EP4689876A1

EP4689876A1 - Physically unclonable function (puf) label, system and methods

Info

Publication number: EP4689876A1
Application number: EP24716328.0A
Authority: EP
Inventors: Tim Liedl; Ulrich Rührmair; Mihir Dass; Gregor POSNJAK; Lena RAAB
Original assignee: Ludwig Maximilians Universitaet Muenchen LMU
Current assignee: Ludwig Maximilians Universitaet Muenchen LMU
Priority date: 2023-03-31
Filing date: 2024-03-28
Publication date: 2026-02-11
Also published as: WO2024200684A1

Abstract

The invention relates to nanomaterial based PUF labels, system and methods. The physically unclonable function (PUF) label (1), comprise a substrate (2) having a surface (3), a plurality of binding patches (4) arranged in a pattern (10; 11) on the surface, each binding patch (4) carrying a random arrangement of one or more nanoparticles (7; 7a; 7b; 7c; 7d; 7e), the pattern (10; 11) being optically detectable as a pattern image of randomly colored patches, the color of each colored patch (4) being determined by the random arrangement of the one or more nanoparticles (7; 7a; 7b; 7c; 7d; 7e) of the respective binding patch (4).

Description

Physically unclonable function (PUF) label, system and methods

The present invention is related to a physically unclonable function (PUF) label, an anticounterfeit system and method using the PUF label and a method for the fabrication of the PUF label.

A Physically Unclonable Function (PUFs) is s a physical object that for a given input and conditions (challenge), provides a physically defined "digital fingerprint" output (response) that serves as a unique identifier (Wikipedia). PUFs can be used as anti-counterfeiting proofs for goods that are threatened by product piracy, ranging from drugs to fashion items. They can also be used for securing passports, banknotes, birth records, financial certificates, and the like. A PUF may be understood as an “object’s fingerprint”.

The concept of using PUF labels as an anti-counterfeit technology is generally known, the theory of physically unclonable functions is found in textbooks. A sub-class of PUF are based on nanomaterials.

A nanomaterial-based physically unclonable function (PUF) device, as for example described by patent US 10,056,905 B, is a type of security device that uses the unique physical characteristics of nanoscale materials to generate an authentication code. The device contains a unique pattern of nanoscale structures that are impossible to replicate accurately due to the inherent variability of the fabrication process at the nanoscale. The PUF device described in patent US10056905 includes a substrate with a pattern of nanoscale structures, such as nanowires or nanotubes, which are randomly oriented and have varying lengths and diameters. To use the device for authentication, a challenge signal is sent to the device, and the response signal generated by the PUF device is compared to a pre-stored reference signal. When a voltage is applied to the device, the nanomaterials exhibit unique electrical properties, which can be measured and used to generate a unique authentication code. This code can be used to verify the identity of the device or user, making it useful for secure applications such as cryptography, anticounterfeiting, and tamper detection. Because the response signal is based on the unique physical properties of the nanomaterials and is therefore unpredictable, the PUF device can be used to securely authenticate devices or individuals. The nanomaterial-based PUF device provides a secure and reliable way to authenticate electronic devices and ensure that they have not been tampered with or cloned. A nanomaterial-based PUF device provides a high level of security and can be used in specific applications, including in the Internet of Things, smart cards, and other valuable electronic devices that require authentication and secure communication. While the PUF device described in patent US10056905 is a promising technology for authentication and security applications, the cost and feasibility of manufacturing and integration in systems may not be suitable for low-cost products.

The technical problem underlying the present invention is to provide a PUF technology, which is based on low-cost parts and suitable for providing anti-counterfeit for low-cost products.

This object is attained by a PUF label according to independent claim 1 , the anti-counterfeit system of claim 10, the method of claim 14 of producing the PUF label and the anticounterfeit method of claim 15. Various preferred embodiments presented are particularly provided by the teachings of the dependent claims and the description.

The physically unclonable function (PUF) label according to the invention comprises a substrate having a surface, a plurality of binding patches arranged in a pattern on the surface, each binding patch carrying a random arrangement of one or more nanoparticles, the pattern being optically detectable as a pattern of randomly colored patches, the color of each colored patch being determined by the random arrangement of one or more nanoparticles of the respective binding patch.

The invention provides that the surface of each PUF label ever produced gives a completely individual optical "response" (a microscopic color pattern) to an optical "query" (illumination). The PUF label according to the invention is secure, low-cost, mass- producible and small. Any article may be provided with such a PUF label, including banknotes, documents, in particular identity documents and other documents requiring a highly secure individual labeling, including secure digital signatures, pharmaceutical containers and packages, etc. The PUF label may also be implemented as a part of authentication processes for a data transfer between instances, wherein such a process may require the interaction with human users or not require user interaction and be fully automated. Such a system, which is also provided by the present invention, includes an imaging apparatus and a computing device as a reading and evaluating device. Measurement and evaluation of the PUF label can be done with a low-cost imaging apparatus, e.g. containing a simple arrangement of lenses and illumination, and be carried out quickly, e.g. within a few seconds. A counterfeiter would need to replicate the PUF label structure of multiple materials to within a nanometer (or better) to outwit the anticounterfeit system (imaging apparatus + computing device, e.g. smartphone or smartphone + add-on lenses + installed app), i.e. to create an effective clone, which is assumed to be substantially impossible.

Preferably, one or more coupling molecules are provided in a layer on top of each binding patch, the one or more nanoparticles being bound to the coupling molecules, which are bound to the binding patch. Preferably, a coupling molecule has one or more first binding sites for binding one or more nanoparticles.

The coupling molecule may be an organic molecule, which preferably is capable of binding to the binding patch. The coupling molecule may have one or more functional groups, e.g. thiol-moieties, which respectively form a first binding site and are capable of binding to the nanoparticle.

Preferably, the coupling molecules are polynucleotide-based nanostructures, in particular DNA-origami, the polynucleotide-based nanostructures having one or more first binding sites for binding one or more nanoparticles.

Preferably, the nanoparticles are plasmonic nanoparticles, which are arranged such on top of the binding patch, in particular bound to the binding patch, that a plasmonic coupling (pc) of at least two of the plasmonic nanoparticles on top of a binding patch is achieved.

The color of each colored patch preferably is determined by the random arrangement of plasmonic nanoparticles of the respective binding patch, in particular by the color of the individual plasmonic nanoparticles and/or the plasmonic coupling between plasmonic nanoparticles of the same binding patch.

The surface of the PUF label, which comprises plasmonic nanoparticles showing a plasmonic coupling, may be referred to as “metasurface” or “plasmonic metasurface”, which may extend over a centimetre scale. Plasmonic nanoparticles, as generally known, interact strongly with light, concentrating far-field radiation into sub-wavelength volumes, giving rise to effects like near-field coupling. A random combination of nanoparticles may bind closely to each other on the binding patch, preferably coupled to the same by a coupling molecule. When the coupling molecules are polynucleotide-based nanostructures, the layer on top of each binding patch may be precisely constructed, in particular by predetermining the distance of first binding sites, which preferably is so close that a plasmonic coupling is achieved, but which preferably is large enough to avoid clustering of nanoparticles. Each nanoparticle patch then exhibits a specific response to incoming light. The scattered light from, e.g., a silver nanosphere is blue, from a gold nanosphere is green and from a gold nanorod is red. Particles that are close to each other (close means: capable of showing plasmonic coupling, the distance being typically not more than a diameter of the nanoparticle) can undergo a plasmonic coupling and thereby create new colors. The obtained metasurfaces are capable of scattering the light at each patch of the pattern with a certain spectrum I color.

The PUF label according to the invention is fabricated by controlling the geometry of the pattern and that fact that particles form a random arrangement on the binding patches, the colored pattern acting as an “object’s fingerprint”.

The PUF label according to a preferred embodiment of the invention is fabricated by controlling the pattern and that fact that particles come very close to each other due to the specific properties of the polynucleotide-based nanostructures. There is, however, no control of where the nanoparticles exactly attach on a binding patch and what and/or at which distance and/or orientation will be the neighbors of each nanoparticle. Therefore, the surface, or respectively the metasurface, obtained is completely random. Substantially nobody can reproduce such a surface, or respectively the metasurface, not even the producer of an original metasurface or PUF label. To reproduce any of these absolutely specific color responses of each chip, the counterfeiter would need to be able to place thousands or millions of nanoparticles with nanometer accuracy, which would take tremendous time and resources.

The plasmonic nanoparticles, preferably are metallic nanoparticles, i.e. nanoparticles containing one or more metals or consisting of one or more metals. The metal may be preferably Gold, Silver, Platinum or other metal. While using plasmonic nanoparticles is a preferred embodiment, the invention is not restricted to using nanoparticles or to employ plasmonic coupling for determining the colors of the patches. The nanoparticles may also be non-metallic, e.g. be semiconductor nanoparticles, polymer nanoparticles with fluorophores or nanoparticles with dyes. Semiconductor nanoparticles, e.g. CdSe nanoparticles, show a size-dependent emission spectrum and color. Such nanoparticles, in particular semiconductor nanocrystals, may be chemically functionalized for binding to the binding patches, suitable products are commercially available.

The term “polynucleotide-based nanostructure” (herein also: “nanostructure”) means, in particular, that the nanostructure comprises at least a portion being a polynucleotide or that the nanostructure is, substantially, formed by a polynucleotide.

The term “one or more first/second binding sites of a polynucleotide-based nanostructure” is to be understood in such a way that the number of first/second binding sites of the labeling nanostructure is predetermined by selection of a nucleotide sequence of the polynucleotide-based nanostructure, in particular before the polynucleotide-based nanostructure is obtained, e.g., from synthesis.

Thereby, the selection of each respective nucleotide sequence, which relates to a binding site, is made such that binding criteria for each respective first binding site are fulfilled in order to bind a detectable nanoparticle to the then obtained polynucleotide-based nanostructure, in particular one detectable label at each first binding site.

These binding criteria may be application-specific criteria. For example, the individual bonds should be chemically predefined. For example, to chemically suit a specific binding site reagent attached to a detectable surface of a nanoparticle functionalizing the nanoparticle for use as a detectable nanoparticle may be employed. The skilled person can check the binding of the detectable nanoparticles to the first binding sites of the nanostructure, for example, by means of conventional electron microscopy, for example after determining the polynucleotide sequences relevant for the binding site and synthesizing the nanostructure accordingly. He can also use the sensitivity of the detection signal to check whether it corresponds to the sensitivity derived from the number of binding sites by measuring known analyte concentrations. Because the binding sites, in particular, a number N of first binding sites are predefined, the binding sites can be arranged on top of the surface of the binding patch using the coupling molecules such that it is basically ensured that a detectable nanoparticle (e.g. a plasmonic nanoparticle) can bind to a first binding site. The maximum number of first binding sites essentially corresponds to the maximum number of detectable nanoparticles, which can be bound to the binding patch.

A nanostructure hereby relates to a polynucleotide-based structural unit, e.g., an architecture of two-dimensional or three-dimensional shapes from polynucleotides comprised of 2 or more, in particular up to 350, 1000 or even 10000 nucleotides and/or oligonucleotides that, in particular through base pair recognition, assemble into an object of defined shape and size, e.g., a DNA Origami as a particular instance of an object having a defined shape and size, whereas the desired shape and size is obtained from the folding route of a given scaffold and a respective set of staple sequences that can fulfill the folding. The shape is thus predetermined by the sequences of the nucleotides and/or oligonucleotides that, as a whole or in part, recognize each other and bind to each other to form the nanostructure. The size is limited by the length of the scaffold used for folding, i.e. , the number of nucleotides and/or polynucleotides used. In some instances, the shape may be described means an aspect ratio, e.g., for a rod-shaped nanostructure, the length : width : height ratio. The polynucleotide-based nanostructure preferably has a molecular mass, which preferably is chosen from the following preferred ranges: 100 kilodalton (kDa) to 100 megadalton (Mda), 250 kDa to 50 Mda, 500 kDa to 20 Mda, 1 to 10 Mda; another preferred range is between 4 to 6 Mda.

Polynucleotide-based structural units include but are not limited to scaffolded deoxyribonucleic acid (DNA) origami, scaffolded ribonucleic (RNA) origami, scaffolded hybrid DNA : RNA origami, scaffold free DNA single - stranded tile (DNA brick) systems, scaffold - free multi - stranded DNA tile systems or RNA tile systems, intramolecularly - folded single - stranded RNA or single - stranded DNA origami, or any DNA/RNA analogues, e.g., LNA, PNA or XNA. Nucleic acid analogues are compounds which are analogous, i.e., structurally similar, to naturally occurring RNA and DNA, used in medicine and in molecular biology research. Typically, the analogue nucleobases confer, among other things, different base pairing and base stacking properties. Examples include universal bases, which can pair with all four canonical bases, and phosphate-sugar backbone analogues such as PNA, which affect the properties of the chain, e.g., PNA can even form a triple helix. Nucleic acid analogues are also called Xeno Nucleic Acid. Artificial nucleic acids include peptide nucleic acid (PNA), Morpholino and locked nucleic acid (LNA), as well as glycol nucleic acid (GNA), threose nucleic acid (TNA) and hexitol nucleic acids (HNA). Each of these is distinguished from naturally occurring DNA or RNA by changes to the backbone of the molecule.

The term “detectable” is to be understood as meaning such that the presence of a nanoparticle and/or a plurality of nanoparticles attached to the binding patch, in particular to a coupling molecule or, respectively, a nanostructure, can be measured, e.g., by visual and/or optical detection.

A (plasmonic) nanoparticle hereby further relates to a molecule and/or a nanoparticle, as a structure having a size, e.g. a maximum diameter, in the order of nanometers or tens of nanometers, typically smaller than 100 nm. The nanoparticle may be functionalized. A functionalized nanoparticle hereby refers to a nanoparticle, which has a functional substituent attached to it. The function of the functional substituent is to bind the nanoparticle to the first binding site but may also be related to the effect of preventing aggregation of the functionalized nanoparticle in solution. The substituent is linked to the nanoparticle’s surface. Linkage of the substituent to the nanoparticle’s surface is achieved at least through a binding group. The binding group at least forms a bond with the substituent and with the surface of the nanoparticle. In a particular embodiment the binding group forms a bond with a metallic, in particular a gold or silver surface of the nanoparticle. The binding group used for attaching the substituent to the nanoparticle can be a thiol moiety of the form R - SH. In a preferred embodiment of a detectable label, the substituent is bond to the metallic, in particular gold or silver surface of the nanoparticle via sulfur.

The plasmonic nanoparticles can be chosen from the group of preferred nanoparticles, including: metal nanoparticles, gold nanoparticles, silver nanoparticles, Platinum nanoparticles, metal coated particles, metallized polystyrene or latex beads, metallized carbon nanotubes, a silver coated particle, a metallized cellulose particle.

The substituent can be any molecule. The substituent preferably further comprises a functional group, e.g., a carboxyl group, a thiol group, an amino acid, a protein, an antibody, a virus or a hormone. The first and/or second binding sites comprise within the polynucleotide-based nanostructure respective different nucleotide or oligonucleotide sequences. Thereby, each respective binding site can individually be “programmed”, i.e., predefined by an unambiguously assignable nucleotide and/or oligonucleotide sequence. During preparation of the nanostructure, these unambiguously assignable nucleotide and/or oligonucleotide sequences are included into a polynucleotide of the nanostructure. A part of these unambiguously assignable sequences peeks out of the prepared nanostructure. Other nucleotides and/or oligonucleotides can be attached to this part or this part is chemically modified with a desired binding reagent further comprised by the binding site to link the nanostructure to a target.

As used herein, a “binding reagent” refers to any substance that binds to a target, a detectable nanoparticle or an analyte with desired affinity and/or specificity. In this sense, a binding reagent is used in analogy to a substituent within the present description of the invention.

Embodiments are described or can be gathered from the description, which can be arbitrarily combined with each other or with other aspects of the present invention, unless such combination is explicitly excluded or technically impossible.

According to a first embodiment of the nanostructure, the polynucleotide-based nanostructure has a predefined size and/or a predefined shape.

The terms “size” and/or “shape” of the nanostructure hereby relate to the two-dimensional or three-dimensional overall shape, i.e., the spatial extension of the labeling nanostructure in analogy to a protein conformation, wherein the spatial arrangement of the polynucleotides and/or oligonucleotides, is obtained in particular through base pair recognition, which assembles the polynucleotides into this object of defined shape and size, thereby determining the overall shape of the object.

E.g., a DNA Origami as a particular instance of an object having a defined shape and size, whereas the desired shape and size is obtained from the folding route of a given scaffold and a respective staple sequence that can fulfill the folding. In this respect, the shape is predetermined by the sequences of the nucleotides and/or oligonucleotides that, as a whole or in part, recognize each other and bind to each other to form the nanostructure.

In this respect, the size is limited by the length of the scaffold used for folding, i.e. , the number of nucleotides and/or polynucleotides used. In some instances, the shape may be described means of an aspect ratio, e.g., for a rod-shaped nanostructure, the length : width : height ratio.

According to a further embodiment of the nanostructure, a) the nanostructure consists of the polynucleotide-based nanostructure, or b) the nanostructure comprises at least one further subcomponent, the at least one further subcomponent having, in particular, a predefined size and/or a predefined shape or not having a predefined size and/or a predefined shape.

According to a further embodiment of the nanostructure, the polynucleotide-based nanostructure comprises one or more scaffold strands.

A scaffold strand hereby relates to a long polynucleotide strand, e.g., a long singlestranded DNA strand, typically comprising 7000 nucleotides. In case of a DNA Origami, the scaffold strand is combined with hundreds of designed short single-stranded DNAs as staples. Each staple has multiple binding domains that bind and bring together otherwise distant regions of the scaffold via crossover base pairing, thereby folding the scaffold in a predefined manner. As the staple sequences can be predefined, the geometry, e.g., the size and shape of the resulting nanostructures is advantageously predeterminable and can thus be adapted to an application specific requirement.

According to a fifth embodiment of the labeling nanostructure, wherein the predefined number N of first binding sites is preferably 1<N<1000, preferably 1<=N<10, preferably 1<=N<100, preferably 10<=N<500, preferably 10<=N<400, preferably 10<=N<350, preferably 10<=N<300, preferably N>2, preferably N>3, preferably N>4, preferably N>5, preferably N>6, preferably N>7, preferably N>8, preferably N>9, preferably N>10. According to a further embodiment, the plasmonic nanoparticle is a functionalized nanoparticle, having a metal core, potentially a silver coating, and a sulfide bond substituent.

The silver coating of each of the functionalized nanoparticles forms a shell around the metal core and the metal core is at least partially covered by the silver shell. This embodiment has the advantage, that under resonant excitation, metal nanoparticles have the unique ability to concentrate the free-space optical field within subwavelength regions, wherein this ability is based on surface plasmon excitation. The overall plasmonic behavior of gold (Au) and silver (Ag) nanoparticles is similar, however, silver is known to give higher field effects due to a lower plasmon damping leading to enhanced optical performances.

The plasmonic nanoparticles can further be chosen for each embodiment from the group of gold nanoparticles, Ag nanoparticles, metal nanoparticles, metallized polystyrene or latex beads. In a further embodiment it is preferred to combine different types of plasmonic nanoparticles at one polynucleotide-based nanostructure.

The nanostructure, preferably, can be a nanoscale object, having in particular a length scale of between 1 to 100 nanometers, and/or may in particular include the features of a nanoscale object as defined in the international patent application PCT/EP2020/082866, in particular in claim 13 thereof, wherein the functionalized nanoparticle defined by claim 1 of said patent application may form a plasmonic nanoparticle for the PUF label. A functionalized nanoparticle is preferably prepared according to the following method, whereas the functionalized nanoparticle comprises or consists of a metal core, a silver coating and a sulfide bond substituent, in an aqueous solution, the method comprising a step of chemical functionalization of a metal nanoparticle in the aqueous solution, wherein the aqueous solution comprises or consists of water and ingredients, wherein the ingredients comprise or consist of the metal nanoparticle, a thiol of the form R-SH, where R represents a substituent, and a silver compound, the substituent being, preferably, organic, and having, preferably, a functional group.

The functional group, respectively preferably, comprises or consists of a carboxyl group (- COOH), an aldehyde group (-CHO), a hydroxyl group (-OH), an amino group (-NH2), an amide group (-CONH), and/or wherein the substituent comprises a carboxyl group, an amino acid, a protein, an antibody, a virus or a hormone, or two or more thereof. The substituent can be any molecule, which is capable of having a binding group (R-SH). Within the present description of the invention, the terms “substituent” and “ligand” have the same meaning, if not defined to the contrary. In particular, the thiol, which includes the substituent, comprises or consists of mercaptopropionic acid (MPA), mercapto methoxy polyethylene glycol (mPEG-SH), PEG-SH, or most preferably DNA-SH. Preferably, the thiol, which includes the substituent, comprises at least one of 2,5,8,11 ,14,17,20- Heptaoxadocosane-22-thiol, or CH3O(CH2CH2O)nCH2CH2SH. The substituent preferably further comprises a functional group, e.g. carboxyl group, an amino acid, a protein, an antibody, a virus or a hormone.

The water of the aqueous solution preferably is a purified water, preferably a distilled water, most preferably a double-distilled water (abbreviated “ddH2O”). Storage of the functionalized nanoparticle according to the invention preferably takes place in aqueous solution, preferably in ddH2O, preferably in the absence of a buffer.

The spectral properties, in particular the color of the nanoparticles, in particular when deposited on top of the binding patch, depend, amongst others, on the size and/or the geometry of the nanoparticles. Therefore, adjusting the size and/or the geometry of the nanoparticles may be used to provide nanoparticles of different color. In particular using a chemical functionalization (of the nanoparticles) depending on the size and/or the geometry (of the nanoparticles) may be employed.

The substrate may consist of, or at least comprise, one or more of the following materials: silicon, glass, polymer, metal. The surface of the substrate is preferably planar, but can also be curved. The substrate preferably is sheet-like or plate-like, but may also have another shape, for example, a cuboid or disk-like shape.

The PUF label may be integrated into the information label. The PUF label and/or the information label may be integrated into another item, which may be the article to be secured by the anti-counterfeit system, or which may be a price tag connected to the article or a brand label connected to, or fabricated with the article.

The pattern preferably, is hexagonal. It may be created by lithography, preferably nanosphere lithography. A way of creating a pattern by nanosphere lithography is described by Shetty RM, Brady SR, Rothemund PWK, Hariadi RF, Gopinath A. Bench- Top Fabrication of Single-Molecule Nanoarrays by DNA Origami Placement. ACS Nano 2021 , 15(7): 11441-11450.

The pattern preferably, is based on a hexagon or another polygon or combinations thereof. It may comprise a regular grid or an irregular grid, wherein preferably the crossing points of the grid (a virtual grid) form the center points of the binding patches. The critical property of the pattern is that the location points for locating the binding patches can be predefined, in order to control the creation of separate points of different colors of a metasurface.

The binding patch of a pattern is suitable to bind one or more nanoparticles, in particular via a coupling molecule, preferable via a nanostructure, in particular DNA-origami. The binding patch may be a surface section of the surface or a patch added to the surface. Each center of a binding patch is preferably located at a crossing point or at a center of a unit cell of the virtual grid, which defines the geometric array of the pattern. The binding patch may be capable to bind the one or more nanoparticles, in particular coupling molecules, preferable nanostructures, while, in particular, the residual area surrounding the binding patches is preferably not capable, and/or has a reduced capability, for binding the one or more nanoparticles, in particular coupling molecules, preferable nanostructures, the residual area also being referred to as a passive surface. Such a residual area may, however, in certain embodiments be capable to unspecifically bind the coupling molecule, but preferably to not bind to a second binding sites of a coupling molecule, in particular the nanostructure. The binding patch may be covered with a covering layer of the surface, in particular by a metal, for example a layer of gold or silver, which may be deployed on the substrate by a deposition method, including sputtering, vapor deposition, or electroplating.

Preferably, the center to center distance of the binding patches of the pattern is a value, in particular a constant value, between 100 nm and 10000 nm, between 100 nm and 5000 nm, between 100 nm and 2000, between 100 nm and 1000 nm, preferably between 100 nm and 600 nm, preferably between 200 nm and 600 nm. Preferably, a maximum, a minimum or an average diameter of the binding patch may be between 100 nm and 1000 nm, preferably between 200 nm and 600 nm.

A random arrangement of particles on a surface, here: on the binding patch, refers to the placement of particles on a surface in a way that does not follow a specific pattern or organization. This arrangement is characterized by the lack of order or structure, and the particles are typically placed in a haphazard or chaotic manner. The randomness may be created by the feature that the first binding sites of the nanostructures are not located and/or oriented in a predefined manner on the binding patches, such that the adsorption and binding of plasmonic nanoparticles to the first binding sites is a statistical event, meaning that not all first binding sites of a binding patch are being occupied by a nanoparticle, that the orientation and/or distance of nanoparticles bound to the nanostructure is not predefined and so on. Moreover, the nanoparticles contained in a stock solution of nanoparticles may be heterogenous and contain different types of plasmonic nanoparticles, each of which may have different properties influencing the color of the particle, taking into account the plasmonic coupling within a binding patch. A random arrangement of particles on a surface refers to a situation where particles can be of different sizes, shapes, and colors, and their arrangement can be irregular or have no apparent pattern. The arrangement of particles can be influenced by various factors, including the properties of the surface, the characteristics of the particles, and external conditions such as temperature, humidity, and pressure. The randomness of the arrangement can be quantified by statistical measures such as the entropy or disorder of the system.

Preferably, a minimum surface-to-surface distance between two of the nanoparticles of the random arrangement is equal or less than the maximum diameter or maximum length of one of the plasmonic nanoparticles, and is, in particular, smaller than 200 nm, preferably smaller than 100 nm, preferably smaller than 50 nm, preferably smaller than 30 nm, preferably smaller than 20 nm, or preferably smaller than 10 nm.

Preferably, a minimum surface-to-surface distance between two of the plasmonic nanoparticles of the random arrangement is equal or less than the maximum diameter or maximum length of one of the plasmonic nanoparticles, and is, in particular, smaller than 20 nm, or preferably smaller than 10 nm. Plasmonic coupling is known to show up, in particular when using such ranges. It is preferred that a fraction, or that the major fraction, or all of the nanoparticles on top of a binding patch are bound to the nanostructures side by side in a distance that allows plasmonic coupling.

A layer of coupling molecules, in particular nanostructures, means that the coupling molecules are arranged within a layer, which is located on top of the binding patch, preferably directly on top of the binding patch. A layer of nanostructures preferably means that the coupling molecule(s), in particular the nanostructure(s), at least partially or completely cover(s) the surface section corresponding to the binding patch. The coverage may be complete or incomplete, and the layer of coupling molecule(s) may also be porous or web-shaped. The layer of coupling molecules, preferably, is restricted to the surface section corresponding to the binding patch, but it may also exceed the surface section corresponding to the binding patch in directions parallel to the surface, as long as the layers of vicinal binding patches are separated; said separation of the binding patches with nanoparticles leads to the desired result containing separate patches/spots of different color.

Preferably, the nanoparticle(s), in particular plasmonic nanoparticle(s), on top of a binding patch may have one or more of the following attributes:

• a randomly determined distance between vicinal nanoparticles, in particular caused by a random occupation of predetermined first binding sites;

• a randomly determined orientation of non-spherical nanoparticles,

• a differing shape of nanoparticles,

• a differing size of nanoparticles,

• a differing material or material composition of nanoparticles.

The invention is also directed to an anti-counterfeit system for providing protection against forgery of an article, comprising

• a PUF label according to the invention, being assigned, in particular connected, to the article, wherein the PUF label may generally be integrated into the article;

• an information label being assigned, in particular connected, to the article, the information label providing access to reference image data, which correspond to a reference image of the optically detectable pattern of randomly colored patches of the PUF label,

• an imaging apparatus for optically detecting and providing a pattern image of the pattern of randomly colored patches of the PUF label, and • a computing device configured, in particular programmed, for using pattern image data, which correspond to the pattern image, and for comparing and detecting a match or a mismatch of the pattern image data and the reference image data, for detecting forgery or originality of the article.

A digital image, in particular the pattern image, is a two-dimensional representation of visual information that is captured from the surface of the PUF label using digital technology, e.g. a camera. The pattern image data may correspond to a set of data representing a (real) pattern image of the surface/the pattern, typically containing information about a matrix corresponding to the pattern image, each pixel of the matrix being characterized by at least one color property, e.g. a hue value. The pattern image data may also correspond to an individual code representing the individual character of the pattern image, e.g. an individual code derived as a mathematical function, which depends on the pattern image. For example, a unique property (UP) of multiple binding sites or each binding site of a pattern image may be determined, and the set of unique properties may be utilized as the pattern image data or for deriving the pattern image data. The pattern image can be analyzed by image processing, performed by the computing device for example, the computing device being programmed to find the binding patches using particle analysis of the image data or an object detection algorithm, and programmed to determine at least one color property of each area, determined by said analysis as representing a binding patch. Such algorithms for image processing are generally known and available, for example by OpenCV. OpenCV (Open Source Computer Vision Library) is a popular open-source library of computer vision and image processing algorithms. It is maintained by the OpenCV community.

The computing device may be programmed for comparing and detecting a match or a mismatch of the pattern image data and the reference image data. To compare two images, the programming may implement, for example, an algorithm using, for example, a Mean Square Error (MSE) of the pixel values of the two images. Similar images will have less mean square error value. Using this method, two images having the same height, width and number of channels may be compared, e.g. the pattern image and the reference image. When being programmed to compare one or more color properties of corresponding binding patches in the same location of the pattern image and the reference image, the respective values are pairwise compared, taking into account a tolerance, and the result may be statistically evaluated over all binding patches contained in the pattern image.

The reference image data, which correspond to a reference image of the optically detectable pattern of randomly colored patches of the PUF label, is preferably obtained and the data is preferably saved in its raw format (TIF, 2D map of RGB values ... ) or as a function thereof, which can also be a hash function.

A unique property of a binding site preferably is a color property of the binding site, in particular any of the following color properties, or respectively, color appearance parameters: hue, lightness, brightness, chroma, colorfulness, RGB values, and saturation.

The UP of one spot I patch could be the three average RGB values that this spot creates in a dark field image - since each patch covers multiple pixels of the camera, it is preferred to utilize an average value of the RGB values of the multiple pixels capturing the spot/patch. An alternative could be the hue value. Usually in color theory, hue is one of the main properties -called color appearance parameters- of a color, defined technically in the CIECAM02 model as "the degree to which a stimulus can be described as similar to or different from stimuli that are described as red, orange, yellow, green, blue, violet,".

Colors can be precisely measured, characterized, and reproduced using various tools and techniques, which may be implemented by the imaging apparatus and which provide a color property as an output.

A spectrophotometer: An instruments that measures the amount of light reflected or transmitted by an object across the visible spectrum of light. The data collected by a spectrophotometer can be used to generate numerical values that describe the color properties of the object, such as its hue, saturation, and brightness.

Colorimeter: This is similar to a spectrophotometer, but measures color based on three primary colors (red, green, and blue) instead of the entire visible spectrum of light.

Color Space: A color space is a mathematical model that defines the range of possible colors. By using a color space, one can precisely specify a color using numerical values that describe its position in the color space. The most common color spaces are RGB (typically used in digital displays), CMYK (typically used in printing), and LAB (typically used in color science and engineering). Color Matching Systems: These are sets of standardized colors that are used to ensure consistent color reproduction across different media and materials. Examples of color matching systems include Pantone, which is widely used in the graphic design and printing industry, and RAL, which is used for paints and coatings.

Color Management Systems: These are software and hardware tools that are used to ensure color consistency across different devices, such as monitors, printers, and cameras. Color management systems use profiles and calibration to ensure that colors are accurately reproduced across different devices and media.

Overall, for a suitable color measurement, a characterization and reproduction of color preferably uses a combination of accurate measurement instruments, standardized color spaces and systems, and color management tools and processes.

Preferably, the PUF label is accompanied by, in particular provided with, a color calibration label, which contains a set of predefined color marks, whose respective color is standardized and known to the computing device, or whose color scheme is presented by an information code as a readable information in the color calibration label or the information label. This way, any systematic deviation between a detected pattern image and the reference image may be compensated by the computing device by calculating a suitable correction factor or respectively, a correction function, determined from the color calibration label. The color marks of the color calibration label may be defined by the plasmonic nanoparticles itself, which are used to create the PUF label. For example, the types of plasmonic nanoparticles contained in a stock solution for creating the PUF label may also be provided as defined stock solution of one respectively purified type of plasmonic nanoparticle, or as a defined set of purified plasmonic nanoparticles of known type, concentration, and/or ratio.

The calibration of the camera of the imaging system may also be performed using color calibration, independent from the article, which means that a color calibration label may be separate from the PUF label and/or the article.

Preferably, the imaging apparatus contains an optical microscope, in particular a darkfield microscopy device. Since the surfaces carry features on a nanometer to micrometer scale, obtaining a useful image will be improved by magnification, e.g. using an optical microscope. Darkfield microscopy is a specialized microscopy technique that allows the visualization of specimens that are not visible under normal brightfield illumination. In darkfield microscopy, the specimen is illuminated with a cone of light from the side, which causes light to scatter off the specimen at various angles. The scattered light is then collected by the objective lens, while the directly transmitted light is blocked. This results in a bright image of the specimen against a dark background.

Plasmonic nanoparticles, in particular metallic nanoparticles bound to a surface can be visualized using darkfield microscopy by taking advantage of the scattering properties of the nanoparticles. Plasmonic nanoparticles, such as gold and silver nanoparticles, exhibit a phenomenon called localized surface plasmon resonance (LSPR), which causes them to scatter light strongly at specific wavelengths. When illuminated with white light, nanoparticles with a certain size and shape will scatter specific colors of light, which can be observed as bright spots in a darkfield microscopy image. The color depends on the angle of polarization, when using polarized light. This can be utilized for obtaining an additional security feature to the anti-counterfeit system, when both, the reference image and the pattern image, are obtained using the same angle of polarization, and preferably, by keeping the angle of polarization secret, or as alternatively also preferred, by providing the angle of polarization to the user of an article labeled by the PUF label. Color may also depend on: material, size, shape, neighboring particles

To obtain darkfield microscopy images of metallic nanoparticles bound to a surface, the sample is illuminated with a white light source and observed under a darkfield microscope. The scattered light from the nanoparticles will appear as bright spots against a dark background. By controlling the size, shape, and concentration of the nanoparticles, it is possible to obtain images with different levels of contrast and resolution. A darkfield microscope is understood to be an optical device capable of producing a darkfield image. A darkfield image, typically, excludes the unscattered beam from the object of interest. As a result, the field around the specimen (i.e., where there is no specimen to scatter the beam) is generally dark, the specimen/object of interest here being the colored patch or the colored pattern composed of the colored patches, respectively.

Preferably, the information label a) contains the reference image data, or b) contains link data including a remote address, in particular an URL, which allows accessing the reference image data being stored in a data storage being remote to the information label and being accessible by the computing device via a data network, in particular the Internet.

Preferably, the information label contains digital signature data forming a digital signature, which is capable of being verified by a public verification key.

A digital signature system using private keys and a public key is a method of verifying the authenticity and integrity of digital messages or documents, and can be used for verifying the authenticity and integrity of the information label. The digital signature system uses two keys: a private key and a public key. The private key is typically owned by the owner (the vendor) of the article comprising the PUF label and the information label. The public key is made available to others who wish to verify the authenticity of the information label. When the owner of the article wants to sign an information label, they use their private key to generate a unique digital signature (the digital signature data). The recipient (for example a pharmacy buying pharmaceuticals) can then use the sender's public key to verify the signature and confirm that the information label was indeed delivered by the owner and has not been tampered with since it was signed. This way, it can be prevented that a counterfeiter successfully imitates thee combination of the PUF label and the information label, which is provided to secure an article.

The private key and/or the digital signature data may also contain information, which is unambiguously derived from the properties of the PUF label, for example, a sequence of color information of neighboring patches of the pattern.

Use case scenario

Use case scenario: A company (e.g. pharmaceutical producer) wants to label an article against counterfeiting. It places the PUF label on the article. Next to the PUF label can be an information label, e.g. a bar code, or also an RFID-tag (which in our case only stores publicly accessible information, no secrets or secret keys of any kind). A client buys the article in a store (for example, a medicine package in a pharmacy). The shop provides an imaging apparatus, containing, for example, a simplified darkfield-microscope, that scans the PUF label and records the optical response, e.g. in the form of obtaining a pattern image. This pattern image will be compared with a reference image (or reference data) either stored within the barcode directly or on a remote data storage of the company, e.g. with a repository stored at the company’s website, at a website of the producer of the labels, or at a third party who stores the information at a secure online repository.

In order to certify that the information label, or respectively the barcode, stems from the pharmaceutical producer (and from no one else), the standard cryptographic tool known as “digital signatures” can be employed in one possible embodiment of the invention. These signatures can be generated only by some legitimate party (in our case, by the pharmaceutical producer), using a secret digital signing key (private key) solely known to this party/the producer. Digital signatures can be verified for validity by everyone on the basis of a “public” verification key. Public here means that this verification key can be known to everyone without harming security. It can be distributed openly to all participants of the scheme and does not need to be protected in hardware against adversarial access. The public nature of the verification key constitutes a very strong security, usability, and practicality advantage in the given situation.

The digital signature is a string that is computed by a randomized algorithm (hereinafter “SIG-ALG”). SIG-ALG gets as input some to-be-signed DATA, which should include the unique properties (UP) of the PUF (e.g., the positions and colors of “glooming points” or centers of emission, or yet other features), or, alternatively, a hash value or other function of these UP (i.e., formally hash(UP) or f(UP)). DATA might also include other product- related information, such as package size and content, or the manufacturing date and place, the best before date, the value of a banknote, biometric features of a passport holder, access card holder, or an owner of something, etc.

SIG-ALG also gets as further input (or is “parameterized by”) the signature scheme’s individual private key SK. Preferably, each legitimate manufacturer or issuer of signatures has its own, individual SK. (I.e., the SK’s of different people/parties differ).

An individual digital signature DS can then be written as DS - SIG-ALG(SK, DATA) or also DS - SIG-ALG_SK(DATA). (The notation in computer science terms means that the value DS is set to the output of the (randomized) algorithm SIG-ALG).

The verification works similarly: There is a (usually, but not necessarily, deterministic) verification algorithm VER-ALG. This algorithm VER-ALG gets as input DATA and the individual public key PK (that is associated with the individual private key SK that was employed). It also gets as input a purported digital signature DS*.

The algorithm VER-ALG with input (DATA, DS*, PK) outputs “ok” or “valid” only if DS* has been generated by SIG-ALG with input (SK, DATA).

PK can be made public and can be known to everyone; PK only allows the verification of a digital signature for validity. The key PK associated with a certain manufacturer (Bayer, or Pfizer, or the Bundesbank, etc.) can therefore be contained in every label verification device. E.g, it can be stored in every smartphone, pharmacy, supermarket, or bank throughout the country. No harm is done if adversaries get to know PK: It only allows them to verify signatures! SK, on the other hand, must ONLY be known to the legitimate manufacturer.

The above means that for the verification procedure, VER-ALG preferably has access to DATA. DATA preferably is stored publicly accessible on the labeled item or product. It can be included in the said barcode or RFID-tag, for example.

While digital signatures are a preferred part of the information label, it is also preferred that a database stores the unique properties of each label. In the verification procedure, one may contact the database and ask for the “right” unique properties of the label that is tested. This has the advantage that “larger” datasets can be used as unique properties, for example the optical features of an entire, larger area of a patch. The storage capacity of the barcode and the RFID-tag used on the label typically are limited; if they store all DATA in an offline verification scenario, then this DATA must necessarily be suitably small and short.

The invention is also directed to a method for the fabrication of the PUF label according to the invention and the embodiments thereof, comprising the steps:

• providing a substrate comprising a surface,

• providing a pattern of binding patches on the surface, in particular by lithography,

• optional: providing, in a layer (5) on top of each binding patch (4), at least one coupling molecule, in particular a polynucleotide-based nanostructure, the coupling molecule having one or more first binding sites for binding a nanoparticle, and preferably having one or more second binding sites for binding the coupling molecule to the surface,

• providing a random arrangement of nanoparticles, in particular plasmonic nanoparticles, on top of each binding patch, the nanoparticles being bound to the binding patch directly or via a coupling molecule,

• the pattern being optically detectable as a pattern of randomly colored patches, the color of each colored patch being determined by the random arrangement of nanoparticles of the respective binding patch.

The invention is also directed to an anti-counterfeit method for examining the originality of an article being provided with a PUF label according to the invention and the embodiments thereof, which, in particular, uses the anti-counterfeit system according to the invention and the embodiments thereof, comprising the steps:

• providing a PUF label according to the invention, being assigned, in particular connected, to an article,

• providing an information label being assigned, in particular connected, to the article, the information label providing access to reference image data, which represent a reference image of the optically detectable pattern of randomly colored patches of the PUF label,

• optically detecting and providing a pattern image of the pattern of randomly colored patches of the PUF label by way of an imaging apparatus,

• comparing and detecting, by way of a computing device, a match or a mismatch of pattern image data, which correspond to the pattern image, and the reference image data, for detecting forgery or originality of the article.

Further advantages, features and applications of the present invention are provided in the following detailed description of the exemplary embodiments and the appended figures. The same components of the exemplary embodiments are substantially characterized by the same reference signs, except if referred to otherwise or if other reference signs emerge from the context. In detail:

Fig. 1 shows a schematic side view of a PUF label according to an embodiment of the invention.

Fig. 2a shows a schematic side view of a nanostructure formed by DNA-origami, which carries metallic nanoparticles, in a PUF label according to Fig. 1.

Fig. 2b shows a schematic side view of plasmonic nanoparticles of different size, shape and material, for the PUF label in Fig. 1.

Fig. 2c shows a schematic top view of a regular pattern of binding patches, for the PUF label in Fig. 1 , wherein the cut along line A corresponds to Fig. 1.

Fig. 2d shows a schematic top view of an irregular pattern of binding patches, for a PUF label according to another embodiment of the invention.

Fig. 3 shows a schematic top view of embodiments of differently shaped polynucleotide-based nanostructures suitable for being used in the PUF label according to the invention.

Fig. 4 shows a schematic perspective view of a PUF label according to another embodiment of the invention.

Fig. 5 shows a scanning electron microscopy image of a surface of a PUF label corresponding to the scheme of a PUF label shown in Fig. 4.

Fig. 6a shows, as the results of subsequently performed manufacturing steps, the schematic perspective view of a substrate, being provided with binding patches arranged in a pattern on the surface, each binding patch being covered by a layer of polynucleotide-based nanostructures.

Fig. 6b shows, as the results of subsequently performed manufacturing steps, the schematic perspective view of a substrate carrying binding patches, as formed in Fig. 6a, the binding patches being arranged in a pattern on the surface, each binding patch being covered by a layer of polynucleotide-based nanostructures, and different plasmonic nanoparticles being bound to the layer. Fig. 7a shows three darkfield microscopy images of the metasurfaces of a PUF label according to other embodiments of the invention, the plasmonic nanoparticles of a metasurface either only comprising one type of nanoparticles, which is spheric gold nanoparticles; or the plasmonic nanoparticles of a metasurface comprising a mixture of spheric gold nanoparticles, spheric silver nanoparticles, and rod-shaped silver nanoparticles, in a high concentration; or the plasmonic nanoparticles of a metasurface comprising a mixture of spheric gold nanoparticles, spheric silver nanoparticles, and rod-shaped silver nanoparticles, in a low concentration.

Fig. 7b shows a histogram representing a statistical evaluation of the pattern images in Fig. 7a, the histogram showing the pixel numbers in the image having a specific hue value.

Fig. 7c shows three darkfield microscopy images of the same section of a metasurface of a PUF label, as shown in Fig. 7a, each image taken by illuminating the metasurface with a light of different polarisation; the diagram shows the hue value of a single patch in the pattern in dependence on the polarisation angle of the light, which illuminates the metasurface.

Fig. 8a schematically shows a top view of the metasurface of a PUF label according to another embodiment of the invention, without illumination of the metasurface.

Fig. 8b schematically shows a top view of the metasurface of Fig. 8a, with illumination by light polarized at 45°.

Fig. 8c schematically shows a top view of the metasurface of Fig. 8a, with illumination by light polarized at -45°.

Fig. 9 shows a larger darkfield microscopy image of the illuminated metasurface of a PUF label according to another embodiment of the invention, the smaller pictures show different versions of a section taken from said image, each version being illuminated by light of different polarisation.

Fig. 10 shows a schematic view of an anti-counterfeit system, using the PUF label shown in Fig. 1 or 5, for providing protection against forgery of an article. Fig. 11a shows the steps of a method for the fabrication of the PUF label according to the invention.

Fig. 11 b shows the steps of an anti-counterfeit method for examining the originality of an article being provided with a PUF label according to the invention.

Fig. 1 shows a schematic side view of a PUF label 1 according to an embodiment of the invention. The physically unclonable function (PUF) label 1 , comprises a silicon substrate 2 having a surface 3, wherein the surface areas 3a corresponding to the binding patches 4 are occupied and the residual area 3b of the surface is passivated to be basically incapable of binding the polynucleotide-based nanostructures and/or the plasmonic nanoparticles. The PUF label 1 also comprises a plurality of binding patches 4 arranged in a pattern 10 on the surface 3, each binding patch 4 being covered by a layer 5 of polynucleotide-based nanostructures 6, the polynucleotide-based nanostructures 6 each optionally having one or more second binding sites 6’ for binding the polynucleotide-based nanostructure 6 to the surface 3, and two or more first binding sites 6” for binding plasmonic nanoparticles 7a, 7b. The layer 5 of each binding patch 4 carries a random arrangement of plasmonic nanoparticles 7a, 7b, which are bound to the layer 5 and are arranged such that a plasmonic coupling (pc) of the metallic nanoparticles 7a and 7b is achieved. Hereby, the pattern is optically detectable as a pattern 10 of randomly colored patches 4, the color of each colored patch 4 being determined by the random arrangement of plasmonic nanoparticles 7a, 7b of the respective binding patch 4. The surface 3 having such a pattern 10 is also referred to as “metasurface”. In the figure, the two binding patches will show different colors, yellow and orange, when illuminated with a light source and imaged by an imaging apparatus.

Fig. 2a shows a schematic side view of a nanostructure 6, here formed by DNA-origami, which has first binding sites 6” for binding metallic nanoparticles 7a, 7b, wherein the minimum surface-to-surface distance d_ss of the two metallic nanoparticles 7a, 7b may be 10 nm or any other suitable distance for establishing a plasmonic coupling (PC) between the nanoparticles. The second binding sites 6’ may be represented by the backbone of the DNA-origami which has the capability to adhere to the silicon oxide surface 3a. However, the second binding sites may also be represented by a more specific section of a nanostructure, respectively a DNA origami, or by one or more additional chemical moieties of the nanostructure.

Fig. 2b shows a schematic side view of nanoparticles, in particular plasmonic/metallic nanoparticles, of different size, shape and material, for the PUF label 1 in Fig. 1. While nanoparticles 7a, 7b and 7c may be all spherical, with possibly differing kind of core and/or shell, i.e. a differing material, the nanoparticles may have also a rod shape, like nanorod 7d and 7e, which, in comparison, have a differing radial diameter along the axial direction of their elongated direction. Other shapes, not shown, of nanoparticles may be easily applied.

Fig. 2c shows a schematic top view of a regular pattern, here a hexagonal pattern, of binding patches 4, for the PUF label 1 in Fig. 1 , wherein the cut along line A corresponds to Fig. 1. The solid lines representing the hexagon cells are virtual lines, drawn for illustrating the hexagonal geometry of the patch pattern, and do not have a material representation in a PUF label 1. A regular pattern may simplify a reliable evaluation of a pattern image, since the individual location of the patches can be derived from the geometric relationships of the patch locations.

Fig. 2d shows a schematic top view of an irregular pattern 11 of binding patches 4, for a PUF label according to another embodiment of the invention. There is, in general, no restriction for the design of the pattern as long as the presentation of differently colored spots has a sufficiently high information density for providing a PUF label. For example, while a clustering of spots/patches, i.e. a zero distance between different patches of the pattern, is generally not desired, it may be acceptable if a certain fraction of the patches are clustered, preferably a minor fraction of all the patches of a pattern. It is preferred that, for a major fraction of the patches of a pattern, each patch has a distance d_ss larger than zero, preferably larger than 20, 50, 100 or 500 nm, to any other patch of the pattern.

Fig. 3 shows a schematic top view of embodiments of differently shaped polynucleotide- based nanostructures 6a, 6b, 6c, 6d, 6e and 6f suitable for being used in the PUF label according to the invention. DNA origami, in particular, offer the advantage of allowing a “programming” the three-dimensional shape of the nanostructure by choosing the appropriate sequence in the polynucleotide backbone.

Fig. 4 shows a schematic perspective view of a PUF label 1 according to another embodiment of the invention. The silicon substrate 2 has a surface 3, which carries a hexagonal pattern of binding patches 4, which was generated by nanosphere lithography. Each patch 4 carries the same type of coupling molecule, here a hollow-triangle shaped DNA origami 6a. The pattern carries a random arrangement of different nanoparticles 7 (7e). In the embodiment, there are no patches, which have an identical arrangement of nanoparticles. However, it may happen from time to time by coincidence that two or more patches have the same arrangement of nanoparticles (type, orientation, relative location on the patch). Reference numeral 90 indicates that the assembly of nanoparticles occurs in solution. The solution can later be removed, the objects can be stored in air. The metasurface can later be covered by a protective layer, e.g. resin or silicon oxide.

Fig. 5 shows a scanning electron microscopy image of a surface of a PUF label 1 corresponding to the scheme of a PUF label shown in Fig. 4. The groups of nanoparticles 7 are each positioned on top of a binding patch 4, which is not visible here for any patch. From the picture it may be derived that the edge-to-edge distance d_ee of two neighboring patches of the hexagonal pattern is roughly 400 nm. The patches will show up in different colors respectively.

Fig. 6a shows, as the results of subsequently performed manufacturing steps, the schematic perspective view of a substrate, being provided with binding patches arranged in a pattern on the surface, each binding patch being covered by a layer of polynucleotide- based nanostructures. In detail, polystyrene nanospheres are deposited on a surface, followed by chemical treatment of the surface by exposing it to a hydrophobic chemical (HMDS), which bonds to the surface at all places not touched by the polystyrene nanospheres. The nanospheres are then removed, resulting in a surface 3 with binding patches 4. The polynucleotide-based nanostructures 6 are then added in Placement buffer (described later), and after a series of washing steps, results in the placement of polynucleotide-based nanostructures 6 on the surface.

Fig. 6b shows, as the results of subsequently performed manufacturing steps, the schematic perspective view of a substrate carrying binding patches 4, as formed in Fig. 6a, the binding patches being arranged in a pattern 10 on the surface, each binding patch 4 being covered by a layer 5 of polynucleotide-based nanostructures 6, and different plasmonic nanoparticles 7a, 7b, 7e being bound to the layer 5. The rightmost images show SEM images of the resulting metasurfaces, with a scale bar of 400 nm. The pictures also show a uniform placement of nanoparticles, which is not a random arrangement arrangement of one or more nanoparticles 7; 7a; 7b; 7c; 7d; 7e and which does not result in differently colored patches 4, as required for the invention.

Fig. 7a shows three darkfield microscopy images of the metasurfaces of a PUF label 1 according to other embodiments of the invention, the plasmonic nanoparticles of a metasurface either only comprising one type of nanoparticles, which is spheric gold nanoparticles in random arrangement; or the plasmonic nanoparticles of a metasurface comprising a mixture of spheric gold nanoparticles, spheric silver nanoparticles, and rodshaped silver nanoparticles, in a high concentration (center image); or the plasmonic nanoparticles of a metasurface comprising a mixture of spheric gold nanoparticles, spheric silver nanoparticles, and rod-shaped silver nanoparticles, in a low concentration (picture on the right).

Fig. 7b shows a histogram representing a statistical evaluation of the pattern images in Fig. 7a. The histogram show the number of pixels (y-axis) in the image having a specific hue value (x-axis). Scale bar is 400 nm.

Fig. 7c shows three darkfield microscopy images of the same section of a metasurface of a PUF label 1 , as shown in Fig. 7a, each image taken by illuminating the metasurface with a light of different polarisation; the diagram shows the hue value of a single patch in the pattern in dependence on the polarisation angle of the light, which illuminates the metasurface. Fig. 8a schematically shows a top view of the metasurface of a PUF label 1 according to another embodiment of the invention, without illumination of the metasurface.

Fig. 8b schematically shows a top view of the metasurface of Fig. 8a, with illumination by light polarized at 45°. Fig. 8c schematically shows a top view of the metasurface of Fig. 8a, with illumination by light polarized at -45°. It is illustrated that the colors of the patches 4 depend on the polarization angle, which is different in case “polaristions 1” and “polarization 2”.

Fig. 9 shows a larger darkfield microscopy image of the illuminated metasurface of a PUF label 1 according to another embodiment of the invention, the smaller picture show different versions of a section taken from a metasurface, each version being illuminated by light of different linear polarisation. The metasurfaces were fabricated a mixture of all three particle types (gold nanospheres, silver nanospheres, silver nanorods).

Fig. 10 shows a schematic view of an anti-counterfeit system 100, using the PUF label 1 shown in Fig. 1 or 5, for providing protection against forgery of an article 105. The article 105 carries the PUF label 1 and preferably also an information label 101 , which is a barcode 101 in the present case, but may be any other kind of information carrier, for example a QR code, written symbols, or even an RFID label. The PUF label 1 , and in particular in combination with the information label 101 , forms a sending instance, when the imaging apparatus 130 of a receiving instance 120 illuminates the PUF label 1 for receiving an image 110, in particular the pattern image as emitted by the PUF label 1. The information label 101 may also be omitted from the system. The imaging apparatus may include a darkfield microscope suitable to receive the colored pattern image from the PUF label. The pattern image is provided by the imaging apparatus in the form of image data to the computing device.

The information label contains link data including a remote address, in particular an URL, which allows accessing the reference image data being stored in a data storage 150 being remote to the information label and being accessible by the computing device via a data network, in particular the Internet. The computing device 140 is configured, in particular programmed, to compare the pattern image with the reference image and to decide if the pattern image corresponds to the reference image. Hereby, the evaluation program may be configured to take into account a tolerance, which allows that not all but only a major fraction of the color spots (=colored patches) of the pattern image correspond to the respective colors of the color spots of the reference image. Moreover, also a systematic deviation may be tolerated when performing a spot by spot comparison. A systematic deviation may be caused by a characteristic of an individual imaging apparatus, because color detection may be not 100% identical, even when using the same type of camera.

Fig. 11 a shows the steps of a method 200 for the fabrication of the PUF label according to the invention.

The method 200 for the fabrication of the PUF label according to the invention, comprising the steps:

• providing a substrate comprising a surface, (201)

• providing a pattern of binding patches on the surface, in particular by lithography, (202)

• optional: providing, in a layer (5) on top of each binding patch (4), at least one coupling molecule, in particular a polynucleotide-based nanostructure, the coupling molecule having one or more first binding sites for binding a nanoparticle, and preferably having one or more second binding sites for binding the coupling molecule nanostructure to the surface,

• providing a random arrangement of nanoparticles, in particular plasmonic nanoparticles, on top of each binding patch, the nanoparticles being bound to the binding patch directly or via a coupling molecule, (203)

• optically detecting the pattern as a pattern of randomly colored patches, the color of each colored patch being determined by the random arrangement of nanoparticles of the respective binding patch. (204)

Fig. 11b shows the steps of an anti-counterfeit method 300 for examining the originality of an article being provided with a PUF label according to the invention. The anti-counterfeit method 300 for examining the originality of an article being provided with a PUF label according to the invention, which, in particular, uses the anti-counterfeit system according to the invention, comprising the steps:

• providing a PUF label of any of the claims 1 to 9, being assigned, in particular connected, to an article, (301)

• providing an information label being assigned, in particular connected, to the article, the information label providing access to reference image data, which represent a reference image of the optically detectable pattern of randomly colored patches of the PUF label, (302)

• optically detecting and providing a pattern image of the pattern of randomly colored patches of the PUF label by way of an imaging apparatus, (303)

• comparing and detecting, by way of a computing device, a match or a mismatch of pattern image data, which correspond to the pattern image, and the reference image data, for detecting forgery or originality of the article. (304)

If necessary, the overall shape and size of the polynucleotide nanostructure can be designed to fit the needs of specific PUF label.

The first and optional second binding sites may expose single stranded polynucleotides. From our experimental data, the following sequences are preferred:

First binding site: 5’ - AAA AAA AAA AAA AAA AAA AA - 3’

Reverse complement (on nanoparticle): Thiol - 5’ - TTTTT TTTTT TTTTT TTTT - 3’

Second binding site: 5’ - ATG TAG GTG GTA GAG AA - 3’

Reverse complement (on surface): Modification - 5’ - TTCTCTACCACCTACAT - 3’

Preferred modifications: thiol or biotin or succinimidyl ester (NHS), or azide or amine or amino

Embodiments of the invention, in particular the embodiments of Figs. 8 to 14, may be implemented using any or all of the following conditions, which may encompass materials, methods, regarding the polynucleotide-based nanostructure, polynucleotide-modified colloidal particle labels, colloidal-particle conjugates, the substrates, all being composed according to aspects of the present invention: i-based nanostructure:

Materials:

Polynucleotide scaffold (M13mp18 based, nucleotide length N preferably between 7249<N<8634)

Polynucleotide staples, pooled

Folding buffer (e.g. 1x Tris-EDTA buffer, pH 7.5 containing 20mM MgCI2 and 5mM NaCI)

Storage buffer (e.g. 1x Tris-EDTA buffer, pH 7.5 containing 5mM MgCI2 and 5mM NaCI)

Methods:

The polynucleotide nanostructures are prepared as follows: the polynucleotide scaffold is mixed with the pooled polynucleotide staples in molar excess (preferably between 5 and 100 times molar excess) and folding buffer. The mixture is heated to 65°C for 15 min and subsequently cooled down to 25°C over the course of preferably 1 hour to 16 hours. The assembled polynucleotide nanostructures are subsequently purified from the excess of polynucleotide staples via size exclusion HPLC. After purification, the buffer is exchanged and the concentration adjusted to the desired value (between 1 to 1000 nmole/liter) via ethanol precipitation or spin filtration. i-modified colloidal metal labels:

Materials:

Colloidal metal nanoparticles of desired diameter, preferably 40 nm

Thiol-modified polynucleotides

Washing buffer, e.g. 0.02% SDS in MQ-water

Methods: Metal nanoparticles are mixed with a molar excess of thiolated DNA and frozen at -20°C for5h. After thawing, samples are washed by iterative centrifugation (5x) and resuspended in 0.02% SDS. In the last step, the pellet is resuspended in a small volume of 0.02% SDS to concentrate the particles. The concentration is determined by UV-VIS spectroscopy at 520 nm.

Buffers used:

Placement buffer. 40mM Tris-HCI, pH 8.35, 40mM Mg²⁺.

Tween buffer. 40mM Tris-HCI, pH 8.35, 40mM Mg²⁺, 0.07% Tween20.

Gel buffer. 40mM Tris base, 20mM acetic acid, 1 mM EDTA, pH 8.3, 11 mM Mg²⁺.

Materials and equipment used :

• Polystyrene nanospheres (diameter - 350nm, 400nm, 600nm, 1000nm) at 1 wt% concentration

• 10mm x 10mm coverslips

• Lobind tubes

• Tweezers

• Hotplate

• Plasma cleaner

• Ultrasonication bath

• Isopropanol

• HMDS (Hexamethyldisilazan, CeHi9NSi2)

• Custom 3D-printed stands to hold the coverslips at a 45° angle.

Substrate preparation and binding site fabrication:

1. Wash ~ 500uL polystyrene nanospheres (PsNs) by centrifuging, removing the supernatant and resuspending them in 50% Ethanol/water (v/v) three times.

2. Resuspend PsNs to 3.5 wt% (~ 150uL) in 50% Ethanol/water (v/v).

3. Mark a small scratch on the coverslips to ensure upright orientation in the event of them flipping over during the binding site fabrication process.

4. Sonicate coverslips in isopropanol for two minutes @ 100% power.

5. Place coverslips in a petri dish and plasma clean @ 50% power i.e. _ W, 5 min,

45sccm O2 flow. 6. Place the coverslips against the 3D-printed stand to maintain a reproducible 45° angle.

7. Drop 8uL of the PsNs suspension to a coverslip. The plasma treatment should render the coverslips hydrophilic and the PsNs suspension should spread over the whole surface.

8. Wait until the PsNs suspension dries completely, around 5 minutes.

9. Heat at 60°C for 5 minutes to dry the coverslips.

10. Place the coverslips in a petri dish and plasma clean @ 20% power i.e. _ W, 2 min, 45sccm O2 flow. This is to clean the surface again (so called ‘descum’ step) and render the surface hydrophilic again.

11 . Add 600uL of HMDS (in a small cup) in the desiccator. Place the coverslips on a plate above the cup, and deposit under a vacuum seal (40mbar) for 20 minutes.

12. Lift-off PsNs by immersing coverslips in DI water and ultrasonicating @100% power for 5 minutes.

13. Heat at 120°C for 5 minutes to stabilize the HMDS-treated surface.

PUF label/Metasurface fabrication:

Materials and equipment required :

• Patterned chips

• Placement buffer

• Tween buffer

• Purified DNA Origami

• Purified DNA-functionalized nanoparticles

• Petri dish

• Parafilm

• Tweezers

• Pipette and tips

• Kimtech wipes

• Ethanol

• DI water

DNA Origami Placement: 1. Place a small strip of parafilm inside a Petri dish.

2. Place the patterned chip on the parafilm. The hydrophobic parafilm often pushes the buffer droplet back onto the chip in the event of spillage.

3. Keep a moistened Kimtech wipe in the Petri dish to help maintain humidity during the incubation period.

4. Dilute the purified DNA Origami with Placement buffer in LoBind tubes.

5. Pipette ~ 60uL of the diluted Origami onto the chip. Incubate for 1 hour.

6. Wash 10x with the Placement buffer by pipetting 60uL fresh buffer onto the chip, mixing 2-3 times and pipetting 60uL off the chip.

7. Wash 5x with the Tween buffer by pipetting 50uL fresh buffer onto the chip, mixing 2-3 times and pipetting 50uL off the chip. Addition of the Tween buffer results in a change of the surface tension and spreading of the droplet. Incubate for 5 minutes.

8. Wash away the Tween buffer by washing for 5 minutes (~ 30 washes) with the Placement buffer by pipetting 80uL fresh buffer onto the chip, mixing 2-3 times and pipetting 80uL off the chip. The indication for the Tween being washed away is that the droplet returns back to its original shape.

If the aim of the experiment is to only place DNA Origami, for example for imaging or troubleshooting purposes, skip to step 15 for the drying procedure.

Adding nanoparticles:

9. Centrifuge the Purified DNA-functionalized nanoparticles. Remove as much supernatant as possible, and then vortex and sonicate the suspension to uniformly redistribute the nanoparticles.

10. Now take the desired amount of this nanoparticle suspension (usually 1-1 OuL) and dilute it with the Placement buffer.

11 . Take away 30uL of the Placement buffer on the chip. Add 60uL of the nanoparticle suspension in Placement buffer to the chip, mixing 2-3 times and leaving the liquid on the chip. Incubate for 1 hour.

12. Take away 30uL of the nanoparticle suspension on the chip. Wash 10x with the Placement buffer by pipetting 60uL fresh buffer onto the chip, mixing 2-3 times and pipetting 60uL off the chip. 13. Wash 5x with the Tween buffer by pipetting 50uL fresh buffer onto the chip, mixing 2-3 times and pipetting 50uL off the chip. Addition of the Tween buffer results in a change of the surface tension and spreading of the droplet. Incubate for 5 minutes. 14. Wash away the Tween buffer by washing for 5 minutes (~ 30 washes) with the

Placement buffer by pipetting 80uL fresh buffer onto the chip, mixing 2-3 times and pipetting 80uL off the chip.

Drying procedure:

15. Successively place the chip into 25%, 50%, 75% and 85% Ethanol/water (v/v) mixture for 10 seconds, 10 seconds, 20 seconds and 2 minutes respectively.

16. Air dry the chips on a Kimtech wipe.

Claims

Patent claims

1. Physically unclonable function (PUF) label (1), comprising a substrate (2) having a surface (3), a plurality of binding patches (4) arranged in a pattern (10; 11) on the surface, each binding patch (4) carrying a random arrangement of one or more nanoparticles (7; 7a; 7b; 7c; 7d; 7e), the pattern (10; 11) being optically detectable as a pattern image of randomly colored patches, the color of each colored patch (4) being determined by the random arrangement of the one or more nanoparticles (7; 7a; 7b; 7c; 7d; 7e) of the respective binding patch (4).

2. The PUF label of claim 1 , wherein one or more coupling molecules are provided in a layer (5) on top of each binding patch (4), the one or more nanoparticles (7; 7a; 7b; 7c; 7d; 7e) being bound to the coupling molecules (6; 6a; 6b; 6c; 6d; 6e; 6f), which are bound to the binding patch (4).

3. The PUF label of claim 2, wherein the coupling molecules are polynucleotide-based nanostructures (6; 6a; 6b; 6c; 6d; 6e; 6f), the polynucleotide-based nanostructures having one or more first binding sites (6”) for binding one or more nanoparticles (7; 7a; 7b; 7c; 7d; 7e).

4. The PUF label of claim 3, wherein the polynucleotide-based nanostructure (6; 6a; 6b; 6c; 6d; 6e; 6f) comprises one or more scaffold strands, in particular a DNA origami.

5. The PUF label of any of claims 1 to 4, wherein the nanoparticles are plasmonic nanoparticles (7; 7a; 7b; 7c; 7d; 7e), which are arranged such on top of the binding patch (4), in particular bound to the binding patch (4), that a plasmonic coupling (pc) of at least two of the plasmonic nanoparticles on top of a binding patch (4) is achieved, the color of each colored patch (4) being determined by the random arrangement of plasmonic nanoparticles (7; 7a; 7b; 7c; 7d; 7e) of the respective binding patch (4), in particular by the color of the individual plasmonic nanoparticles (7; 7a; 7b; 7c; 7d; 7e) and/or the plasmonic coupling between plasmonic nanoparticles (7; 7a; 7b; 7c; 7d; 7e) of the same binding patch (4).

6. The PUF label of any of the previous claims, wherein the pattern (10) has a hexagonal symmetry.

7. The PUF label of any of the previous claims, wherein the center-to-center distance (despatches) of the binding patches of the pattern is a value, in particular a constant value, between 100 nm and 5000 nm, preferably between 100 nm and 1000 nm, preferably between 200 nm and 600 nm.

8. The PUF label of any of the previous claims, wherein a minimum surface-to-surface distance between two of the plasmonic nanoparticles of the random arrangement is equal or less than the maximum diameter or maximum length of one of the plasmonic nanoparticles, and is, in particular, equal or less than 20 nm, preferably equal or less than 10 nm.

9. The PUF label of any of the previous claims, wherein the nanoparticles on top of a binding patch may have one or more of the following attributes:

• a randomly determined distance between vicinal plasmonic nanoparticles, in particular caused by a random occupation of predetermined first binding sites;

• a randomly determined orientation of non-spherical nanoparticles,

• a differing shape of nanoparticles,

• a differing size of nanoparticles,

• a differing material or material composition of nanoparticles.

10. An anti-counterfeit system for providing protection against forgery of an article, comprising a PUF label of any of the previous claims, being assigned to the article, an information label being assigned to the article, the information label providing access to reference image data, which correspond to a reference image of the optically detectable pattern of randomly colored patches of the PUF label, an imaging apparatus for optically detecting and providing a pattern image of the pattern of randomly colored patches of the PUF label, a computing device configured, in particular programmed, for using pattern image data, which correspond to the pattern image, and for comparing and detecting a match or a mismatch of the pattern image data and the reference image data, for detecting forgery or originality of the article.

11. The anti-counterfeit system of claim 10, wherein the imaging apparatus contains a darkfield microscopy device, preferably comprising a light source and/or a polarizer for polarizing the light emitted by the light source and/or the light received from the surface of the PUF label.

12. The anti-counterfeit system of claim 10 or 11 , wherein the information label a) contains the reference image data, or b) contains identification data for unambiguously identifying the PUF label, the identification data being usable for identifying the reference image data within a set of multiple reference image data, which each correspond to a unique PUF label; or c) contains link data including a remote address, which allows accessing the reference image data being stored in a data storage being remote to the information label and being accessible by the computing device via a data network, in particular the Internet.

13. The anti-counterfeit system according to any of claims 10 to 12, wherein the information label contains a digital signature.

14. Method for the fabrication of the PUF label of any of the previous claims 1 to 9, comprising the steps:

• providing a substrate comprising a surface,

15. Anti-counterfeit method for examining the originality of an article being provided with a PUF label of any of the previous claims 1 to 9, which, in particular, uses the anticounterfeit system of any of claims 10 to 13, comprising the steps: providing a PUF label of any of the claims 1 to 9, being assigned, in particular connected, to an article, providing an information label being assigned, in particular connected, to the article, the information label providing access to reference image data, which represent a reference image of the optically detectable pattern of randomly colored patches of the PUF label, optically detecting and providing a pattern image of the pattern of randomly colored patches of the PUF label by way of an imaging apparatus, comparing and detecting, by way of a computing device, a match or a mismatch of pattern image data, which correspond to the pattern image, and the reference image data, for detecting forgery or originality of the article.