2.1. Physical and Antimicrobial Properties of PMMA/PDDA/Amphiphile Dispersions
SEM micrographs for the dried PMMA/PDDA/amphiphile dispersions after dialysis revealed uniform spherical NPs as synthesized in the presence of CTAB (
Figure 1a), DODAB (
Figure 1b) or lecithin (
Figure 1c).
The micrographs on
Figure 1 submitted to the ImageJ software yielded the mean diameter (D) of the dry NPs in the three dispersions (
Table 1). As compared to the hydrodynamic diameter (Dz) obtained previously by dynamic light scattering (DLS) [
17], the D values were smaller than the Dz ones for the NPs in water (
Table 1).
The zeta-potentials (ζ) after dialysis remained positive, high and very similar to the ζ values for the PMMA/PDDA NPs obtained in absence of amphiphiles. The PDDA outer shell was responsible for the surface potential at the shear plane of the NPs in water dispersions (
Table 1). In absence of PDDA, PMMA/DODAB, PMMA/CTAB, or PMMA/lecithin NPs exhibited zeta-potentials equal to 50, 26, and −21 mV, respectively [
17]. In these cases, the amphiphile determined the surface potential at the shear plane of the NPs.
Figure 2 shows the effect of NaCl on Dz of PMMA/PDDA/DODAB, PMMA/PDDA/CTAB, and PMMA/PDDA/lecithin NPs. The effect of increasing [NaCl] was decreasing Dz possibly due to decreasing the thickness of the outer PDDA shell in water and screening of the positive charges on the outer polymer chains (
Figure 2). This was consistent with the drying effect on particle size shown in
Table 1 with D being generally smaller than Dz. The collapse of the outer and stretched PDDA polymer shell can take place either by drying the NP (
Table 1) or by increasing the ionic strength thereby causing the collapse of the outer shell (
Figure 2). Consistently, minimum Dz values were equal to the mean diameter D of the dry particles obtained by SEM (
Table 1;
Figure 2). Above the minimum Dz, further increasing [NaCl] lead to loss of colloidal stability and NPs aggregation with a substantial increase in Dz due to NPs aggregation (not shown). This occurred above 50, 20 and 50 mM NaCl for PMMA/PDDA/DODAB, PMMA/PDDA/CTAB, and PMMA/PDDA/lecithin NPs (
Figure 2). Therefore, these simple experiments and considerations allowed determining the thickness of the outer PDDA shell in the different NPs as (Dz–D)/2 or as (Dz–Dz
minimum)/2. Furthermore, both determinations agreed (
Table 1). Therefore, one can be confident that the shell thickness was properly obtained.
Previous data for PMMA/PDDA NPs obtained in absence of emulsifiers as compared to the antimicrobial effect of PDDA alone revealed that both similarly killed
E. coli and
S. aureus but PDDA was much more effective against
C. albicans than the PMMA/PDDA NPs [
19]. For the coatings cast from PMMA/PDDA NPs onto glass coverslips at 0.25 mg PDDA, the tests against bacteria (
E. coli or
S. aureus) yielded complete loss of cell viability but tests against
C. albicans were not performed [
17]. In this work, the antimicrobial properties of PMMA/PDDA/amphiphile dispersions were obtained for bacteria and fungus over a range of PDDA concentrations (
Figure 3). The minimal microbicidal concentration (MMC) determined for each dispersion was shown in
Table 2. The real potency of the dispersions was established over orders of magnitude using a logarithmic scale for the cell viability as a function of PDDA concentration. Microbial viability decreased by 10
7–10
8 viable colony forming unities (CFU) upon interaction with the NPs for 1 h (
Figure 3;
Table 2).
From
Figure 3 and
Table 2, the most sensitive microorganism to the dispersions was
E. coli, followed by
S. aureus and
C. albicans. The PMMA/PDDA/CTAB dispersions were the most effective dispersions killing the three strains by 7–8 logs in 1 h (
Figure 3,
Table 2). At this point, one should recall that all NPs exhibited hydrodynamic diameters of about 200 nm with the exception of the PMMA/PDDA/CTAB NPs with 90 ± 1 nm as Dz. Possibly, for this nanometric size especially; penetration of the NPs through the fungus cell wall and the cell membrane damaged these structures.
For other assemblies with PDDA, PDDA doses required for killing
S. aureus were always higher than those required for killing
P. aeruginosa, a gram-negative bacterium [
14]. Tetraalkyl ammonium compounds have been recognized as efficient blockers of the potassium channels KcsA of the Gram-negative
E. coli [
30].
Whereas the Gram-negative bacteria are very sensitive to all cationic compounds and assemblies, the Gram-positive
S. aureus is less sensitive via a mechanism of resistance. A sensor system for cationic antimicrobial molecules in
Staphylococcus sp. causes resistance to cationic antimicrobial agents [
31]. A short extracellular loop with a high density of negatively charged amino acid residues would attract and interact with cationic antimicrobial compounds. Transduction of this interaction signal would trigger the D-alanylation of teichoic acids and the lysylation of phosphatidylglycerol, resulting in a decreased negative charge of the cell surface and membrane, respectively. Thereby there would be a decreased attraction for cationic molecules.
In order to evaluate the concentration of free and nanoparticle bound PDDA in the dispersions, the dispersions were dialyzed after synthesis and then centrifuged at 9,000 g in order to precipitate the NPs and separate them from their supernatant. The supernatant was then used to determine chloride concentration by micro-titration as previously described (see Methods). Since chloride is the PDDA counter-ion, its concentration is directly related to the PDDA concentration in mg·mL
−1. Another possible way for evaluating [PDDA] in the supernatant was its biological activity determined from inhibition of growth on
S. aureus seeded agar plates (
Figure 4). The inhibition zones against
S. aureus depended on PDDA concentration in the wells numbered from 1 to 8; increasing [PDDA], increased the inhibition zone though the curve was not linear. Nevertheless, it was possible to estimate [PDDA] over a range of low concentrations (0–1.0 mg·mL
−1). After considering the dilution factor for the PMMA/PDDA/DODAB NPs after dialysis which was 2.8 the [PDDA] from the inhibition zone was 2.8 mg·mL
−1 in reasonable agreement with the [PDDA] determined from chloride microtitration which was 2.5 mg·mL
−1. One should notice that PDDA cannot permeate the dialysis membrane so that the PDDA in the dialyzed dispersions will still be available to contribute to the inhibition zones. DODAB also would not permeate the dialysis membrane if it were not bound to the PMMA matrix as described before [
27].
Table 3 shows the [PDDA] in the supernatants of the 4 different NPs dispersions determined by the two methods: chloride microtitration and inhibition zones against
S. aureus.PDDA, DODAB and CTAB are all compounds with halides as counterions. Dispersions with DODAB or CTAB and PDDA may yield overestimated PDDA concentrations by the halide method. However, DODAB, despite not being permeable through the dialysis membrane, displayed high affinity for PMMA polymeric matrix and did not leach to the outer medium from PMMA coatings [
27]. In addition, DODAB does not move through the agar and could not contribute to the inhibition zone observed for PMMA/PDDA/DODAB dialyzed NPs. Thus, for PMMA/PDDA/DODAB NPs after dialysis, both the microtitration and the determination of inhibition zones would give a reliable analysis of PDDA concentrations. On the other hand, CTAB was found to leach from PMMA coatings [
27] and readily permeate the dialysis membrane (
Figure 5). Thus CTAB would not be found in the supernatant of the dialyzed PMMA/PDDA/CTAB NPs—determination of PDDA concentration in the supernatant of dialyzed PMMA/PDDA/CTAB NPs would be reliable and not overestimated by the micro-titration method. Regarding the determination of inhibition zones, eventually, the lack of linear dependence of inhibition zones on [PDDA} would eventually yield poor determinations of concentration as indeed observed for PMMA/PDDA/lecithin dispersions. The composition of soybean lecithin includes negatively charged phospholipids and fatty acids [
32], which would occupy the monomer droplet/water interface during PMMA synthesis eventually facilitating the incorporation of the positively charged PDDA in the PMMA NPs. However, the results on
Table 3 show that the lecithin amphiphiles did not increase [PDDA]
nanoparticles. In summary, the use of amphiphiles during PMMA/PDDA NPs synthesis did not improve PDDA attachment, entanglement, adsorption and/or mechanical immobilization to the NPs. It was in absence of surfactants, that the highest amount of PDDA became attached to the NPs yielding the most well-structured PDDA shell around the PMMA core (
Table 1 and
Table 3;
Figure 2). In the presence of amphiphiles, by determining [PDDA] in the supernatant of centrifuged NPs it was found that most PDDA molecules became poorly attached to the NPs. The amphiphiles indeed reduced the interaction of PDDA with the NPs and diminished the thickness of the NP shell.
Figure 5 shows the different mobilities of DODAB or CTAB on
S. aureus seeded agar plates. There was a lack of inhibition zones against
S. aureus induced by 2mM DODAB in form of bilayer fragments [
33] on
Figure 5a. Inhibition zones occurred for 2 mM CTAB on
Figure 5b and for PMMA/CTAB NPs at 2 mM CTAB before dialysis on
Figure 5c. For PMMA/CTAB NPs after dialysis no inhibition zone was observed (
Figure 5d). Whereas CTAB moved through the agar, DODAB bilayer fragments remained inside the limits of the sample wells. CTAB was no longer found in the PMMA/CTAB dialyzed NPs so that inhibition zones were not observed (
Figure 5d).
In order to confirm that the location of PDDA in the PMMA/PDDA dispersions was mainly in the water phase (
Table 3), the antimicrobial activity against
E. coli was determined in the supernatant of centrifuged dispersions and compared to the total activity of the original dispersions.
Figure 6 shows this comparison for dispersions and their supernatants against
E. coli. The activity of the supernatants was slightly smaller than or equal to the one obtained using the original dispersions. This result was consistent with the predominant location of PDDA in the water phase and its poor association with the PMMA/PDDA/amphiphile NPs. The best association between PDDA and PMMA was achieved in absence of surfactants.
2.2. Physical and Antimicrobial Properties of Coatings Prepared from PMMA/PDDA/Amphiphile Dispersions
Figure 7 shows the macroscopic aspects of the PMMA/PDDA/amphiphile coatings obtained after casting and drying the NPs dispersions. The microscopic aspects of these coatings determined by SEM were also shown as the SEM micrographs previously shown as
Figure 1. In general, their features were very similar to those previously described in detail for the PMMA/PDDA films in absence of amphiphiles on the three different substrates [
17]. A uniform distribution of spherical nanoparticles of very similar size disposed in general side-by-side with some holes and discontinuities occurring at a low frequency showed the uniformity of the coatings.
The PMMA/PDDA/amphiphile coatings spread and adhered better to the hydrophilic surfaces such as the silicon wafers and the glass coverslips than to the hydrophobic polystyrene surfaces (
Figure 7). The outer cationic PDDA shell on each NP interacted better with the anionic surfaces of silicon wafers or glass whereas the repulsion between the hydrophobic polystyrene surfaces; the hydrophilic PDDA shell of the NPs originated poor adherence and cracks on the PMMA/PDDA/amphiphile coatings (
Figure 7). The coffee ring effect antagonized by the Marangoni effect was similar to effects previously described for coatings obtained from polymeric particles [
34,
35] as were the cracks visible for the coatings on the polystyrene surfaces [
36]. Using lecithin as emulsifier resulted in some phase separation in the coatings that did not occur for CTAB and DODAB as emulsifiers (
Figure 7).
In order to estimate the thickness of the coatings, simple calculations were performed as follows. The approximate area A for each coating on the hydrophilic substrates could be estimated considering spherical shape for the film with a radius of 0.5 cm: A = 0.785 cm
2. The solid content for each dispersion given in
Table 1 was previously determined gravimetrically [
17] so that the total mass of solids for each coating obtained from 0.05 mL could easily be calculated for each film. Considering each coating as a cylinder its volume will be its base area multiplied by its height. From the radius for the dry particles and the PMMA density (1.18 g·cm
−3), the volume and mass of each particle can be obtained. The total number of particles could be obtained from the total mass of the coating divided by the mass of one particle. The volume of the cylindrical films would then be given by the volume of each particle multiplied by the number of particles in each coating. Once known the volume of the cylinder, its thickness i.e., its height, could be estimated from the assumption of uniform coating and the division of the cylinder volume by its base area. From these simple calculations, the thicknesses for each film were estimated and added to
Table 4.
The coatings of PMMA/PDDA/amphiphile obtained by casting of the NPs dispersions onto silicon wafers were also evaluated regarding their wettability from contact angle determinations (
Table 4). These characteristics of the hybrid films compared to those of pure PMMA coatings revealed higher wettability for the hybrid coatings than the one determined for PMMA films (
Table 4).
Coatings obtained by casting the PMMA/PDDA dispersions yielded lower contact angles than those obtained by spin-coating. Possibly, some molecules of the hydrophilic PDDA immobilized as an outer layer of the PMMA/PDDA nanoparticle imparted a more hydrophilic character to the film surface than the one of the spin-coated PMMA/PDDA (
Table 4).
For coatings obtained by casting and drying the PMMA/PDDA/amphiphile dispersions onto the silicon wafers, there was a consistent decrease of hydrophobicity meaning a decrease of contact angles Ө
A in the following order: Ө
A DODAB > Ө
A lecithin > Ө
A CTAB (
Table 4). The presence of residual amphiphile possibly at the surface of the PMMA core reduced the hydrophilicity of the coatings in the expected order from CTAB to lecithin to DODAB. The more hydrophobic amphiphile was DODAB which has a small hydrophilic polar head and double-chained saturated hydrocarbon tails. Lecithin has a mixed composition of phospholipids and fatty acids, some of them unsaturated [
32] thereby displaying an intermediate hydrophobicity between the double-chained DODAB and the single-chained surfactant CTAB.
Rugosity (R) is a measure of small-scale variations of amplitude in the height of a surface given by the ratio between the real surface area (A) and the geometric surface area (A
g) [
37]. In the case of the NPs coatings from dispersions, R can be easily estimated due to the uniform size of the spherical NPs and the uniform general features of the coatings both microscopically (
Figure 1) and macroscopically (
Figure 7). The surface area for spherical NPs is given by 4Π(D/2)
2, where D is the diameter of the dry NP. One needs to calculate the total number of NPs in a cross-section of the coating. The geometric area of the coating (A
g) is 0.785 cm
2. The film thickness divided by D yields the number of NPs layers per film. From the total number of NPs per coating divided by the number of NPs layers, the number of NPs per layer could be calculated. Since the area of the NPs is known multiplying this area per the number of NPs per layer divided by 2 will give the real surface area of the outer layer of the coating (A). Thus, R can be obtained from A/A
g (
Table 4). The thicker the film, the rougher is its surface as depicted from the results on
Table 4. Thus, in order to reduce the film rugosity, one should reduce the total amount of solid deposited. High rugosity might also mean a larger surface area to interact with the microbes and higher activity against the microbes.
As previously described, high conversion rates for the NPs synthesis depended on the presence of stabilizers, such as the amphiphiles used in this work, with the function of stabilizing the droplet/water interface during the polymerization [
17]; only about 10% of the monomer mass added was converted into polymer in absence of the emulsifiers [
19]. Here 2 mM amphiphile effectively lead to ≥80% monomer-to-polymer conversion but this was at the expenses of the core-shell nanoparticle structure; the PDDA shell around each PMMA core was practically lost due to the presence of amphiphile at the core-water interface of the NPs.
Coatings from the NPs were also tested for their antimicrobial activity. The coatings contained about 0.20 mg of PDDA and reduced cell viability to zero in many instances (
Figure 8). The PMMA/PDDA/CTAB coatings yielded a larger reduction in final CFU countings for the three microorganisms tested at 0.18 mg of PDDA (
Figure 8). The low NPs size and high rugosity of the coatings (R) also affected the activity. The roughest coating cast from the smallest NPs was the one with the highest activity, namely the coating cast from the PMMA/PDDA/CTAB NPs (
Table 4;
Figure 8).
At 2 mM lecithin, since lecithin contains phospholipids and fatty acids [
32,
38], a net negative charge (zeta-potential equal to −21 mV) was obtained for PMMA/lecithin NPs synthesized in absence of PDDA even after exhaustive dialysis [
17]. At 2 mM DODAB or CTAB, the PMMA/DODAB or PMMA/CTAB NPs exhibited positive zeta-potentials [
17]. However, the lower zeta-potential for PMMA/CTAB in comparison to the one for PMMA/DODAB NPs was deemed consistent with the reported affinity of DODAB for the PMMA polymeric matrix that did not occur for CTAB—CTAB diffused to the outer water medium from PMMA coatings [
27,
28]. In summary, although the three stabilizers indeed improved monomer-to-polymer conversion, PDDA imparted the electro-steric repulsion between the MMA droplets during NP synthesis and represented an additional stabilizing factor for the PMMA/PDDA NPs. The use of cationic stabilizers such as DODAB and CTAB reduced PDDA shell surrounding the NPs (
Figure 2).
In this work, one of the most interesting findings referred to the role of NPs size on the microbicidal activity.
Figure 9 illustrated this finding showing that the frequency of multipoint attachment of the cells to the coatings cast from NPs can increase with the decrease in NPs size. Thereby the coatings made from small PMMA/PDDA/CTAB NPs would be more effective in killing the cells than those made from large PMMA/PDDA/DODAB or PMMA/PDDA/lecithin NPs.
The reason for the small size of PMMA/PDDA/CTAB NPs in the dispersion can be related to the conical shape of the CTAB molecule, which favors high curvature for the MMA droplets during polymerization. The lipids DODAB or lecithin with molecular shapes closer to cylinders would not favor the curvature of the MMA droplets so that final particle size would be larger than the one for stabilizers with a conical shape.
By adding amphiphiles such as CTAB, DODAB and lecithin as surfactants, active as stabilizers during the NPs synthesis, a major question was raised regarding the effect of the stabilizers on the core-shell NPs structure. Would the quaternary ammonium, cationic amphiphiles hamper the location of cationic PDDA as a shell surrounding each PMMA NP core? Apparently, this indeed happened since the shell thickness was substantially reduced in the dispersions with the cationic amphiphiles as shown in
Figure 2 and
Table 1. Using the cationic amphiphiles as stabilizers, changed PDDA location from the NP shell to bulk water phase. This reduced the PDDA shell surrounding NPs PMMA core. The negatively charged lecithin possibly found also as closed bilayers in dispersion also might have withdrawn PDDA from the NPs shell to a certain extent thereby reducing the shell thickness (
Figure 2;
Table 1).
Other questions referred to the eventual contribution of CTAB or DODAB to the antimicrobial activity of the dispersions and coatings. The reason for this question was the reported activity of DODAB and CTAB as antimicrobial agents [
8,
14,
16,
29,
39,
40,
41,
42]. In this work, the antimicrobial properties of these ternary systems both as latexes dispersions in water and as coatings were determined. PDDA resulted in the most important microbicidal agent due to its location as an outer shell of the core-shell PMMA/PDDA NPs and/or its release to the water medium of amphiphile stabilized NPs. Residual CTAB or DODAB would either remain in the NPs polymer matrix or leach from the NPs and through the dialysis membrane leaving the dispersion during dialysis and barely contributing to the microbicidal activity (
Figure 5). PDDA did not leave the dispersions by dialysis so that free PDDA antimicrobial activity remained significant for dialyzed dispersions and their coatings.