Synthetic non-classical luminescence generation by enhanced silica nanophotonics based on nano-bio-FRET

Fluorescent silica nanoparticles (NPs–(SiO2–Fluo)) were synthesized based on the classical Störber method for cyanobacteria labelling. Modified mono-coloured SiO2 NPs with fluorescein (Fl) and rhodamine B (RhB) were obtained (NPs–(SiO2–Fl) and NPs–(SiO2–RhB)). Moreover, multi-coloured SiO2 NPs, via the incorporation of both emitters (NPs–(SiO2–RhB–Fl)), were tuned for optimal emissions and the biodetection of cyanobacteria. NPs–(SiO2–Fl) and NPs–(SiO2–RhB–Fl) were optimized for detection via laser fluorescence microscopy and in-flow cytometry with laser excitation and fluorescence detection. By TEM, homogeneous SiO2 NPs of 180.0 nm in diameter were recorded. These sizes were slightly increased due to the covalent linking incorporation of fluorescent dye emitters to 210.0 nm with mono-coloured fluorescent modified amine-organosilanes, and to 340.0 nm in diameter with multi-coloured dye incorporation. NPs–(SiO2–Fluo) showed variable emission depending on the dye emitter concentration, quantum yield and applied luminescent pathway. Thus, mono-coloured NPs–(SiO2–Fl) and NPs–(SiO2–RhB) showed diminished emissions in comparison to multi-coloured NPs–(SiO2–RhB–Fl). This enhancement was explained by fluorescence resonance energy transfer (FRET) between Fl as a fluorescent energy donor and RhB as an energy acceptor produced within the nanoarchitecture, produced only in the presence of both fluorophores with the appropriate laser excitation of the energy donor. The depositions of the nano-emitters on cyanobacteria by non-covalent interactions were observed by TEM and laser fluorescence microscopy. For multi-coloured NPs–(SiO2–RhB–Fl) labelling, bio-FRET was observed between the emission of the nano-labellers and the natural fluorophores from the cyanobacteria that quenched the emission of the whole nano-biostructure in comparison to mono-coloured NPs–(SiO2–Fl) labelling. This fact was explained and discussed in terms of different fluorescence energy transfer from the nanolabellers towards different natural chromophore coupling. In the presence of NPs–(SiO2–RhB–Fl) and NPs–(SiO2–RhB), the emission was coupled with lower quantum yield chromophores; while upon the application of NPs–(SiO2–Fl), it was coupled with higher quantum yield chromophores. In this manner, for enhanced luminescent nanoplatform tracking, the multi-coloured NPs–(SiO2–RhB–Fl) showed improved properties; but more highly luminescent bio-surfaces were generated with mono-coloured NPs–(SiO2–Fl) that permitted faster cyanobacteria detection and counting by laser fluorescence microscopy, and by in-flow cytometry with laser excitation and fluorescence detection.


Introduction
Bacterial detection has been shown to be a high-impact research eld within biology, biochemistry and clinical chemistry that is still a challenge due to needs related to biodetection coupled to higher level of information collected from individual biostructures. In this manner the signal collected should transduce from the total biostructure surface to arrive at molecular event detection within the biostructure. For these reasons the advent of bioimaging, by different techniques and chemical approaches, is in progress with high impact on new advanced instrumentation and new products on the market available for researchers as well as to professionals from different elds. In addition new phenomena from physics coupled to chemical properties permit new research studies and developments.
In order to realize bioimaging, it could be applied different approaches, but it should be mentioned that uorescence due to its intrinsic high sensitivity is widely used and well developed. However, it is still a challenge for targeted enhanced applications.
From the literature it could be mentioned tuning in vivo Gram positive and negative bacteria with red uorescent dyes for imaging, at 650 nm emission wavelengths, 1 where it was studied and overcame many aspects related to the specic incorporation of organic molecules within membranes and background signalling. Moreover, other strategies as applications of different luminescent nanoarchitectures should be highlighted. Examples are self-assembled quantum dots as uorescence resonance energy transfer (FRET) donors in the presence of uorescent modied saccharide membranes of Escherichia coli bacteria. 2 Others are uorescence energy transfer inhibition bioassays for cholera toxin based on galactosestabilized gold nanoparticles and amine-terminated quantum dots. 3 In these examples, it was highlighted the importance of the main role of the control of energy transfer by light stimulation for biodetection applications. In addition studies related to high-energy electromagnetic elds generated from metallic surfaces within the near eld at the nanoscale and their interactions with uorescent molecules showed enhanced emissions named as metal enhanced uorescence (MEF), 4 which permitted new biolabelling approaches based on natural uorescent molecular sensing within biomembranes in the presence of deposited silver nanoparticles at the right distance. 5 In this manner, the importance of studying different energy pathways within biostructures via quantum experimental approaches was shown as well. Recently it was reported how energy transfer by photosynthetic proteins within bacteria produced modications from the excited state of excitonic superpositions to the basal state with energy migration, suggesting the quantum role of non-classical energies in natural photosynthetic systems. 6 If the focus is on nanomaterials for biolabelling applications, one should mention the use of biocompatible hybrid nanomaterials from different synthetic and natural sources as for example silica and gold nanomaterials with optical transparent 7 and optically active properties respectively in addition to biocompatible properties that permit silica being considered as an inorganic collagen, 8 and gold nanoparticles applied for laser-assisted therapy for controlled CRISPR delivery in vivo. 9 In addition should be highlighted silica nanomaterials and applications based on nanophotonic luminescent nanoparticles with the incorporation of organic uorescent dyes 10 within silica nanocomposites for biomedical imaging, 11 multiple homogeneous immunoassays based on quantum dots-gold nanorods by FRET nanoplatforms, 12 gold core-shell silica nanoparticles for biosensing 13 by MEF, silica waveguides by resonant uorescent core-shell nanoparticles by MEF, 14 and drug delivery applications via a controlled silica porosity. 15 So, silica has been widely applied and is still a key nanomaterial for nanophotonic developments. However, new biomaterials are being developed, for example based on controlled nano-aggregated biomolecules with uorescent properties for bioimaging applications. 16 In particular for this research study, our interest was in cyanobacteria due to their environmental implications and optical active properties. Cyanobacteria or blue-green algae are the dominant phytoplankton group in eutrophic freshwater bodies worldwide. Moreover, climate change has contributed to increases in cyanobacteria occurrence in surface waters, and the risk of harmful algae blooms. 17 Therefore, many countries or jurisdictions have implemented specic water quality regulations to protect public health and safety. Drinking water quality guidelines related to cyanobacteria are based on maximum acceptable concentrations of toxins (e.g., microcystin-LR) in treated water (e.g., 1.0 mg L À1 proposed by WHO in 1999) or high levels of cyanobacterial cells (e.g., $100 000 cells per mL) in water supplies (WHO, 2011). 18 Despite the time required for identication, conrmation, and enumeration of cyanobacterial cells, direct microscopic enumeration is the simplest and most cost-effective method still used. In natural samples, this method involves some limitations such as the weak contrast of cells against the background, high species diversity, variable morphology of individual cells, and complexity of cell aggregates or units (colonies, entangled laments etc.).
Indirect quantication methods also have been developed to estimate cyanobacterial cell concentrations in water, such as ow cytometry, antibody-mediated immunouorescence microscopy assays, PCR-uorescent fragment detection, qPCR molecular probes using sandwich hybridization, and in situ uorescence. However, all of them have some drawbacks, and cost-effective, fast, and reliable cyanobacterial cell identication and enumeration methods are thus much needed. 19 Therefore, imaging-based enumeration methods appear to be promising for rapid and low-cost water quality monitoring of cyanobacteria, and uorescent silica nanoparticles could be a means to improve the detection limits and sensitivity of these methods.
Moreover, these types of bacteria were evaluated as optically active biostructures 20 that interact by non-classical light pathways with luminescent nanoplatforms for potential biotechnological applications 21 as well.
For these reasons, our interest was focused in the design and synthesis of tunable hybrid nanomaterials based on uorescent silica nanoparticles with enhanced properties based on FRET for biolabelling applications.
In this manner, for this research communication uorescent emission properties were tuned by the incorporation of well overlapped spectroscopic properties and optimal quantum yields from molecular donor-acceptor pairs within silica nanoparticles. These nano-emitters were applied for noncovalent cyanobacteria labelling and detection by enhanced uorescence imaging recorded with laser uorescence microscopy. Then, evaluation was by in-ow cytometry with nanobiostructure detection and counting with laser excitation and uorescence detection.
Transmission electron microscopy (TEM), JEM-1230, JEOL, with an operating voltage of 200 kV, was used for determination of nanoparticle size.
UV-visible and spectrouorimetric determinations were carried out with a Varian UV-50 Cary 50 Conc. and a Cary-Eclipse respectively. Lifetime measurements were done with a PicoQuant FluoTime 2000.
The ow cytometer was from BD, model FACSCalibur, with laser excitation at 488.0 nm and 555.0 nm with standard lters at 533/30 and 585/40 for Alexa Fluor 488-A and AF5555-a.
An ultrasonic bath (Branson 2510) was used for the dispersion of the reagents and colloids. Centrifugation was done using an Eppendorf Centrifuge 5804 (range 7500-8000 rpm).
Data analysis was performed with Origin (Scientic Graph system), version 8.
Ultra-ltrated and deionized water was obtained using a Millipore apparatus.
Micrometer multicolor beads from BD Company were used as control particles for in-ow cytometry.
The wild population of Microcystis aeruginosa was concentrated from an atoxic bloom collected in San Roque water reservoir (Cordoba, Argentina).

General procedures
Silica nanoparticles were synthesized based on the classical Störber method. 22,23 In order to do that, the TEOS concentration was adjusted by variable volume addition of mL aliquots into basic ethanol solution (pH ¼ 9.00) adjusted with concentrated ammonium hydroxide. For the TEOS reaction, variable mL aliquots of TEOS were added, maintaining constant the ratios of reagents as described in the following. The ratio of TEOS/ ethanol/H 2 O/NH 4 OH was 150/2300/80/620. For example, for a typical synthesis of varied sizes of silica nanoparticles, 20.0, 40.0 and 80.0 mL were added of a concentrated solution of TEOS Fluorescent silica nanoparticles were obtained by incorporation into the described silica nanoparticle synthesis uorophores covalently bonded to modied organosilanes dissolved in ethanol (Scheme 1). The covalent linking of Fl and RhB was done by activation of their carboxylic groups with NHS/ EDC and nucleophilic attack from the amine group of APS. 13 For a typical reaction, 2 mg of RhB or Fl was added in the presence of 10 times higher concentrated APS and NHS/EDC in 2.0 mL total volume of ethanol. Thus, APS-Fl and APS-RhB were obtained.
The silica nanoparticles were centrifuged at 8500 rpm and redispersed in anhydrous ethanol.
The concentrations of uorophores loaded within the nanoparticles were determined by static uorescence. These values were obtained from the difference of the known APS-uorophore concentration added and the concentration of the non-incorporated APS-uorophores collected from the supernatant of the colloidal dispersion media of the synthesis. So, calibration curves were made for the concentration of APS-uorophores used for the synthesis of the different uorescent silica nanoparticles. At the same time the determined values were corroborated by absorption measurements. Moreover, additional controls for dye content within coloured silica nanoparticles were developed by disintegrating the nanostructures in strong (pH ¼ 1) acid media and sonication for a period of 4 h, and then quantifying the respective uorescent dyes incorporated.
Aer the synthesis of the uorescent silica nanoparticles, their emission was controlled to regular levels before use. The control was done by static uorescence and laser uorescence microscopy. To do that, the samples were centrifuged and controlled the presence of nanoparticles in the supernatant by dynamic light scattering (DLS). Thus was evaluated the uorophore leakage from the cleaned supernatant, and the standard emission intensity of nanoparticles from the re-suspended sample. No leakage was recorded for the samples used in the different experiments. And the percentage of variation within different measurements for the same samples was below AE5%. In this manner, were investigated samples with low background emissions determined by laser uorescence microscopy.
Fluorescence emission spectra were measured with an excitation wavelength equal to the wavelength of maximum absorption of the uorescent dyes (l exc ¼ 480.0 and 539.0 nm for Fl and RhB respectively). In order to conrm that the maximal emission uorescence was measured in these conditions, the excitation wavelength was evaluated by measuring 3D uorescence emission spectra. For emission and excitation uorescence spectra, the excitation and emission bandwidths were set at 10 nm. The PMT gain was medium. All the measurements were performed at (25.0 AE 0.1) C, with the temperature of the cell compartment being controlled with a Haake K10 circulator under continuous stirring.
The uorescence lifetime decay measurements of the uorescent silica nanoparticles were performed in ethanol.
The cyanobacteria bloom Microcystis aeruginosa was concentrated from a sample collected in San Roque water reservoir (Cordoba, Argentina). Cell counts of Microcystis aeruginosa were performed with a standard optical microscope using a haemocytometer and then followed by measuring optical density (OD) at 600 nm (OD of 0.1 corresponds to a concentration of 10 8 cells per mL). The conservation of these samples was performed in diluted (1/10) phosphate-buffered saline (PBS buffer) aqueous solution used for typical DNA hybridization assays.
From this concentrated colloidal dispersion of bacteria, dilutions were done at intermediate and diluted concentration levels (0.2 and 0.05 OD values at 600 nm respectively). Each inoculum was examined with the microscope to conrm its composition and the dominance of Microcystis aeruginosa within the sample (>99% cells counted correspond to these cyanobacteria).
For cyanobacteria-nanoparticle interaction, a dispersion of cells was prepared from the bloom sample of Microcystis aeruginosa. The bacterial concentrations were determined by measuring OD at 600 nm at intervals of 30 min (OD of 0.1 corresponds to a concentration of 10 8 cells per mL). In this manner, from a concentrated dispersion of cyanobacteria in aqueous media, variable dilutions were prepared depending on the number of biostructures intended to be determined by the optical microscopy techniques used. Thus, by bright-eld confocal microscopy as control, from individual bacteria were obtained micro-aggregates of bacteria. For uorescent labelling, the dispersions were in contact with variable additions of mL aliquots of concentrated uorescent silica nanoparticles for a 4 h period of time (Scheme 2). For typical cell labelling, 0.5 mL of concentrated sample in water was added into 2.0 mL of colloidal dispersion (total volume of 2.5 mL). The concentrations of uorescent nanoparticles as nanolabellers were within the interval of 9 Â 10 8 to 10 10 NPs per mL depending on the cyanobacterial bloom concentration used.
Aer that the samples were observed by laser uorescence microscopy with a minimal volume added (1 drop of 20 mL) on a microscope glass slide (covered aer addition with a coverglass). For the in-ow cytometry analysis, contour plots of sidescattered light (SSC; proportional to cell granularity or internal complexity) vs. forward-scattered light (FSC; proportional to cell-surface area or size) were used to characterize distributions of uorescent event detections. Laser excitations at 488.0 nm and 555.0 nm with emission lters of 530/30 nm and 585/42 nm were used.
The nano-imaging was recorded by variable look-up table image edition (LUT) of brightness and contrast parameters. For bio-and nano-imaging, green, red-green and re LUT were applied, depending on the degree of detail tracked. For images generated with higher contrast with the background, green and re LUT were used; while for differentiated intensities recording, red-green LUT was used. For the image edition, the background signal was subtracted that never overcame 10% of the higher intensity collected.

Characterization of mono-coloured and multi-coloured uorescent silica nanoparticles
Silica nanoparticles were obtained by the Störber method with varied diameters depending on the added TEOS concentrations.
Sizes of 380.0, 250.0 and 200.0 nm were observed by TEM. From all batches of colloidal dispersions were observed homogeneous silica nanoparticles (Fig. 1a), with well-dened spherical shapes ( Fig. 1b) accompanied by small trimer and dimer formations ( Fig. 1c and d). The addition of the uorescent dyes was done by their conjugation with APS for incorporation within the polymerized TEOS organosilane by covalent linking.
The uorescent hybrid nanoarchitectures produced increased in size in the interval of 10 to 40% depending on the added conjugated uorophore concentrations. For example, silica nanoparticles of 180-200 nm ranges of diameters were incorporated with [RhB] and [Fl] of 0.39 and 0.27 mM respectively, which produced increased sizes of 20% for monocoloured to 40% for multi-coloured nanoparticles (Fig. 2).
Moreover, the one distribution of well-shaped Gaussians recorded from these nanoparticles showed the good dispersibility in ethanol as well as in aqueous colloidal dispersions. In order to verify the stability of the obtained nanoparticles, different measurements were recorded within a 10 minute period of time. Thus, the intensities and sizes of nanoparticles were stable in the mentioned period of time. For longer periods of time the intensities diminished due to the reduced number of detected nanoparticles. However, it should be highlighted the fast dispersibility of these samples by just shaking them. The zeta-potential measurements were in the interval of À20 to 30 mV. These measurements correlated with typical values from well-dispersible free gold and silver coresilica shell nanoparticles as well. 4,24 Moreover, the sizes determined by DLS correlated with determinations by TEM images (Fig. 3). Moreover, it should be mentioned that the addition of the uorescent dyes did not modify the original spherical shapes of the different hybrid nanoparticles. For NPs-SiO 2 -RhB-Fl, sizes within the 330-  Then by laser uorescence microscopy, the nanoparticles obtained were evaluated. In this manner were recorded strong variable emissions from uorescent nanoparticles with variable dimensions depending on the uorophores incorporated and the emission pathways involved. From uorescent silica nanoparticles modied with Fl (SiO 2 -Fl) were recorded strong emission intensities from reduced sizes close to individual nanoparticle dimensions determined by TEM and DLS. Moreover NPs-(SiO 2 -Fl) nanoparticles showed good dispersibility and homogeneous nanostructures were detected (Fig. 4a); however more enhanced uorescent nanoparticle detections were recorded from multi-coloured uorescent silica nanoparticles (Fig. 4b).
About the stability and well-dispersible characteristics of the nanoparticles obtained within colloidal dispersions, it should be mentioned that the re-dispersion of decanted nanoparticles was done easily by simply manual shaking. So, for example, a fast re-dispersion of decanted nanoparticles in glass vials with a clear transparent and limpid solution was modied to bright yellow-orange colloidal dispersions for NPs-(SiO 2 -RhB-Fl), yellow for NPs-(SiO 2 -Fl), and purple for NPs-(SiO 2 -RhB) nanoparticles. Moreover, this phenomenon was observed as  This journal is © The Royal Society of Chemistry 2020 RSC Adv., 2020, 10, 20620-20637 | 20625 well by a change of opalescence from non-coloured silica nanoparticles.
The NPs-(SiO 2 -Fl) emission intensities showed bright and clear dots at 225.0 nm (inset (i) in Fig. 4). These nanoparticles produced a strong green uorescent core surrounded with gradually diminished red uorescence intensities (inset (ii) in Fig. 4) from dual-coloured red-green LUT image edition. However their intensities were lower than those of multicoloured silica nanoparticles with the incorporation of RhB and Fl (SiO 2 -RhB-Fl) (inset (iii) in Fig. 4). The multi-coloured NPs-(SiO 2 -RhB-Fl) nanoparticles showed increased emissions of at least 35% accompanied by the generation of more enhanced uorescent surfaces (inset (iv) of Fig. 4) in comparison to mono-coloured silica nanoparticles. The enhanced emitter surface core green highlight generated from multicoloured NPs-(SiO 2 -RhB-Fl) nanoparticles was explained by an improved uorescence energy routing through the 3D silica nanostructure aer interaction with the laser beam. In this manner were recorded higher emission intensities from more highly luminescent surfaces that generated bigger sizes of nanoparticles recorded by laser uorescence microscopy ( Fig. 4) than by TEM (Fig. 3).
Moreover, it should be mentioned that by single uorescence nanoparticle analysis, the mono-coloured and multicoloured uorescent nanoparticles showed homogeneous distributions of emission intensities as it was previously described. For this reason, at this point it should be claried that the observed emission differences from the nanoparticles within colloidal dispersions ( Fig. 4a and b) were attributed to nanoparticles randomly detected in Brownian motion in different planes and deep within a conned mL-volume drop added on the glass microscope slide.
Moreover, it should be highlighted that the distributions of dimeric SiO 2 nanoparticles based on non-covalent interactions previously mentioned (insets (iii) and (i) of Fig. 4) were explained by an optimal ratio of sizes and interaction strength. Due to the chemistry involved in the developed silica nanoparticles, the contributing forces were polar non-covalent interactions from the hydroxyl groups of silanol accompanied as well by attractive van der Waals interactions. These noncovalent interactions generated from nano-surfaces could be explained by Hamaker constants. 25,26 The Hamaker constant considers the ratio of non-covalent interaction force and the available nano-surface in contact. 27 In this manner, for example, were reported forces between dimers of larger sized polystyrene beads in the range between 0.3 and 50 pN in the presence of controlled ionic strength. 28 Thus, the higher frequency of dimeric forms of the silica nanoparticles obtained by us was explained as due to an optimal ratio of available surface and forces for intermediate particle sizes of 200-300 nm, and not just obtained by Brownian motion and encountering. This fact prompted our interest to study potential applications of dimeric forms by chemical modication of the nano-surfaces with short molecular spacers 29-31 for non-classical light generation 32 and nano-resolution. 33 In addition, the enhanced surfaces accompanied by higher emission properties from SiO 2 -RhB-Fl nanoparticles were explained by FRET. 34 Both uorophores showed well overlapped spectroscopic properties from the emission of the energy donor (Fl) to the absorption of the energy acceptor (RhB) (Fig. 5). In addition, for the uorescent energy donor 35 were reported three times higher quantum yields than for the energy acceptor. 36 This enhancement occurred only when the samples were excited at 488.0 nm that corresponded to the maximal absorption of Fl as uorescent energy donor. While for 543.0 nm and 555.0 nm laser excitations, for only the energy acceptor RhB stimulation, diminished emissions were recorded in comparison to 488.0 nm laser excitation. Moreover, controls of monocoloured silica nanoparticles showed drastic reductions of their emissions in comparison to multi-coloured nanoparticles. These phenomena via FRET pathways were previously studied by us. 37 The observations achieved by laser uorescence microscopy were recorded with emission lters in the interval of emission wavelengths of 510-625 nm for 488.0 nm laser excitation to record the complete emission band of both uorophores considering all the uorescence emission phenomena as individual uorophore emission and coupled phenomena via FRET. Similar observations were recorded when the interval of the emission lters was changed to longer emission wavelengths, as 575-650 nm (with 488.0 nm laser excitation), to diminish the contribution of the Fl uorescent reporter and increase the contribution of the emission from the FRET pathway. However, for 543.0 nm and 555.5 nm laser excitation of only the RhB uorescent reporter with longer interval of emission window, at 575-650 nm, their emission intensities were drastically diminished in comparison to 488.0 nm laser excitation.
In order to tune the optimal emissions considering the wellknown quenching by intermolecular energy homo-transfer for Fl 38 and RhB, 39 the concentration of both uorescent laser dyes was varied. Then in optimal concentration conditions for maximal emissions, and neglecting reduction of emission by quenching in conned volumes, the ratio of concentrations between RhB and Fl (ratio of RhB : Fl) was varied. In this manner with optimal excitation of Fl as uorescent energy donor, variable uorescence emission was recorded depending  (Table 1). As could be observed from these results, both emissions bands had higher emission than for mono-coloured SiO 2 NPs.
Moreover, for RhB : Fl ratios of 3 : 0.1 and 3 : 0.5, ratios of intensities between the emissions of Fl in the presence and absence of RhB (Fl/Fl ref ) of 3.0 and 1.9 respectively were observed (Table 1). So, the tendency showed a diminution and opposite direction in comparison to lower concentration of RhB as energy acceptor. This trend was produced by the quenching effect from dimeric species of both uorophores at higher concentrations of RhB and Fl.
In this way should be highlighted the ratio of intensities between RhB and Fl (RhB/Fl) emission bands increasing the RhB concentrations from RhB : Fl ¼ 1.0 : 0.5 to 3.0 : 0.5, and 1.0 : 0.1 to 3.0 : 0.1 that generated a reduction of their emissions (Table 1). So, considering only increasing the concentration of RhB as energy acceptor, the emissions were reduced caused by homo-transfer and quenching. It is known that for higher concentrations of these dyes, homo-transfer was    This journal is © The Royal Society of Chemistry 2020 RSC Adv., 2020, 10, 20620-20637 | 20627 produced due to the close intermolecular proximity accompanied by the formation of quenched dimeric species. 40 This phenomenon was shown by other similar derivatives such rhodamine 6G adsorbed on titanium oxide nanoparticles depending on the added concentration. 41 However, tuning concentrations to obtain the right ratio of RhB : Fl was required to obtain FRET pairs. The RhB/Fl emission bands were controlled and maximized within conned silica nanoparticles avoiding the quenching contribution.
Then, with excitation at 515.0 nm, for optimal excitation of RhB as uorescent energy acceptor and partially Fl as uorescent energy donor, noted were diminished emissions accompanied by small intensity increase with addition of higher Fl concentrations. Moreover for higher RhB concentrations a quenching effect was observed. In this manner it was shown how by controlling the excitation wavelength it was possible to activate or deactivate different uorescent emission pathways within a conned volume at the nanoscale with incorporation of different emitters (Fig. 7).
In this manner these nanoparticles showed excellent properties for nano-tracking and biolabelling applications based on their tuneable emission properties. Moreover it should be highlighted that these silica nanoplatforms showed potential chemical surface modications for bioconjugation by covalent and non-covalent interactions based on their polar surfaces given by hydroxyls of silanol groups. At this point, it should be mentioned for example that already reported were interactions of aminated and thiosulfonated modied silica nanoparticles with Escherichia coli, 42 permitting homogeneous deposition of the nanoparticles over these biostructures. In this study was shown the implication of non-covalent interactions such as polar interactions, van der Waals forces, and strong hydrogen bonding between the biomolecules placed on the bacterial membrane and the modied hydroxyl groups and free silanols.
For these reasons their applications were evaluated as nanolabellers for cyanobacteria labelling.

Bioimaging based on uorescent cyanobacteria nanolabelling
In order to apply these uorescent silica nanoparticles for cyanobacteria labelling, variable aliquots of bacteria were added within concentrated conditions of the different optimized uorescent nano-labellers. To do that were chosen smaller nanoparticle sizes of 180-200 nm due to their improved resolution at the nanoscale accompanied by strong emission intensities and stronger non-covalent interactions between nanoparticles based on their van der Waals interactions predicted from Hamaker constants. 43 In addition, exo-cellular polysaccharides produced from cyanobacteria 44 generated, in the absence of nano-labellers, strong inter-cyanobacteria interactions. While in the presence of the modied silica nanoparticles additional strong hydrogen bridges could be involved in their interactions and targeted depositions over the biostructures. So, based on strong van der Waals, polar and noncovalent interactions, and hydrogen bridges between cyanobacteria and silica nano-labellers, their interactions were evaluated by different microscopy methodologies. In this manner smaller uorescent nano-labeller sizes showed better bio-and nano-surface ratio. In this manner could be deposited a higher number of nanoparticles per biostructure. However, it should be mentioned that larger sizes interacted as well with the cyanobacterial biostructures.
Thus was observed by laser uorescence microscopy the generation of bioimaging from small cyanobacterial aggregates formed from tetramers to higher nano-bio-aggregates. Cyanobacterial labelling with NPs-(SiO 2 -Fl) showed enhanced emission from labelled bio-surfaces with optimal laser excitation at 488.0 nm. LUT edition image with dual red-green color permitted obtaining the variations of the emission intensities from the nano-biostructures (Fig. 8a). The stronger green intensities corresponded to homogeneous NPs-(SiO 2 -Fl) nanolabeller depositions; while the multi-coloured silica nanoparticles produced quenched emissions as well as in the presence of labelled cyanobacteria with NPs-(SiO 2 -RhB) (Fig. 8b) at both laser excitations applied. By optimal excitation of RhB at 543.0 and 555.0 nm, homogeneous low emission was recorded from the labelled biosurfaces (Fig. 8b) as was observed for NPs-(SiO 2 -RhB) biolabelling. But, from non-labelled cyanobacteria drastically diminished bioimaging was recorded due to their intrinsic low uorescence emissions. 45 The deposition of silica nanolabellers was corroborated by TEM (inset image (i) of Fig. 8) by increased contrast from labelled bacteria in comparison to non-labelled cyanobacteria (inset image (ii) of Fig. 8).
Moreover, DLS measurements showed higher distribution of labelled bacterial sizes than non-labelled cyanobacteria. Smaller average sizes within 5000.0-7500.0 nm range and larger aggregates as well were recorded by DLS (Fig. 9). The sizes were veried by TEM that corresponded to labelled single, dimeric and trimeric cyanobacteria (inset (i) of Fig. 9), and larger aggregates (inset (ii) of Fig. 9). The labelled biostructure sizes were larger than the non-labelled ones (inset (iii) of Fig. 9) measured by DLS. The labelled cyanobacteria showed higher contrasted images than non-labelled biostructures (inset (iv) of Fig. 9). This fact was explained by the presence of coloured silica nanoparticles with higher electron density from the highly conjugated chromophores incorporated within the silica nanoparticles.
In addition, it should be added that the sizes measured by DLS of the nano-biostructure aggregates with NPs-(SiO 2 -Fl) corresponded to those observed by laser uorescence microscopy (Fig. 10a). From these hot-spots only were generated stronger emissions. In optimized conditions, non-aggregated free nano-labellers were observed. While, in the absence of the cyanobacteria were observed single free NPs-(SiO 2 -Fl) nanolabellers, and from dimeric to tetrameric species as well (Fig. 10b), but not observed in any case were similar shapes, aggregates and sizes as was observed for the nano-biostructures (Fig. 10b).
The free nanolabellers and small nanoaggregates observed were conrmed in colloidal dispersion in the absence of cyanobacteria by DLS measurements. Dimeric and trimeric species of NPs-(SiO 2 -Fl) nanoparticles were determined (Fig. 11a), as well as single nanoparticles (Fig. 11b). These colloidal This journal is © The Royal Society of Chemistry 2020 RSC Adv., 2020, 10, 20620-20637 | 20629 dispersions were stable within 10 minutes. Aer this period of time, they showed diminished intensities from decantation in colloidal dispersion. Therefore, this effect diminished the detection of the dispersed nanoparticles in the colloidal dispersion. However, by simple manual shaking they were redispersed. So, the colloidal dispersions were well stable and dispersible.
In this manner was conrmed the use of the uorescent silica nanoparticles as nano-labellers by non-covalent interactions for cyanobacterial bioimaging.
The strong non-covalent interactions were explained by hydrogen bonding that showed higher strengths than other non-covalent interactions 46,47 with a high dependence of the electronic donor-acceptor lengths from the functional groups or atoms involved. 48 This could explain some observations made by us with different real water samples with variable composition and possible interferences. However further experiments should be done in order to study the effect of possible interferences from real matrixes, as well as the possible incorporation of the nanolabellers within the biostructure.
From the literature the uorescence technique showed high sensitivity for biolabelling, with an example application being the use of mannose-uorescent functionalized polymer for Escherichia coli labelling and detection. 49 Moreover, modied silica nanoparticles being aminated and thiosulfonated were used for non-covalent depositions on Escherichia coli 42 as   11 The distribution of smaller sizes in a colloidal dispersion by DLS of (a) dimeric nanoarchitectures of fluorescent silica NPs; and (b) individual fluorescent silica NPs. Size distribution measurements were carried out within 10 minutes. The pink, orange, and red lines correspond to 1, 5, and 9 minutes, respectively. previously discussed. Moreover, it should be highlighted the recent development by us of ultraluminescent gold core-shell silica nanoparticles based on MEF for individual Escherichia coli detection. 13,50 In addition, cyanobacterial non-specic labelling with quantum dots 51 on lamentous structures of cyanobacteria was reported as well. In this way, it should be mentioned that cyanobacteria could generate biolm formation by their exocellular polysaccharide production and higher aggregates as well. 52 Thus, in the context of antifouling nanoparticle applications, non-covalent deposition of silanized magnetic nanoparticles has been reported 53 and PEG-silver nanoparticles 54 as well as other types of nanomaterials. 55 So, to the best of our knowledge, the utilisation of silica nanoparticles for non-covalent luminescent biolabelling applications of cyanobacteria has not been reported yet; however non-covalent depositions of different nanoparticles and nanomaterials were already reported.

Static and time-resolved uorescence characterization by mono-and multi-coloured nano-silica cyanobacteria labelling
Based on the biolabelling methodology described and due to variable emission properties observed for the different monocoloured and multi-coloured uorescent silica nanoparticles, their static and time-resolved emissions were evaluated quantitatively.
First, it should be mentioned that variable absorption and low-intensity emission were reported for different types of cyanobacteria. 56 In particular the wild population of Microcystis aeruginosa studied showed a higher absorption band in the UV region around 350.0 nm. In concentrated conditions static uorescence measurements showed 4 times higher emission intensities than cyanobacteria labelled with NPs-SiO 2 -Fl nanolabellers and even higher than with NPs-(SiO 2 )-RhB and multicoloured NPs-(SiO 2 -RhB-Fl) nanoparticles (Fig. 12a). This fact was clearly explained by the optimal excitation of uorescent photosystem and negligible absorption from the uorescent reporters incorporated within the different silica nanoparticles. This fact showed the important role of the uorophores and control of the light recorded from the non-labelled biostructure and labelled biostructure by the excitation wavelength applied. But, by the optimal excitation of Fl as a uorescent energy donor reporter at 480.0 nm, up to ten times higher emission intensities were recorded with the application of NPs-(SiO 2 -Fl) nanoparticles than non-labelled cyanobacteria. However, the attendant enhanced uorescence emission from multi-coloured NPs-(SiO 2 -RhB-Fl) nano-labellers observed from free nanoparticles was not observed. Instead were recorded diminished emissions with only 25% of the increase in comparison to nonlabelled cyanobacteria. While with NPs-(SiO 2 -RhB) nanolabellers was observed 10% reduction in comparison to nonlabelled cyanobacteria (Fig. 12b). The main fact to explain this was the diminution of the nano-biostructures with NPs-SiO 2 -(RhB-Fl) nano-labellers in comparison to NPs-(SiO 2 -Fl) in the absence of the uorescent energy acceptor RhB.  In order to complete this study for a well understanding of the emission pathways, uorescence lifetime decays were measured of the free nano-labellers and labelled cyanobacteria.
For multi-coloured and mono-coloured nanoparticles were recorded bi-exponential uorescent lifetime decays (s 1 and s 2 ). For mono-coloured silica nanoparticles were recorded a shorter component related to scattering (s 1 ) and longer decay (s 2 ) related to the conned uorophores within the silica nanoarchitecture (Table 1). These values correlated with reported literature for free Fl 35 and RhB 36 uorophores slightly modied by scattering. While for multi-coloured silica nanoparticles with optimal excitation of Fl as uorescent energy donor reporter, a s 2 shortening of 15% was obtained. This uorescence lifetime decay shortening was accompanied by 25% and 35-40% emission intensity increases recorded by static uorescence and laser uorescence microscopy respectively. This enhancement in the presence of both emitters accompanied by diminished uorescence lifetime decays supported the improved emission pathway based on FRET already observed and discussed in connection with laser uorescence microscopy.
Then from labelled cyanobacteria were recorded multiexponential decays related to the emission from the different components of the nano-biostructures formed by various chromophores and emitters from the nano-labellers and photosynthetic systems of cyanobacteria. For non-labelled cyanobacteria was recorded tetra-exponential decay tting (Fig. 13a); s 1 and s 2 of 1.53 and 6.120 ns correlated with shorter and longer uorescent lifetime decays, while s 3 and s 4 of 30.62 and 61.24 ns were related to modied emission from more aggregated bacteria (Table 2). As was already reported for these types of cyanobacteria, variable uorescent lifetime decays were collected depending on their chromophore compositions. 57 However faster decays related to 0.75 to 3.0 ns interval values were reported for different molecular composition of photosynthetic systems, as well as, depending on their state of aggregation, longer components being recorded. 58 So, the tetraexponential tting showed clearly reduced uorescence lifetime decays (Fig. 13b) in comparison to non-labelled cyanobacteria (Fig. 13a). Thus, for cyanobacteria labelled with NPs-(SiO 2 -Fl) nanoparticles (SiO 2 -Fl-cysbact) were recorded diminished shorter and longer uorescent lifetime decays. Values of s 1 , s 2 , and s 3 of 0.02, 0.120, and 0.99 ns respectively were determined, and s 4 of 6.10 ns ( Table 2). Noted are the reduced values of s 2 and s 3 for NPs-(SiO 2 -Fl)-cysbact in comparison to s 1 and s 2 for non-labelled cyanobacteria. In addition it should be mentioned that these reductions of lifetime decays were accompanied by 35-40% higher emission intensities from nano-biosurfaces than from free NPs-(SiO 2 -Fl) nano-labellers, with up to 10 times higher emissions than non-labelled cyanobacteria. However, from labelled multi-coloured cyanobacteria with NPs-(SiO 2 -RhB-Fl) nano-labellers, drastic reduction of emission intensities was recorded accompanied by uorescence lifetime decay shortening. The values of s 1 , s 2 , and s 3 were even lower than with NPs-(SiO 2 -Fl) nano-labelling (Table 2); however the transferred energy was conducted within a radiative pathway with a lower quantum yield that generated the reduction in intensity.
This fact was correlated with uorescence emission bands recorded from time-gated uorescence in intact blue-green and red algae from B-phycoerythrin as intermediate chromophore in their complex photosynthetic systems that showed two emission bands centered at 530.0 and 645.0 nm selectively excited at 540.0 nm. 59 From this it was observed that up to 4 times higher emissions could be obtained from the emission band centered at 530.0 nm than at 645.0 nm. The mentioned differences showed the relative quantum yields between the chromophores involved in the emissions from non-labelled cyanobacteria that in the presence of the mono-coloured and multi-coloured nano-labellers coupled different bacterial chromophores via a bio-FRET pathway. In the presence of monocoloured NPs-(SiO 2 -Fl) nano-labellers clear enhancements were found explained by the coupling based on the welloverlapping shorter emission band with the cyanobacterial chromophore with higher quantum yield centered at 530.0 nm. While in the presence of the mono-coloured NPs-(SiO 2 -RhB) and multi-coloured NPs-(SiO 2 -RhB-Fl) nano-labellers, the diminished emission was due to the coupling with the lower quantum yield chromophore with the longer wavelength band centered at 645.0 nm. In this manner in the presence of labelled cyanobacteria with RhB emitter was recorded a diminished emission pathway conducted via bio-FRET, which in the presence of RhB-Fl pair was slightly enhanced but it was not proportional to the enhancements recorded for free NPs-(SiO 2 -RhB-Fl) nano-labeller.
To the best of our knowledge, there has been no previous report of an enhanced bio-structure like this one by controlling targeted emissions through different quantum yielding natural photo-receptors. However, this is a high-impact research eld within biophotonics such as already recently reported for living lasers 60 based on the incorporation of green uorescent proteins in cells. Thus the importance was shown of the design of nano-engineered materials as optical gain media 61 for bioapplications. In a similar manner, it could be mentioned the synthesis and fabrication of new materials for the design of miniaturized devices and instrumentation for targeted light delivery 62 for sensing and bioimaging 63 within conned biostructures and tissues.
In summary, the reported luminescent nanoplatforms showed variable and controlled emissions depending on the uorescent emitter reporters incorporated within the silica nanoparticles. In the presence of the RhB-Fl FRET pairs, enhanced emissions were observed in comparison to monocoloured nanoparticles. However these enhancements were coupled via a bio-FRET pathway with 4-5 times lower cyanobacteria chromophore quantum yield that produced clear diminished emission from multi-coloured and mono-coloured silica nanoparticles in the presence of RhB as uorescence energy acceptor and emitter (Scheme 2(i) and (ii)). And with mono-coloured NPs-(SiO 2 -Fl) nano-labelling was coupled the higher bacterial chromophore quantum yield that produced enhanced bio-FRET emissions (Scheme 2(iii)).

Cyanobacteria detection by in-ow cytometry and laser uorescence microscopy
Then was evaluated the application of the developed nanobiolabelling methodology by in-ow cytometry with laser excitation and uorescence detection. In order to do that, variable distributions were recorded of SSC and FSC (SSC: Side Scattered Light; and FSC: Foward Scattered Light) from the different labelled cyanobacteria and non-labelled biostructures (Fig. 14).
As is known, it should be mentioned that the SSC parameter is a measurement of mostly refracted and reected light that occurs at any interface within a cell where there is a change in refractive index. 64 The SSC is collected at approximately 90 degrees to the laser beam by a collection lens and then redirected by a beam splitter to the appropriate detector. In this manner, the SSC parameter is proportional to cell granularity or internal complexity and it registers cleaner uorescent event detections from samples with diminished background signalling. Moreover, the FSC is a measurement of mostly diffracted light and it is detected just off the axis of the incident laser beam in the forward direction by a photodiode. FSC provides a suitable method of detecting particles greater than a given size independent of their uorescence and is therefore oen used in immune-phenotyping to trigger signal processing. 65 In addition variable uorescent event detection counts were collected from the different samples with standard uorescence parameters of Alexa-Fluor-488-A (laser excitation at 488.0 nm with emission lter placed at 530/30 nm) and AF-555-A (laser excitation at 555.0 nm with emission lter at 585/42 nm) (Fig. 14).
For cyanobacteria labelled with NPs-(SiO 2 -RhB-Fl) and NPs-(SiO 2 -Fl) nanoparticles, different distributions of SSC values were recorded. For cyanobacteria labelled with NPs-(SiO 2 -RhB@Fl), SSC values were recorded up to Â10 5 (Fig. 14a); while for those labelled with NPs-(SiO 2 -Fl), values of 1 Â 10 5 and higher were collected (Fig. 14b). In this manner was recorded a higher number of uorescent event detection counts of cyanobacteria labelled with NPs-(SiO 2 -Fl) (Fig. 14b(i)) than with NPs-(SiO 2 -RhB-Fl) (Fig. 14a(i)). This fact was explained in terms of the enhanced bio-FRET pathway by the emission of the This journal is © The Royal Society of Chemistry 2020 NPs-(SiO 2 -Fl) towards the higher quantum yielding photoreceptor of cyanobacteria previously discussed.
In addition, from the analysis of FSC was observed a higher number of detection values within a larger interval of FSC values for cyanobacteria labelled with NPs-(SiO 2 -RhB-Fl) (Fig. 14a), and lower FSC values within a shorter range of values for those labelled with NPs-(SiO 2 -Fl) (Fig. 14b). This fact was supported by the larger sizes of the deposited NPs-(SiO 2 -RhB-Fl) than NPs-(SiO 2 -Fl) nano-labellers.
For non-labelled cyanobacteria, the uorescent event detection distributions as well as the detection event counting were different in comparison to the previously discussed labelled biostructures. Smaller detection surfaces were obtained from SSC vs. FSC plots (Fig. 14c), accompanied by a lower number of detection counts (Fig. 14c(i)) than for cyanobacteria labelled with NPs-(SiO 2 -RhB-Fl) (Fig. 14a, a(i)) and NPs-(SiO 2 -Fl) (Fig. 14b, b(i)). Moreover, it should be highlighted that the free mono-coloured NPs-(SiO 2 -Fl) and multi-coloured NPs-(SiO 2 -RhB-Fl) nano-labellers were detected in reduced sized areas within the distribution of non-labelled cyanobacteria (highlighted yellow oval in Fig. 14c).
In this manner it was possible to apply the nano-biolabellers for cyanobacteria detection and counting with a versatile in-ow technique such as cytometry. However, further experiments should be done to validate this methodology.
In addition in order to evaluate applications by detection of these biostructures at low concentrations based on uorescent bioimaging, colloidal dispersions were prepared with lower cyanobacteria concentrations. In this manner, by application of optimal NPs-(SiO 2 -Fl) nano-labellers, smaller bacterial aggregates were recorded ( Fig. 15a and b) formed by dimers to tetramers of cyanobacteria (Fig. 15c). From these nanobiosurfaces, strong luminescent intensities were generated that permitted faster detection and counting with 3D uorescence plots (Fig. 15d).
Finally it should be highlighted that using the developed methodology, individual ultraluminescent nano-biostructures were collected by in-ow cytometry as well as their detections by laser uorescence microscopy. And at this point it should be mentioned that the non-classical light collected was from conned intermolecular interactions within silica nanoplatforms that permitted targeted light delivery via bio-FRET on biostructures that generated varied controlled emissions depending on the emitters incorporated within the nanostructures. In this way, these results showed potential applications of this methodology within conned microuidics chips 66 for biodetection and biophotonics applications. 67

Conclusions
Variable silica nano-emitters were developed, formed via different uorescent emitters with different quantum yields in order to tune their emission intensities based on FRET. From mono-coloured silica nanoparticles with the incorporation of Fl or RhB uorescent dyes, strongly emitting 220.0 nm luminescent nanoparticles were recorded, with proportional quantum yields; while from multi-coloured silica nanoparticles, enhanced emissions were collected from 240.0 nm nanoparticles based on FRET. Increases were seen until 40% for multi-coloured silica nanoparticles.
These nanoparticles were evaluated for cyanobacteria labelling by non-covalent interactions and detected by laser uorescence microscopy and in-ow cytometry. Thus, by TEM and laser uorescence microscopy, the interactions were observed between the polar silica surfaces and polysaccharides naturally produced by the cyanobacteria that permitted targeted nanoparticle depositions.
By laser uorescence microscopy, variable intensities from the luminescent nano-biostructures were recorded depending on the nano-labeller applied. Highly luminescent multicoloured NPs-(SiO 2 -RhB-Fl) generated diminished emissions in comparison to NPs-(SiO 2 -Fl). This fact was explained by a bio-FRET coupling pathway with low quantum yields of natural chromophores from the biostructure. While, the application of NPs-(SiO 2 -Fl) nano-labellers produced higher emission by bio-FRET coupling with natural chromophores with higher quantum yields (known as photosystem I and II respectively).
Optimal biolabelling conditions were applied for in-ow cytometry experiments for cyanobacteria detection and counting. From cyanobacteria labelled with NPs-(SiO 2 -Fl) greater uorescent event counts were recorded than with multicoloured NPs-(SiO 2 -RhB-Fl) and NPs-(SiO 2 -RhB) nanolabellers. These results correlated with the results obtained via static uorescence and laser uorescence microscopy.
In this manner light delivery was achieved and controlled from conned uorophores within silica nanoplatforms for biolabelling applications with optically active biomaterials. Finally, it should be mentioned that the enhanced emissions were recorded by a targeted bio-FRET pathway considering the intrinsic and natural chromophore compositions of the biostructures assayed. Thus, future applications to other types of biostructures were opened up based on light delivery from conned emitters.

Conflicts of interest
There are no conicts to declare.