In silico process for identifying the design of a nanostructured plasmonic sensor

Abstract

An in silico process for identifying the design of a nanostructured plasmonic sensor comprises the following operating steps: selecting an analyte; characterising the analyte with computational techniques; selecting at least one two- or three-dimensional nanostructured substrate having plasmonic characteristics; characterising each substrate with computational techniques; performing a mapping of the potential of interaction between the analyte and each substrate so as to identify, for each region and / or each atom of each substrate, a value of interaction with the analyte; calculating the spectral signal of the analyte adsorbed onto each substrate for at least one region and / or atom of each substrate; analytically evaluating an enhancement factor of each spectral signal calculated; and comparing each enhancement factor with a corresponding predefined minimum threshold value so as to evaluate whether the associated nanostructured plasmonic sensor is configured to perform detection of the analyte with single-molecule sensitivity.

Description

[0001] IN SILICO PROCESS FOR IDENTIFYING THE DESIGN OF A NANOSTRUCTURED PLASMONIC SENSOR

[0002] DESCRIPTION

[0003] Technical field

[0004] The present invention is related to an in silico process for identifying the design of a nanostructured plasmonic sensor.

[0005] In particular, given a specific analyte of interest, the present invention provides a multi- scale in silico process, based on principles of quantum mechanics and classic electrodynamics, for identifying the best chemical composition of a nanostructured substrate with a plasmonic character, the morphology thereof on a macroscopic and microscopic scale, and the best geometry for the reciprocal disposition of the substrate / analyte pair (optionally in suspension in a solvent), with the aim of achieving detection of the analyte itself by means of spectroscopic techniques with single- molecule sensitivity.

[0006] Prior art

[0007] At present, nanostructured plasmonic sensors are produced for the purpose of exploiting Raman spectroscopy - “SERS” (or, optionally, other types of spectroscopy) to detect, from small samples, even smaller amounts of an analyte contained in the sample.

[0008] For example, nanostructured plasmonic sensors can be used to identify the presence of specific pathogens within a few drops of blood taken from a patient, or the presence of polluting agents in a small amount of water.

[0009] When a laser beam strikes an analyte, the molecules can become excited, giving rise to inelastic scattering, also known as Raman scattering. When the ray strikes the molecules, photons are released which have a frequency differing from the frequency of the incident laser. Generally, however, the scattering signal is weak and is difficult to detect with the most widely available spectroscopy apparatus.

[0010] Nanostructured plasmonic sensors, by virtue of their particular morphology and the plasmonic characteristics of the material making up the substrate, are advantageously capable of amplifying the normal scattering signal so as to make it more easily detectable with common spectroscopy apparatus.

[0011] The production of a nanostructured plasmonic sensor (i.e. the choice of the morphology of the substrate and the material making it up) usually takes place with the known heuristic “trial and error” method.

[0012] Research laboratories usually possess a limited number of utilisable plasmonic materials and, above all, possess a limited number of apparatus for processing such materials in order to produce a substrate useful for a nanostructured plasmonic sensor. Therefore, with the aforesaid “trial and error” method, the laboratories establish, based on a desk analysis, one or more utilisable materials and in which nanostructures to model them with the aim of obtaining nanostructured plasmonic sensors to be tested at length in order to obtain usable results, i.e. to obtain spectroscopic output signals that are intense enough to be detected, while also having a sufficiently high degree of certainty such as to be able to affirm that the desired analyte has been detected rather than other molecules present in the starting solution / sample to be analysed.

[0013] Summary

[0014] In this context, the technical task at the basis of the present invention is to propose an in silico process for identifying the design of a nanostructured plasmonic sensor that overcomes the abovementioned drawbacks of the prior art.

[0015] In particular, one object of the present invention is to provide an in silico process for identifying the design of a nanostructured plasmonic sensor that is capable of identifying one or more substrates that are particularly suitable for detecting an analyte of interest, preferably with the use of a spectroscopy technique.

[0016] Another object of the present invention is to provide an in silico process for identifying the design of a nanostructured plasmonic sensor that is capable of defining the best zones where the analyte should be adsorbed onto a substrate in order to simplify the identification thereof, preferably with spectroscopy techniques.

[0017] A further object of the present invention is to provide an in silico process for identifying the design of a nanostructured plasmonic sensor that is capable of detecting an analyte with single-molecule sensitivity and further capable of discriminating whether the detection refers to a molecule of the analyte of interest or a different molecule. One object of the present invention is also to propose an in silico process for identifying the design of a computer-implementable nanostructured plasmonic sensor which, with a high degree of reliability, has a reduced computational “cost” in terms of the power and / or computing times of the computer.

[0018] The stated technical task and specified objects are substantially achieved by an in silico process for identifying the design of a nanostructured plasmonic sensor, which comprises the technical features disclosed in the independent claim. The dependent claims correspond to further advantageous aspects of the invention.

[0019] It should be understood that this summary introduces a selection of concepts in simplified form, which will be further expanded on in the detailed description given below.

[0020] The invention relates to an in silico process for identifying the design of a nanostructured plasmonic sensor configured to detect at least one molecule of an analyte of interest capable of being adsorbed onto a substrate of the sensor itself.

[0021] In particular, the process comprises the following operating steps: selecting an analyte to be detected by means of the nanostructured plasmonic sensor; morphologically and conformationally characterising the analyte with computational techniques with the aim of defining the molecular and / or supramolecular structure thereof, extracting a geometric model thereof in terms of defined spatial coordinates, identifying the degrees of molecular freedom, carrying out a conformational study thereof and, in the cases of analytes in the condensed phase (homogeneous or inhomogeneous), carrying out a conformational study in the presence of the external environment; selecting at least one substrate of the nanostructured plasmonic sensor having a two- or three-dimensional nanostructured morphology and produced with a material having plasmonic characteristics; morphologically and conformationally characterising each substrate with computational techniques in order to extract a geometric model in terms of defined spatial coordinates; performing a mapping of a potential of interaction between the analyte and the substrate so as to identify, for each region and / or each atom of the substrate, a value of interaction with the analyte and, in particular, the regions and / or atoms of the substrate of greatest interaction with the analyte; calculating the spectral signal (generated by the electric and / or magnetic and / or nuclear response or any combination of the latter) of the analyte adsorbed onto each substrate (optionally also in the presence of a solvent) for at least one atom and / or region of each substrate; analytically evaluating an enhancement factor of each spectral signal calculated; comparing each enhancement factor with a corresponding predefined minimum threshold value so as to evaluate whether the associated nanostructured plasmonic sensor is configured to perform detection of the analyte with single-molecule sensitivity; generating an output signal identifying at least one design for a nanostructured plasmonic sensor capable of detection with single-molecule sensitivity, wherein at least one enhancement factor is greater than said minimum threshold value, said at least one enhancement factor referring to an atom of a corresponding substrate for which the value of the interaction potential is greater than a value of minimum interaction. In other words, the aforesaid in silico process defines a computational protocol capable of answering the following questions by associating a series of computational simulations that enable a specific answer to be obtained:

[0022] - which substrate with a plasmonic character is ideal for the analyte of interest? Is the chemical composition of the plasmonic substrate based on graphene? On a plasmonic metal? On an alloy of metals with a plasmonic character?

[0023] - what morphology of the substrate allows the maximum increase in the value of the electromagnetic field in proximity to the analyte? Does the morphology of the plasmonic substrate have to be nanostructured? With the creation of particular shapes / conformations as metal nanoprisms or nanotubes, or sheets of graphene in the shape of a star or triangle? Is it necessary or advisable for the substrate to have structural defects?

[0024] - In what point of the geometry of the substrate is the maximum interaction with the analyte obtained? Does the zone of maximum probability of interaction coincide with the region of maximum increase in the field of interaction with the analyte? Is it possible to modify the morphology of the substrate to make the zone of maximum probability of interaction coincide with the region of maximum increase in the field of interaction with the analyte? Is it possible to modify the substrate so as to increase the spectral signal in the absorption region?

[0025] - Is the sample immersed in a solvent? If so, what is the ideal solvent for the analyte / substrate pair based on the previous points / questions?

[0026] In other words, the aforesaid process is advantageously capable of providing an in silico protocol for identifying at least one substrate that is particularly suitable for defining a nanostructured plasmonic sensor useful for detecting at least one molecule of an analyte of interest by exploiting spectroscopy techniques, such as, for example, Raman spectroscopy.

[0027] Brief description of the drawings

[0028] Additional features and advantages of the present invention will emerge more clearly from the approximate, and thus non-limiting, description of a preferred but not exclusive embodiment of an in silico process for identifying the design of a nanostructured plasmonic sensor.

[0029] The accompanying drawings illustrate different aspects and features of the aforesaid process, in which: figure 1 illustrates, according to a schematic view, a flow diagram of the in silico process in accordance with the invention; figure 2 illustrates a same analyte adsorbed in different regions of a same substrate and the respective Raman spectra calculated in accordance with the process of the invention (example 1 ); figure 3 illustrates a same analyte adsorbed onto different morphologies of a same substrate and the respective Raman spectra calculated in accordance with the process of the invention (example 2); figure 4 illustrates a same analyte adsorbed onto different substrates and the respective Raman spectra calculated in accordance with the process of the invention (example 3); figures 5, 6 illustrate different analytes adsorbed onto a same substrate and the respective Raman spectra calculated in accordance with the process of the invention (example 4); figure 7 illustrates a comparison between the experimental spectra and the Raman spectra calculated in accordance with the process of the invention (example 5).

[0030] With reference to the drawings, they serve solely to illustrate embodiments of the invention for the purpose of better clarifying, in combination with the description, the inventive principles at the basis of the invention.

[0031] Detailed description of at least one embodiment

[0032] The present invention relates to in silico process for identifying the design of a nanostructured plasmonic sensor configured to detect at least one molecule of an analyte of interest capable of being adsorbed onto a substrate of said sensor.

[0033] In particular, figure 1 illustrates a schematic representation of the process, which comprises a plurality of operating steps.

[0034] To begin with, the first step of the process comprises selecting an analyte to be detected by means of the nanostructured plasmonic sensor.

[0035] For example, the analyte can be a polluting particle present in the air or in water, or a pathogenic agent present in a patient’s blood.

[0036] Therefore, the nanostructured plasmonic sensor will be developed according to the analyte it is intended to detect. The constructive features and intrinsic properties of the nanostructured plasmonic sensor (deriving from the materials and / or morphology that define it) will depend on the analyte, the intrinsic properties of the latter, and its morphology.

[0037] Following the choice of the analyte, the process envisages carrying out a step of morphologically and conformationally characterising the analyte with computational techniques with the aim of identifying, for each conformer of the same analyte, at least the minimum energy conformations.

[0038] For the purpose of identifying one or more nanostructured plasmonic sensors that perform particularly well in detecting the aforesaid analyte (i.e. are capable of detecting the analyte with a minimum single-molecule sensitivity), it is useful to define for the latter the molecular and / or supramolecular structure, extract therefrom a geometric model in terms of spatial coordinates however defined (Cartesian, internal), define the degrees of molecular freedom thereof, and subject it to a conformational study, possibly in the presence of an external environment if it is an analyte in a condensed phase (homogeneous or inhomogeneous).

[0039] Preferably, the computational characterisation of the analyte enables a text file to be obtained, for example a file in “txt” format, containing a set of morphological and conformational information and other information related to the properties of the analyte, its axes / planes of symmetry and, moreover, its minimum energy conformations.

[0040] Generally speaking, a molecule tends to take on the conformation in which the energy is the lowest possible.

[0041] The energy can be described as an analytic function of the nuclear coordinates E(R) . In order to be able to obtain the optimal conformations of the analyte in interaction with the substrate (hence the set of coordinates R), recourse is had to a process of minimisation of the energy function E(R). This minimisation process is carried out using numerical techniques, i.e. with an iterative process, which requires, however, knowledge of the derivative of E(R) . The minimum of the aforesaid function is obtained when the derivative is zero.

[0042] However, it needs to be considered that the minimisation process makes it possible to obtain a relative minimum closer to the starting conformation, and not to an absolute minimum.

[0043] For the purpose of sampling a large number of starting conformations of the analyte, it is possible to use computational simulation techniques based on “molecular dynamics”, “Monte Carlo” techniques, or other stochastic methods.

[0044] Advantageously, molecular dynamics techniques or other stochastic methods are useful for modelling the analyte also when it is in the presence of a solvent.

[0045] Preferably, each possible orientation of the analyte adsorbed onto the substrate is sampled using molecular dynamics techniques.

[0046] In accordance with one aspect of the invention, the step of morphologically and conformationally characterising the analyte is performed with a quantum-mechanical analytical technique.

[0047] Preferably, the step of morphologically and conformationally characterising the analyte is performed by exploiting quantum-mechanical computational methods based on density functional theory (DFT).

[0048] In accordance with another aspect of the invention, in the presence of a solvent, the step of morphologically and conformationally characterising the analyte preferably also comprises the use of hybrid methods based on quantum mechanics and classical mechanics: the aforesaid quantum mechanics / molecular mechanics (“QM / MM”) methods. In this case, the analyte and the solvent, if present, are treated on a quantum- mechanical level, preferably with the aforesaid DFT. This spatial region is referred to as “QM”.

[0049] The remaining part of the solvent, by contrast, is preferably described according to laws of classic physics. Therefore, this spatial region is defined “MM”.

[0050] The interaction between these two different spatial regions (the “QM” part and “MM” part) is taken into consideration, preferably, by exploiting the most modern theoretical models. The hybrid “QM / MM” methods have a reduced computational cost compared to purely quantum-mechanical approaches. Therefore, the use of hybrid “QM / MM” methods enables the analyte to be characterised morphologically and conformationally, also in interaction with a much greater number of solvent molecules than could be achieved with purely quantum-mechanical approaches alone.

[0051] Following the step of selecting and characterising the analyte, the process envisages selecting at least one substrate of the nanostructured plasmonic sensor having a two- or three-dimensional nanostructured morphology and produced with a material having plasmonic characteristics.

[0052] Essentially, the substrate / analyte pair defines the nanostructured plasmonic sensor.

[0053] Therefore, in order that a nanostructured plasmonic sensor capable of detecting the analyte of interest, preferably with single-molecule precision, may be produced, the substrate itself must have a nanostructured morphology and, moreover, it must have plasmonic characteristics.

[0054] Some examples of two- or three-dimensional geometrical structures / morphologies having any shape, size and chemical nature whatsoever can be: cones, stars, tips, colloidal aggregates of metals, etc. Such structures are to be considered produced / producible on a nanometric scale for the purposes of producing the substrate of the nanostructured plasmonic sensor.

[0055] Some of the possible materials with plasmonic characteristics that can be used for the substrate are gold, silver, graphene, or a gold / silver alloy.

[0056] The substrate can be produced by means of any combination of the possible nanostructured morphologies (of which a non-exhaustive list has been provided above by way of example) and of plasmonic materials (i.e. with plasmonic characteristics, of which a non-exhaustive list has been provided above by way of example).

[0057] Preferably, in a research laboratory in which the process in accordance with the invention is applied, the selection of the substrate will initially be guided by the choice of plasmonic materials already in the laboratory’s possession (i.e. ready-to-use materials whose properties are known and which one is accustomed to handling for the manipulation thereof on a nanostructured level) and of the nanostructured morphologies that are regularly produced in the aforesaid laboratory and which can thus be produced with the instruments that laboratories are already provided with.

[0058] In this manner, once the analyte has been bound chemically (or by chemical adsorption) or physically onto the substrate, the nanostructured plasmonic sensor will enable the detection thereof using spectroscopic techniques, such as, for example, Raman spectroscopy, or IR spectroscopy.

[0059] Although the initial choice of the chemical nature of the substrate and its morphology is determined by the operator, preferably based on the instruments available to the laboratory concerned, the in silico process in accordance with the invention is capable of testing different geometries of the substrate in question. Moreover, if the interaction between the analyte and the substrate in question is particularly weak, the aforesaid process can test substrates of varying chemical nature, different from the one initially selected by the user, so as to maximise the interaction between the analyte of interest and the substrate.

[0060] Following the choice of at least one substrate, the process envisages carrying out a step of morphologically and conformationally characterising each substrate with computational techniques, in order to extract a geometric model thereof in terms of spatial coordinates however defined (Cartesian, internal).

[0061] Following the morphological and conformational characterisation of each substrate, the process envisages identifying, for each substrate, the plasmonic peak that corresponds to the maximum electron absorption of each substrate.

[0062] Preferably, in accordance with one aspect of the invention, the step of determining the plasmonic peak (i.e. the step of morphologically and conformationally characterising each substrate) is performed using the atomistic model known as "ωFQ" . The "ωFQ" model can be modified with specific correction parameters as a function of the two- or three-dimensional morphology of the substrate and / or the presence of interband interactions of the substrate material.

[0063] In particular, the aforesaid "ωFQ" model allows for the description of the plasmonic response of any type of plasmonic substrate, following a process of optimisation of the parameters defining the model. It should be noted that, when the "ωFQ" model is used, the abovementioned plasmonic peak is completely described by the charges and, possibly, by the dipoles present on the plasmonic substrate, at the specific external frequency at which the aforesaid peak occurs.

[0064] The "ωFQ" model is preferably used to study the plasmonic excitation of a metal substrate in the presence of an external radiation field (for example the one induced by an external laser).

[0065] The publications cited below analyse the aforesaid "ωFQ" model and other features thereof: “T. Giovannini, M. Rosa, S. Corni, C. Cappelli. Nanoscale 2019, 1 1 , 6004- 6015”; “L. Bonatti, G. Gil, T. Giovannini, S. Corni, C. Cappelli. Front. Chem. 2020, 8, 340”; “T. Giovannini, L. Bonatti, M. Polini, C. Cappelli. J. Phys. Chem. Lett. 2020, 1 1 , 7595-7602”; “P. Lafiosca, T. Giovannini, M. Benzi, C. Cappelli. J. Phys. Chem. C 2021 , 125, 23848-23863”; “L. Bonatti, L. Nicoli, T. Giovannini, C. Cappelli. Nanoscale Adv. 2022, 4, 2294-2302”.

[0066] The"ωFQ" model assigns a complex charge to every atom of the substrate, the value of the charge changing as a function of the interaction thereof with the other charges of the system and with the external radiation field. This model, based on concepts deriving from the Drude model of conduction and quantum tunnelling, is used to model plasmonic substrates that have no interband absorption in the incident frequencies of interest.

[0067] Some examples in this regard could be sodium and graphene.

[0068] The final equation of the "ωFQ" model that needs to be solved in order to obtain the charges associated with every atom of the system is expressed as (Aq- z(ω)I)q = R. The term Aqrepresents the charge-charge interaction among all the atoms making up the system, the term z(ω) contains the terms of classical conductivity (i.e. derivable from the Drude model), which depend on the frequency of the external field (to) and, finally, the term R includes the contribution of the electric potential associated with the external electric field on the charges of the substrate.

[0069] Upon solving the system of linear equations, it is possible to derive the set of complex charges (q) corresponding to every atom of the substrate alone at a given frequency. The same system can also be solved in the case of modelling of two-dimensional substrates, for example graphene.

[0070] In such a case, the previously described physical parameters are suitably modified in order to consider the structural particularities of the substrate of interest.

[0071] It should be noted that the linear system is composed of a number of linear equations equal to the number of atoms considered in the substrate, since the "ωFQ" model is completely atomistic. Therefore, the solution of the system can be obtained with the use of computational numerical techniques.

[0072] In the event that the material making up the substrate shows interband excitations, for example in the case of gold, silver or other metals comprising “d”-type electrons, the "ωFQ" model can be suitably modified with correction factors in order to correctly model such models as well. The new model thus derived is normally defined as "ωFQFμ" .

[0073] The "ωFQFμ" model assigns to every atom of the substrate a complex dipole in addition to a complex charge. In this case, the value of both the charges and dipoles is modified as a function of the interaction with the other charges and dipoles of the substrate and with the external radiation field. The publication “T. Giovannini, L. Bonatti, P. Lafiosca, L. Nicoli, M. Castagnola, P. Grobas lllobre, S. Corni, C. Cappelli, submitted, 2022, ACS Photonics, 2022, 9, 3025” describes such a case.

[0074] The choice of one model rather than another is determined solely by the physics of the material of the substrate under analysis.

[0075] Therefore, the "ωFQ" model and the variants thereof are advantageously capable of potentially dealing with all the substrates having a plasmonic response, irrespective of the chemical nature of the same.

[0076] The final equation for obtaining the charges and dipoles is the following: wherein Aq, Aq, Tqqand Tqqrespectively represent the charge-charge interaction (similarly to before), the charge-dipole interaction, the dipole-charge interaction and the dipole-dipole interaction among the various atoms making up the substrate.

[0077] The term z'(ω) contains the contribution solely of interband transitions to the polarizability of the system, whilst z(ω) (similarly to before) contains the terms of the classic conductivity of the Drude type, dependent on the frequency of the external field. The term Eextrepresents the contribution of the external electric field on the dipoles of the substrate, whilst the term R contains the contribution of the electric potential associated with the external field on the charges of the system.

[0078] The simultaneous solution of the system (in this case composed of the quadruple of the previous linear equations) provides the set of values of the charges q and of the dipoles μ present on every atom of the system, at a given frequency of the external field incident on the system itself.

[0079] By using the "ωFQ" model or, depending on the substrate, the "ωFQFμ" variant thereof, the process of the invention is advantageously capable of reproducing the absorption spectrum of a generic plasmonic substrate. In particular, if, with the "ωFQ" model or the "ωFQFμ" variant thereof, the absorption spectrum of the geometry selected during the step of morphologically and conformationally characterising a specific substrate does not show any plasmonic response, the process envisages modifying the morphological or conformational characteristics of said substrate or, ultimately, selecting a further substrate.

[0080] Following the step of selecting the analyte, selecting at least one substrate and the associated characterisation steps, the process comprises performing a mapping of a potential of interaction between each conformer of the analyte and each region and / or each atom of each substrate so as to identify, for each region and / or each atom of each substrate, a value of interaction with the analyte.

[0081] Specifically, the process in accordance with the invention comprises calculating the potential of interaction between the analyte and substrate for each relative position between the conformer of the analyte and substrate so as to evaluate the intensity of interaction and, therefore, the capacity of adsorption of the analyte onto the substrate. Preferably, the analyte is in a condensed phase, either homogeneous or inhomogeneous.

[0082] In accordance with one aspect of the invention, the step of performing a mapping of an interaction potential is achieved using electrostatic calculation techniques and, where necessary, by introducing non-electrostatic effects such as dispersive and repulsive interactions.

[0083] Preferably, to perform the mapping, one can use docking approaches known in the literature in other contexts, which allow the energy of interaction between analyte and substrate to be mapped for each reciprocal disposition between the analyte, substrate, and solvent.

[0084] In accordance with another aspect of the invention, the step of performing a mapping of an interaction potential comprises a “scoring” sub-step wherein, for each substrate, a hierarchical interaction scale is defined for each region and / or each atom of a same substrate. In particular, the hierarchical scale is configured to assign to each region and / or each atom of a same substrate a numerical index representative of the probability and / or force of interaction between the analyte and the substrate.

[0085] Mapping the interaction potential over the entire surface of the substrate allows multiple values of analyte / substrate interaction to be obtained.

[0086] By normalising these values, it is possible to establish a hierarchical scale which, for example, extends between a minimum value of “1 ”, representative of the region and / or atom in which the electrostatic interaction with the analyte (and thus the adsorption thereof) is minimal, and a maximum value of “10”, representative of the region and / or atom in which the interaction with the analyte is maximum.

[0087] Once the corresponding value of the hierarchical scale has been assigned to each region and / or each atom and a value of minimum interaction has been established, it is possible to divide the regions and / or atoms of the same substrate into a set with greater interaction and a set with less interaction and thus minimally interesting, as the possibility of absorption of the analyte is reduced or excessively low.

[0088] In contrast, the regions and / or atoms included in the set with greater interaction will be used for the subsequent mapping step, as they ensure a higher probability of adsorption of the analyte.

[0089] Advantageously, the definition of a hierarchical scale and thus the definition of a set with greater interaction allows for reducing the number of regions and / or atoms that need to be considered for the next part of the process, i.e. for the step of calculating the spectral signal, which is more costly from a computational standpoint, as we shall see below. Following the selection and characterisation of the analyte and substrate considered separately, the process comprises a step of calculating the spectral signal of the analyte adsorbed onto each substrate for at least one atom of each substrate.

[0090] In other words, this spectral calculation step is performed in order to model the complex analyte-substrate system. Both the analyte (and any solvent, if present) and the substrate, in fact, reciprocally alter their states due to their interaction as a result of the adsorption of the analyte onto the substrate.

[0091] In this step, therefore, it is envisaged to perform a new morphological and conformational characterisation of the analyte and substrate, but in this case by considering their interaction.

[0092] Therefore, the effect of the plasmonic excitation of the substrate on the spectral signal of the analyte is preferably modelled with the inclusion of a coupling term, preferably at an electrostatic level.

[0093] In accordance with a preferred aspect of the invention, the step of calculating the spectral signal is performed after the mapping step so as to calculate the spectral signal solely for the regions and / or atoms of the substrate included in the set with greater interaction.

[0094] The calculable spectral signal of the analyte may be tied to the electric and / or magnetic and / or nuclear response or any combination thereof.

[0095] In accordance with another aspect of the invention, in order to obtain the spectral properties of interaction between the analyte and the substrate, the step of calculating the spectral signal comprises: determining a ground-state electronic density of the analyte adsorbed onto each substrate; determining the density of the analyte adsorbed onto each substrate, in particular as a function of the modifications induced by the external field.

[0096] The model used for the spectral calculation is preferably defined at a quantum- mechanical and molecular mechanical level, wherein the analyte is described at a quantum-mechanical level, whereas the substrate is treated at a classic level by means of the "ωFQ" model.

[0097] In particular, a first energy factor making reference to the analyte adsorbed onto the substrate is determined with a quantum-mechanical analytical technique, preferably using the density functional theory “DFT”, a second energy factor making reference to the substrate onto which the analyte is adsorbed is determined with an atomistic model based on concepts of classical physics, preferably the "ωFQ" model, suitably adapted for the ground state or for the treatment of the linear response, and a third energy factor making reference to the interaction between the substrate and the analyte is determined with electrostatic calculation techniques.

[0098] Advantageously, the quantum-mechanical density functional theory “DFT” enables an accurate description of the electron structure of the analyte with a modest computational cost.

[0099] Advantageously, the time-dependent “TD-DFT” formulation likewise enables treatment of the linear response of an analyte at an excellent cost-accuracy compromise.

[0100] For the purpose of performing the calculation of the ground-state electronic density, the complex analyte / substrate system must be considered in the absence of the external radiation field. To this end, it will thus be necessary to consider the condition ω=0.

[0101] Therefore, the expression related to the energy of the complex analyte / substrate system is: ε = EQM+ EQM / MM+ EMMwhere the term EQMrefers to the energy associated with the analyte (represented with the letters “QM” in the present formulas) and calculated with quantum-mechanical techniques, the term EMMrefers to the energy associated with the substrate (represented with the letters “MM” in the present formulas) and preferably calculated with the “FQ” model (equivalent to "ωFQ" in the condition where ω=0 ) or a variant thereof depending on the type of substrate, and the term EQM / MMrefers to the energy associated with the analyte / substrate interaction, which is preferably calculated at the electrostatic level.

[0102] If the substrate is treated with the "ωFQ" model, given the condition ω=0 , the term EQM / MMwill consider the corresponding static force field “FQ” and can thus be rewritten as follows:

[0103] In contrast, if the substrate is treated with the "ωFQFμ" model, again using the condition ω=0 and the corresponding static force field “FQFμ ”, the term EQM / MMcan be rewritten as follows:

[0104] In both reformulations of the term EQM / MM, the term N is the number of atoms of the substrate and, therefore, the variable “i” ranges from 1 to N.

[0105] Preferably, the charges qiand the dipoles μipresent in the previous equations can be determined by following approaches described, for example, in the publication “T. Giovannini, F. Egidi, C. Cappelli. Chem. Soc. Rev., 2020, 49, 5664-5677” and other bibliographic references cited therein.

[0106] Preferably, the determination of the spectral properties of the analyte exploits the equations of the linear response time-dependent density function theory “LR-TDDFT”, founded on the concepts of the time-dependent version of DFT, also called “TD-DFT”. According to this theory, the first-order electronic density can be written on the basis of the molecular orbitals occupied in the ground state (GS) and the unoccupied onesΦαobtainable with the solution of the ground state equations as:

[0107] The determination of the terms must consider the fact that the analyte is subject to an external potential at a certain frequency, describable with the following equation:

[0108] Vpert(r, ω) = Vext(r, ω) + Vloc(r, ω)

[0109] In order to model the interaction with the substrate, it is useful also to add, to the potential generated by the external radiation field Vext(r, ω) at a certain frequency ω , the local field Vloc(r, ω) generated by the complex charges of the substrate (or by the complex dipoles and charges, in the event of the presence of interband excitations), both obtained using the "ωFQ" model (or, respectively, the "ωFQFμ" model) considered at the aforesaid frequency value ω.

[0110] In the case of treatment with the "ωFQ" model, the local potential Vloc(r, ω) can be rewritten as follows:

[0111] In the case of treatment with the model "ωFQFμ" , by contrast, the local potential Vloc(r, ω) can be rewritten as follows:

[0112] Considering once again the solution procedures of the linear response density function theory, it is possible to write a set of coupled equations which, once solved, determine the electronic density (preferably approximated to the first order) of the analyte in interaction with the external field generated during spectroscopic analysis and with the field induced by the plasmon of the substrate.

[0113] In particular, the system of linear equations to be solved is the following:

[0114] Where ω the various terms of the previous equation are defined as: where ω is the frequency of the external field generated for the performance of the spectroscopic analysis, Γ is the lifetime of the excited state of the analyte, εαe εiare the energies of the ground state orbitals, respectively unoccupied and occupied.

[0115] In addition, the equation shown here below makes it possible to obtain the perturbed first-order electronic density, from which the spectral properties of the analyte can be obtained by following the calculation techniques of quantum chemistry

[0116] In other words, the step of calculating the spectral signal of the analyte adsorbed onto the substrate is divided into two parts, more precisely into two different calculation sub- steps.

[0117] In the first part the ground state (GS) energy of the complex analyte / substrate system is calculated. In the second part, by contrast, starting from the result of the first, the linear response of the same complex analyte / substrate system is calculated. Obviously, such calculations are repeated for each selected point of each selected substrate and thus for each respective complex analyte / substrate system.

[0118] In the second calculation step, the linear response density function theory “LR-TDDFT” is preferably used, since, in the presence of an external field suitable for defining an interaction potential, this theory is capable of modelling the response of the analyte adsorbed onto the substrate to the external field with precision and with a relatively low computational effort. With the appropriate corrections, given the presence of the plasmonic substrate, the aforesaid theory is capable of providing the molecular response of the analyte to the external field, which is amplified due to the plasmonic characteristics of the complex system.

[0119] Advantageously, the polarizability is closely correlated to the linear response of the external field generated and, moreover, the polarizability derivative with respect to the normal modes / Cartesian coordinates makes it possible to calculate the spectral signal and thus to construct the spectral graphs of interest to be verified with experimental measurements (as will also be illustrated below).

[0120] Preferably, the polarizability derivative with respect to the normal modes / Cartesian coordinates is numerically determined, that is, by constructing small displacements along the normal modes / Cartesian coordinates, calculating the polarizability for every new structure and, finally, estimating the derivative with numerical methods based on the finite differences. Therefore, using this procedure, the calculation of polarizability must be repeated for all the structures created, which are directly proportional to the number of normal modes / Cartesian coordinates of the analyte.

[0121] Following the determination of the desired spectral signal, the process envisages carrying out the following steps: analytically evaluating an enhancement factor of each spectral signal calculated; comparing each enhancement factor with a corresponding and predefined minimum threshold value so as to evaluate whether the associated nanostructured plasmonic sensor is configured to perform detection of the analyte with single-molecule sensitivity; generating an output signal identifying at least one design for a nanostructured plasmonic sensor capable of detection with single-molecule sensitivity, wherein at least one enhancement factor is greater than said minimum threshold value, said at least one enhancement factor referring to an atom of a corresponding substrate for which the value of the interaction potential is greater than a value of minimum interaction. In other words, if the enhancement factor derived from the spectral signal of a complex analyte / substrate system is greater than a corresponding minimum threshold value, the associated nanostructured plasmonic sensor will be capable of detecting the analyte adsorbed onto the substrate thereof with single-molecule sensitivity.

[0122] The procedure of calculating the enhancement factor and comparing it with the minimum threshold value is advantageously carried out for each spectral signal calculated, i.e. for each selected substrate.

[0123] Preferably, for each selected substrate solely the atoms belonging to the set with maximum interaction are considered, i.e. atoms associated with an index of the hierarchical scale greater than a minimum reference index.

[0124] In fact, as better explained in the examples below, the zone of the substrate in which the atoms having a greater value of interaction with the analyte (i.e. a higher probability of adsorption of the analyte itself) are located are not necessarily also the zones of the substrate in which the spectral signals show the highest enhancement factor, i.e. an enhancement factor useful for detecting the analyte adsorbed with certainty and with single-molecule sensitivity.

[0125] At this point, for each zone of each substrate for which the spectral signals have been calculated (and, if the related step is performed, depending on the hierarchical index that such zones have), a researcher will be able to evaluate which selected substrate is the best for producing the nanostructured plasmonic sensor.

[0126] The zone which has the highest hierarchical index and also the best spectral signal should probably be evaluated as the best choice on which to base the construction of the nanostructured plasmonic sensor. On the associated substrate, in fact, the aforesaid zone is the one in which the analyte has the greatest probability of being adsorbed and / or bound and in which the spectral signal is best (i.e. particularly detectable).

[0127] If the most amplified spectral signal (i.e. the best one) is not located in the zone with the highest hierarchical index, the choice of the best substrate for producing the nanostructured plasmonic sensor will be subjective for each researcher. On the one hand, the zone with the best (or in any case high) hierarchical index might be preferred, as it assures the presence of an adsorbed analyte. On the other hand, one might prefer a spectral signal that is particularly amplified, and thus visible, but which requires, however, a more complex operation for the detection of the analyte, for example with the use of a nanometric probe placed in the zone of the substrate in which the analyte is adsorbed.

[0128] Alternatively, if different zones of a substrate, and / or different morphologies with which to produce a substrate and / or different materials for producing a substrate have been evaluated and considered, it is possible that the best substrate for detecting an analyte with single-molecule sensitivity will be determined by applying an analysis algorithm, for example analysis software. The analysis algorithm is preferably structured so as to compare the different mapped interaction potentials and the different enhancement factors obtained, in order to select the substrate with a design capable of assuring the best compromise between the value of analyte / substrate interaction (that is, in other words, the capacity of the analyte to bind to a specific atom / group of atoms of the substrate) and the best enhancement factor (that is, in other words, the amplification that the spectral signal obtains and thus the capacity to detect the adsorbed analyte). In particular, such algorithms evaluate the factor of the increase in the electric field produced by the plasmon in proximity to the surface of the nanostructured substrate in order to obtain a quantification of the enhancement of the spectroscopic signal of the analyte in its interaction with the aforesaid nanostructure of the substrate.

[0129] In accordance with a further aspect of the invention, the process comprises a further step of selecting at least one substrate of the nanostructured plasmonic sensor in order to select at least one new further substrate having a two- or three-dimensional nanostructured morphology and / or a material with plasmonic characteristics that are different compared to at least one of the previous substrates, when each enhancement factor is lower than the corresponding predefined minimum threshold value.

[0130] In other words, if none of the spectral signals show an acceptable enhancement factor (i.e. one that is higher than the corresponding predefined minimum threshold), the process comprises selecting at least one new substrate, based on the same analyte to be detected, to be characterised according to what was explained previously.

[0131] In still other words, if it is deemed necessary and / or useful to select a new nanostructured plasmonic substrate, it will be possible to repeat the entire process (with the exclusion of the first steps regarding the analyte, i.e. the selection and characterisation of the analyte) in order to study the spectral properties of the new complex analyte / substrate system, both with quantum-mechanical calculation techniques (for the analyte) and with calculation techniques related to classic physics (as regards the substrate).

[0132] The first steps regarding the analyte need not be repeated, since the previously obtained results remain valid because they were calculated in reference solely to the analyte, thus without considering any interaction with the substrate.

[0133] In the light of the detailed description just laid out, some case studies are presented below as examples of the in silico process for identifying the design of a nanostructured plasmonic sensor configured to detect at least one molecule of an analyte of interest capable of being adsorbed onto a substrate of said sensor. The aforesaid example case studies are also an object of protection of the present invention, to the same extent as the abovementioned process.

[0134] Example 1 - Analyte adsorbed onto diverse regions of a same substrate

[0135] Example 1 addresses the adsorption of a molecule of pyridine (i.e. the analyte of interest) in four different regions (identified by the letters “a”, “b”, “c” and “d” in figure 2) of a silver nanoparticle having a truncated cuboctahedral shape (i.e. the substrate).

[0136] Figure 2 also shows the four respective Raman spectra, each of the which is associated with a region of adsorption of the analyte. The Raman spectra are also identified by the letters “a”, “b”, “c” and “d” so as to enable a correct association between the Raman spectrum and the region of adsorption of the analyte onto the substrate. In addition, for each Raman spectrum the corresponding parameters Maximum Enhancement Field (MEF) and Average Enhancement Field (AEF) are also shown. The aforesaid parameters are indicators capable of quantifying the efficiency of the specific analyte / substrate geometry. They are preferably calculated on the basis of the Raman spectrum obtained in the gas phase. In particular, a calculation is made of the ratio between the Raman intensity of a specific band of the molecular system adsorbed onto the plasmonic substrate and the corresponding intensity calculated for the molecular system in the gas phase. In other words, the parameter “MEF” indicates the maximum increase obtained considering all the individual bands of the simulated spectrum (in figure 2 and in the following figures, the band that shows the maximum value of “MEF” is indicated by an asterisk), whilst the parameter “AEF” refers to the average enhancement field over all the Raman bands.

[0137] Analysing figure 2, in particular the Raman spectra, the region of adsorption in which it is most likely to obtain the maximum enhancement is the one in which the pyridine is adsorbed onto a vertex of the nanoparticle (i.e. the region indicated with the letter “a”). In particular, for that adsorption region, both of the abovementioned parameters show to be greater, by about two orders of magnitude, than the corresponding parameters of the other adsorption regions considered.

[0138] Example 2 - Analyte adsorbed onto different morphologies of a same substrate Example 2 addresses the adsorption of a molecule of pyridine (i.e. the analyte of interest) at a vertex of a gold nanoparticle with a radius of about 40 A (i.e. the substrate) as the morphology of the substrate itself varies.

[0139] Figure 3 shows three different morphologies of the substrate and the respective Raman spectra calculated considering, as mentioned, the adsorption of the analyte at the vertex at the top.

[0140] The morphologies considered for the present example 2 are: truncated cuboctahedron (“cTO”), icosahedron (“Ih”) and ino-dodecahedron (“l-Dh”) identified, respectively, by the letters “a”, “b” and “c”.

[0141] Through the analysis of the aforesaid Raman spectra and the respective parameters “AEF” and “MEF” (shown in each graph in figure 3), one deduces that the substrate morphology influences both of the indices considered. In particular, the geometry that provides the highest average enhancement field (“AEF”) and maximum enhancement field (“MEF”) is the truncated cuboctahedron “cTO” identified by the letter “a”.

[0142] Example 3 - Analyte adsorbed onto substrates having a different chemical nature

[0143] Example 3 addresses the adsorption of a molecule of pyridine (i.e. the analyte of interest) at the vertex of a silver nanoparticle, a gold nanoparticle and, finally, on an internal portion of a graphene disk. The aforesaid metal nanoparticles and graphene disk are different substrates having different chemical characteristics, but they are all characterised by a plasmonic behaviour.

[0144] In particular, in the first two cases the substrate has an icosahedral shape and a radius of about 40 A, whereas in the case of graphene an ideal disk (i.e. free of surface defects) having a radius of 40 A was considered.

[0145] Figure 4 illustrates the aforesaid substrates onto which the analyte is adsorbed and the respective calculated Raman spectra accompanied by the corresponding parameters “AEF” and “MEF”. More precisely, the silver nanoparticle is shown at the top, the gold nanoparticle is shown in the middle, while the graphene disk is shown at the bottom.

[0146] From the graphs and the respective parameters shown in figure 4, one deduces that not only the morphology, but also the chemical nature of a substrate determines the increase in the electric field near the surface thereof (note the calculated values of the parameters “AEF” and “MEF”) and the spectral fingerprint of the molecule under analysis. In fact, for each substrate considered it is possible to identify different Raman peaks, each of the shows a different intensity as a function of the substrate itself.

[0147] In conclusion, the in silico process of the present invention is advantageously capable of enabling a distinction between the Raman spectra generated by a given analyte, in this example pyridine, adsorbed onto different substrates.

[0148] Example 4 - Different molecular systems adsorbed onto the same substrate

[0149] Example 4 addresses the adsorption of a molecule of pyridine (identified by the letters “PY”) and a molecule of methotrexate (identified by the letters “MTX”) on a graphene disk having a diameter of 14 nm.

[0150] Figure 5 shows the respective Raman spectra of the molecules of pyridine and methotrexate adsorbed onto the graphene substrate, on the left and right respectively. In particular, it is possible to note two different spectral fingerprints for the two (different) molecules adsorbed onto a same substrate. Pyridine shows the highest average enhancement field (AEF), whereas methotrexate shows the highest maximum enhancement field (MEF).

[0151] In addition, unlike the molecule of pyridine, the molecule of methotrexate possesses various degrees of torsional freedom. Consequently, it is necessary to adequately sample the conformational space to search for the conformers which, in their interaction with the given substrate (in this case the graphene disk), are more stable. Advantageously, for the purpose of considering the aforesaid possibility of the presence of different conformers and, therefore, identifying the most stable ones for the given substrate of interest, the process in accordance with the invention envisages being able to perform advanced molecular dynamics calculations and a consequent clustering process.

[0152] Figure 6 shows, on the left, two conformers of methotrexate, superimposed on each other in order to highlight the stereometric differences. In addition, figure 6 also shows the respective Raman spectra and the respective parameters “AEF” and “MEF” calculated for the aforesaid conformers. More precisely, the Raman spectrum of the conformer identified as “MTX1 / GD” is shown at the top and the Raman spectrum of the conformer identified as “MTX2 / GD” is shown in the middle. Shown at the bottom is the average Raman spectrum calculated by weighting the spectrum of each conformer based on the number of occurrences of that given conformer during the molecular dynamics process.

[0153] In conclusion, in the present example 4 the in silico process of the present invention is advantageously capable of enabling a distinction between different analytes adsorbed onto the same substrate. Even more advantageously, the present example 4 also allows discriminating between different conformers of the same analyte adsorbed onto a substrate.

[0154] Example 5 - Comparison with experimental data

[0155] Example 5 shows a comparison between the data calculated using the aforesaid in silico process of the present invention and the corresponding experimental spectrum. In particular, as shown in figure 7, the molecule of pyridine adsorbed onto a silver nanoparticle is considered for this purpose.

[0156] Like example 1 , example 5 shows a molecule of pyridine adsorbed in four different regions (identified by the letters “a”, “b”, “c” and “d” in figure 2) of a silver nanoparticle having a truncated cuboctahedral shape.

[0157] The experimental spectrum is reproduced from J. Chem. Phys., 1988, 88, 7942-7951 , and was measured at 514 nm by adsorbing pyridine onto an electrode made of silver. The comparison between the experimental data and the computationally calculated data shows high agreement between the spectra and, in particular, for the configurations in which the molecule of pyridine (i.e. the analyte) is adsorbed onto the faces of the truncated cuboctahedron of the silver nanoparticle.

[0158] Finally, the invention also relates to a computer program capable of implementing the various steps of the previously described process, in particular, the computational calculation steps, in order to calculate the various Raman spectra (or analogous spectra of interest) of the analyte / substrate system of interest.

[0159] In this manner, the use of said computer program is advantageously capable of providing a user, for example a researcher, with the nanostructured plasmonic sensor that is most suitable for detecting a specific analyte, preferably with single-molecule sensitivity.

[0160] In other words, the in silico process and the computer program are advantageously capable of signalling which nanostructured plasmonic substrate performs best in the detection, preferably single-molecule detection, of a specific analyte.

[0161] The process of identifying the design of a nanostructured plasmonic sensor also allows a further practical advantage to be obtained: it can reduce the calculation times and power necessary to obtain the desired output compared to the techniques of the prior art.

[0162] A first example for showing the aforesaid savings in calculation times involves calculating the SERS spectroscopic signal of an adsorbed molecule of pyridine at a distance of 3.5 angstroms on a graphene sheet of small dimensions (having a diameter of approximately 1 .7 nanometres and number of atoms equal to 96).

[0163] In accordance with the process according to the present invention, the SERS signal can be calculated using the QM / ωFQ method, whereby the pyridine is treated at a QM level (B3LYP / DZP), whilst the graphene sheet can be described with the classic ωFQ model.

[0164] The processes in accordance with the prior art, by contrast, envisage treating the entire system at the QM level (B3LYP / DZP). In both cases the external frequency was set at 532 nm, i.e. the wavelength of the laser beam usually used experimentally for the measurement of SERS spectra, while a lifetime parameter of 0.10 a.u. was also imposed.

[0165] The calculation, both for the process in accordance with the invention and for the known processes of the prior art, was performed thanks to the following hardware architecture: Intel(R) Xeon(R) Gold 5120 CPU @ 2.20GHz, 28 cores, 128 GB of RAM. The computation times for calculating the SERS signal in the former case (i.e. in accordance with the invention) were in the order of several tens of minutes, more precisely about 30 minutes, whereas in the latter case (the process in accordance with the prior art) took several days, more precisely a good six days.

[0166] The savings in terms of computing power (i.e. of computational cost) mainly derive from the savings obtainable thanks to the treatment of the plasmonic substrate. A second example of calculation involves, by contrast, considering metal nanoparticles with an icosahedral morphology with growing dimensions, for example with a radius ranging between 8.23 angstroms (and a number of atoms equal to 147) and 13.72 angstroms (and a number of atoms equal to 561 ).

[0167] In this second example, solely the plasmonic response of the nanostructured substrate was evaluated, in other words in the absence of an analyte.

[0168] The calculation was thus performed with the process in accordance with the invention, i.e. with the model called ωFQFμ , in order to be able to compare the results with those obtained in accordance with the known processes of the prior art, i.e. with a completely quantum-based treatment of the nanostructured substrate.

[0169] In this second case as well, the resulting computation times are significantly different considering the two different approaches.

[0170] On the one hand, the process in accordance with the invention makes it possible to maintain computation times ranging from a few seconds (in the case of nanoparticles of a smaller size) to a few minutes, about 7 minutes (in the case of nanoparticles of a larger size).

[0171] In contrast, the calculation approach based on a completely quantum-mechanical treatment (i.e. in accordance with the processes of the prior art) provided results with computation times ranging from a thousand hours or so, about 1 152 hours, and tens of thousands of hours, about 21 ,504 hours (the present computation times are CPU processing times, i.e. “CPU hours”).

[0172] More precisely, the computation times obtained with a completely quantum- mechanical approach have been retrieved from the literature “Kuisma et al., Phys. Rev. B, 2015, 91 , 1 15431 ” and were obtained thanks to a high-performance computing centre, using the following number of processors: 64 (Ag147), 256 (Ag309), 512 (Ag561 ).

[0173] In contrast, the computation times with the application of the ωFQFμ model were achieved using the following architecture: Intel(R) Core(TM) i7-1 185G7 @3.00GHz, 32Gb RAM, 4 core (Dell XPS 13” portable computer of 2021 ).

[0174] In conclusion, the data reported above show the significant computational savings brought by the application of the process in accordance with the invention, i.e. with the application of a mixed approach for treating the substrate and the analyte.

[0175] In addition, it should be underscored that the advantage is not limited to a reduction in computation times, but also extends to the possibility of using personal computers definable as commercial, easily to be found on the market and, therefore, of not needing high-performance computing centres (which, moreover, enormously limit the size of the systems that can be studied).

[0176] The present invention also relates to a method for producing a nanostructured plasmonic sensor configured to detect at least one molecule of an analyte of interest. In particular, the aforesaid method envisages applying the following steps: carrying out the previously described in silico process so as to obtain at least one output signal identifying a design applicable to a substrate (i.e. the morphology and plasmonic material with which to produce it) whereby it is possible, for at least one atom of the substrate having an acceptable value of potential interaction - greater than a value of minimum interaction - to obtain detection with single-molecule sensitivity; selecting at least one substrate, among the substrates identified in the aforesaid output signal, which has a two- or three-dimensional nanostructured morphology and can be produced with a material having plasmonic characteristics and, as mentioned, for which at least one enhancement factor greater than a predefined minimum threshold value is present (for at least one atom thereof of interaction with the analyte); producing, with one or more nanomanufacturing techniques, a substrate having the aforesaid design, i.e. with the aforesaid two- or three-dimensional nanostructured morphology and the aforesaid material having plasmonic characteristics, so that the associated nanostructured plasmonic sensor has a single-molecule detection sensitivity to the analyte.

[0177] Nanomanufacturing techniques differ and each of them can be advantageously used to produce a specific substrate design, i.e. for a specific morphology and / or for modelling a specific material.

[0178] The aforesaid method of production can thus advantageously make it possible to produce (using known nanomanufacturing techniques) a nanostructured plasmonic sensor having the correct design useful for ensuring a single-molecule detection sensitivity to the analyte of interest it is intended to detect and which, in an operating step, is adsorbed onto one or more atoms of the aforesaid substrate.

[0179] In other words, by means of known spectroscopy techniques, for example surface- enhanced Raman spectroscopy (SERS), the aforesaid nanostructured plasmonic sensor (produced with the design characteristics obtained thanks to the aforesaid in silico process) is advantageously capable of enabling detection of the analyte of interest with single-molecule precision.

[0180] In still other words, the aforesaid sensor ensures that the spectroscopy technique selected by the researcher is able to detect even a single molecule adsorbed onto the surface of the nanostructured plasmonic substrate. More precisely, that sensor ensures a precision of analyte detection in the zones of the substrate (i.e. atoms and / or groups of atoms) for which an analyte / substrate interaction potential is present that is greater than a value of minimum interaction (i.e. the probability of interaction is sufficiently high, comparable to the standard values of energies of molecular adsorption onto the selected substrate, i.e. in the order of kcal / mol) and, in addition, for which an enhancement factor is present that is greater than a pre-established minimum threshold value (i.e. the amplification of the signal is such as to exceed the tolerance value of the spectroscopy instrument used and thus to render the same easily detectable signal).

Claims

CLAIMS1. An in silico process for identifying the design of a nanostructured plasmonic sensor configured to detect at least one molecule of an analyte of interest capable of being adsorbed onto a substrate of said sensor, said process comprising the following operating steps: selecting an analyte to be detected by means of the nanostructured plasmonic sensor; morphologically and conformationally characterising the analyte with computational techniques; selecting at least one substrate of the nanostructured plasmonic sensor having a two- or three-dimensional nanostructured morphology and produced with a material having plasmonic characteristics; morphologically and conformationally characterising each substrate with computational techniques; performing a mapping of a potential of interaction between the analyte and each atom of each substrate so as to identify, for each atom of each substrate, a value of interaction with the analyte; calculating the spectral signal of the analyte adsorbed onto each substrate for at least one atom of each substrate; analytically evaluating an enhancement factor of each spectral signal calculated; comparing each enhancement factor with a corresponding predefined minimum threshold value so as to evaluate whether the associated nanostructured plasmonic sensor is configured to perform detection of the analyte with single-molecule sensitivity; generating an output signal identifying at least one design for a nanostructured plasmonic sensor capable of detection with single-molecule sensitivity, wherein at least one enhancement factor is greater than said minimum threshold value, said at least one enhancement factor referring to an atom of a corresponding substrate for which the value of the interaction potential is greater than a value of minimum interaction.

2. The process according to claim 1 , wherein said step of morphologically and conformationally characterising the analyte is performed with a quantum-mechanicalanalytical technique.

3. The process according to claim 2, wherein said step of morphologically and conformationally characterising the analyte is performed with density functional theory (DFT).

4. The process according to any preceding claim, wherein said step of morphologically and conformationally characterising each substrate is performed with an atomistic model based on concepts of classical physics.

5. The process according to claim 4, wherein said atomistic model is the "ωFQ" model, wherein said "ωFQ" model can be modified with specific correction parameters as a function of the two- or three-dimensional morphology of the substrate and / or the presence of interband interactions of the substrate material.

6. The process according to any preceding claim, wherein said step of performing a mapping of an interaction potential is performed with electrostatic calculation techniques, optionally modified to take account of dispersion and repulsion effects.

7. The process according to any preceding claim, wherein said step of performing a mapping of an interaction potential comprises a scoring sub-step wherein, for each substrate, a hierarchical interaction scale is defined for each region and / or each atom of a same substrate, said hierarchical scale being configured to assign to each atom of a same substrate a numerical index representative of the probability and / or force of interaction between the analyte and the substrate.

8. The process according to any preceding claim, wherein said step of calculating the spectral signal comprises: determining an electronic density of ground state of the analyte adsorbed onto each substrate; determining an electronic density of the analyte adsorbed onto each substrate modified by the external field, wherein a first energy factor making reference to the analyte adsorbed onto the substrate is determined with a quantum-mechanical analytical technique, preferably with the density functional theory (DFT), a second energy factor making reference to the substrate onto which the analyte is adsorbed is determined with an atomistic model based on concepts of classical physics, preferably the "ωFQ" linear model suitably adapted for the ground state or for the treatment of the linear response, and a thirdenergy factor making reference to the interaction between the substrate and the analyte is determined with electrostatic calculation techniques.

9. The process according to any preceding claim, wherein said step of morphologically and conformationally characterising the analyte comprises using methods of classical physics, of the continuous or atomistic type, to describe the interaction between the analyte and any solvent in which said analyte is immersed.

10. The process according to any preceding claim, wherein each possible orientation of the analyte adsorbed onto the substrate is sampled by means of molecular dynamics techniques.

11. The process according to any preceding claim, comprising a further step of selecting at least one substrate of the nanostructured plasmonic sensor in order to select at least one new further substrate having a two- or three-dimensional nanostructured morphology and / or a material with plasmonic characteristics that are different compared to at least one of the previous substrates, when each enhancement factor is lower than said corresponding predefined minimum threshold value.

12. A method for producing a nanostructured plasmonic sensor configured to detect at least one molecule of an analyte of interest, comprising the steps of: carrying out the in silico process in accordance with any one of claims 1 to 1 1 ; selecting at least one substrate with a two- or three-dimensional nanostructured morphology and made of a material with plasmonic characteristics having at least one enhancement factor greater than a predefined minimum threshold value; producing, with one or more nanomanufacturing techniques, said at least one substrate with said two- or three-dimensional nanostructured morphology and with the material having plasmonic characteristics so that the associated nanostructured plasmonic sensor has a single-molecule detection sensitivity to the analyte.

Images