The invention relates to a transfer method of a high viscosity functional material, such as a conductive paste, on a substrate.
Interconnection of discrete bare chip components or microelectronic components (chips) on a substrate is a process for which precision material deposition with sufficient accuracy is necessary to allow for miniature electrical or other type of interconnects so that when a chip component becomes attached to a placement position all electrical connection and fixation to the substrate can be made by a conductive material or paste. An example of a conductive adhesives is a thermal or UV settable resin loaded with silver particles, or a solder paste comprising solder particles and flux. A common method for selectively depositing a functional material on a substrate is via printing. Typically a functional material needs to be formulated with other materials before the functional material can be printed on a substrate. A formulation can e.g. be formed by dispersing the functional material in a solvent or liquid; thus, the formulation is generally wet. The formulation is often referred to as an ink or paste, depending on the viscosity. A challenge exists to transfer such functional materials, since ideally, instead of printing a functional ink material on a substrate, selectively depositing only the pure functional material on a substrate is most preferable. Such materials e.g. consist of more than 50 vol% e.g. even 90 vol% of a solid state material such as a conductive polymer. Due to their highly viscous nature, e.g. higher than 50 Pa/s to even as much about 1000 Pa/s (10^6 cps). Such materials will not significantly deform due to gravity and flow, but substantial maintain the intended aspect ratio and form definition in the transfer. In US20170268100 a transfer method is described, also known as photonic printing, where a transparent transfer plate having cavities formed therein, that are filled with a functional material. The transfer plate is irradiated with pulsed light to heat a layer of light-absorbing material in order to generate gas at an interface between the layer of light absorbing material and the functional material, to release the functional material from the cavities. The problem with this method is ununiform ink release due to shadow effects. Especially for high aspect ratio cavities that are closely packed, the light only reaches the sides with difficulty. Uniform release of the ink results in poor print qualities, especially when high resolution patterns are transferred over relatively large distances. Furthermore, to properly transfer feature sizes of 5 - 10 µm without overheating the ink, the heat flux needs to be at least 40 kW/cm2 in the order of microseconds, which are unrealistic operation conditions for flashlamps. Another problem is that the method is only suitable for global transfer of structures designed via the patterned cavities provided in the substrate, as the light is not selective. Ideally patterns can be selectively formed and transferred, depending on a required pattern.
The prior art assembly has failed to provide an adequate method for high resolution bonding of fine pitch transfers on a substrate providing enough flexibility and reliability, for which the invention seeks a solution. To this end, it is proposed to provide a device and transfer method of a high viscosity functional material, such as a conductive paste, onto a receiving substrate (20).
In an aspect the invention pertains to transfer device for transferring a viscous functional material, such as a conductive paste, onto a receiving substrate. The device comprises a plate having a cavity surface that includes at least one cavity. The plate, e.g. a plate patterned a plurality of cavities, further comprises at least a thermal insulator layer formed of a thermally insulating material, e.g. glass (silica or the like). The at least one cavity is formed in the thermal insulator layer. The at least one cavity is provided with a thermal conductor. The thermal conductor is formed of a material having a high thermal conductivity, e.g. silicon or a metal such as copper or aluminum and the like. The thermal conductor extends along, and is restricted to, the cavity surface, e.g. as a layer or coating covering bottom and side wall surfaces of the cavity. The cavity can thus be understood to be defined or formed by said thermal conductor in the thermal insulator layer. The transfer device further comprises at least a resistive device that is in thermal contact with the thermal conductor and control circuitry connected to said resistive heating device. Thermal contact may be provided by direct contact. Alternatively thermal contact may be provided via one or more intermediate layers. The control circuitry is arranged to heat the resistive heating device, so as to, in use, generate a gas at an interface between said cavity surface in the cavity and a functional material provided in said cavity by heating the thermal conductor to transfer said functional material from said at least one cavity by said gas generation onto a receiving substrate. The resistive device is provided as a resistive layer parallel to the plate.
Advantageously, transfer devices as disclosed herein mitigate transfer, drying add/or curing of material of residual material present outside the transfer pattern, e.g. residues deposited next to a cavity a filling cycle with a doctor blade, cf the hemispherical droplet on the surface of the thermal insulator layer as depicted in
This advantage, in particular over devices having uniform or global heating such as devices heated by light pulses, is realized because, in use, the device generates thermal contrast between transfer areas of the device, the cavities having a heat spreader, and non-transfer areas, e.g. the surface of the thermal insulator next to the cavities.
The patterned plate can be filled with the functional material in a convenient and/or rapid way, e.g. by spreading a paste across the plate with a doctor blade.
To maximize the heat contrast the thermal diffusivity of the material forming the heatspreader divided by the thermal diffusivity of the material forming the thermally insulating layer of the plate is preferably as large as possible. A minimum ratio can be understood to about 10, a better ratio is >50, preferably the ratio is at least 100. For example, copper and glass form a suitable pair of materials.
In another of further aspect, the invention pertains to a method of transferring a high viscosity functional material, such as a conductive paste, onto a receiving substrate. The method comprising the steps of: providing a device according to the invention; e.g. a plate having a cavity surface that includes at least one cavity as disclosed herein; and providing a functional material in the at least one cavity, having a material composition that, when heated by the heating device, generates a gas at an interface between said cavity surface in the cavity and said functional material, to transfer said functional material from said at least one cavity by said gas generation onto the receiving substrate. Said functional material may be provided with a solvent and a solid state composition of more than 50 vol % solid state material. Said functional material may be a conductive ink, e.g. conductive adhesive material, such as an electroconductive polymer, a metallized polymer or a solder paste or adhesive.
In an aspect, a direct write method is provided for high speed positioning die bonding material for discrete chip components on substrates, which can be operated in a reel to reel manufacturing fashion. In particular, the method can be used for high resolution deposition of highly viscous materials that can be deposited in thick layers with high aspect ratio features. With the attainable resolution of the disclosed methods and systems, that a resolution spot size of transferred die bonding material may be attained with a spot diameter of the transferred bonding material that is smaller than 50 micron.
The resistivity of the conductors may be 10-7 Ohmm or lower. By using the equation below the required resistivity of the conductor can be calculated based on the resistivity of the resistor and layer thicknesses. The transfer method uses an electric pulse from circuit 14 to rapidly heat a thin film resistor 132 to transfer a functional material 50 from carrier onto an acceptor substrate. For example, the control circuitry 14 is controlled to heat said heating device 13 in less than 100 microseconds at an average power of more than 10 kW/cm2. The release method may be more effective if the control circuitry 14 allows for a pre heating step, prior to the gas generation step.
The thin film heating devices 13 are in this example only located in the recessed cavity parts of the transfer plate 10. When electric current flows through all the resistors, joule heating causes the solvent (or polymer) within the functional material to evaporate. This can be compared to the Leidenfrost effect, only in this case the heating is selective and for a very short time period (<100 µs). The patterned plate can be filled rapidly e.g. by using a doctor blade, whereupon the functional pattern is transferred onto a planar or non-planar substrate. The transfer plate material preferably has a low thermal conductivity in order to reduce heat sinking when the heating device is triggered. The lower the thermal conductivity of the plate (and paste), the more rapid and efficient the solvent can be evaporated. The functional material is heated at a heating rate of at least 0.5, preferably 5 K/microsecond or even as high as 50 K/microsecond in order to generate said gas at the interface but preferably as high as possible without damaging the resistor stack. This thermal shock means that the thermal expansion coefficient of the materials in the stack are closely matched to that of the plate 10 itself. Too high stresses in the thin film may lead to damage or delamination.
In some cases it might be preferred to have different heating rates to control the ejection characteristics of the paste. For example, it might be beneficial when the paste releases on the sides a little bit earlier than the top. This can be achieved by embedding the resistor in a selectively coated heat reflective layer 133-1, 133-2 or thermal insulation layer 133-1 or 133-2 (see
Where ρr is the resistivity of the resistor in Ohmm, A is the resistor area (m2), U is the voltage over the resistor in volts (V), P is the power in watts (W) and tr is the thickness in meters (m).
Typically the resistivity value is in the order of 10-2 - 106 Ohmm to enable high heat fluxes with thin layers. It is important that during the entire heating process the resistance through the resistor is always higher than the resistance through the conductor.
E.g. in
When using a glass or quartz carrier plate, the typical fluence to release a functional paste material can be in the order of 0.4 J/cm2 when the pulse width is 100 µs. This depends on the type of solvent used in the paste and how fast the resistor is being heated.
The power requirements are quite demanding for this process. For a 100 µs pulse, the required average power level may be around 10 kW/cm2. However, the threshold fluence to release an ink can go down significantly when the average power level increases. The thin film resistor also allows to pre-heat the functional material before it is being ejected. This can be used to print hot melt polymers for example.
In order to deliver this power, a predefined voltage and current density is supplied to the plate. As the resistance across the entire thin film resistor is dependent on the patterns size and resistor material, the required voltage on the capacitors may vary from pattern to pattern. When the resistance of the resistor material is known, the total surface area of the thin film resistor can be calculated. Knowing the total surface area, the total required energy can be determined. Then Ohm’s law is used to calculate the electrical power and voltage for the circuit 14. It is important that the voltage drop only occurs over the resistive layer, not in-plane of the conductors. This means that the conductors need to be highly conductive, many orders of magnitude higher e.g. 100 or more, than the resistive coating. A suitable example for an electrode layer may be molybdenum, e.g. 500 nm Mo that is very thin, to reduce thermal stress due to heating. A suitable example of a resistor may be SiC or GaN. These materials have comparable expansion coefficients. The resistive device 13 may be capped by a hardened layer e.g. a SiN layer, to reduce wear.
When the resistance is higher through the conductive layer, then joule heating will occur in this layer, which is undesired. For semiconductors (undoped) the resistivity drops significantly when the temperature increases. This negative temperature coefficient (NTC) effect could cause problems with respect to heating homogeneity. If the thickness of the resistor is not perfectly uniform, then heating will occur faster where the layer is thinnest. Hence the resistance will drop even further at that location and the current will primarily flow through that part of the resistor.
Therefore it is preferred to have a positive temperature coefficient (PTC), meaning that the resistance increases and power consumption decreases with temperature. To obtain a PTC with a resistivity in the range of 10-2 - 106 Ohmm it is possible to use a doped semiconductor (e.g. silicon, silicon carbide and gallium nitride). However, typically the resistivity of PTC materials is relatively low, which would mean that the resistance requirements of the conductive layer will become more demanding. Ideally the resistor is made out a of a material with a high resistivity (10^0 - 10 \^2 Ohmm) which has a PTC behavior between 20 C and 400 C.
Besides a single resistor (doped semiconductor) material, it is also possible to use a stack with multiple layers. For example Schottky barriers can be added.
Depending on the size of plate and the layer thicknesses and electrical resistivity of the resistor material can be calculated. If the voltage drop is still too high (for example when the plate is very big), through hole vias 143 can be added to connect the conductors to highly conductive busbars 144 at the backside of the plate e.g. to an alternative circuit 144.
In contrast to the
As it is difficult to design local heaters for complex printing patterns, resistive layer preferably covers a full area of the plate 10 for ease of manufacturability, but may also be patterned in smaller areas on plate 10.
To make sure the resistive layer 132 does not overheat and damage the plate where there is no heat spreader 122, the carrier substrate plate 10 may be provided in some embodiments from a high thermally conductive material, such as silicon. This allows heat from resistive layer 132 to diffuse into the carrier where there is no printing pattern, which limits the peak temperature of the heater itself, and avoids high peak temperature (overheating) at bottom of the plate.
Also, when there is residual paste on this surface, it is desired to prevent this material from transferring, drying or curing.
In order to further create high temperatures in the heat spreaders/ thermal conductor 122, the transfer device 100 further comprises a heat flux reflector part 133 aimed at directing heat flux towards the cavity 12. Heat flux reflector part 133 may be a thermal insulator, opposite the cavity at least, a part that has a relative high thermal insulation contrast relative to the further plate 10, e.g. of glass (Silica or the like) heat flux reflector parts embedded in a silicon plate. Such heat flux reflector parts may have a thickness of e.g. 3-5 micron, but other dimensions (up to 100 micron) may also work. Thus the heat flux reflector provides a thermal buffer on top of the cavity 12 to direct the heat to the downwards to the cavity 12. This creates a high thermal contrast between the patterned and un-patterned parts of the plate 10.
It will be clear that the plate 10, like for non-planar embodiments, may be formed of a material having a low thermal conductivity, e.g. glass (silica or the like). Forming the plate of a material having a low thermal conductivity increases thermal contrast between an outer surface of the plate and the cavity having a thermoconductive material that extends along and is restricted to the cavity. Forming the plate of a material having a low thermal conductivity further reduces, heat sinking, transfer of heat into the plate when the heating device is triggered, which improves transfer of heat towards the cavity for a given energy pulse as compared to devices having a plate formed of a thermoconductive material.
Highly conductive busbars 144 (e.g. of copper) can be added into the plate 10 to reduce the required voltage for achieving a high heat flux, which is advantageous as high operating voltages may increase a chance for electrical breakdown between adjacent heaters. In the example, busbars 144 draw a planar current by alternately wiring anode and cathode busbars e.g. distanced over the cavity structures 12 and at a distance therefrom to prevent thermal conduction to the busbars, as indicated in the Figure. Depending on the conductive nature of the plate resistive layer 132 may be electrically shielded by additional electrical insulator layers 134 on opposite sides.
In an example, the thermal conductor 122 completely forms the cavity, which may for example be at least 5 micron deep, and may be in thermal contact only at the bottom of the cavity, which may be planar conforming the planar resistive layer 132.
A cavity having a thermal conductor 122 that extends along the cavity and is limited thereto, e.g. as shown, can be manufactured by a number of methods. These include microfabrication methods such as local deposition e.g. by masking. Advantageously, the cavities with the thermal conductor 122 can be cost-effectively and accurately manufactured by a process including the following steps: i) patterning the thermal insulator layer of the plate parallel to the resistive layer to form cavities; ii) depositing a thermal conductor, e.g. a blanket layer, of desired thickness covering the thermal insulator layer and extending along the cavities; and iii) removing the thermal conductor outside the cavities along the plate, e.g. by grinding or polishing.
However, using thin conductive layers 131 may cause problems when the pattern is scaled up to larger areas. Preferably a homogeneous current density distribution is achieved with a minimal voltage drop. When the printing plate 10 becomes larger, the voltage drop may become much more significant. As the conductivity of the conductors and layer thickness in the cavities is limited, a solution to reduce the sheet resistance is by adding highly conductive busbars or increased thickness of the conductive layers away from the cavity 12. The busbars 144 may be embedded into the printing plate to keep a flat topology.
A thermally insulating gap 145 may span a distance between the cavity and conductors: e.g. in an embodiment of copper conductors on a glass substrate, in relation to specific values of thermal diffusivity:
A minimum gap distance is about ~ 1 µm or higher, preferably, larger than 10 µm, or even larger than 100 µm.
The higher the heat flux in the cavity the smaller the distance can be without any heat sink effects that influence the printing process.
It is preferred to operate at high voltages to reduce the current density. However, the maximum voltage is limited by the dielectric strength and thickness of the insulator layer. Too high voltages will lead to electrical breakdown between the two electrodes. As a non limiting example, at a film thickness of 1 µm, silicon dioxide can have a dielectric strength of 0.5 GV/m. This means that the at this thickness the maximum voltage would be 500 V. High dielectric strength materials are desired as it allow to use high voltages while keeping the stack as thin as possible. The thinner the stack, the fewer topology and thermal problems next to the cavities. For example, a thin undesired thin film of ink can be transferred or unwanted shear forces due to insufficient edge heating occur when the insulator film is relatively thick.
As a printing plate 10 is typically less than 1 mm thick, the dimensional stability of such a plate may be challenging, especially for larger plates (low aspect ratio, thickness versus width). This makes it difficult to control the printing gap over a large area. To control this gap, ideally the printing plate is pulled against a rigid and dimensionally stable vacuum chuck 30. This vacuum chuck 30 can be flat or curved (convex or concave) to hold the printing plate in position. Ideally the vacuum chuck has holes 301 for the electrically conductive pins to make contact with the via’s in the printing plate 10. The vacuum chuck 30 can be cooled to maintain a stable printing temperature and has a vacuum sealing plate 302 sealed by seal 304 and having vacuum connection 303 that provide for electrical connections 146 to the control circuitry 14.
For automation, spacer pins 142 may generate a trigger signal when all of them are in touch with the printing plate 10. Connecting control circuitry 14 for top and bottom conductors and respective vias 143 are separated in separate circuitry boards 145, This means that the printing process can be immediately initiated when the correct sample holder is in the correct position. This speeds up the process and it ensures printing will only occur when there is a sample holder underneath. Triggering can be done electronically by means of a resistance measurement between the spacer pins 142. The spacer pins need to be electronically connected and make electrical contact with triggering pads on the printing plate 10.
The electrical requirements for the print head 100 depend on the thermal diffusivity of the printing plate material. During the heating process heat will dissipate into the ink 50 and plate. The higher the thermal diffusivity, the higher the heat flux needs to be to generate a peak temperature at the ink/plate interface at which the solvent explosively evaporates. Typically this means going from room temperature to around 150 - 400 C, depending on the solvent type in the ink.
Evaporation timescale requirements:
When the boiling point of the inks’ solvent is approached too slowly (> 1 ms), then the generated bubbles tend to be too big, which leads to more residue of solids at the interface between two bubbles. The faster the ink is being heated, the smaller the bubble sizes and will result in less residue (solids in the ink). It is important to limit the amount of residue because it enables printing multiple times in a stable manner without intermediate cleaning steps. For this reason high a high heat flux is preferred.
Heat flux requirements for a glass/quartz printing plate:
At lower thermal diffusivities the required heat flux goes down (plastics), and at higher thermal diffusivities (metals) the required heat flux to locally heat the ink will go up significantly.
Besides using the resistive heaters for ejection, they can also be used to change the material properties of the functional ink. Heating will occur on a longer time scale to avoid boiling effects and bubble formation.
While the invention has been illustrated and described in detail in the drawings and foregoing description, such illustration and description are to be considered illustrative or exemplary and not restrictive; the invention is not limited to the disclosed embodiments. In particular, unless clear from context, aspects of various embodiments that are treated in various embodiments separately discussed are deemed disclosed in any combination variation of relevance and physically possible and the scope of the invention extends to such combinations.
Other variations to the disclosed embodiments can be understood and by those skilled in the art in practicing the claimed invention, from a study of the drawings, the disclosure, and the appended claims. In the claims, the word “comprising” does not exclude other elements or steps, and the indefinite article “a” or “an” does not exclude a plurality. A single unit may fulfill the functions of several items recited in the claims. The mere fact that certain measures are recited in mutually different dependent claims does not indicate that a combination of these measured cannot be used to advantage. Any reference signs in the claims should not be construed as limiting the scope. In addition, the functional material may have a wide-range of viscosity, possible to transfer (from 50 Pa.s to 1000 Pa.s (non-exhaustive)). Contrary to stencil printing and screen printing it is a non-contact direct write method wherein the distance to the substrate surface is preferably kept higher than 1 micron. The transfer accuracy of the functional ink is dependent on the printing gap. Typically the printing gap needs to be in the same order as the feature size to obtain a good printing accuracy. When a small printing gaps are required, it becomes technically more difficult to align the print head with the sample holder. Therefore it is preferred to control the gap by using spacer pins and press the print head and sample holder together. It is preferred to have at least three degrees (heave, roll, pitch) of freedom between the print head and sample holder. In an example pastes can be formed as follows:
with viscosities ranging between 100 en 1000 Pa*s. In contrast to conventional screen print paste and inks shear thinning may be minimized e.g. by adding fumed silica to prevent flow of the past on the acceptor substrate. This can also be prevented by curing/drying the ink e.g..by laser irradiation, which can be conveniently realized by a high solids content and low solvent content.
Number | Date | Country | Kind |
---|---|---|---|
20171361.7 | Apr 2020 | EP | regional |
Filing Document | Filing Date | Country | Kind |
---|---|---|---|
PCT/NL2021/050270 | 4/23/2021 | WO |