Systems and methods for modifying substrates - Patents.com

Abstract

Various embodiments of the present technology generally relate to planarizing a substrate. More specifically, some embodiments of the present technology relate to versatile systems and methods for precision surface topography optimization, known as planarization on a nominally planar substrate. In some embodiments, a method for planarizing a patterned substrate using an inkjet can determine global and nanoscale topography and pattern information of the patterned substrate. Based on the global and nanoscale topography and pattern information, a droplet pattern can be determined and then dispensed onto the patterned substrate. A gap between the patterned substrate and a superstrate can be closed so that the dispensed droplets form a substantially continuous film. The substantially continuous film can be cured, and the superstrate can be separated from the patterned substrate with the substantially continuous film.

Description

[Technical field]

[0001] (CROSS REFERENCE TO RELATED APPLICATIONS) This application claims priority to U.S. Provisional Application No. 62 / 779,091, filed December 13, 2018, which is incorporated by reference in its entirety for all purposes.

[0002] (Statement Regarding Federally Sponsored Research) This invention was made with Government support under Grant No. ECCS1120823 awarded by the National Science Foundation. The United States Government has certain rights in this invention.

[0003] (Technical field) Various embodiments of the present technique relate generally to planarizing a substrate, and more particularly, to versatile systems and methods for precision surface topography optimization, known as planarization, on a nominally planar substrate. [Background technology]

[0004] Processes such as Programmable Adaptive Inkjet for Nanoscale Thin Films (PAINT) can be used to deposit films of tailored thickness with near-zero material waste. PAINT can deposit films over large areas with virtually no dependency on substrate type, thickness, or material choice. By design, PAINT can also virtually isolate systematic parasitic effects such as low spatial frequency surface topography and inkjet drop volume variations, preventing them from compromising the final film thickness.

[0005] However, in such processes, where the surface of the substrate is of primary importance, such as in planarization, surface profile metrology is required to obtain a map of the nominal and parasitic topography on the substrate, which is required to measure the topography of the final surface closest to the "superstrate" (used to encourage the inkjet-dispensed liquid organic material droplets to bond laterally across the substrate) immediately prior to performing the PAINT process. Summary of the Invention

[0006] Systems and methods for substrate planarization are described. More specifically, some embodiments of the present technology relate to versatile systems and methods for precision surface topography optimization, known as planarization on a nominally planar substrate. In some embodiments, a method for planarizing a patterned substrate using an inkjet uses global and nanoscale topography and pattern information of the patterned substrate. For example, the topography and pattern information can be loaded from a data-encoded digital file. Based on the global and nanoscale topography and pattern information, a droplet pattern can be determined and then dispensed onto the patterned substrate. A gap between the patterned substrate and the superstrate can be closed such that the dispensed droplets form a substantially continuous film. The substantially continuous film can be cured, and the superstrate can be separated from the patterned substrate with the substantially continuous film. According to various embodiments, the gap can be closed at a first station and the curing occurs at a second station. Curing can occur at the end of a desired process time selected to be long enough so that air bubbles are substantially alleviated, yet short enough so that undesirable parasitics do not affect planarization performance.

[0007] In some embodiments, the droplet pattern can be determined from process models, optimization schemes, and / or experimental data. For example, the droplet pattern can be calculated using model-based optimization with inputs including global and nanoscale topography and pattern information, superstrate geometry, superstrate material properties, planarization material properties, inkjet droplet resolution, inkjet nozzle pitch, and tolerance information. A surface of the superstrate can be substantially more convex than a surface of the substrate at a forward spreading location. In some embodiments, the superstrate includes a sacrificial film, and the method further includes ablating the sacrificial film to initiate peeling to separate the superstrate from the substrate.

[0008] In some embodiments, planarization performance in a sub-region of a semiconductor device die can be prioritized. For example, the sub-region of the semiconductor device die may correspond to a portion of the semiconductor device die where substantially higher lithographic resolution is required during a subsequent lithographic process. As another example, the sub-region can be aligned perpendicular to the scanning direction of a photolithography scanner during exposure, although the focus of the photolithography scanner is a downstream process that benefits from surface topography control. In some embodiments, the inkjets can be aligned parallel to the sub-region.

[0009] Some embodiments provide for mitigation of hot spots resulting from bubbles or voids in a nanoimprint lithography process. Some methods can include identifying hot spots on a substrate. The hot spots can be caused by particles, bubbles, or voids. For example, the hot spots can be caused by particles that are still trapped between the superstrate and the substrate. The hot spots can be identified using a high resolution imaging system that images in the liquid phase or in the solid phase after UV curing. Then, based on the characteristics of the hot spots, a compensation solution can be identified and implemented to mitigate the hot spots. The compensation solution can include evaporation control, pressure zone chucks, digital micromirror devices, or programmable heater arrays.

[0010] A method for mitigating systematic hot spots resulting from a nanoimprint lithography process is provided according to various embodiments. The method can include identifying the hot spots prior to the start of the nanoimprint lithography process. Based on the characteristics of the hot spots, a compensation solution can be identified and implemented to mitigate the hot spots. In some embodiments, the compensation solution can be applied only to an area within the vicinity of the hot spots. The compensation solution can include evaporation control, a pressure zone chuck, a digital micromirror device, or a programmable heater array. The pressure zone chuck can be a programmable pressure zone chuck configured to squeeze on the superstrate, on the substrate, or on both sides of the vicinity of the hot spots. The programmable heater array can be configured to induce thermocapillary flow in the vicinity of the hot spots. The compensation solution can include selecting a thickness variation of the superstrate to mitigate systematic hot spots to achieve a desired planarization.

[0011] In some embodiments, hot spots can be classified based on the characteristics of the hot spots. A drop pattern that compensates for the hot spots can be calculated. Then, the drop pattern can be applied. The hot spots can be systematic or non-systematic hot spots. Systematic hot spots can include variable process time, variable pattern density, or contact line pinning. Systematic hot spots can result from, for example, upstream, in-process, or downstream processes (e.g., in a nanoimprint lithography process or a planarization process).

[0012] In some embodiments, the selected thickness variation may be a smooth function of spatial coordinate or a step function of spatial coordinate. The thickness variation may be a function identified by an alignment mark. In some embodiments, the alignment mark may be aligned to less than 10 microns relative to an alignment mark on the substrate. The function of thickness variation may be selected based on an alignment mark that is aligned to better than 1 micron relative to an alignment mark on the substrate.

[0013] Embodiments of the present technology also include a computer-readable storage medium that includes a set of instructions for causing one or more processors to perform the methods, method variations, and other operations described herein.

[0014] Some embodiments provide a system for planarization comprising a processor, a substrate shape modulating chuck under control of the processor, a superstrate shape modulating chuck under control of the processor, and a memory, which may store a set of instructions that, when executed by the processor, causes a machine to identify a time-varying shape of the substrate and the superstrate that maximizes a relative curvature between the substrate and the superstrate for all time periods between an initiation of contact between the substrate and the superstrate and a completion of substantially conformal contact. In some embodiments, at least one of the curvatures of the superstrate and the substrate is concave at at least one time between an initiation of contact between the substrate and the superstrate and a completion of substantially conformal contact.

[0015] Some embodiments can include a camera having a camera pixel size of at least 100 μm, at least 10 μm, or at least 1 μm. The camera can be RGB or grayscale. In some embodiments, the camera can be line scan, area, or telecentric. In some embodiments, the particle center location can be used to detect a small area around that point to look for defects caused by damage on the superstrate that may have propagated at the contact area with the particle.

[0016] A method for detection of systematic hot spots resulting from a nanoimprint lithography process can include fabricating a dummy template having features larger than a nanoscale resolution template. The dummy template can substantially mimic physical characteristics of the nanoscale resolution template. The mimicked physical characteristics of the nanoscale resolution template can result in substantially similar defects. The nanoimprint lithography process can be performed using the dummy template.

[0017] While multiple embodiments are disclosed, still other embodiments of the present technology will become apparent to those skilled in the art from the following detailed description, which shows and describes exemplary embodiments of the present technology. As will be recognized, the present technology is capable of modification in various aspects, all without departing from the scope of the present technology. Accordingly, the drawings and detailed description are to be regarded as illustrative in nature and not limiting. [Brief description of the drawings]

[0018] An embodiment of the present technology will be described with reference to the accompanying drawings. [Figure 1A] 1 provides an example of a flattening metric. [Figure 1B] 1 provides an example of a flattening metric. [Diagram 2] 1 is a plot showing the variation of shrinkage with geometric geometry. [Diagram 3] Shown is a levitating droplet versus a penetrating droplet. [Figure 4] FIG. 1 illustrates an example of planarization for scanning photolithography showing substantially more accurate planarization performance in the x-direction than in the y-direction in accordance with one or more embodiments of the present technology. [Diagram 5] 1A-1C illustrate different droplet spreading scenarios illustrating the desired relative shapes or curvatures of the superstrate and substrate. [Figure 6] 1 is a flowchart illustrating a sequence of operations for planarizing a substrate, in accordance with one or more embodiments of the present technique. [Figure 7] 1 is a flowchart illustrating a sequence of operations for identifying and compensating for hot spots in accordance with some embodiments of the present technology. [Figure 8] 1 is an example of a computing system that may be used to perform various calculations and / or control devices in some embodiments of the present technology.

[0019] The drawings are not necessarily drawn to scale. Similarly, for purposes of discussion of some embodiments of the present technology, some components and / or operations may be separated into different blocks or combined into a single block. Moreover, while the present technology is susceptible to various modifications and alternative forms, specific embodiments are shown by way of example in the drawings and are described in detail below. However, the intention is not to limit the present technology to the specific embodiments described. On the contrary, the present technology is intended to encompass all modifications, equivalents, and alternatives falling within the scope of the present technology as defined by the appended claims. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0020] Various embodiments of the present technology generally relate to planarization of substrates. More specifically, some embodiments of the present technology relate to versatile systems and methods for precision surface topography optimization, known as planarization on nominally planar substrates. According to various embodiments, these techniques include obtaining a planar top film that can match low and medium spatial frequency substrate topography in some cases while planarizing high spatial frequency features. Various embodiments of this process are referred to herein as Programmable Adaptive Inkjet of Nanoscale Thin Films (PAINT), as described in U.S. Patent No. 9,415,418, entitled "Programmable Deposition of Thin Films of a User-Defined Profile with Nanometer Scale Accuracy," which is incorporated herein by reference in its entirety for all purposes.

[0021] PAINT uses inkjets to dispense droplets of precursor monomers onto a substrate. The substrate surface may be pretreated to enhance spreading of the monomer and / or adhesion of the polymerized material. If the inkjets have multiple nozzles, the desired substrate area can be covered with the required droplets in a few seconds or less, using a scanning stage to drive the inkjets relative to the substrate while still controlling the volume and position of each dispensed droplet. For each desired film thickness profile, the optimal droplet volume and position can be obtained from the model.

[0022] After droplet dispensing, an optimally flexible superstrate, curved with the aid of back pressure or gravity, can be lowered such that initial contact with the droplet is made at its front side. This creates a liquid front that spreads outward rapidly and joins with the droplet to produce a continuous film. This substrate-fluid-superstrate "sandwich" can then be allowed to evolve for a desired duration, after which the monomers are cured by light or thermal energy to crosslink into polymers. The superstrate can then be separated from the sandwich, leaving a thin polymer film on the substrate.

[0023] Although it has been assumed that the superstrate is located above the substrate while the process is taking place on the substrate surface, the principles of the present technology apply to embodiments where the relative positions of the two surfaces are reversed, for example, where the substrate is located above the superstrate. Similarly, in the inkjet step, it may be the substrate or the superstrate onto which the droplets are dispensed, depending on the relative positions of these surfaces. Also, in the separation step, either the superstrate or the substrate may be separated from the other without changing the nature of the process. In the following, the substrate is described as having a surface that needs to be "painted", but the principles of the present technology are not so limited and may include other surfaces that need to be "painted".

[0024] The use of the appropriate superstrate and substrate combination is important. Generally, the properties of the substrate are immutable and cannot be altered for the process. Thus, typically, only the properties of the superstrate can be altered. However, in the general case, it is both the substrate and the superstrate that affect the process dynamics, and thus is presented as a combination, even though in practice only one of the two is free to be altered relative to the other. This superstrate and substrate combination must have an "optimum flexibility"; however, the stiffness must be high enough to encourage the superstrate / substrate to laterally bond to the monomer droplets rather than capture the individual droplets as islands wrapped around them, and low enough that the strain energy stored in the superstrate and substrate due to deformation does not significantly affect the dynamic behavior of the thin film fluid prior to the curing or crosslinking of the monomer. It must also be low enough and insensitive to substantially reduce the presence of parasitic or undesirable topographical signatures.

[0025] In general, either the superstrate or the substrate, or both, may be non-rigid depending on the application. Flexible substrates or superstrates may include sheets, rolls, or other manifestations of plastic materials such as polycarbonate, polyethylene terephthalate, etc., as well as rolls, sheets, or other manifestations of thin glass-like materials (e.g., Willow glass from Corning). However, for purposes of the description of the present technique, the substrate is assumed to be rigid (e.g., a silicon wafer) and held against a chuck. The superstrate, on the other hand, may be bent. The substrate may also be bent without disturbing the basic concept behind the PAINT process.

[0026] According to various embodiments, the superstrate may be constructed of a material such as glass, ceramic, or polymer with a flat surface and an appropriate thickness to provide the desired flexibility. Other embodiments may include the use of a thin, flexible membrane attached to a more rigid substrate with engineered or naturally occurring pores (such as anodized aluminum oxide (AAO)). The substrate may be attached to a superstrate chuck with two vacuum zones (one along the outer annulus, for example, and one toward the inside). Vacuum control is transferred through the porous substrate to the thin membrane attached to it.

[0027] If a superstrate is used in the spreading step, all vacuum zones are engaged to allow the thin film to be attached to the rigid substrate. After spreading is accomplished, the inner vacuum zone can be disengaged (or even allowed to blow positive pressure) so that the thin film can be supported only by the outer vacuum zone. This serves to substantially reduce the effective thickness (and therefore stiffness) of the superstrate that is desirable for the development and capture of pre-equilibrium transients (discussed further below). In some embodiments, the thickness of the thin film superstrate can be from about 100 nm to about 100 μm for some materials (e.g., polymers, glasses, ceramics, etc.). The porous substrate can have a thickness of 10 μm to 5 mm in some embodiments.

[0028] Other superstrate embodiments can be realized by having thicknesses that vary at different superstrate cross sections depending on which cross sections of the superstrate are involved in forming a sandwich with the fluid and substrate, which allows the thickness of the superstrate to be a control knob in the presence of substrate topography with varying density.

[0029] Other characteristics of the superstrate include promoting the spreading and dissolution of outgassed materials before curing, followed by separation after curing. This typically requires the superstrate to have good wetting properties with the precursor liquid, followed by dewetting properties with the post-cure polymer. Such properties can be obtained by coating the superstrate with a thin film of oxide or metal. The surface of the superstrate can also be treated.

[0030] Another step in this process is alignment, as described in U.S. Patent Application Publication No. 2017 / 0333940, entitled "Precision Alignment of the Substrate Coordinate System Relative to the Inkjet Coordinate System," which is incorporated by reference in its entirety for all purposes.

[0031] Some representative materials that may be used as the deposited film include etch barrier solutions, inkjettable compositions such as mv-Cur from Microresist Technology, Inc., and Monomat® from Canon Nanotechnologies, Inc. The substrate may also need to be pretreated to adjust wettability. Some materials used to promote adhesion between the substrate and the deposited film include ValMat® and Transpin® from Canon Nanotechnologies, Inc.

[0032] Fluid flow in domains with lateral length scales much larger than the height (thin film) can be solved using lubrication models that assume that the flow is primarily parallel to the surface and that the normal pressure gradient is zero. Typically, this leads to nonlinear models that can be linearized for lower computational cost and better understanding of the process mechanisms. Linearization results in a characteristic process time scale that is given by:

[0033]

number

[0034]

number

[0035] where h0 is the average film thickness, R is the horizontal length scale, typically the radius of the deposition region, a and b are exponents that depend on the exact nature of the process, and D eff is the effective bending stiffness of the superstrate (if the substrate is rigid, and vice versa), which depends on Young's modulus E, the thickness b of the superstrate, and Poisson's ratio v. If the substrate is also flexible, the effective bending stiffness will include a combination of both. In general, τ PAINT A larger value of is desirable because it provides a longer time to capture the pre-equilibrium transient. From equation (1), D eff The smaller the τ paint Therefore, appropriate process conditions (high values ​​of τ paint (meaning) requires low superstraight stiffness.

[0036] This term (τ paint ) to obtain the dimensionless process time (t * ) by the spreading time (τ) of the actual process fluid and τ paint It can be defined as the ratio of

[0037]

number

[0038] The objective of the various embodiments is to minimize lateral fluid redistribution so that the final film thickness has a strong correlation with the initial material distribution, so that the "programmable" nature of PAINT can be achieved by predefined positions and volumes of the inkjetted fluid droplets. In other words, the lubrication model reveals the fact that it is essential to capture pre-equilibrium transients in the evolution of the sandwich, since the equilibrium state allows only one possible steady-state solution (which is a flat top in the case of planarization), which is typically undesirable and destroyed by parasitic low spatial frequency topography of the substrate. This defeats the purpose of programmable deposition of thin films. This concept of "capturing pre-equilibrium transients that are substantially correlated with the inkjet fluid droplets" while minimizing the effects of substrate and superstrate topographies at different spatial length scales is central to the PAINT process.

[0039] In the field of planarization, a proper definition of the average film thickness h0 becomes important when high frequency spatial topography such as nanoscale patterns is present, because what is needed here is to overcome this high frequency topography while also minimizing any fluid distribution due to the mid- and low-spatial frequency topography. Thus, the average film thickness in equation (1) can be selected as a suitable average between the maximum and minimum film thickness values. For example, the minimum film thickness can be the residual layer thickness, and the maximum film thickness can be the residual layer thickness plus the maximum feature height. The suitable average can be a statistical measure of central tendency, such as an area-weighted average, or it can be the most conservative bound, which in this case can be the maximum film thickness. The definition of the average film thickness can include a measure of film thickness variation, as follows:

[0040]

number

[0041] However, h approx is an approximate measure of the average film thickness, h0 is the true average film thickness, and s is the standard deviation of the film thickness. Alternatively, the overall process time constant τ paintcan be kept substantially the same in areas of different pattern density by varying the thickness of the superstrate.

[0042] The transient characteristics of the model can be analytically determined by performing a linear analysis using the following equation:

[0043]

number

[0044] where r is in the nominal substrate surface coordinate system. It is important not to factor in the presence of substantial substrate topography to obscure the linearization of the model. However, establishing an appropriate substrate coordinate system that accounts for the substrate topography is necessary to ensure that the linearization produces feasible results. For example, linearization results in the presence of patterns or high spatial frequency substrate topography leads to the presence of the term s, i.e., the standard deviation, in the definition of the approximate average film thickness.

[0045] The linearized model can be solved analytically, thus greatly reducing the computational complexity and allowing solving a key aspect of PAINT, namely the optimal location and volume of the fluid droplet for a desired film thickness profile. In one embodiment of the process, only the droplet location may be optimized, keeping the droplet volume fixed at the minimum volume dispensed by the inkjet. Also, if the low spatial frequency topography is substantially similar across different wafers, it does not need to be measured every time. In one embodiment of the process, the optimal droplet volume and location can be obtained from an inverse optimization scheme around the lubrication model without any input to the low frequency nanotopography, especially when the dimensionless process time is low. In another embodiment of the process, the optimal droplet volume and location can be obtained from an inverse optimization scheme around images captured in real time from the process. In another embodiment, a hybrid approach can be chosen, where the lubrication model is augmented with a model based on real time data, a statistical model, or an empirical model.

[0046] In addition to inverse optimization for an optimal film thickness profile based on process mechanics, an important aspect of the PAINT process includes a wrap functional optimization scheme. The purpose of the functional optimization scheme is to correlate the desired film thickness with the desired functional performance of the process and the substrate. For example, in semiconductor planarization, it may be important to ensure that the planarization film does not end up being anti-reflective or has other desired optical properties. The computational cost of this higher level of constraints can slow down the inverse optimization scheme, and depending on how computationally expensive it is, an approach can be taken where the optical performance calculations are performed intermittently at a computationally inexpensive frequency, rather than for each suboptimal film thickness profile obtained during the optimization run. It may also be important to ensure that the thickness of the planarization film does not interfere with subsequent etchback.

[0047] Various embodiments of the present technology provide a wide range of technical effects, advantages, and / or improvements over computing systems and components. For example, various embodiments include one or more of the following technical effects, advantages, and / or improvements: 1) global (or near-global) planarization in the presence of nanoscale features; 2) integrated use of optimization and inverse optimization techniques to identify customized droplet patterns; 3) automated in-situ hotspot detection and mitigation integrated into processing manufacturing workflows; 4) use of customized algorithms for planarization; 5) cross-platform integration of machine learning to more efficiently identify hotspots and corresponding hotspot mitigation solutions; and / or 6) system designs to enable high-yield, high-throughput planarization through innovations in: (i) superstrate design; (ii) superstrate and substrate handling; (iii) UV curing mechanisms; (iv) substrate and superstrate separation techniques; and (v) parallel processing of multiple substrates on the same tool.

[0048] In the following description, for purposes of explanation, numerous specific details are set forth in order to provide a thorough understanding of embodiments of the present technology. However, it will be apparent to one skilled in the art that embodiments of the present technology may be practiced without some of these specific details.

[0049] The techniques introduced herein can be implemented as dedicated hardware (e.g., circuits), programmable circuits appropriately programmed with software and / or firmware, or a combination of dedicated and programmable circuits. Thus, the embodiments may include a machine-readable medium having stored thereon instructions that may be used to program a computer (or other electronic device) to execute a process. The machine-readable medium may include, but is not limited to, a floppy disk, an optical disk, a compact disk read-only memory (CD-ROM), a magneto-optical disk, a ROM, a random access memory (RAM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), a magnetic or optical card, a flash memory, or any other type of medium / machine-readable medium suitable for storing electronic instructions.

[0050] Phrases such as "in some embodiments," "according to some embodiments," "in an illustrated embodiment," "in another embodiment," and the like generally mean that the particular feature, structure, or characteristic that follows the phrase is included in at least one implementation of the technology, and may be included in more than one implementation. Moreover, such phrases do not necessarily refer to the same embodiment or different embodiments.

[0051] (Flattening performance) Planarization is a key process step in the field of semiconductor device fabrication. Maintaining the depth of focus within desired lithographic tolerances allows photolithography to be performed on multiple planar layers without loss of feature resolution. Its use is common in both front-end semiconductor device processing of transistor circuits as well as back-end semiconductor device processing applications such as packaging of fabricated circuits. Although process requirements may vary depending on the underlying patterns and desired process performance, some general aspects of the use of inkjet to achieve such planarization are described below.

[0052] The performance of a planarization process is generally captured in two metrics: (i) the degree of planarization, and (ii) the planarization length. The degree of planarization represents the extent to which a step pattern is planarized and is quantified as the difference in step height before and after planarization. The planarization length represents the distance over which some degree of planarization is achieved.

[0053] 1A-1B show these two metrics and how they relate to the planarization process. An ideal process would achieve 100% planarization with infinite planarization length to achieve what is known as global planarization. However, existing processes are not ideal and are inadequate to achieve global planarization. For example, spin coating can only achieve what is called planarization (i.e., a very short planarization length on the order of the feature pitch).

[0054] Chemical mechanical planarization (CMP) is the semiconductor industry's work horse for planarization today, but can typically only achieve local planarization (i.e., planarization length that is die order and highly dependent on pattern density within the die). For example, in the CMP process, the dependency of planarization performance on pattern density necessitates the use of techniques such as dummy fill, which consists of fabricating non-functional patterns with the same density as the functional patterns. This fabrication of non-functional patterns adds complexity to the photolithography process and can be detrimental to the electrical performance of integrated circuits (ICs). Furthermore, at current semiconductor manufacturing technology nodes below 14 nm, where photolithography processes are already highly complex (e.g., utilizing double and quadruple patterning techniques), substantial savings can be achieved if the need for dummy fill can be avoided without compromising planarization performance.

[0055] To this end, the primary goals of several embodiments of the planarization processes discussed in this disclosure are to: 1. Global (or near-global) planarization in the presence of varying features (e.g., varying density, pitch, feature size, etc.); 2. Compatibility with existing lithography processes, both front-end and back-end. These goals are achieved by optimizing both the software and the hardware behind the PAINT process.

[0056] (flattening parasitic) The ability of the PAINT process to achieve near global planarization can be compromised by the presence of several parasitic phenomena or factors ("parasitism"). One of these parasitics is the shrinkage of the UV-curable monomers used in the planarization process upon polymerization. The amount of monomer shrinkage depends on the geometric geometry of the underlying features. For example, a polymerized film shrinks less across a narrow trench compared to film shrinkage across a much wider trench.

[0057] In some embodiments of the present technology, the parasitics can be obtained using classical thin film shrinkage models derived from polymer elasticity and viscoelasticity theories and used to obtain optimal droplet volume and location. In other embodiments, the parasitics are obtained by experiment and used to obtain optimal droplet volume and location. Furthermore, a multi-step planarization process can be used to overcome the parasitics and obtain the desired planarization performance, although the first planarization process may provide moderate planarization performance and one or more subsequent steps are required to obtain higher planarization performance.

[0058] Bulk shrinkage characteristics can be obtained from experiments on unpatterned substrates and used as a parameter in obtaining shrinkage on features. An example of the variation of shrinkage with feature geometry is shown in FIG. 2. In plot 200 of FIG. 2, the x-axis shows the ratio of width to depth (i.e., width / depth) of a generic trench, and the y-axis shows the percentage of shrinkage observed upon polymerization of a thin film covering a trench, assuming a maximum shrinkage of 0.1 (i.e., 10%) in the absence of a trench. Here, the initial thickness t i and final thickness t f The shrinkage S of a thin film having f / t i It is defined as:

[0059] Another parasitic effect that can affect the planarization performance is the presence of a pre-existing film of a different material on the pattern that is deposited using various processes such as spin coating, vacuum deposition, etc. Each of these processes results in undesirable variations in the thickness of the corresponding film that changes the final pattern that needs to be deposited and planarized. It is important to have a robust model of these processes in the presence of different patterns or to obtain the final pattern information experimentally. Having a robust computational or analytical model of these thin film deposition processes in the presence of patterns or having experimental data from these processes allows for compensation of parasitics. For example, spin coating of an adhesion layer that promotes the spreading and bonding of the inkjet droplets may be necessary before the planarization step.

[0060] As mentioned above, spin coating results in smoothing of the features, which causes sharp edges or corners to lose their sharpness. Spin coating may also cause the sidewall angles of the features to become gentler (i.e., less steep), which changes how the inkjet droplets behave or flow on the pattern. It is important to capture this effect in order to spread the droplets to form a continuous film. In some embodiments, the impact of such pre- and post-processing steps on planarization performance is compensated for by placing droplets with optimal volumes at optimal locations. In one embodiment, the impact of such pre- and post-processing steps on planarization performance is determined by experimentation. While the above discussion is limited to thin film deposition processes, other exemplary pre-processing (upstream) steps include CMP, etching, wafer dicing that results in changes in wafer thickness, and exemplary post-processing (downstream) steps include etching, CMP, etc.

[0061] In general, the goal of a planarization process is to obtain a desired level of planarization performance. If this goal is to be computationally achieved by an optimization routine, it may require an objective function that can be a combination of the desired planarization length and the desired degree of planarization. In various embodiments, the desired planarization performance metric is substantially similar across the substrate. However, planarization performance may be more challenging in certain local regions, or sub-regions of a die, where each die is a device, due to higher lithographic resolution or other process constraints in the sub-region. There may be as many as 50 to 50,000 dies on a wafer, and the dies may or may not be identical. In such cases, the planarization performance requirements may be increased for these sub-regions, thus spatially adapting the goal of the planarization process from one sub-region to another within the same process step. In one embodiment, this can be done by appropriately modifying the objective function to minimize a weighted error of the actual planarization performance relative to the desired planarization performance, where the weight is higher in the regions requiring more accurate planarization performance.

[0062] If computational costs need to be managed, increased performance requirements in these sub-regions can be compensated for by reducing performance requirements in other sub-regions where yield is acceptable and no yield issues are expected by modifying the desired metric. The compensation or improvement of the desired planarization performance may be related to the input pattern geometry, such as the pattern GDS file. In one embodiment, the target related to the planarization performance can be spatially variable, where the source of spatial variation can be input pattern information or prior information about defect hot spots, where hot spots are defined as areas of undesired planarization performance or yield loss.

[0063] (Model-based or experimental prediction of droplet spreading) The formation of a continuous film on a patterned surface precedes the development of a thin liquid film as described above, and predictions of droplet spreading can be made using either model-based techniques or experiments. The following discussion focuses primarily on model-based predictions. However, experimental data can be used to support the model or used alone. In one embodiment, the understanding of how individual droplets combine can be supported by analytical, empirical, or statistical techniques, including a combination of one or more of these methods. Analytical methods simultaneously solve the dynamics of one or more droplets as they are squeezed under the action of a superstrate. The dynamic behavior of these droplets can include phenomena such as wetting, evaporation, etc.

[0064] The presence of a patterned substrate modifies the wetting and spreading behavior of these droplets compared to flat substrates. On a substantially flat substrate, the wetting of the droplets is substantially continuous. However, the presence of patterns can introduce discontinuities in the motion of these droplets, as described in "Dynamics of low capillary number interfaces moving through sharp features, Physics of Fluids 17, 122104 (2005)," which is incorporated by reference in its entirety for all purposes. These discontinuities arise from contact line pinning, i.e., the interface between the droplet and the surrounding environment, on the substantially sharp corners of these features. Contact line pinning is also known as hysteresis.

[0065] As discussed above, in early process steps such as spin-coating, some of the features may lose sharp corners and have less steep sidewall angles. This would allow the contact line to navigate these corners more easily with the aid of gravity, thus minimizing the effects of contact line pinning. In one embodiment, the determination of contact line pinning is complemented with the evolution of thin films and made by analogy with the phenomenon of friction and backlash in mechanical systems. This stick-slip behavior may result in a delayed onset of motion in the presence of external forces on the droplet that are a combination of gravity, capillarity, inertia, and viscosity. Some embodiments take advantage of the lack of hysteresis from features that do not have substantially sharp transitions. Such smooth transitions can be obtained if one or more spin-coating steps are performed prior to the PAINT process.

[0066] The spin-coated material can be substantially similar to the planarization material, or it can be a functional material that (i) improves adhesion between the underlying substrate or film and the planarization material (e.g., Transpin), or (ii) provides the required optical, thermal, or process properties, such as etch resistance (e.g., spin-on carbon). If spin-coating is not feasible, a multi-step PAINT process can be performed in which the desired planarization performance is achieved in multiple steps rather than in a single step. This is particularly useful when the planarization requirements vary substantially across the substrate, or when substantially high aspect ratio features need to be planarized, or when undesirable process parasitics such as shrinkage cause undesirable loss of process performance, or when stringent yield requirements require a substantially robust process operating window. If the overall film thickness becomes too large as a result of the multi-step PAINT process, an etch-back can be performed after each PAINT step (see, for example, U.S. Patent No. 8,394,282, entitled "Adaptive Nanotopography Sculpting," which is incorporated herein by reference in its entirety for all purposes).

[0067] (Stick-slip behavior of droplets) In the absence of any features, assuming that the characteristic lateral feature dimensions are much smaller (less than 10X) than the diameter of the droplet, the contact angle of a droplet sitting on the substrate will be given by the bulk contact angle of the liquid. In the presence of features, the equilibrium contact angle is also governed by the pattern geometry (or a roughness factor incorporating pattern geometry information). This pattern geometry may make the wetting droplet more wet (Wenzel state) or may transition to a Cassie-Baxter state where the droplet is substantially hydrophobic (see, for example, Wetting and Roughness, David Quere, Annual Review of Materials Research 2008 38:1, 71-99, incorporated herein by reference in its entirety for all purposes). Hydrophobicity is undesirable as it does not naturally wet the droplet and can occur when substrate features are deep, but it allows the droplet to be moved more easily when positional accuracy needs to be improved.

[0068] In one embodiment, such droplet movement can be facilitated by methods such as acoustic vibration or shape modulation of the substrate via a piezoelectric actuator array (see, e.g., U.S. Pat. No. 7,307,697, incorporated herein by reference in its entirety for all purposes), thermocapillary motion in the presence of a temperature gradient across the substrate, or precisely controlled tip / tilt of the substrate, as hydrophobic droplets tend to levitate on top of features without penetrating.

[0069] On the other hand, hydrophilic droplets may impale themselves inside the feature, thereby becoming "attached" to the feature, as shown in example 300 of FIG. 3. The transition from levitation to impalement is a function of the pattern geometry, the material of the substrate, and the pressure applied to the droplet (see, for example, Impalement of fakir drops, M. Reyssat, JM Yeomans and D. Quere, 20 December 2007, EPL (Europhysics Letters), Volume 81, Number 2, which is incorporated by reference in its entirety for all purposes). Thus, in regions where there is a transition from one feature density to another, the droplet stick-slip behavior may be significantly different. In some embodiments, it is desirable to have the droplet substantially adhere to the feature to prevent undesired droplet movement.

[0070] In one embodiment, the droplet behavior in this transition region can be examined by first applying pressure with the superstrate to exceed the threshold pressure required for droplet impingement, and then retracting the superstrate. This retraction does not pull the droplet away, as it is substantially attached to the substrate. The droplet that is not impinged on the substrate and is still sliding can then be moved. This process can be repeated several times to allow the droplet to reach the desired location. Alternatively, a multi-step planarization process can be performed that mitigates parasitic errors from one step to the next, especially since the pattern is substantially smoothed after the first step, thereby preventing undesired droplet movement.

[0071] Another manifestation of altered droplet wetting behavior occurs through droplet anisotropy, particularly in the presence of substantially one-dimensional features such as lines and spaces. In this situation, the droplet may spread preferentially in one direction (along the line) versus the other. In one embodiment, the equilibrium elongation of an individual droplet dispensed on such a feature can be determined by analytical formulation or experimental data, and thus can help determine the droplet spreading (see, for example, Anisotropic Drop Morphologies on Corrugated Surfaces, H. Kusumaatmaja, RJ Vrancken, CWM Bastiaansen and JM Yeomans, Langmuir, 2008, 24 (14), pp 7299-7308, incorporated herein by reference in its entirety for all purposes).

[0072] If abnormal droplet behavior, such as anisotropic spreading or stick-slip, is present, a multi-step planarization process can be performed in which the planarization performance is substantially improved from one step to the next. The first planarization step will substantially fill the feature and smooth out the discontinuities that lead to the abnormal droplet spreading. This will allow the planarization film in the subsequent step to encounter features with substantially reduced anisotropy, i.e., the "rough" pattern geometry that leads to contact line pinning. This relatively smooth topography of the substrate can then be measured and substantially planarized in the next planarization step.

[0073] (Computational Modeling) The analytical approaches described above rely on modeling of individual droplets. These analytical models can be simulated using tools such as molecular dynamics, lattice Boltzmann techniques, or numerical solutions of the Navier-Stokes equations. In a typical planarization scenario, tens of thousands of droplets may be dispensed, which would require the use of substantial computational horsepower to solve the spreading and wetting behavior of individual droplets to form an initial film. Such computational resources may be available in the form of supercomputers or clusters of servers, but there may be situations where such resources are not available. In such scenarios, the use of analytical models can be augmented with empirical and statistical techniques. These techniques rely solely on experimental data, or analytical insights from modeling a few droplets can be combined with experimental data to develop statistically relevant models that describe the behavior of a few droplets in the presence of substrate patterns.

[0074] Statistical relevance can be determined with the help of methods of uncertainty quantification such as Bayesian statistics, confidence intervals, etc. The selection of the relevant model (analytical, empirical, statistical) and the number of parameters in each model can be made based on the desired accuracy and precision requirements. These precision requirements may not be spatially uniform, e.g., more stringent in regions where there are transitions in the substrate pattern. In one embodiment, the determination of the initial film from several drops is made with the help of an empirical technique in which each drop is modeled as a disk with a height and radius commensurate with the volume of the drop. This is useful when the drop volume is less than 1 pl and the drops are highly wetting.

[0075] The curved sidewalls of the disk may not be perfectly steep and may be modeled as having a slope. Additionally, the sidewall profile may be described by other functional expressions, including trigonometric series, polynomial series, step functions, error functions, etc. In each of these expressions, the choice of the number of parameters is important and can be adjusted based on the desired accuracy of the results. The tradeoff is computational expense, given that adding more parameters increases the time to calculate the results. In one embodiment, a substantially accurate expression may be selected for initial offline calculations, while an expression with fewer parameters may be selected for subsequent or real-time calculations.

[0076] The inverse optimization scheme may rely on the input of the desired film thickness and the starting substrate topography. For high-throughput planarization with process times less than 10 seconds and a sufficiently optimal superstrate, the global substrate topography may not contribute much to the evolution of the superstrate-fluid-substrate sandwich thin film. This means that the computational model may not incorporate the global substrate topography as an input, which would eliminate the need to measure this topography on every wafer that is planarized.

[0077] The computational model may have the ability to linearize or reduce the complexity of the nonlinear thin film fluid model that governs the behavior of the fluid trapped between the two plates. Linearization involves having the average film thickness dependent on small changes in film thickness over an area of ​​the substrate, typically defined as the characteristic length scale of the film thickness. In the presence of substrate patterns, the film thickness changes rapidly locally. This complicates the linearization, and the average film thickness is no longer an accurate measure of the characteristic film thickness. The average film thickness may need to be augmented with additional terms, such as the film thickness variance, given by the standard deviation of the film thickness.

[0078] The optimization scheme relies on minimizing the error between the actual film thickness and the desired film thickness. If the process time is short and there is a strict requirement to have no voids, the error metric may be expanded to include the presence or absence of voids, the number of voids that exceed the tolerance dimensions, the total area of ​​voids, or a combination of these. It is also important to simulate the full nonlinear model with the optimal parameters generated by the inverse optimization scheme on the linearized model, since if the linearization of the full nonlinear model is not accurate, the error can be high for planarization.

[0079] (Bubble reduction) A significant issue in combining resist droplets while forming a continuous film is that of bubble mitigation. The presence of such bubbles and voids can cause defects in the final planarized film and can occur when gas trapped in the bubbles cannot escape through the substrate or superstrate due to low porosity. This is also an issue in areas where droplets may be spaced apart from each other due to film thickness and underlying feature constraints. Mitigation of bubble trapping may be achieved by increasing the time it takes for the droplets to combine or by providing a gas such as CO2 to the environment that can dissolve quickly through the resist and the surfaces that contact the resist.

[0080] (Inkjet accuracy and precision) One of the important aspects of PAINT is the determination of the evolution of a thin film fluid between a substrate and a superstrate, where a thin fluid film is formed using individual droplets. In this disclosure, embodiments of the PAINT process are described that include substrates with nanoscale features. The evolution of the thin film and any model for it depends on the order of film thickness, which defines the relevant process time scale. Some embodiments may use different film thickness values ​​for different regions of the substrate, assuming that the presence of patterns locally changes the film thickness value. This would follow the accuracy requirements for planarization of different regions of the substrate.

[0081] This can be augmented by the ability to divide the substrate into various domains, each domain having a well-defined average film thickness value that correlates substantially well with the tolerance and accuracy requirements at that location. Because the underlying substrate can have substantially similar patterns on several substrates, the domain division can also be hard-coded in the superstrate design by having different superstrate thicknesses at different locations that also correspond to increased accuracy requirements. For example, in pattern density transition regions, i.e., regions where the pattern density changes substantially, it may be advantageous to have higher accuracy requirements to address the complexity of this pattern. At the same time, inkjets may introduce inaccuracies in the film thickness evolution through inaccurate droplet ejection onto the substrate given the inherent variability of the inkjet printhead.

[0082] Typically, this variability is expressed as the precision of the drop placement relative to a complete grid. This precision is higher in the direction along the print head, i.e. along the inkjet nozzle row, and lower in the direction perpendicular thereto. Thus, areas requiring higher precision during the planarization process may require more precise inkjetting, which may occur substantially along the direction of the nozzle row. This can be achieved by orienting the substrate such that these critical features are substantially aligned along the direction of the inkjet nozzles. Such orientation and alignment can be achieved using a suitable rotation stage in conjunction with a wafer chuck, and is possible with prior knowledge of the substrate pattern. Techniques for achieving precision orientation and alignment of the substrate relative to a coordinate system are discussed in the above-mentioned U.S. Patent Application Publication No. 2017 / 0333940, entitled "Precision Alignment of the Substrate Coordinate System Relative to the Inkjet Coordinate System." Similarly, some embodiments may have more analog control of the inkjet drop volume at specific locations.

[0083] For example, in a typical inkjet, the dispensed drop volume is an integer multiple of the inkjet drop resolution. Thus, if the inkjet drop volume is 6 pL, the inkjet can dispense drops of 6 pL, 12 pL, 18 pL, etc. This leads to discrete changes in drop volume that may prevent the achievement of a desired level of precision in the critical area. In this regard, in one embodiment, it is desirable to be able to have more analog control over the inkjet drop volume (e.g., 6 pL, 7 pL, etc.) in such critical areas in order to obtain a more desirable initial film thickness profile that can be substantially correlated with the final film thickness profile.

[0084] This means that the overall inkjet drop volume resolution may differ from the minimum drop volume from the inkjet system. This analog control can be achieved using one or more of the representations described herein. In one representation of this embodiment, a single inkjet array is used, but the waveform is modified when dispensing drops near these critical sites to obtain the desired analog control over inkjet drop volume and velocity, and therefore placement accuracy and precision (see, e.g., Synder et al, “Automated tuning of high-order waveforms for picoliter resolution jetting of rheologically challenging materials,” Precision Engineering, incorporated herein by reference in its entirety for all purposes and available online at http: / / www.sciencedirect.com / science / article / pii / S0141635918300114). In another representation of this embodiment, the inkjet system consists of multiple arrays of inkjet nozzles, with each inkjet nozzle array calibrated to a different drop volume resolution and minimum drop volume. In another representation, the above two approaches can be combined. In other words, the nozzle or set of nozzles in each head identified with the highest accuracy is used in areas having high precision or accuracy requirements.

[0085] Multi-nozzle printheads are usually aligned along a single direction of the substrate. This means that the accuracy of the inkjet is typically higher in the direction parallel to the nozzle row. This situation is advantageous if the planarization process is performed in response to scanning photolithography as a downstream process. Planarization errors in the direction parallel to the scan direction can be tolerated to a much greater extent than errors in the cross-scan direction. Scanning photolithography can substantially tolerate surface topography errors that look like extruded surfaces. This means that any cross-section of a surface along the scan direction is substantially similar to a different cross-section of the same surface along the scan direction. Thus, aligning the inkjet nozzles parallel to the cross-scan direction allows for higher accuracy in that direction. In general, what is meant here is that if scanning photolithography is a downstream process to planarization, the accuracy requirements can be substantially higher in one direction than the other, as shown in FIG. 4.

[0086] 4 illustrates planarization 400 for scanning photolithography, showing that planarization performance should be substantially more accurate in the x-direction rather than the y-direction. In other words, the substrate profile cross sections AA', BB', ..., EE' are substantially close to linear, while the lines ABCDE and A'B'C'D'E', while substantially parallel, may not be substantially linear.

[0087] (Control of droplet and continuous thin film states) Achieving high droplet position accuracy and precision is a key requirement for the PAINT process when planarizing substrate topography. This is because the behavior of the droplet on the patterned substrate is affected by where the droplet is relative to the pattern geometry. As discussed above, the droplet may exhibit contact line pinning, levitation, or stabbing based on its interaction with the patterned substrate. As long as this behavior is consistent across different substrates, it can be incorporated as a feature to obtain the droplet volume and location. This can be done through precision experiments with send-ahead wafers that can provide information on systematic hot spots or regions with repeatable but undesirable planarization performance. These systematic hot spots may be determined from simulations of the system or a combination of simulations and precision experiments with send-ahead wafers. These systematic hot spots can be caused by, but are not limited to, the following mechanisms: unexpected flow asymmetries (which can result from anisotropic features such as gratings, contact line pinning, patterns on the substrate), inaccuracies in droplet volume, droplet position, errors in aligning the superstrate with the substrate, shrinkage, evaporation, etc.

[0088] Hot spots can be detected in experimental systems with multiple AFMs / AFM arrays specifically configured to address them. However, there can be inaccuracies that can impair planarization performance if there are random hot spots such as droplets on one substrate that are substantially pinned, but not on a different substrate. To achieve the desired precision and accuracy of droplet position and volume, several different strategies can be employed. One strategy can be to use a partial vacuum in the vicinity of the droplet that needs to be altered. The use of a partial vacuum can locally change the volume of the droplet and can also induce displacements on the substrate, including overcoming contact line pinning.

[0089] A partially evacuated or pressurized environment can also be used to enhance or suppress evaporation of the dispensed droplets to further control the volume, position, or wetting angle of the droplets governed by contact line hysteresis on the patterned surface. A partially evacuated or pressurized environment has been used to control undesirable effects from evaporation and can be done with the aid of solvent vapor. This can also be used when the droplets are bonded to a continuous film. The surrounding environment can be controlled to affect parasitics such as evaporation from the open side of the superstrate-fluid-substrate sandwich. A complementary strategy can be to use an individually addressable single air jet or an array of air jets above the superstrate or below the substrate, preferably after the formation of the continuous film. The extra air pressure from these air jets can locally deform the superstrate or substrate, thereby effectively providing a control knob to control the flow of the continuous film.

[0090] A combination of air pressure and vacuum may also be used in the form of a pressure zone chuck, an embodiment of which is disclosed in U.S. Patent No. 6,982,783, which is incorporated herein by reference in its entirety for all purposes. This zone chuck may be used to control the transient relative shape or curvature of the substrate and superstrate such that droplets are forced to merge and spread to avoid trapping air bubbles or voids between the initiation of contact and the achievement of substantial conformality between the substrate and superstrate.

[0091] Different relative shapes or curvatures 500 of the substrate and superstrate are shown in FIG. 5. The best case scenario is a substantially convex superstrate and a substantially convex substrate. If one of the two surfaces is substantially concave or flat, it is desirable to make the other surface substantially more convex. This shape control is particularly relevant when droplet spreading and film formation occurs in areas of curvature on the superstrate and substrate, such as in areas where there is substantially fixed contact between the respective chucks and one surface, such that shape control of the other surface overcomes this curvature, allowing droplet spreading and film development with substantial bubble mitigation.

[0092] Another strategy could be to use controlled thermal energy. Heat can be used to partially evaporate droplets and change their volume, while acoustic energy can be used to locally displace them to meet droplet precision requirements or to affect thin film flow after formation of a continuous film. Thermal energy can be provided from the substrate via a set of individually addressable micro-heaters on the substrate (e.g., as shown in U.S. Patent Application Publication No. 2017 / 0131640, which is incorporated by reference in its entirety for all purposes for high precision thermal actuation and temperature control), or via localized infrared light focused by a superstrate using a spatial light modulator such as a digital micromirror device (DMD).

[0093] Alternatively, the superstrate can be designed with a layer of transparent conductive electrodes or metal nanoscale mesh patterns that can heat the superstrate when supplied with electric current. The heat is transferred to the droplets or continuous film on the substrate, allowing them to move by thermocapillary effect or evaporate to change volume. The transparent electrodes or metal mesh patterns are substantially transparent and do not block UV exposure through the superstrate. Electrical energy can also be used directly to induce thin film flow or displacement of individual droplets via a phenomenon called electrowetting. This requires the use of electrode patterns on the superstrate, which are then coated with a transparent dielectric that ultimately contacts the droplets or film. In this disclosure, the concept of controlling individual droplet volume in flight or on the substrate, or by modifying the piezo inkjet waveform, is referred to as analog droplet volume control.

[0094] Droplet position accuracy can also be improved by moving the substrate closer to the inkjet. Measurement of the droplets on the substrate can be performed by imaging the droplets onto the substrate, aligning them against a precise XY grid to determine the droplet location and volume before contacting the superstrate. Imaging with sufficient resolution across a full wafer, especially one with a diameter of 300 mm, can lead to prohibitively expensive or cumbersome optics, considering that this step needs to occur quickly to allow a sufficiently rapid transition to superstrate contact. Thus, metrology can be prioritized in areas with known or anticipated yield problems. Coarse metrology can be performed on the remainder of the wafer.

[0095] (Super Straight Design) The presence of nanoscopic features on the substrate with steep gradients and variable densities can complicate the evolution of the thin continuous film of liquid sandwiched between the superstrate and the substrate. Model-based predictions, when involved, can be highly nonlinear and computationally expensive, so it may be desirable to linearize and simplify the model. Thus, the model can be augmented with empirical experimental evidence to determine the optimal geometric parameters required for the superstrate.

[0096] For example, the thickness of the superstrate can be determined based on a film thickness norm, where the norm can be a 2-norm, an infinity norm, etc. This thickness of the superstrate can be verified against experiments to determine an optimal value based on a defined planarization metric. In other embodiments, the substrate that needs to be planarized can be patterned with features that are substantially larger than the nanoscale features typically found in semiconductor planarization, because thin film lubrication defines an approximate spatial wavelength below which the fluid will substantially planarize and above which the fluid can conform to the underlying topography. Thus, any features smaller than this critical wavelength do not need to be patterned for testing, for example, to determine the optimal superstrate thickness. In this embodiment, the test wafer or send-ahead wafer is patterned with substantially larger features without compromising the planarization performance. This would allow testing of the planarization performance at a substantially lower cost, considering that it is much cheaper to fabricate substrates with microscale features than substrates with substantially small nanoscale features.

[0097] The thickness of the superstrate is a key process parameter that defines the performance of this planarization process because it has an exponential effect on the time it takes for the liquid film to planarize a feature. Thus, any variation in the thickness of the superstrate can induce substantial variations in the planarization performance, given that the process time is typically dictated by throughput or other constraints. It is therefore imperative to ensure that undesirable variations in the thickness of the superstrate are minimized. The selection of the thickness variation tolerance can be determined from the allowable tolerance in the process time.

[0098] Alternatively, the thickness variation in the superstrate can be measured with high precision before performing the planarization process. This spatial thickness variation can be incorporated into models or experiments to minimize any parasitic behavior during the process. In one embodiment, the total thickness variation (TTV) of the superstrate is kept within a sufficient tolerance such that the bending stiffness variation is within 10% of the desired stiffness variation, which is inversely proportional to the time it takes the liquid film to planarize the feature. In one embodiment, the TTV of the superstrate is kept within a sufficient tolerance such that the bending stiffness variation is within 5% of the desired stiffness variation, which is inversely proportional to the time it takes the liquid film to planarize the feature. In one embodiment, the TTV of the superstrate is kept within a sufficient tolerance such that the bending stiffness variation is within 1% of the desired stiffness variation, which is inversely proportional to the time it takes the liquid film to planarize the feature.

[0099] Although the superstrate may have undesired thickness variations, the thickness of the superstrate may have intentional spatial variations. For example, planarizing a wafer can take anywhere from a few seconds to almost a minute. In this process, the fluid film originating from the initial contact point (assumed to be the center in this embodiment, but may be other points, lines, or regions, including those substantially closer to the edge) and extending radially outward to the edge of the wafer will be exposed to the superstrate for a longer period of time than the edge. Additionally, other parasitic center-edge variations may exist due to systematic errors in upstream processes or expected systematic signatures from downstream processes.

[0100] The thickness of the superstrate can be varied to compensate for such variations. For example, in the case of unequal process times, the superstrate can be intentionally made thin at the initial contact location or region, with the thickness gradually increasing toward the final contact location or region. A PAINT process can be used to create this graded thickness superstrate (see, for example, U.S. Patent No. 9,415,418, entitled "Programmable Deposition of Thin Films of a User-Defined Profile with Nanometer Scale Accuracy," which is incorporated by reference in its entirety for all purposes).

[0101] In one embodiment, the time it takes for the film to form a continuous film substantially bubble-free (referred to as the spreading time) may take 1 to 10 seconds. In one embodiment, the time between the formation of the continuous film and the UV curing of the planarizing film (referred to as the waiting time) may take 1 to 60 seconds. Because there may be substantial variation in the total process time (combining the spreading time and waiting time), it is important to maintain the thickness variation of the superstrate within a desired tolerance. For example, in one embodiment, the thickness of the superstrate is designed with a maximum total process time in mind. This is because keeping the superstrate thicker than the thickness defined by the maximum total process time may result in undesirable parasitics in certain areas of the substrate.

[0102] When there is a pattern density variation on the substrate, the thickness of the planarization film may be different in various regions of the wafer, which may result in different process time scales across the wafer if the thickness of the superstrate is kept substantially similar. Typically, the pattern density variation has an abrupt transition from one region of the substrate to another, with each region being on the order of a few square millimeters. This means that any superstrate thickness variation should match the pattern density transition and can be patterned onto the superstrate by photolithography or laser processing based on the GDS file of the patterned substrate. The idea here is that the process time scale is a function of the product of the thickness of the superstrate and the thickness of the planarization film. Thus, if this product is kept substantially similar across the wafer by varying the thickness of the superstrate, the overall process time scale across the substrate will remain similar. This approach can potentially minimize parasitics that may result from unequal process time scales.

[0103] To keep process time scales substantially similar across wafers with regions of different pattern density, one strategy is to spatially vary the thickness of the superstrate in a corresponding manner, as described above. This can be possible using photolithography if the desired change in superstrate thickness from one region to another is less than about 30 microns. If the thickness change is higher, other techniques such as laser machining or cutting / polishing (for superstrate materials such as quartz, fused silica, etc.) may be required based on the desired spatial variation in thickness corresponding to the regions of different pattern density. Laser machining can also be performed such that the density of the workpiece is equivalent to the desired stiffness required for the superstrate. The superstrate can have alignment marks that can be used to align it to the substrate. However, such alignment marks can align the superstrate with the substrate to an accuracy of 10 microns and better than 1 micron.

[0104] Planarization requires the presence of micro- and nano-scale features on the substrate. In one embodiment, the planarization material is deposited using inkjet directly onto the patterned substrate, or onto any film deposited on the patterned substrate that would normally substantially match the features on the substrate. In another embodiment, the planarization material is deposited using inkjet onto the superstrate. Since the superstrate is substantially flat without patterned features, inkjetting the material onto the superstrate rather than the substrate overcomes problems resulting from the dispensing of droplets onto features with steep profiles that cause discontinuities in the joining of the droplets prior to forming the planarization film.

[0105] (In situ measurement) The implementation of PAINT for planarization can be combined with in-situ metrology to determine the performance of the planarization process in real-time, near real-time, or offline conditions. This in-situ metrology can be optically based (e.g., thin film reflectance measurement) or tip-based (e.g., AFM-based) metrology. The AFM system can be a single AFM or an AFM array that is individually addressable for positioning in hot spot areas. The optical techniques for metrology or for the detection of particle events (disclosed below) can rely on the use of an imaging system, where such an imaging system can be a line-scan or area camera that may be coupled with a telecentric lens that minimizes the signature from off-axis images. For high resolution, a line-scan camera may be the preferred embodiment, since these cameras can be installed in-line and used in conjunction with a linear array of UV LEDs to benefit from the motion of the substrate as the UV curing is performed in a separate station. Using a grayscale camera can also yield higher resolution, but if a hyperspectral signature is desired, an RGB camera may also be used for this purpose. The pixel size of such cameras may be less than 100 μm, less than 10 μm or as small as 1 μm.

[0106] This is particularly useful when the model behind the process is based on some empirical experimental evidence (often referred to as a grey-box model when combined with a physical understanding of the process, or a black-box model when independent of the physics of the process). The metrology can be performed at one or multiple locations on the substrate, depending on the requirements. Alternatively, the determination of the optimal measurement location can be done using intelligent sampling. Furthermore, in-situ metrology can also be used to monitor and "sense" the planarization process, which can be useful for continuous system improvement, especially for grey or black-box models. This can then facilitate predictive analysis and virtual metrology, i.e., the process of predicting impending defects or failures due to factors such as inkjet volume changing with systematic parasitics, and the process of determining where and what to measure, respectively.

[0107] Particle contamination can also be a problem during this process. The presence of particles on the substrate or superstrate can lead to voids and loss of yield if the superstrate cannot force droplets adjacent to the particles to join. The size of the voids is typically much larger than the size of the particles themselves, but can be as much as 1000 times larger than the particles when the superstrate is fused silica and the particles are sharp polymers (see, for example, Singhal S, Grigas MA, Sreenivasan SV. Mechanics-Based Approach for Detection and Measurement of Particle Contamination in Proximity Nanofabrication Processes. ASME. J. Micro Nano-Manuf. 2016;4(3):031004-031004-7. doi:10.1115 / 1.4033742, which is incorporated by reference in its entirety for all purposes). The size of these voids can depend on both the thickness of the superstrate and the elastic modulus of the superstrate. It is therefore important to control unwanted variations in the thickness of the superstrate to enable in situ detection and accurate characterization of such particle events.

[0108] In situ detection of particles during spreading can be captured by discontinuities or anomalies in the flow of the liquid front. These anomalies can be captured in real time via image processing techniques and used to make decisions regarding the size and origin of the particle contaminants. Optical metrology should be performed prior to separation to detect particle events and also to inspect the superstrate for damage caused by such events in areas proximate to the particle events and can be done via either a camera array or a single camera. The system can scan the substrate in a single direction (line scan) or multiple directions (area). The optical metrology system detects particle events and uses the exclusion zone and transition zone information to detect at least one of particle size, particle height, or force applied to the superstrate.

[0109] (Superstrate separation) The separation of the superstrate from the substrate is also a critical step in the PAINT process and should be done with minimal damage to the planarization film. For this purpose, a shape control system as disclosed above may be useful. Another way to enable separation in this situation is to have a shield with pins that can move in and out of the hollow center. This shield can be moved out when it is necessary to expose to UV light and move back so that it substantially contacts the backside of the superstrate to enhance support during separation. The pins can be augmented with the help of an adhesive such as Geckskin (see, for example, http: / / geckskin.umass.edu). Alternatively, the superstrate chuck can have foldable wings that can be attached to the edges of the superstrate during separation with the help of an adhesive such as Geckskin. A third strategy may be to use a sacrificial polymer film on the superstrate that can be partially or substantially evaporated with heat at the edge. This evaporation will create a gap at the edge and allow the separation of the superstrate from the substrate.

[0110] Separation of a superstrate from a substrate can be difficult when there are regions of bending in the shape or curvature of the superstrate or substrate, or if the superstrate can only be held on a narrow annular region that is substantially smaller than the diameter of the entire superstrate due to process constraints. During separation, adhesion between the chuck and the superstrate along this narrow annulus must overcome adhesion between the entire superstrate region and the planarizing polymer film.

[0111] This can be made possible by enhancing the adhesion between the chuck and the superstrate through mechanisms such as Geckskin mentioned above, or by reducing the interfacial strength of the superstrate-polymeric film laminate. The latter can be achieved by coating the top plate with an anti-stick material such as PTEE. In this situation, once the superstrate and the substrate are "bonded" through the planarizing film, defect-free generation of the delamination "crack" at the right interface becomes crucial. One strategy to generate the crack can be the use of acoustic agitation. This is because multiple interfaces exist across a wide range of dissimilar materials during the separation sub-process. By selecting the appropriate acoustic frequency, vibrations can be induced at the desired interface between the superstrate and the polymeric film, thereby generating the crack.

[0112] Cracks can be initiated by modifying the superstrate design to have more compliance at the edges, since the energy release rate is inversely proportional to the effective elastic modulus of the interface, and higher compliance, i.e., lower stiffness, means a higher energy release rate and easier delamination. Compliance can be obtained by reducing the thickness of the superstrate at the edges (since the energy release rate is also inversely proportional to the exponent of the superstrate thickness) or by adding a thin surface at the edge with a material that is substantially less stiff than the bulk superstrate material. The need to have a thinner or more compliant superstrate directly conflicts with the need to have a thicker superstrate to compensate for unequal spreading times, assuming the initial contact point is at the center and the final contact point is at the edge of the wafer. This conflict can be managed by prioritizing the edge thickness of the superstrate in the crack initiation region and then prioritizing the superstrate thickness in the remainder of the superstrate for optimal spreading.

[0113] Separation of the superstrate from the substrate can also occur at different locations. In one embodiment, the superstrate and substrate are brought into contact with one another at one location in the tool. The superstrate is mounted on a superstrate chuck at this location in the tool, except that the superstrate chuck is an annular ring with a cavity in the middle to allow for UV light curing. After curing is complete, the superstrate is released by reducing the vacuum such that the substrate and superstrate "sandwich" is now substantially supported by the substrate chuck.

[0114] A thin film of planarizing material bonds the two surfaces together. The substrate chuck is then moved to a second location inside the tool, where the superstrate is positioned under a chuck or mount that can provide substantially more support to the superstrate than the annular ring chuck in the first location. This is because the superstrate chuck in the second location does not need to have a substantially larger cavity in the center to allow for UV curing of the substrate. The superstrate chuck in this second location adheres to the superstrate on the backside with the aid of vacuum or mechanical gripping (e.g., the Geckskin fingers mentioned above). The superstrate is then peeled off the substrate, leaving the planarizing film behind on the substrate.

[0115] In another embodiment of the above concept, in one station, the substrate and superstrate are contacted forming a continuous planarizing film therebetween. This is followed by UV curing in another station, followed by separation in a third station or in the same initial station where the contact of the substrate and superstrate was made. It is understood that the planarizing film is substantially liquid when it is transferred from the initial station to the UV curing station. Since the average film thickness is small, the liquid film essentially acts like a liquid adhesive between the superstrate and the substrate without redistribution. This is further ensured by keeping the thickness of the superstrate optimally low. The liquid planarizing film can be partially cured at the edge of the initial station by a ring of UV light source or LEDs protruding substantially at the edge. This solidifies the film at the edge, thereby forming a seal around the sandwich. This process can be parallelized by having at least two substrate chucks, such that while the substrate is in the UV curing station, a second substrate is brought under the first station to begin carrying out the process.

[0116] (Nanoimprint Lithography) One of the key defect mechanisms in nanoimprint lithography is the entrapment of air bubbles or voids. These typically arise when two or more drops fail to substantially combine or substantially fill the nanoscale features of the template. These defects are often quite repeatable and can be classified as systematic errors rather than random errors. The filling of nanoscale features typically relies on the phenomenon of capillary pinning, where the nanoscale feature causes the moving meniscus to encounter an abrupt transition in the template. The meniscus can be forced to cross the abrupt transition and thus fill the feature, but only if there is significant capillary pressure that overcomes the resistance presented by that transition. If the features are sufficiently tall, the meniscus may not be able to completely wet them, thus leading to entrapment of voids.

[0117] One way to force the liquid to wet the features is to apply pressure to the template, for example with a pressure zone chuck, to further squeeze the liquid film. However, this can lead to problems with distortion of the template and therefore the overlay specifications. Another method is to locally heat the liquid so that a local temperature change occurs in the liquid. This temperature change creates a pressure difference that can force the liquid to navigate the nanoscale features. The local heating of the liquid can be focused on areas with such systematic defects, resulting in a 1mm 2 This can be done from the substrate side via individually addressable micro-heaters over areas less than 100 nm, however this heat can cause the substrate to expand and thus disrupt overlay specifications.

[0118] Another method of heating the substrate is by using a spatial light modulator such as a digital micromirror device (DMD). Current DMDs have pixel sizes that are sub-10 microns, so areas as small as the pixel size can be selectively heated. Light projection can be through the template or through the substrate. The preferred wavelength should be infrared, since resists strongly absorb UV and are crosslinked at these wavelengths. For example, monomers such as methyl methacrylate have absorption peaks in the IR region as well. In this region, both quartz and silicon can be transparent, thereby providing a way to effectively heat the liquid resist composition without crosslinking it. Alternatively, UV wavelengths can also be used if a sufficiently low power UV light can be provided so that even if the resist crosslinks, it does not crosslink completely and can still flow. However, such light must come from the template side, since silicon strongly absorbs UV light.

[0119] Similar to planarization, the flow of fluids in nanoimprint lithography can also be determined by extended thin film lubrication. One way to obtain information about hot spots during the process can be obtained with the help of a dummy template. Such a dummy template can be patterned with a less precise resolution than the actual template used in nanoimprint lithography, and such feature resolution can be less than 100 nm. The key here is that the length scale of the features on the dummy template should be sufficient to capture important physical properties such as bulk flow, as with the actual nanoscale template. This will substantially reduce the cost of capturing defect data for the nanoimprint lithography process. This will also be possible because the fluid will essentially not depend on the presence of features below a certain length scale, and will interact primarily with features above that length scale. Thus, a template with features above this critical length scale, in one embodiment, is sufficient to capture defects resulting from bulk fluid properties, given the combination of fluid properties and process time.

[0120] 6 is a flow chart illustrating a sequence of operations 600 for substrate planarization in accordance with one or more embodiments of the present technology. As shown in FIG. 6, a determining operation 610 can determine the topography and pattern information (e.g., global and nanoscale) of the patterned substrate. For example, in some embodiments, the topography of the patterned substrate can be obtained using an instrument such as a Zygo interferometer. The pattern information, which may include pattern density, anisotropy, area, and geometric information such as height / depth, width, diameter, etc., is typically available as input. If the pattern information is not available as input, the pattern information can be determined from a microscopy / inspection instrument (e.g., SEM, AFM, etc.).

[0121] Based on the topography and patterned substrate information, a drop pattern operation 620 can determine a drop pattern that can be dispensed during a dispense operation 630. In some embodiments, an inverse optimization approach can be used to determine the drop pattern. However, an inverse optimization scheme solves a model that represents the system using the drop pattern. Once a given drop pattern represents the system within an acceptable range of the desired results, the drop pattern is determined to be optimal. In this approach, system parameters (e.g., inkjet nozzle pitch, rheology of the fluid, existing topography of the substrate, etc.) can be provided as inputs as they affect the model representation of the system. For example, the presence of substrate topography in a region will require more material to be deposited in that region to fill the existing topography. This will affect the optimal drop pattern, allowing more fluid to be deposited in that region.

[0122] The closing operation 640 can close the gap between the patterned substrate and the superstrate to form a substantially continuous film of dispensed droplets. For example, in some embodiments, the substrate and superstrate can be mounted on a chuck. The chuck can be brought relatively close to each other via an actuator, such as a spring-loaded voice coil. The curing operation 650 can harden the substantially continuous film produced by the closing operation 640. For example, in some embodiments, the curing can be performed with the aid of UV light. This depends on the formulation composition of the liquid material. The curing step hardens the liquid material into a polymeric film and essentially freezes the evolution of the superstrate-fluid-substrate sandwich. Thus, the curing step should typically be performed after the desired sandwich evolution is achieved.

[0123] Once the curing operation 650 is complete, a separation operation 660 can separate the superstrate and the patterned substrate. Separating the superstrate from the cured film deposited on the patterned substrate can be accomplished in a variety of ways as described above. Strategies include, but are not limited to, the use of shields, sacrificial materials, Geckskin, etc. After a crack is initiated, the superstrate chuck and substrate chuck can be moved relatively far apart from each other using similar actuators (e.g., spring-loaded voice coils) as described above.

[0124] FIG. 7 is a flow chart illustrating a series of operations 700 for identifying and compensating for hot spots, according to some embodiments of the present technique. As shown in FIG. 7, a monitoring operation 710 can monitor the substrate for hot spots. According to various embodiments, hot spots can be monitored using in-situ metrology techniques such as optical cameras, microscopes, and AFMs, as described above. Hot spots can be identified based on their signature. For example, hot spots caused by particles have a typical signature of a circular area with no pattern, except that the circular area has a diameter that can approach 1000 times the particle size. Other hot spots can be identified as straight lines of low / no film thickness, which can be due to, for example, a misfiring nozzle on an inkjet. This can necessitate a complete replacement of the inkjet if the nozzle is substantially damaged. In general, hot spots represent a substantial deviation from ideal process performance outside a given tolerance window across a given area.

[0125] Decision operation 720 can determine whether a hot spot has been identified. If decision operation 720 fails to identify a hot spot, it branches to monitor operation 710 where the system can continue to monitor for hot spots. If decision operation 720 identifies a hot spot, it branches to decision operation 730 where an evaluation is made as to whether the identified hot spot is a systematic hot spot. A classic example of a systematic hot spot is the missing nozzle mentioned above. A systematic hot spot can be considered a repeatable error in the system caused by a non-random event. For example, particle encounters are random hot spots. Typically, machines are repeatable, and therefore, if the same signature is observed across multiple fields, it can be determined that there is something in the machine that is inaccurately tuned. This inaccurate parameter can be related to the tool design itself, or to the process parameters (such as low process times leading to the presence of bubbles and voids).

[0126] If the determine operation 730 determines that the hot spot is a systematic hot spot, the determine operation 730 branches to an identify operation 740 where a compensation solution is identified. As previously described, if a similar signature is observed across multiple fields, it can be determined that it is a systematic hot spot. An inject operation 750 can then implement the compensation solution. According to various embodiments, the systematic hot spot can be compensated for by properly adjusting the inaccurate parameters. For example, if a droplet is incorrectly placed or pinned on a substrate topography feature due to systematic inkjet misfiring that is difficult to fix due to the inkjet geometric geometry, the compensation solution can use thermocapillary or other mechanisms to locally move the droplet.

[0127] If determining operation 730 determines that the hot spot is not a systematic hot spot, determining operation 730 branches to identifying operation 760, which identifies potential failure mechanisms from non-systematic hot spots. Potential failure mechanisms from non-systematic hot spots, such as particles, include loss of deposited film or patterning due to incomplete fluid fill or even damage to the superstrate for IAP or template for NIL. Such hot spots can be identified, for example, using real-time in situ optical inspection techniques.

[0128] Compensation operation 770 determines a compensation solution to be implemented by implementation operation 780. Compensation solutions for random hot spots may require in-situ modification since these may not be repeatable and may cause yield loss. For example, inkjets have a stochastic drop placement accuracy, so drops may be placed within a circular area around an ideal spot. If this circular area encompasses an area of ​​pattern density variation, random hot spots may be expected as the drops interact with different features. Such hot spots may be compensated for by first identifying the hot spots and then implementing a solution such as thermocapillary, acoustic, etc. The compensation solution depends on the nature of the hot spots.

[0129] Systematic hot spots require a deeper understanding of the root cause and therefore may require systematic changes in tool design and process parameters, whereas random hot spots may require temporary intervention. In some cases, the same compensation solution can be proposed for both random and systematic hot spots. For example, thermocapillary action may be used to locally change the droplet distribution in the case of random inkjet droplet placement variations, and for droplet loss due to systematic misfiring nozzles.

[0130] (Overview of an Exemplary Computer System) Aspects and implementations of the present technology are described in the general context of various steps and operations. Various of these steps and operations may be performed by hardware components or embodied in computer-executable instructions that may be used to cause a general-purpose or special-purpose processor (e.g., in a computer, server, or other computing device) that is programmed with the instructions to perform the steps or operations. For example, the steps or operations may be performed by a combination of hardware, software, and / or firmware.

[0131] 8 is a block diagram illustrating an example machine representation of a computerized implementation of the flattening system. A controller 800 may communicate with entities including one or more users 825, client / terminal devices 820, user input devices 805, peripheral devices 810, optional co-processor devices (e.g., cryptographic processor devices) 815, and a network 830. Users may interface with the controller 800 via terminal devices 820 over the network 830.

[0132] A computer may use a central processing unit (CPU) or processor to process information. A processor may include a programmable general-purpose or special-purpose microprocessor, a programmable controller, an application specific integrated circuit (ASIC), a programmable logic device (PLD), embedded components, combinations of such devices, and the like. A processor executes program components in response to user and / or system generated requests. One or more of these components may be implemented in software, hardware, or both hardware and software. A processor communicates instructions (e.g., operational instructions and data instructions) to enable various operations.

[0133] The controller 800 may include a clock 865, a CPU 870, memories such as a read only memory (ROM) 885 and a random access memory (RAM) 880, and a co-processor 875, etc. These controller components are connected to a system bus 860, which is connected to an interface bus 835. Furthermore, a user input device 805, a peripheral device 810, a co-processor device 815, etc. may be connected to the system bus 860 via the interface bus 835. The interface bus 835 may be connected to a number of interface adapters, such as a processor interface 840, an input / output interface (I / O) 845, a network interface 850, a memory interface 855, etc.

[0134] The processor interface 840 can facilitate communication between the coprocessor device 815 and the coprocessor 875. In one implementation, the processor interface 840 can facilitate encryption and decryption of requests or data. The input / output interface (I / O) 845 facilitates communication between the user input device 805, the peripheral device 810, the coprocessor device 815, etc., and the components of the controller 800 using protocols such as protocols for handling audio, data, video interfaces, wireless transceivers, etc. (e.g., Bluetooth, IEEE 1394a-b, serial, Universal Serial Bus (USB), Digital Visual Interface (DVI), 802.11a / b / g / n / x, cellular, etc.). The network interface 850 can communicate with the network 830. Through the network 830, the controller 800 can be accessible to the remote terminal device 820. The network interface 850 can use a variety of wired and wireless connection protocols such as direct connection, Ethernet, wireless connections such as IEEE 802.11a-x, etc.

[0135] Examples of the network 830 include the Internet, a local area network (LAN), a metropolitan area network (MAN), a wide area network (WAN), a wireless network (e.g., using the wireless application protocol WAP), a secure custom connection, and the like. The network interface 850 may, in some aspects, include a firewall that can govern and / or manage authorization to access / proxy data within a computer network and track various levels of trust between different machines and / or applications. The firewall may be any number of modules having any combination of hardware and / or software components that can enforce a given set of access rights between a particular set of machines and applications, between machines, and / or between applications, for example, to regulate the flow of traffic and resource sharing between these varying entities. The firewall may further manage and / or utilize access control lists that detail permissions, including, for example, access and operation rights of objects by individuals, machines, and / or applications, and circumstances under which permission rights exist. Other network security functions that may be included or performed by the firewall function may be, for example, but are not limited to, intrusion prevention, intrusion detection, next generation firewall, personal firewall, and the like, without departing from the novel techniques of the present disclosure.

[0136] The memory interface 855 may communicate with multiple storage devices such as the storage device 890, a removable disk device, etc. The memory interface 855 can use a variety of connection protocols such as Serial Advanced Technology Attachment (SATA), IEEE 1394, Ethernet, Universal Serial Bus (USB), etc.

[0137] User input devices 805 and peripheral devices 810 may be connected to the I / O interface 845 and potentially other interfaces, buses, and / or components. User input devices 805 may include card readers, fingerprint readers, joysticks, keyboards, microphones, mice, remote controls, retina readers, touch screens, sensors, etc. Peripherals 810 may include antennas, audio devices (e.g., microphones, speakers, etc.), cameras, external processors, communication devices, radio frequency identifiers (RFIDs), scanners, printers, storage devices, transceivers, etc. Co-processor devices 815 may be connected to the controller 800 via an interface bus 835 and may include microcontrollers, processors, interfaces, or other devices.

[0138] Computer executable instructions and data may be stored in memory accessible by the processor (e.g., registers, cache memory, random access memory, flash, etc.). These stored instruction codes (e.g., programs) may interlock with the processor components, motherboard, and / or other system components to perform desired operations. The controller 800 may use various forms of memory, including on-chip CPU memory (e.g., registers), RAM 880, ROM 885, and storage 890. Storage 890 may use any number of tangible, non-transitory storage devices or systems, such as fixed or removable magnetic disk drives, optical drives, solid-state memory devices, and other processor-readable storage media. The computer executable instructions stored in the memory may include one or more program modules, such as routines, programs, objects, components, data structures, etc., that perform particular tasks or implement particular abstract data types. For example, the memory may include an operating system (OS) component 895, modules and other components, database tables. These modules / components may be stored and accessed from storage devices, including from external storage devices accessible via an interface bus.

[0139] The database component can store programs that are executed by a processor to process stored data. The database component may be implemented in the form of a relational, scalable, and secure database. Examples of such databases include DB2, MySQL, Oracle, Sybase, etc. Alternatively, the database may be implemented using a variety of standard data structures, such as arrays, hashes, lists, stacks, structured text files (e.g., XML), tables, etc. Such data structures may be stored in memory and / or structured files.

[0140] The controller 800 may be implemented in a distributed computing environment where tasks or modules are performed by remote processing devices linked through a communication network such as a local area network (LAN), a wide area network (WAN), or the Internet. In a distributed computing environment, program modules or subroutines may be located in both local and remote memory storage devices. Distributed computing may be used to load balance and / or aggregate resources for processing. Alternatively, aspects of the controller 800 may be electronically distributed over the Internet or other networks (including wireless networks). Those skilled in the art will recognize that parts of the system may reside on a server computer and corresponding parts on a client computer. Data structures and data transmission specific to aspects of the controller 800 are also encompassed within the scope of the present disclosure.

[0141] (Conclusion) Unless the context clearly indicates otherwise, throughout the specification and claims, the words "comprises," "comprising," and similar phrases are to be construed in an inclusive sense, rather than an exclusive or exhaustive sense; i.e., "including, but not limited to." As used herein, the words "connected," "coupled," or any variation thereof, mean any direct or indirect connection or coupling between two or more elements. The coupling or connection between the elements may be physical, logical, or a combination thereof. Furthermore, the words "herein," "on," "under," and similar phrases, when used herein, refer to the single application as a whole, and not to any particular portion of the application. Where the context permits, words in the above detailed description using the singular or plural may include the plural or singular, respectively. The word "or," in connection with a list of two or more items, includes all of the following interpretations of the word: any of the items in the list, all of the items in the list, and any combination of the items in the list.

[0142] The above detailed description of examples of the present technology is not intended to be exhaustive or to limit the present technology to the precise form disclosed above. Although specific examples for the present technology are described above for illustrative purposes, various equivalent modifications are possible within the scope of the present technology, as one of ordinary skill in the art will recognize. For example, while processes or blocks are presented in a given order, another implementation may perform a routine having steps or use a system having blocks in a different order. Some processes or blocks may be deleted, moved, added, subdivided, combined, and / or modified to provide alternatives or subcombinations. Each of these processes or blocks may be implemented in a variety of different ways. Also, while processes or blocks are sometimes shown as being performed in sequence, these processes or blocks may instead be performed or implemented in parallel, or may be performed at different times. Furthermore, any specific numbers referred to herein are merely examples. Another implementation may employ different values ​​or ranges.

[0143] The teachings of the technology provided herein may be applied to other systems, not necessarily those described above. Elements and operations of the various examples described above may be combined to provide further implementations of the technology. Some alternative implementations of the technology may include fewer elements as well as additional elements to the implementations described above.

[0144] These and other changes can be made to the technology in light of the detailed description above. The above description describes certain examples of the technology and describes the best modes contemplated, but no matter how detailed the above may appear in text, the technology can be practiced in many ways. The details of the system may vary considerably in its specific implementation while still being encompassed by the technology disclosed herein. As mentioned above, a specific term used when describing a certain characteristic or aspect of the technology should not be construed as suggesting that the term is redefined herein to be limited to any particular feature, characteristic, or aspect of the technology with which the term is associated. In general, the language used in the following claims should not be construed to limit the technology to the specific examples disclosed herein, unless the detailed description section above explicitly defines such language. Thus, the actual scope of the technology encompasses not only the disclosed examples, but also all equivalent ways of practicing or implementing the technology under the claims.

[0145] In order to reduce the number of claims, certain aspects of the technology are presented below in certain claim forms, but applicants contemplate various aspects of the technology in any number of claim forms. For example, only one aspect of the technology is recited as a computer-readable medium claim, but other aspects may be embodied as a computer-readable medium claim as well, or in other forms, such as embodied in a means-plus-function claim. Claims intended to be treated under 35 U.S.C. § 112(f) begin with the phrase "means for," but the use of the phrase "for" in any other context is not intended for treatment under 35 U.S.C. § 112(f). Accordingly, applicants reserve the right to pursue additional claims after the filing of this application to pursue such additional claim forms in this application or any continuing application.

Claims

1. 1. A method for inkjet deposition onto a patterned substrate, comprising: determining global and nanoscale topography and pattern information of the patterned substrate; determining a droplet pattern based on the global and nanoscale topography, the pattern information, and a desired thickness profile of the deposited film; dispensing the droplet pattern onto the patterned substrate; closing gaps between the patterned substrate and a superstrate resulting from dispensed droplets to form a substantially continuous film, wherein a thickness of the superstrate is designed to vary spatially to mitigate systematic hot spots; curing the substantially continuous film; and separating the superstrate and the patterned substrate from the substantially continuous film on the patterned substrate; The method according to claim 1, further comprising:

2. The method of claim 1 , wherein the droplet pattern is determined from a process model, an optimization scheme, and experimental data.

3. 10. The method of claim 1, further comprising loading a digital file encoded with data representing the global and nanoscale topography and the pattern information of the patterned substrate.

4. 2. The method of claim 1, wherein the droplet pattern is calculated using model-based optimization with inputs including the global and nanoscale topography and pattern information, superstrate geometry, superstrate material properties, inkjetted material properties, inkjet droplet resolution, inkjet nozzle pitch, and tolerance information.

5. 10. The method of claim 1, wherein curing occurs at the end of a desired process time, the desired process time being selected to be long enough so that air bubbles are substantially mitigated and short enough to prevent undesirable infestation.

6. 1. A method for mitigating hot spots resulting from a film deposition process, comprising: identifying systematic hot spots on the substrate; identifying a compensation solution based on the characteristics of the hot spot, the compensation solution including varying a thickness of a superstrate and selecting at least one of an evaporation control, a pressure zone chuck, a digital micromirror device, and a programmable heater array; implementing the compensation solution to mitigate the hotspot; The method according to claim 1, further comprising:

7. 1. A method for mitigating hot spots resulting from a nanoimprint lithography process, comprising: Identifying systematic hot spots prior to or during the nanoimprint lithography process; identifying a compensation solution based on the characteristics of the hot spot, the compensation solution including varying a thickness of a superstrate and selecting at least one of an evaporation control, a pressure zone chuck, a digital micromirror device, and a programmable heater array; implementing the compensation solution to mitigate the hotspot; The method according to claim 1, further comprising:

8. The system for film deposition comprises: A processor; a substrate shape modulating chuck under the control of said processor; a superstrate shape modulating chuck under the control of said processor; a memory having stored thereon a set of instructions which, when executed by the processor, causes the machine to identify a time-varying shape of the substrate and the superstrate to maximize a relative curvature between the substrate and the superstrate over an entire period between an initiation of contact of the substrate and the superstrate and a completion of substantially conformal contact, wherein a thickness of the superstrate is designed to vary spatially to mitigate systematic hot spots, and at least one of the curvatures of the superstrate and the substrate is concave at at least one time between an initiation of contact of the substrate and the superstrate and a completion of substantially conformal contact; A system comprising:

9. 10. The system of claim 8, further comprising a camera having one of the following characteristics: RGB pixels, grayscale pixels, a line scan configuration, an area scan configuration, a telecentric configuration, and a camera pixel size of at least 100 μm, at least 10 μm, or at least 1 μm.

10. 10. The method of claim 1, wherein the desired film thickness profile is selected to optimize functional behavior of the patterned substrate, including desired substrate optical properties.

11. The method of claim 1 , wherein the patterned substrate is etched back.

Images