New Hybrid Nanocomposite Films for Optical Diagnostics and Optical Temperature Sensing: Synergy Among α‐Synuclein, Gold and Upconverting Nanoparticles

TL;DR

Researchers created a hybrid nanocomposite film using alpha-synuclein, gold, and upconverting nanoparticles for optical diagnostics and temperature sensing.

Contribution

The study introduces a novel hybrid nanocomposite film with potential for optical diagnostics and nanothermometry.

Findings

The film exhibits strong upconversion emission under 980 nm laser excitation.

The nanocomposite film functions reliably as a primary optical nanothermometer in various media.

Image-based optical thermometry with micrometer-scale resolution was successfully demonstrated.

Abstract

The propensity of the unstructured protein α‐synuclein to undergo a conformational transition to fibrillar aggregates was harnessed to prepare an upconverting organic–inorganic hybrid film composed of Au nanoparticles (NPs) and Yb3+, Er3+‐activated CaF2 NPs, with α‐synuclein serving as a structural linker. The 2D nanomaterial appears as a quasi‐monolayer film, composed of distributed AuNPs and tightly packed CaF2:Yb,Er NPs. The film shows excellent upconversion emission arising from the Er3+ ions following excitation of the Yb3+ ions under 980 nm laser radiation. Optical thermometry investigations were conducted in various media (air, H2O, and D2O) to assess the performance of the film as an optical nanothermometer. The thermometric calibration curves obtained within the physiological temperature range (25–60°C) support its potential application in 2D optical nanothermometry. By evaluating the energy gap between the two thermalized levels and calibrating the upconversion at one temperature, the nanocomposite film was found to perform reliably as a primary thermometer. Furthermore, image‐based optical thermometry has demonstrated that upconversion emission can be successfully employed for local temperature estimation, thereby enabling image‐based 2D optical thermometry with micrometer‐scale spatial resolution.

Gold and upconverting nanoparticles coated with alpha‐synuclein were exploited to produce a quasi‐monolayer hybrid film. The film was then characterized, and its behaviour as a primary optical nano‐thermometer was studied. The film was also used in a proof‐of‐concept study of image‐based optical nanothermometry exploiting a home‐built up‐conversion fluorescence microscope and different filter combinations.

Figures

Schematic representation of the aim of the research.

(A) Hydrodynamic diameter of AuNPs in water (black) and α‐Syn‐AuNPs dispersions (red). (B) Hydrodynamic diameter of UCNPs (black) and α‐Syn‐UCNPs (red) dispersions. (C),(D) NMR analysis of the mode of protein adsorption to NPs. 1H,15N‐HSQC spectra of 50 µM [15N]α‐Syn in the presence of CaF2 NPs at the reported molar ratios. Chemical shift perturbations, CSP, are displayed in panel C. Intensity perturbations are displayed in panel D and were determined as the ratio between the intensity I of a given sample in the presence of NPs and the intensity I0 of a sample containing only protein. Data are reported as a function of the primary protein sequence. The domain organization of α‐Syn is schematically depicted on top of the figures. E) Visualization of hybrid layer formation. Test tubes contained aqueous solutions of UCNPs, α‐Syn, UCNPs@α‐Syn, and UCNPs@α‐Syn in the presence of chloroform. The solid layer formed at the solvent interface is highlighted with a red circle. F) Thioflavin‐T fluorescence assay. Samples contained UCNPs@α‐Syn (grey bar) and UCNPs@α‐Syn+chloroform (green bar). Measurements were carried out on three replicates; data are reported as mean ± s.d.

(a) ABF‐STEM images of the α‐Syn‐Au, CaF2NPs film. (b) HAADF‐STEM of α‐Syn‐Au, CaF2NPs film (upper left) and elemental mapping. (c) Absorption spectra of AuNPs (black line) and AuNPs@α‐Syn (red line) water dispersions and freestanding α‐Syn‐Au, CaF2NPs film (blue). Inset: picture of the α‐Syn‐Au, CaF2NPs film. (d) Upconversion image of the α‐Syn‐Au, UCNPs film upon laser excitation (λ = 980 nm), (1): bright‐field picture of the film; (2) UC from the portion of nanofilm irradiated by the laser (dotted red line). (e) UC spectrum from the irradiated portion of film (dotted red line) (λ exc = 980 nm, laser irradiance = 200 kW/cm2). Bands assignment: (1) 2 H 9/2 → 4 I 15/2; (2) 2 H 11/2 → 4 I 15/2; (3) 4 S 3/2 → 4 I 15/2; (4) 2 H 9/2 → 4 I 13/2; (5) 4 F 9/2 → 4 I 15/2. (f) Energy level scheme for Yb3+ and Er3+ ions and UC processes. (g) (3, 4) Bright field picture of the film; (5, 6) UC image from the portion of film delimited with a dotted blue line in (4); (7, 8) merge of the bright field and UC images: 7:3+5; 8:4+6. (h) UC spectra of the αSyn‐CaF2UCNPs film at different laser irradiance values. (i) Emission intensities of the blue (400‐410 nm), green (500–570 nm) and red (630–700 nm) UC bands of α‐Syn‐CaF2UCNPs film vs laser irradiance.

UC spectra (λ exc = 980 nm, laser I = 20 kW/cm2) for the α‐Syn‐Au, CaF2NPs film immersed in D2O (A), H2O (E) and air (I) for different temperatures: 2 H 11/2 → 4 I 15/2 (left) and 4 S 3/2 → 4 I 15/2 bands (right); LIR(T) vs measured T (black dots) and linear fit (red line) for the film in D2O (B), H2O (F) and air (J); Relative thermal sensitivities S rel vs measured T for the film in D2O (C), H2O (G) and air (K). (D) Calculated T, from Equation (7), vs measured T for the film in D2O (D), H2O (H) and air (L). The red dotted line is a guide for the eye.

(a) Scheme of the microscope setup for the thermometry‐imaging experiment; (b) UC spectra of the film collected with (black line) and without (green line) 532 nm long‐pass filter; (c) UC image for the nanocomposite film at different T collected without (TOT) and with (LP532) the 532 nm long‐pass filter and image of the 525 nm centered emission calculated subtracting the signal of LP532 image to the TOT image with intensity increased by 10 times to observe the signal; (d) LIR(T) image (see Equation 11) vs calculated T; (e) Detail of the 2 H 11/2 (left) → 4 I 15/2 and 4 S 3/2 → 4 I 15/2 (right) UC bands at the calculated T.

Full text

Introduction

1

Bottom‐up assembly strategies offer robust routes to create organized nanomaterials that exhibit tailored and emergent properties [1]. In particular, molecular self‐assembly can direct the fabrication of diverse multiscale structures, including fibers, sheets, and gels [2]. Biomolecules represent attractive building blocks for the self‐assembly of functional materials, including organic–inorganic nanocomposites, capable of establishing multiple noncovalent interactions such as electrostatic forces, hydrogen bonds, and hydrophobic interactions [3]. Among biomolecules, proteins and peptides are prominent due to their varied amino acid compositions that confer unique chemical properties. The combination of protein building blocks with inorganic nanomaterials further expands the chemical space and functional possibilities of composite systems [4, 5, 6].

Alpha–synuclein (α‐Syn) is an intrinsically disordered polypeptide composed of 140 amino acids, organized in three domains: an amphipathic N‐terminal region (amino acids 1–60) with a propensity for α‐helical structure formation; a central non‐amyloid component (NAC) domain (a.a. 61‐95) with hydrophobic character; an acidic C‐terminal region (a.a. 96–140). The inherent stability, water solubility, and ease of preparation by recombinant expression have made α‐Syn arguably the most popular model biomolecule for research on unstructured polypeptides. The protein solution state is best described as an ensemble of dynamically interconverting conformers [7, 8]. The amphiphilic character gives the protein the ability to associate with lipid membranes and other surfaces, in some cases facilitated by the acquisition of secondary α‐helical structure in the N‐terminal region [9, 10, 11]. The interaction of α‐Syn with nanomaterials has recently been documented [12, 13, 14, 15]. α‐Syn was found to preferentially bind to charged nano‐objects, with the binding strength and orientation influenced by the surface charge. These interactions often lead to a disordered and dynamic protein corona that can vary in thickness and organization depending on factors such as the protein‐to‐NP molar ratio and NP surface chemistry.

Despite its high solubility in water, α‐Syn is known to undergo structural transitions from a monomeric disordered state to insoluble higher‐order structures enriched in β‐sheets [16], also observed in the brains of patients affected by Parkinson's disease. In addition to its pathological relevance, fibril formation by amyloidogenic peptides and proteins is of considerable interest from a molecular point of view, and for its potential exploitation in material science [17, 18]. Characterized by a cross‐β structure, where β‐sheets stack perpendicular to the fiber axis, amyloid fibrils exhibit exceptional mechanical strength and affinity for various surfaces [19]. This behavior offers a promising platform for creating functional 2D materials by incorporating inorganic NPs and tailoring functionalities into a composite film [6, 20, 21].

Trivalent lanthanide (Ln^3+^) ion‐activated nanomaterials have attracted significant attention across various technological fields, including security, sensing, photovoltaics, and biomedicine, due to their unique magnetic and optical properties [22, 23, 24]. Among these, upconverting nanoparticles (UCNPs) are particularly intriguing for their upconversion (UC) properties, i.e., their ability to absorb photons in the near‐infrared region (NIR) and emit radiation at higher energies, in particular in the visible or ultraviolet. The luminescence of Ln^3+^ ions, including the UC emission, is sensitive to temperature fluctuations, and it can be exploited in optical nanothermometry [25]. Indeed, since the Ln^3+^ energy levels are differently populated according to Boltzmann's distribution law, variation of the temperature induces changes in the level populations and, in turn, in the emission intensities from these populated levels [26]. Emissions in the green region from Er^3+^ ions, arising from different populations of excited energy levels, form the basis of effective thermometric functions. The ^2^ H 11/2 and ^2^ S 3/2 excited levels of the Er^3+^ ions, separated by an energy difference around 700 cm^−1^ [27], are exceptionally well‐suited for optical thermometry, whereby temperature can be determined from the ratio of the emission intensities of these two excited states (ratiometric method) [26, 28, 29]. Yb^3+^ and Er^3+^ co‐doped UCNPs are of particular interest due to their strong UC emission upon NIR excitation, resulting from efficient sensitization of the Er^3+^ ions by Yb^3+^ ions, upon excitation at around 980 nm [30].

Among inorganic nanomaterials, CaF_2_ NPs have emerged in the last years as excellent hosts for luminescent Ln^3+^ ions [31], due to several advantages: i) the similar ionic radii of Ca^2+^ and Ln^3+^ ions (e.g. for an 8‐coordination site r(Ca^2+^) = 1.12 Å, r(Yb^3+^) = 0.985 Å, r(Er^3+^) = 1.004 Å[32]) facilitate an efficient incorporation of Ln^3+^ ions into the CaF_2_ host; ii) the low phonon energy of the fluoride matrix minimizes multiphonon non‐radiative energy transfers, enhancing the luminescence intensity of the embedded Ln^3+^ ions; iii) CaF_2_ NPs can be synthesized through facile and environmentally friendly methods; iv) good chemical stability and biocompatibility make CaF_2_ NPs dispersible in biological fluids and suitable for their use in various biomedical applications.

To develop novel 2D nanomaterials with multiple functionalities, we prepared and investigated upconverting organic–inorganic hybrid films using the α‐Syn protein as an organic scaffold embedding UCNPs and AuNPs.

We point out that Lee and co‐workers [21] developed free‐standing gold nanoparticle films exploiting the β‐sheets interactions of the α‐Synuclein protein, and their plasmonic behavior has been evidenced by absorption spectroscopy. The plasmonic behavior of gold nanoparticles makes them really interesting in various technological fields, opening these films to multiple application possibilities ranging from photoelectronics, to energy, catalysis and sensing [33, 34].

Inspired by this latter work, we leveraged α‐Syn's self‐assembly properties to form a new nanocomposite film by incorporating Yb^3+^, Er^3+^ codoped CaF_2_ UCNPs for their interesting luminescent properties and AuNPs for the structural strength and color useful for film observation and handling. The excellent luminescence in the optical region of this nanocomposite film was thoroughly investigated, with a focus on its thermometric properties for the development of advanced nanomaterials as optical diagnostic agents and nanothermometers. Exploiting their upconversion, we also investigated image optical thermometry, using a straightforward home‐built experimental setup that incorporated optical filters and a CMOS camera. In fact, this kind of image thermometry is still in its infancy [35], but it has enormous potentialities and implications for temperature sensing in biological systems with high spatial resolution. To the best of our knowledge, the development of such a hybrid plasmonic‐upconverting nanofilm has not been reported in the literature [36].

An illustrative representation of the research aim is sketched in Figure 1.

Schematic representation of the aim of the research.

Results and Discussion

2

Building Units for the Nanofilm: AuNPs and CaF2:Yb,Er UCNPs

2.1

AuNPs

2.1.1

The STEM (Annular Bright Field, ABF, and High Angle Annular Dark Field, HAADF) images of the AuNPs, Figure S1A, display NPs of round‐shaped morphology with an average particle size of 6.7 ± 1.8 nm. HR‐STEM images, Figure S1B–D, show several crystal planes, which confirm the crystalline nature of the AuNPs. Based on DLS measurements, the hydrodynamic diameter of the AuNPs was determined to be 8 ± 2 nm (Figure S2). The measured ζ‐potential of –51 ± 4 mV indicates excellent colloidal stability of the AuNPs in water dispersion (see inset Figure S3). The absorption spectrum of a water dispersion of citrate‐capped AuNPs is characterized by a broad absorption band that peaks around 515 nm due to the surface plasmon resonance (Figure S3).

CaF2:Yb,Er UCNPs: Structure, Morphology and Elemental Analysis

2.1.2

The XRPD pattern of the CaF_2_:Yb,Er UCNPs is shown in Figure S4A, confirming the formation of single‐phase cubic CaF_2_ (Fm‐3m space group). The increase in intensity of the (200) reflection with respect to the reference pattern confirms the successful incorporation of the dopants in the crystal lattice [37, 38]. STEM images of the UCNPs show quite monodispersed NPs mainly round‐shaped (Figure S4B–D), with a particle size of 8.2 ± 1.7 nm (Figure S4B, inset). Elemental mapping using the EDX technique (Figure S4D clearly indicates a homogeneous distribution of the most abundant elements (Ca, Yb and F).

ICP‐MS analysis indicates a molar ratio between Ca:Yb:Er similar to the nominal one, with a slightly larger amount of Yb and Er than the nominal one (see Table S3).

CaF2:Yb,Er UCNPs: Colloidal Properties

2.1.3

The hydrodynamic diameter of the CaF_2_:Yb,Er UCNPs is 12 ± 3 nm (Figure S5), slightly larger than that obtained from the STEM images, indicating that the capping citrate ions are present on the NP surface. The measured ζ‐potential, −31 ± 5 mV indicates excellent colloidal stability of the UCNPs (insets in Figure S10) and demonstrates the presence of carboxylate moieties due to the citrate groups on the NP surface.

CaF2:Yb,Er UCNPs: NIR Absorption

2.1.4

The absorption spectrum in the NIR range (860‐1000 nm) of a water dispersion of UCNPs shows typical features due to the Yb^3+^ ions, attributed to the ^2^ F 7/2 → ^2^ F 5/2 transition of Yb^3+^ ions (Figure S6), in particular a relatively sharp absorption band around 976 nm, in analogy with CaF_2_ NPs [39, 40] and for Yb^3+^ doped single crystals [41]. A weak band around 966 nm partially overlapped with the strongest one is observed, attributed to absorption from Yb^3+^ ions in crystal sites with O _ h _ symmetry [41]. A broad band in the 920–950 nm range is present, due to several transitions from the Stark components of the ^2^ F 7/2 ground state and of the ^2^ F 5/2 excited level of Yb^3+^ ions. Clustering of Yb^3+^ ions, due to the high concentration in the host, can also contribute to absorption in the NIR region [42].

CaF2:Yb,Er UCNPs: Energy Gap Between 4

S 3/2 and 2 H 11/2 Levels

2.1.5

From the excitation spectrum for the UCNPs in the 470–550 nm region (Figure S7) the energy barycenters ν∼bar of the bands have been calculated according to the formula:

$$[eqn]$$

where I represents the excitation intensity, ν∼ represents the wavenumbers (in cm^−1^), and the integrals extend over the corresponding excitation band. Using formula (1), the ν∼bar of the bands corresponding to the ^4^ I 15/2 → ^4^ S 3/2 and ^4^ I 15/2 → ^2^ H 11/2 transitions have been calculated as (18524 ± 15) cm^−1^ and (19242 ± 15) cm^−1^, respectively. Since both transitions start from the ^4^ I 15/2 ground state, the energy barycenters of the two ^4^ S 3/2 and ^2^ H 11/2 levels coincide with the two calculated ν∼bars. The average energy gap ΔE between the two ^4^ S 3/2 and ^2^ H 11/2 thermalized levels is calculated as the difference between the barycenters:

$$[eqn]$$

A value of 718 ± 30 cm^−1^ was obtained for ΔE. The corresponding energy gap between the two thermalized levels from the absorption bands of a colloidal dispersion (Figure S8) is in full agreement with that obtained from the excitation spectrum. It is worth noting that the ΔE between the two thermalized levels is similar to that found by Sanz–Garcia et al. [43] and Chouahda et al. [44] for CaF_2_ NPs and single crystals, respectively. This average energy gap value is of paramount importance in verifying the behavior of the thermometer according to the Boltzmann law regime. As reported by Pedroni et al. [39], the incorporation of sodium ions into the CaF_2_ lattice symmetrizes the local environment of Ln^3+^ and therefore reduces the emission intensities from the lanthanide ions. For this reason, we avoided the presence of sodium ions in the preparation of NPs, and a potassium salt was employed as the precursor of the citrate capping agent.

CaF2:Yb,Er UCNPs: UC Emission Spectra

2.1.6

The UC spectrum of the CaF_2_:Yb,Er UCNPs in powder form, after excitation at 980 nm shows typical bands of Er^3+^ ions, excited through an energy transfer mechanism (ETU process) from Yb^3+^ ions acting as sensitizers (Figure S9). The shoulder around 560 nm indicates the occurrence of the ^2^ H 9/2→^4^ I 13/2 transition [45].

A comparison of the UC spectra of the CaF_2_:Yb,Er UCNPs in H_2_O and D_2_O is shown in Figure S10. Since D_2_O has a lower vibrational energy (maximum vibrational energy around 2600 cm^−1^) than H_2_O (maximum vibrational energy around 3600 cm^−1^), multiphonon relaxation processes (MRP) of the excited states of the lanthanide ions due to the solvent are more effective in H_2_O than D_2_O [46]. We observed a much higher overall UC intensity for colloidal dispersions in D_2_O than in H_2_O, maintaining the same concentrations of NPs (Figure S10), indicating that a substantial fraction of the dopant lanthanide ions is located at the surface of the NPs, in contact with the solvent. In the case of H_2_O as a solvent, the UC spectrum is substantially similar to that found for powders (Figure S9); the green emission bands around 550 nm are enhanced for the D_2_O dispersion with respect to the H_2_O one (Figure S10). This green UC increase is due to different population dynamics of the ^2^ H 11/2, ^4^ S 3/2 and ^4^ I 9/2 emitting levels [47, 48] because the less energetic O‐D vibrations decrease the MRP from both the ^4^ S 3/2 and ^4^ F 9/2 levels to the respective lower‐lying energy levels. Nonetheless, the MRP processes are less efficient for the ^4^ S 3/2 green‐emitting level, due to its greater energy gap from the corresponding lower‐lying level (^4^ F 9/2), increasing the population of the ^4^ S 3/2 level.

CaF2:Yb,Er UCNPs: UC Emission Decays

2.1.7

To obtain more information on the dynamical properties of the excited levels, emission decay curves for the green and red transitions have been measured for the CaF_2_:Yb,Er UCNPs in powder form and as colloidal dispersions in H_2_O and D_2_O. The decay curves are shown in Figure S11a (powder form), in Figure S11b (in H_2_O) and in Figure S11c (in D_2_O). The decay curves show a non‐exponential behavior, typical for nanostructured upconverting materials [49, 50, 51] The emission intensity I_ UC _ of the decay curves was fitted with a biexponential profile:

$$[eqn]$$

where A _ i _ and τ_ i _ are the weight and decay time of the i^ th ^ contribution. Intensity‐weighted average lifetimes, τ_ av _, are calculated by:

$$[eqn]$$

From biexponential fits of the decay curves (Equation (3)), we obtained the decay times, the corresponding weights, and the intensity‐averaged lifetimes from Equation (4) (Table S4). As expected, the effective lifetimes are much shorter than those measured for similar CaF_2_:Yb,Er samples, single crystals [52] or ceramic powders [53]. On the other hand, CaF_2_:Yb,Er NPs of the same composition prepared by hydrothermal synthesis [40] show similar lifetimes. Moreover, although the variation of the average lifetimes for the green emission is not very pronounced for the three samples (for the powder, H_2_O and D_2_O dispersions), the average lifetime of the red emission decreases when passing from the powder to the H_2_O and D_2_O dispersions, indicating a higher multiphonon relaxation of the Er^3+^ excited levels for the colloidal dispersions than for powders, due to solvent quenching.

Nanofilm Building Mechanism: Insight into the Interaction Between α‐Syn and NPs

2.2

Perturbation of the Colloidal Properties of the NPs

2.2.1

Before the preparation of nanohybrid films, we set out to explore the interaction between the prepared NPs and the selected biomolecule, α‐Syn. The initial characterization of the NP/α‐Syn interaction utilized the DLS technique to observe the variations of the colloidal properties of NPs after protein corona formation. As shown in Figure 2A, the hydrodynamic diameter of AuNPs in water dispersion was 8 ± 2 nm, while it increased to 15 ± 5 nm for the α‐Syn‐AuNPs, indicating the formation of a protein corona. Concurrently, the ζ‐potential of the AuNPs increased to –11.8 ± 0.5 mV, demonstrating that the formation of the protein corona reduced the net surface charge of the NPs. Similarly, the CaF_2_:Yb,Er NPs exhibited an increase of the hydrodynamic diameter to 21 ± 5 nm (Figure 2B) and a ζ‐potential increase to –3.8 ± 0.7 mV. In both cases, the hydrodynamic diameter increased by approximately 8 nm. Despite the decrease in net surface charge, no aggregation was observed in either sample.

(A) Hydrodynamic diameter of AuNPs in water (black) and α‐Syn‐AuNPs dispersions (red). (B) Hydrodynamic diameter of UCNPs (black) and α‐Syn‐UCNPs (red) dispersions. (C),(D) NMR analysis of the mode of protein adsorption to NPs. 1H,15N‐HSQC spectra of 50 µM [15N]α‐Syn in the presence of CaF2 NPs at the reported molar ratios. Chemical shift perturbations, CSP, are displayed in panel C. Intensity perturbations are displayed in panel D and were determined as the ratio between the intensity I of a given sample in the presence of NPs and the intensity I0 of a sample containing only protein. Data are reported as a function of the primary protein sequence. The domain organization of α‐Syn is schematically depicted on top of the figures. E) Visualization of hybrid layer formation. Test tubes contained aqueous solutions of UCNPs, α‐Syn, UCNPs@α‐Syn, and UCNPs@α‐Syn in the presence of chloroform. The solid layer formed at the solvent interface is highlighted with a red circle. F) Thioflavin‐T fluorescence assay. Samples contained UCNPs@α‐Syn (grey bar) and UCNPs@α‐Syn+chloroform (green bar). Measurements were carried out on three replicates; data are reported as mean ± s.d.

Protein Orientation on the AuNPs and UCNPs

2.2.2

Previous studies have demonstrated the utility of NMR spectroscopy in elucidating protein‐nanoparticle interactions [54, 55]. Here, we applied site‐resolved NMR to determine the mode of binding of α‐Syn to the UCNPs. To exploit HN‐correlation spectra, the protein was enriched in nitrogen‐15 by recombinant expression in an isotope‐enriched medium. NMR analysis of [^15^N]α‐Syn in the presence and absence of NPs revealed that NP addition did not significantly alter peak positions (Figure 2C) but perturbed peak intensities (Figure 2D). A region‐specific effect was observed, with stronger signal attenuation in the N‐terminal protein region compared to the central and C‐terminal sequences. These findings indicate that the amphiphilic N‐terminus adsorbed to the NP surface while the acidic domain remains flexible due to electrostatic repulsion with the negatively charged capping molecules. The central NAC domain was possibly partially immobilized or experienced restricted motion due to its proximity to the N‐terminal portion, which served as an anchoring point on the NP surface.

Induction of Protein Structural Conversion

2.2.3

We verified the ability of the NP/α‐Syn pair to undergo film formation in a test tube. Adding chloroform to an aqueous solution of UCNPs and α‐Syn resulted in a visible solid layer forming at the interface between the solvents (Figure 2E). To probe the chloroform‐induced conformational transition of α‐Syn, we applied the thioflavin‐T (ThT) fluorescence‐based assay. ThT is a benzothiazole dye that exhibits strong fluorescence upon interaction with cross‐β‐sheet‐rich protein fibrils. The observation of an intense fluorescence signal from the chloroform‐treated solution was a clear indication of the conformational conversion of α‐Syn into amyloid‐like structures (Figure 2F). On the basis of the observed mode of interaction, the NAC domain and other relevant amino acid stretches appeared to retain sufficient dynamics to undergo conformational changes.

Nanofilm: α‐Syn‐Au,CaF2 UCNPs

2.3

Absorption Spectra

2.3.1

The UV–vis absorption spectra of water dispersions of AuNPs covered by α‐Syn and of the α‐Syn‐Au,CaF_2_NPs nanofilm are shown in Figure 3c, clearly indicating a retained plasmonic behavior, as the freestanding AuNPs (Figure S3). Nonetheless, a small redshift of the absorption maximum from 516 to 525 nm on passing from the AuNPs to AuNPs@α‐Syn is noted (Figure 3c), suggesting a small aggregation of the protein‐covered AuNPs, resulting in a corresponding shift of the surface plasmon resonance to lower energies. A further shift to lower energies is also found for the α‐Syn‐Au,CaF_2_NPs nanofilm, for the small distance among the AuNPs and further aggregation of the AuNPs, as expected for a condensed phase. This behavior is in excellent agreement with the morphology of the nanofilm, shown in the ABF‐STEM and HAADF‐STEM images (Figure 3a,b and, for α‐Syn‐AuNPs film, Figure S12), which evidences a fairly uniform distribution of the AuNPs in the film layer as well as their nice small size dispersion.

(a) ABF‐STEM images of the α‐Syn‐Au, CaF2NPs film. (b) HAADF‐STEM of α‐Syn‐Au, CaF2NPs film (upper left) and elemental mapping. (c) Absorption spectra of AuNPs (black line) and AuNPs@α‐Syn (red line) water dispersions and freestanding α‐Syn‐Au, CaF2NPs film (blue). Inset: picture of the α‐Syn‐Au, CaF2NPs film. (d) Upconversion image of the α‐Syn‐Au, UCNPs film upon laser excitation (λ = 980 nm), (1): bright‐field picture of the film; (2) UC from the portion of nanofilm irradiated by the laser (dotted red line). (e) UC spectrum from the irradiated portion of film (dotted red line) (λ exc = 980 nm, laser irradiance = 200 kW/cm2). Bands assignment: (1) 2 H 9/2 → 4 I 15/2; (2) 2 H 11/2 → 4 I 15/2; (3) 4 S 3/2 → 4 I 15/2; (4) 2 H 9/2 → 4 I 13/2; (5) 4 F 9/2 → 4 I 15/2. (f) Energy level scheme for Yb3+ and Er3+ ions and UC processes. (g) (3, 4) Bright field picture of the film; (5, 6) UC image from the portion of film delimited with a dotted blue line in (4); (7, 8) merge of the bright field and UC images: 7:3+5; 8:4+6. (h) UC spectra of the αSyn‐CaF2UCNPs film at different laser irradiance values. (i) Emission intensities of the blue (400‐410 nm), green (500–570 nm) and red (630–700 nm) UC bands of α‐Syn‐CaF2UCNPs film vs laser irradiance.

We point out that we did not observe particular reabsorption of the Er^3+^ emission or variation of the UC emission due to the AuNPs in the nanocomposite film, also because of the much lower amount of the AuNPs with respect to the UCNPs (see below).

Morphology and Elemental Analysis

2.3.2

The STEM images of the α‐Syn‐AuNPs point out that the film is constituted by a quasi‐monolayer of AuNPs, with good homogeneity. The darker areas in Figure S12 show parts of the film with a certain degree of superposition of AuNPs layers, probably caused by local film folding during the deposition on the TEM grid or by an uneven distribution of AuNPs in the film. Nonetheless, the presence of α‐Syn surrounding the NPs is demonstrated by the spacing between the AuNPs within the film. Moreover, the good homogeneity of the AuNPs is demonstrated by the HAADF‐STEM image of the α‐Syn‐AuNPs film (Figure S12). The ABF‐STEM and HAADF‐STEM images of the α‐Syn‐Au, CaF_2_NPs film (Figure 3a,b), combined with the energy‐dispersive X‐ray (EDX) mapping (3b), reveal that the UCNPs form a densely‐packed layer, overlapped by layers of AuNPs. These distinct NPs distribution patterns, with closely packed UCNPs and AuNPs confined to certain areas while maintaining a consistent spacing, may be attributed to different interactions between the two types of NPs and α‐Syn.

From EDX analysis, we obtained the elemental percentages of the film components (see Table S7). We point out that the Er^3+^ amount was too small to be detected by the EDX technique. The Au/(Ca+Yb+F) mass ratio resulted to be 0.33 ± 0.04, much less than the nominal one (see Experimental section).

Optical Microscopy Images and UC

2.3.3

Optical microscopy images of the α‐Syn‐Au, CaF_2_NPs (Figure S13) revealed a quite homogeneous film structure at micrometer size, with few spots due to local AuNP aggregation. In fact, the absence of evident purple or blue regions confirmed the excellent dispersion of AuNPs in the film.

The UC emission properties of the synthesized α‐Syn‐Au, CaF_2_NPs film were investigated using the home‐made fluorescence microscope (Figure S16). In particular, the UC emission was measured using laser radiation at a wavelength of 980 nm as an excitation source, focused by a microscope objective. The emitted signal was detected with a CMOS camera, after passing through a shortpass cutoff filter (λ_ cutoff _ = 900 nm). A picture of the UC emission from a portion of the film is shown in Figure 3d2, evidencing that the irradiated area is about 125 × 20 µm^2^. The UC spectrum corresponding to this emission, shown in Figure 3e, is virtually identical to the UC emission of the UCNPs (see Figure S9).

The UC images for the α‐Syn‐Au, CaF_2_NPs film (Figure 3g) reflect the microscopic structure of the film, composed by a diffuse luminescence superimposed by brighter spots in which the UCNPs appear more concentrated. Nonetheless, we point out that the NIR laser intensity was not perfectly homogeneous in the irradiated area, so it could partially induce UC intensity inhomogeneities.

Power Study of the UC Emission

2.3.4

The UC emission intensity (I_ UC ) for the α‐Syn‐Au, CaF_2_NPs film was studied vs laser irradiance (I) between 2.5 and 200 kW/cm^2^ (Figure 3h) using the custom‐built fluorescence microscope (Figure S16). The slope of log(I UC ) vs log(I) for the blue emission (in the 400–410 nm range) (Figure 3i), was calculated as 2.7 ± 0.2, and therefore the blue UC emission arises from a 3‐photon process; the UC signal was not detectable with our experimental setup for laser irradiances <10 kW/cm^2^. Moreover, the log(I UC _) vs log(I) curve for the green emission (in the 500–570 nm range), arising from both the ^2^ H 11/2 and ^4^ S 3/2 levels, has a slope of 2.0 ± 0.1, indicating a 2‐photon UC process. Interestingly, for I⩾40 kW/cm^2^, an emission band due to the ^2^ H 9/2 → ^4^ I 13/2 transition is observed (Figure 3h). Although this transition is due to a 3‐photon UC process, no difference in the log‐log curve slope was detected for data at lower irradiance values. This behavior could be explained by a decrease in the slope at high irradiance values, due to the increase in the population and subsequent saturation of the different emitting levels of Er^3+^ ions [56]. Moreover, a slope of 1.9 ± 0.1 was detected for the log‐log curve slope for the red emission (in the 630–700 nm range), indicating that the ^4^ F 9/2 energy level is populated by different 2‐photon processes. Also in the latter case, no deviation from linearity was observed in the considered laser irradiance range. It is worth mentioning that although the irradiance spans a range from 2.5 to 200 kW/cm^2^, the UC behavior is far from saturation, as we did not observe a plateau at high irradiances. Moreover, we did not observe any heating of the sample due to the laser beam in our experimental conditions, as clearly evidenced from the negligible change in the relative intensities of the two thermalized bands in the 500–570 nm range in the UC spectra, as a function of the laser irradiance (Figure 3h). Furthermore, we point out that the Au NPs in our nanocomposite exhibit negligible absorption at 980 nm, as evidenced from the absorption spectrum (Figure 3c), and they can efficiently dissipate heat by conduction, which is very high in metals. Moreover, in water dispersions, e.g. biological fluids, heat dissipation of the film is favoured and as a result, we did not observe any local heating of our film in water dispersions. In fact, our investigation on optical thermometry for the nanofilm clearly demonstrates this behavior (see below).

UC Emission Decay

2.3.5

The decay curves for the nanofilm (Figure S14) show a non‐exponential behavior, similar to that observed for the UCNPs, and they were fitted with a biexponential profile (Equation (3)) and the results are reported in Table S8. The calculated average lifetimes for the film in the various environments closely resemblethose found for the UCNPs, clearly indicating that the UCNPs in the film have been preserved upon passing from the colloidal dispersion to the nanofilm.

Optical Thermometry and Primary Thermometers

2.3.6

The UC emission of the α‐Syn‐Au, UCNPs film was studied as a function of temperature between 25 and 60°C, a useful range for applications in real biological environments, in particular for in vivo experiments. We investigated the thermometric features of the film exploiting the emissions of the Er^3+^ ions from the ^2^ H 11/2 and ^4^ S 3/2 excited levels. Furthermore, we have carefully considered the possibility of exploiting the 2D nanostructured system as a primary thermometer, which would be very useful to avoid laborious calibrations of the thermometric system [57, 58, 59]. Assuming we have two levels, |1> and |2>, of increasing energy, the behavior of a primary thermometer is directly related to the population of these two levels, according to Boltzmann's law:

$$[eqn]$$

where N 1 and N 2 are the populations of the thermally coupled |1> and |2> levels, respectively, ΔE is the energy gap between the two levels, k _ B _ is the Boltzmann constant, T is the temperature. g 1, g 2 are the degeneracies of the |1> and |2> levels, respectively. Since the emission intensities from the excited levels depend on the corresponding populations, the ratio between these emission intensities (Luminescence Intensity Ratio, LIR), depends on the temperature and can be expressed by:

$$[eqn]$$

where I 1 and I 2 represent the integrated emissions from the |1> and |2> levels, respectively, B accounts for the level degeneracies, emission branching ratios, and spontaneous absorption coefficients of the two transitions [26]. Since B is constant with temperature, rearranging Equation (6) it is possible to correlate the temperature T with LIR(T) according to the following equation:

$$[eqn]$$

Therefore, if the ΔE and LIR(T 0)) values for a given temperature T 0 are known, it is in principle possible to estimate any other temperature simply using Equation (7). Although this aspect is fascinating, it is a crucial requirement for a correct application of Equation (7) that the populations of the two levels can be fully described by the Boltzmann law, Equation (5) in the entire range of temperatures. This is not always the case, as pointed out by several authors [27, 60, 61], as other processes (i.e. multiphonon or cross‐relaxation processes between lanthanide ions) could be responsible for altering the population of the two “thermalized” states. For our 2D nanostructured system, we evaluated ΔE using the energy barycenters of the two ^2^ H 11/2 and ^4^ S 3/2 levels from the excitation spectra, as described above.

For the sake of comparison, the thermometric performances of the α‐Syn‐Au, CaF_2_NPs film were evaluated in several environments: as a free‐standing film or immersed in solvents such as D_2_O and H_2_O, to highlight the differences in their performances and highlight possible applications.

Notable advantages of conducting thermometric measurements in D_2_O as a solvent are the significant enhancement of the UC intensity, as described above, and also the minimization of local heating, since the absorption of 980 nm radiation in D_2_O is negligible [47]. Thanks to these features, we excited the α‐Syn‐Au, CaF_2_NPs film with quite low laser irradiances (I = 20 kW/cm^2^). We note that this low laser irradiance minimizes the UC intensity of the ^2^ H 9/2 → ^4^ I 13/2 transition, as observed in other systems, which is detrimental to optical thermometry based on the Er^3+^ emission [45]. The UC spectra for the α‐Syn‐Au, CaF_2_NPs film in D_2_O as a function of temperature are shown in Figure 4A, evidencing the increase of the UC intensity for the ^2^ H 11/2 → ^4^ I 15/2 transition with respect to the ^4^ S 3/2 → ^4^ I 15/2 one, reflecting a corresponding relative increase of the ^2^ H 11/2 population. From the integrated emissions for the two thermalized bands and from their ratio (ratiometric method, see Equation (6), we obtained the experimental LIR(T) in the range 25–60°C, shown in Figure 4B, where the uncertainties were evaluated from three replicate measurements in the same experimental conditions. The LIR(T) exhibits a linear temperature dependence within the investigated temperature range and, from a linear fit, a slope of (0.00152 ± 0.00003) K^−1^ was found. An important parameter for nanothermometry is the relative thermal sensitivity S _ rel _, defined as:

$$[eqn]$$

*UC spectra (λ exc = 980 nm, laser I = 20 kW/cm2) for the α‐Syn‐Au, CaF2NPs film immersed in D2O (A), H2O (E) and air (I) for different temperatures: 2 H 11/2 → 4 I 15/2 (left) and 4 S 3/2 → 4 I 15/2 bands (right); LIR(T) vs measured T (black dots) and linear fit (red line) for the film in D2O (B), H2O (F) and air (J); Relative thermal sensitivities S

rel vs measured T for the film in D2O (C), H2O (G) and air (K). (D) Calculated T, from Equation (7), vs measured T for the film in D2O (D), H2O (H) and air (L). The red dotted line is a guide for the eye.*

The relative values of the thermal sensitivities S _ rel _ range between 1.25% and 0.85%, see Figure 4C), where a maximum value of 1.25 % K^−1^ was found for T = 25.0°C. The uncertainties in the S _ rel _ values were estimated by considering the uncertainties in LIR(T). The slope and the S _ rel _ values align well with those reported in the literature for different nanosized optical thermometers based on Er^3+^ emission [62, 63]. Moreover, the temperature uncertainty ΔT was calculated as:

$$[eqn]$$

where (ΔLIR) represents the error in the LIR(T) values. The average value of ΔT in the measured range was 1.54°C, which is similar to the values reported in the literature for Yb, Er doped UCNPs [29, 62].

To highlight the performance as a primary thermometer, we calculated the temperature for different ln(LIR(T)LIR(T0)) values, assuming that Equation (7) is satisfied for all temperatures in the range under investigation. We point out that the experiment with lower uncertainty in temperature measurements is at room temperature (25.0°C), i.e. when the system is reasonably in the best thermal equilibrium with the environment. Therefore, we considered T 0 = 298 K as the reference temperature, LIR(T 0) = LIR(25.0°C) and ΔE = 718 ± 30 cm^−1^ as input parameters for the calculation of the T (T _ c _) in Equation (7). The calculated vs measured temperatures are shown in Figure 4D, where the uncertainties were estimated by considering the error propagation using the corresponding uncertainties of the parameters in Equation (7). From this behavior, an excellent calculated/measured T agreement was found within the experimental errors, indicating that our nanocomposite film behaves as a primary thermometer in the investigated temperature range. Rearranging Equation (6), we obtain the following Equation (10), which shows the correlation between the logarithm of the LIR and 1T:

$$[eqn]$$

By the plot of ln (LIR(T)) against 1T (Figure S15c), it is possible to determine from the slope of the fit (−ΔEkB) the experimental thermometric value of ΔE between the thermalized levels. The experimental thermometric ΔE (729 ± 13 cm^−1^) is quite similar to that determined from the excitation spectra, indicating good behavior as a primary thermometer in D_2_O.

UC measurements for the film in H_2_O were performed at laser irradiances around one order of magnitude higher than for D_2_O, to enhance the UC intensity, reduce the acquisition time, and therefore minimize laser heating due to absorption by H_2_O during the measurement (Figure 4E). The LIR(T) vs T plot is shown in Figure 4F, denoting a linear behavior, with a slope of (0.00152 ± 0.00007) K^−1^, virtually identical to that found for the film in D_2_O, although affected by a larger error, mainly due to the lower S/N ratio. For the film immersed in H_2_O as a solvent, the S _ rel _ has a maximum value of 1.05% at 25°C (Figure 4G) and the minimum temperature uncertainty ΔT results in 7.5°C. Although the minimum temperature uncertainty is quite high, we highlight that in our case, the thermometric performances were evaluated for a quasi‐monolayer film of UCNPs, with a very few emitting UCNPs present in the light path, reflecting in a relatively low emission intensity. It is then reasonable that the obtained ΔT is much higher than that found for samples in powder form or even for concentrated colloidal dispersions. From Figure 4H, it is observed that the calculated temperature slightly deviates from the measured one calculated using the primary thermometer model (Equation (7)), and the effect is more pronounced for temperatures higher than 25°C. This behavior is reasonable because the room temperature was chosen as a reference, and it can likely be attributed to other relaxation processes that can alter the population of the thermalized levels with respect to Boltzmann's law (Equation (5), as MPR, due to the solvent. As in the case of D_2_O, we plotted ln(LIR(T)) against 1T (Figure S15b). The experimental thermometric ΔE (671 ± 22 cm^−1^) is quite different from the one found in the excitation spectra, which determines a non‐ideal behavior as a primary thermometer in H_2_O. Nonetheless, the behavior of the nanocomposite film as a primary thermometer is almost fulfilled in the investigated temperature range.

Measurements in air were performed at higher laser irradiance (200 kW/cm^2^) to enhance UC emission and improve the S/N ratio (Figure 4I). As shown in Figure 4J, the value of LIR(T) against the measured temperature shows a good linear behavior, and from a linear fit, a slope of (0.00150 ± 0.00006) K^−1^ was found, similar to that obtained from the case of immersion in H_2_O and D_2_O as solvents. In this case, the S_ rel _ has a maximum value of 1.1% at 25°C (Figure 4K) and a ΔT of 5.4°C in the measured range. Again, also in this case, the relevant ΔT value is due to multiple factors, mainly to the small amount of emitting UCNPs in the excitation/emission light path. In this case, as in the case of H_2_O, the calculated T via the primary thermometer method (Equation (7)) slightly deviates from the measured T (4L), with a decreasing trend as the temperature increases starting from the calibration point (25°C), but remaining compatible with the estimated uncertainties. The experimental thermometric ΔE (643 ± 24 cm^−1^) obtained from the plot ln(LIR(T)) against 1T (Figure S15a) is quite different from that found from the excitation spectra, which determines a non‐ideal behavior as a primary thermometer in air.

Image‐Based Optical Thermometry

2.3.7

As a proof‐of‐concept experiment, we explored the possibility of evaluating the nanofilm temperature by exploiting images of green UC emission collected with our home‐made experimental setup, using the cooled color CMOS camera and two long‐pass filters (Figure 5a). In particular, a 633 nm short‐pass filter was added in the emission path to remove the red UC emission (emission at wavelengths longer than 630 nm), due to the ^4^ F 9/2 → ^4^ I 15/2 transition, from the collected color image. Two color images were collected, one with and one without a 532 nm long‐pass filter (5c) to alternatively collect the total green UC emission (in the 510‐580 nm range, without the 532 nm long‐pass filter) or the one derived only from the ^4^ S 3/2 → ^4^ I 15/2 transition (in the 510–535 nm range, with the 532 nm long‐pass filter). The experimental setup, therefore, permits the measurement of the entire green emission (denoted as TOT, Figure 5b) and the ”cold” band centered at 548 nm (denoted as LP532) as color images. We point out that the images were collected with a 16‐bit depth, i.e. on an intensity scale from 0 to 65536 arbitrary units. Moreover, since we used a color CMOS, the pixel intensities are weighted by the response curves of the three color components (blue, green, and red) of the camera, and these components are merged in the output image (Figure S17). Since we focus on the collection of the green UC emission, we applied a Debayerization procedure to the images, to separate the three color components. For the image processing, we kept only the green component, which is the most representative in the spectral range under investigation. The experiments were carried out using the nanocomposite film in air by heating the sample at three representative temperatures: room temperature (24°C), 41°C and 61°C. The acquired images of the nanocomposite film as a function of the T are shown in Figure 5c, where images of the sample with (LP532) and without (TOT) the 532 nm filter are shown, together with their difference (TOT ‐ LP532). Although their difference results in an overall lower intensity, it is clearly visible after a rescale of the intensity axis (image brightness). The temperature of the sample was estimated by using the calibration curve determined for the nanocomposite in air, shown in Figure 4J. Therefore, by acquiring the UC spectra with the monochromator and CCD system (Figure 5e)and using this calibration curve, we estimated the temperature of the film. Then, we defined a LIR(T)_ image _ based on the images shown in Figure 5d by integrating the total intensity of the emission as:

$$[eqn]$$

where I _ LP532_ I _ TOT _ represents the spatially integrated intensity of the images with and without the 532 nm filter. The LIR(T)_ image _ vs the estimated T is shown in Figure 5d. Although the data are limited in number, they exhibit a linear behavior, and from a linear fit we obtained a slope of (1.4 ± 0.2) · 10^−4^. This slope value differs significantly from the one obtained from the calibration using the spectra (1.50 ± 0.06) · 10^−3^. This difference is reasonable, and imost likely due to the fact that the intensities of the images are filtered by the green response curve of the color CMOS camera (Figure S17). Although preliminary, this experiment lays the groundwork for image‐based optical thermometry, which involve rapid image collection using a relatively inexpensive CMOS camera and image processing.

(a) Scheme of the microscope setup for the thermometry‐imaging experiment; (b) UC spectra of the film collected with (black line) and without (green line) 532 nm long‐pass filter; (c) UC image for the nanocomposite film at different T collected without (TOT) and with (LP532) the 532 nm long‐pass filter and image of the 525 nm centered emission calculated subtracting the signal of LP532 image to the TOT image with intensity increased by 10 times to observe the signal; (d) LIR(T) image (see Equation 11) vs calculated T; (e) Detail of the 2 H 11/2 (left) → 4 I 15/2 and 4 S 3/2 → 4 I 15/2 (right) UC bands at the calculated T.

In the literature, optical thermometry based on images is a research topic still in its infancy, with only a few noticeable examples.

For instance, Liu et al. [64] developed a method for real‐time, wide‐field optical temperature mapping using single‐shot photoluminescence lifetime imaging thermometry and NaGdF_4_:Er^3+^, Yb^3+^/NaGdF_4_ core–shell NPs as phosphors. Although this methodology is very interesting, the thermal relative sensitivity was estimated to be 0.39%–0.43% for the green emission, which is slightly lower than that obtained in the present case for the Er^3+^ green emission (1.1%–0.8%, see Figure 4). Moreover, the total time needed to form a lifetime map was of 48 min; in our case, significantly less time, on the order of seconds, was required to collect the emission signal with a very good signal‐to‐noise ratio, and a few minutes were needed for processing the data and obtaining the estimated temperature. Furthermore, our experimental setup is much simpler than the one used by Liu et al. [64]

A ratiometric luminescence thermometer based on Ba_3_(VO_4_)2 activated with Mn^5+^ and Nd^3+^ for deep tissue thermal imaging, investigated by Piotrowski et al. [65], revealed a thermal sensitivity of 1% K^−1^, similar to that obtained in the present investigation. Differently from our case, the thermal images were taken with a thermal camera, on a sample of around 1x5 cm size.

An important investigation was also carried out by Martinez et al. [66], who combined hyperspectral microscopy with the upconversion properties of Er^3+^‐doped NaYF_4_ nanoparticles to produce thermal images. To this end, a ratiometric method involving the emissions derived from the two thermalized bands of Er^3+^ ions in the green region was employed, analogous to the interesting investigation. The hyperspectral image consisted of 696 lines, with a mean scan time of 3.4 s for each line, totaling approximately 40 min. Although the method proposed by Martinez et al. is interesting, in our case, we were able to acquire emission intensities for all pixels simultaneously with the CMOS camera in a few seconds, simply by working with optical filters.

Conclusions

3

Multifunctional films with potential applications in nanothermometry hold promise for the advancement of nanomaterial science and biomedicine [67, 68, 69].

In this work, we demonstrated that the fibril‐forming ability of α‐synuclein, driven by a conformational transition under specific conditions, can be exploited to fabricate a luminescent hybrid film composed of AuNPs, CaF_2_:Yb,Er NPs, and α‐synuclein, with the protein serving as a structural linker.

The CaF_2_:Yb,Er were prepared via a co‐precipitation method followed by hydrothermal treatment in a microwave reactor, which produced nanocrystalline NPs exhibiting bright upconversion luminescence.

In‐solution NMR experiments revealed that the N‐terminal domain of α‐Syn adsorbed onto the surface of NPs. The NAC domain exhibited partial immobilization, likely because of its proximity to the adsorbed N‐terminus. In contrast, the acidic domain remained flexible, attributed to electrostatic repulsion from the negatively charged NPs surface. Organic solvent treatment induced a structural conversion to beta‐sheet material.

The produced films were characterized by TEM, which revealed the presence of AuNPs distributed throughout the film structure and tightly packed with CaF_2_:Yb,Er NPs.

We demonstrated the potential of these quasi‐monolayer films as optical primary nanothermometers, requiring no calibration within the entire application temperature range, just by determining the energy difference between the two thermalized levels with excitation spectroscopy and calibrating the emission at a single temperature. The relative sensitivity of the films is comparable to other optical nanothermometers based on Yb and Er‐doped UCNPs, while the ΔT_ min _ is higher due to the low amount of UCNPs which determines a low signal‐to‐noise ratio.

We also conducted a proof‐of‐concept experiment to evaluate the feasibility of using an optical imaging‐based approach for local temperature determination via upconversion emission in a 2D nanosystem. Following image acquisition and processing, the temperature could be rapidly estimated by exploiting a previously obtained calibration curve, either directly or derived from a primary thermometer methodology.

We point out that the present approach has, in principle, the potential to enable image‐based optical thermometry with pixel‐level spatial resolution, corresponding to a micrometric scale. The definition of spatial resolution of temperature mapping is quite difficult. An interesting approach could be the one used by Martinez et al. [66], correlating the spatial resolution of the measurement with the temperature uncertainty and the maximum temperature gradient.

Finally, we note that the AuNPs incorporated into the film confer plasmonic properties to the resulting film. The strong absorption can, in principle, be exploited to generate heat if the sample is irradiated with radiation within the plasmonic band. This property suggests that the nanocomposite film can serve as a multifunctional tool for both heating and temperature monitoring, albeit using different types of radiation, paving the way to applications in biomedicine.

Experimental Section

4

Preparation of the Au NPs

4.1

The gold NPs (AuNPs) were prepared according to Martinez et al. [70] by a modified Turkevich method in the presence of Ag^+^ ions, which narrow the size distribution of the NPs and leads to the formation of quasi‐spherical NPs [71]. Briefly, 1.6 mL of a solution of sodium citrate (510 mM) (⩾ 99%, Sigma–Aldrich), 250 µL of a solution of AgNO_3_ (10 mM) (⩾ 99%, Sigma–Aldrich), and 500 µL of a solution of HAuCl_4_·3H_2_O (250 mM) (⩾ 49%, Thermo Scientific) were added to 5.6 mL of Milli‐Q water and stirred for 5 min. The obtained solution was then quickly poured into 117 mL of boiling Milli‐Q water and heated under reflux for 1 h. The suspension was then cooled down to room temperature. The AuNPs were stable at room temperature for several months.

Preparation of CaF2:Yb,Er Upconverting Nanoparticles

4.2

The Yb^3+^, Er^3+^ doped CaF_2_ NPs (CaF_2_:Yb,Er UCNPs or simply UCNPs) were synthesized exploiting a microwave‐assisted hydrothermal method similar to that reported by Milan et al. [72] Briefly, 2.340 mL of a solution of CaCl_2_ (1.00 M), 0.600 mL of a 1.00 M solution of YbCl_3_· 6H 2 O (purity ⩾ 99.99%, Thermo Scientific) and 0.060 mL of a 1.00 M solution of ErCl_3_ (purity 99.99%, Alfa Aesar) and 18 mmol of solid potassium citrate tribasic monohydrate (purity >99%, Alfa Aesar) were poured in a G‐30 glass microwave vial containing 12 mL of deionized water and left stirring until complete dissolution. Then, 2.143 mL of a 3.5 M solution of NH_4_F (purity ⩾ 98.0%, Sigma–Aldrich) was added to the mixture and left stirring for one minute. The resulting solution was then heat treated at 190°C in a microwave reactor (Anton–Paar Monowave 400) for 20 min and rapidly cooled to room temperature. The NPs were then collected by centrifugation at 7000 g for 10 min after the addition of 10 mL of acetone, washed three times with a solution of water and acetone (1:2 in volume), and stored in pellet form under acetone.

Preparation of the α‐Synuclein

4.3

The α‐Syn codifying gene was inserted into a pET‐28a(+) vector (Novagen). E. coli BL21(DE3) cells were transformed with the α‐Syn plasmid and grown in LB medium at 37°C to OD600 of 0.4. Protein expression was induced by adding IPTG at a final concentration of 0.1 mM. The culture was kept at 37°C for 5 h, and then bacterial cells were collected and lysed. The lysate was boiled for 10 min and centrifuged at 14000 g for 20 min at 4°C. The soluble fraction was treated with 35% ammonium sulfate to precipitate the contaminants. After centrifugation, the supernatant was treated with 55% ammonium sulfate (Reagent Grade) to precipitate α‐Syn. The pellet was resuspended in a buffer solution, filtered, loaded into an anionic exchange column (Resource Q, Amersham Bioscience), pre‐equilibrated with 20 mM Tris‐HCl buffer, pH 8, and eluted with a linear gradient from 0 to 500 mM of NaCl. The fractions containing α‐Syn were collected and stored at −20°C. Analogously, the ^15^N‐isotope‐labeled protein was produced by growing E. coli cells in an M9 minimal medium containing 1 g/L of ^15^NH_4_Cl (Reagent Grade) as a sole source of nitrogen.

Preparation of the α‐Syn‐Nanoparticles Films

4.4

The hybrid α‐Syn‐Nanoparticles films were prepared by a modification of the procedure described by Lee et al. [67] This method involves exposing α‐Syn to chloroform, inducing a conformational transition that promotes β‐sheet‐based intermolecular interactions, ultimately stabilizing a free‐standing hybrid thin film. We exploited this method to produce films containing both AuNPs and CaF_2_:Yb, Er UCNPs. We point out that the simultaneous presence of AuNPs with the UCNPs in the free‐standing film is crucial for its proper handling, as it results red coloured, due to the AuNPs. In fact, the film produced with only the UCNPs would not be colored but completely invisible, making its handling much more difficult. First, the AuNPs or CaF_2_:Yb, Er UCNPs were incubated with the α‐Syn to permit the formation of a protein corona and after removal of the protein in excess, the pH is adjusted to 4.5. Then, the α‐Syn‐Au, UCNPs suspension was incubated on a polycarbonate film (PCF) in a humid chamber at 40°C to induce interactions among the NPs and their deposition on the PCF. After proper washing and drying, the NPs layer was immersed in chloroform to initiate the α‐Syn fibril formation, therefore releasing the α‐Syn‐Au, UCNPs layer from the PCF.

The α‐Syn‐Au, UCNPs films were prepared using either AuNPs or a 1:1 (w/w) mixture of AuNPs and CaF_2_:Yb, Er UCNPs. The starting α‐Syn‐AuNPs dispersion was prepared by incubating 2.5 mL of AuNPs dispersion (0.2 mg/mL) with 71 µLof α‐Syn 1 mM; The resulting dispersion was dialyzed for 12 h at 4°C. The α‐Syn‐UCNPs dispersion was prepared by incubating the CaF_2_:Yb, Er UCNPs at 4°C with α‐Syn, in a 2:1 weight ratio, to induce the formation of a protein corona surrounding the NPs. To this aim, 0.5 mL of a 1 mg/mL colloidal dispersion of CaF_2_:Yb, Er UCNPs were incubated with 71 µL of α‐Syn 1 mM at 4°C for 12 h.

The α‐Syn‐Nanoparticles films were obtained using the following procedure. First, a 1 mm polycarbonate film (PCF) was deposited on a microscope slide by drop‐casting a saturated polycarbonate solution on a microscope glass slide as a support. The α‐Syn‐AuNPs films were then prepared by placing 50 µL of α‐Syn‐AuNPs suspension on the PCF with 5 µL of a citrate buffer solution (500 mM) at pH = 4.5. The hybrid α‐Syn‐Au, UCNPs films were then prepared by mixing 50 µL of α‐Syn‐AuNPs dispersion with 10 µL of α‐Syn‐UCNPs one, together with 6 µL of citrate buffer solution onto the PCF, to form a single droplet. The droplets were then placed in an oven, in a humid environment, and incubated at 40°C for 3 h to induce aggregation of the NPs and deposition onto the PCF. The excess water was removed using a micropipette, and the aggregated α‐Syn‐Au, UCNPs were washed twice with deionized water before drying at room temperature. The deposited α‐Syn‐Au, UCNPs on the PCF were then immersed in chloroform to release the hybrid films (see Video S1). After the removal of residual chloroform, the films were transferred onto a glass cover slip or a TEM grid.

Experimental Setup for Investigating the α‐Syn‐Nanoparticles Interactions

4.5

NMR Setup and Analysis

4.5.1

NMR spectra were recorded on a Bruker Avance III 600 spectrometer, operating at 600.13 MHz proton Larmor frequency, equipped with a triple resonance TCI cryoprobe. Measurements were performed at 15°C. ^1^H,^15^N heteronuclear single quantum coherence (HSQC) spectra were recorded with 8 scans, 512 increments, and a recycle delay of 1.2 s. HSQC spectra were processed with Topspin 3.6.4 and analyzed with Sparky (T. D. Goddard and D. G. Kneller, University of California, San Francisco). Intensity and chemical shift perturbation data plots were generated with GraphPad Prism 8 (GraphPad Software Inc., La Jolla, CA, USA).

Thioflavin‐T Assay

4.5.2

The transition of α‐Syn into a β‐sheet‐rich conformation was assessed by adding the fluorescent dye thioflavin‐T (20 µM) to samples containing 1) NPs+ α‐Syn or 2) NPs+ α‐Syn+ chloroform. NP and α‐Syn concentrations were 6 and 213 µM, respectively. The solutions were incubated for 4 h in static conditions and 2 h with shaking after the addition of CHCl_3_ at room temperature. Control experiments indicated that ThT fluorescence intensity was only slightly elevated in chloroform compared to its level in water. Measurements were carried out on a Tecan Infinite M200 PRO fluorescence microplate reader, at least in triplicate, in a flat‐bottom 96‐well black microplate; the excitation and emission wavelengths were 450 and 482 nm, respectively, and the gain was set to 100. Data are plotted as mean ± s.d. using GraphPad Prism 8.

Structural and Morphological Characterization

4.6

X‐Ray Powder Diffraction

4.6.1

The X‐Ray Powder Diffraction (XRPD) patterns were collected with a Rigaku SmartLab SE diffractometer using a Copper‐anode X‐ray tube (K_α_, λ = 1.54 Å) and an XSPA‐400ER 2D/1D/0D hybrid pixel array direct photon counting detector. The XRPD pattern was collected in a Bragg–Brentano geometry with a scan rate of 10°/min and a step size of 0.02°. The crystalline phase was analyzed by comparing the measured XRPD pattern with the one found in the ICSD database. The crystallite size was estimated using the Scherrer equation (see Supporting Information).

Transmission Electron Microscopy and Elemental Analysis

4.6.2

High‐resolution TEM (HRTEM), Scanning TEM (STEM) and EDX mapping were acquired using a JEOL JEM‐F200 TEM equipped with a cold FEG gun and operating at 200 kV in HREM. The AuNPs and UCNPs samples were prepared by dropping 5 µL of NPs dispersion (about 0.1 mg/mL) on a Formvar–coated copper grid (300 mesh, MEDIA System Lab). The α‐Syn‐Au,CaF_2_NPs films were directly deposited on carbon‐coated copper grids (300 mesh, MEDIA System Lab) for the TEM investigation. EDS were performed by a windowless Jeol, JED‐2300 series, SDD detector.

Colloidal Property Measurements

4.6.3

Hydrodynamic sizes and ζ‐potentials of AuNPs, UCNPs, α‐Syn‐AuNPs and α‐Syn‐UCNPs were measured using the Dynamic Light Scattering technique with a Malvern Zetasizer Nano ZS equipped with a He‐Ne laser at 633 nm on water dispersions. For hydrodynamic diameter and ζ‐potential measurements, disposable micro–cuvettes (Brand) or folded capillary cells (Malvern, DTS1070) were used, respectively.

ICP‐MS Analysis

4.6.4

The α‐Syn‐Au,CaF_2_NPs film was first treated with aqua regia in order to digest the NPs and dissolve the metallic ions in the solution for 1 h and then diluted with 1% HNO_3_. The ICP‐MS analysis was performed using a iCAP RQ ICP‐MS ‐Thermo Scientific.

Spectroscopic Measurements

4.7

Upconversion Spectra and Images

4.7.1

The UC emission spectra of the NPs and the α‐Syn‐AuNPs, UCNPs film were collected by exciting at 980 nm with a diode laser (MDL III 980, CNI) as the source, focused using a 20× microscope objective. The emitted radiation was collected in backscattering mode and directed through a 925 nm longpass dichroic mirror (Thorlabs) to a monochromator (Shamrock‐500i, Andor) equipped with a CCD camera (DU420A‐BVF, ANDOR) cooled at –80°C. The optical resolution of the emission spectra is 0.7 nm.

A custom‐built fluorescence microscope was employed to measure UC images (see Figure S1). The UC emission was detected by a cooled CMOS camera (ZWO, ASI533MC Pro Color). UC emissions from powders or nanocomposite films were measured on a microscopy slide or coverslip, respectively.

Optical Thermometry

4.7.2

Optical thermometry experiments were conducted using the same experimental setup described for UC measurements. The film sample was directly deposited on the inside face of a quartz cuvette, then filled with D_2_O or H_2_O. The UC spectra were collected in a temperature range between 25 and 60°C at 5°C intervals. The temperature was measured with a Pt thermocouple (temperature uncertainty of 0.1°C).

Conflicts of Interest

The authors declare no conflict of interests.

Supporting information

Supporting File 1: smll72424‐sup‐0001‐SuppMat.pdf.

Supporting File 2: smll72424‐sup‐0002‐VideoS1.mp4.