Synthesis Monitoring, Characterization and Cleanup of Ag-Polydopamine Nanoparticles Used as Antibacterial Agents with Field-Flow Fractionation

Advances in nanotechnology have opened up new horizons in nanomedicine through the synthesis of new composite nanomaterials able to tackle the growing drug resistance in bacterial strains. Among these, nanosilver antimicrobials sow promise for use in the treatment of bacterial infections. The use of polydopamine (PDA) as a biocompatible carrier for nanosilver is appealing; however, the synthesis and functionalization steps used to obtain Ag-PDA nanoparticles (NPs) are complex and require time-consuming cleanup processes. Post-synthesis treatment can also hinder the stability and applicability of the material, and dry, offline characterization is time-consuming and unrepresentative of real conditions. The optimization of Ag-PDA preparation and purification together with well-defined characterization are fundamental goals for the safe development of these new nanomaterials. In this paper, we show the use of field-flow fractionation with multi-angle light scattering and spectrophotometric detection to improve the synthesis and quality control of the production of Ag-PDA NPs. An ad hoc method was able to monitor particle growth in a TLC-like fashion; characterize the species obtained; and provide purified, isolated Ag-PDA nanoparticles, which proved to be biologically active as antibacterial agents, while achieving a short analysis time and being based on the use of green, cost-effective carriers such as water.


Introduction
Advances in nanotechnology have opened up new horizons in nanomedicine through the synthesis of new nanomaterials. Among the different antibacterial approaches, metal nanoparticles (NPs) are increasingly being used as an alternative to antibiotics or in combination with them. The advantages of these strategies are numerous, including providing activity based on different and simultaneous mechanisms such as oxidative stress induction, metal ion release, and non-oxidative mechanisms; the prevention of microbial drug resistance; and their potential use for carriers of antibiotics [1]. A wide range of nanomaterials have been demonstrated to possess antimicrobial effects, including iron (III) oxide, zinc oxide, magnesium oxide, silver, gold, copper and copper oxide, calcium oxide, titanium dioxide, and cadmium oxide. Among these, silver NPs (AgNPs) show promise for use as antimicrobials able to kill both Gram-negative and Gram-positive bacteria [2] due to both their internalization in the cell due to their small size and their release of silver ions [3].
for hydrodynamic size distribution analysis [30,31]. However, DLS gives no information on particle shape and density distribution and, in case of samples with complex, multimodal PSD, the accuracy of DLS-based PSD analysis may be intrinsically limited. Static, multi-angle LS (MALS) gives independent information on NP molar mass (M r ) and rootmean-square (rms) radius values [32]. Consequently, it may provide information on the conformation and structure of NPs. As in the case of DLS, the detection accuracy of MALS is reduced for NPs with complex and multimodal PSD. These techniques are not suitable for directly giving precise information on dispersed nanoparticles, specifically when many species coexist and samples present a relatively high complexity and heterogeneity. The hyphenation of DLS or MALS detection to size-based separation methods can thus enhance the accuracy of the size analysis of complex NP samples.
The use of size-based separation techniques, such as field-flow fractionation (FFF), with spectroscopic and light scattering detection has already been shown as a unique platform able to characterize different NP in terms of their size, dimension, shape, and functional properties [33][34][35][36][37]. Recently, FFF was applied to the quantification of AgNPs in complex samples such as food samples and biological samples [38][39][40][41][42].
The micro-volume variant of FFF, hollow-fiber flow-field flow fractionation (HF5), was demonstrated to be capable of achieving a high performance and low dilution at the same time for the analysis of particles of different natures; this method also allows for applications where a disposable device is needed to avoid cross-contamination [43][44][45][46][47][48][49][50][51].
In HF5, separation is achieved without a stationary phase and by an external flow-field, perpendicular to the parabolic flow in an empty, porous capillary channel. The retention of NPs is inversely proportional to the hydrodynamic diffusion coefficient of the analyte and, consequently, to its M r or hydrodynamic size. Selectivity is particularly high in the high-M r and nanometer-size range [35]. HF5 coupled on-line with multiple detectors is able to separate and characterize populations such as proteins, colloids, polymers, and particulate materials up to about 100 µm in size. HF5 merges the advantages of being a miniaturized technique with the absence of the need for specific optimization for the separation of various nanoparticles depending on its sample properties, such as purity, density, solubility, hydrophobicity, solution conductivity, and particle isoelectric charge.
HF5 has already been applied to the deep characterization of silver nanoparticles, providing fundamental information on their size/morphological characterization and ion release quantification in various conditions (e.g., dilution, preservation media, coated surface) [50,52,53]. In this work, we present the use of an HF5-based analytical platform for the characterization of Ag-PDA NPs in dispersion, both during the synthesis steps to monitor particle formation and to characterize, separate, and collect purified particles. The goal is to provide a low-time, low-cost approach which can overcome the limitations of offline techniques and is able to integrate information about particle synthesis and formation; provide product characterization; and enable the prompt purification of isolated, biologically active nanoparticles. Indeed, the effectiveness of the purified AgPDA NPs in inhibiting bacterial growth was demonstrated in vitro against S. aureus and E. coli reference strains, as opposed to the lack of activity found for unpurified samples. The proposed approach can be used for synthesis optimization and for the quality control of the final AgPDA NPs product.

Reagents and Chemicals
All reagents were purchased from Sigma Aldrich (St. Louis, MO, USA). MilliQ water was employed for the mobile phase and synthesis medium.

Synthesis of Ag-PDA Nanoparticles
The synthesis of Ag-PDA nanoparticles was performed with an optimized method elaborated from the literature [52][53][54]. An initial screening of synthesis conditions was carried out to define the best reaction mixture (see Supplementary Materials: Figure S1 and Table S1), which was then used for this study. Briefly, ammonia aqueous solution (25%) was added to a mixture of ethanol in water (30% v/v). Following this, AgNO 3 was added to achieve a final concentration of 0.5 mM. Finally, Dopamine-HCl was added drop by drop and the solution was kept under stirring. The final concentration of each species in the nanoparticle synthesis mix is reported in Table 1. Table 1. Concentration of reagents and medium used in Ag-PDA synthesis.

Ag (AgNO 3 ) (mM)
Dopamine-HCl (mM) The stirring solution turned first yellow, indicating the beginning of AgNP formation [55], and then became black, indicating the beginning of the formation of polydopamine (melamin) composites. The sampling of the solution was performed during the color changes (15 min, 1 h, 2 h) and every two hours starting from 22 h to 30 h.

FFF-DAD-MALS
HF5 analyses were performed using an Agilent 1200 HPLC system (Agilent Technologies, Santa Clara, CA, USA) consisting of a degasser, an isocratic pump, and an Agilent 1100 DAD UV/Vis spectrophotometer combined with an Eclipse ® DUALTEC separation system (Wyatt Technology Europe, Dernbach, Germany). The system was connected to an 18-angle multiangle light scattering detector model DAWN HELEOS (Wyatt Technology Corporation, Santa Barbara, CA, USA). The HF5 cartridge (Wyatt Technology Europe) is commercially available and has a 10 kDa cutoff [56]. ChemStation version B.04.02 (Agilent Technologies) data system for Agilent instrumentation was used to set and control the instrumentation and for the computation of various separation parameters, complete with the Wyatt Eclipse @ ChemStation version 3.5.02 (Wyatt Technology Europe). ASTRA ® software version 6.1.7 (Wyatt Technology Corporation) was used to handle signals from the detectors (MALS and UV) and to compute the sample rg values.
After samples were injected in the HF fiber, the separation process was performed in four steps: (1) Focus. During focus, a flow of mobile phase was split into two different streams entering from inlet and outlet. This step was used to stabilize flows.
(2) Focus-injection. The flow settings remained unvaried. The sample entered the channel through the inlet and the flow settings allowed us to focus analytes into a narrow band.
(3) Elution. After sample injection, the flow settings changed and a flow of the mobile phase entered the channel only by the inlet. Part of it came out transversely through the channel pores (cross-flow), while the rest (channel flow, Vc) reached the detectors. The strength of the hydrodynamic file applied to nanoparticles to achieve their separation could be regulated by modifying the intensity of the cross-flow while analytes were eluted along the fiber towards the detectors. This parameter could be modified throughout the analysis to generate a decreasing cross-flow (namely, gradient). (4) Elution-injection. The cross-flow was set to zero and the mobile phase flowed along the fiber to the detectors, allowing for any remaining sample due to the cross-flow action inside the channel to be released. Additionally, the flow was redirected in the injection line to clean it before the next injection. The flow conditions for the HF5 method developed are detailed in Table 2, Where longitudinal flow is indicated by Vc and cross/focus flow as Vx.
Multi-angle light scattering (MALS) was used to calculate the gyration radius of eluting species, since it allows for the absolute determination of the particle root mean square radius of gyration (Rg) by measuring the net intensity of light scattered by such particles at a range of fixed angles. The polidispersity index of the two populations obtained from the FFF analysis was also obtained from the LS data. The separation method was used to characterize the synthesis and fractionate Ag-PDA nanoparticles. The injection volume was set to 10 µL for synthesis monitoring and 50 µL for particle characterization and collection. All analyses were carried out in Milliq water as the mobile phase. For antimicrobial evaluation, a non-separative, non-filtrating (flow injection analysis, FIA) and non-separative, filtrating (Focus-FIA) method were also employed. An FIA is a shortened, non-separative method: the sample is injected into the channel in the absence of cross/focus flow and reaches the detector without separation. This allows 100% of the sample to be collected. Instead, in a Focus-FIA, the sample components smaller than the membrane cutoff are filtered out and only the colloidal portion of the sample goes through the detector, while unreacted dopamine and Ag ions are removed.

SEM Analysis
For morphological examinations, a few drops of each sample of interest were deposited onto a metallic stub and allowed to dry. A Philips XL-20 scanning electron microscope operating at 15 kV was used for this. Samples were sputter-coated with gold before examination. Energy dispersive X-ray spectrometry (EDX) measurements were also performed to evidence the presence of silver.

X-ray Powder Diffraction Analysis
For X-ray investigations, a few drops of each sample of interest (FFF fractions 1, 2) were deposited onto a recessed silicon glass and allowed to dry. X-ray powder diffraction analyses were carried out by means of a Philips X'Celerator powder diffractometer equipped with a graphite monochromator in the diffracted beam. CuKα radiation (λ = 1.54 Å; 40 mA, 40 kV) was used. The 2θ range was from 30 • to 80 • , with a step size of 0.0668 • and a scan rate of 500 s/step.

Antibacterial Activity
The antibacterial activity of the FIA, Focus-FIA, and method fractions was evaluated in vitro against Staphylococcus aureus (ATCC 25923) and Escherichia coli (ATCC 25922), which were selected as representative strains for Gram-positive and Gram-negative bacteria. The effectiveness of the samples in inhibiting bacterial growth was assessed by a standardized microdilution broth method using a 96-well plate [57]. Briefly, the bacterial suspensions were prepared at 0.5 McFarland, diluted 1:200 in Mueller-Hinton Broth (Sigma-Aldrich), and incubated with dilutions of each sample, starting from a ten-fold dilution of the volume collected from the fractionation process. Experiments included controls used to measure the bacterial growth in regular medium (positive control) and to check the background turbidity of the reagents and the sterility of the procedures (negative controls). The microplate was incubated at 37 • C and bacterial growth was monitored by measuring the Optical Density at 630 nm (Multiskan Ascent microplate reader, Thermo Fisher Scientific Inc., Waltham, MA, USA). Percentage values of samples in the different experimental conditions were determined relative to the positive growth control.

Results
The synthesis of pure, unmodified, and reliable nanoparticles is necessary to ensure their safe and standardized use; this is particularly true for clinical applications. In our work, we addressed the critical steps of the synthesis of a clinically relevant nanocomposite carrying both an antibacterial agent and a biocompatible material. The polymerization dynamics of dopamine are complex, time consuming, and often yield a mix of co-products. Likewise, the deposition of silver in the particulate is scarcely controlled and requires adequate method optimization and cleanup and thorough offline characterization. A series of post-synthesis steps is often performed, including cycles of ultracentrifugation and washing. However, these steps could modify the structure of the material; induce aggregation; and induce the release of silver, which is prone to dissolution [50,58].
Within this framework, we devised a separation method able to (1) monitor the synthesis of the nanocomposite in a TLC-like fashion and (2) provide the characterization and purification of isolated Ag-PDA using the same optimized separation method coupled to spectroscopic and size characterization. Indeed, the presence of non-destructive, orthogonal detectors, such as a DAD detector and multi-angle light scattering, offered the simultaneous size determination and identification of the different species obtained during the conjugation, and allowed the reaction to be carried out to completion, avoiding the waste of time and chemicals. ing both an antibacterial agent and a biocompatible material. The polymerization dyna of dopamine are complex, time consuming, and often yield a mix of co-products. Likew the deposition of silver in the particulate is scarcely controlled and requires adeq method optimization and cleanup and thorough offline characterization. A series of p synthesis steps is often performed, including cycles of ultracentrifugation and wash However, these steps could modify the structure of the material; induce aggregation; induce the release of silver, which is prone to dissolution [50,58].
Within this framework, we devised a separation method able to 1) monitor the thesis of the nanocomposite in a TLC-like fashion and 2) provide the characterization purification of isolated Ag-PDA using the same optimized separation method couple spectroscopic and size characterization. Indeed, the presence of non-destructive, ort onal detectors, such as a DAD detector and multi-angle light scattering, offered the ultaneous size determination and identification of the different species obtained du the conjugation, and allowed the reaction to be carried out to completion, avoiding waste of time and chemicals.  Following this, the reaction was monitored until (i) a peak was visible and intense and (ii) the peak was also detectable from the MALS analysis, which occurred after a reaction time of about 20 h, which was consistent with the literature [54][55][56]. After this, the real-time  Following this, the reaction was monitored until (i) a peak was visible and intense and (ii) the peak was also detectable from the MALS analysis, which occurred after a reaction time of about 20 h, which was consistent with the literature [54][55][56]. After this, the real-time analysis was performed every two hours and showed the process of formation of Ag-PDA nanoparticles (Figure 2). In Figure 2a, the UV absorption at three timepoints (22, 24, 26 h) is overlaid: the intense band at 5 min is common for the three profiles, but becomes slightly lower with the increase in the reaction time; with the same trend, the peak tails with increasing absorption (black arrows). This behavior could go undetected using only UV detection, but the formation of a new, different species is made clear by the observation of the LS trace (Figure 2b-d). In fact, the light scattering signal is more sensible to size increases and allows us to observe the insurgence of a second species. Moreover, the gyration radius is calculated (overlaid on the panels), showing that the species under formation has a good monodispersion. The measured radii ranged from (38 ± 5) nm (at 22 h) to (45 ± 4) nm after 26 h; this is in line with the time retention shift of the emerging species, corresponding to particle growth. A similar dataset regarding Ag-PDA synthesis would be almost impossible to obtain by only relying on offline techniques such as imaging, requiring a more cost-and time-intensive experiment. On the other hand, "faster" techniques such as DLS and spectra acquisition would have been unable to identify the formation of different species, particularly in the early stages of particle formation.

Online Characterization of Ag-PDA
The reaction was monitored until the relative intensity of the two bands reached a plateau. This was observed after a reaction time of 30 h, when the second band was also resolved from the first. At later times, no difference was observed, indicating that the two In Figure 2a, the UV absorption at three timepoints (22,24,26 h) is overlaid: the intense band at 5 min is common for the three profiles, but becomes slightly lower with the increase in the reaction time; with the same trend, the peak tails with increasing absorption (black arrows). This behavior could go undetected using only UV detection, but the formation of a new, different species is made clear by the observation of the LS trace (Figure 2b-d). In fact, the light scattering signal is more sensible to size increases and allows us to observe the insurgence of a second species. Moreover, the gyration radius is calculated (overlaid on the panels), showing that the species under formation has a good monodispersion. The measured radii ranged from (38 ± 5) nm (at 22 h) to (45 ± 4) nm after 26 h; this is in line with the time retention shift of the emerging species, corresponding to particle growth. A similar dataset regarding Ag-PDA synthesis would be almost impossible to obtain by only relying on offline techniques such as imaging, requiring a more cost-and time-intensive experiment. On the other hand, "faster" techniques such as DLS and spectra acquisition would have been unable to identify the formation of different species, particularly in the early stages of particle formation.

Online Characterization of Ag-PDA
The reaction was monitored until the relative intensity of the two bands reached a plateau. This was observed after a reaction time of 30 h, when the second band was also resolved from the first. At later times, no difference was observed, indicating that the two peaks detected (one at 5 min, the second at 10 min, dashed selections in Figure 3a) consisted of two distinct species likely reaching equilibrium. peaks detected (one at 5 min, the second at 10 min, dashed selections in Figure 3a) consisted of two distinct species likely reaching equilibrium. The scattering profile (Figure 3b) is relatively more intense for the second band, which is compatible both with the increased hydrodynamical size and the presence of a stronger scatterer such as silver. The radius calculation also highlights the difference between the two species. The first is eluted at the void time and displays an unusually high gyration radius (35 ± 6 nm), which could be an artifact due to the presence of intertwined polymeric chains lacking an organized 3D structure. These chains could be eluted with an inverse mechanism due to their elongated morphology and could penetrate the parabolic flow by steric hindrance rather than diffusion [59]. This hypothesis is also supported by a high polydispersity index (PDI = 1.105), obtained from the LS data. A PDI value of 1.000 corresponds to a monodispersed sample. The second band instead had a gyration radius of (47 ± 5) nm. In order to verify the hydrodynamical size of Ag-PDA nanoparticles, a size standard (polystyrene beads, geometrical radius = 51 nm) was injected (not shown). The resulting retention time was found to be identical to that of band 2, indicating that the estimated rh of this species the same. In this case, the particles appeared to be monodispersed, with a PDI of 1.016.
Further insights into the morphology and arrangement of the PDA-silver nanocomposite can be obtained by studying its shape factor. This value, calculated as the rg/rh ratio, indicates where the mass of a particle is located with respect to its center of mass and is directly linked to morphology. Numerically, a shape factor of 0.8 is typical for solid spheres (such as polystyrene beads), increasing to 1 for hollow spheres and reaching higher values for more elongated structures such as rods (1.4) and random coils (1.6). A value closer to 1 is typical for either core-shell systems or prolate ellipsoids. The value found for particles from fraction 2, which is 0.9, can be justified by the presence of nanosilver localizing on the surface of the nanoparticle and creating a denser shell and by a more elongated structure than that of a regular sphere [60]. The size, shape, and polydispersity results obtained for the two species are shown in Table 3. The scattering profile (Figure 3b) is relatively more intense for the second band, which is compatible both with the increased hydrodynamical size and the presence of a stronger scatterer such as silver. The radius calculation also highlights the difference between the two species. The first is eluted at the void time and displays an unusually high gyration radius (35 ± 6 nm), which could be an artifact due to the presence of intertwined polymeric chains lacking an organized 3D structure. These chains could be eluted with an inverse mechanism due to their elongated morphology and could penetrate the parabolic flow by steric hindrance rather than diffusion [59]. This hypothesis is also supported by a high polydispersity index (PDI = 1.105), obtained from the LS data. A PDI value of 1.000 corresponds to a monodispersed sample. The second band instead had a gyration radius of (47 ± 5) nm. In order to verify the hydrodynamical size of Ag-PDA nanoparticles, a size standard (polystyrene beads, geometrical radius = 51 nm) was injected (not shown). The resulting retention time was found to be identical to that of band 2, indicating that the estimated rh of this species the same. In this case, the particles appeared to be monodispersed, with a PDI of 1.016.
Further insights into the morphology and arrangement of the PDA-silver nanocomposite can be obtained by studying its shape factor. This value, calculated as the rg/rh ratio, indicates where the mass of a particle is located with respect to its center of mass and is directly linked to morphology. Numerically, a shape factor of 0.8 is typical for solid spheres (such as polystyrene beads), increasing to 1 for hollow spheres and reaching higher values for more elongated structures such as rods (1.4) and random coils (1.6). A value closer to 1 is typical for either core-shell systems or prolate ellipsoids. The value found for particles from fraction 2, which is 0.9, can be justified by the presence of nanosilver localizing on the surface of the nanoparticle and creating a denser shell and by a more elongated structure than that of a regular sphere [60]. The size, shape, and polydispersity results obtained for the two species are shown in Table 3.
The two FFF-separated species were also different in terms of their spectroscopical behavior (Figure 3c,d). The first one had a spectrum typical for polydopamine, while the second displayed a local maximum at 390 nm, compatible with the presence of nanosilver embedded in the particle. This information is crucial for the intended use of such a material: indeed, the presence of a biocompatible polymer which could be deposited as a film would hinder the application of a silver delivery system, while the identification and separation of the two species would lead to the production of an effective antiseptic candidate [61]. Based on this evidence, two fractions were collected (one from min 4 to 6 and one in the range of 8.5-11.5 min) and subjected to further offline characterization.

Purification and Offline Characterization
The two fractions were imaged through scanning electron microscopy (SEM). The first species collected (Figure 4a) displayed a more amorphous structure, indicating the lack of a tridimensional organization, as expected from the results of the FFF characterization. The second fraction, in contrast, consisted of well-defined, spherical particles with a diameter of about 100 nm (Figure 4b).

Hydrodynamic
Radius (nm) The two FFF-separated species were also different in terms of their spectroscopical behavior (Figure 3c,d). The first one had a spectrum typical for polydopamine, while the second displayed a local maximum at 390 nm, compatible with the presence of nanosilver embedded in the particle. This information is crucial for the intended use of such a material: indeed, the presence of a biocompatible polymer which could be deposited as a film would hinder the application of a silver delivery system, while the identification and separation of the two species would lead to the production of an effective antiseptic candidate [61]. Based on this evidence, two fractions were collected (one from min 4 to 6 and one in the range of 8.5-11.5 min) and subjected to further offline characterization.

Purification and Offline Characterization
The two fractions were imaged through scanning electron microscopy (SEM). The first species collected (Figure 4a) displayed a more amorphous structure, indicating the lack of a tridimensional organization, as expected from the results of the FFF characterization. The second fraction, in contrast, consisted of well-defined, spherical particles with a diameter of about 100 nm (Figure 4b). Energy dispersive X-ray analysis (EDX) (Figure 4, insets) shows the presence of silver only for the second fraction (green points), as hypothesized from the absorption maximum observed in Figure 3d. If batch characterization would have left room for speculation about the presence of free silver in the suspension due to the presence of unreacted AgNO 3 or silver nanoparticles alone, the use of FFF separation and purification removes all doubts. In fact, the fractionation process also includes a filtration step (i.e., focusing; see M&M), which would eliminate free ions present in the reaction mix, and only preserve Ag if the metal took part in the nanostructure.
While the presence of silver in the nanostructure was confirmed, it was necessary to determine the nature of silver in the second fraction. This was achieved by X-ray analysis, which is shown in Figure 4c together with the spectrum of AgNO 3 as a reference.
The results confirm the presence of metallic nano silver, indicating that the sample corresponds to purified, functionalized Ag-PDA nanoparticles. Indeed, the corresponding powder X-ray diffraction pattern shows characteristic peaks of Ag crystal found at 38.092, 44.214, 64.409, and 77.344 on the 2θ scale, corresponding, respectively, to the (1 1 1), (2 0 0), (2 2 0), and (3 1 1) crystal planes. There was no signal corresponding to Ag + .

Antibacterial Activity
In order to verify the antimicrobial activity of the freshly synthesized Ag-PDA nanoparticles and evaluate the effect of the FFF purification, a set of samples was produced to represent the reaction mix: free dopamine and potential unreacted Ag + ions (A), the colloidal portion made of both polymeric PDA and Ag-PDA particles (B), the purified polymeric population (C) and the purified Ag-PDA particles (D). This was achieved by the collection of HF5 fractions via two non-separative methods (FIA and Focus-FIA, as described in Section 2.3) and the previously employed separation method. The isolation process is schematized in Figure 5.
Energy dispersive X-ray analysis (EDX) (Figure 4, insets) shows the presence of only for the second fraction (green points), as hypothesized from the absorption mum observed in Figure 3d. If batch characterization would have left room for sp tion about the presence of free silver in the suspension due to the presence of unr AgNO3 or silver nanoparticles alone, the use of FFF separation and purification re all doubts. In fact, the fractionation process also includes a filtration step (i.e., foc see M&M), which would eliminate free ions present in the reaction mix, and only pr Ag if the metal took part in the nanostructure.
While the presence of silver in the nanostructure was confirmed, it was necess determine the nature of silver in the second fraction. This was achieved by X-ray an which is shown in Figure 4c together with the spectrum of AgNO3 as a reference.
The results confirm the presence of metallic nano silver, indicating that the s corresponds to purified, functionalized Ag-PDA nanoparticles. Indeed, the correspo powder X-ray diffraction pattern shows characteristic peaks of Ag crystal found at 3 44.214, 64.409, and 77.344 on the 2θ scale, corresponding, respectively, to the (1 1 1 0), (2 2 0), and (3 1 1) crystal planes. There was no signal corresponding to Ag + .

Antibacterial Activity
In order to verify the antimicrobial activity of the freshly synthesized Ag-PDA particles and evaluate the effect of the FFF purification, a set of samples was produ represent the reaction mix: free dopamine and potential unreacted Ag + ions (A), th loidal portion made of both polymeric PDA and Ag-PDA particles (B), the purified meric population (C) and the purified Ag-PDA particles (D). This was achieved collection of HF5 fractions via two non-separative methods (FIA and Focus-FIA, scribed in Section 2.3) and the previously employed separation method. The isolatio cess is schematized in Figure 5. The collected samples A-D were diluted to the same final concentration, as calc from their relative area at 300 nm, and subjected to antibacterial activity tests.
The inhibitory activity of the different fractions was assayed in vitro agains Gram-positive and Gram-negative bacteria by a broth microdilution method. Ba The collected samples A-D were diluted to the same final concentration, as calculated from their relative area at 300 nm, and subjected to antibacterial activity tests.
The inhibitory activity of the different fractions was assayed in vitro against both Grampositive and Gram-negative bacteria by a broth microdilution method. Bacterial growth of S. aureus and E. coli was monitored at different times, and the results demonstrated that the only biologically active fraction was D, consisting of purified Ag-PDA nanoparticles, as reported in Figure 6a. The lack of the activity of fraction C, where only polydopamine is present (as confirmed both by online and offline characterization), was also confirmed, as expected. Interestingly, fractions A and B, where Ag is present both as an ion and as unpurified Ag-PDA nanoparticles, were found to be ineffective. This could be explained by the presence of polydopamine in the suspension, which hinders interaction between nanosilver and bacterial cell walls, confirming the necessity of particle purification. strated that the only biologically active fraction was D, consisting of purified Ag-PDA nanoparticles, as reported in Figure 6a. The lack of the activity of fraction C, where only polydopamine is present (as confirmed both by online and offline characterization), was also confirmed, as expected. Interestingly, fractions A and B, where Ag is present both as an ion and as unpurified Ag-PDA nanoparticles, were found to be ineffective. This could be explained by the presence of polydopamine in the suspension, which hinders interaction between nanosilver and bacterial cell walls, confirming the necessity of particle purification.
Growth data analysis showed that Gram-negative E. coli (Figure 6a, left) had a higher sensitivity than Gram-positive S. aureus (Figure 6a, right). Indeed, the cell wall structure plays a crucial role in the antibacterial activity of Ag-based compounds; it is well known that Gram-negative bacteria are more susceptible due to their narrower cellular walls compared to Gram-positive strains [62]. The purified Ag-PDA nanoparticles in fraction D completely inhibited E. coli when used at the highest dilution (1:10 with respect to the volume collected from the fractionation process) and with up to 24 h of incubation. Given this result, serial two-fold dilutions of fraction D were also assayed against E. coli. The results confirmed the effectiveness of the sample in inhibiting bacterial growth and highlighted the specific interaction of the HF5-isolated Ag-PDA nanoparticles with bacterial membranes, leading to dose-dependent activity (Figure 6b).

Conclusions
In our work, we addressed the critical steps of the synthesis of a clinically relevant nanocomposite carrying both an antibacterial agent (nanosilver) and a biocompatible material (polydopamine), Ag-PDA. Given the complexity of the synthesis mechanisms and the difficulty of controlling its process, a series of post-synthesis steps is often performed after this process, including cycles of centrifugation and washing. However, these steps can be ineffective, modify the structure of the material, induce aggregation, and induce the release or dissolution of silver.
Within this framework, we devised a fast, non-destructive method able to 1) monitor the synthesis of the nanocomposite in a TLC-like fashion and 2) Provide the characterization and purification of isolated Ag-PDA using the same, optimized separation method based on miniaturized FFF coupled to spectroscopical and size characterization. Orthogonal UV and light scattering detectors offered the simultaneous size determination and identification of the different species obtained during the conjugation, which were confirmed through conventional, offline characterization. The comparison between unfractionated, filtered, and isolated nanoparticles confirmed the antibacterial effectiveness of Growth data analysis showed that Gram-negative E. coli (Figure 6a, left) had a higher sensitivity than Gram-positive S. aureus (Figure 6a, right). Indeed, the cell wall structure plays a crucial role in the antibacterial activity of Ag-based compounds; it is well known that Gram-negative bacteria are more susceptible due to their narrower cellular walls compared to Gram-positive strains [62]. The purified Ag-PDA nanoparticles in fraction D completely inhibited E. coli when used at the highest dilution (1:10 with respect to the volume collected from the fractionation process) and with up to 24 h of incubation.
Given this result, serial two-fold dilutions of fraction D were also assayed against E. coli. The results confirmed the effectiveness of the sample in inhibiting bacterial growth and highlighted the specific interaction of the HF5-isolated Ag-PDA nanoparticles with bacterial membranes, leading to dose-dependent activity (Figure 6b).

Conclusions
In our work, we addressed the critical steps of the synthesis of a clinically relevant nanocomposite carrying both an antibacterial agent (nanosilver) and a biocompatible material (polydopamine), Ag-PDA. Given the complexity of the synthesis mechanisms and the difficulty of controlling its process, a series of post-synthesis steps is often performed after this process, including cycles of centrifugation and washing. However, these steps can be ineffective, modify the structure of the material, induce aggregation, and induce the release or dissolution of silver.
Within this framework, we devised a fast, non-destructive method able to (1) monitor the synthesis of the nanocomposite in a TLC-like fashion and (2) Provide the characterization and purification of isolated Ag-PDA using the same, optimized separation method based on miniaturized FFF coupled to spectroscopical and size characterization. Orthogonal UV and light scattering detectors offered the simultaneous size determination and identification of the different species obtained during the conjugation, which were confirmed through conventional, offline characterization. The comparison between unfractionated, filtered, and isolated nanoparticles confirmed the antibacterial effectiveness of Ag-PDA, especially towards Gram-negative E. coli, but also highlighted how the isolation of these particles proved crucial to avoid interference and the loss of activity. The FFF multidetection approach allowed the reaction to be carried out to completion, avoiding the waste of time and chemicals; allowed fast on-line characterization; and provided isolated, purified, and biologically active Ag-PDA particles in water, ensuring their safe use as biocompatible antibacterials.