Background of the invention

The present invention relates to an optical sensor.

The present invention further relates to method of manufacturing the same.

Optical sensors, such as time-temperature sensors, are proposed as a smart label that shows the accumulated storage history of a product, for example the time-temperature history of the product. Such sensors are commonly used in food, pharmaceutical and medical products to indicate previous exposure to (excessive) temperature. Such an exposure is visible as a change of color or other optical characteristic.

WO2010084010 discloses such an optical sensor. The sensor comprises a substrate and a polymeric layer comprising polymerized liquid crystal monomers having an ordered morphology, wherein the color, the reflectivity or the birefringence of the sensor changes due to a change of the morphology, wherein said change of the morphology is caused by physical contact with a chemical agent such as a gas or liquid, a change of

temperature, or passage of time.

It is a disadvantage of the known optical sensor that the change in morphology upon occurrence of a condition to be detected is relatively small. Therewith the visibility of the change in optical characteristic is limited.

DEFINITIONS For the purpose of this patent application, unless separately defined hereinbefore or elsewhere in this patent application, terms, including abbreviations used herein, have the meaning as defined hereunder,

'embossing' is the process of creating either raised or recessed relief images or designs in paper, synthetic or natural polymers and other materials, 'embossed film' refers to a film in which either raised or recessed relief images or designs have been created.

'sensor' and 'indicator' have the same meaning, i.e. indicating change.

'ΤΤΓ refers to both time-temperature indicator and time-temperature integrator.

'RM' refers to reactive mesogen, i.e. polymerisable mesogen.

'chiraP refers to non-superimposable on its mirror image.

'LCP' refers to liquid crystalline polymers.

'SCLCP' refers to side chain hquid crystalline polysiloxane.

'CLC or 'ChLC refers to cholesteric liquid crystal.

'LMWLC refers to low molecular weight (non-reactive) liquid crystal.

'cholesteric liquid crystal' refers to a liquid crystal with a helical structure and which is therefore chiral.

'SMP' refers to shape memory polymer.

'IPN' refers to interpenetrating polymer network.

'POM' refers to Polarized Optical Microscopy.

'elastomer' refers to a polymer with elastomeric behavior, i.e. entropy driven reversible stress-strain behaviour.

'nematic phase' refers to nematic hquid crystal phase characterized by molecules that have no positional order but tend to point in the same direction (along the director).

'isotropic phase' refers to disorder of molecules in all directions,

'switchable polymer' refers to a stimulus-sensitive polymer having shape memory or which is capable of undergoing transformation from one phase to another phase under influence of a stimulus, e.g. a change in temperature. Such may induce a transition from a cholesteric phase to an isotropic phase at a certain temperature.

'cholesteric phase' refers to the nematic state superimposed with a twist including the long axis of the molecules induced by the incorporation of a chiral group to give a helical twist to the orientation of the director,

'suitable' refers to what a person skilled in the art would consider technically required for the purpose, which is without undue burden technically feasible and for which no inventive effort or undue

experimentation is required to arrive at.

'optics' refers to the behaviour of visible, ultraviolet, and infrared light, 'photonic' refers to the practical application of optics,

'photonic time-temperature indicator' refers to an irreversible time- temperature photonic sensor.

'food sensor' refers to a photonic time-temperature indicator or other photonic indicators indicating other mitigating factors for food safety such as pH, humidity, C02, 02, etc.

'orthogonal polymer' refers to a polymer which has disparate physical or mechanical properties to the cholesteric liquid crystal polymer network and which does not interfere with the properties of the latter.

For the definition of other terms, not defined above or hereinafter, reference is made to pubhshed patent specifications and or published scientific papers including theses, in which such terms have already been defined. These can without undue effort be found on the internet.

Summary of the invention It is an object to provide an improved optical sensor that provides a stronger optical response to its exposure history. It is a further object to provide a method of manufacturing the improved optical sensor.

According to the invention a method of manufacturing the optical sensor is defined in claim 1. The improved optical sensor is defined in claim 10.

The claimed method includes at least the following steps.

preparing a composition comprising one or more reactive liquid crystal mesogens;

printing depositing the composition as an array of mutually non- touching array elements on a surface of a carrier;

curing the composition printed as an array of array elements therewith obtaining an array of array elements formed by the cured composition;

at a temperature above a glass transition temperature of the cured composition, exerting a pressure transverse to the surface to emboss the array of array elements;

allowing the array of array elements to cool down and releasing the transverse pressure.

In sharp contrast to the known method, the composition is printed as an array of array elements. A free space is defined between the array elements. In the subsequent embossing step the array elements can expand laterally within the free space. Therewith the array of array elements can be compressed to a much larger extent with a same pressure than is the case with a continuous layer. As a result the change in optical characteristics of the array is better visible than that in a continuous layer of the same composition.

As a further technical effect the printing of the composition as array elements has the effect that the molecules therein upon evaporation of the solvent align spontaneously in the direction required for inducing the optical response upon detection of a violated condition. Therewith so called alignment layers such as rubbed poly(imides) can be omitted. It is

conjectured that the alignment of the molecules is affected by flows in the liquid composition which are induced by evaporation of the solvent at the surfaces. When printing the composition as mutually non-touching array elements, in particular as circular symmetric array elements, the flows include also a lateral component. Therewith the lateral flows enforce a planar alignment of the molecules, resulting in an improved alignment and thereby improved optical contrast. This is not the case if the liquid is deposited as a continuous layer, in which case evaporation occurs solely from the top surface. Due to the confinement in this case the flows are more limited and only directed vertically. Moreover, when printing the

composition as identical mutually non-touching array elements, evaporation occurs reproducibly between the array elements themselves, resulting in consistent flows, molecular alignment, and appearance across the sensor dimensions..

In an embodiment of the method, allowing the array of array elements to cool down and releasing the transverse pressure comprises coohng down the array of array elements while maintaining said pressure, until the array of array elements has assumed a temperature below the glass transition temperature.

In practice it is sufficient if allowing the array of array elements to cool down and releasing the transverse pressure takes place simultaneously. For example, if the device that exerts the pressure is also the device that heats the array of array elements, the material of the array cools down almost immediately and is vitrified, before substantial shape (colour) change can occur. This facilitates a rapid production for example in a roll to roll process.

Dependent on a required thickness to be achieved each array element may be deposited as one or more layers of droplets of the composition. Typically the number of layers is less than 5, for example one or two layers depending on desired thickness e.g. in the range of 2-40 mu, preferably in the range of 5 - 20 mu.

In an embodiment a layer in an array element is deposited as a number of droplets in the range of 20 to 3000. A substantially lower number, for example less than 10 results in a larger variation of the dimensions of the array elements. It has been found that best results are achieved with array elements having lateral dimensions below 1 mm, or even below 0.5 mm, for example in the order of 0.2 or 0.3 mm. Therewith a substantially higher number of droplets, e.g. more than 5000 would not be favorable.

The claimed optical sensor may be configured as a time-temperature sensor. In that case the glass-transition temperature should be equal to the threshold temperature allowed for the product. In an alternative

embodiment, the cured composition used for the array elements may have a glass-transition temperature range so that a relaxation array elements to their original shape occurs at a relatively slow pace at temperatures in the lower side of the glass-transition temperature range and at a high pace in the pace at temperatures in the higher side of the glass-transition

temperature range. Therewith the morphological change is indicative for an integrated value of the temperature as a function of time.

In an embodiment the composition further comprises a secondary polymer that is compatible to the one or more reactive liquid crystal mesogens. In an embodiment, the secondary polymer is orthogonal and is not covalently attached to the cholesteric liquid crystalline polymer network, but physically interpenetrated therein. The secondary polymer is for example either crystalline in nature or liquid crystalhne or amorphous in nature. This polymer has a suitably low glass transition temperature, for example but not limited to a poly(dimethylsiloxane) or derivatives thereof. For different applications of these sensors one needs to be able to change the transition temperature easily and at will, without disturbing the properties of the cholesteric network (the liquid crystalline behaviour). By changing the material properties of this polymer (chemistry) or amount added

(volume fraction), the material properties of the IPN can be easily tuned.

In still another embodiment the composition used may include a further component that is sensitive to physical contact with a chemical agent such as a gas or liquid, for example a chemical agent selected from the group consisting of water (vapour), amines, sulfides, phosphides, CO, CO2, NO, NO2, oxygen.

This may be achieved in that the cured composition comprises breakable bonds that are specifically sensitive to the agent to be sensed.

Breakable bonds in this context are bonds that can be reversible broken and formed in the presence and the absence of a reagent. Breakable bonds can be secondary bonds like hydrogen bonds. But it can also be even weaker bonds based on dipole-dipole interactions or dip ole -induced dipole interactions between molecules. However the bonds must be strong enough to enable the formation of an ordered liquid-crystalline phase at the temperature range of interest. In this context, a secondary bond refers to a bond that involves attraction between molecules but that does not involve transfer or sharing of electrons. Another type of breakable bond that can be opened reversibly in the presence of a reagent are for instance primary bonds with higher binding strengths such as ionic bonds. In contrast, for example, primary bonds in the form of covalent bonds do not generally open reversibly. Covalent bonds are formed between atoms sharing electrons and are the building blocks for most organic compounds. They are stable and when opened with special reagents or temperature they seldom go back to the original structure. Secondary bonds such as hydrogen bridges are generally much weaker than primary bonds but are still strong enough to recombine when the conditions change and are therefore preferred in many embodiments. Ionic bonds are stronger but can be reversibly broken by a reagent that reacts even stronger with the cation or anion. Examples of such specifically sensitive breakable bonds are hydrogen bonds that can be broken by the presence of a base.

The cured composition comprising such a specifically sensitive breakable bond can be configured such that its glass transition temperature is relatively high so that upon occurrence of a morphological change it is immediately clear that it is due to the presence of the agent to be sensed. In that case a separate optical sensor may be used for monitoring the

temperature. This is advantageous in that is immediately clear which of the conditions was violated, which helps to improve the storage and transport facilities. Alternatively, the cured composition may have a glass transition temperature or a glass transition temperature range corresponding to the threshold temperature to be observed for the product. In that case a single indicator can be used for both purposes, and if the quality of the product is at risk for any of these causes it becomes visible.

The composition may optionally comprise a solvent to lower a viscosity of the composition so as to facihtate its deposition. Alternatively, or in addition a solvent may be selected to control other physical properties of the composition, such as its surface tension or its adherence to the surface of the substrate. The method may comprise a further step to cause or allow the solvent to evaporate subsequent to deposition of the array of array elements and before curing.

Upon embossing, a pressure transverse to the surface of the carrier is exerted. In practice the array elements have a semi-spherical profile in their height caused by the surface tension of the composition used. Due to the transverse pressure applied in embossing the largest transverse

deformation of the array elements during embossing occurs in their central portion, whereas a peripheral portion surrounding the central portion may be left substantially undeformed and a ring-shaped portion between the central portion and the peripheral portion is laterally expanded due to the material that is displaced outwards from the central portion during embossing.

The deformation caused by embossing the array elements is typically also visible in that the ring shaped portion shows a height gradient in a direction radially outwards that is higher than a height gradient within the peripheral portion or the central portion. I.e. in a direction radially outwards, the height first decreases slowly, than in the ring shaped portion rapidly decreases and in the peripheral portion radially beyond the ring shaped portion decreases more slowly.

The deformation caused by embossing is typically also clearly observable in terms of the optical characteristics of the array elements. In the initial state of the optical sensor, i.e. as long as the condition to be monitored is not violated, the specific deformation will have the effect that a reflection band of the central portion is blue shifted with respect to a reflection band of the peripheral portion. In contrast thereto, the specific deformation will have the effect that the intermediate ring shaped portion has a reflection band that is red shifted with respect to the reflection band of the peripheral portion. BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1A and FIG. IB schematically show a first and a second embodiment of a method according to the present invention,

FIG. 2 shows an exemplary set of components for use in a method according to the invention,

FIG. 3 shows another exemplary set of components for use in a method according to the invention,

FIG. 4 shows a further exemplary set of components for use in a method according to the invention, FIG. 5A, 5B, 5C show examples of array configurations in

embodiments of an optical sensor according to the invention,

FIG. 6A, 6B, 6C, 6D, 6E shows subsequent steps in the method according to FIG. 1A,

FIG. 7 A, IB, 7C are photographs of samples of the optical sensor according to the invention,

FIG. 7D schematically shows further aspects of the sample of FIG.

7A,

FIG. 8 shows a single array element in the sample of FIG. 7A in more detail in an image obtained by interferometry,

FIG. 9A, 9B and 9C are optical reflectance micrographs of arrays obtained with mutually different compressive forces in the embossing step,

FIG. 10A show from left to right optical microscopy pictures taken from an array element immediately before embossing, after embossing, and subsequent to heating the embossed array element to a temperature above its glass transition temperature,

FIG. 10B shows the corresponding SEM images,

FIG. IOC shows the interferometric profile of an array element according to a diagonal cross-section indicated in the leftmost part of FIG. 10A,

FIG. 11A, 11B, l lC respectively show POM micrographs obtained in reflection mode, obtained with 0°-90° crossed polarization, and with ±45° in crossed polarization respectively.

FIG. 12A, 12B, 12C show SEM Micrographs of three cross-sectioned array elements. Therein FIG. 12 A, FIG. 12B, and FIG. 12C respectively show the state immediately before embossing, after embossing and heated above the glass transition temperature,

FIG. 13 shows a time-temperature response of sample products. DETAILED DESCRIPTION OF EMBODIMENTS

FIG. 1A schematically shows a method of manufacturing an optical sensor. In a first step Si of the method shown in FIG 1A, and step S10 of the method shown in FIG. IB, a composition comprising one or more reactive hquid crystal mesogens is prepared. By way of example the one or more reactive liquid crystal niesogens may comprise the substances denoted as RM-1 to RM-3 and SCLP in FIG. 2. Further examples are the substances denoted as LC756, RM257 and RM105 in FIG. 3. FIG. 3 in addition shows components (a porogen 5CB, and a precursor for a secondary polymer) that are specifically used in the method depicted in FIG. IB.

FIG. 2 shows an example of a composition including an achiral side chain liquid crystal polymer (SCLCP) as a chiral dopant and further comprising various reactive mesogens (RMs), chiral and achiral. It was shown that the reflective wavelength of these cholesteric mixtures could be tuned by the concentration of chiral RM. Other components may be added to the composition such as a curing agent and a surfactant.

It was found that compositions comprising only cliacrylates as the reactive mesogens showed a reversible decrease in reflection band intensity. The level of decrease was found to be strongly dependent on the

concentration of diacrylate. To increase the temperature-response for coatings in the visible region a chiral diacrylate with a high helical twisting power (RM-3) may be included to fabricate coatings reflecting in the visible region with only 15 wt% diacrylate.

To facihtate processing, the composition may further comprise an organic solvent, such as tetrahydrofuran (THF) for example in a ratio of 1 unit of weight for the original composition and 1 to 1.5 weight units of the solvent THF. Also other organic solvents may be used such as toluene, dodecane, chlorobenzene, hexane etc. The organic solvent may be included to improve the fluid properties (viscosity mainly) for printing. In some cases, it is possible to print without a solvent, if the liquid crystal mixture has sufficient low viscosity. This may for example be achieved by printing at an elevated temperatures, in particular in a temperature range wherein the liquid crystal mixture assumes an isotropic phase. This composition and its use to obtain continuous layers is described in more detail in Appendix A of the previously filed, not yet published application PCT2017-070837.

In the example of FIG. 3, a chiral molecule (LC756) with

polymerizable end groups, for example 5 wt. % in the undiluted composition is used to induce cholesteric (or chiral nematic) liquid crystalline phase. The undiluted composition may further comprise one or more diacrylate reactive mesogens and one or more monoacrylate reactive mesogens.

More details on a photonic polymer material to be manufactured based on such a composition are published in Moirangthem et al, "Photonic Shape Memory Polymer with Stable Multiple Colors", ACS Appl. Mater. Interfaces 2017, 9, 32161-32167, DOI: 10.1021/acsami.7bl0198. As further discussed in this paper, a broad glass transition temperature Tg can be achieved by combining the cured composition with a second polymer in a semi-interpenetrating network. The secondary polymer as proposed therein is obtained with the precursor BA shown in FIG. 3. To enable the secondary polymer to form a semi-interpenetrating network, semi-IPN, a porogen (5CB for example) is added to the composition comprising the chiral molecule and the reactive mesogens.

In an example the composition may comprise 5 wt% of the chiral molecule (LC756), 35 wt % of the monoacrylate RM105 as a reactive mesogen, 29% of the diacrylate RM257 as a reactive mesogen, 30 wt % of the porogen 5CB and 1 wt% of the curing agent Ignacure 651. The formulation is heated into a (more) mobile hquid crystalline phase, therewith obviating a solvent for this purpose.

In the cited publication the composition is deposited as a continuous layer between two substrates, and subsequently cured. The curing step does not affect the porogen 5CB as it is a non-polymerizable substance. In a subsequent step the porogen is removed by extracting with an organic solvent, such as tetrahydrofuran. After this removal the continuous layer is treated with the precursor for the secondary polymer, e.g. benzyl acrylate (BA) monomer mixed with 1 wt. % of photoinitiator (Irgacure 651), and treated by UV-radiation to cause the formation of the secondary polymer from its precursor, therewith obtaining an interpenetrating network.

An overview of components as used in various embodiments of the invention is presented in Table 1 below.

Table 1: Overview of components.

Compponent Name Description Source

LC756 Chiral dopant BASF

SCLCP Chiral dopant

RM257 diacrylate type reactive mesogen Merck

RM-1 diacrylate type reactive mesogen

RM-2 diacrylate type reactive mesogen

RM-3 diacrylate type reactive mesogen

monoacrylate type reactive

RM105 mesogen Merck

5CB porogen Merck

BA benzyl acrylate precursor for secondary polymer Sigma Aldrich

Biosolve

THF Tetrahydrofurane solvent Equipments

Irgacure 651 Curing agent CIBA

As shown in FIG. 5A, 5B, 5C, the, optionally diluted, composition so obtained may then be printed, in a further step S2, using inkjet printing for example, to obtain an array of array elements 20 on a surface 12 of a carrier 10, for example a carrier of a glass or a polymer. The surface may be coated for example with a polyimide. It is noted that apart from inkjet printing also other printing methods could be used such as micro contact printing, flexoprinting, offset printing and screen printing. Also deposition methods other than printing are conceivable. However, inkjet printing is preferred as it allows for a very accurate dosing of the amount of the (diluted)

composition as a plurality of droplets. Also, dependent on a required thickness to be achieved each array element may be deposited as one or more layers of droplets of the composition.

In particular flexographic printing, also denoted as flexoprinting, is suitable to achieve a high speed production. For production at a lower speed but with a high accuracy, Inkjet printing is particularly suitable.

Typically it suffices to deposit one or two layers to achieve a desired thickness e.g. in the range of 5 - 20 mu. A layer in an array element may be deposited as a number of droplets in the range of 20 to 3000 for example.

The array elements 20 can for example have a square, a hexagonal or a circular base on the surface 12 of the substrate 10, as shown in FIG. 5A, 5B, 5C respectively. In practice a central portion of a top surface of the array elements, facing away from the substrate surface is substantially

hemispherical, independent of the base, due to a surface tension of the deposited liquid. This is schematically shown in the cross-section in FIG. 6A corresponding to the V-V in FIG. 5 A.

In case the deposited composition comprises a solvent, this is allowed to evaporate subsequent to deposition of the array of array elements and before curing. Dependent on the type of solvent evaporation is faster or slower. For a high speed process a relatively volatile solvent may be used, whereas otherwise a less volatile solvent may be more suitable.

As shown schematically in FIG. 6B, in step S4 the composition printed as an array of array elements is cured to obtain an array of array elements formed by the cured composition. Typically the printed

composition comprises a curing agent, so that the composition can be controllably cured. The curing agent will typically be UV-sensitive, so that the curing can be initiated by applying UV radiation. As an alternative it may be conceived to use a composition that is curable by heat or even without external activation. However, using a UV curable mixture is advantageous in that the composition can be stored for a long time, and in that it allows to rapidly initiate the curing process after deposition and optional evaporation of the solvent while avoiding that curing reactions take place when the composition is stored or deposited.

In step S5, illustrated in FIG. 6C, a temperature of the array of array elements is kept above a glass transition temperature of the cured

composition and a pressure is exerted to the array of array elements transverse to the surface of the substrate. Therewith the array of array elements is embossed. The fact that the cured composition is deposited as an array of array elements makes it possible to use a flat stamp. An example thereof, is a self-ahgning flat stamp 30 as shown in FIG. 6D. Therein the left portion shows a first side view and the right portion shows a second side view according to 6R in the left-half. This self-aligning stamp 30 exerts a uniform pressure with its bottom surface 32 as a proportional to a weight of a load placed on its top surface 32. Experiments are scheduled using a pneumatic press of type hpdSystem Nano-V available from KBA-Metronic GmbH. Also a calendar type system is feasible, wherein the intermediate product of FIG. 6C is forced between two spaced rolls.

In step S6, illustrated in FIG. 6E, the array of array elements, is allowed to cool down and the transverse pressure is released. One option is to maintain the pressure until the array of array elements is fully cooled down, for example by cooling the stamp. In practice it was found that the release of the heated stamp in an otherwise cool environment is sufficient to substantially maintain the deformed shape obtained with the embossing in step S5. In this way the process can be performed significantly faster, which is advantageous when implementing the method in a roll to roll process.

An example of the method is now described in more detail with reference to FIG. 4. A composition was prepared with the following specification 6OBA, 21.9 wt%; 6OBA-M, 21.9 wt%; LC756, 4.5 wt%;

C6BP(equivalent to EM 105), 38 wt%; C6M (equivalent to RM 82), 13 wt%; photoinitiator (Irgacure 369), 0.6 wt%; and 0.1 wt% of inhibitor (BHT, 3,5- di-t-butyl-4-hydroxytolueen) acting as a radical scavenger in order to prevent polymerization until this is initiated by exposure of the

photoinitiator to UV. The composition was diluted by adding 1.2 weight unit of THF to 1 weight unit of the prepared composition. More details of these compositions are described in Davies, D. J. D.; Vaccaro, A. R.; Morris, S. M.; Herzer, N.; Schenning, A. P. H. J.; Bastiaansen, C. W. M. Adv. Funct.

Mater. 2013, 23 (21), 2723.

The diluted composition was printed on a polyimide coated surface 12 of a rubbed glass substrate 10 using inkjet printing, in block patterns with a total area of 6 mm2, e.g. 6x lxlmm (FIG. 7A), 24x 0.5x0.5mm (FIG. 7B), and 96x 0.25x0.25mm (FIG. 7C). For this purpose a Fuji Dimatix printer (DMP- 2800) using 10 pL droplet size cartridges was used. Droplet spacing was set to 30 pm. The samples were printed around the isotropization temperature of the liquid crystal phase, 40C. By way of example the block pattern of FIG. 7A is shown in more detail in FIG. 7D. As shown in FIG. 7D, in this experimental setting, the array elements 20 are printed with a square cross- section, i.e. d2=dl= 1 mm for FIG. 7A, d2 = dl=0.5 mm for FIG. 7B and d2 = dl=0.25 mm for FIG. 7C. As discussed with reference to FIG. 5A-C the array elements 20 may alternatively be provided with another cross-section, e.g. circular, hexagonal, triangular, rectangular etc. It is further noted that the array elements 20 are disposed relatively far apart from one another in this experimental setting. I.e. for the purpose of this experiment an intermediate space between subsequent array elements is left that is of the same order of magnitude of the size of the array elements. In practice the intermediate space will be substantially smaller to achieve a high coverage of the surface that contributes to a visual response to changed circumstances. It should however be avoided that the array elements touch each other as this would change the morphology of the individual array elements.

Subsequent to deposition, the sample was placed on a hot plate of 80 °C for 5 min to accelerate evaporation of the solvent from the array elements. Alternatively, a slower evaporation of the solvent can be achieved by leaving the sample at the printing temperature (40 °C). Upon

evaporation of the solvent the remaining composition was photo- polymerized using UV radiation. For this purpose the sample was exposed to radiation of a high intensity EXFO lamp (Omnicure model series 2000) for 100 s in a nitrogen atmosphere. Optionally a post-curing step can be performed, for example 3 h at 130 °C under nitrogen atmosphere. In each case, it was observed that the array of array elements exhibit a green-yellow colour arising from the cholesteric reflection band. Importantly, the reflected colour is homogenous across the dimensions of the label, and between different arrays and sizes.

Closer inspection using interferometry reveals that the array elements have a spherical height profile, likely driven by the minimization of surface energy prior to curing. The results of this interferometric inspection are shown in FIG. 8.

The height of the array elements can be tuned by block size, droplet density as well as by the selection of the properties of the substrate polyimide type, as shown in Table 2 below, and was varied between 7 and 14 μηι. In Table 2, the first column specifies the applied pattern, wherein A denotes the pattern of FIG. 7 A, B denotes the pattern of FIG. 7B and C denotes the pattern of FIG. 7C. The type of coating used for the surface of the substrate is indicated in the second column, where PI 1 refers to a polyimide coating of type JSR Optmer AL24101 and PI2 refers to a polyimide coating of type JSR Optmer AL1051. The center to center droplet spacing as indicated in the third column determines the droplet density. The droplet spacing is close enough to allow the droplets to merge within the boundary of an array element. The fourth column indicates the average top height of the array elements 20 and the fifth column indicates the standard deviation in the average top height, both in micron. Table 2: Profile analysis of inkjet printed CLC-arrays. The averaged top height is characteristic for the profile, standard deviation is characteristic for the homogeneity of the profiles over the whole array.

As becomes apparent from the results presented in Table 2, the array elements are obtained with a high precision. I.e. the standard deviation of the height is only a few percent of the average height of the array elements. It was observed that the variation in top height across the surface is substantially lower than a variation in height across a surface of a

continuously deposited layer.

The arrays as specified with reference to table 2, were subsequently embossed to induce a shape memory effect having the accompanying colour shift. To this end, the arrays were heated above Tg, here 50 °C, compressed using a flat stamp, cooled below Tg, and the stamp was removed, as schematically illustrated in FIG. 6E. The stamp load was varied between 10, 20, and 40 kg, resulting in an approximate force of 1, 2, and 4 N per array element respectively. Using a 1 N force per array element, a slight colour shift was observed across the array due to the deformation of the array elements. Using a higher force resulted in a larger colour shift. By way of example optical reflectance micrographs of the arrays so obtained are shown in FIG. 9A, 9B and 9C for the array of pattern type C using the substrate type PIl are shown for an applied force of 1, 2 and 4 N respectively. In each case, the results obtained revealed an even color shift between the individual array elements towards the blue, indicating an even distribution of compressive deformation across the array of array elements. Such a regular and sizeable compression could not be achieved using a

homogeneous barcoated layer. In addition, the average indentation was obtained for each sample using profilometry . The results of this

measurement are shown in Table 3 below. Therein the first column shows the load applied to stamp expressed in kg. The second column shows the force per array element in N and the third column shows the average pressure in MPa, as determined by the force exerted per array element divided by the contact surface between the surface of the stamp and the an array element. The fourth column shows the maximum indentation that was achieved in %. I.e. this is amount with which the top height of the array elements is reduced as a percentage of the top height immediately before the embossing step. As shown in the fifth column of this table, in each of these cases a complete recovery of the array elements was obtained after a heating above the glass transition temperature.

Table 3: The effect of increasing force per array element during

. The maximum indentation was determined by interferometry for each square.

Interestingly, only 30 MPa compressive stress is needed for a relative vertical shrinkage of 8% of an array element, while at least 130 MPa is

25 needed to achieve the same (or less) effect on a homogeneous layer of the same cured polymer composition, using a spherical stamp. It is conjectured that the presence of the cured polymer composition as a pattern of array elements allows for a lateral displacement of the network, while such a lateral displacement is restricted in homogenous layers.

The photonic shape memory effect was investigated more closely on the level of individual array elements using optical microscopy, scanning electron microscopy (SEM), and profilometry.

The results are described in more detail with reference to FIG. 10A, 10B, IOC.

FIG. 10A shows from left to right optical microscopy pictures taken from an array element 20 immediately before embossing, after embossing, and subsequent to heating the embossed array element to a temperature above its glass transition temperature.

The optical microscopy measurement showed that the array element 20 upon embossing (shown in the middle of the FIG. 10A) has a central portion 20c with a reflection band that is blue shifted with respect a reflection band of peripheral portion 20p surrounding the central portion. In addition a ring shaped portion 20r between the central portion 20c and the peripheral portion 20p was observed with a reflection band that is red shifted with respect to the reflection band of the peripheral portion 20p. Upon heating above the glass transition temperature Tg, this pattern disappears.

FIG. 10B shows corresponding SEM images of the array element 20 in the three states. In each of the three states the array element 20 has a substantially square base. In the first state, directly before embossing the surface of the array element facing away from the surface of the substrate is substantially hemispherical. In the second state a substantially flattened central portion 20c is formed that is surrounded by a peripheral portion 20p, while a ring shaped portion 20r is formed between the central portion 20c and the peripheral portion 20p. In the third state, after heating the array element 20 above the glass transition temperature Tg, the shape of the array element is restored to its shape immediately before embossing.

Finally FIG. IOC shows the interferometric profile of the array element 20 according to a diagonal cross-section indicated in the leftmost part of FIG. 10A. Therein curve a shows the profile as determined in the state immediately before embossing, and curve b shows the profile as determined in the state after embossing. Curve a shows corresponds to the height profile of a substantially hemispherical shape. After embossing a flattened portion is formed in the middle of the array element. As can be seen therewith a central portion 20c is formed wherein the height is reduced in comparison to that of the height of the array element in the first state. In a peripheral portion 20p, the shape is substantially unaffected. It can further be seen that cured material from the central portion is displaced to a ring-shaped portion 20r. As a result thereof the height of the array is increased in this portion. The compression in the central portion 20c results in the observed blue shift of the reflected light therein. The vertical network swelling in the ring shaped portion causes the red-shift. As noted, upon heating the array element 20 above the glass transition temperature of the cured composition, the shape of the array element is restored to that before embossing. Accordingly curve a is also representative for the profile in the third state.

The volume of the array elements does not change significantly, rather, a simultaneous contraction and lateral expansion is observed. The lateral expansion is due to displacement of the CLC network. This was confirmed by polarized optical microscopy (POM) image data obtained with a Leica CTR 6000 optical microscope as discussed with reference to FIG. 11A, 11B and 11C. In these figures, the left and the right image in each row respectively show the array element in its state directly before embossing and after embossing at temperature of 75C with a pressure of 4N/array element. The array element is of the type printed in the 0.50 mm x 0.50 mm block pattern on a type I substrate wherein the embossing temperature 75C, pressure 4N/array element.

FIG. 11A shows the POM micrographs obtained in reflection mode, FIG. 11B shows the results obtained with 0°-90° crossed polarization, FIG. l lC shows the results obtained with ±45° in crossed polarization. The POM micrographs clearly shows that the embossing induces axisymmetric linear birefringence due to lateral network reorientation induced by compressive stress.

FIG. 12A, 12B, 12C show a SEM Micrographs of three cross-sectioned

CLC-squares based on 0.25x0.25 block pattern representing different stages. Pictures were taken near a central position within an array element.

Therein FIG. 12A, FIG. 12B, and FIG. 12C respectively show the state immediately before embossing, after embossing and heated above the glass transition temperature. The helical pitches in the embossed state were estimated using the solid white bars representing 2 μιη and thin dotted lines to guide the eye by averaging 15 layers. These SEM images confirmed the corresponding compression in helical pitch of the cholesteric between 8 and 11% from the original 310 nm in line with profUometry observations presented in Table 3.

The array of array elements was evaluated as a time temperature integrator label. The average hue (representative for colour) was extracted from the center of the array elements during various isothermal

experiments performed using an optical microscopy operating in reflectance mode. In this experiment different specimens of the label so obtained were exposed to mutually different temperatures in a range between 35 °C and 60°C, during a time interval of 60 minutes. The results are presented in FIG. 13.

Various alternative embodiments are possible, for example using an alternative method, depicted schematically in FIG. IB, the cured composition forming the array elements may comprise a secondary polymer as an interpenetrating network. In the alternative method of FIG. IB, steps S10, S20, optional step S30, S40, S50 and S60 substantially correspond to steps SI, S2, optional step S3, S4, S5 and S6 respectively.

An important difference between step S10 and Si is however that the composition prepared in step S10 includes a porogen, i.e. a non- polymerizable component, such as 5CB (See FIG. 3). When in step S40 the composition forming the deposited array of elements is cured, the porogen remains liquid. Subsequent to the curing step S40, the porogen is removed in step S42, for example by extracting with an organic solvent such as tetrahydrofuran. Therewith a porous matrix is obtained. In a further step S44 the porous matrix is filled with a precursor (e.g. BA in FIG. 3) for the secondary polymer. Then in a second curing step S46 the secondary polymer is formed as an interpenetrating network inside the porous matrix.

Subsequently the steps S50 and S60 can be applied in the same manner as in the steps S5 and S6 described with reference to FIG. 1A. It is noted that according to the cited publication of Moirangthem et al. the composition is deposited as a continuous layer on a substrate. According to the present invention, the composition is deposited as an array of mutually distinct array elements. In this embodiment the deposition in step S20 as an array of array elements in combination with the steps S42 and S44 have a synergic effect. I.e. dure to the deposition of the composition as mutually distinct array elements in step S20 the cured matrix containing the porogen has a higher contact surface than is the case with a continuous layer formed by the cured matrix. As a result the subsequent steps S42 of extracting the porogen and of S44 of treating with the precursor for the secondary polymer can take place more efficiently.

In another embodiment of the claimed optical sensor the cured composition forming the array elements is a semi-interpenetrating polymer network composed of a hquid crystal elastomer (LCE) and a liquid crystal network ((LCN > 15 wt%), which can be obtained with the components of FIG. 2. By way of example a first coating mixture Ml is obtained by mixing an achiral side chain LC polysiloxane (SCLCP) elastomer as the LCE with a chiral diacrylate dopant RM-1 (mixture Ml, 21 wt% RM-1). To ensure good mixing, RMs with a similar mesogenic moiety as the SCLCP elastomer were chosen. Furthermore, a photo initiator (Irgacure 651, 1 wt%) and a perfluoro surfactant (1 wt%) were added to be able to photo polymerize and align the mixture at the polymer air interface, respectively. The mixture may be diluted with a solvent, e.g. toluene, for example with equal weights of the mixture and the solvent in the diluted composition. Then the diluted composition may be used in the subsequent steps of S2-S6 in the method of FIG. 1A. The reflective wavelength of the semi-IPN coating can be tuned by varying the concentration of RM-1. For example alternative mixtures M2, M3 may be prepared with a contribution of 25.5 wt% and 14.9 wt%.

Therewith a reflection in a reflection band at a shorter and a longer wavelength are achieved. In order to reflect in the visible wavelength regime, while maintaining a low crosslink density, the component RM-1 may be replaced by a combination of the components RM-2, an achiral diacrylate, and RM-3, a chiral diacrylate with a high helical twisting power (FIG. 2). The total amount of these two components RM-2 and RM-3 may be modest, for example 15 wt% in the undiluted composition. The contribution of the component RM-3 to the undiluted composition is for example a few wt%, e.g. 2-4 wt%. Exemplary compositions were prepared with 2.7 wt% and with 3.3 wt% of RM-3.