Method for assembling conductive particles into conductive pathways and sensors thus formed

Abstract

A sensor is achieved by applying a layer of a mixture that contains polymer and conductive particles over a substrate or first surface, when the mixture has a first viscosity that allows the conductive particles to rearrange within the material. An electric field is applied over the layer, so that a number of the conductive particles are assembled into one or more chain-like conductive pathways with the field and thereafter the viscosity of the layer is changed to a second, higher viscosity, in order to mechanically stabilize the material. The conductivity of the pathway is highly sensitive to the deformations and it can therefore act as deformation sensor. The pathways can be transparent and is thus suited for conductive and resistive touch screens. Other sensors such as strain gauge and vapor sensor can also be achieved.

Claims

1. A method for forming a sensor on a substrate, the method comprising, in the following order: a. forming a layer of a mixture comprising a matrix and conductive particles on a substrate, the mixture having a first viscosity which allows the conductive particles to rearrange within the layer; b. applying an electric field over the layer with electrodes, so that a number of the conductive particles are assembled and aligned with the electric field, thus creating one or more anisotropic conductive pathways in the layer, wherein at least one of the electrodes that applies the electric field to assemble and align the conductive particles is not in direct contact with the layer while the electric field is applied; c. changing the viscosity of the layer to a second viscosity, said second viscosity being higher than the first viscosity, in order to mechanically stabilize the layer and preserve the one or more anisotropic conductive pathways.

2. The method in accordance with claim 1, further comprising, after step c: d. totally or partly removing the matrix from the layer.

3. The method in accordance with claim 1, wherein the conductivity of the pathways is changed if the matrix is deformed.

4. The method in accordance with claim 1, wherein the conductive particles comprise at least one material selected from the group consisting of carbon, metal, metal oxides, ceramics, and piezoelectric material.

5. The method in accordance with claim 1, wherein the conductive particles are conductive from quantum tunneling effects.

6. The method in accordance with claim 1, wherein the number of conductive particles in step a) is below a percolation threshold.

7. The method in accordance with claim 1, wherein the electrodes that apply the electric field are a pair of alignment electrodes that are at fixed position relatively to a substrate on which the layer is applied.

8. The method in accordance with claim 7, wherein one of the alignment electrodes is in direct contact with the layer while the electric field is applied with the alignment electrodes.

9. The method in accordance with claim 7, wherein the alignment electrodes are insulated from the layer.

10. The method in accordance with claim 1, wherein the electric field applied over the layer with electrodes is either an AC or a DC-electric field on the order of 0.05-35 kV/cm.

11. The method in accordance with claim 1, wherein the conductive particles have an aspect ratio range of 1-20.

12. The method in accordance with claim 1, wherein the conductive particles comprise irregular graphitic particles, spherical carbon black (CB) particles, disc-like particles, or conical carbon particles (carbon nanocones CNCs).

13. The method in accordance with claim 1, wherein the matrix is a thermoset polymer, a thermoplastic polymer system, a lyotropic system or a mixture thereof.

14. The method in accordance with claim 1, wherein the matrix comprises a UV-curable polymer.

15. The method in accordance with claim 1, wherein the matrix comprises an elastomer.

16. The method in accordance with claim 1, wherein the electric field applied over the layer with electrodes is either an AC or a DC-electric field on the order of 0.1-10 kV/cm.

17. The method in accordance with claim 1, wherein the conductive particles have an aspect ratio of 1-10.

18. The method in accordance with claim 1, wherein the conductive particles have an aspect ratio of 1-5.

19. The method in accordance with claim 1, wherein the conductive particles have an aspect ratio of 1-4.

20. The method in accordance with claim 1, wherein the conductive particles comprise carbon.

21. The method in accordance with claim 1, wherein the conductive particles are irregular graphitic particles.

22. The method in accordance with claim 1, wherein the conductive particles are spherical carbon black (CB) particles.

23. The method in accordance with claim 1, wherein the conductive particles are conical carbon particles (carbon nanocones CNCs).

24. The method in accordance with claim 1, wherein the electrodes that apply the electric field are a pair of alignment electrodes that are moved relatively to a substrate on which the layer is applied.

25. The method in accordance with claim 24, wherein one of the alignment electrodes is in direct contact with the layer while the electric field is applied with the alignment electrodes.

26. The method in accordance with claim 24, wherein the alignment electrodes are insulated from the layer.

Description

LIST OF DRAWINGS

(1) FIG. 1 shows the method of forming aligned particle wires in between two electrodes. The arrow in the Figure indicates the direction of the process.

(2) FIG. 2A shows alignment with electrical contacts between electrodes.

(3) FIG. 2B shows alignment with electrical contacts between electrodes.

(4) FIG. 2C shows alignment without electrical contacts between electrodes.

(5) FIG. 2D shows alignment without electrical contacts between electrodes.

(6) FIG. 3 shows an AFM image of a single aligned particle wire on the generic surface.

(7) FIG. 4 shows schematics of a single wire on an AFM cantilever.

(8) FIG. 5 shows schematics of a hybrid touch screen with controllers connected to a processor of a PC, mobile phone or the similar.

(9) FIG. 6A shows schematics of a strain gauge having a strain sensitive pattern (S) between two terminals (T).

(10) FIG. 6B shows schematics of a strain gauge when the matrix is tensioned.

(11) FIG. 6C shows schematics of a strain gauge when the matrix is compressed.

(12) FIG. 7 shows dendritic pathways useful for capacitive sensing.

(13) FIG. 8A illustrates schematically an experimental procedure for forming a strain sensor, where a low particle fraction mixture was spread between the electrodes.

(14) FIG. 8B illustrates schematically an experimental procedure for forming a strain sensor, where the low particle fraction mixture was aligned by an E-field between the electrode tips.

(15) FIG. 8C illustrates schematically an experimental procedure for forming a strain sensor, where substrate was bent and the resistance of the strings was measured as a function of deflection.

(16) FIG. 8D illustrates schematically an experimental procedure for forming a strain sensor, where W(x) is the vertical deflection and L is the length of beam clamped at both ends.

(17) FIG. 9A shows the impedivity of an aligned CB string in cured polymer.

(18) FIGS. 9B shows the phase angle of an aligned CB string in cured polymer.

(19) FIG. 10A illustrates the resistivity of an aligned string of CB particles as a function of deflection.

(20) FIG. 10B illustrates the relative change in resistance for the aligned string as a function of deflection.

(21) FIG. 11 illustrates the relationship between resistance and compression of a sensor in accordance with another embodiment of the invention.

(22) FIG. 12 is an image of multiple aligned strings of CNCs in an embodiment of the invention.

(23) FIG. 13A is a micrograph of the assembly of a string of CNCs before the electrical field was applied.

(24) FIG. 13B is a micrograph of the assembly of a string of CNCs at 45 seconds after 45 seconds.

(25) FIG. 13C is a micrograph of the assembly of a string of CNCs after 1 minute and 20 seconds.

(26) FIG. 13D is a micrograph of the assembly of a string of CNCs after 2 minutes and 20 seconds.

(27) FIG. 14 illustrates a string of CNC particles in an embodiment of the invention.

(28) FIG. 15 is a comparative figure schematically showing sample fraction versus conductivity for aligned CB and aligned CNC;

(29) FIG. 16 is a comparative figure showing current to voltage or an aligned sample and a non-aligned sample;

(30) FIG. 17 shows the direct current through the sample grows linearly with increasing voltage up to 100 mV and no sign of hysteresis is observed on cycling.

(31) FIG. 18A shows AC impedivity of aligned CNC film and isotropic CNC film.

(32) FIG. 18B shows the phase angle of aligned CNC film and isotropic CNC film.

(33) FIG. 19A illustrates resistivity as a function of deflection for an aligned CNC string versus a nonaligned sample. FIG. 19B illustrates the relative change in resistance as a function of deflection for the aligned string.

DETAILED DESCRIPTION OF THE INVENTION

(34) In all embodiments, the method comprises the mixing of infusible conductive particles and fluid matrix. The matrix contains at least polymer and potentially solvent. The electric field aligns the conductive particles mixed in this fluid. Control of the viscosity of this mixture is by curing the polymer matrix, e.g. by lowering its temperature or by evaporating solvent off.

(35) The resultant aligned material retains anisotropic properties and has directional electrical conductivity. In this way, aligned conductive microstructures are formed of originally infusible particles.

(36) When the substrate of the aligned pathway is deformed, the conductivity changes. When a conductive or dielectric body is close to the pathway, the capacitance changes.

(37) The sensor is manufactured by performing the steps of a. forming a layer of the mixture, the mixture having a first viscosity which allows the conductive particles to rearrange within the layer; b. applying an electric field over the layer, so that a number of the conductive particles are assembled and aligned with the field, thus creating one or more conductive pathways; c. changing the viscosity of the layer to a second viscosity, said second viscosity being higher than the first viscosity in order to mechanically stabilise the layer and preserve the one or more conductive pathways.

(38) The matrix may be totally or partly removed from the layer after step c. The steps may be repeated to create several layers. The conductive pathways in one layer can be connected to the pathways in other layers. The field in step b) can be changed and moved.

(39) In another embodiment the matrix is partly removed by using a solvent or heat. The conductive pathways are exposed. The matrix is replaced with a polymer having mechanical properties more preferable for its use as a sensor.

(40) The resulting sensor device with one or more conductive pathways formed from conductive particles in a matrix can have a number of particles in the pathways being below the number of particles that constitutes a percolation threshold if the particles were homogenously distributed in the matrix.

(41) The invention will be further described by the following examples. These are intended to embody the invention but not to limit its scope.

Example 1

(42) This example concerns the applicability of the alignment method, the use of alignment for formation of individual aligned chains in the predetermined positions.

(43) The employed conductive particles were carbon black (CB) from Alfa Aesar, carbon nano cones CNC from n-Tec AS (Norway) and iron oxide (FeO.Fe2O3) from Sigma-Aldrich.

(44) The employed polymer matrix was a two component low viscosity adhesive formed by combining Araldite AY 105-1 (Huntsman Advanced Materials GmbH) with low viscosity epoxy resin with Ren HY 5160 (Vantico AG).

(45) The conductive particles were mixed in the adhesive by stirring for 30 minutes. Due to the high viscosity of mixture, efficient mixing is possible only up to 20 vol-%. of particles.

(46) Estimated percolation threshold of these materials are at 2 vol-%. The mixtures are conductive above and insulators below this threshold. Particle loads of 1/10 of the estimated percolation threshold were used.

(47) The particles in the matrix were aligned using an AC source. In this example the alignment procedure 1 kHz AC-field (0.6-4 kV/cm, rms value) was employed for >10 minutes for >1 mm electrode spacing and <10 minutes for <1 mm electrode spacing.

(48) The curing was performed immediately afterwards at 373 K for 6 minutes.

(49) The electrode area is kept sufficiently small to allow only a single pathway of particles. Alternatively, the particle fraction is lowered. This is shown in FIG. 1.

(50) In one embodiment of this example, metal particles, silver flakes (Sigma-Aldrich) of size 10 m, was used instead of carbon particles.

Example 2

(51) In FIG. 1b is shown removal of electrodes after alignment and thus freestanding aligned film even in the case where the matrix is adhesive. The alignment also occurs if the electrodes do not touch the material and so the alignment can be performed from the distance. When the material and electrodes are moved, continuous or stepwise, with respect to each other during the alignment, this allows continuous alignment processing and different geometries. Three possible options for the alignment settings are illustrated in FIG. 1b that shows aligned film with (A-B) and without (C-D) electrical contacts between electrodes (a) and material (b). In the case (A) the aligned film forms permanent connection between the electrodes. In the case (B) the electrodes and material are only loosely joined together and can be moved apart after alignment. In the case (C) there are insulating layers (c) between the material and electrodes and they are easily moved apart after the alignment even in the case where the material is an adhesive. In this case the obtained material is a multilayer consisting of aligned layer (b) and two insulating layers (c) In the case (D) the alignment is carried out from the distance and the mutual location of electrodes and film can be additionally moved during the alignment. For illustrative purposes the placement of the electrodes are shown such that alignment occurs in the z-direction. Alignment in the x- and y-direction or an arbitrary direction can be achieved by relative movement of the field, such as moving the distant electrodes.

(52) FIG. 2 shows a picture of a single pathway consisting of an aligned row of particles.

Example 3

(53) This example concerns the applicability of the alignment method, the use of alignment for formation of individual aligned chains in the predetermined positions.

(54) The procedure was otherwise similar to that in Example 1 but instead of generic surface the particle chain is aligned on the AFM cantilever. When the cantilever is bending, influenced due to the changing surface forces, the aligned pathway gets microscopically stretched and the particles become disconnected from each other. This influences the conductivity through the particle chain. This allows the use of aligned chain as a sensor of surface properties of the surface studied by the cantilever. This setting is illustrated in FIG. 3, where the upper image to the right illustrates a connected path, and the lower image (bent state) illustrates a disconnected path.

Example 4

(55) A resistive touchscreen placed in front of the display is created from two layers of film with conductive pathways in x and y direction, and with the matrix reduced so that the pathways are exposed. When contact is made to the surface of the touchscreen, the two sheets are pressed together. The horizontal and vertical pathways that when pushed together, let the controller register the precise location of the touch from any object, e.g. finger, stylus, pen, hand, by forming a contact.

(56) In an alternative embodiment of this example the pathways in the x and y directions are formed as two layers in a single film. During operation of a four-wire touchscreen, a uniform, unidirectional voltage gradient is applied to the first layer, using the two wires to electrodes at each end of the sheet. The horizontal and vertical lines that are in the deformed area will be broken because the carbon particles in the conductive pathways will separated. When the sheet is pressed, the controller measures the voltage as distance along the first sheet, providing the X coordinate. When this coordinate has been acquired, the uniform voltage gradient is applied to the second layer using the two other wires to ascertain the Y coordinate. These operations occur within a few milliseconds, registering the touch location as contact is made.

(57) Such a touchscreens typically have high resolution (40964096 DPI or higher), providing accurate touch control.

(58) Due to the low particle loading the touchscreens will be more transparent, as the pathways will be practically invisible.

Example 5

(59) A capacitive touchscreen is created from one layer of film with conductive pathways aligned in any direction. The pathways forms a capacitor that holds charge, e.g. from a voltage applied to the edges of the layer creating a controlled capacitor. When contact is made to the surface of the touchscreen, or a finger of a dielectric or conductive body is close to the touchscreen, the electric field across the touchscreen is changed

(60) The capacitive controller connected to the screen calculates the X and Y coordinates from the change in the capacitance as measured from the four corners of the film.

(61) In another embodiment two or more layers are used, with eight or more wires to the corners of the film, four and four connecting to each layer, thus creating higher resolution when the controller switches between reading the layers, or if multiple controllers are used.

Example 6

(62) A hybrid screen is manufactured with resistive and capacitive layers as showed in FIG. 5. There could be more than one resistive and capacitive layer. In this embodiment both the position of proximity to and pressure on the sensor from e.g. a finger or stylus will be detected by the controllers and sent to the processor of the PC, mobile phone or similar device. In FIG. 5, the reference numbers indicates the following: 4:1Display 4:2Transparent capacitive sensor 4:3Transparent resistive sensor 4:4Capacitive controller 4:5Resistive controller 4:6Processor

Example 7

(63) Touch sensors to be used as in examples 5, 6 and 7 are formed using glass instead of a polymer as matrix.

Example 8

(64) A strain gauge is formed by using a matrix that is an elastic polymer. As shown in FIG. 6A-6C the conductive pathways have been formed to a pattern suitable for a strain gauge by moving the field relative to the matrix. In a simpler embodiment the pattern is a straight line, or any other suitable pattern.

(65) In FIG. 6A, the pathways form a strain sensitive pattern (S) between two terminals (T). When the matrix is tensioned (FIG. 6B), the area of the pathways narrows and the resistance increases, resulting in higher resistance between the terminals (T). When the matrix is compressed (FIG. 6C), the area thickens and the resistance decreases, resulting in a lower resistance between terminals (T).

Example 9

(66) An electronic hygrometer, a humidity sensor is manufactured using capacitive sensors similar to that in example 5, but where the change in capacity is due to a change in the amount of water present in the matrix. In another embodiment the change in conductivity is measured. In one embodiment the matrix is made from cellulose. Temperature must also be measured, as it affects the calibration of these humidity sensors.

(67) In one embodiment the alignment is in the z-plane, perpendicular to the substrate, and thus forming a structure similar to a carbon nanotube array. The alignment can also be in the x, y plane. The absorption and desorption of chemical vapours by the polymer matrix cause changes in the inter-tube distance or the electrical properties of the matrix and the conductance or capacitance changes.

Example 10

(68) In order to produce a matrix with increased capacitance a procedure similar to that in example 1 was used, but the alignment was terminated before the chains reached from electrode to electrode. FIG. 7 shows so obtained electrodes with dendritic surface. This creates pathways that can hold more charge than the conductive pathways else produced.

(69) The capacitance of this structure will be sensitive to deformation along the direction indicated by an arrow in FIG. 7.

Example 11

(70) The example with the procedures similar to any of the claim 1, 2, 3, 4, or 5 but instead of other previously employed particles, spiky particles used in the QTCs are used.

Example 12

(71) In this example aligned particles are used in a nanomechanical cantilever. This means that the cantilever is highly miniaturized and that instead of bulk layer the properties of single particles dominate.

Example 13

(72) In this example a flexible sensor shaped as a thin sheet, coating or film is formed. It could be formed as part of the structure or as a material that is added on the whole or a part of the inner or outer surface of a body of a ship's hull, an aircraft or another vehicle or part there of (such as a car's engine), industrial machinery or on buildings, such as bridges or houses. It could also be used for packaging or as part of clothing, furniture or electronic equipment such as computers.

Example 14

(73) This example is similar to example 13 but in addition the sensor function as at least one of an anti-static coating, a thermal conductor, an antenna and a shielding for electro-magnetic waves due to the properties of the aligned pathways and the matrix that makes up the sensor.

Example 15

(74) A micro-mechanical strain sensor made of carbon black (CB) particles assembled into a single wire in a polymer matrix was produced.

(75) The experimental procedure is shown in FIG. 8A-8D. CB particles (Alfa Aesar) were dispersed in UV-curable urethane methacrylate-based thermoset polymer Dymax 3094 (Dymax Corporation, CT). The particle fraction was 0.1 vol. %. This dispersion was spread to form a <10 m layer on top of tiplike gold electrodes (FIG. 8A).

(76) These electrodes were prepared using UV-lithography on a 250 m thick silicon substrate covered by a insulating silicon oxide layer. Their thickness, width and mutual spacing were 100 nm, 3 m and 100 m, respectively. In the next step an alternating electric field of amplitude 3 kV/cm with a frequency of 1 kHz was applied over the sample using a custom-made voltage source. This led to the assembly of a particle string in between the electrode tips, FIG. 8B, in less than two minutes. The material was subsequently UV-cured by a mercury lamp. The resistance over the electrodes was monitored using a Keithley 2000 multimeter.

(77) FIGS. 8A-8D illustrate the experimental procedure. In FIG. 8A, a low particle fraction mixture was spread between the electrodes and aligned by an E-field into single strings by alternating electric field between the electrode tips in FIG. 8B. In FIG. 8C it is illustrated how this substrate was bent and the resistance of the strings was measured as a function of deflection. In FIG. 8D, W(x) is the vertical deflection and L is the length of the beam clamped at both ends.

(78) The electromechanical properties of so prepared strings were studied by clamping the substrates under two clamps and bending them, as seen in FIGS. 8C and 8D.

(79) The substrates were bent at the substrate centre, which leads to the stretching of the string on the surface. The resistivity of the string was measured as a function of the vertical deflection W(x) FIG. 8D. The surface strain and the resistivity increase with increasing bending. The strain on the surface corresponds to the strain in the string and the resistance through the string is increased with this strain.

(80) FIG. 9A and FIG. 9B show the impedivity and phase angle , respectively as function of frequency of the aligned and cured CB particle string. The impedivity is nearly constant up to 1 kHz and decreases for higher frequencies. The phase angle begins to deviate from zero at 100 Hz, indicating a contribution from the capacitive conductivity.

(81) FIG. 9A illustrates impedivity and FIG. 9B illustrates phase angle of an aligned CB string in cured polymer.

(82) FIG. 10A shows the resistivity of the single string in a cured polymer matrix as a function of deflection. Also shown are corresponding data for the nonaligned film containing 12 vol. % of CB particles, i.e., a fraction well above the percolation threshold. The strain corresponding to a given deflection is also presented in the graph and is the relative displacement of particles in the polymer with deformation. The vertical dotted lines mark the change of deflection direction. The samples were bent from a relaxed state at 0 m to a deflected state at 50 m, then back to the relaxed state, and lastly bent to 50 m once more. These measurements were done subsequently with the same samples. For the aligned sample, the data show an increase in resistivity with 50 m deflection. The original resistivity is restored on the release and the increase is again seen with next deflection.

(83) FIG. 10A illustrates the resistivity p of an aligned string of CB particles (circles) with the initial 0.1 vol. % particle loading, and a non-aligned sample with the 12 vol. % particle loading, (squares) as function of deflection D. The dotted lines mark the change of deflection direction. FIG. 10B illustrates the relative change in resistance for the aligned string as a function of deflection. The open triangles and diamonds show the first and third deflection, respectively. The solid squares show the first release. Dashed, solid and dash dotted lines are corresponding linear fits.

(84) FIG. 10B shows the relative change in resistance for the first and second deflection plus the first release of the aligned sample corresponding to FIG. 10A. A gauge factor estimated for the aligned string is about 150 with an error margin of 10%, as estimated from these slopes using Eq. 1 (R/R=K.Math.S, where K is the gauge factor and S is the strain). This is significantly higher than that of a nonaligned, high-particle fraction sample, which did not show any measurable effect due to the stretching. This means that the alignment has significant benefits in both the conductivity enhancement and in the strain sensitivity.

(85) Results reported earlier (by Gammelgaard et al.) show that a SU8 polymer mixed with high concentration (16%) of isotropic CBs have a gauge factor of 15-20. Thus a gauge factor about 10 times higher may be obtained with an aligned single string compared to an isotopic high particle fraction sample.

(86) In conclusion, single strings of CB particles were aligned by dielectrophoresis in UV-curable Dymax 3094 polymer matrix and shown to be a promising candidate as a strain sensor. By deforming these single strings in-plane, a reversible change in resistivity was observed, similar to what has been reported before 8 for non-aligned CB-SU8 polymer composite when the particle fraction exceeds the percolation threshold. A gauge factor of 150 was found, exceeding the value of 15-20 reported previously. Detection of significant gauge factors for nonaligned CB-Dymax 3094 composites with the particle fraction exceeding the threshold was not possible.

(87) Higher gauge factors could be achieved by using different particle sizes, particles size distributions or by improving the conductivity of the nanoparticles.

Example 16

(88) In another embodiment, carbon black particles were aligned in an elastomeric matrix. The particles were carbon black particles 0.0004 g, and the elastomer Dow Corning 734 Flowable Sealant (silicone based elastomer), 0.6700 g. A solvent, 2-Butanone, 0.5941 g were used to decrease the viscosity of the elastomer. The particle concentration was 0.03 wt %.

(89) The resulting elastomeric matrix including aligned CB particles displayed behaviour where the resistance of the conductive path of CB particles decreased with compression of the elastomer.

(90) FIG. 11 illustrates the resistance through the conductive path versus compression of out-of-plane aligned CB articles in the matrix of Dow Corning 734. This means that the sample is sheetlike and the aligned strings are formed parallel to its surface normal, thus connecting two largest surfaces through the sheet. The electrode had an initial spacing of about 150 m. The data correspond to three subsequent compressions. It is seen how the resistance decreases with each compression and resumes a higher value when the compression is released. An alternating electric field of 1.5 kV/cm for 5 min. The sample was humidity-cured for 24 hours. The resistances were measured with Keithley 2000 millimeter.

Example 17

(91) Alignment of single strings of CNCs in Dymax 3094 polymer was performed. CNCs were mixed with urethane methacrylate based Dymax 3094 with a particle fraction of 0.1 vol. %. This particle fraction is an order of magnitude lower than expected percolation threshold (2 vol. %). The low particle fraction suppresses aggregation thereby rendering a uniform mixture with the particle size below 3 Since the size of CNC particles are between 100 nm and 3 the CNCs are believed to be nearly perfectly dispersed. The alignment was done by spreading a thin (1-10 m) layer of this dispersion over the tip-like electrodes and applying an alternating electric field between the electrodes followed by UV-curing, as seen in FIG. 8A-8C (details can be found in the Experimental section below).

(92) FIG. 12 Multiple aligned strings of CNCs in a cured Dymax 3094 polymer on interdigidated electrodes.

(93) FIG. 13A to 13D show micrographs of the assembly of a string of CNCs over time. The applied field is E=4 kV/cm over an electrode spacing of 100 m. Originally isotropic mixture had the particle fraction of about 0.1 vol. %. FIG. 13A illustrates particles dispersed in the polymer before the electric field was applied. Snapshots of the alignment process after 45 seconds FIG. 13B and 1 minute 20 seconds FIG. 13C. A complete string was formed within 2 minutes 20 seconds FIG. 13D.

(94) Hence, a conducting string between two electrodes with a spacing of 100 m was produced in less than 150 seconds.

(95) Prior to the alignment, the resistance of the mixture is in the M range. The alignment causes the resistance to drop over three orders of magnitude to the k range. For instance, the aligned string shown in FIG. 13D had a resistivity of p=40 m.Math.m. UV-curing of the polymer composite will locks the aligned carbon particles, making electrical characterization of the aligned strings possible. The alignment and conductivity are maintained upon curing.

(96) FIG. 14 shows a close-up of an aligned string after curing with a resistivity of =32 m.Math.m. This is an UV-cured single string of aligned CNC particles spanning an electrode gap of 100 m.

EXPERIMENTAL

(97) Materials and Sample Preparation

(98) The samples contained carbon particles dispersed in a polymer matrix. The employed CNC material was supplied by n-TEC AS (Norway) and it contained about 70% discs, 20% nanocones and 10% carbon black. The material had been heat treated to 2700 C. prior use. The employed CB was supplied by Alfa Aesar. The polymer used was Dymax 3094 Ultra Light-Weld (Dymax Corporation, CT) supplied by Lindberg & Lund AS (Norway). This is an urethane methacrylate based UV-curable thermoset polymer. Carbon particles were dispersed in the polymer by stirring at 150 rpm for 15 minutes, which leads to a uniform dispersion with the particle size less than 10 m.

(99) The alignment procedure is shown in FIGS. 8A-8C. The gold electrodes were made by UV-lithography on a 250 m thick silicon wafer covered by a 300 nm thick insulating silicon oxide layer, and consisted of two 100 nm thick and 3 m wide fingers facing each other with the spacing d ranging from 10 to 100 m.

(100) A layer (<10 m) of dispersion was smeared on top of the electrodes (FIG. 8A). An electric field of 4 kV/cm with a frequency of 1 kHz was applied over the electrodes (FIG. 8B). The dielectrophoresis effect causes the particles to move towards the two electrode tips, forming continuous strings at the edge of the tips. The strings would grow from each electrode tip until they met at the halfway point of the electrode gap forming a continuous conducting string (FIG. 8C). The alignment occurs within 1-3 minutes, depending on particle concentration and the applied electric field. The particles will stay in place after the electric field is turned off, but the characterization or moving of the sample may destroy the aligned string. The polymer is therefore subsequently UV-cured by a mercury lamp for 5 minutes, locking the particles into place.

(101) For the DC conductivity measurements, 20 similarly prepared parallel samples of CNCs and CB particles were aligned in Dymax 3094. The resistance was monitored by a Keithley 2000 multimeter over the alignment electrodes immediately after alignment without UV-curing.

(102) Electrical and Electromechanical Characterization

(103) The electromechanical experiment is illustrated in FIG. 8D. The samples were clamped at both ends and a micrometer screw was used to control a small blade at the center of the substrate. The blade would begin to bend sample thus stretching the carbon string on its upper surface. The samples were deflect by a given deflection and the relaxed, and then bent yet again several times in a continuous manner. An IV curve was measured for every deflection point and used for determining the resistance for each point. The resistance through the sample increases with increasing bending.

(104) Electrical Properties of Aligned Strings

(105) The DC conductivity of 20 identically aligned samples prepared in accordance with the above were measured for both CNCs and CB particles. The distributions of these conductivities are shown in FIG. 15. FIG. 15 illustrates the sample fraction versus conductivity for aligned CB (striped) and CNC (black) in Dymax 3094 polymer. The CNCs have higher conductivity but have a smaller probability to produce a conducting string. Only fifty percent of the aligned single strings of CNC were conducting and the nonconducting strings are not included in the shown probability distribution.

(106) The figure is normalized so that the probability of producing a conductive string is unity in both cases. The CNC strings have a higher conductivity than the CB strings.

(107) The CNC particle strings have conductivities ranging from 25 to 500 S/m, with 90% falling in the region below 100 S/m. The CB particles have conductivities ranging from 1 to 22 S/m, with 95% falling in the region below 6 S/m. Though the conductive CNC particle strings have a higher conductivity, only 10 out of 20 prepared samples conducted any measureable current. These nonconductive strings are not included in FIG. 15. However, all 20 the CNC particle strings appeared visually complete and were therefore expected to be conductive. This implies that tiny, optically invisible mismatches between particles are enough to prevent a conductive pathway. One reason for this difference between CNCs and CB may stem from the different particle topology. Another reason may be the polydisperse nature of the CNC particles. The cones and discs might have difficulties creating good topologically matching connection between each other due to the variable shape of the particles.

(108) The conductivity of an intact string provides an estimation for the uppermost conductivity of the earlier reported multi-string samples like the one shown in FIG. 12. As the former is significantly higher (for CNCs 25-500 S/m, see FIG. 15) than the conductivity of multi-string samples normalized to the volume fraction of particles (0.1-1 S/m), this indicates that a large part of seemingly intact strings in the multi-string samples (FIG. 12) are actually broken. The conductivity gap between single strings and multi-string samples could potentially be reduced by optimizing the preparation of larger samples for example using even and vibration free preparation setups.

(109) FIG. 16 plots the current-voltage curves of an aligned low particle fraction sample and a nonaligned high particle fraction sample. The solid line is from an aligned sample prepared from the isotropic CNC-polymer mixture with a concentration of 0.1 vol. %, while the dashed line represents the nonaligned sample with a concentration of 13 vol. %. The direct current through both samples grow linearly with increasing voltage up to 100 mV and no sign of hysteresis is observed on cycling. The nonaligned sample has a marginally higher conductivity than the aligned one but they are both of the same order of magnitude.

(110) FIG. 17 The direct current through the sample grows linearly with increasing voltage up to 100 mV and no sign of hysteresis is observed on cycling.

(111) FIGS. 18A and 18B plot the impedivity and phase angle vs. frequency for both an aligned sample and a non-aligned, high particle-fraction sample. AC impedance data of aligned CNC strings (squares) and isotropic CNC film with high particle fraction (13 vol %) in Dymax 3094 polymer (triangles). FIG. 18A impeditivity as a function of frequency, FIG. 18B phase angle as a function of frequency.

(112) Both samples behave very similarly when the frequency is increased. The impedivity is nearly constant up to 1 kHz, but falls noteworthy between 1 kHz and 10 kHz. The phase angle begins to deviate from zero at 100 Hz, indicating the rise of capacitive conductivity. The similarity between aligned and nonaligned samples implies that the aligned string behaves essentially as bulk, heavily loaded CNC composite. These data are also consistent with the AC-impedance of multi-string samples indicating the Ohmic nature of the strings. However, the data differs from the data of nonaligned CNC polymer mixtures at low particle fraction where the phase angle begins to differ from zero much earlier (>10 Hz), pointing to the ionic conductivity of the polymer.

(113) Electromechanical Properties of Aligned Strings

(114) Next, electromechanical measurements of the CNC strings in cured matrix were performed. The samples were clamped at both ends and deflected gradually in the centre as shown in FIG. 8D. The deflection leads to the stretching of the surface layer and thereby causes an increasing strain of the aligned film and presumably moves the particles with respect to each other. The measurements were conducted by deflecting the samples several times and measuring the resistivity for every deflection point.

(115) FIG. 19A illustrates resistivity as a function of deflection for an aligned CNC string (open squares) and nonaligned sample with high 13 vol. % particle fraction (solid square). The dotted lines mark the change in deflection direction. FIG. 19B The relative change in resistance as a function of deflection for the aligned string. The open triangles show the first deflection while the open diamonds represent the third deflection. The solid squares show the first release. Dashed, dash dotted and solid lines are corresponding linear fits.

(116) FIG. 19A shows the so obtained resistivities for an aligned sample as a function of deflection. The corresponding data of nonaligned sample is shown for comparison. The aligned sample was prepared from the dispersion with a particle fraction of 0.1 vol. % and is the exact same sample as shown in FIG. 14. The nonaligned sample had a particle concentration of 13 vol. %. The resistivity at each deflection point was calculated by linear fits to separate IV curves measured at each point. This method is justified by the ohmic behaviour of strings for DC and low measurement frequencies. The top axis shows the induced strain of the strings defined as the relative displacement of particles due to an applied external force. The measurements were done in a continuous manner and the vertically dotted lines mark the change of deflection direction. When the deflection goes from 0 to 50 m to 30 m in FIG. 19A, it means that the sample was deflected from a relaxed state at 0 deflection to a deflected state at 50 m and relaxed to a less deflected state at 30 m. The deflection has a significant and reversible effect on the resistivity of the aligned sample, the resistivity increasing with the increasing deflection and strain. In contrast, the resistivity of nonaligned sample does not show any measureable effect.

(117) FIG. 19B shows the relative change in resistance calculated from the first and second deflections as well as the first release of the aligned sample (FIG. 19A). These data gives a gauge factor of 50 with an error margin of 10%, as estimated from the slopes in FIG. 19B, using Eq. 1. This shows that the alignment has a significant effect making both conductive and piezoresistive CNC materials possible. The aligned string performs well against its standards of comparison. The obtained gauge factor is higher than those of typical thick-film resistors, whose gauge factors range from 3-30 and also compares well to that of an optimized silicon piezoresistive cantilever. The gauge factor is also notably higher than typical values reported for aligned multi-walled carbon nanotubes, for which values of about 1.5 and 3 have been reported for particle concentrations of 0.75 wt-% and 0.5 wt-%, respectively.

(118) In this work, the strain was calculated at the top surface of the substrate, and it was assumed that the strain experienced by the composite layer with the particle string, was the same. This is only correct if the composite layer thickness is well below the substrate thickness, but this condition was fulfilled in our experiments.

(119) It will be understood that a person skilled in the art may readily envisage numerous alternatives to the above-described example embodiments. In particular, the described features may be varied or combined to form new embodiments.