Introduction

Demand for responsive polymeric materials is continuously growing. Chromophoric advanced tools focused the attention of scientists and technologists due to their extensive uses in biotechnology and biomedical sciences. Their role as drug-carriers, anti-cancers, antimicrobials, sensors, and markers is well documented1,2,3,4,5,6. Specifically, chromophores exhibiting aggregation-induced emission (AIEgens) are the subject of keen interest7,8. AIEgens9,10,11 are luminophores active in the aggregate state and concentrated solutions. Therefore, they are active in aqueous and/or physiological environments where they aggregate.

The combination of AIE activity and stimuli responsiveness in a single platform provides an effective strategy for developing novel polymeric tools for modern bio-applications. Furthermore, thanks to the state-of-the-art bioimaging technologies, macro or nanoformulation of monitorable bioactive AIE-polymers provide theranostic probes for therapy and diagnostics12,13.

Antimicrobial resistance is one of the major challenges in human history. Macromolecules have been efficiently and extensively employed in the biomedical field as antimicrobial probes14,15. Due to the current global COVID-19 pandemic, interest in developing novel polymeric scaffolds with superior antibacterial and antiviral activities dramatically increased16. Antimicrobial polymers are desirable as surface coatings preventing viable bacteria adhesion17, for tissue engineering applications18, and as antibiofilm in healthcare and food packaging19. Fully organic antimicrobial polymers are abundant and well-documented20,21. In addition, many formulations for antimicrobial surfaces are based on the hybrid combination of an inorganic antimicrobial moiety with an organic one22. Many pharmaceutical agents in current clinical use containing metals23, and polymer-metallic composites are widely employed in the smart-packaging industry24 for the production of active biofilms25 and for the synthesis of nanosized theranostic instruments with high activity and broad response spectrum25,26. From the rational design of the structure–activity relationship, novel bio-responsive hybrid materials have been synthesized where the pivotal role of the metal cation is fully recognized27,28,29,30,31,32.

Widely used in organic synthesis33,34, Schiff bases are known as biologically performant molecules35,36. Very recently, the antimicrobial properties of a series of molecules containing salicyl-aldehyde groups were examined from both the experimental and the theoretical points of view37,38. The manageable Schiff base synthesis is the starting point to produce organic polymers able to chelate a suitable metal central cation. In many of the resulting hybrid materials, the biological/pharmacophore effects can be improved, including antifungal, antimicrobial, antiproliferative, anti-inflammatory, and antiviral activity21,27,33,39.

Palladium organometallic complexes are widely employed in bioinorganic chemistry40,41. Organopalladium tools are good inactivators of the world’s best-characterized, ubiquitous, and adaptable organism, Escherichia coli42,43. Escherichia coli is the most representative microbial species of the faecal coliform group, in turn, part of the larger family of total coliforms, present in the digestive tracts of animals, in wastes, in plants, and in soil material. Therefore, total coliform counts give a general indication of the sanitary condition of water, aqueous environments, and humid surfaces. Even better, testing for coliform bacteria can be a reasonable indication of whether other pathogenic bacteria are present44.

In previous works, we systematically investigated orthopalladated push–pull Schiff bases as stable chromophores and/or bioactive molecules45,46,47. The antagonist response of some low-weight palladium complexes to Escherichia coli bacteria gave interesting results that prompted us to expand the study towards polymeric systems.

This article reports the design and synthesis of a series of hybrid metallopolymers obtained by grafting the orthopalladated fragment Pd-L (see Fig. 1) on the O, N chelating sites of two Schiff base polymers (P1 and P2, see Fig. 1) with a different flexible spacer. The two organic polymers are AIE active materials with a solid-state light-yellow emission. They can be moulded and formed in thin layers and display by-contact antimicrobial activity against the total coliform group. To preserve emission and solubility, we grafted the orthopalladated fragment one site out of sixteen (samples named A) or one site out of eight (samples named B) chelating units, obtaining two hybrid series (see Fig. 1). The grafted polymers still display a solid-state emission gradually red-shifted depending on the organometallic fragment content. A comprehensive exploration of chemical-physical properties, optical behaviour, and antimicrobial assays was performed. Specifically, a good-to-excellent activity was recorded against the total coliform group, which significantly increased from the organic to the hybrid polymers. By correlating experimental data and theoretical DFT investigation, we could argue the pivotal role of the coordination core as an active electron bridge across the ligand system.

Figure 1
figure 1

Synthetic route to the hybrid polymers.

Experimental section

Materials

Commercially available starting products were supplied by Sigma Aldrich. Compound D was obtained as described in Refs.48,49.

1H NMR spectra were recorded in 1,1,2,2-tetrachlorethane-d2 with a Bruker Advance II 400MHz apparatus by Bruker Corporation, Billerica, MA, USA. Optical observations were performed using a Zeiss Axioscop polarising microscope by Carl Zeiss, Oberkochen, Germany, equipped with an FP90 Mettler microfurnace by Mettler-Toledo International INC MTD, Columbus, OH, USA. The decomposition temperatures (measured as 5 wt% weight loss) and phase transition temperatures and enthalpies were measured under nitrogen flow 50 mL/min, using a DSC Mettler Toledo DSC3+ at a scanning rate 10 °C min−1. TGA experiments were performed by a thermobalance Perkin Elmer Pyris Diamond TGA under airflow 25 mL/min, at a scanning rate 20 °C min−1. Absorption and UV–Visible emission spectra were recorded using JASCO F-530 and FP-750 spectrometers by JASCO Inc., Easton, MD, USA, at a scanning rate of 200 nm min−1, and on a spectrofluorometer Jasco FP-750 by JASCO Inc., Easton, MD, USA (excitation wavelengths set at the absorption maxima of the samples) at scanning rate 125 nm min−1. Thin films of the samples were prepared using an SCS P6700 spin coater operating at 600 rpm for 1 min (I step) and at 3000 rpm for 30 s (II step). Photoluminescence quantum efficiency values were recorded on quartz slides using a Fluorolog 3 spectrofluorometer by Horiba Jobin Instruments, SA, within an integrating sphere provided by an optical fibre connection. Morphological characterization was performed by a Scanning Electron Microscope (SEM) Leo 1530 Gemini by Zeiss on the spun films. The operating voltage was 6 kV for all the measurements performed. The polymers P1 and P2 intrinsic viscosity (η) at 130.0 °C in o-dichlorobenzene was measured with an Ubbelohde viscometer. Vapour-pressure osmometry (Knauer apparatus) of o-dichlorobenzene solutions at 130 °C was utilized to estimate the average molecular weights (MW). X-ray powder diffraction data were collected in reflection geometry using the Empyrean (PANalytical) diffractometric apparatus using Nifiltered Cu Kα radiation (λ = 0.15418 nm).

Organic polymers

Polymers P1 and P2 were obtained following a procedure described by us50. Polymerization was performed in solution starting from two different dianiline monomers (with different flexible spacers for P1 and P2) and the same dialdehyde monomer. Formulas of the precursors are reported in Fig. S1. As an example, in the synthesis of P1 a total of 4.397 g of the dialdehyde monomer bis(4-formyl-3-hydroxyphenyl) 4,4′-(dodecane-1,12-diylbis(oxy))dibenzoate was dissolved in 40 mL of boiling o-dichlorobenzene and 2.477 g of the dianiline monomer 4,4′-(dodecane-1,12-diylbis(oxy))dianiline was added under a nitrogen atmosphere. The reaction mixture was kept under vigorous stirring at boiling temperature, then poured into boiling n-hexane. The precipitated polymer was washed twice in n-hexane and oven-dried at 100 °C. Yields are virtually quantitative. The reaction time for the polycondensation was modulated to obtain medium molecular weight, suitable viscosity, and good adhesion properties. Specifically, we kept the reaction for 20′ in boiling o-dichlorobenzene. Intrinsic viscosity measurement in o-dichlorobenzene at 130 °C provided values of 2.50 dL/g and 1.65 dL/g for P1 and P2, respectively. The 1H-NMR spectrum of P1 and P2 (Figs. S2, S3) are consistent with the expected formula. Terminal groups are barely detectable at 11.20 ppm (OH proton) and 9.89 ppm (CHO proton) for both P1 and P2 spectra. The ratio between the areas of the diagnostic peaks of the polymers and the peaks due to the terminal groups at least ten monomer units in the polymeric chain. Vapour-pressure osmometry by a Knauer apparatus was used to measure the average molecular weights (MW) of P1 and P2 in o-dichlorobenzene at 130 °C. In both cases, we obtained a value ranging from 12 × 103 to 13 × 103. A detailed analysis of the thermal behaviour of the samples was performed in our previous article. The samples employed in this article replicate the expected thermal behaviour (see “Antibacterial activity” section). X-ray diffraction pattern recorded on P1 and P2 samples confirm a small amount of crystallinity50. In both cases, the solid samples melt to a mesophase (anisotropic nematic liquid) observable in crossed polarizers optical microscopy as a schlieren texture.

Synthesis of the grafted polymers

The grafted polymers were obtained by bridge-splitting reaction of the complex D on a preformed chain of P1 or P2 polymer (see Fig. 1). The grafting reaction was carried out by the same procedure. As an example, we describe the reaction to obtain P1-Pd(B). To 1.030 g of P1 (molecular unit: C64H74N2O10, 1030.53) dissolved in 15 mL of boiling 1,1,2,2-tetrachlorethane (TCE) 0.108 g of D (0.125 mmol) dissolved in 3 mL of hot TCE. To the reaction solution, 200 mg of K2CO3 was added, and the boiling mixture was kept under stirring for 2 h. The crude product was filtered and precipitated by pouring methanol at room temperature. The compound was washed in methanol twice and dried at 80 °C overnight under a vacuum. Yield is virtually quantitative. The 1H-NMR spectra of all grafted polymers were reported in Supplementary Materials (Figs. S4S7). Referring to the NMR pattern, the insertion of the Pd-L fragment is revealed by some diagnostic signals, although scarcely intense due to the low grafting degree. Compared to NMR spectra of the unbonded polymers, NMR spectra of all the hybrid polymers show two very small signals in the region from 5 to 6.5 ppm, due respectively to the proton in ortho and in para to the C-Pd bond, and a singlet is in the region from 2.5 to 3 ppm due to NMe2 protons. For comparison, the spectrum of compound D showing the same unique pattern can be seen in Fig. S8.

Molecular modelling

Ab-initio calculations were performed in the Density Functional Theory (DFT) approximation using the Linear Combination of Atomic Orbitals (LCAO) formalism. Then, the molecular orbitals were expanded in Gaussian Atomic Orbital (GAO) functions centred in the nuclei.

HOMO–LUMO analysis is obtained from the ground and first excited state calculations; in particular, the latter is achieved by adopting the Time-Dependent Density Functional Theory (TDDFT) formalism.

The software adopted is ORCA 5.0.351. A B3LYP hybrid functional52, with a 20% fraction of the exact Hartree Fock exchange, was used as previously described53,54. The dispersion interactions are evaluated by adopting the D3 Grimme empirical correction Double-z (for the light atoms) and Large Triple-z (for the metal atom) basis set was used. In this case, the def-SVP and def-TZVP basis sets of the Karlsruhe group were adopted55. This basis set was previously tested by us53,54,56. Therefore, they can be considered a good choice in terms of computational performance and accuracy of the results, providing a reliable comparison with the experimental data without renouncing a smart calculation.

In vitro antibacterial activity methods

The reference sample for colony development is urban wastewater collected on the Tyrrhenian side of Italy. The sample contains a collection of different bacteria, viruses, and parasites. As expected, microbiological analysis evidenced an abundance of Escherichia, Enterobacter, Citrobacter, and Klebsiella.

The standard 7010 C APAT IRSA-CNR method was followed for the collection, storage, and preparation of the samples and to evaluate the bacterial colonies. The wastewater was diluted 1: 1000 times to allow the optimal number of colonies to be read. A 100 mL aliquot was taken and placed in membrane filtration for each test. The wastewater sample was treated by filtering a water content of 100 mL on a nitrocellulose membrane with a porosity of 0.45 µm. The filtration system consists of a ramp with supports and containers (filter funnels) made of stainless steel, using an electrically operated vacuum pump or a water pump as a suction system. All operations are performed in sterile conditions, the supports and containers being sterilized in an autoclave or sterile disposable. Bacteria larger than the filter’s pores remain trapped on the surface of the filter itself. After filtration, each membrane is placed in a plate containing a specific culture medium that allows the growth of the bacteria.

In our case, the medium used for total coliforms was chromogenic coliform agar. Each polymer sample was dispersed in chloroform at a concentration equal to 1 mg/mL, i.e., 0.1% W/V. Each emulsion of the polymer samples undergoes sonication for 30′. On each 6 mm Petri dish agarised with m-Endo agar (Less), 5 mg of the polymer was homogeneously dispersed on the surface. The dispersion was carried out using a polymer suspension deposited on the agar, then evaporated under aseptic conditions in a biological hood.

The plates prepared with known concentrations of the tested complex were incubated at 37 °C for 24 h for both the investigated microorganisms. After incubation, the results were read by comparing the blank or without complex plates and those with the complex, observing a significant decrease in the number of colonies.

Results

Chemistry of the grafted polymers

Developing novel optical and biological-responsive tools meets the expectations of technological and scientific demands. Biologically performant macro or nano-sized AIE polymers11,57,58,59,60 play a key role among the modern monitorable theranostic tools61,62. Hybrid polymers obtained by grafting metalloorganic fragments on preformed chelating polymers combine the easy and tuneable synthesis with the specific features brought by the coordination fraction.

The grafting process synthetic strategy provides an easy-not-trivial alternative to the common practice of doping active molecules in a polymeric host matrix. Using preformed polymeric chains with suitable chelating sites leads to well-defined MW material where the active units are chemically bonded. Moreover, the grafting percentage can be turned to optimize desirable properties. Metals used as bridges along a conjugated organic system are expected to induce a more efficient electron π-conjugation path with significant spectroscopic and even biological differences between free ligands and metal complexes63,64,65.

The Pd–C organometallic bound was exploited to firmly graft non-symmetric coordination moieties on preformed polymeric chains in different percentages66,67. In the first synthetic step, the Schiff base L in Fig. 1 undergoes orthopalladation resulting in the aceto-bridged dinuclear species D. By the bridge-splitting reaction of D directly on the O,N chelating sites of P1 or P2 substrates, we grafted the Pd-L complex moiety on the polymeric skeletons (see Fig. 1), in different percentages.

As for the aceto-bridged dinuclear species D, the coordination moiety proved to have an excellent antimicrobial potential48. The ligand L has a π-electron conjugated system, dying properties, and proven antimicrobial activity, which increases in some related complexes68,69. Locked in the orthopalladated ring, the iminic bond of L benefits from stabilization. At the same time, the reaction of the orthopalladated Pd-L fragments with the chelating M sites of P1 and P2 produces a second stable, rigid cycle. The big differences between organic and hybrid polymers must be ascribed to the electron charge transfer involving the whole coordination core, as discussed in “Theoretical analysis” section70.

The organic polymers P1 and P2 are medium MW Schiff base polymers obtained by polycondensation reaction of a salicylic di-aldehyde and a di-aniline. Two equivalent O, N chelating sites for each monomer unit derive from the polycondensation reaction. The flexible spacers guarantee solubility and processability. Two different di-aniline were employed with different flexible chains. The most articulated triethylene glycol (TEG) spacer used in the synthesis of P2 ensues greater (about 30%) solubility compared to P1. Hence the indisputable advantage of better processability. Both polymers were expected to be emissive and pathogen antagonists34,36,37,71. Moreover, P2 was expected to be a better pathogen antagonist than P1. PEG (polyethylene glycol) based polymers exhibit a broad spectrum of antimicrobial activities and are employed as long-lasting antimicrobial surface coatings able to reduce the adhesion of bacteria72,73,74,75,76.

By tuning the grafting percentage of the organometallic fragment, we did search for the best balance between spectroscopic and biological properties. Specifically, we set the Pd-L moiety content as 0.25 mol ratio (Pd-L/Organic polymer) in type (A) samples and 0.125 mol ratio (Pd-L/organic polymer) in type (B) samples. Therefore, as two chelating sites are available per monomer unit both in P1 and P2, we grafted one site in sixteen in the case of the series named (A) and one site in eight in the series named (B) (see Fig. 1). As a percentage, type (A) samples were obtained using dinuclear complex D, about 5% w/w of the total, and type (B) samples using dinuclear complex D, about 10% w/w of the total. By not exceeding 10% D, we planned to achieve good antimicrobial ability without sacrificing the solubility and the AIE pattern.

Antibacterial activity

Many scientific studies have focused on the removal of bacteria77. Palladium (II) complexes from Schiff bases ligands are known for their significant activity against Staphylococcus aureus and Gram-negative bacteria, even higher compared to their free respective ligands. The group of total coliforms are rod-shaped, Gram-negative, mainly aerobic, and non-sporing bacteria. The group is a collection of Gram- negative bacteria, including Citrobacter, Enterobacter, Hafnia, Klebsiella, some species of Serratia, and Escherichia coli. They are considered classic indicators of contamination to evaluate the hygienic conditions of home and work environments and food and water.

The total coliform test provided us with extensive information on the inhibitory ability of our polymers. Moreover, we could use a real sample of urban wastewater without any pre-treatment.

A total coliform inhibition test was performed to investigate the ability of both organic polymers and hybrid polymers to mitigate or prevent bacterial growth. The colony-forming units (CFU) were evaluated at the same concentrations of polymer stock suspension (see “In vitro antibacterial activity methods” section). The CFU abatement results are summarized in Table 1.

Table 1 CFU abatement results for organic and hybrid polymers.

Due to the Schiff base M active moiety (see Fig. 1), unfunctionalized P1 and P2 samples were found to affect the growth of the total coliforms sample. As expected, the presence of a TEG spacer makes the biocide activity of P2 higher (quite double) with respect to P163,64,65,70.

All the hybrid polymers show over 50% inhibition of CFU, higher than the respective unfunctionalized polymer. By increasing the grafting percentage, the biocidal activity increases for both grafted series, i.e., samples (A) have lower activity than samples (B). Both P1-Pd(A) and P1-Pd(B) reach quite 100% CFU abatement. This is a remarkable result as compared with recent literature data78.

Contrary to what was expected, P2-Pd derivatives show antibacterial activity significantly lower than Pd1-Pd derivatives (Table 1). Structure-dependent antimicrobial activities were discussed in significant articles79,80,81. In addition to the surface effects due to the different spacers82, we could assume a role of the metal-chain interactions on the biofilm growth. Ethylene glycol (EG)-functionalized palladium (II) complexes are well known83,84. In our case, a crosslinked structure through weak interactions of the oxygen atoms of the TEG chains to the tetracoordinate metal centres may occur. As a result, the tetracoordinate core may distort, and the glycolic chains may change configuration with an unforeseen global effect on the biocidal activity85,86. The polymers’ inhibition efficiency is summarised in histograms in Fig. 2c. On the left of the same figure, picks of the Petri dishes treated with the wastewater sample are reported in the case of the most active polymeric sample P1-Pd(B) (Fig. 2a) compared with the control (Fig. 2b).

Figure 2
figure 2

Petri dishes of the wastewater sample for P1-Pd(B) (a) and control (b). Percentage inhibition efficiency of all the polymers by standard 7010 C APAT IRSA-CNR method (c).

Phase behaviour, optical response, and stability

The main-chain organic polymers are yellow semicrystalline materials with a nematic mesophase (see thermal behaviour summarised in Table 2). DSC curves of P1 and P2 (Figs. S9, S10) display high melting points of the crystalline fraction (above 220 °C) with a polymorphic behaviour in the solid state (two melting phenomena reported in Table 2, the least intense in brackets). Decomposition occurs with isotropization.

Table 2 Thermal behaviour of organic and hybrid polymers.

The introduction of the orthopalladated units forcing the ligand sites in a local planar or semi-planar arrangement causes a loss of the structural order both in the solid state and in the mesophase. The grafted polymers with lower palladium content P1-Pd(A) and P2-Pd(A) (see DSC curves in Figs. S9, S10, respectively) are semicrystalline materials as the organic precursors (see Table 2). After melting, both samples retain a nematic mesophase, clearly observable in crossed polarizers optical microscopy as a poorly mobile schlieren texture. The samples with higher palladium content P1-Pd(B) and P2-Pd(B), display different behaviour. P2-Pd(B) is a non-crystalline sample retaining a nematic mesophase observable as a poorly mobile schlieren texture above Tg (above 140°C, see DSC curve in Fig. S10). Decomposition occurs before isotropization. P1-Pd(B) is an amorphous material fluid and isotropic above Tg (above 170 °C, see DSC curve in Fig. S9), missing the mesophasic behaviour above Tg. This material can be deposed as an orange film that is highly transparent (see Fig. 2), a desirable requirement for some coating applications17,72. SEM analysis confirmed the homogeneous non-structured nature of the film.

All the polymeric samples show remarkable thermal stability with decomposition above 300 °C. By TGA analysis, we also obtained the Pd (0) content of the metallated polymers. Fixed the percentage of dimer compound D at 5% and 10% (and consequently the percentage of Pd-L fragment), the expected Pd content varies between just over 1% and just over 2% for type A and type B samples, respectively. As can be detected from Table 2, the grafted polymers show quite nominal palladium content proving the grafting route an efficient synthetic way to obtain hybrid polymers.

Organic polymers and grafted polymers in natural light are from yellow to orange materials (see Fig. 2). The intriguing spectroscopic pattern of P1 and P2 was never explored before50. Here we checked their absorption/emission properties, highlighting their potential as fluorophores. Specifically, we found that P1 and P2 polymers are poorly emissive in a diluted solution. Conversely, they act in the AIE mode displaying yellow emission in the aggregate phase. Absorbance and emission maxima measured on spin-coated thin films (sample photographed in Fig. S11), RGB emission colour space, Stokes shift, and photoluminescence quantum yields (PLQYs) for both organic and hybrid polymers are summarised in Table 3.

Table 3 Spectroscopic behaviour of organic and hybrid polymers.

As expected, Refs.77,87,88 the non-emissive Pd-L grafted moiety causes a decrease in the emission performance. Specifically, by increasing Pd-L content from type (A) samples to type (B) samples, the emission band broadens and undergoes a redshift (Figs. S12, S13). At the same time, PLQYs drop. Therefore, P1 and P2 show medium-to-good PLQYs. Type (A) samples still retain medium PLQYs. In addition, Stokes shifts are large (183 nm and 189 nm, respectively) as desirable to minimize emissions-absorption overlap89,90. In type (B) samples PLQYs lower to about 2%. Samples with higher palladium content than type (B) were not examined due to the non-emissive pattern and poor solubility (see solubility in Table 2). Interestingly, the yellow/orange fluorescence emission of organic and hybrid polymers (compared in Fig. 3) can be naked eye observed by using the standard 365 nm excitation wavelength of a commercial UV lamp.

Figure 3
figure 3

Eye-catcher of the polymers’ emission under a UV–Visible lamp at 365 nm. The ovals were cut out from the real thin film samples and placed from the appropriate side of the tridimensional RGB colour space diagram (red-green-yellow edge).

The spin-coated films of all grafted samples show strong adhesion to the glass support and resist common organic solvent redissolving and mechanical scratching.

A systematic analysis of the stability in ordinary (dry and wet) conditions of use was performed on the grafted samples. A set of thin films of type (A) and type (B) polymers spin-coated onto quartz slides were kept at room temperature and under daylight for over 2 months. Their habitus and absorbance/emission spectroscopic pattern remains unaltered. Another set of thin films of type (A) and type (B) polymers spin-coated onto quartz slides was immersed in aqueous solutions buffered at slightly acidic/basic pH (respectively at 5.0, 5.5, 6.5, 7.5, 8.0 pH values) for 2 weeks. As a result, the films do not detach from the quartz holder, and their habitus and spectroscopic absorbance/emission pattern remains unaltered. Finally, the 1H-NMR spectrum of all grafted samples recorded before and after 2 weeks resulted unaltered. In addition to quarts, various commercially available polymers were used as coating substrates.

Interestingly, the films of the cyclopalladated polymers do not detach from commercial samples of PE, PMMA, PET, PVC, PS, PP and their habitus and spectroscopic pattern remain unaltered for months. As a practical check, P1-Pd(A) was coated onto a flexible commercial PE slide and on one face of a commercial laboratory flask (BD falcon for cell culture) made of PS. The deposition of a thin layer was trivially made by brushing a polymer solution in 1,1,2,2-TCE, as the solubility of P1-Pd(A) in 1,1,2,2-TCE at 50–60 °C is still sufficient for brushing the solution onto a support. In both cases, the emission can be well appreciated using a UV–Visible lamp at 365 nm (see Fig. 4).

Figure 4
figure 4

P1-Pd(A) deposed on a commercial PE flexible slide (a), on a BD falcon flask in PS (b), and turned into a die-cast film (c), irradiated by a UV–Visible lamp at 365 nm.

Solubility is a relevant parameter for the polymers’ processability and coating operations. In our case, all polymers are insoluble in low-boiling organic solvents. Therefore, the hybrid polymeric layers are stable in contact with common solvents such as acetone, chloroform, ethyl acetate, and hexane. On the other hand, the organic polymers and the (A) type polymers can be dissolved in boiling 1,1,2,2-TCE or ortho dichlorobenzene (o-DCB), and their concerted solutions can be managed at 60 °C for the brushing/coating time. Contrarily, in the (B) type polymers, solubility drops, and the polymers cannot be easily processed. Finally, all the polymers can be extruded in fibres or die cast in plates by heating the solid material above Tg (200 °C). For example, in Fig. 4c is shown a small die-cast plate of P1-Pd(A) obtained from the solid sample using a heating press at 190 °C for 5 min. In all cases, the casting process can be carried out without damage.

The diagram in Fig. 5 summarizes the main characteristics of the polymers in terms of monitorability (i.e., emission intensity quantified by PLQY value) and bacterial inhibition efficiency. The better performance (evidenced in the blue oval) is noticeable of P1-Pd(A) polymer which exhibits high values of both PLQY and inhibition efficiency.

Figure 5
figure 5

Summary of the main features of the polymers. The blue points represent the PL quantum yield values measured on polymer thin films (from Table 3). The red points represent the antibacterial inhibition efficiency against total coliform bacteria (from Table 1).

Theoretical analysis

The theoretical models, P and P-Pd, are simplified moieties resembling the aromatic part of the polymer (representative of both P1 and P2) ending with methyl groups, respectively, before (P) and after (P-Pd) coordination of the orthopalladated fragment (see Fig. 6). By the classical molecular orbital representation of the HOMO to LUMO transition, it could be hard to describe “which electrons” go from “which occupied orbitals” to “which unoccupied orbitals”. To overcome the issue, the excitation process was examined by the electronic charge transfer between the first excited state and the ground state using a more intuitive representation. The process was evaluated considering the differential electron density. The differential electron contributions (rdiff = rexc − rgr) is the electron transfers from the first excited state and the ground state. The electron differences between HOMO and LUMO are represented as the density plot in Fig. 6, where are clearly detectable the charge transfer process and the involved molecular sites.

Figure 6
figure 6

The isosurface plot is set to 0.002 |e|/a0 iso-level density. The magenta lobes are the positive contribution to the differential electron density rdiff; the green lobes are the negative isovalues. In (a) is represented the P precursor, in (b) is represented P-Pd complex moiety. The enlarged detail of the differential electron density on the metal cation is in the inset (c).

A significant difference is detected for P compared to P-Pd. This is because the main electron charge transfer contribution in the HOMO–LUMO excitation of the precursor P entails a redistribution of the electronic path involving the oxygen of the chain as the charge donor (negative differential densities represented as the green lobes in Fig. 6a) and the aromatic groups (positive differential densities represented as the magenta lobes in Fig. 6a).

In the case of P-Pd (Fig. 6b), the excitation process involves the charge transfer from the organometallic moiety Pd-L to the O, N chelating core of P. A role can be ascribed to Pd (II) cation as an electronic bridge between the organic chelating site of P and the orthopalladated ligand L. The frontiers orbitals localized on the metal cation are shown in the inset, Fig. 6c.

Finally, the energy gap (eV) in the HOMO–LUMO transition of P compared to P-Pd underwent a significant decrease from 3.64 eV of P precursor to 2.03 eV of P-Pd complex moiety.

Conclusions

In the search for new bio-responsive tools, hybrid materials obtained by the integration of a coordination fragment into an organic matrix play a role. By grafting an orthopalladated fragment to the O, N chelating site of two Schiff base polymers, we obtained two series of organic-metallorganic polymers. The Pd-L moiety was added in two different percentages with the aim of preserving the AIE activity and improving the bacterial inhibition ability of the organic polymers. While the most concentrated grafted samples suffer a sudden decline of PLQYs, the most diluted grafted samples still retain medium PLQYs. All hybrid samples display good-to-excellent activity against the total coliform group of urban wastewater, increased as compared to the related organic polymers. The best balance between PL performance and antimicrobial activity was obtained with the P1-Pd(A) sample, which shows almost total bacterial abatement and retains 24% PLQY in the solid state. These results suggest such a hybrid set as a stable, monitorable, and antimicrobial polymer for functional coatings.