Potato virus A particles – A versatile material for self-assembled nanopatterned surfaces

Potato virus A (PVA) is a plant-infecting RNA virus that produces flexible particles with a high aspect ratio. PVA has been investigated extensively for its infection biology, however, its potential to serve as a nanopatterning platform remains unexplored. Here, we study the liquid crystal and interfacial self-assembly behavior of PVA particles. Furthermore, we generate nanopatterned surfaces using self-assembled PVA particles through three different coating techniques: drop-casting, drop-top deposition and flow-coating. The liquid crystal phase of PVA solution visualized by polarized optical microscopy revealed a chiral nematic phase in water, while in pH 8 buffer it produced a nematic phase. This allowed us to produce thin films with either randomly or anisotropically oriented cylindrical nanopatterns using drop-top and flow-coating methods. Overall, this study explores the self-assembly process of PVA in different conditions, establishing a starting point for PVA self-assembly research and contributing a virus-assisted fabrication technique for nanopatterned surfaces.


Introduction
Self-assembly of nanoparticles into organized structures has demonstrated its potential to benefit various technical fields, including electronics (Fendler, 2001), optics (Murphy et al., 2005), sensing (Thorkelsson et al., 2015), energy storage (Frey et al., 2009), and biomedical applications (Thanh and Green, 2010).Precise control over the self-assembly process relies upon the structural properties and the surface chemistry of the unit blocks, as well as the external conditions during the process (Linko et al., 2022).In general, bottom-up fabrication methods can offer an alternative route to overcome the challenge of top-down photolithography in producing high spatial resolution patterns at low cost (Nie and Kumacheva, 2008).Over the last two decades, virus nanoparticles (VNPs) have attracted focused attention and emerged as promising candidates for nanotechnology applications, including nanopatterning (Pokorski and Steinmetz, 2011;Korpi et al., 2020).As a component, when compared to the more commonly employed synthetic self-assembling materials such as block copolymers, VNPs provide superior assembly properties due to their uniform size, available surface functionalization possibilities, stability, and inexpensive purification on a large scale (Lee et al., 2009).
VNPs come in various symmetrical shapes, including rod-like and spherical structures, as well as a broad range of sizes ranging from 10 nm to a few microns.At the laboratory scale, VNPs can be prepared in milligram to gram-level quantities with high uniformity in composition, shape, and size (Pokorski and Steinmetz, 2011).The protein shell of VNPs, known as the capsid, acts as a protective layer for the genome and enables VNPs to resist temperature and pH changes.Since the capsid consists of several protein subunits that can be covalently or non-covalently bonded with a large variety of functional groups, VNPs provide various surface modification and functionalization opportunities (Smith et al., 2013;Chen et al., 2016;Kostiainen et al., 2013;Liljeström et al., 2017).Therefore, VNPs have been expandingly engineered for tissue engineering (Zhao et al., 2015), drug delivery (Chung et al., 2020;Esfandiari et al., 2016), and vaccine design (Plummer and Manchester, 2011;Sulczewski et al., 2018).In the context of nanoscale patterning, generating a thin film of well-organized VNPs on a solid substrate remains challenging.Drop-casting (Tseng et al., 2006;Wargacki et al., 2008), spin-coating (Rink et al., 2017), and dip-coating (Yoo et al., 2011;Atanasova et al., 2015) are the most common methods for depositing VNPs.To obtain a stable film, a substantial electrostatic attraction between VNPs and substrate is usually required (Rong et al., 2011).However, this can lead to low ordering of the VNP patterns due to the initial strong electrostatic interaction with the substrate, leading to kinetically trapped structures (Knez et al., 2004;Lee et al., 2007).Furthermore, the coffee ring (Kaliyaraj Selva Kumar et al., 2020) and the edge effects (Puetz and Aegerter, 2004) associated with the above-mentioned coating techniques hamper the preparation of uniform structures and make them less appealing.
Most studies on the self-assembly of rod-like VNPs have been limited to tobacco mosaic virus (TMV) and M13 bacteriophage.The selfassembly of TMV (18 nm in width × 300 nm in length) has been known as a head-to-tail assembly process, in which the dipolar ends of the TMV's helical structure interact hydrophobically and complementarily with each other (Rong et al., 2011;Niu et al., 2006;Bruckman et al., 2007).Specifically, an acidic environment with pH of less than 4 was shown to promote the self-assembly of TMV into a fiber-like structure by reducing the repulsive carboxylate interaction (Santos et al., 2004;Lu et al., 1996).M13 (6 nm in width × 880 nm in length) is another standard model for the study of rod-like VNPs self-assembly.The lyotropic LC behavior of M13 is the reason behind its highly ordered organization (Park et al., 2021;Lee et al., 2003;Chung et al., 2011).In addition, different complex liquid crystal (LC) structures have been shown as a function of evaporation (Park et al., 2021), concentration (Lee et al., 2003), or the interplay between meniscus constraints and material flux during the coating process of dip-coated M13 films (Chung et al., 2011).
Unlike rigid rod-like TMV and semi-rigid M13, potato virus A (PVA) has a highly flexible structure, which can provide diverse opportunities for self-assembled patterns.PVA belongs to the genus Potyvirus, which is considered a primary plant pathogen family from an economic perspective (Valli et al., 2015).PVA VNPs can be prepared in planta with high reproducibility and uniform dimensions (Valli et al., 2015).They obtain a flexuous rod-like structure with a high aspect ratio, typically 680-900 nm in length and 11-13 nm in width (Martínez-Turiño and García, 2020).Their capsid comprises approximately 2000 protein subunits enclosing the genomic RNA (Martínez-Turiño and García, 2020), in which the exposed N-terminus can tolerate small amino acid insertion (Martí et al., 2022).Analogously to other VNPs, the exposed amino acid side chains in the capsid layer of PVA allow versatile chemical modification and functionalization.PVA is an extensively studied potyvirus and the number of the research works is constantly increasing, yet most have focused on the general biological aspects.Though there are few reports available on surface and structural characterization of PVA (Nault et al., 2015;Ksenofontov et al., 2018;Shtykova et al., 2021), the only applicative aspect of PVA that has been studied is as enzyme nanocarriers (Besong-Ndika et al., 2016).
Here, we expand the use of PVA and report on their self-assembly behaviour and utilization for surface nanopatterning.In order to be feasible, relatively large amounts of PVA VNPs need to be produced.We show the optimization of the purification process based on a pre-existed protocol developed by Gabrenaite-Verkhovskaya et al. (2008) Compared to the previous protocol, PVA VNPs with better quality and higher yield are obtained, and we discuss their stability with different pHs, buffer strengths, added salts as well as concentrations.The self-assembly of PVA in water and buffer solution results in liquid crystalline (LC) behavior and a model was developed to explain the effect of surface charge on the orientation of nanoparticles.Thin films of PVA with varied surface morphologies could be deposited onto a substrate by drop-top or flow-coating methods.These methods are discussed in detail regarding their feasibility and applicability.It should be noted that self-assembly of VNPs in this article refers exclusively to intermolecular self-assembly of VNPs.This work provides a straightforward pathway for generating surface patterns, expanding the knowledge of nanoparticle self-assembly to a new type of VNPs.

PVA VNP purification protocol
PVA VNPs were produced in Nicotiana benthamiana plants (NCBI: txid4100).The plants were grown at 22 • C under 16 h (light) and 8 h (dark) photoperiod in environmentally controlled greenhouses.Young plants at the 4-leaves stage were infected with Agrobacterium carrying PVA infectious complementary DNA (icDNA).The infections were then allowed to spread systemically for three weeks.Infected leaves were then harvested and stored at −80 • C. Full length infectious PVA VNPs were subsequently purified from the infected leaves following the protocol described in Fig. S1.

PVA stability assay
PVA samples of concentration 5 × 10 11 particles/μL were mixed with test subjects (pH, buffers, or salts) of desired molarities.Samples were incubated at room temperature for 96 h.After the incubation, cDNA was prepared from the particles using the H Minus First Strand cDNA synthesis kit (Product No. K1632, Thermo Fisher Scientific) with random hexamer primers.The particle concentration was further quantified via qPCR using a Maxima SYBR Green qPCR Master mix (Thermo Fisher Scientific) and a CFX96 TouchTM Real-Time PCR Detection System.PVA RNA quantification was carried out by amplifying the coat protein (CP) region of PVA RNA using primers-forward-5 ′ -CAT GCC CAG GTA TGG TCT TC-3 ′ , reverse-5 ′ -ATC GGA GTG GTT GCA GTG AT-3'.Percentage stability was calculated by comparing incubated particle quantity with control particles (cDNA prepared at 0 h).

PVA particle visualization
PVA particles and self-assembled structures were visualized using a Tescan Mira 3 scanning electron microscope (SEM); a Bruker Multimode 8 (Bruker) atomic force microscope (AFM) in non-contact mode using Silicon tips (NCHV-A, Bruker); and Jeol JEM-1400 and JEM-2800 transmission electron microscopes (TEM).α-CP antibody was coated on the TEM grid surface by incubating the carbon surface of the grid with the antibody at 1:100 dilution for 1 h at room temperature.

PVA liquid crystal analysis
Liquid crystal domains of PVA were visually analyzed using an Olympus BX53 M polarized optical microscope (POM).
Small and wide-angle X-ray scattering (SAXS, WAXS) data is obtained using a Xenocs Xeuss 3.0 SAXS/WAXS system (Xenocs SAS, Grenoble, France).The system includes a microfocus X-ray source (sealed tube) with a Cu target and a multilayer mirror which yields a parallel beam with a nominal wavelength of 1.542 Å (combined Cu K-α 1 and Cu K-α 2 characteristic radiation).The source operates at 50 kV and 0.6 mA.The beam is collimated by a set of variable slits and the beam size at the sample was 0.4 mm during the experiment.The system does not include a beam stop, which enables direct measurement of sample transmission.The background scattering from sample holder is normalized and subtracted from the data according to sample transmission.The data is acquired using an area detector (Eiger2 R 1 M, Dectris AG, Switzerland) that was in the evacuated chamber.The sample-to-detector distance was calibrated by measuring the diffraction from a known LaB 6 standard sample.
The samples were sealed in borosilicate Mark tubes (diameter 1.5 mm, wall thickness 10 μm, Hilgenberg GmbH, Germany) using hot glue and the tubes were in the evacuated sample chamber during the measurement (pressure <1 mBar).In addition, reference samples (empty Mark tube, solvent) were measured in the same conditions.Scattering contribution from the empty chamber, the capillary, and the solvent were sequentially subtracted from the azimuthally averaged data.

In silico analysis of charge distribution of PVA CP in different pH
Predicted model of PVA CP (strain B11, GenBank accession number AJ296311) was generated using I-TASSER server for protein structure and function prediction.The charge distribution model of PVA CP was generated using the APBS-PDB2PQR software hosted by a server (https://server.poissonboltzmann.org/).The model was visualized using the software ChimeraX 1.2.5.Isoelectric pH (pI) of PVA CP and net charge of the CPs at different pH conditions were determined using online server PROTEIN CALCULATOR v3.4 hosted by http://protcalc.sourceforge.net/.

Drop-top coating method
First, the TEM grids were coated with primary antibody against PVA CP (α-CP, SASA, UK) by incubating the grids with 1:100 diluted Petri dish solution for 1 h.The grids were then washed to remove excess antibodies.Meanwhile, a 20 μL drop of 3-5 mg/mL PVA in water or buffer was allowed to rest for 5 min on a Petri dish placed over ice.A α-CP coated grid was then instantly placed on the top of the drop with the help of a tweezer and removed.Finally, the grid was washed on the top of a drop of water or buffer exhaustively to reveal the monolayer of PVA VNPs bound to the grid.

Statistical analysis
Statistical significances were calculated between the marked samples and their corresponding controls using Student's t-Test.Only those cases where the calculated p-values were below 0.05 (p < 0.05) were considered significantly different from control.

Purification of PVA VNPs
In order to employ PVA VNPs in material science applications, it is essential to obtain high-quality purified PVA VNPs with sufficiently high yield.Therefore, we developed further the protocol for PVA purification to suit this purpose.The protocol steps are presented in Fig. 1A.The existing protocol for PVA purification was reported in the work of Gabrenaite-Verkhovskaya et al. (2008) In their method, the particles were divided into two fractions, soluble fraction (F) and precipitated fraction (P2).The F-fraction mainly comprised broken/defective VNPs, while P2 fraction had mostly intact VNPs.However, we found that the P2-fraction contained co-precipitated impurities that were very difficult to eliminate.Earlier, Choi et al. (1977) reported that potyvirus turnip mosaic virus (TuMV) was also prone to aggregation with host-derived contaminants during purification and this was the single most critical reason behind the loss in the particle yield.To deal with this issue, we included a chelating agent EDTA, a dispersing agent NaOH and a reducing agent β-mercaptoethanol in sodium phosphate buffer (pH 7.4).Compared to the original protocol for purification (Fig. 1B), the new protocol significantly altered the distribution of PVA VNPs within the sucrose gradient (Fig. 1C).PVA coat protein (CP) bands in Fig. 1C clearly demonstrate that the particle concentrations gradually increased with increasing sucrose concentration, and the maximum amount was concentrated in fractions 5 and 6.The precipitate, on the other hand, was largely depleted of VNPs.VNPs concentrated from pooled fractions 5 and 6 were intact, of appropriate size, devoid of insoluble debris (Fig. 1D, see also Fig. S2) and suitable for subsequent studies.

Stability behavior of PVA VNPs
Though there are several studies on PVA biology, there lies a substantial knowledge gap in understanding the properties of PVA VNPs from a material science perspective.Our initial objective was to study the stability of PVA VNPs in different pH, buffer strengths, salts, and concentrations.Since RNA is highly unstable in room temperature solutions and no protective measures for stabilization of the viral RNA (vRNA) were employed, we assume that the remaining RNAs after the incubation period originated from intact PVA VNPs.Therefore, the RNA quantity can be used as a direct indicator of intact PVA quantity.Therefore, after 96 h of incubation at room temperature, we prepared complementary DNA (cDNA) from the PVA samples and calculated the percentage of remaining vRNA by comparing it with the cDNA samples prepared at 0 h.Considering RNA stability as a measure of particle stability, Fig. 2 reveals that the VNPs were the most stable at pH 8.8 and 9.5 (Fig. 2A) and low buffer/ionic concentration (Fig. 2B and C).The effect of all the salts tested here at 0.1 M had a negative impact on the particle stability, however the effect of divalent metal salts i.e., CaCl 2 , MgCl 2, and MnCl 2 were more prominent (Fig. 2C and D).

PVA self-assembly behavior
Next, we studied the self-assembly behavior of PVA in water (PVAwater ) and pH 8 buffer (PVA buffer ).As for buffer, we chose ammonium acetate buffer (0.1 M) to avoid the artifacts caused by the concentration and crystallization of salts (as compared to HEPES).Our drop-casting protocol consisted of dispensing a drop of 3 mg/mL PVA in either solvent onto a Si wafer and allowing it to dry (Fig. 3A).Then, we investigated the surface features of the resulting films using SEM and AFM.As observed through both of these methods, the surface of PVA water film demonstrated bundles of multiple PVA VNPs arranged in all lateral directions (Fig. 3B,D).On the other hand, SEM and AFM images of PVAbuffer film surface revealed a parallelly arranged PVA particles (Fig. 3C,E) forming a relatively smooth film.
Each PVA particle is made up of approx.2000 copies of coat proteins.
Being proteinaceous in nature the VNPs carry positive and negative charge in buffer below and above pI respectively, while at pI the VNPs are neutral.As the structure of PVA VNPs are not yet resolved, we did an in silico modelling of PVA CP to visualize its charge distribution at different pHs (Fig. 3A).Calculated pI of PVA CP was predicted to be 7.02.Since the pH of deionized water is usually slightly below pH 7 and buffer was of pH 8, we then theoretically estimated net charges at pH 6, 7, and 8. Consistent with our assumption, the charges were 5.7, 0.1, and −2.7 units respectively (Fig. 3A).We characterized the LC behavior of PVA water and PVA buffer by polarized optical microscope (POM) and small-angle x-ray scattering (SAXS), which are complementary methods as POM describes alignment on micrometer level and above whereas SAXS provides information about particle shape and packing in the 1-100 nm range (Dogic and Fraden, 2000).
Apart from the two PVA water samples with a particle concentration of 50 mg/mL, the SAXS profiles of PVA water and PVA buffer samples correspond directly to the form factor P(q) of the particles.Thus, according to SAXS, the structure of these samples was weakly correlated, or the correlation distance is larger than which can be observed with the SAXS instrument used.However, the PVA water sample with the highest concentration shows clear deviation from P(q), as a correlation maximum (a "shoulder" at small q) was observed, which indicates a periodic positional order (Fig. 4D and E).The correlation maximum represents the typical center-to-center distance d cc in the lyotropic LC phase, hence these samples exhibited both orientational order (POM, discussed below) and positional order (SAXS).Moreover, the position of the correlation maximum (q ~0.012-0.014Å −1 ) translates to a d cc of 45-52 nm.Here we assume that the rod-like virus had an effective density similar to that of water.Providing that the virus concentration was 50 mg/mL, this condition would yield a theoretical d cc of 51 nm, which corresponds to a continuous 2D close-packing of rods (Fig. S3a).This result is in strong agreement with an assumption of repulsion induced LC ordering.However, the underestimation of the unknown effective density makes this result slightly inaccurate.The effective density is most probably a little higher than that of water, which means that the theoretical d cc of 51 nm should be higher.
According to POM, all samples exhibited local LC nematic ordering with orientational order (Figs.S3b and c).This conclusion is based on the observation that all samples showed bright areas when observed between crossed polarizers.In particular, POM images of PVA water LC at 15 mg/mL concentration demonstrated clear signature of chiral nematic LC phase i.e., fingerprint like alternating bright and dark bands (Fig. 4B).(Dogic and Fraden, 2000) Higher magnification POM images reveal large alternating domains of bright and dark bands for PVA water compared to thin filament-like continuous phase for PVA buffer (Figs.S3b  and c), which demonstrates the difference between their LC behaviour.Alternating bright and dark bands, i.e., a "fingerprint pattern" is typical for a chiral nematic LC, where the director of the nematic phase alternates between being perpendicular to the direction of observation (bright area) and being parallel to the direction of observation (dark area).The pitch of the chiral nematic LC, which is double the period of the fingerprint, is more than 100 μm, meaning that the LC phase is not strongly twisted, thus d cc is expected to be higher.Also, PVA buffer showed both bright and dark areas when observed between crossed polarizers (Fig. 4C and Fig. S3c).Interestingly, the characteristic domain size (area with uniform brightness) was less than 10 μm, showing that the alignment of the particles had no long-range correlation.Therefore, the local arrangement is more likely due to the high aspect ratio of the particles, rather than the long-range interaction.
Chiral nematic LC phase has earlier been reported for semi-flexible viruses like M13 and Fd phage (Dogic and Fraden, 2000).On the other hand, pH dependent transition of chiral nematic to nematic LCs at low and high pH has been reported for cellulose nanocrystals (CNCs).Electrostatic properties of the solution and their interaction with CNCs were deemed responsible for this phenomenon (Li et al., 2019).Previously, phenylboronic acid (PBA) functionalized M13 phage was shown to transit from chiral nematic LC to nematic LCs, based on whether the pH of the solution was above or below the pKa of PBA (Cao et al., 2014).
In our case, compared to water, buffer at pH 8 could influence the charge distribution of the VNPs (positive to negative surface charge), electrostatic properties of the solution, and protonation of histidine (Pogostin et al., 2019).It is clear that the ionic strength of the solution affected the electrostatic repulsion between particles.Debye length λ D , which is the characteristic range of electrostatic interaction in solution, strongly depends on the ionic strength.An ionic strength of 0.1 M would yield λ D ~1 nm, and for pure water (ionic strength 10 −7 M) λ D ~ 1 μm.In other words, the electrostatics interactions were significant in pure water and minimal in buffer.The aforementioned factors play a role in differential LC behavior of PVA in water and buffer and are responsible for the LC behaviors behind contrasting self-assembly of PVA water and PVA buffer (Fig. 3B-E).

Translating PVA LCs to self-assembled nanopatterned thin film
As discussed earlier, top layer of PVA buffer drop-casted films showed smooth and aligned stripe patterns consisting of individual concatenated PVA VNPs.We hypothesize that, as the droplets begins to evaporate, concentration of PVA partially increase in the surface of the drop where the effect of evaporation is higher.During this process adjacent VNPs assemble in the most favorable fashion, i.e., chiral nematic LC for PVA water and nematic LC for PVA buffer .In order to translate the LC assemblies from the droplet surface into solid state nanopatterned thin film, TEM grids coated with PVA α-CP antibody were utilized.This droptop coating method is illustrated in Fig. 5A.The resulting PVA surface pattern from PVA water and PVA buffer samples were imaged with SEM, as shown in Fig. 5B-C.Consistent with our assumption, we observed in the PVA water samples bundles of PVA in all directions (Fig. 5B).PVA buffer , on the other, hand generated more directionally oriented PVA pattern (Fig. 5C).
We further washed the grids extensively to remove excess PVA from the grids to reveal the monolayer of PVA VNPs directly bound to the antibodies.Transmission electron microscopy (TEM) images of the PVA monolayer also agrees with our hypothesis.Owing to changing directors of the LC planes in PVA water , the VNPs are pointing towards all in-plane directions (Fig. 5D, also see Fig. 4D).The FFT diffractogram (Fig. 5D inset) further confirms the isotropic assembly of PVA monolayer on the grid.Similarly, TEM image and FFT diffractogram for PVA buffer also corroborates with our hypotheses and show anisotropic assembly of monolayer PVA via drop-top method (Fig. 5D).Though some misdirected VNPs could be seen across PVA monolayer, it could be attributed to imperfections in LC phases that got bound to the α-CP antibodies and could not be removed later even with extensive washes.We also validated that the anisotropic self-assembly behavior of PVA is a pHdependent phenomenon by comparing the PVA monolayer morphologies from PVA dissolved in ammonium acetate/HEPES buffer at pH8 (Fig. S4).

Towards PVA-based biotemplating applications
Drop-top coating method successfully generated organized selfassembled structures with PVA buffer , but its practical use could be limited due to difficult scale-up.Therefore, we explored flow-coating as an alternative self-assembly technique for organizing PVA VNPs over a larger surface.Earlier Moghimian et al. (2014) reported better alignment of M13 phage onto amorphous carbon surface (i.e., TEM grid) compared to SiO 2 surface.The method applied therein comprised initial adsorption of M13 particles onto substrate surface followed by washing at an angle of 45 • .In our study, we coated Si-wafer surface with 20 nm carbon film via spluttering.Then we dispensed PVA buffer along the carbon coated surface as described in Fig. 6A.The substrate here was devoid of α-CP antibodies to bind the VNPs, and the contact time of the solution to the surface was minimum due to vertical flow.Therefore, higher concentration (15 mg/mL) of PVA buffer was used in this study to ensure retention of sufficient amount of PVA VNPs on the substrate.Moreover, since PVA buffer showed nematic LC phase at 15 mg/mL range (Fig. 4C), it was deemed as a good concentration for depositing aligned PVA VNPs even with low contact time.The PVA deposition pattern on the surface was analyzed with AFM, which demonstrated the formation of a periodic surface pattern with local alignment (Fig. 6B).These results are thus aligned with the results done earlier with filamentous M13 particles (Moghimian et al., 2014) and highlight the importance of both solution-state self-assembly of the VNPs and the interactions between the deposition substrate and the VNPs in question for the final thin film morphology.
When taking into account all the conducted nanopatterning experiments, all the three coating methods including drop-casting, droptop deposition and flow-coating offer ways for creation of relatively smooth aligned in-plane cylindrical nanopatterns.Out of these methods, drop-top and flow coating arguably enable better control over thin film thickness and are more suitable for different types of layer-by-layer applications, as compared to drop-casting, which may also suffer from thickness variations due to coffee ring effects.Furthermore, when comparing drop-top deposition against flow-coating, the latter opens up wider opportunities in terms of tunability and scalability.

Conclusions
The PVA purification process presented in this work is superior to the previously reported protocols.By adding chelating (EDTA), disaggregating (NaOH), and reducing agents (β-mercaptoethanol) to the PVA solution, we were able to obtain higher yield of intact PVA VNPs, which is an important factor for all down-stream applications.The produced PVA particles were the most stable in low ionic-concentrated buffers with pH > 8, and without the addition of salts.Fig. 4. Analysis of PVA self-assembly behavior in water and buffer.A) PVA CP charge distribution in different pHs.Since the structure of PVA particles is not yet resolved in atomic detail, we predicted the structure of PVA CP using I-TASSER.Charge distribution was calculated using APBS-PDB2PQR software.Color code: blue, white and red denotes positive, neutral and negative charges respectively.B) LC behavior of 15 mg/mL PVA water visualized by POM.C) LC behavior of 15 mg/mL PVA buffer visualized by POM.Scale bars represents 100 μm.D,E) SAXS profile for PVA in water and buffer respectively.The dashed line represents a theoretical model of P(q) for a flexible cylinder with radius 6.3 nm, length 400 nm, and Kuhn length 30 nm generated using SasView 4.2.0 software.The arrow indicates the position of the correlation peak.
Investigation on the LC behavior of PVA suggested that the organization of particles in water was affected by the electrostatic repulsion between particles, leading to a chiral nematic phase, and thus patterns of isotropic orientation were obtained in the drop-casted film after drying.In contrast, buffer with a pH above isoelectric point resulted in pure nematic phase dominated by entropic alignment, therefore creating a drop-casted film with anisotropic morphology.Self-assembled thin films with aligned nanopatterns could also be achieved with drop-top and flow-coating methods.The presented investigation on the self-assembly of PVA as well as their applicability in thin film deposition contribute towards the development of novel VNP nanostructures.As compared to nanopatterning with synthetic block copolymers, with PVA VNPs, similar pattern dimensions can be reached combined with more versatile possibilities for controlled surface functionalization.Thus, these materials provide an unexplored platform for fabricating highly controlled nanoscale organic-inorganic functional materials and devices.
S.De et al.

Fig. 1 .
Fig. 1.Development of purification protocol for obtaining high-quality PVA particles suitable for material science applications.A) Schematic representation of the optimized purification protocol.B) Particle distribution in the old protocol.Fr1 -Fr6 denotes six consecutive 2 mL fractions collected from 5 to 40% sucrose gradient.C) Particle distribution in the new protocol.Arrowheads on the western blots (B and C) denote the position of monomeric PVA CP.In addition, Fr5 -Fr6 in C show faint bands near 70 and 100 kDa (marked with asterisk) supposedly coming from dimer and trimers of CP.D) TEM image of the purified particles.

Fig. 2 .
Fig. 2. Stability of PVA particles in different solution conditions.A) Stability in 0.1 M buffers with different pH.B) Different molarity of HEPES buffer at pH 8. C) Different ionic concentrations of KCl.D) Different salts at 0.1 M * denotes a significant difference (p < 0.05) from control.

Fig. 3 .
Fig. 3. Self-assembly behavior of drop-casted PVA films.A) Schematic illustration of the drop-casting of PVA solution (3 mg/mL) on Si wafer.B,C) SEM images of drop-casted PVA water and PVA buffer films, respectively.D,E) AFM images of drop-casted PVA water and PVA buffer film respectively.

Fig. 5 .
Fig. 5. PVA templating using drop-top coating method.A) Schematic model depicting drop-top coating method.B,C) SEM image of the PVA particles bound to the grid right after they were removed from the top of the corresponding drops in PVA water and PVA buffer respectively.D,E) TEM images of the PVA monolayer after extensive washing of PVA monolayer bound to the grid by α-CP antibody.Insets show FFT diffractograms produced by (D) isotropic PVA water and (E) anisotropic PVA buffer .