corresponding to the maximum adhesion time used in this study.

Topography of the films and their indentation modulus
Typical AFM images and section analysis of the coatings are shown in Figure 5.As can be observed, the two PHA samples show semi-crystalline structures.The PHBHV films are organized in large spherulites of about 20 mm diameter that could be observed at a larger scale (Guennec et al. 2021).The image in Figure 5a shows crystalline lamellae organized in a radial distribution, starting from the center of the spherulites and spreading up to the intersection with other spherulites.Consequently, due to the presence of this high crystalline content, the Z variation on the PHBHV surface was found to be about 80 nm over 1 mm of distance (x direction).In comparison, PHAmcl films consist of smaller crystallites (Figure 5b), arranged together in a fairly smooth film with typical Z variation (over 1 mm) of about 8 nm, i.e. 10 times less than for the PHBHV film.Finally, the PDMS chains polymers are organized in a typical mesh-like structure with typical pore sizes of about 15 nm, yielding a relatively smooth polymer film with Z variation (over 1 mm) about 6 nm, i.e. comparable to the height variations of the PHA-mcl coating (Figure 5c).AFM height images of PEG doped coatings are given in supplementary materials (Figure S1) but were not different from those obtained on the neat polymer films.This finding could be explained by the observation conditions (ambient conditions, 23 � C, RH 50%) of the dried films but can also be due to the small amount of PEG additives in the film.The variation of the RMS for the coatings, as a function of the area, is presented in Figure 6.Results show that the RMS of both PHA-mcl and PDMS samples were found to be about 3-5 nm over a large scanning area (up to 100 mm 2 ), whereas, for the PHBHV sample, the RMS started at 25 nm and increased up to 100 nm for a 100 mm 2 .The nanomechanical properties of the coatings were investigated using the PFQNM technique, in their dry state.Results, presented in Figure 7, show a significant difference between the indentation modulus of the PHBHV and the PHA-mcl coatings, with more than 100 orders of magnitude.This finding can be related to the high crystalline content of the PHBHV film.With an indentation modulus of 15 MPa, despite the presence of crystallites within the film, the PHA-mcl coating had a modulus of only �10 times that of the PDMS (indentation modulus of 1 MPa).The obtained values for the indentation modulus of the PHA surfaces were in good agreement with literature values (Sofi� nska et al. 2019).Similarly, as for the topographical analysis, no difference in the indentation modulus was observed between the neat and PEG-doped coatings.

Discussion
The physico-chemical characterization of the films at the local scale allows us to compare three important parameters that are essential in designing polymer surfaces with effective anti-biofouling properties.The results obtained on the polymers showed: (i) a significant adhesion of the three organisms to the glass, which confirms their adhesion capacities, (ii) a greater efficiency of the neat PHA-mcl matrix, greater than that of the neat PDMS reference, and the neat PHBHV film, and (iii) a major impact of the addition of amphiphilic PEG copolymers in the three matrices that boost the coatings efficiencies.The best performances are similarly obtained with the PHA-mcl and PDMS doped coatings with a markedly reduced adhesion rate, whatever the organism.Characteristics known to influence the anti-adhesion effectiveness of surfaces include hydrophobicity, roughness, and Young's modulus.For the sake of clarity, the following discussion will initially focus on the differences between the neat PHA-mcl and PDMS coatings, followed by the characteristics of the highly-crystallized  PHBHV film and finally the impact of PEG addition in the different matrices.
Surfaces with an RMS value below the critical value of 6 nm, as defined in Yang's review (Yang et al. 2022), are generally known to have a negative or no correlation with the amount of adsorbed cells.It is accepted that a reduction in surface roughness may act against cell adhesion, by reducing the available surface area (Volant et al. 2021) although this behavior is not systematically observed, especially on nanostructured designed surfaces (Singh et al. 2011;Liu et al. 2016).Both the PHA-mcl and PDMS films were found to be smooth surfaces, with RMS values between 2 and 4 nm over a large area (100 mm 2 ) which is in good agreement with the literature by G€ okaltun et al. (2019) andSofi� nska et al. (2019).Their similar roughnesses suggest that this parameter is not the primary factor elucidating the superior performance of PHA-mcl in comparison to PDMS.
Stiffness is another parameter mentioned as influencing bacterial adhesion and biofilm formation (Brady and Singer 2000).If previous studies have attempted to investigate the relationship between bacterial adhesion and surface elasticity, the literature data remains limited and disputed.The prevailing trend is when a cell encounters a soft substrate, its adhesion seems to be restricted; although a clear explanation for this phenomenon has not yet been found.For example, using poly(ethylene glycol) dimethacrylate (PEGDMA) and agar hydrogels as coatings and Escherichia coli and Staphylococcus aureus as model bacteria, Kolewe et al. (2015) found that bacterial adhesion increased with material stiffness.Another study using agarose hydrogels demonstrated a positive correlation between adhesion and stiffness (Gu� egan et al. 2014).However, studies using polydimethylsiloxane (PDMS) have mostly shown a negative correlation between stiffness and adhesion (Song and Ren 2014;Song et al. 2017;Straub et al. 2019;Arias et al. 2020); except for the study conducted by Peng et al. (2019).It should be noted that studies of more or less cross-linked PDMS films have two main biases: (i) the number of free chains and (ii) the number of chain ends.These highly mobile molecular entities play a crucial role in interactions with microorganisms that condition bioadhesion (Valentin et al. 2019).In this study, PDMS and PHA-mcl coatings elastic moduli ranged from the MPa range, with a higher modulus for the PHA-mcl coating (10 times superior to the indentation modulus of the PDMS coating).Despite this difference, it should be noted that these polymer films have elastic modulus typical of elastomers and low Tg polymer materials in contrast with some conventional plastics with elastic moduli in the GPa range, such as polystyrene (2.5 GPa) or polyethylene (0.8 GPa).Both coatings show interesting anti-adhesive properties with better performances obtained for PHA-mcl films.Since both materials are within a range of low modulus, one should better consider the difference in modulus between the two objects in interaction, i.e. the coating surface and the outer membrane of the microorganisms.An estimation of bacteria cell envelope Young's modulus is given to be around 50-200 MPa (Tuson et al. 2012), thus slightly higher than the modulus of the two soft coatings.
The last parameter to be considered is the hydrophobicity of the surfaces.When considering hydrophobic coatings, their low surface energy eliminates the possibility of strong polar interactions with biomolecules or microorganisms.Baier (1970) has proposed that polymer films with surface energy in the range of 20-30 mN/m were the best candidates for preventing bioadhesion.In this study, the PDMS and PHA-mcl films have similar surface energy within this range, which could explain their good performance but not their difference.
At this point, one can infer that the strong performance of PHA-mcl and PDMS can be attributed to their low roughness, low Young's modulus, and surface energy falling within the favorable range of 20-30 mN m −1 (Baier 1970).The issue is that these three parameters do not appear to be the parameters capable of interpreting the variation in adhesion between PDMS and PHA-mcl.In addition, the antiadhesive properties of PHBHV are similar to those of PDMS, despite its high roughness (over 6 nm), modulus (in the GPa range), and surface energy (over 30 mN m −1 ).The three polymer films differ in that PDMS is purely amorphous (but reticulated), PHA-mcl is weakly crystallized and PHBHV is highly crystallized.The question then arises as to the influence of crystallization on adhesion.A literature review confirms an influence of partial crystallinity on cellular response (Park and Cima 1996;Smith et al. 2020;Kołbuk et al. 2022).These few studies show that the rate of cell (fibroblasts) proliferation was faster in the predominantly amorphous sample.Nevertheless, in most of the studies, it is challenging to discern the primary factor responsible for enhanced cell adhesion (Kołbuk et al. 2022).Yet crystallinity is a characteristic of the structure that influences the surface properties of semicrystalline polymers related to cell adhesion like for example profile roughness, T g , as well as types and strength of chemical bonds (Reiter and Strobl 2007).To illustrate the influence of crystallization, a few of these parameters can be discussed.For example, RMS roughness measures variations in height relative to a mean line.However, at a similar RMS value, it is evident that the nanoscale surface morphology of a crystallized polymer differs from that of an amorphous polymer.Amorphous polymers have random, entangled chains, while semi-crystalline polymers are structured.Other parameters than RMS could be used to characterize such surface topographies like for example the peak to peak distance or peak density, as reported by L€ udecke et al. ( 2016) when studying nano-structured Titanium substrates.The presence of surface irregularities, like peaks, is thought to promote the number of contact points between microorganisms and the surface (Li et al. 2023).
The glass transition temperature is also an important parameter influenced by crystallinity.In semi-crystalline polymers, T g depends on the degree of crystallinity, in that its mobility is drastically restricted by geometric constraints at the interface with the crystals (Monnier et al. 2017).Concerning the three polymers, regardless of the origin of these varying T g values, their distinct T g indicates varying degrees of polymer chain mobility at room temperature.However, the exact organization of polymer chains at the subsurface of the coatings, including free-end chains, dangling chains, or cross-linking chains is not known.Differences in the polymer organization between the PDMS and PHA samples are likely to exist to explain their anti-adhesion performances (reticulated polymer network vs. semi-crystalline polymers).Semi-crystalline coatings can be seen as heterogeneous surfaces with crystallized and amorphous parts that generate ambiguous properties (topography, mechanical properties, chain mobility).
The addition of an amphiphilic copolymer of PEG type within the polymer matrices strongly increased their efficiency while no variation in the RMS or in polymer surface modulus was measured.By using the captive bubble, it is possible to demonstrate a variation in the contact angle caused by PEG addition.It is generally described that the anti-adhesive effect observed with PEG grafted chains on surfaces is due to the entropic effect resulting from incompressibility of the swollen and highly mobile PEG chains (T g of −100 � C) upon the approach of other molecules (Szleifer 1997).This interpretation is not perfectly suited to the studied systems, since the concentration of PEG polymer is very low and the PEG copolymer is not grafted but added to the hydrophobic polymer matrix.In this case, the notion of heterogeneous/ ambiguous surface is more suited to describe the PEG-doped surfaces.Indeed, these surfaces remain globally hydrophobic but are more hydrophilic than their native polymer films.This underlines the presence of PEG chains in low amount on the surface of the films, creating a chemical heterogeneity.It seems that the chemical composition of the surface at the microorganism scale, particularly its heterogeneity, is a significant factor that interferes with the adhesion of bacteria and microalgae.Heterogeneity can be generated by hydrophilic/hydrophobic spaces via the PEG copolymer on the surface, or by crystalline/amorphous parts.Indeed, many questions arise regarding the perception of a surface by a microorganism: what are the parameters perceived?What is the spatial resolution?In what range is intensity?(Szleifer 1997)

Conclusion
This study explored the potential of polyhydroxyalkanoate films to resist the adhesion of marine microorganisms, such as bacteria and diatoms.In this study, two PHAs, one scl, and the other mcl, were compared to PDMS.Results showed a better efficiency of the matrix based on PHA-mcl in comparison with the PHBHV sample and the reference PDMS sample.The analysis of antifouling results is grounded in key physico-chemical properties known to influence the bioadhesion of coatings, including surface energy, roughness, and mechanical characteristics.However, these parameters alone do not provide a comprehensive explanation for all the results, prompting a discussion of PHA crystallinityrelated properties.Overall, the anti-adhesion performances of the polymer coatings were boosted by adding PEG copolymer.This research underscores the significance of surface physical and chemical heterogeneity, related to semi-crystalline or amphiphilic coatings, in determining anti-adhesive properties.Furthermore, the potential of medium chain length PHA films to resist the initial adhesion of microorganisms is highlighted.

Figure 1 .
Figure 1.Surface coverage by Bacillus 4J6 after 2 h of incubation.(a) Representative confocal microscopy images of the adsorbed bacteria (marked with Syto 9 Green) on raw polymer films.Scale bar is 20 mm.(b) Graphic representation of the percentage of coverage for the coatings.

Figure 2 .
Figure 2. Surface coverage by Phaeodactylum tricornutum after 72 h of incubation.(a) Representative confocal microscopy images of the adsorbed diatoms on raw polymer films.Scale bar is 20 mm.(b) Graphic representation of percentage of coverage for the coatings.

Figure 3 .
Figure 3. Surface coverage by Navicula perminuta after 72 h of incubation.(a) Representative confocal microscopy images of the adsorbed diatoms on raw polymer films.Scale bar is 20 mm.(b) Graphic representation of percentage of coverage for the coatings.

Figure 4 .
Figure 4. (a) Surface energy components (in mN/m) for the native polymer films, as estimated from Owens-Wendt approach by measuring contact angles with liquids of different polarity (water, formamide, and diiodomethane).(b) Graph representing the water contact angles obtained on the native polymer films, as measured with the sessile drop method (white part) and the water contact angle, as measured with the captive bubble method (dashed part), on the two PHA samples, either native and doped with PEG (1%).

Figure 5 .
Figure 5. AFM height images of native polymer surfaces (a: PHBHV, b: PHA-mcl, and c: PDMS) showing topographical aspects of films and corresponding profiles showing the Z variation along a line.The roughness of the polymer surfaces is determined from at least five different representative AFM images on each sample.

Figure 6 .
Figure 6.Comparison of roughnesses of the polymer films as a function of the scanned area (1, 25, and 100 mm 2 ), corresponding to the size of the adsorbed micro-organisms (1 mm for bacteria and 10 mm for diatoms).

Figure 7 .
Figure 7. (a) Example of AFM height image of the PHA-mcl film scan area of 1 � 1 mm.(b) Corresponding indentation modulus PFQNM image, calculated using the DMT model-the Z bar is 38 MPa.(c) Section analysis of indentation modulus image showing the variation in E along the PHA-mcl surface.(d) Graphic representing the indentation modulus (in MPa) for the polymer films.

Table 1 .
Characteristics of the PHA samples.

Table 2 .
Role of the coatings in this study.