Method and system for generating lithographic patterns

Abstract

A lithography pattern generation system (1) and method comprising an incident particle beam (21) having the most likely wavelength (λ) from a particle source (2), and a mask (4) having through-holes (3) arranged within the incident particle beam (21). The lithography pattern generation system (1) generates a single nanometer feature size pattern (51) in which the through-holes (3) are arranged aperiodically within the mask. A computer implementation method for generating a lithography mask (4) model is also disclosed.

Description

Technical Field

[0001] The present invention relates to microlithography and the manufacture of integrated circuits and other devices, such as quantum devices that cannot be mass-produced with existing technologies. More specifically, the present invention is associated with exposing a mask with a particle beam, generating a particle pattern based on the information encoded in the mask, and exposing a target, such as a resist on a silicon wafer, with the particle pattern.

Background Art

[0002] The manufacture of integrated circuits currently relies on pattern generation using mask-based photolithography. An enlarged version of the desired chip pattern or a part of the desired chip pattern is printed on a substrate (photomask). The pattern on the mask may be distorted considering aberration effects and the like. Light (photons) passes through or reflects off the photomask using optical components (refractive lenses or reflective lenses) to create a reduced image of the pattern on a silicon wafer. The silicon wafer is coated with a photosensitive material (resist), and the photosensitive material reacts with the photons to form a permanent imprint of the image in the resist. This is then used as a physical mask in subsequent manufacturing processes. The two main light sources used in the semiconductor industry today are deep ultraviolet (DUV) with a wavelength of 193 nm and extreme ultraviolet (EUV) with a wavelength of 13.5 nm.

[0003] The ongoing goal is to create patterns with smaller feature sizes and higher information density, commonly referred to as minimum pitch (the minimum distance between the centers of two features). For a standard optical system, the achievable minimum pitch is half the wavelength of the projected beam, if the process is carried out in air or vacuum (refractive index 1, Abbe reference). DUV performance is further enhanced by using wavefront engineering techniques such as phase-shift masks, off-axis illumination, or optical proximity correction, achieving minimum pitches of approximately 30% of the wavelength. The final pitch in DUV lithography is achieved using immersion lithography techniques, where a high refractive index fluid is introduced between the final lens and the wafer. Combining these techniques allows DUV to operate at an effective wavelength of approximately 135 nm, resulting in a minimum pitch of approximately 38 nm. Further pitch reductions reduce throughput and increase complexity, as they require multiple exposures to create a single chip pattern.

[0004] The most advanced technology is extreme ultraviolet (EUV) photolithography, which should be able to generate patterns with a minimum pitch or minimum feature size of approximately 6.75 nm according to the Abbe standard with a refractive index of 1 (vacuum), using photons (electromagnetic waves) with a wavelength of 13.5 nm.

[0005] However, due to the high energy of photons in EUV lithography, the pattern generation process in resists is mediated by photogenerated secondary electrons that can travel several nanometers before inducing a reaction. Current experiments and theories show that the blurred radius of secondary electrons for EUV is about 3 nm, which limits the achievable feature size to about 6 nm. This means that it is not possible to fabricate small quantum dots, as well as quantum devices based on individual atoms and molecules, with EUV. Moving to wavelengths shorter than 13.5 nm only exacerbates the secondary electron problem.

[0006] In "Microlithography by Using Neutral Metastable Atoms and Self-Assembled Monolayers" (Science 269(5228), 1255-1257(1995) https: / / arxiv.org / abs / https: / / www.science.org / doi / pdf / 10.1126 / science.7652572), lithography using metastable atoms is proposed as an alternative to photolithography. Pattern generation in thiol-based resists is demonstrated in a proximity lithography setup using a beam of metastable argon atoms transmitted through a lattice. Other experiments have used optical masks to generate patterns with metastable atoms, and in some cases, atoms are deposited directly onto the substrate. However, it has not been demonstrated that atomic beams can be used to generate any desired complex pattern with a small minimum pitch.

[0007] Several experiments have been conducted to focus atomic beams into small points using solid lenses, mirrors, or fields. However, this is only applicable to serial programming and is therefore unsuitable for mass production.

[0008] In Fujita, J., Morinaga, M., Kishimoto, T., Yasuda, M., Matsui, S., Shimizu, F.: Manipulation of an atomic beam by a computer-generated hologram. Nature 380(6576), 691-694(1996). https: / / doi.org / 10.1038 / 380691a0, a desired pattern having metastable neon atoms is generated on a screen by transmitting an atomic beam through a solid mask consisting of a distribution of nearly circular through-holes with a diameter of 30 nm etched into a silicon nitride film. All openings are the same size and are positioned on the film at locations defined by a regular square lattice structure. The inventors call this technique lattice-based binary holography. The setup operates in all system dimensions within the Fraunhofer regime (far-field) where Fresnel number F << 1. The mask is designed using the Fourier transform of the desired pattern, and as a result, the diffraction process that takes place during transmission through the mask yields the desired pattern appearing as multiple copies on each side of the zeroth-order peak. The pattern copy on one side of the zeroth-order peak is rotated 180° relative to the pattern copy on the other side of the zeroth-order peak (see Figure 4 in the original paper).

[0009] Nesse, T., Simonsen, I., Holst, B.: Nanometer-Resolution Mask Lithography with Matter Waves: Near-Field Binary Holography. Phys. Rev. Applied 11, 024009 (2019). https: / / doi.org / 10.1103 / PhysRevApplied.11.024009 discloses theoretical research on lattice-based binary holography by applying a mask, using a small mask with a square and hexagonal lattice structure, where the distance between the mask and the patterned surface is only 40 μm, and the pores are very small, less than 1 nm.

[0010] The prior art lattice-based binary holography disclosed in some of the publications mentioned above and shown in Figure 1 is not suitable for the rapid generation of patterns with feature sizes below the current level of technology. In particular, it is not suitable for creating small features with the large masks required for the large write fields (typically 29 mm × 33 mm) used in the semiconductor industry. [Prior art documents] [Non-patent literature]

[0011] [Non-Patent Document 1] Berggren, KK, Bard, A., Wilbur, JL, Gillaspy, JD, Helg, AG, McClelland, JJ, Rolston, SL, Phillips, WD, Prentiss, M., Whitesides, GM: Microlithography by Using Neutral Metastable Atoms and Self-Assembled Monolayers.Science 269(5228), 1255-1257(1995)https: / / arxiv.org / abs / https: / / www.science.org / doi / pdf / 10.1126 / science.7652572.https: / / doi.org / 10.1126 / science.7652572 [Non-Patent Document 2] Fujita, J., Morinaga, M., Kishimoto, T., Yasuda, M., Matsui, S., Shimizu, F.: Manipulation of an atomic beam by a computer-generated hologram. Nature 380(6576), 691-694(1996). https: / / doi.org / 10.1038 / 380691a0 [Non-Patent Document 3] Nesse, T., Simonsen, I., Holst, B.: Nanometer-Resolution Mask Lithography with Matter Waves: Near-Field Binary Holography.Phys.Rev.Applied 11,024009(2019).https: / / doi.org / 10.1103 / PhysRevApplied.11.024009 [Overview of the project]

[0012] To improve the performance of semiconductor devices and enable mass production in the realm of quantum devices, there is a need for high-speed generation of patterns with feature sizes below the limits of conventional technologies.

[0013] The object of the present invention is to disclose a system and method for rapidly generating patterns having feature sizes down to the nm range.

[0014] One of the main advantages of the present invention over the prior art is the ability to generate nm features at nm pitch across a large writing field, which is not possible with the prior art. Furthermore, nm features can be written using holes larger than the feature size, something not demonstrated in binary holography. This facilitates the fabrication of masks.

[0015] A further advantage of the present invention may be that the exposure time can be defined more precisely than in grid-based binary holography because the intensity of the pattern is not uniformly distributed throughout and between patterns in grid-based binary holography. Furthermore, in the present invention, only one copy of the desired pattern is produced during exposure, simplifying the generation process before and after pattern generation.

[0016] The invention that solves the above problems is a lithography pattern generation system and a lithography pattern generation method as described in the independent claim.

[0017] Herein, the above and further advantages of the systems and methods provided in accordance with the present invention will be described in more detail, by example, with reference to the accompanying drawings. The drawings are not necessarily to scale. Instead, certain features may be shown in an exaggerated size, or in a somewhat simplified or schematic manner, and certain prior art elements may be omitted for the purpose of illustrating these principles rather than cluttering the drawings with details that do not contribute to understanding the principles of the present invention. [Brief explanation of the drawing]

[0018] [Figure 1] This figure illustrates pattern generation based on conventional lattice-based binary holography. Multiple copies of a target pattern (510A, 510B) are generated by the diffraction of atomic waves (210) passing through a binary holographic mask (400) having holes (300) of the same size. The holes are positioned to correspond to the lattice structure. The lattice has a pixel-like structure, such as 512 × 512 pixels, where each pixel point may be completely or partially open (through-hole) or completely closed. The periodicity of the lattice structure leads to the generation of several copies of two rotated versions of the target pattern with varying intensity at the pattern target distance (L2) behind the mask. [Figure 2] This figure schematically shows a lithography pattern generation system (1) according to one embodiment of the present invention. The lithography pattern generation system (1) comprises an incident particle beam (21) having the most likely wavelength (λ) from a particle source (2), and a mask (4) positioned in the incident particle beam (21) between the particle source (2) and a pattern target (5). The pattern (51) generated on the pattern target is produced from particles diffracted by the mask (4). The source distance (L1) from the particle source (2) to the mask (4) and the pattern target distance (L2) from the mask (4) to the pattern target (5) are shown, respectively. Furthermore, the mask (4) comprises through-holes (3) having a hole width (w) and mask extension (a), i.e., the maximum distance between through-holes (3) exposed to the incident particle beam (21). [Figure 3] FIG. 1 schematically shows in one dimension the main components and parameters of a lithography pattern generation system (1) according to an embodiment of the present invention. The reference signs are the same as those used in FIG. 2. [Figure 4] FIG. 4 graphically shows the Fraunhofer diffraction patterns of light wave (A) and metastable helium matter wave (B) having the same wavelength through the same 5 nm thick SiN double slit mask having 8 nm openings separated by 8 nm walls. The vertical axis is the normalized intensity and the horizontal axis is the position in units of (μm). The matter wave has a dispersive force interaction with a dielectric mask that is not applicable to the light wave, resulting in a broad diffraction pattern having a much higher population of much higher diffraction orders. [Figure 5.1] FIG. 7 shows in one dimension the atomic flux [atoms / s / m2] in the vertical direction for each example of Tables 1 and 2 corresponding in sequence. [Figure 5.1] FIG. 10 shows in one dimension the atomic flux [atoms / s / m2] in the vertical direction for each example of Tables 1 and 2 corresponding in sequence. [Figure 5.3] FIG. 13 shows in one dimension the atomic flux [atoms / s / m2] in the vertical direction for each example of Tables 1 and 2 corresponding in sequence. [Figure 5.4] FIG. 16 shows in one dimension the atomic flux [atoms / s / m2] in the vertical direction for each example of Tables 1 and 2 corresponding in sequence. [Figure 5.5] FIG. 19 shows in one dimension the atomic flux [atoms / s / m2] in the vertical direction for each example of Tables 1 and 2 corresponding in sequence. [Figure 5.6]For each example of Table 1 and Table 2 that correspond in sequence, it is a diagram showing the atomic flux [atoms / s / m2] one-dimensionally in the vertical direction. The horizontal line is the position in [nm]. The lower graph shows the flux considering the dispersive force (Di), while the upper graph corresponds to the pattern seen when the mask is irradiated with a light beam of the same wavelength as the atomic beam without considering these forces. As will be further explained below for the atomic beam, it can be seen that by selecting appropriate system parameters to adjust the dispersive force, a significant reduction in the 0th-order background and a higher contrast of the features can be obtained. [Figure 6a] It is a diagram showing an example of a 2D generated pattern and a mask applied to the right side of each pattern in a cross-sectional view. [Figure 6b] It is a diagram showing an example of a 2D generated pattern and a mask applied to the right side of each pattern in a cross-sectional view. [Figure 6c] It is a diagram showing an example of a 2D generated pattern and a mask applied to the right side of each pattern in a cross-sectional view. The scales in the horizontal and vertical directions are both in nm units. The obtained feature sizes are 1.9 nm in Fig. 6a, 2.5 nm in Fig. 6b, and 1.1 nm in Fig. 6c. [Figure 7] It is a diagram showing an example of a mask made of the mask material indicated by the black square. The hatched area is the absorption area of the physical through-holes through which particles do not pass, as will be explained below. Only the remaining white area contributes to the diffraction pattern. This diagram corresponds to the situation where the center of the mask and the center of the source aperture are aligned.

Embodiments for Carrying Out the Invention

[0019] In the following description, various examples and embodiments of the present invention are described in order to provide a more complete understanding of the present invention to those skilled in the art. In the context of the various embodiments, the specific details described with reference to the accompanying drawings are not intended to be construed as limitations. Rather, the scope of the present invention is defined by the appended claims.

[0020] The terms used herein, although used in conjunction with the detailed description of specific embodiments of the invention, are intended to be interpreted in their broadest and most reasonable manner.

[0021] For clarity, the terms used in this document are defined below. -Binary Holograms and Binary Holography: Traditionally, a binary hologram refers to a hologram that has regions (apertures) that are completely transparent or completely opaque to an incident optical (scalar wave) beam. In the case of matter waves, the situation is slightly different, as dispersion forces between the matter wave and the dielectric mask may prevent some or all of the matter wave from penetrating the through-holes in the mask. Furthermore, the phase is perturbed across the apertures. For the purposes of this document, the terms binary hologram and binary holography also apply to such holograms (masks) when exposed by matter waves. Previous literature on binary holography (theoretical and experimental) did not consider dispersion force interactions. The experimental paper (NATURE cited earlier) had a target pattern with large features where the effects of dispersion forces were not clear. -Contrast(c): (Imax-Ibackground) / (Imax+Ibackground), where Imax and Ibackground represent the maximum and minimum intensities. -Features: The smallest piece of information within a pattern. In lithography, this is usually an identifiable object within a pattern, such as a line or dot. -Feature size: The size of the feature. Feature sizes can vary in different directions, but in the context of this document, the minimum feature size is the most important. - Fresnel number (F): where JPEG2026509824000002.jpg1724, - Full width at half maximum (d¹ / ²): The distance between points where the intensity reaches half of its maximum value. - Lattice period: (g). The minimum distance between two through-holes in lattice-based binary holography. The lattice period is greater than or equal to the minimum hole width. - Mask extension (a): The maximum distance between two through holes (3) in the plane of the mask (4). - Minimum hole width (w): The minimum width of the through-hole (3) in the mask (4). The through-hole (3) may have different shapes and sizes. - Aperture diameter (da): For supersonic sources, the diameter of the aperture of the particle source (2), defined as the proportion of the virtual source held by the collimated aperture. The virtual source is defined in the literature as the intensity distribution that can be obtained by backtracking atoms exiting the ejection surface to a minimum area. The ejection surface is the point where particles can be said to move in a free molecular flow without interacting with each other. See, for example, https: / / www.researchgate.net / publication / 356510630_Neutral_Helium_Microscopy_SHeM_A_Review. In the case of an outflow source, the aperture of the particle source is simply a hole between the particle reservoir and the vacuum, or a collimated aperture located downstream of that hole. -Pattern target (5): An object positioned in the diffraction particle pattern on the opposite side of the mask (4) from the particle source (2). For integrated circuit manufacturing, this is typically a silicon wafer coated with a resist. Suitable resist formulations for helium atoms are presented in the literature, e.g., the paper by Berggren et al. mentioned above. - Pitch: The distance between the centers of two adjacent features. - Resolution: The minimum pitch that can be transferred to the substrate. - Source distance (L1): The distance between the particle source (2) and the mask (4). More specifically, it can be said to be the distance from the aperture of the particle source (2) to the front surface of the holographic mask (4). -Pattern target distance (L2): The distance from mask (4) to pattern target (5). More specifically, it can be identified as the distance from the back of mask (4) to pattern target (5). -Throughput: The number of substrates or wafers that can be exposed per hour, and therefore a measure of the efficiency of the lithography process. -Wavelength (λ): The most likely wavelength of particles in the incident particle beam (21) from the particle source (2). The particle source may have a distorted wavelength distribution.

[0022] Conventional binary holography, a prior art technique used with atomic beams as initially mentioned and shown in Figure 1, is based on a holographic mask with uniformly sized holes arranged according to the underlying lattice structure. This gives some significance to the generated pattern.

[0023] Firstly, different diffraction orders have different intensities, determined by the size and shape of the through-holes as well as the lattice period, which means that the exposure time cannot be optimized for all parts of the resulting pattern.

[0024] Secondly, as shown in Figure 1, the different orientations of the target patterns (510A, 510B) complicate the manufacturing processes before, during, and after pattern generation.

[0025] Thirdly, the pattern target distance (L2) must be large enough so that the zero-order beam and the target pattern do not overlap with each other.

[0026] The minimum pattern target distance (L2) can be derived from the angular position θ of the primary peak in the pattern target plane relative to the center of the primary peak, which is determined by the underlying grid structure. The angle θ is measured with respect to the surface normal of the mask plane (see Figure 1). sinθ = λ / g

[0027] The center position of the target pattern closest to the zeroth-order peak within the pattern target plane (the center position of the first-order peak) is as follows (small-angle approximation is applied here): (λ / g)L2

[0028] For the first-order diffraction to not overlap with the zero-order diffraction, the center of the first-order peak must be at a distance 'a' from the center, regardless of the distance to the mask, so as not to overlap with the zero-order peak having mask extension (a). This is because the first-order peak consists of all atoms that travel in a straight line. Therefore, (λ / g)L2≧a→L2≧ga / λ≧wa / λ It can be seen that the larger the mask, the larger the minimum pattern target distance L2 must be.

[0029] In the first independent system embodiment ES1-1, the present invention relates to a lithography pattern generation system (1), - An incident particle beam (21) having a wavelength (λ) from a particle source (2), -A mask (4) having through holes (3) positioned within an incident particle beam (21), the lithography pattern generation system (1) is configured to generate a pattern (51) on a pattern target (5).

[0030] In the system embodiment ES1-2 of ES1-1, particles in the incident particle beam (21) reach the mask (4) as incident particle waves, and the mask (4) is configured to induce a phase shift of the particle wavefront as it propagates through the mask.

[0031] In the system embodiment ES1-3 of ES1-1 or ES1-2, the induced phase shift is at least π, 3 / 2π, or 5 / 3π over a distance of 0.016 times the minimum hole width (w) of the through hole (3) from the mask (4).

[0032] The minimum hole width (w) is the minimum hole width of the through-hole (3) in the mask (4).

[0033] In any of the system embodiments ES1-1 to ES1-3, specifically ES1-4, particles in the incident particle beam (21) reach the mask (4) as incident particle waves, and the mask (4) is configured to modulate the amplitude of the particle wavefront as it propagates through the mask.

[0034] In any of the system embodiments ES1-1 to ES1-4, specifically ES1-5, the lithography pattern generation system (1) is configured to generate a single copy of the pattern (51).

[0035] In the system embodiment ES1-6 of ES1-1 or ES1-5, the lithography pattern generation system (1) is configured to generate a pattern (51) at a target distance (L2) from a mask (4) on the opposite side of the incident particle beam (21).

[0036] In any of the system embodiments ES1-1 to ES1-6, specifically ES1-7, the through-hole (3) adjusts the dispersion force interaction between the incident particle beam (21) and the mask (4) to achieve an effective aperture size (w) smaller than the physical hole width (w) of the through-hole (3). e It is adjusted to have ).

[0037] In any of the system embodiments ES1-1 to ES1-7, specifically ES1-8, the parameter values ​​for the physical hole width (w), mask material, and mask thickness are selected to induce a phase shift of the incident particle wavefront by modifying the dispersion force interaction between the incident particle beam (21) and the mask (4).

[0038] In any of the system embodiments ES1-1 to ES1-8, specifically ES1-9, the mask (4) is doped or implanted with neutral particles, ions, and electrons to correct dispersion force interactions and / or Debye force interactions in order to correct phase shifts across the wavefront.

[0039] In any of the system embodiments ES1-10 from ES1-1 to ES1-9, the lithography pattern generation system (1) is configured to reduce the background signal (the proportion of straight-traveling signals) such that the proportion of diffracted beams is greater than that of light waves propagating through the same mask (4) at the same wavelength in the axial direction through the through-hole (3) in order to increase the contrast (c) of the pattern (51).

[0040] Generally, the overall transmission signal of a particle is reduced compared to the optical signal. However, the reduction in the background signal (the proportion of the signal that passes straight through) is stronger than the reduction in the diffraction signal.

[0041] In any of the system embodiments ES1-1 to ES1-10, in ES1-11, the through-hole (3) has a minimum hole width (w) and a mask extension (a) which is the maximum distance between two through-holes in the mask (3) exposed by the incident particle beam (21), and the following is satisfied: JPEG2026509824000003.jpg1427

[0042] In any of the system embodiments ES1-11 to ES1-11, specifically ES1-12, the particle source (2) has an aperture diameter (da) smaller than the source distance (L1) between the particle source (2) and the mask (4).

[0043] In any of the system embodiments ES1-13 from ES1-1 to ES1-12, the mask extension (a) is less than the source distance (L1) between the particle source (2) and the mask (4).

[0044] In any of the system embodiments ES1-1 to ES1-13, specifically ES1-14, the mask and the pattern target (5) are arranged in parallel.

[0045] In any of the system embodiments ES1-1 to ES1-13, specifically ES1-15, the mask and the pattern target (5) are arranged rather than in parallel.

[0046] In any of the system embodiments ES1-1 through ES1-15, specifically ES1-16, the mask is curved.

[0047] In any of the system embodiments ES1-1 to ES1-16, specifically ES1-17, either the mask (4) or the pattern target (5) is tilted with respect to the particle source (2).

[0048] In any of the system embodiments ES1-1 to ES1-17, specifically ES1-18, the center of the particle source (2) is shifted relative to the center of the mask (4).

[0049] A tilted mask can be configured to introduce asymmetry into the diffraction pattern, while a curved mask can be configured to equalize the population of diffraction orders and thus support illumination of the entire pattern target (5).

[0050] In any of the system embodiments ES1-1 to ES1-18, specifically ES1-19, the minimum hole width (w) is at least twice, three times, or four times the minimum feature size of the pattern (51).

[0051] In any of the system embodiments ES1-1 to ES1-19, specifically ES1-20, the mask (4) is a binary holography mask.

[0052] In system embodiment ES1-21, which is one of ES1-1 to ES1-20, the mask (4) is made of a dielectric material.

[0053] In any of the system embodiments ES1-1 to ES1-21, specifically ES1-22, the mask (4) is silicon nitride, Si3N x Or it contains Si3N4.

[0054] In any of the system embodiments ES1-1 to ES1-22, specifically ES1-23, the through-holes (3) are not arranged according to a predetermined periodic grid structure of the substrate on the mask (4).

[0055] In any of the system embodiments ES1-1 to ES1-23, specifically ES1-24, the through-hole (3) has a non-uniform minimum hole width (w).

[0056] In any of the system embodiments ES1-1 to ES1-24, specifically ES1-25, the through-hole (3) has a minimum hole width (w) of 5 nm, 4 nm, 3 nm, 2 nm, or 1 nm.

[0057] Pore ​​sizes smaller than 5 nm can be achieved with 2D materials.

[0058] In any of the system embodiments ES1-1 to ES1-25, specifically ES1-26, the through-hole (3) has a maximum hole width (w) of 100 nm.

[0059] In any of the system embodiments ES1-1 to ES1-26, specifically ES1-27, the through-holes (3) have different hole widths (w).

[0060] In any of the system embodiments ES1-1 to ES1-27, specifically in embodiment ES1-28, the through-hole (3) has a different shape.

[0061] In any of the system embodiments ES1-1 to ES1-28, specifically ES1-29, the center-to-center distance (c) between adjacent through holes (3) is non-uniform.

[0062] In any of the system embodiments ES1-1 to ES1-29, specifically ES1-30, the mask thickness is 500 nm, 100 nm, 50 nm, 20 nm, 10 nm, or less than 7 nm.

[0063] In any of the system embodiments ES1-1 to ES1-30, specifically ES1-31, the mask thickness is less than the minimum hole width (w).

[0064] In any system embodiment ES1-32 from ES1-1 to ES1-31, the particle beam (21) includes metastable atoms or molecules.

[0065] In any of the system embodiments ES1-1 to ES1-32, specifically ES1-33, the particle beam (21) contains metastable noble gas atoms.

[0066] In any of the system embodiments ES1-1 to ES1-33, ES1-34, the particle beam (21) contains metastable helium atoms.

[0067] In any system embodiment ES1-35 from ES1-1 to ES1-34, the wavelength (λ) is smaller than any of the minimum feature size (df), minimum pitch, minimum hole width (w), and mask thickness.

[0068] In any of the system embodiments ES1-1 through ES1-35, specifically ES1-36, the wavelength (λ) is less than 1.0 nm, 0.5 nm, or 0.2 nm.

[0069] In any of the system embodiments ES1-1 to ES1-36, specifically ES1-37, the lithography pattern generation system (1) is configured to generate a pattern on a pattern target (5) in a single exposure.

[0070] In any of the system embodiments ES1-1 to ES1-37, specifically ES1-38, the lithography pattern generation system (1) includes a particle source (2).

[0071] In any of the system embodiments ES1-1 to ES1-38, specifically ES1-39, the lithography pattern generation system (1) includes a pattern target (5).

[0072] In any of the system embodiments ES1-1 to ES1-39, ES1-40, the particle source (2) is supersonic or effluent.

[0073] In system embodiment ES1-41 of ES1-1 to ES1-40, the mask (4) is fabricated based on a mask model generated according to one of the following computer implementation methods EC1-1 to EC1-6.

[0074] In the first independent method embodiment EM1-1, the present invention relates to a method for generating a pattern (51) by lithography, - This includes exposing a mask (4) having through holes (3) to an incident particle beam (21) from a particle source (2).

[0075] In the EM1-2 embodiment of the EM1-1 method, particles in the incident particle beam (21) reach the mask (4) as particle waves, and the mask (4) is configured to induce a phase shift of the particle waves as they propagate through the mask.

[0076] In the EM1-3 embodiment of the EM1-1 or EM1-2 method, the induced phase shift is at least π, 3 / 2π, or 5 / 3π over a distance of 0.016 times the minimum hole width (w) of the through hole (3) from the mask (4).

[0077] In any of the method embodiments EM1-1 to EM1-3, EM1-4, this method is - This includes placing a mask (4) in front of the pattern target (5) and generating a pattern (51) on the pattern target (5).

[0078] In any of the method embodiments EM1-1 to EM1-4, in embodiment EM1-5, the pattern target (5) is positioned at a target distance (L2) from the mask, the through-holes (3) have a minimum hole width (w), and the mask (4) has mask extension (a) between the through-holes (3) exposed by the incident particle beam (21), and the following conditions are met: JPEG2026509824000004.jpg1428

[0079] In any of the method embodiments EM1-1 to EM1-5, in embodiment EM1-6, the through-hole (3) has an effective hole width (w) smaller than the minimum hole width (w) due to dispersion forces between the incident particle beam (21) and the mask (4). e ) has.

[0080] In any of the method embodiments EM1-1 to EM1-6, specifically EM1-7, the method is performed using a lithography pattern generation system (1) from any of the ES1-1 to ES1-40.

[0081] In any of the method embodiments EM1-1 to EM1-7, in embodiment EM1-8, the method includes generating a mask model according to any of EC1-1 to EC1-6.

[0082] In any of the method embodiments EM1-1 to EM1-8, EM1-9, the method includes manufacturing a mask based on a mask model generated in any of the computer implementation method embodiments EC1-1 to EC1-6.

[0083] Next, with reference to Figure 2, a specific independent embodiment of the present invention will be disclosed.

[0084] In this embodiment, the present invention is a lithography pattern generation system (1) comprising an incident metastable helium atomic beam (21) having the most likely wavelength (λ) from a particle source (2), and a mask (4) having through holes (3) positioned within the incident particle beam (21).

[0085] In contrast to the mask (400) in Figure 1 for a conventional lithography pattern generation system in which through-holes (300) are arranged according to a predetermined underlying grid structure independently of the target pattern, the through-holes (3) according to the present invention are not arranged on the mask according to a predetermined underlying grid structure (4). Furthermore, the holes (3) may have different shapes and different minimum hole widths (w). A single pattern (51) with a minimum feature size of 1 nm or less is generated on a pattern target (5) positioned at a target distance (L2) from the mask (4).

[0086] This invention utilizes the perturbation of particle waves due to dispersion force interactions between a dielectric mask and the particle itself as the particle passes through holes in the mask. Dispersion forces, particularly the Casimir-Polder force, are caused by quantum mechanical ground state fluctuations of the electromagnetic field in the absence of charge. Due to the electric field fluctuations, metastable helium atoms are polarized for a short time. The resulting induced dipole moment then interacts with the mask via dipole-dipole interactions. These forces are exerted over a distance r -3Although it decays according to a power law, it plays a significant role on the nanometer scale. The interaction is stronger near the edge of the hole and weaker towards the center. This reduces the number of atoms that travel in a straight line. In the Fraunhofer diffraction regime, this has the effect of a reduced 0th-order d peak and a broadened diffraction pattern, as shown in Figure 4.

[0087] In this embodiment, metastable helium atoms pass through through-holes in a silicon nitride film mask. Dispersion forces prevent a very large proportion of metastable helium atoms with a wavelength of 0.1 nm from penetrating holes less than 2 nm in diameter in a silicon nitride mask with a thickness of 5 nm.

[0088] Dispersive force interactions have a positive effect on the contrast of the generated patterns.

[0089] The difference in contrast can be illustrated by examining a one-dimensional set of key system parameters. Table 1 below shows examples of parameter values ​​for realizing a system considering dispersion forces and a scalar wave system (assuming metastable helium and a 5 nm thick silicon nitride mask). Table 2 shows the parameter values ​​obtained for the two systems. It is noteworthy from Table 1 that all examples are within the range L2 ≤ wa / λ, and that the first-order diffraction overlaps with the zero-order diffraction of the conventional technique. JPEG2026509824000005.jpg35130 (Table 1) Parameters used to compare the feature size of a system considering dispersion forces with the feature size of a system that does not consider these features. A standard 5nm thick SiN wafer and a dispersion force profile of metastable helium atoms are used. The pattern consists of three slits: one at each edge of the mask and one in the center. Furthermore, although the Fresnel number varies considerably, the condition L2 ≤ wa / λ is satisfied for all examples, as can be seen from the last column. JPEG2026509824000006.jpg30131 (Table 2) Results for the system parameters in Table 1 regarding the full width at half maximum: (d1 / 2)(-) and contrast: (c)(-) for the corresponding parameters (d1 / 2)(Di) and (c)(Di) for no dispersion power and with dispersion power.

[0090] Two columns are particularly worth examining regarding contrast. Considering dispersion forces, we can see a significant increase in contrast across all examples with comparable feature sizes. These results are also shown graphically in Figures 5.1 through F.6.

[0091] The present invention makes it possible to reduce the most likely particle (λ) in an incident particle beam compared to other limiting dimensions such as feature size (df), minimum distance between features (d1 / 2), mask stretch (a), and source and pattern target distances (L1, L2).

[0092] The most likely wavelength (λ) can be selected to be small compared to the desired feature size (df), so the pattern can be generated using a beam with supersonic or velocity diffusion of the outflow source. The slight blurring this introduces to the pattern is not a problem. The wavelength is given by the velocity, and vice versa, and is given by de Broglie's formula.

[0093] The contrast of the diffraction pattern can be altered by changing the effect of dispersion forces by one of the following means: changing the size of the holes, changing the shape of the holes, changing the thickness of the mask material (film), changing the material of the mask, doping the mask by electron or ion implantation or surface functionalization, or changing the metastable atoms or molecules.

[0094] Figures 6a, 6b, and 6c show cross-sections of three 2D patterns generated in one embodiment of the present invention using the settings listed in Examples 1, 2, and 3 in Table 3 below. On the right side of each figure is a diagram of the applied mask, which has a size of 10 μm × 10 μm with 10 nm circular through-holes. A standard 5 nm thick SiN wafer and a dispersion force profile of metastable helium atoms are used. The repeating pattern is generated by through-holes arranged in a 3 × 3 grid with a hole spacing of 6 μm. It is worth noting that the hole width (w) of the mask is 4 to 9 times larger than the feature size of the generated pattern. JPEG2026509824000007.jpg18130 (Table 3) Information regarding the settings and characteristics of the 2D patterns generated in the examples shown in Figures 6a, 6b, and 6c. The parameters are the same as those in Tables 1 and 2.

[0095] The desired 2D pattern is periodic, giving a periodic mask structure as seen in Figures 6a, 6b, and 6c.

[0096] The diffraction pattern is determined by Kirchhoff's diffraction formula, which uses the transmission function of neutral matter waves. JPEG2026509824000008.jpg33151

[0097] Wave amplitude and distance JPEG2026509824000009.jpg 831 source points Final point from JPEG2026509824000010.jpg627 JPEG2026509824000011.jpg628 has a wave propeller JPEG2026509824000012.jpg730. This equation must be read from right to left: A matter wave is in the source plane along distance. Points in JPEG2026509824000013.jpg54 within the mask plane It was propagated to JPEG2026509824000014.jpg68 JPEG2026509824000015.jpg68, dispersion force, here representing the interaction strength between the mask material and the particles. JPEG2026509824000016.jpg77, Mass of beam particles JPEG2026509824000017.jpg47, wavelength JPEG2026509824000018.jpg87, Mask thickness JPEG2026509824000019.jpg65, Planck constant The image JPEG2026509824000020.jpg65 collects the phase shift across the wavefront due to the Casimir-Polder interaction at the non-delay limit, which depends on the shape of the mask and is represented by an integral across the surface of the mask. The propagation is then traced along the propagation length to the pattern target surface. Continue until the point in JPEG2026509824000021.jpg55 JPEG2026509824000022.jpg68. Interference is from source s to pattern target. It is described by the superposition of all light paths up to JPEG2026509824000023.jpg54, determined by the square of the absolute value. Each particle interferes only with itself, and therefore with other source points Different initial conditions, such as different wavelengths (and therefore different particle velocities), result in incoherent superposition. Therefore, interference of all states occurs across the source region and most likely at different wavelengths. JPEG2026509824000025.jpg65 and wavelength distribution Wavelength distribution with JPEG2026509824000026.jpg69 It must be incoherently averaged so that it yields an integral over JPEG2026509824000027.jpg832. The above formula needs to be modified to account for inclined or curved masks and / or pattern targets.

[0098] Within the mask plane, there are two integrals, one of which is coherent superposition. This concerns JPEG2026509824000028.jpg68, one of which is a phase shift. This concerns JPEG2026509824000029.jpg610. The first integral ( The integral (over JPEG2026509824000030.jpg68) represents the optical path and is located over the region of the hole. The other integral ( The image (JPEG2026509824000031.jpg610) describes the interaction with the mask material and thus represents the bulk material. Figure 7 shows these two different regions of a cross-section of a mask with regularly spaced spherical pores. The mask material is indicated by the dark region, and the region for phase shift integration is shown. JPEG2026509824000032.jpg610. The physical pores are the white and gray regions, where the gray regions are absorption regions defined as areas through which atoms do not pass and therefore do not contribute to the optical path. Only the white regions are related to coherent superposition.

[0099] Kirchhoff's diffraction equation describes the most common interference case. Different approximations exist for the propeller, resulting in simplifications that are effective in different regimes. These are summarized in Table 4. It is found that the propeller simplifies the Fourier transform in the Fraunhofer and Fresnel regimes due to the elimination of denominator dependence used in all previous publications on material wave lithography. The wave propagation function (g) between the mask and target to achieve resolution... L2 ), and the Fraunhofer regime between the source and mask to achieve stability with respect to source extension and wavelength variation (g L1 ) take this into consideration. JPEG2026509824000033.jpg56153JPEG2026509824000034.jpg52153 (Table 4) Overview of wave carriers in different regimes (Fraunhofer regime, Fresnel regime, wave carrier).

[0100] The characteristics of the mask (4), such as hole width (w), mask stretching (a), hole arrangement, and mask material and thickness, determine the generated pattern. However, in real-world situations, the generated pattern, or the desired pattern, is determined by the circuit design of the manufactured item, such as an integrated circuit. Therefore, it is necessary to manufacture a mask based on the desired pattern. This is often called the inverse problem.

[0101] As mentioned above, existing binary holography theories for scalar waves use a grid of holes of equal size as the basis for the mask and assume system dimensions and wavelengths that satisfy the well-known optical conditions of the Fraunhofer approximation. In the Fraunhofer approximation, the inverse problem is reduced to the Fourier transform of the desired pattern, thereby obtaining the required mask. Sampling the mask according to the Nyquist-Shannon sampling theorem yields the bandwidth of its modes, and therefore the smallest possible pitch. In one dimension, this can be done analytically as described in the prior art. This may also work for matter waves if the holes in the mask are large enough that dispersion force interactions can be ignored.

[0102] However, the above is because the dispersion force interaction between the particle and the diffracting object induces a complex phase distribution of the matter wave, as explained above, making it impossible to adapt the matter wave and mask to the smaller holes required for a small pitch, which is not taken into account in conventional mask generation theories.

[0103] According to one embodiment of the present invention, an alternative method for manufacturing a mask structure based on machine learning is disclosed below.

[0104] To approximate a general solution to the inverse problem, a deep convolutional neural network or any model representing a differentiable nonlinear function is used. This neural network is trained on a large dataset of randomly generated examples that map masks to patterns by applying Kirchhoff's diffraction formula as described above. The mask can be represented as a binary sequence or a finite-length array, where 1 represents a through-hole in the mask and 0 represents a blocked space. The pattern may also be represented in one embodiment by a similar binary sequence or array, and the pattern target can be defined as a fine grid, where the distance between grid lines is smaller than the desired feature size.

[0105] Once an approximate solution, such as a binary mask, is obtained, it can be further refined by using a genetic algorithm or any other optimization algorithm based on maximizing the fitness or reward function until a solution with the desired accuracy is produced. Both algorithms without gradients and algorithms with gradients can be used.

[0106] The genetic algorithm is a probabilistic classical evolutionary algorithm that dynamically changes to optimize the fitness function F. The genetic algorithm is inspired by Darwin's theory of evolution, which describes how genes in a population evolve according to their capacity for reproduction and mutation. In the genetic algorithm, the population is not an actual population of organisms, but rather a set of solutions to a specific optimization problem (known as chromosomes). Furthermore, mutations are not necessarily the result of a natural stochastic process, but occur according to different rules that can be determined during implementation. For each chromosome in the population, a fitness value can be calculated using a predefined set of formulas of the researcher's choice. Finally, the reproductive process of life is simulated by combining different chromosomes (parents) into offspring of the next generation. The genetic algorithm is a very robust optimization algorithm, particularly well-suited to using sequences of categorical variables, such as bases present in DNA. The mask used here is entirely defined by a sequence or array of binary values.

[0107] The fit function is the mean absolute error of the generated diffraction pattern plus the reciprocal of a small numerical constant α. JPEG2026509824000035.jpg21122

[0108] JPEG2026509824000036.jpg98ri) is the absolute square of the discretized wave function on a grid of radial coordinates ri, which measures the distance to the center of the diffraction pattern. Ψ(ri;x;d) is the wave function dependent on the position x=(x1;x2;:::) and thickness d=(d1;d2;:::) of the mask aperture.

[0109] Deep neural networks can be implemented using available software libraries such as TensorFlow, which has a Keras module available from https: / / www.tensorflow.org / or PyTorch.

[0110] Genetic algorithms can be implemented using pygad:https: / / pygad.readthedocs.io / en / latest / #. Other optimization algorithms can be implemented using, for example, nevergrad:https: / / facebookresearch.github.io / nevergrad / .

[0111] In an independent embodiment EC1-1, the present invention relates to a computer implementation method for generating a lithography mask model, -Generating a training mask with through holes, wherein the properties of the training mask are output variables to the machine learning algorithm. -Generating training patterns based on the generated training mask, where the characteristics of the training patterns are the input parameters of the machine learning algorithm. - This includes supplying a desired pattern to a machine learning algorithm to generate a lithography mask structure based on a desired pattern.

[0112] The characteristics of a training mask can depend on the characteristics of the particle source, such as wavelength, aperture size, and velocity diffusion.

[0113] The lithography mask model generated by this method is a model of the actual mask (4) that will be generated later based on the model, while the training mask is a mask model used solely for training a machine learning algorithm. The desired pattern is, for example, the result of the design of the circuit that will be generated when the actual mask (4) is exposed to the incident particle beam (21), resulting in the generated pattern (51). The generated pattern (51) should be as identical as possible to the desired pattern.

[0114] In a dependent embodiment EC1-2 that can be combined with EC1-1, the position, width, and shape of the through-holes are characteristics of the training mask that are generated randomly or partially randomly. Therefore, this method allows for many degrees of freedom when creating the training mask. This is also reflected in the generated mask model and the final mask, as a grid or repeating structure is generally not required to create the desired pattern unless the pattern itself is periodic.

[0115] In a dependent embodiment EC1-3, which can be combined with EC1-1 or EC1-2, the machine learning algorithm includes a supervised machine learning algorithm.

[0116] In a dependent embodiment EC1-4, which can be combined with any of EC1-1 through EC1-3, the supervised machine learning algorithm includes a deep neural network.

[0117] In a dependent embodiment EC1-5, which can be combined with any of EC1-1 to EC1-4, a computer implementation method for generating a lithography mask model includes manufacturing a mask (4) included in any of ES1 to ES40 based on the generated lithography mask model.

[0118] After the mask structure is generated, a physical mask can be manufactured according to the parameters obtained as a result of the machine learning process described above. Mask manufacturing may be carried out according to the prior art, for example, by etching holes, which is outside the scope of this application. However, as described above, the present invention makes it possible to obtain a small feature size with a relatively large hole size compared to the prior art, facilitating mask manufacturing and resulting in faster and cheaper mask manufacturing.

[0119] In a dependent embodiment EC1-6, which can be combined with any of EM1-1 to EM1-5, a computer implementation method for generating a lithography mask model includes manufacturing a mask (4) included in any of ES1 to ES40 based on the generated lithography mask model.

[0120] In exemplary embodiments, various features and details are shown in combination. The fact that some features are described in relation to a particular example should not be construed to mean that those features must be included together in all embodiments of the invention as needed. Conversely, features described with reference to different embodiments should not be construed to be mutually exclusive. Embodiments that incorporate any subset of the features described herein, and that are not explicitly interdependent, are contemplated by the inventors and are part of the intended disclosure, as will be readily apparent to those skilled in the art. However, an explicit description of all such embodiments would not contribute to understanding the principles of the invention, and consequently, for the sake of simplicity or brevity, some substitutions of features have been omitted.

Claims

1. Lithography pattern generation system (1), An incident particle beam (21) having a wavelength (λ) from a particle source (2), A lithography pattern generation system (1) comprising: a mask (4) having through-holes (3) disposed within the incident particle beam (21); configured to generate a pattern (51) on a pattern target (5); particles in the incident particle beam (21) reaching the mask (4) in the incident particle wave; the mask (4) being configured to induce a phase shift across the wavefront of the particle wave as the particle wave propagates through the mask; the mask (4) having a mask extension (a) which is the maximum distance between two through-holes in the mask exposed by the incident particle beam (21); and configured to operate in a Fresnel diffraction regime where the Fresnel number F, defined as F = a² / (L²*λ), is 1 or greater, where L² is the target distance, λ is the wavelength of the incident particle beam, and a>=L².

2. The lithography pattern generation system (1) according to claim 1, wherein the induced phase shift is at least π, 3 / 2π, or 5 / 3π over a distance of 0.016 times the minimum hole width (w) of the through hole (3) from the mask (4).

3. A lithography pattern generation system (1) according to claim 1, configured to reduce a background signal stronger than the diffraction order compared to the diffraction of a light wave propagating at the same wavelength in the axial direction through the through-hole (3) through the same mask (4), thereby increasing the proportion of beams that are diffracted rather than passing straight through the contrast (c) of the pattern (51).

4. The lithography pattern generation system (1) according to claim 1, wherein the minimum hole width (w) is at least two, three, or four times the minimum feature size of the pattern (51).

5. The lithography pattern generation system (1) according to claim 1, wherein the mask (4) is a binary holography mask.

6. The lithography pattern generation system (1) according to claim 1, wherein the through holes (3) are not arranged at non-periodic positions on the mask (4) according to a predetermined periodic grid structure of the substrate.

7. The lithography pattern generation system (1) according to claim 1, wherein the through-hole (3) has a non-uniform minimum hole width (w).

8. The through-hole (3) has a minimum hole width (w) of 5 nm, 4 nm, 3 nm, 2 nm, or 1 nm, and / or The lithography pattern generation system (1) according to claim 1, wherein the through-hole (3) has a maximum hole width (w) of 100 nm.

9. The through-hole (3) has a different shape, and / or The lithography pattern generation system (1) according to claim 1, wherein the center-to-center distance (c) between adjacent through holes (3) is non-uniform.

10. The mask thickness is less than 500 nm, 100 nm, 50 nm, 20 nm, 10 nm, or 7 nm, and / or The lithography pattern generation system (1) according to claim 1, wherein the mask thickness is less than the minimum hole width (w).

11. The lithography pattern generation system (1) according to claim 1, wherein the wavelength (λ) is smaller than any of the minimum feature size (df), minimum pitch, minimum hole width (w), and mask thickness.

12. A lithography pattern generation system (1), An incident particle beam (21) having a wavelength (λ) from a particle source (2), A lithography pattern generation system (1) comprising: a mask (4) having through holes (3) disposed within the incident particle beam (21); the lithography pattern generation system (1) is configured to generate a pattern (51) on a pattern target (5); particles in the incident particle beam (21) reach the mask (4) in the incident particle wave; the mask (4) is configured to induce a phase shift across the wavefront of the particle wave as the particle wave propagates through the mask; the system is configured to operate in a Fresnel diffraction regime; and the system is configured to generate a single representation of a desired pattern (51) on the pattern target (5) that substantially does not contain a conjugate image pattern.

13. The lithography pattern generation system (1) according to claim 12, comprising any of the features described in claims 2 to 11.

14. A method for generating a pattern (51) by lithography, Exposure of a mask (4) having through-holes (3) to an incident particle beam (21) from a particle source (2), wherein particles in the incident particle beam (21) reach the mask (4) in particle wave form, and the mask (4) is configured to induce a phase shift of the particle wave as it propagates through the mask, generating a pattern (51) at a target distance (L2) from the mask, and the mask (4) has a mask extension (a) which is the maximum distance between two through-holes in the mask that are exposed by the incident particle beam (21), A method comprising operating in a Fresnel diffraction regime in which the Fresnel number F, defined as F = a² / (L²*λ), is 1 or greater, wherein L² is the target distance, λ is the wavelength of the incident particle beam, and a ≥ L².

15. A method for generating a pattern (51) by lithography, Exposure of a mask (4) having through holes (3) to an incident particle beam (21) from a particle source (2), wherein particles in the incident particle beam (21) reach the mask (4) in particle wave form, and the mask (4) is configured to induce a phase shift of the particle wave as it propagates through the mask. A method comprising operating in a Fresnel diffraction regime, wherein the method is configured to produce a single representation of a desired pattern (51) on the pattern target (5) that substantially does not contain a conjugate image pattern.

16. A computer implementation method for generating a lithography mask model, The process involves generating a training mask having through holes, wherein the characteristics of the training mask are output variables to a machine learning algorithm. The process involves generating a training pattern based on the generated training mask, wherein the characteristics of the training pattern are the input parameters of the machine learning algorithm. A method comprising supplying a desired pattern to a machine learning algorithm to generate the lithography mask model based on the desired pattern.

17. A computer implementation method for generating a lithography mask model according to claim 16, wherein the position, width, and shape of the through-holes are characteristics of the training mask that are generated randomly or partially randomly.

18. A computer implementation method for generating a lithography mask model according to claim 16, wherein the machine learning algorithm includes a supervised machine learning algorithm.

19. A computer implementation method for generating a lithography mask model according to claim 16, wherein the supervised machine learning algorithm includes a deep neural network.

20. A computer implementation method for generating a lithography mask model according to claim 16, comprising manufacturing a mask (4) according to any one of claims 1 to 12, 14, or 15 based on the generated lithography mask model.

21. A method for generating a pattern (51) by lithography according to any one of claims 14 to 15, comprising manufacturing a mask based on the mask model generated in claim 16.

Images