WO2019183736A1

WO2019183736A1 - Nanoprinting-based security document and method for its authentication

Info

Publication number: WO2019183736A1
Application number: PCT/CH2018/000011
Authority: WO
Inventors: Patrick Galliker
Original assignee: Orell Füssli Sicherheitsdruck Ag
Priority date: 2018-03-27
Filing date: 2018-03-27
Publication date: 2019-10-03

Abstract

A security element (1) for a security document comprises a plurality of spots (2) having, at least in a primary direction, extensions (x) of less than 20 μm, in particular of less than 10 μm. The extension (y) along a secondary direction, which is perpendicular to the primary direction, can be larger. The security element (1) is scanned by means of a device having a larger resolution along the primary direction than along the secondary direction.

Description

Nanoprinting-based security document and method for its authentication

Technical Field The invention relates to a security document comprising at least one security element, with said security element having a plurality of spots. It also relates to methods for manufacturing such a security document and for testing its authenticity. Background Art

These days, ink-jet printers are capable of achieving printing resolutions sufficient to create graphical prints that are essentially flawless to a human observer. The resolution of ink-jet printers can largely mimic the results achieved with high-resolution contact printing techniques, such as flexography or gravure printing. Ink-jet printers can be used to counterfeit branded 2D artwork and, more severely, banknotes and other security documents.

In the case of ink-jet printing, it is the size, impact-spreading, and wettability of a droplet on the substrate that are the major limitations for what can be achieved in terms of details. For small droplets, another limitation is placement accuracy, which is generally worse than 10 pm. Better results can be achieved with technologies such as Aerosol Jet. However, due to its high cost, this technology has limited reach and is therefore only available in industrial environments.

Generally, even consumer inkjet printers can readily achieve sub 50 pm feature sizes and resolutions beter than 100 pm. This may already be sufficient to create brilliant artwork and trick end-users into accepting counterfeit products, banknotes or documents.

Hence, banknotes and other security documents often rely on the application of specialized inks. Such inks may be used to protect security documents either by creating effects that a human interrogator can authenticate using his own senses, or they may only be authenticated by the use of dedicated hardware. The sec ond type, so-called machine-readable security features (MRPs), may e.g. be based on NIR absorbing inks or fluorescent dyes.

The use of such special materials has so far been an effective way of protecting security documents from forgery, but since many of the physical characteristics of the employed materials are easily imitated, their security is now questioned. Furthermore, even though the special effects created by MRFs may not be readily visible by the human eye, their mere presence generally is, which generates constraints while designing the security documents.

Disclosure of the lnvention

The problem to be solved by the present invention is therefore to provide a security document and methods as described above that offer an improved security.

The problem is solved by the subject matter of the independent claims.

Accordingly, the invention relates to a security document comprising at least one security element, wherein said security element includes a plurality of spots. Each of said spots has. at least in a primary direction, an extension of less than 20 pm. A security document with such fine features in at least one direction is practi- cally impossible to print using printing methods available to the public.

Advantageously, the extension of the spots along the primary direction is less than 10 pm for even better security.

The spots are e.g. formed by dyes, which can e.g. include visible dyes, infrared-absorbing dyes, or fluorescent dyes.

The security element can be designed to be visually imperceptible, which makes it even less likely to be copied.

For example, the spots can have a coverage of no more than 10%, in particular no more than 1 % within each visible segment of the security element (with the term“visible segment” as defined below). This renders the security element hard or even impossible to see by the naked eye.

Along the primary direction, the spacing between neighboring spots is advantageously smaller than 1 mm, in particular smaller than 250 pm, in particular smaller than 50 pm, in particular smaller than 20 pm. Such a small spacing between plots is hard to achieve by conventional printing means.

On the other hand, the spacing between neighboring spots along the primary direction is advantageously larger than 1 pm, in particular at least 10 pm, which makes it possible to detect it by optical means.

To ease the machine-assisted detection of the security document, and in particular for aligning the detector in respect to the security document, the pri- mary direction can be parallel to at least one edge of the security document. In a“secondary direction”, which is perpendicular to the primary direction, the spots have an extension and/or a spacing that is larger than their extension and/or spacing in the primary direction, in particular at least three times larger than the extension and/or spacing in the primary direction. In other words, the resolu- tion of the features is higher in the primary direction than in the secondary direction. This is based on the understanding that it is sufficient to have high resolution along the primary direction for establishing that the feature is genuine, while a lower resolution in the secondary direction can make a detection of the security document easier.

Further, spots with an extension that is longer along the secondary direction than along the primary direction allow for an improved signal -to-noise ratio when they are scanned by a detector moving, relatively to the security element, along said secondary direction.

However, advantageously, the secondary extension and/or second ary spacing is smaller than 1 mm, in particular smaller than 0.5 mm, i.e. the security document does have some structure along the secondary direction too, which e.g. al- lows to encode information and/or to provide repetitive structures for redundancy. Also, such a structure can render the security element harder to see.

In one embodiment, at least some of the spots comprise several distinct, inked areas arranged in a row along the secondary direction perpendicular to said primary direction. This is based on the idea that elongate spots can be generated using several drops of inks deposited in a row.

Advantageously, the spots are arranged at points of a lattice having first and second lattice directions. Such a regular arrangement is easier to manufac ture and to detect.

The lattice can have a primitive cell (as defined below) containing more than one point for said spots. This makes the security element more difficult to imitate unless the counterfeiter has access to print heads having custom-tailored geometries.

For the same reason, the lattice may be non-perpendicular, i.e. said first and said second lattice directions are non-perpendicular. Advantageously, the an gle between said first and said second lattice directions is between 30° and 60°.

The second direction of the lattice is advantageously perpendicular to the primary direction, which makes it possible to manufacture the security element by relatively displacing a carrier (for receiving the spots) perpendicularly to an elon- gate print head with the print head having a plurality of nozzles. The security element can store information. For example, only a subset of the points of said lattice can be covered by said spots, in which case said subset of points defines information encoded in the security element.

In an advantageous embodiment, at least part of the spots are fluo- rescent. This allows detection of the spots by illumination with a shorter excitation wavelength while monitoring the emissions at a longer emission wavelength. Using suitable filters for detection, a high contrast ratio can be obtained at the emission wavelength, which makes it easier to detect the small spots.

Advantageously, the emission wavelength is larger than 400 nm, i.e. it is in the visible or IR for easier detection. In particular, it is in the IR range, i.e. above 750 nm to make it invisible to the eye. On the other hand, it is advantageously shorter than 1200 run for being detectable by silicon-based light sensors.

In another embodiment, at least part of the spots are IR absorbing and provide a detectable infrared contrast against their background. In that case, the reflectivity of said spots differs from the reflectivity of the background of the security element by at least a factor 2, and this should be the case for at least one wavelength > 750 nm, in particular for at least wavelength between 750 nm and 1200 n .

In yet another advantageous embodiment, the security element is ta pered towards its edges for rendering it harder to see. In that case, the security eie- ment comprises an edge region and a center region, wherein tire edge region has a width exceeding 100 pm, in.particular exceeding 500 pm, and wherein, within any visible segment of said edge region, the spots on their background have a coverage smaller than an average coverage of said center region.

The security element can also be used to encode grayscale infor- mation. In that case, the security document comprises at least two spots that differ from each other in size and/or amount of dye by at least 10%, in particular by at least 50%.

The invention also relates to a method of testing the authenticity of the above security document. This method comprises at least the following steps:

- Optically scanning, by means of a scanning device having, at least along said primary direction, a spatial resolution better than 20 pm, in particular better than 10 pm, at least a part of the security element in order to generate scan data.

- Analyzing said scan data for the presence of the spots of the security elements.

Advantageously, during the optical scanning, the security document is moved along a secondary direction perpendicular to the primary direction. The movement is relative, i.e. it is possible to move the scanning device, the security document, or both.

By moving the scanning device perpendicularly to the primary di rection, the movement can be comparatively fast while it is still possible to spatially resolve the structure of the spots along the primary direction.

The scanning device need only be optimized for high resolution in the primary direction, i.e. the spatial resolution of the scanning device along the secondary direction can be weaker (poorer) than the spatial resolution along the primary direction.

Advantageously, the scanning device comprises a linear array of sensor pixels and optics imaging the substrate onto said sensor pixels. During optical scanning, this linear array is advantageously arranged parallel to the primary direc tion. In other words, the spatial resolution of detection along the primary direction is provided by the spacing of the array of sensor pixels and the magnification of a focus ing optics part.

The scanning can, for the reasons above, be carried out for at least one wavelength that is

- above 400 nm, in particular larger than 750 ran, and/or

- below 1200 n .

The invention also relates to a method for manufacturing the above security element. This method comprises at least the following steps:

- Moving a carrier for the security element in a direction of move ment, relative to a print head having a plurality of print nozzles: The movement is rel ative in the sense that the carrier, the head, or both may be moving.

- Ejecting ink from the print head in order to print the spots.

The print head advantageously comprises at least one row of print nozzles. The direction of movement is transversal, in particular perpendicular, to this at least one row in order to generate the security element quickly.

The print head may comprise a first and a second row of print nozzles. As seen in the direction of movement (i.e. in a projection along that direction onto a plane perpendicular to that direction), the nozzles of said second row are placed between the nozzles of said first row. In other words, the nozzles of the two rows are staggered.

Advantageously, the primary direction of the security element is transversal, in particular perpendicular, to the direction of movement. In that case, the fine resolution along the primary direction can be maintained even if the movement is fast. Alternatively, the primary direction can be parallel to said direction of movement. In that case, very high resolution can be achieved by slowing down the movement and/or increasing the ejection frequency while the print head does not require to have closely spaced nozzles.

The carrier mentioned above can be one of the following:

- The carrier can be the substrate of the security document to be manufactured: In this case, the security element is directly printed onto the security document.

- The carrier can be separate from a substrate of the security docu- ment and atached to the substrate after the spots have been printed onto said carrier:

In this case, the security element is printed onto a separate part, such as a transfer foil, which is then applied to the security document.

The spots can, as mentioned, be arranged at the points of a lattice having first and second lattice directions. In that case, the method advantageously comprises the step of encoding information in the security element by covering only a subset of the points of said lattice by spots and/or by generating spots that differ from each other in size and/or amount of dye by at least 10%, in particular by at least 50%.

In particular, the encoded information is unique to the document (e.g. it can be derived from a serial number of the document, or it can be derived from from the name of a bearer of the document if the document is a personal document), or it is at least unique for each of several subsets of a plurality of documents (in which case it can e.g, encode the denomination of a banknote or the issuer of the document). This allows to increase the security and/or to use the security element for gaining information about the document.

Brief Description of the Drawings

The invention will be better understood and objects other than those set forth above will become apparent when consideration is given to the following detailed description thereof. This description makes reference to the annexed drawings, wherein:

Fig. la shows a first configuration of a security element,

Fig. lb shows a second configuration of a security element,

Fig. 2 illustrates possible placement errors,

Fig. 3 illustrates an embodiment for encoding information in the security element,

Fig. 4 shows an embodiment having elongated spots, Fig. 5 shows an embodiment where each spot of formed by several, distinct dots,

Fig. 6 illustrates an embodiment having a lattice with a primitive cell having more than one spot,

Fig. 7 shows a possible embodiment of the nozzle of a print head in sectional view,

Fig. 8 shows a printing device,

Fig. 9 shows a first configuration of nozzles on a print head,

Fig. 10 shows a second configuration of nozzles on a print head, Fig. 11 shows a third configuration of nozzles on a print head,

Fig. 12 shows a fourth configuration of nozzles on a print head,

Fig. 13 shows a fifth configuration of nozzles on a print head,

Fig. 14 shows a sixth configuration of nozzles on a print head,

Fig. 15 shows an illustrative embodiment of a security document, Fig. 16 shows a printed image of spots having varying density.

Modes for Carrying Out the Invention Definitions:

A“spot” according to the present invention is an inked area of any shape. It may e.g. be circular or elongate. Also, optionally, a spot may be formed by at least two inked subparts separated by non-inked gaps, in particular as long as the gaps are sufficiently small to make the subparts appear as a single spot in the detector.

A“security document” can be any document that is likely to be counterfeited. Advantageously, the term relates to banknotes, identity documents (such as passports and ID cards), checks, certificates, access documents, coupons representing monetary value, etc. It has a substrate that may be flexible or non-flexible, e.g. of paper, plastics or a laminate of such materials,

The term“visually imperceptible” designates a security element that is invisible to the unaided eye. Advantageously, a security element is considered to be visually imperceptible when no more that 10%, in particular no more than 5%, of the general population is able to see the security element by the unaided eye.

The term“distance” between two spots designates the distance be- tween the centers of the two spots.

The term“spacing” between two spots designates the distances between the closest borders of the spots. A“lattice” is a two-dimensional, repetitive structure of points. The lattice is a regular tiling of the security element by a primitive cell, with the primitive cell defining one or more points. The lattice has first and second non-parallel lattice directions, along which the primitive cell repeats itself. In this context, the primitive cell is the minimum area cell that can be found to generate the lattice. In a most sim ple embodiment, the primitive cell contains one point, in which case the points lie on a regular grid whose grid-lines correspond to the first and second lattice directions. Note: Advantageously, not all points of the lattice need to be occupied by a spot of the security element, but the points designate the possible locations of the spots.

The term“visible segment” designates any section of the security element having a size of 100 pm x 100 pm. The term“coverage” of a“visible segment” designates the percentage of area in such a segment that is covered by the inked area of the spots.

Introduction:

The security element described here is able to combine the subtle ties of specialized ink materials with the distinguished patterning capabilities of a high-resolution printing process. The printing technology under discussion, termed NanoDrip printing in the following, is conceptually very similar to inkjet printing but achieves an up to G000 times higher printing resolution. The security element is de signed to integrate the high resolution of NanoDrip printing into a simple geometrical layout that is digitally variable for any security document and optimized to the implied throughput restrictions, especially for banknotes, which are often produced at web speeds of up to 3 m/s.

The fineness of the security element allows to potentially render it invisible to the human eye, such that it does not interact with the majority of other security and design elements on the security document. It is therefore possible for the disclosed element to be integrated into almost any security document without the need for design adaptions.

The disclosed security element advantageously consists of a 2D lattice of individual, non-overlapping spots of deposited material.

All spots advantageously have an extension no larger than 20 pm, in particular not larger than 10pm, along at least a common primary direction. This primary direction is advantageously either parallel or perpendicular to an edge, in partic ular the longest edge, of the security document.

The direction perpendicular to the primary direction is called“the secondary direction”. The spots are advantageously made of a material that can be detected by an appropriate hardware detector with high contrast to the surrounding envi ronment.

Because of its small size, a single spot remains invisible to the hu- man eye, even if the material is made of an otherwise highly visible material, such as metal.

The material in its basic form, before being aggregated within the spot, is advantageously dispersible or dissolvable at very fine granularity inside a liquid, in order to make an ink for NarioDrip printing.

When dispersed or dissolved, the maximum size of solid species contained inside the liquid is advantageously smaller than 2 pm, preferably smaller than 500 run, most preferably smaller than 100 nm.

For example, the solid material contained within the ink can be flu- orescent, including fluorescent nanoparticles (i.e. quantum dots) or organic fluores- cent dyes.

The material contained within the ink can also exhibit electromagnetic absorption and/or polarization characteristics that are different from the respective characteristics of the surrounding background of the security element.

Any other form of material that creates contrast against its back- ground can be used as well, such as a magnetic or radioactive material, as long as this material it detectable with sufficient resolution. Sufficient resolution means that it is physically possible, using available sensor technology, to reproduce the real geometry with a spatial resolution of better than 2(3 pm, in particular better than 10 pm, at least along the primary direction.

Such high detection resolution can be easily achieved when using light as a means of detection. However, the wavelength of the employed light needs to be shorter than 10 pm. Advantageously, it is shorter than 2 pm and longer than 100 nm.

Hence, the used material needs to be selectively detectable at this particular range of wavelengths, most preferably the material is selectively detectable at wavelengths in the NIR range, i.e. 400 nm - 2500 nm, more preferably 750 nm - 2500.

Infrared wavelengths above 750 nm, in particular 750 nm - 2500 nm, are advantageous because such wavelengths transmit easily through soil, which is a regular pollutant on security documents. For example, the material can comprise quantum dots (QDs) that are excited at e.g. 750 nm and emit at e.g. 800 run. Fluorescent materials are of ad vantageous because they can be detected with very high contrast, making it possible to easily discern the small spots from their background.

In order for the spots to be environmentally stable and to improve their adherence to their substrates, the inks may be supplemented with additives. Such additives are made to be transparent to the detection wavelength, such that they do not interact with the detection process. For example, the additives may be UV-curable resins that can be cross-linked and hardened by UV light.

By being smaller than 20 pm, in particular smaller than 10 pm, along at least the primary direction, the spots can be distinguished by an appropriate sensor from spots created by an inkjet printer or even from the spots created by an Aerosol printer.

It is advantageous for the spots to be arranged within a 2D matrix at positions defined by the lattice points of a rectangular or oblique lattice. Such a lattice has first and second lattice directions.

In this case, the secondary direction (which is, by definition, orthog onal to the primary direction) is advantageously parallel to one of the lattice directions. In the following, it is assumed (without loss of generality) that the secondary direction is parallel to the second latice direction. The first lattice direction is only parallel to the primary direction of the security element if the lattice is a rectangular lattice. If the lattice is of the oblique type, e.g. with an angle between the first and second lattice directions between 30° and 60°, then the primary direction of the security element is not parallel to the first lattice direction.

Any lattice point of the lattice can either be covered or not be covered by a spot. It is defined that spots aligned along the first lattice direction are situated on the same "first lattice row", while spots aligned along the second lattice direction are situated on the same "second lattice row".

The situation of whether a lattice point is covered or not covered with a spot can be translated into a basic unit of digital information, i.e. a bit. It is understood that each of the disclosed security elements can be formed with a different arrangement of spots on its lattice points. For example, by creating a 2D matrix with 50 times 50 lattice points, it is possible to store 2.5 kfi of digital information, and by creating a 2D matrix with 50 times 100 lattice points, it is possible to store 5 kB.

As shown in Figs. 1 , 2, and according to their orientation within the lattice, the distance between the centers of any two spots along the primary direction is a constant value x or an integer multiple of said constant value x, and the distance between the centers of any two spots along the secondary direction (being perpendicular to the primary direction) is a constant value y or an integer multiple of said constant value y.

However, due to technical constraints, the position of the spots will 5 generally deviate from their intended positions given by the theoretical lattice points, as shown in Fig. 2. The maximum deviation of the spots from their intended position into any direction depends on technical factors of the NanoDri process.

For spots smaller than the detection resolution along the primary or secondary direction (i.e. the spot appears larger than it actually is), the distance cho- _lo- sen between any two spots along the respective direction has to be larger than the detection resolution plus two times the maximum deviation achieved by the employed printing system.

For spots larger than the detection resolution along the primary or secondary direction, the distance chosen between any two spots along the respective is direction has to be larger than the spot size along said direction plus two times the maximum deviation achieved by the employed printing system.

This constitutes a minimum requirement if the spots are to be individually discemable via the output of the detection sensor. In case the detection resolution is different in the primary and in the secondary directions, the distance between 0 the spots has to be adjusted accordingly. For example, if the resolution is lower in the secondary direction than it is into the primary direction, then the distance between spots along the primary direction has to be smaller than their distance along the secondary direction.

Different sensor resolutions along the primary and secondary direc-₅ tion can originate if, for example, a sensor is used to image the 2D barcode on a fast- moving banknote. In this case, the effective resolution along the movement direction may be substantially lower than the resolution perpendicular to the movement direc tion due to the finite frame rate of the detector, i.e. due to blurring.

In general, the value of x, i.e. the distance along the primary direc-o tion, is best kept close to above-defined minimum distance in order to create a large amount of information within a given area while still being able to discern any neighboring spots from each other.

The intended value of x and/or the spacing between spots is advantageously smaller than 1 mm, preferably said value is smaller than 250 pm, and better₅ said value is smaller than 50 pm, even better smaller than 20 pm.

At least along the primary direction, the small size of the spots, and potentially also their distance from each other, are a characteristic for validating the authenticity of the security element. So in case the detection resolution is not the same along all directions, then the security element should be oriented on the security document in such a way that the primary direction of the security element aligns with the direction at which the effective resolution of detection is maximum.

For example, when detecting a security element on a banknote in an automatized manner, banknotes are regularly accelerated to high velocities of several meters per second either along their long edge or along their short edge. Hence, any details perpendicular to the moving direction of the banknotes can potentially be detected with the highest resolution provided by the imaging system (i.e. sensor and fo cusing elements) while details parallel to the movement direction are blurred due to the fast movement of the banknote and the finite frame rate of the detector. Therefore, the primary direction of the security element on a banknote should be chosen such that it is perpendicular to the (relative) movement direction of the banknote during authenticity verification, which is either along the short edge or along the long edge of the banknote.

If a security document is subject to the detection by more than one detector, wherein at least two detectors have their maximum resolution along different directions, it is possible to create at least two security elements on the banknote having different primary directions. The at least two different primary directions of the at least two security elements are chosen such that they align with the maximum detection resolution of the applicable detectors.

It is possible to densely arrange the spots along the primary direction while still being able to minimize visibility by using oblique lattice arrangements instead of rectangular lattice arrangements. In this case, it is advantageous to keep the density of spots along the same first lattice row larger than three times the diameter of the spot, more preferably the distance between any two adjacent spots on the same first lattice row is kept larger than ten times the diameter of the spot.

If the security element is formed with a rectangular lattice arrangement, this means that x needs to be larger than three times the size of the spot along the primary direction, and, preferably, x is larger than five times the size of the spot along the primary direction. However, if the security element is formed with an oblique lattice arrangement, then x needs to be larger than three times the size of the spot along the primary direction.

If the security element has its primary direction aligned with the maximum detector resolution, it may be that the size of the respective spots will be detected at an inferior resolution along the secondary direction than along the primary direction. For the example of a moving document, the maximum effective resolution of an optical detector along the secondary direction is not defined by the actual reso lution of employed optics and sensors, but it also depends on the inverse of the detec tor frame rate divided by the movement velocity of the banknote during detection, i.e. by how much the effect of blur can be reduced. As a result, it is may be impossible to validate the authenticity of a security element by detecting the size of a spot along the secondary direction. Hence, the value of y may be chosen larger than the value of x.

A value of y larger than a value of x is advantageous since such an increase in the value of y can be favorable in terms of signal detection, i.e. to increase the signal-to-noise ratio (SNR). To increase SNR, spots may in fact be intentionally formed longer than 10 pm and even longer than 20 pm along the secondary direction.

The definition of a spot also includes geometries that are elongated along the secondary direction. Furthermore, the definition of a spot also includes geometries that consist of at least two separated subparts being symmetrically arranged around their latice point. If a spot is formed from at least two subparts, the distance between the centers of any two subparts of a spot is chosen equal as or smaller than a fraction of the value of y, that fraction being equal to the number of subparts that are contained within the spot.

Intentionally increasing the value of y can also have the benefit of reducing the visibility of the security element because less spots can formed within a given area.

In order to keep the security element essentially invisible, it is advantageous to prevent spots from becoming too densely arranged within a line or within a given area. Spots may become faintly visible especially if they appear cumu latively to the eye, meaning that multiple spots are arranged within a distance that is smaller than the resolution of the human eye, i.e. ~75pm. Visibility of the security element will occur particularly if material is densely arranged within an area. Therefore, the average material area-density of the security feature within an area segment of 100 pm x 100 pm should advantageously be less than 10%, preferably less than 1%.

Generally, if the value of y is much larger than the value of x, material may not be uniformly distributed over the whole area of the security element.

This means that certain regions of the security element can more easily reach visibility than other regions of the security element. This, on the other hand, can limit the minimum feasible value of x, for example. However, the security feature may be cov- ered by material more uniformly if spots are formed within an oblique lattice arrange ment than within a rectangular lattice arrangement, particularly when introducing an angle of 45°, or between 30° and 60°, between the first and second lattice direction. Since the detector resolution along the secondary direction is often insufficient to validate the high printing resolution of the NanoDrip process, machine- validation may be based only on assessing the size of spots along the primary direction. However, if detection is blurred due to fast movement of the security document, it is understood that the size of any non-circular spot along its primary direction is not solely defined by the width of that spot at its center position. Instead, the blurred image collected by the sensor will also reflect placement deviations, for example. Essentially, the size of the spot along the primary direction will be equal to the distance between the outermost parts of any subpart of the spot along the primary direction. As a result, the size measured along the primary direction of a printed deposit may have to be substantially smaller than the intended size of the whole spot, in order to offset for placement deviations.

The security element may also comprise sub-arrays. This may be particularly beneficial whenever a strongly angled oblique lattice forms the basis of the security element. The 2D matrices formed from such lattice arrangement will be stretched along the secondary direction. Hence, they require more space on a security document along their secondary direction than a security element that is based on a rectangular lattice arrangement. The size of a security element with a strongly angled oblique lattice can be reduced by creating several sub-arrays, wherein each sub-array only contains a fraction of the total number of second lattice rows, as compared to the full security element. The at least two sub-arrays are arranged parallel to each other along the primary direction of the security element, with zero offset, i.e. the first lat tice rows of the at least two sub-arrays are aligned with each other. The distance be tween the at least two sub-arrays is chosen such that, within the total security element, the distance between any two adjacent second lattice rows is constant. Within the total security feature, the spots belonging to the same first lattice row within a sub-array will also belong to the same first lattice row within the total security element.

Preferably, the spots are generated at an instance of time when specific information can be collected about the individual security documents. For bank- notes, for example, the security element is preferably created alongside with the same production step that yields the serial number of the respective banknotes, or at an even later time. This allows encoding the serial number of the individual banknote into the security element.

In practice, the production of security documents often takes place in a fast-moving process where the substrate quickly passes along a production line. For example, the production of banknotes takes place in a sheet-by-sheet fashion where the banknote substrates are moved by rotating rolls across different processing equipment, at a velocity of up to 3 meters per second. At such extremely high veloci- ties, the creation of accurately placed microscopic spots, with a different information content for each banknote, is challenging.

To create the security element, a NanoDrip print head may be em- ployed as disclosed in W02016120381. The ejection of the droplet from a given noz zle of such a NanoDrip print head is stimulated by an extraction electrode associated with said nozzle. Droplets ejected by the process are created in an electrohydrody namic process, where each ejected droplet carries a strong electrical charge. Once a droplet has been ejected, it will quickly enter a uniform electric field between the print head and an acceleration electrode below the substrate. The electric field accelerates the droplets towards the substrate. The electric potential applied to the liquid, the extraction electrode, and the acceleration electrode can follow the guidelines disclosed in W02016120381. For example, the liquid contained within the print head can be electrically grounded while the ejection of liquid is stimulated by applying a minimum electric potential to the extraction electrode. The electric potential applied to the extraction electrode in order to stimulate droplet ejection depends on the spe cific design of the print head and on the particular ink. As an example, given that noz zles with a diameter of 10pm are being employed, it may be sufficient to apply a voltage of approximately 200V to stimulate droplet ejection. Furthermore, for the print head disclosed in W02016120381, the extraction electrodes should be covered by an additional shielding electrode that may be operated at an electric potential similar to that applied to the extraction electrode. By increasing the voltage applied between the extraction electrode and the liquid, while keeping the electric potential of the shielding electrode constant, one can reduce the size of ejected droplets, generally by more than a factor of ten as compared to the diameter of the nozzle itself. A reduction of the droplet size to 2 pm may be achieved by increasing the electric potential at the extraction electrode by only ~35%, i.e. to -270V.

For a practical distance of 1.5 m between print head and substrate, the acceleration electrode, positioned in close proximity underneath the substrate, may be exposed to an electric potential of e.g. 4.7 kV in order to obtain a uniform electric field with a strength of about 3 MV/m. Due to the resulting uniform electric field, droplets with 2 pm diameter will be quickly accelerated to velocities above 10 m/s and reach the substrate in a straight trajectory, with about 55 m/s final velocity. Because of the small size of the droplet, the kinetic energy can be absorbed by viscos- ity without a major spreading or even splashing. As a result of the impact, the droplet may spread to a final size of approximately 3.5 pm. The initially spread droplet may leave behind a circular material deposit of approximately the same size, i.e. 3.5 pm. No mechanical stimuli are required when creating the droplets in that manner. The droplets are solely a result of electrically charging the liquid and pulling the charged liquid from the nozzle by an electric force until eventually a droplet is released. As ejected droplets are continuously accelerated, they can be placed onto the substrate with high accuracy, despite their small size. In the current example, the inherent placement precision at zero substrate velocity can be below 2pm. However, at the presence of movement-induced airflow, the absolute placement precision will be further reduced. In the case of printing onto banknote, with substrates moving at 3 m/s, the expected electric acceleration force is almost forty times higher than the average lateral force acting onto the droplets by a laminar airflow. Hence, over a distance of 1.5 mm it is expected that a droplet will be deflected by only 40pm. Furthermore, if airflow is laminar, such drag-induced deflection will be largely systematic which means that it only affects the absolute placement position but not the relative position between any two deposited droplets. In general, the absolute placement pre- cision is not as critical as the relative placement precision between individual impact positions.

To enable easy detection of the security element on the document, the absolute precision in forming spots on the substrate needs to be better than 10mm, preferably better than 5mm, most preferably better than 2mm.

Depending on the particular ink being used, ejection of droplets from the NanoDrip print head can take place at frequencies of more than 100 kHz but it any case it is preferable to keep the ejection frequency below 1 MHz in order to assure that successively ejected droplets do not excessively repulse each other.

The shape of an individually ejected droplet is not necessarily spherical but— depending on ejection conditions— it may also be elongated or even form a continuous jet. The jet or elongated droplets can maintain their shape until impact, which means that for a fast-moving substrate the ejected liquid from a single ejection cycle may not form a circular deposit but instead it may form a deposit that is elongated along the movement direction of the substrate, i.e. a line.

To form the disclosed security element on a moving substrate, the print head may be employed to form a pattern in a similar process as disclosed in WO2016169956. In essence, nozzles are formed at predefined positions on the print head, so to generate deposits on the substrate in relation to their position on the print head. For the purpose of creating the security element of the present invention, at least one row of at least two individually addressable nozzles may be arranged on the print head, wherein the orientation of the at least one nozzle row can be oriented ei ther perpendicularly or at least transversal ly to the relative movement direction of the substrate during printing. At every instance in time the at least two nozzles of a given nozzle row may be selectively activated, wherein any activated nozzles will eject a droplet. Because of the straight trajectory of droplets and their extreme placement precision on the substrate, the droplets will form material deposits that directly relate to the relative positions of the nozzles on the print head.

It is possible to control the relative position of the nozzles on the print head with an accuracy better than 1 pm if the print head is manufactured by conventional photolithography-based microfabrication.

By performing selective nozzle-activation at a constant main inter- val, a 2D matrix of spots can be printed in a purely digital manner.

As defined herein, the main interval directly relates to the distance between two spots along the direction of the substrate movement. The distance between two spots is given by the duration of the main interval multiplied by the movement velocity of the substrate. As an example, at a substrate velocity of 3 m/s and at a main interval of 20 ps, spots will be separated along the movement direction of the substrate by 60 pm. The main interval must be chosen such that spots are formed at a distance that correlates with either the value of x or the value of y, in which case either the primary or the secondary direction of the security element will be aligned with the relative movement direction of the substrate during printing.

If the movement direction of the substrate is aligned with the intended secondary direction of the security element, the at least one nozzle row may be oriented on the print head either perpendicular or at a less than 90° angle, preferably at a less than 80° angle, with respect to the movement direction of the substrate. If the orientation of the at least one nozzle row is perpendicular to the movement direction of the substrate, security elements with a rectangular lattice can be formed, and if the at least one nozzle row is angled, an oblique lattice will be obtained.

If only a single nozzle row is used, the distance between any two adjacent nozzles, measured along the intended primary direction of the security element, must be equal to the value of x.

If the value of x is smaller than the feasible production-resolution of nozzles on the print head (at a given angle), the number of nozzle rows can be increased and the required nozzles can be distributed between them (see Figs. 1 1, 13,

15).

If at least two parallel nozzle rows are formed in this way, the sepa- ration between any two adjacent nozzles of the same nozzle row, along the intended primary direction of the security element, will be increased to a value of x multiplied by the number of parallel nozzle rows. The nozzle rows can be formed such that the nozzles of the at least two nozzle rows are evenly staggered, i.e. if the nozzles of all nozzle rows are projected onto a single nozzle row, then the distance between any two adjacent nozzles of said projected nozzle row will be equal to the value x, The distance z between two parallel nozzle rows along the intended secondary direction of the security element can be chosen equal to the value of y. In this way the nozzles of any of the at least two parallel nozzle rows can be selectively activated at the same time wherein the at least two nozzle rows will create their spots on different first lattice rows (or, alternatively speaking, as the two spots of a primitive cell of the lattice). Hence, the nozzle row that will be reached first by the front of the moving substrate (called the“leading nozzle row” in the following) will always create spots on a given first lattice row first, while the remaining spots of that first lattice row are successively printed in the following main intervals by the following one or more nozzle rows. If a number of m parallel nozzle rows are formed, then m main intervals are required to complete all spots of a first lattice row.

Importantly, while the leading nozzle row will create the spots of the first first lattice row of the security element (in direction of substrate movement), the remaining nozzle rows will still be deactivated until the first first lattice row of the security element will reach their position along with the substrate movement as well. According to the same logic, the leading nozzle row will be the first to be deactivated once it has created its spots on the last first lattice row of the security element·, while all following nozzles will only be deactivated in one of the following main intervals.

When printing a security element that has its secondary direction aligned to the movement direction of the substrate, one may use the nozzle to either eject spherical, elongated, or jet-like droplets (as explained above), as long as the droplets form deposits in such a way that the elongated part is deposited in a straight line parallel to the movement direction of the substrate. Furthermore, the main interval needs to be chosen such that the distance between two spots along the movement direction is equal to the value of y.

However, as explained above, in certain situations it can also be useful to create spots having several subparts. The subparts can be created by activating a given nozzle at least twice during each main interval. This means that if a noz zle is destined to be activated during a given main interval, then said nozzle will eject not one but several droplets during that main interval. These "sub-intervals" are cho- sen such that droplets are formed at distances on the substrate that are equivalent to the intended distance of the respective subparts along the secondary direction. Furthermore, the multiple liquid ejections during a main interval have to be executed such that the center of the respective spot coincides with the intended lattice point. This implies that the ejections of any two sub-intervals of the same number (e.g, the first sub-intervals of subsequent main intervals) need to be separated by a duration that is equivalent to a multiple integer of the main interval.

When printing a security element that has its secondary direction aligned to the relative movement direction of the substrate, and if such security element consists of several sub-arrays, the position of nozzles on the print head may simply be formed by treating each sub-array as an individual security element.

When printing a security element that has its primary direction aligned to the relative movement direction of the substrate, the at least one nozzle row is advantageously arranged on the print head perpendicular to the movement direction of the substrate. Each nozzle row contains a minimum number of nozzles that is equivalent to the number of lattice points defined along the secondary direction of the security element. However, the number of nozzles within a nozzle row may be increased to an integer multiple of said minimum number in order to be able to form an equivalent integer multiple of subparts to each spot.

All of the nozzles destined to print subparts to the same spot can always be activated at the same time, and hence there is no need for these nozzles to be individually addressable. Rather, all nozzles destined to print subparts of a mutual spot can receive the same electrical signals and thereby form a "nozzle group".

Two adjacent nozzles of the same nozzle group are separated from each other, along the intended secondary direction of the security element, by a distance that is equal to the intended distance of subparts within the spot. The centers of any two nozzle groups of the same nozzle row are situated on the intended lattice points, i.e. they are separated from each other, along the intended secondary direction of the security element, by a multiple integer of the value of y.

When printing a security element having its primary direction aligned to the movement direction of the substrate, the nozzles are best operated such that they produce individual spherical droplets. The ejection of spherical droplets as- sures that the spot will be formed as a single circular deposit, i.e. with minimum pos sible size along the intended primary direction of the security feature.

As the lattice points along the primary direction are generally chosen denser than the lattice points along the secondary direction, the main interval of ejection can be very short if the movement direction is aligned to the primary direc- tion of the security element.

If the ejection of a certain ink is not possible at the required main interval, one may arrange at least one more nozzle row in parallel to the first nozzle row, In this case, each nozzle of the at least two nozzle rows will be selectively activated during an extended interval. The duration of the extended interval is chosen as an integer factor longer than the main interval, wherein the integer factor is equivalent to the number of parallel nozzle rows that have been formed. Essentially, any two spots of neighboring lattice points along the primary direction can then be printed by nozzles belonging to two different nozzle rows.

The at least two parallel nozzle rows may also be formed on the print head at an offset from each other, along the orientation of the nozzle rows, i.e. the intended secondary direction of the security element. The offset between the at least two nozzle rows should be chosen such that the nozzle rows are equally staggered. Hence, if the nozzle rows are all to be projected onto a single nozzle row, the distance between any two adjacent nozzles will be constant. To achieve such equal staggering, the offset value between any two adjacent nozzle rows is equal to a frac tion of the value of y, the fraction being equal to the number of parallel nozzle rows

Nozzle rows defined in this way can create a security element that consists of several sub-arrays (i.e. primitive cells with several points) with an oblique lattice arrangement. The number of second lattice rows contained within each sub-array is equal to the number of parallel offset nozzle rows. The distance z along the in- tended primary direction between any two parallel nozzle rows can be equal to the value of x. In this way, the selective activation of nozzles from all. nozzle rows can be executed at the same time, and during each main interval a single sub-array is created.

When printing a security element that has its primary direction aligned to the relative movement direction of the substrate, an oblique lattice arrangement without sub-arrays can only be achieved if the number of parallel offset nozzle rows is equal to the number of second lattice rows contained within the security element. However, security elements with oblique lattice arrangements and several subarrays are generally more desirable than a single array only, particularly because more individually addressable nozzles are required on the print head.

If several nozzle rows are to be formed on the print head, indep end- ent of whether the relative substrate movement is in primary or secondary direction, the main interval of the different nozzle rows may be intentionally offset, i.e. the selective activation of nozzle rows does not take place at the same time. This allows nozzles to be operated by use of the double-actuation principle as disclosed in WO2016169956. According to the double actuation principle, each nozzle of a nozzle row will only eject droplets through mutual activation of two different signals, one being a signal dedicated to a full nozzle row, and another signal being dedicated to an individual nozzle. The signals dedicated to individual nozzles of a nozzle row are interconnected to the nozzles of other nozzle rows as well if all of the interconnected nozzles are situated at the same position within their respective nozzle row. If the main intervals between the nozzle rows are shifted, it is possible to activate only one nozzle row signal at a time. If only a first nozzle row signal is activated, then droplet ejection is only possible for nozzles of said nozzle row, even though the nozzles of other nozzle rows receive the same individual nozzle signals as well. However, once the first nozzle row signal is deactivated, the second nozzle row signal can be acti- vated, and hence ejection of droplets from said second nozzle row can be executed. The benefit of this technique is that individual nozzle signals can be shared between the nozzle rows, such that the number of nozzle rows scales like A+B instead of A-B, wherein A is the number nozzle rows, and B is the number of individually addressable nozzles within each nozzle row.

When shifting the main interval of different nozzle rows, the dis- tance z between any adjacent nozzle rows is chosen differently from the above suggested values x, and y, respectively. Because if the distance is chosen exactly equal to the distance between secondary or first lattice rows, the nozzle rows cannot be acti vated at the same time and, hence, they cannot be operated at the maximum possible ejection interval. Instead, the distance z should be chosen as a factor of x or y, respec- lively, that is not equal to one. Rather, the factor is chosen as 1 + 1/U, where U is equal to the number of parallel nozzle rows. In this case, the selective activation between any two adjacent nozzle rows will be offset by a time difference that is equal to a fraction of 1/U of the main interval.

The spots created during each main interval will add up to form the 2D matrix. The size of this 2D matrix depends on the number of individually addressable nozzles / nozzle groups that are arranged within the at least one nozzle row and on the number of main intervals that are being exercised for each security element.

For example, if the print head contains 50 individually addressable nozzles that are arranged within a single nozzle row, one can cover 50 lattice points during each main interval, if 50 intervals are executed with this print head, a 2D matrix with 50 x 50 lattice points will be created.

As stated before these 50 x 50 lattice points essentially correspond to a digital storage of 50 x 50 bits, and hence the security element can be used to store specific information about the security document it protects. This variable infor- mation stored within the security element can provide a basis for further security-improvement, particularly if the link between the specific security document and the 2D matrix of spots (e.g. the serial number) is securely encrypted, i.e. if it cannot be understood without proprietary translation. Storing such information also allows to collect multiple data points at once (e.g. the serial number). Hence, there is no need for separate reading cycles to obtain these data points.

Like a QR code, the information within the 2D matrix of spots can be encoded with redundancy. This means that a large portion of the spots may eventually become unreadable without loss of information.

Furthermore, the security element may be printed at least twice onto the security document to create further redundancy. In this way, a failing security ele- ment can be compensated by its multiple copies.

It should be noted though that a security feature may fail in the sense that it becomes impossible to properly read its data, yet it may still be possible to verify its size-constraints (i.e. whether it has been printed by NanoDrip printing or inkjet printing). Hence, even a partially failing security element can still be validated in its most essential characteristics and thereby add to protection of the security document against forgery.

If there are several security elements distributed over the. whole se curity document, or a major part thereof, it will not be possible for a forger to cut iso lated parts of the security document and reassemble them with another security docu- ment.

As stated before, multiple security elements may also be arranged in different orientation to enable reading along different orientations.

As mentioned, the security element may encode a serial number in the case of a banknote or it may encode the name of the wearer in case of a passport.

The 2D data matrix may be created along the design guidelines of the well-known QR-Code or DataMatrix-Code. The information contained within the security element may also be formed according to other design guidelines known to those skilled in the art, or according to undisclosed proprietary design guidelines.

In any case, it is preferable to create the 2D data matrix of spots ac~ cording to design guidelines that allow for error correction.

The detection of the real size along the secondary may be suppressed.

The distance between spots in the primary direction may be increased beyond the smallest possible value because a larger distance between two spots may allow a sensor to collect a higher signal-to-noise ratio from said spot.

In one example, if a line sensor is being used, the intended value of one of the variables x or y is smaller than 1 mm, preferably said value is smaller than 0.5 mm and most preferably said value is smaller than 0. 2 mm. The intended distance of the other variable x or y is smaller than 10 mm, and preferably said distance is smaller than 5 m , and most preferably said distance is smaller than 2 mm.

Because the imaging sensor itself will be constrained to a maximum detection resolution, the spacing between any two spots cannot be infinitesimally small to be still discernible. If the spots are smaller than the resolution of the detection, then the distance chosen between any two spots needs to be the sum of the detection resolution and two times the maximum deviation achieved by the employed printing system. If the spots are equal or larger in size than the resolution of the detec- tion, the distance chosen between any two spots needs to be the sum of spot size and two times the maximum deviation achieved by the employed printing system and it has to be larger than the resolution of a detector system used to differentiate the two spots from each other.

The size of the spots is given by the size of ejected droplets. The size of the ejected droplets is preferably larger than 50 nm but smaller than 10 pm.

In addition, the size of the spots can be chosen smaller than the optical resolution of the detector that is used to visualize the spots. For example, if the op tical resolution of the detector is 8 pm, the size of the spots can be smaller than 8 pm.

The spots are best made of material that can be detected with high contrast with respect to its background, at a resolution of beter than 20 pm, in partic ular better than 10 pm. Such high-resolution, high-contrast detection can be achieved using electromagnetic radiation as a means of detection. In this case, the spots can be made of a material that interacts with said electromagnetic radiation differently than their background.

For example, the spots may be made of a material that is more ab sorptive than their environment, in which case, for monochrome radiation, the spots will appear darker than their environment. The spots can also be made of a fluorescent material that emits light of a longer wavelength than the excitation light. When filtering the excitation wavelength by an appropriate optical filter, only the Stoke shifted emission wavelength is detected, and hence only for the fluorescent material a signal is registered on the detector.

For a resolution better than 10 pm, the wavelength of the detected electromagnetic radiation needs to be shorter than 15 pm. This is due to the fact that only light with a wavelength of shorter than ~1S pm can be used to detect at a resolu- tion of better than 10 pm using conventional air-immersion optics.

Signal-to-noise ratio (SNR) is an important parameter when detecting fast moving objects. The longer a sensor pixel can collect information from one of the moving spots, the higher the respective SNR is. The distance over which a sensor pixel can collect information along the secondary direction is evaluated by projecting the size of the pixel from the image plane (i.e. the surface of the sensor) to the object plane (i.e. the surface of the high security on which the security element is contained). The size of said projected pixel along the secondary direction on the object plane corresponds to the collection distance of said pixel.

An imaging optics having a smaller magnification in the secondary direction and higher magnification in the primary direction can advantageously be used for improving SNR without affecting the resolution in the primary direction. Such imaging optics can e.g. be based on cylindrical lenses with different focal lengths.

Examples:

Fig. 1 shows two embodiments of security elements I, where microscopic spots 2 of circular shape are arranged at each lattice point of a rectangular lattice (a) and an oblique lattice (b). The lattice has first lattice rows 3 and second lattice rows 4. However, there is also defined a primary and a secondary direction of the security element 1, which indicate preferred orientations during detection of the security element 1. The primary direction is perpendicular to the secondary direction, and the latter is parallel to the second lattice rows 3. The primary direction of the security ele- ment is indicated with bold double-arrows.

As can be seen, the primary direction is only parallel to the first lat tice rows 3 for rectangular lattice arrangements.

When detecting the security element, the direction of maximum resolution of the detector is advantageously alighted with the primary direction.

The distance between two spots along the primary direction is indicated by x, and the distance between two spots along the secondary direction is indicated by y. D defines the extension of the spot 2 along the primary direction. D is smaller than 20 pm, in particular smaller than 10 pm.

Fig. 2 illustrates that, in practice, the position of spots 2 within the security element 1 is not perfect because the manufacturing process introduces a deviation from a perfect arrangement. The square boxes 5 formed around each lattice point indicate the maximum deviation in the spot 2 position from its ideal position.

As shown in Fig. 3, the disclosed security element 1 can be formed without every lattice point being covered by a spot 2. This can be used to encode information in the security element. Each lattice point is e.g. used to encode one bit. For example, a missing spot may indicate a bit to be zero, while the presence of a spot in- dicates the bit to be 1.

Fig. 4 shows spots 2 with elongated geometry along the secondary direction. The extension D', D", D" of such an elongated spot along the primary direc- tion is not necessarily equal to the width of the structure along its center point, as it is the case with a point. Instead, the size of the spot along the primary direction is defined by the distance between its outmost borders along the primary direction. While D', D", D'" needs to be kept smaller than 20 pm. in particular smaller than 10 pm, the size L of the spot 2 along the secondary direction can be larger than D and it can be larger than 10 pm or 20 pm.

In Fig. 5, each spot 2 consists of several sub-parts 2a, 2b, 2c, which are geometrically separated from each other. Here, the sub-parts 2a, 2b, 2c are made of circular structures. As it is the case with an elongated spot, the formation of a spot from several sub-parts 2a, 2b, 2c can lead to deviations of the spot size D', D" from the size of a sub-part, here being represented by the diameter of the individual circular structures.

Fig. 6 shows a security element 1 that consists of five sub-arrays la, lb, lc, Id, le, each of which is formed with an oblique lattice and equal values of x and y. Each sub-array la, lb, lc, Id, le consists of two second lattice rows 4 and four first lattice rows 3. The sub-arrays la, lb, lc, Id, le are parallel to each other along their second lattice rows 4, wherein the distance u between any two adjacent sub-arrays la, lb, lc, ld, l e is equal to the value of x. The dotted boxes indicate that, within the total security element 1 , the included spots are considered to belong to the same first lattice row 3 of their respective sub-arrays la, lb, lc, ld, le.

The security element of Fig. 6 can also be considered to consist of a lattice whose primitive cell contains two points to receive spots.

Fig, 7 shows a NanoDrip print head 6 according to prior art. The print head 6 is equipped with at least one nozzle 7. The nozzle is supplied with liquid 8. By applying a voltage between the liquid and an extraction electrode 9, a liquid meniscus 10 can be formed. If the applied voltage is large enough, the liquid menis cus 10 will eventually release a droplet 1 1. The droplet II can be smaller than the nozzle 7 itself. In order to improve the reliability of the printing process, an additional shielding electrode 12 can be formed, the shielding electrode 12 covering the extrac tion electrode 8.

Fig. 8 illustrates how an acceleration electrode 13 can be placed below a substrate 14 for accurately depositing droplets 1 1 on said substrate when using a print head according to Fig. 7. Print head 6 and acceleration electrode 13 are parallel and form a uniform electric field between each other. Droplets 11 ejected from the at least one nozzle 7 of print head 6 will be accelerated to high velocities such that even at large distances between substrate 14 and print head 6, in the order of a milli- 5 meter, are crossed quickly, and it is possible to deposit said droplets 11 with a relative accuracy of better than 10 pm with respect to each other. This means that the droplets 11 will be deposited onto substrate 14 in an arrangement that largely reflects the ar rangement of nozzles 7 on print head 6.

In order to eject and accurately deposit the droplets 11, a multi- io channel voltage source 15 is used for supplying signals via electrical channels 16a, 16b, 16c, and 16d. These include an electric liquid channel l6a, an electric extraction electrode channel 16b, an electric shielding electrode channel l6c, and an electric acceleration electrode channel 16d. Generally, liquid channel 16a is at electric ground, the extraction and shielding electrode channels are at similar values around 200V, and is the acceleration electrode channel is at several kilovolts. During deposition of the droplets 11, substrate 14 may be moving (indicated by a bold arrow).

Fig. 9 illustrates the process for forming security element 1 with a print head according to Fig. 7. The view is from the direction of the print head onto the substrate 14, with all parts of the print head besides the nozzles 7 and its extrac- ₂₀ tion electrode 9 faded out. In this way, one can see how the security element 1 is formed on a moving substrate 14. The direction of movement is indicated with a bold arrow. In this figure, the movement direction of substrate 14 is aligned with the sec ondary direction y of security element 1. The extraction electrode 9 is contacted by the respective electric channel 16a to the voltage source 15.

25 In this example, a security element 1 is created where each spot 2 is made of several subparts 2a, 2b, 2c. The nozzles 7 are arranged along one row per pendicular to the direction of movement of substrate 14, and the distance between any two adjacent nozzles 7 is equal to the value of x. The nozzles 7 are activated during an interval that is shorter than the main interval of ejection. In this way, several circu ses lar deposits 2a, 2b, 2c are created for each spot 2, the circular deposits 2a, 2b, 2c being aligned along the direction of movement. Because all nozzles 7 are arranged along the primary direction, security element 1 is formed with a rectangular lattice.

In Fig. 10, security element 1 is formed of two sub-arrays, consisting of spots la, lb with spots 2 made of elongated structures, the spots 2 being ar- ₃₅ ranged within an oblique lattice. To create such a security element 1, the nozzles 7 and the respective extraction electrodes 9 are formed such that they are arranged along a row that is equally angled with respect to the direction of movement as the first lattice rows of the oblique lattice. To create two sub-arrays la, lb, lc, there are two rows of nozzles 7. Along the direction of movement, the two rows of nozzles 7 have no offset with respect to each other.

Fig. 11 shows how a security element 1 can be formed from nozzles 7 arranged within two rows, the two rows being offset along the direction of move- ment. This allows creating denser patterns of spots 2 in a direction perpendicular to the direction of movement as compared to a print head having its nozzles 7 arranged within a single row only. The nozzles 7 that are first to reach the moving substrate 14 are the first ones to print three spots of the security element 1, and once the same first lattice row 3 is reached by the nozzles 7 of the second row, another three spots 2 are printed. The distance between and two nozzles 7 is two times larger than the value of x.

Fig. 12 shows how to print a security element 1 that has its primary direction aligned with the movement direction of substrate 14. Here, each spot 2 is created from two subparts 2a, 2b. To create the two subparts 2a, 2b, two nozzles 7 have to be formed for each first lattice row 3 of the security element 1. The two nozzles 7 are symmetrically arranged with respect to their associated first lattice row 3, which means that the centers formed between the two nozzles 7 associated with two adjacent first lattice rows 3 are separated by a value that is equal to y. Nozzles 7 associated with the same first lattice row 3 are connected to the same electrical channel 6, and hence they are not addressable separately from each other. Their separation from each other is equal to the value yi.

Fig. 13 shows how two rows of nozzles 7 are used while printing onto a substrate 14 that moves along the primary direction x of security element 1. A second row of nozzles 7 is used in order to increase the interval of selective ejection of any nozzle to an extended main interval. In this example the extended interval is twice as large as the main interval because nozzles 7 are separated along the movement direction of the substrate 14 by a value that is equal to x. In this way, all six nozzles 7 can be selectively activated at the same time at an interval that is equal to the extended main interval. The two nozzle rows are offset from each other along the secondary direction, and therefore sub-arrays la, lb, lc with an oblique lattice arrangement are created. The sub-arrays la, lb, lc have two second lattice rows 4, which is equal to the number of nozzle rows. Six electric channels 16a are required, one to address each nozzle 7.

Fig. 14 shows how the number of electric channels 16a may be reduced by employing extraction electrodes 9’ that work according to the double actuation principle disclosed in WO 2016/169956. Here, only three electric channels 16a 23

are required for addressing the individual three nozzles 3 of each row because the same electric channel 16a is connected to the two nozzles 7 that are at the same posi tion within their respective row. To still be able to individually address these nozzles 7 separately, an additional two electric channels 16a’ are connected to the extraction 5 electrodes 9⁵ of all nozzles 7 of the same row. Ejection of droplets form a nozzle 7 only occurs if both electric channels 16a, 16a’ of a given extraction electrode 9’ are activated. Hence, at an instance of time only one row channel 16a’ can be activated, while the individual nozzles 7 are addressed by their respective electric channels 16a. Because the nozzles 7 of two rows cannot be activated at the same time, the distance io z has been increased to 1.5 times the value of x. In this way the selective activation of the nozzles 7 belonging to the different rows is shifted by a duration that is 0,5 times the main interval. To individually control each nozzle, only five electric channels 16a, 16a’ are required. This is only one less than in Fig. 13, but the efficiency of this method strongly increases with an increasing number of rows and contained nozzles, is The largest benefit is achieved for square arrangements.

Fig. 15 shows a security document 17 having several security elements 1. The arrows within the security elements 1 indicate the secondary directions of the security elements. The secondary direction of each security element 1 is generally preferred as a direction for reading the security element 1 with a detector.

zo The security document also contains visual elements 18a, 18b.

These can include non-specific visual elements 18a that are common to a series of security documents. They can also include specific visual elements 18b that are unique for each security document. Each security document 17 may obtain its own specific visual element 18b, e.g. a serial number or the name of its holder. The information 25 contained within the specific visual element 18b can be encrypted inside the security element(s) 1.

The security element(s) 1 can be placed at any position on the secu rity document 17, even on top of existing visual structures 18a, 18b, because the secu rity element 1 is not visible to the human eye and does interfere with the other visual elements 18 a, 18b.

Fig. 16 shows a gray-scale image being printed of spots having different size and/or amount of dye. This illustrates how information can be encoded into the security element by varying the density or weight of the spots. Such information can e.g. be a grayscale image, or it can be a machine-readable code.

Notes: Printing and scanning of the security element are best achieved by moving the print head or scanning device in respect to the security element. In such context, the term“movement” is always understood to imply a relative movement. Hence, when it is e.g. stated that the print head is moved, it may be that the print head is actually moving while the security element is at rest, or the print head is at rest while the security document is moving, or both of these parts are moving.

While there are shown and described presently preferred embodiments of the invention, it is to be distinctly understood that the invention is not limited thereto but may be otherwise variously embodied and practiced within the scope of the following claims.

Claims

1. A security document comprising at least one security element (1), wherein said security element (1) comprises a plurality of spots (2), wherein said spots (2) have, at least in a primary direction, extensions of less than 20 pm, in particular of less than 10 pm.

2. The security document of claim 1 where said security element (1) is visually imperceptible.

3, The security document of any of the preceding claims wherein, within any visible segment of said security element (1), said spots (2) have a coverage of no more than 10%, in particular no more than 1%.

4. The security document of any of the preceding claims wherein, along said primary direction, a spacing between neighboring spots (2) is smaller than 1 mm, in particular smaller than 250 pm, in particular smaller than 50 pm, in particular smaller than 20 pm

and/or at least 1 pm, in particular at least 10 pm.

5. The security documents of any of the preceding claims wherein said primary direction is parallel to at least one edge of said security document.

6. The security document of any of the preceding claims wherein, in a secondary direction perpendicular to said primary direction, said spots (2) have an extension and/or a spacing that is larger than the extension and/or spacing in the primary direction, in particular at least three times larger than said extension and/or spacing in said primary direction.

7. The security document of claim 6 wherein the extension of said spots (2) in said secondary direction and/or the spacing of said spots (2) in said secondary direction is smaller than 1 mm, in particular smaller than 0.5 mm.

8. The security document of any of the preceding claims wherein at least some of said spots (2) comprise several distinct, inked areas arranged in a row along a secondary direction perpendicular to said primary direction.

9. The security document of any of the preceding claims wherein said spots (2) are arranged at points of a lattice having first and second lattice directions.

5

10. The security document of claim 9 wherein a primitive cell of said lattice contains more than one of said points.

1 1. The security document of any of the claims 9 or 10 wherein said xo first and said second lattice directions are non-perpendicular, and in particular

wherein an angle between said first and said second lattice directions is between 30° and 60°.

12. The security document of any of the claims 9 or 10 wherein said is second lattice direction is perpendicular to said primary direction.

13. The security document of any of the claims 9 to 12 wherein only a subset of the points of said lattice is covered by said spots (2), wherein said subset of points defines information encoded in said security element (1).

0

14. The security document of any of the preceding claims wherein at least part of said spots (2) are fluorescent.

15. The security document of claim 14 wherein said at least part of 5 said spots (2) are fluorescent at an emission wavelength

above 400 nm, in particular larger than 750 nm, and/or

below 2500 nm, in particular below 1200 nm.

16. The security document of any of the preceding claims wherein a₀ reflectivity of said spots (2) for at least one wavelength > 750 run, in particular for at least wavelength between 750 run and 1200 run, differs by a reflectivity of a background of said security element (1) by at least a factor 2.

17. The security document of any of the preceding claims wherein5 said security element (I) comprises an edge region and a center region, wherein said edge region has a width exceeding 100 pm, in particular exceeding 500 pm, and wherein, within any visible segment of said edge region, said spots (2) have a coverage smaller than an average coverage of said center region.

18. The security document of any of the preceding claims compris- ing at least two spots (2) that differ from each other in size and/or amount of dye by at least 10%, in particular by at least 50%.

19. A method for testing an authenticity of the security document of any of the preceding claims comprising the steps of

optically scanning, by means of a scanning device having, at least along said primary direction, a spatial resolution better than 20 pm, in particular bet ter than 10 pm, at least part of said security document to generate scan data, and

analyzing said scan data for a presence of said spots (2).

20. The method of claim 19 wherein, during said optical scanning, said security document is moved, relative to said scanning device, along a secondary direction perpendicular to said primary direction.

21. The method of any of the claims 19 or 20 wherein a spatial reso- lution of said scanning device along a secondary direction perpendicular to said primary direction is weaker than the spatial resolution along said primary direction.

22. The method of claim 21, wherein said scanning device comprises a linear array of sensor pixels and optics imaging the substrate onto said sensor pixels wherein, during said optical scanning, said linear array is arranged parallel to said primary direction.

23. The method of any of the claims 19 to 22 wherein said optical scanning is carried out at at least one wavelength

above 400 run, in particular larger than 750 nm, and/or

below 2500 nm, in particular below 1200 nm.

24. A method for manufacturing the security document of any of the claims 1 to 18 comprising the steps of

moving a carrier (14) for said security element (1) in a direction of movement, relative to a print head (6) having a plurality of print nozzles (7), and ejecting ink from said print head (6) in order to print said spots (2).

25. The method of claim 24 wherein said print head (6) comprises at least one row of print nozzles (7) and wherein said direction of movement is trans versal, in particular perpendicular, to said at least one row.

26. The method of claim 25 wherein said print head (6) comprises a first and a second row of print nozzles (7), wherein, as seen in said direction of movement, the nozzles (7) of said second row are placed between the nozzles (7) of said first row.

27. The method of any of the claims 25 or 26 wherein said primary direction is transversal, in particular perpendicular, to said direction of movement.

28. The method of any of the claims 25 or 26 wherein said primary direction is parallel to said direction of movement.

29. The method of any of the claims 24 to 28 wherein said carrier

(14) is

a substrate of said security document or

separate horn a substrate of said security document and attached to said substrate after said spots (2) have been printed onto said carrier.

30. The method of any of the claims 24 to 29 wherein said spots (2) are arranged at points of a lattice haying first and second lattice directions and wherein said method comprises the step of encoding information in said security element (1) by covering only a subset of the points of said lattice by spots (2) and/or by generating spots (2) that differ from each other in size and/or amount of dye by at least 10%, in particular by at least 50%,

and in particular wherein said information is unique to said docu- ment or unique for each of several subsets of a plurality of documents.