Planarisation of iii / v substrates, in particular indium phosphide substrates

Abstract

The present application relates to the planarisation of semiconductor wafers, in particular indium phosphide wafers. It was surprisingly found that the use of an α-hydroxy carboxylic acid during the polishing process leads to surfaces with particularly advantageous local and global flatness parameters, which is of great advantage for subsequent epitaxy processes and photolithography processes for the production of electronic components.

Description

[0001] HF 525-P38868PC00

[0002] 1

[0003] Planarization of III / V substrates, especially indium phosphide substrates

[0004] The present invention relates to a method for producing improved surfaces of III-V substrates, as well as III-V substrates themselves, wherein indium phosphide (InP) substrates are of particular importance.

[0005] BACKGROUND OF THE INVENTION

[0006] Indium phosphide (InP) plays a vital role as a semiconductor material today, being an important component in optoelectronic and microelectronic devices. Many optical components, such as lasers for data transmission, are manufactured using indium phosphide. Other applications include solid-state light emission, microwave communication, fiber optic communication, microwave and millimeter-wave devices, and radiation-resistant solar cells.

[0007] Indium phosphide (InP) is an III-V compound semiconductor material that combines indium (In) as a group III element and phosphorus (P) as a group V element.

[0008] Compared to germanium and silicon, InP has a direct bandgap structure, high efficiency in converting electrical to optical energy, and high electron mobility. It can also be easily processed into a semi-insulating material, making it suitable for high-frequency millimeter and terahertz wave devices and circuits.

[0009] Semiconducting InP materials also have many applications; for example, such materials are suitable for optoelectronic components, e.g., edge emitter lasers for data transmission or infrared photodiodes for sensing the environment.

[0010] HF 487 / CG / / March 20, 2025 To manufacture InP devices, for example, an epitaxial layer is produced on a monocrystalline InP wafer as a substrate – with a specific composition and layer stacking sequence, depending on the desired functionality of the device. Electronic devices are then fabricated from the grown material, i.e., the epitaxial layer. A large number of electronic devices can also be produced on the wafer at once.

[0011] To manufacture electronic components, it may be necessary to etch certain structures into the epitaxial layer. The masks for etching are made from resists (so-called photoresists) that are applied to the entire surface of the epitaxial layer (for example, by rotary coating) before being dried under heat so that the solvent and water contained within evaporate.

[0012] In areas where the epitaxial layer is not to be etched, the resist layer is cured by exposure to light. This allows the image of a photomask to be transferred onto a light-sensitive photoresist. Exposure is carried out in stepper units. An exposure area defined by the stepper unit, e.g.,

[0013] A 20 mm x 20 mm area is exposed simultaneously. Within this exposure area, the substrate must be flat so that all areas of this surface lie at the focal point of the exposure optics (focal plane). Any tilting of the entire wafer surface relative to the wafer back surface held by the wafer holder can be compensated for by the exposure unit. From one exposed area to the next, the surface of the wafer or the epitaxial layer may be slightly raised or lowered. This difference is corrected by lowering or raising the wafer. No correction for the tilt of the wafer surface is made between measurement fields.

[0014] The unexposed areas of the photoresist are then dissolved, for example, by a developer solution (alternatively, the exposed areas can also be dissolved). This can be done by spraying, dipping, or dripping. Thus, the chemical properties of the photoresist in the exposed areas are altered by a photochemical reaction—depending on the type of photoresist, it becomes either soluble or insoluble in the developer solution. Afterward, drying and even tempering can be performed to improve the cross-linking of the photoresist remaining on the epitaxial layer. This produces a so-called lithographic mask, which enables further processing by chemical and physical processes, such as introducing material into the open windows, etching depressions beneath the open windows, or introducing foreign atoms (e.g.,...Doping and ion implantation) or the selective deposition of materials. The photoresist layer thus acts as a masking layer to protect certain areas before the subsequent process step.

[0015] After further processing is completed, the photoresist layer is completely removed, for example with the help of oxygen plasma or a wet chemical solution.

[0016] This process is used, for example, to manufacture semiconductors for microchips, processors, etc.

[0017] The basis for InP wafers are InP crystal ingots. These crystals can be produced using various methods. One example is the vertical gradient freeze (VGF) process. In this process, the crystal seed is not moved; instead, the crystal grows upwards in a crucible from the top of the seed. The crucible itself remains stationary, but the temperature field above the seed exhibits a positive temperature gradient and is shifted upwards by controlling the heater.

[0018] Another example is the Vertical Bridgman (VB) method. Here, the crucible containing the seed is moved vertically relative to the heater assembly. Combinations of the VGF and VB methods are also possible.

[0019] Another example is the Liquid Encapsulated Czech ralski-V (LEC) process. For this, a seed crystal is immersed from above into a melt and, while rotating around its own axis, is pulled out of the melt so that a single crystal grows downwards from this seed.

[0020] The single crystal produced during crystal growth is removed from the crucible. By removing the conical and end sections, a cylinder is formed, sometimes with an irregular outer surface. If necessary, the outer surfaces are ground to ensure a constant cylinder diameter along its entire length. Finally, the crystal is separated into individual wafers using wire lapping, i.e., sawn apart.

[0021] After wire lapping, the wafers exhibit irregularities such as...

[0022] Thickness variations, some with a large range, are present. Furthermore, the material is damaged on the surface and to a certain depth by the lapping process, which leads to microcracks.

[0023] The subsequent polishing process is therefore intended to compensate for large-scale irregularities, bring the wafer to the target thickness, compensate for local irregularities and produce a mirror-like surface so that an undisturbed and defect-free epitaxial layer can grow on the wafer.

[0024] The polishing process typically consists of two successive sub-processes: planarization and finalization.

[0025] In the sub-process of planarization, a relatively large amount of material is removed relatively quickly in order to level out the unevenness over a large area and to reduce local unevenness.

[0026] In the subsequent sub-process of finalization, relatively little material is removed relatively gently, so that any remaining damage at a certain depth (subsurface damage) is removed and the wafer is mirror-polished without generating new damage at a certain depth.

[0027] The present invention focuses on the subprocess of planarization, in which, as mentioned, global and local irregularities on a wafer are to be eliminated in order to produce as few errors as possible in a subsequent epitaxy process using photolithography.

[0028] The local flatness of a wafer with a grown epitaxial layer must therefore be better than a defined threshold. It follows that the local flatness of the wafer itself must also be better than a defined threshold. Several prior art documents deal with methods and means for wafer planarization.

[0029] Document CN103100965B discloses a method for the two-sided polishing of individual wafers. The described process comprises three stages (coarse polishing, medium polishing, and fine polishing), each using different chemicals and polishing cloths. This procedure is very complex and costly. The coarse polishing compound consists of a suspension of basic silicon dioxide in deionized water and sodium dichlorocyanurate in a volume ratio of 1:10:(1.0–1.8) with a pH of 10–11. The medium polishing compound consists of a suspension of basic silicon dioxide in deionized water and sodium hypochlorite in a volume ratio of 1:15:(0.1–0.5), with a pH between 8 and 9. The fine polishing agent contains a suspension of basic silicon dioxide in deionized water in a volume ratio of 1:(40-60) with a pH value of 7-8. The removal rates are relatively low.

[0030] Document EP3258481A1 describes a two-sided pre-polishing process. A mixture containing SiÜ2 and INSEC IPP is used as the polishing slurry. A solution called NCW 1001 is used for surface activation. The SUBA IV polishing cloth is treated with pellets. This process is specifically designed to address surface roughness on InP wafers.

[0031] WO2018198718A1 describes a process for manufacturing wafers with a specific roll-off angle at the edge. The process involves lapping the wafer as a first step.

[0032] CN106346318A describes a thinning and polishing process for indium phosphide wafers and a chemical corrosion apparatus. The thinning and polishing process comprises the following steps: physical grinding: placing an InP wafer on a grinding wheel, physically grinding and thinning the InP wafer from 300-500 pm to 180-240 pm; chemical corrosion: placing the physically ground InP wafer in the chemical corrosion apparatus and chemical corrosion in concentrated hydrochloric acid; chemical-mechanical polishing: placing the chemically corroded InP wafer on a nylon fiber cloth and performing chemical-mechanical polishing.

[0033] JPS58145604A describes a polishing solution containing bromine in methanol. A two-layer polyurethane-polyester polishing cloth is used.

[0034] CN102172885A describes global thickness variations of wafers in the range between 0.63 and 1.5 pm.

[0035] With known polishing solutions and processes, the problem can arise that the wafer surface is not flat across the entire exposure field of the lithography. In this case, part of the area can fall outside the focus of the exposure optics, and the photoresist is not sufficiently polymerized in these areas and is subsequently partially removed.

[0036] Compared to the prior art, it can be considered an object of the present invention to provide a wafer which exhibits the best possible flatness parameters (with few variations) and a low density of defects (scratches, pits, etc.) over a given area, in order to create the conditions for virtually defect-free lithography. Furthermore, it can be considered an object of the present invention to provide a method for planarizing InP wafer surfaces by which, at a sufficiently high ablation rate, the smallest possible flatness deviations on the wafer surface are achieved.

[0037] SUMMARY OF THE INVENTION

[0038] This problem is solved by a method according to claim 1, a III-V wafer according to claim 10 or 14, a group of wafers according to claim 17, and a use according to claim 19. Further advantageous embodiments of the present invention are the subject of the corresponding dependent claims.

[0039] Interferometer measurements were performed to characterize the surface properties of the corresponding wafers; details of the measurements are explained in the description and examples.

[0040] Without limiting the invention, the following points are set out to describe the main aspects, preferred embodiments and special features of the present invention:

[0041] 1. Method for polishing III / -semiconductor wafers (W), in which the semiconductor wafer (W) is brought into contact with a polishing cloth (P) with at least one side to be polished, wherein the semiconductor wafer (W) and the polishing cloth (P) perform a rotational movement relative to each other,

[0042] wherein a polishing agent (M) is added between wafer (W) and polishing cloth (P), comprising: deionized water, particles, and at least one α-hydroxycarboxylic acid.

[0043] 2. Method according to point 1, wherein the α-hydroxycarboxylic acid is selected from the group consisting of citric acid, tartaric acid, glycolic acid, lactic acid, malic acid and mandelic acid.

[0044] 3. Method according to one of points 1 or 2, wherein the polishing compound contains between 0.1 vol% and 1.0 vol% α-hydroxycarboxylic acid.

[0045] Active complexation takes place at a concentration of 0.1 vol% α-hydroxycarboxylic acid in the polishing solution.

[0046] From a concentration of 1.0 vol% α-hydroxycarboxylic acid in the polishing solution, there is an inhomogeneous reactivity in the polishing solution.

[0047] 4. The method according to point 3, wherein the polishing compound contains at least 0.4 vol% α-hydroxycarboxylic acid. 5. The method according to any of the preceding points, wherein the polishing compound contains between 1 vol% and 10 vol% particles.

[0048] 6. Method according to one of the preceding points, wherein the particles preferably contain SiC>2, TiC , Al2O3, MnÜ2 and / or CeÜ2.

[0049] 7. Method according to point 6 or 7, wherein the polishing compound contains between 6 vol.% and 8 vol.% particles.

[0050] 8. Method according to one of the preceding points, wherein the main fraction of particles is in a range of 10 nm to 30 nm particle diameter.

[0051] 9. Method according to one of the preceding points, wherein the specific surface area of ​​the particles is in a range between 100 m 2 / g up to 150 m 2 / g lies.

[0052] 10. Method according to one of the preceding points, wherein the polishing agent still contains active chlorine.

[0053] Active chlorine comprises various chlorine species that have an oxidizing effect.

[0054] 11. Method according to point 10, wherein the polishing agent contains between 1 g / l and 10 g / l of active chlorine.

[0055] 12. Method according to one of the preceding points, wherein the polishing agent has a pH value between 2 and 4.

[0056] 13. Method according to one of the preceding points, wherein the polishing cloth consists of non-woven polyurethane.

[0057] 14. Method according to any of the preceding points, wherein the polishing cloth has a thickness between 1 mm and 2 mm. 15. Method according to point 14, wherein the polishing cloth has a thickness between 1.2 mm and 1.4 mm.

[0058] 16. Method according to any of the preceding points, wherein several semiconductor wafers with at least one side to be polished are brought into contact with a polishing cloth, wherein several semiconductor wafers are inserted into a carrier, wherein the carrier moves on a circular path and rotates simultaneously.

[0059] 17. Method according to point 16, wherein the polishing cloth itself also rotates around its central axis.

[0060] 18. Method according to point 16 or 17, wherein several semiconductor wafers with two sides to be polished are brought into contact with a polishing cloth each.

[0061] 19. Procedure according to point 18, wherein both polishing cloths themselves also rotate about their central axis.

[0062] 20. Method according to one of the preceding points, wherein the polishing cloth(s) is / are structured in at least one part.

[0063] 21. Method according to point 20, wherein a wafer protrudes between 10 and 15 mm into the structured area of ​​the polishing cloth during its movement.

[0064] 22. Method according to one of points 20 or 21, wherein structuring comprises thermal densification of the relevant areas of the polishing cloth(s).

[0065] 23. A method according to any one of points 20 to 22, wherein the structuring comprises the mechanical removal of material from corresponding areas of the polishing cloth(s). 24. A method according to any one of points 20 to 23, wherein in structured areas the thickness of the polishing cloth(s) is 500 pm - 800 pm less than in non-structured areas.

[0066] 25. Method according to one of points 20 to 24, wherein the structuring includes hydrophilizing the relevant areas of the polishing cloth(s).

[0067] 26. Method according to one of points 20 to 25, wherein 1% to 18% of the surface of the polishing cloth(s), preferably approximately 7.1%, is structured.

[0068] 27. Method according to one of points 20 to 25, wherein 2% to 12%, preferably 7.1%, of the surface of the polishing cloth(s) is structured.

[0069] 28. Method according to any of the preceding points 16 to 27, wherein the support(s) have a lesser thickness than the wafer(s) held by the support.

[0070] 29. Method according to any of the preceding points 16 to 28, wherein the support(s) has a maximum thickness of 2 / 3 of the wafer(s) held by the support.

[0071] 30. Ill / V wafer having two opposing surfaces, wherein in a measurement field on at least one surface of the wafer the distance between the highest and lowest point of the surface is a maximum of 1.067 pm.

[0072] During the measurement, the other surface is drawn in from below by vacuum.

[0073] This parameter is the LTV value (local thickness variation).

[0074] 31. III / V wafer having two opposing surfaces, wherein in a measuring field on at least one surface of the wafer the distance between the highest and lowest point of the surface is between 0.745 pm and 1.067 pm, preferably between 0.304 pm and 1.067 pm.

[0075] 32. III / V wafer according to point 30, wherein in a measurement field on at least one surface of the wafer the distance between the highest and lowest point of the surface is a maximum of 0.754 pm.

[0076] 33. III / V wafer according to point 32, wherein in a measurement field on at least one surface of the wafer the distance between the highest and lowest point of the surface is between 0.304 pm and 0.754 pm.

[0077] 34. III / V wafer according to point 30, wherein in a measurement field on at least one surface of the wafer the distance between the highest and lowest point of the surface is a maximum of 0.495 pm.

[0078] This wafer preferably has a diameter of between 70 and 80 mm.

[0079] 35. III / V wafer according to point 34, wherein in a measurement field on at least one surface of the wafer the distance between the highest and lowest point of the surface is between 0.304 pm and 0.495 pm.

[0080] This wafer preferably has a diameter of between 70 and 80 mm.

[0081] 36. Ill / V wafer having two opposing surfaces, wherein in a measurement field on at least one surface of the wafer the maximum distance between the highest point above the shifted focal plane and the lowest point below the shifted focal plane is a maximum of 0.66 pm.

[0082] This parameter is the LTIR value (local total indicated reading).

[0083] 37. III / V wafer having two opposing surfaces, wherein in a measurement field on at least one surface of the wafer the maximum distance between the highest point above the displaced focal plane and the lowest point below the displaced focal plane is between 0.478 pm and 0.66 pm, preferably between 0.236 pm and 0.66 pm.

[0084] 38. III / V wafer according to point 36, wherein in a measurement field on at least one surface of the wafer the maximum distance between the highest point above the shifted focal plane and the lowest point below the shifted focal plane is a maximum of 0.48 pm.

[0085] 39. III / V wafer according to point 38, wherein in a measurement field on at least one surface of the wafer the maximum distance between the highest point above the shifted focal plane and the lowest point below the shifted focal plane is between 0.236 pm and 0.48 pm.

[0086] 40. III / V wafer according to point 36, wherein in a measurement field on at least one surface of the wafer the maximum distance between the highest point above the shifted focal plane and the lowest point below the shifted focal plane is a maximum of 0.407 pm.

[0087] This wafer preferably has a diameter of between 70 and 80 mm.

[0088] 41. III / V wafer according to point 38, wherein in a measurement field on at least one surface of the wafer the maximum distance between the highest point above the shifted focal plane and the lowest point below the shifted focal plane is between 0.251 pm and 0.48 pm.

[0089] This wafer preferably has a diameter of between 70 and 80 mm.

[0090] 42. A 13 / V wafer having two opposing surfaces, wherein in a measurement field on at least one surface of the wafer the maximum distance between the highest or lowest point above or below the shifted focal plane and the shifted focal plane is a maximum of 0.544 pm. This parameter is the LFPD value (local focal plane deviation).

[0091] 43. III / V wafer having two opposing surfaces, wherein in a measurement field on at least one surface of the wafer the maximum distance between the highest or lowest point above or below the displaced focal plane and the displaced focal plane is between 0.38 pm and 0.544 pm, preferably between 0.158 pm and 0.544 pm.

[0092] 44. III / V wafer according to point 42 or 43, wherein the size of the measuring field is between 5 mm x 5 mm and 15 mm x 15 mm.

[0093] 45. III / V wafer according to point 42, wherein in a measurement field on at least one surface of the wafer the maximum distance between the highest or lowest point above or below the shifted focal plane and the shifted focal plane is a maximum of 0.313 pm.

[0094] 46. ​​III / V wafer according to point 45, wherein in a measurement field on at least one surface of the wafer the maximum distance between the highest or lowest point above or below the shifted focal plane and the shifted focal plane is between 0.158 pm and 0.313 pm.

[0095] 47. III / V wafer according to point 45 or 46, wherein the size of the measuring field is between 5 mm x 5 mm and 15 mm x 15 mm.

[0096] 48. III / V wafer according to point 42, wherein in a measurement field on at least one surface of the wafer the maximum distance between the highest or lowest point above or below the shifted focal plane and the shifted focal plane is a maximum of 0.307 pm.

[0097] This wafer preferably has a diameter of between 70 and 80 mm.

[0098] 49. III / V wafer according to point 46, wherein in a measurement field on at least one surface of the wafer the maximum distance between the highest or lowest point above or below the shifted focal plane and the shifted focal plane is between 0.158 pm and 0.307 pm.

[0099] This wafer preferably has a diameter of between 70 and 80 mm.

[0100] 50. III / V wafer according to point 48 or 49, wherein the size of the measurement field is between 5 mm x 5 mm and 15 mm x 15 mm.

[0101] 51. Ill / V wafer having two opposing surfaces, wherein in a measurement field of 20 mm x 20 mm on at least one surface of the wafer the maximum distance between the highest or lowest point above or below the shifted focal plane and the shifted focal plane is a maximum of 0.367 pm.

[0102] This parameter is the LFPD value (local focal plane deviation) with a measurement field size of 20 mm x 20 mm.

[0103] 52. III / V wafer having two opposing surfaces, wherein in a measuring field of 20 mm x 20 mm on at least one surface of the wafer the maximum distance between the highest or lowest point above or below the displaced focal plane and the displaced focal plane is between 0.225 pm and 0.367 pm, preferably between 0.116 pm and 0.367 pm.

[0104] 53. III / V wafer according to point 51, wherein in a measurement field of 20 mm x 20 mm on at least one surface of the wafer the maximum distance between the highest or lowest point above or below the shifted focal plane and the shifted focal plane is a maximum of 0.305 pm.

[0105] 54. III / V wafer according to point 53, wherein, in a measurement field of 20 mm x 20 mm on at least one surface of the wafer, the maximum distance between the highest or lowest point above or below the displaced focal plane and the displaced focal plane is between 0.116 pm and 0.305 pm. 55. I111 / V wafer according to point 51, wherein, in a measurement field of 20 mm x 20 mm on at least one surface of the wafer, the maximum distance between the highest or lowest point above or below the displaced focal plane and the displaced focal plane is a maximum of 0.305 pm, preferably a maximum of 0.285 pm.

[0106] This wafer preferably has a diameter of between 70 and 80 mm.

[0107] 56. III / V wafer according to point 55, wherein in a measuring field of 20 mm x 20 mm on at least one surface of the wafer the maximum distance between the highest or lowest point above or below the shifted focal plane and the shifted focal plane is between 0.116 pm and 0.305 pm, preferably between 0.116 and 0.285.

[0108] This wafer preferably has a diameter of between 70 and 80 mm.

[0109] 57. IIIA / -wafer, which has two opposing surfaces, wherein the taper value is a maximum of 0.588 pm.

[0110] 58. IIIA / -wafer having two opposing surfaces, wherein the taper value is between 0.286 pm and 0.588 pm, preferably between 0.033 pm and 0.588 pm.

[0111] The taper value is the amount of the slope in the plane averaged over the surface of the wafer.

[0112] 59. IIIA / -Wafer according to point 57, wherein the taper value is a maximum of 0.554 pm.

[0113] 60. IIIA / -wafer according to point 59, wherein the taper value is between 0.033 pm and 0.554 pm, preferably a maximum of 0.43 pm. 61. III / V-wafer having two opposing surfaces, wherein the distance between the highest point of the wafer above the focal plane and the lowest point of the wafer below the focal plane is a maximum of 1.535 pm.

[0114] This parameter is the global TIR value (total indicated reading), i.e., based on the entire wafer.

[0115] 62. III / V wafer having two opposing surfaces, wherein the distance between the highest point of the wafer above the focal plane and the lowest point of the wafer below the focal plane is between 1.027 pm and 1.535 pm, preferably between 0.491 pm and 1.535 pm.

[0116] 63. III / V wafer according to point 61, wherein the distance between the highest point of the wafer above the focal plane and the lowest point of the wafer below the focal plane is a maximum of 1.125, preferably a maximum of 1.028 pm.

[0117] 64. III / V wafer according to point 63, wherein the distance between the highest point of the wafer above the focal plane and the lowest point of the wafer below the focal plane is between 0.491 pm and 1.125 pm.

[0118] 65. III / V wafer according to point 61, wherein the distance between the highest point of the wafer above the focal plane and the lowest point of the wafer below the focal plane is a maximum of 0.984 pm.

[0119] This wafer preferably has a diameter of between 70 and 80 mm.

[0120] 66. III / V wafer according to point 65, wherein the distance between the highest point of the wafer above the focal plane and the lowest point of the wafer below the focal plane is between 0.491 pm and 0.984 pm.

[0121] This wafer preferably has a diameter of between 70 and 80 mm.67. 111 / V wafer, which has two opposing surfaces, wherein the distance between the point with the greatest distance to the focal plane and the focal plane is a maximum of 0.996 pm.

[0122] This parameter is the global FPD value (focal plane deviation), i.e., relative to the entire wafer.

[0123] 68. IIIA / -Wafer, which has two opposing surfaces, wherein the distance of the point with the greatest distance to the focal plane and the focal plane is between 0.519 pm and 0.996 pm, preferably between 0.341 pm and 0.996 pm.

[0124] 69. IIIA / -Wafer according to point 67, wherein the distance of the point with the greatest distance to the focal plane and the focal plane is a maximum of 0.649 pm.

[0125] 70. IIIA / -Wafer according to item 69, wherein the distance of the point with the greatest distance to the focal plane and the focal plane is between 0.341 pm and 0.649 pm.

[0126] 71. IIIA / -Wafer, which has two opposing surfaces, wherein the distance between the highest and lowest point of a surface is a maximum of 1.419 pm.

[0127] This parameter is the global TTV value (total thickness variation), i.e., related to the entire wafer.

[0128] During the measurement, the other surface is drawn in from below by vacuum.

[0129] 72. IIIA / -wafer having two opposing surfaces, wherein the distance between the highest and lowest point of a surface is between 1.029 pm and 1.419 pm, preferably between 0.535 pm and 1.419 pm. 73. III / V-wafer according to point 71, wherein the distance between the highest and lowest point of a surface is a maximum of 1.047 pm.

[0130] During the measurement, the other surface is drawn in from below by vacuum.

[0131] 74. III / V wafer according to point 73, wherein the distance between the highest and lowest point of a surface is between 0.535 pm and 1.047 pm.

[0132] 75. III / V wafer according to item 71, wherein the distance between the highest and lowest point of a surface is a maximum of 0.964 pm.

[0133] This wafer preferably has a diameter of between 70 and 80 mm.

[0134] 76. III / V wafer according to point 75, wherein the distance between the highest and lowest point of a surface is between 0.535 pm and 0.964 pm.

[0135] This wafer preferably has a diameter of between 70 and 80 mm.

[0136] 77. Ill / V wafer, which has two opposing surfaces, with a maximum WARP value of 2.687 pm.

[0137] 78. Ill / V wafer, which has two opposing surfaces, wherein the WARP value is between 1.225 pm and 2.687 pm.

[0138] 79. III / V wafer according to item 77, wherein the WARP value is a maximum of 2.211 pm.

[0139] 80. III / V wafer, according to item 79, wherein the WARP value is between 0.917 pm and 2.211 pm.

[0140] 81. III / V wafer having two opposing surfaces, wherein the Bow value is a maximum of 1.408 pm. 82. 111 / V wafer having two opposing surfaces, wherein the Bow value is between -0.636 pm and 1.408 pm.

[0141] 83. III / V wafer according to item 81, wherein the Bow value is a maximum of 1.495 pm.

[0142] 84. III / V wafer, according to item 83, wherein the Bow value is between -1.022 pm and 1.495 pm.

[0143] 85. Ill / V wafer, which has two opposing surfaces, where the Sori value is a maximum of 3.091 pm.

[0144] 86. Ill / V wafer, which has two opposing surfaces, wherein the Sori value is between 1.074 pm and 3.091 pm.

[0145] 87. III / V wafer according to item 85, wherein the Sori value is a maximum of 2.263 pm.

[0146] 88. III / V wafer, according to item 86, wherein the Sori value is between 0.901 pm and 2.263 pm.

[0147] 89. Ill-V wafer according to one of points 30 to 88, wherein the Hl element is selected from Ga, AI and In, and the V element is selected from P and As.

[0148] 90. III-V wafer according to one of points 30 to 89, wherein the Hl element is In, and the V element is P.

[0149] 91. III-V wafer according to the previous item 90, wherein the wafer is n-type semiconducting.

[0150] 92. III-V wafer according to the preceding items 30 to 91, wherein the wafer is sulfur-doped and has a charge carrier concentration in the range of 2 to 8 × 18 cm⁻¹ -3 lies.93. Ill-V wafer according to the previous points 30 to 92, wherein its specific resistance is in the range of 5E-4 to 2E-3 Qcm.

[0151] 94. III-V wafer according to the preceding items 30 to 93, wherein the mean dislocation density is 500 cm³ -2or less.

[0152] 95. III-V wafer according to one of points 30 to 94, wherein the wafer diameter is between 75 and 155 mm, and the wafer thickness is 575 pm to 650 pm.

[0153] 96. III-V wafer according to one of points 30 to 95, wherein the wafer diameter is between 75 and 155 mm, and the wafer thickness is 425 pm to 575 pm.

[0154] 97. Group of Ill / V wafers having two opposing surfaces, wherein in a measurement field on at least one surface of the wafer the mean distance between the highest and lowest point of the surface is a maximum of 0.894 pm.

[0155] This parameter is the LTV value (local thickness variation).

[0156] The standard deviation should preferably be a maximum of 0.0882 pm.

[0157] 98. Group of Ill / V wafers according to point 97, wherein in a measurement field on at least one surface of the wafer the mean distance between the highest and lowest point of the surface is a maximum of 0.653 pm.

[0158] The standard deviation should preferably be a maximum of 0.044 pm.

[0159] 99. Group of III / V wafers according to point 97, wherein in a measurement field on at least one surface of the wafer the mean value of the distance between the highest and lowest point of the surface is a maximum of 0.412 pm. The standard deviation is preferably a maximum of 0.043 pm.

[0160] This group of wafers preferably has a diameter of between 70 and 80 mm.

[0161] 100. Group of Ill / V wafers having two opposing surfaces, wherein in a measurement field on at least one surface of the wafer the mean value of the maximum distance between the highest point above the shifted focal plane and the lowest point below the shifted focal plane is a maximum of 0.573 pm.

[0162] This parameter is the LTIR value (local total indicated reading).

[0163] The standard deviation should preferably be a maximum of 0.052 pm.

[0164] 101. Group of III / V wafers according to point 100, wherein in a measurement field on at least one surface of the wafer the mean value of the maximum distance between the highest point above the shifted focal plane and the lowest point below the shifted focal plane is a maximum of 0.370 pm

[0165] The standard deviation should preferably be a maximum of 0.020 pm.

[0166] 102. Group of Ill / V wafers having two opposing surfaces according to point 100, wherein in a measurement field on at least one surface of the wafer the mean value of the maximum distance between the highest point above the shifted focal plane and the lowest point below the shifted focal plane is a maximum of 0.336 pm.

[0167] The standard deviation should preferably be a maximum of 0.052 pm.

[0168] This group of wafers preferably has a diameter of between 70 and 80 mm.103. Group of III / V wafers which have two opposing surfaces, wherein in a measurement field on at least one surface of the wafer the mean value of the maximum distance between the highest or lowest point above or below the displaced focal plane and the displaced focal plane is a maximum of 0.445 pm.

[0169] This parameter is the LFPD value (local focal plane deviation).

[0170] The standard deviation should preferably be a maximum of 0.042 pm.

[0171] 104. Group of Ill / V wafers according to point 103, wherein in a measurement field on at least one surface of the wafer the mean value of the maximum distance between the highest or lowest point above or below the shifted focal plane and the shifted focal plane is a maximum of 0.270 pm.

[0172] The standard deviation should preferably be a maximum of 0.023 pm.

[0173] 105. Group of III / V wafers according to point 103, wherein in a measurement field on at least one surface of the wafer the mean value of the maximum distance between the highest or lowest point above or below the shifted focal plane and the shifted focal plane is a maximum of 0.238 pm, preferably a maximum of 0.199 pm.

[0174] The standard deviation is preferably a maximum of 0.031 pm or, in the preferred case, 0.030 pm.

[0175] 106. Group of Ill / V wafers according to one of points 103 to 105, wherein the size of the measurement field is between 5 mm x 5 mm and 15 mm x 15 mm.

[0176] 107. Group of Ill / V wafers having two opposing surfaces, wherein in a measurement field of 20 mm x 20 mm on at least one surface of the wafer the mean value of the maximum distance between the highest or lowest point above or below the shifted focal plane and the shifted focal plane is a maximum of 0.279 pm.

[0177] This parameter is the LFPD value (local focal plane deviation) with a measurement field size of 20 mm x 20 mm.

[0178] The standard deviation should preferably be a maximum of 0.042 pm.

[0179] 108. Group of Ill / V wafers according to point 107, wherein in a measurement field of 20 mm x 20 mm on at least one surface of the wafer the mean value of the maximum distance between the highest or lowest point above or below the shifted focal plane and the shifted focal plane is a maximum of 0.258 pm.

[0180] The standard deviation should preferably be a maximum of 0.024 pm.

[0181] 109. Group of IIIA / -wafers according to item 107, which have two opposing surfaces, wherein in a measurement field of 20 mm x 20 mm on at least one surface of the wafer the mean value of the maximum distance between the highest or lowest point above or below the shifted focal plane and the shifted focal plane is a maximum of 0.184 pm.

[0182] The standard deviation should preferably be a maximum of 0.040 pm.

[0183] This group of wafers preferably has a diameter of between 70 and 80 mm.

[0184] 110. Group of IIIA / -wafers having two opposing surfaces, wherein the mean taper value on one surface of the wafer is a maximum of 0.441 pm

[0185] The standard deviation is preferably a maximum of 0.082 pm. The taper value is the magnitude of the slope in the plane, averaged over the surface of the wafer, over the entire surface of the wafer.

[0186] 111. Group of Ill / V wafers according to item 110, wherein the mean taper value is a maximum of 0.371 pm.

[0187] The standard deviation should preferably be a maximum of 0.040 pm.

[0188] 112. Group of Ill / V wafers having two opposing surfaces, wherein the mean taper value on one surface of the wafer is a maximum of 0.280 pm.

[0189] The standard deviation should preferably be a maximum of 0.082 pm.

[0190] This group of wafers preferably has a diameter of between 70 and 80 mm.

[0191] 113. Group of Ill / V wafers having two opposing surfaces, wherein the mean distance of the highest point of the wafers above the focal plane and the lowest point of the wafers below the focal plane is a maximum of 1.258 pm.

[0192] This parameter is the global TIR value (total indicated reading), i.e., based on the entire wafer.

[0193] The standard deviation should preferably be a maximum of 0.102 pm.

[0194] 114. Group of Ill / V wafers according to item 113, wherein the mean distance of the highest point of the wafer above the focal plane and the lowest point of the wafer below the focal plane is a maximum of 0.936 pm.

[0195] The standard deviation here is preferably a maximum of 0.121 pm.115. Group of Ill / V wafers which have two opposing surfaces, wherein the mean of the distance of the highest point of the wafers above the focal plane and the lowest point of the wafers below the focal plane is a maximum of 0.730 pm.

[0196] The standard deviation should preferably be a maximum of 0.098 pm.

[0197] This group of wafers preferably has a diameter of between 70 and 80 mm.

[0198] 117. Group of Ill / V wafers having two opposing surfaces, wherein the mean distance of the point with the greatest distance to the focal plane and the focal plane is at most 0.721 pm

[0199] This parameter is the global FPD value (focal plane deviation), i.e., relative to the entire wafer.

[0200] The standard deviation should preferably be a maximum of 0.108 pm.

[0201] 118. Group of Ill / V wafers according to point 117, wherein the mean of the distance of the point with the greatest distance to the focal plane and the focal plane is a maximum of 0.547 pm.

[0202] The standard deviation should preferably be a maximum of 0.083 pm.

[0203] 119. Group of Ill / V wafers having two opposing surfaces, wherein the mean distance of the point with the greatest distance to the focal plane and the focal plane is at most 0.502 pm.

[0204] The standard deviation is preferably a maximum of 0.068 pm. This group of wafers preferably has a diameter of between 70 and 80 mm.

[0205] 120. Group of Ill / V wafers having two opposing surfaces, wherein the mean distance between the highest and lowest point of a surface is a maximum of 1.189 pm.

[0206] This parameter is the global TTV value (total thickness variation), i.e., related to the entire wafer.

[0207] The standard deviation should preferably be a maximum of 0.084 pm.

[0208] 121. Group of III / V wafers according to item 120, wherein the mean distance between the highest and lowest point of a surface is a maximum of 0.930 pm.

[0209] The standard deviation should preferably be a maximum of 0.068 pm.

[0210] 122. Group of Ill / V wafers having two opposing surfaces, wherein the mean distance between the highest and lowest point of a surface is a maximum of 0.760 pm.

[0211] The standard deviation should preferably be a maximum of 0.103 pm.

[0212] This group of wafers preferably has a diameter of between 70 and 80 mm.

[0213] 123. Group of Ill-V wafers according to one of points 97 to 122, wherein the Hl element is selected from Ga, AI and In, and the V element is selected from P and As.

[0214] 124. Group of III-V wafers according to one of points 97 to 123, wherein the Hl element is In, and the V element is P. 125. Group of III-V wafers according to the previous point 124, wherein the wafers are n-type semiconducting.

[0215] 126. Group of III-V wafers according to the preceding items 97 to 125, wherein the wafers are sulfur-doped and have a charge carrier concentration in the range of 2 to 8 × 18 cm⁻¹ -3 lies.

[0216] 127. Group of III-V wafers according to the preceding points 97 to 126, wherein the resistivity is in the range of 5E-4 to 2E-3 Qcm.

[0217] 128. Group of III-V wafers according to the preceding items 97 to 127, wherein the mean dislocation density is 500 cm³ -2 or less.

[0218] 129. Group of III-V wafers according to one of points 97 to 128, wherein the wafer diameter is between 75 and 155 mm, and the thickness is 575 pm to 650 pm.

[0219] 130. Group of III-V wafers according to one of points 97 to 129, wherein the wafer diameter is between 75 and 155 mm, and the thickness is 425 pm to 575 pm.

[0220] 131. Use of an α-hydroxycarboxylic acid in a polishing solution for semiconductor wafers to prevent flocculation in wastewater.

[0221] 132. Use as per point 131, wherein the polishing solution is used for polishing InP wafers.

[0222] Preferred wafers or groups of wafers have at least one of the parameters shown. Preferably, they have at least two, and preferably more than two, of the parameters shown. DESCRIPTION OF THE FIGURES

[0223] Preferred embodiments of the present invention are described in more detail below with reference to the accompanying figures.

[0224] Fig. 1 shows a device for planarizing multiple wafers (on both wafer sides), both in the bottom view and in the front view.

[0225] Fig. 2 shows a device for planarizing multiple wafers (on one wafer side), both in the bottom view and in the front view.

[0226] Fig. 3 shows a diagram illustrating the path of a point on a wafer relative to the polishing cloth (simulation).

[0227] Fig. 4 schematically shows the measurement of the global thickness variation (distance between the highest and lowest point of a surface) of a wafer.

[0228] Fig. 5 schematically shows the measurement of the taper value of a wafer.

[0229] Fig. 6 schematically shows the measurement of the TIR value of a wafer (distance of the highest point of the wafer above the focal plane and the lowest point of the wafer below the focal plane).

[0230] Fig. 7 schematically shows the measurement of the focal plane deviation (global). The focal plane deviation is the distance between the point furthest from the focal plane and the focal plane itself. Fig. 8 schematically shows the measurement of the local thickness deviation (distance between the highest and lowest points of the surface in a measurement field).

[0231] Fig. 9 schematically shows the measurement of the local TIR value (maximum distance between the highest point above the shifted focal plane and the lowest point below the shifted focal plane in a measurement field).

[0232] Fig. 10 schematically shows the measurement of the Bow value.

[0233] Fig. 11 schematically shows the measurement of the warp value.

[0234] Fig. 12 schematically shows the measurement of the Sori value.

[0235] Fig. 13 shows the global thickness deviation, taper value, TIR value and focal plane deviation for six groups of InP wafers.

[0236] Fig. 14 shows the local thickness variation, the local TIR value, the local deviation from the shifted focal plane, and the local deviation of the shifted focal plane with a measurement field of 20 x 20 mm for six groups of InP wafers.

[0237] Fig. 15 shows the warp value, the bow value and the sori value for several groups of InP wafers.

[0238] Fig. 1 shows a device for the two-sided planarization of wafers, shown above as a view from below, and below as a view from the front.

[0239] The wafers W are mounted in carriers 13 between an upper polishing plate 14 and a lower polishing plate 15. The upper polishing plate 14 and the lower polishing plate 15 are each covered with a polishing cloth 20 (not shown). The upper polishing plate 14 and the lower polishing plate 15 rotate, and the carriers 13, which hold the wafers W, also rotate. This occurs because an outer toothed ring 11 rotates clockwise, while an inner toothed ring 12 rotates counterclockwise, thus driving the carriers 13. Furthermore, a polishing compound distribution device 16 is shown in the front view. This device has several polishing compound inlets, allowing the polishing compound to be injected directly into the area between the upper polishing plate 14 and the lower polishing plate 15.

[0240] The chemical-mechanical polishing process utilizes the oxidizing effect of the polishing agent, the mechanical friction of the polishing cloth 20 (not shown), and the mechanical component of the polishing agent (particles) to polish both sides of the wafers W simultaneously. The carriers 13 hold the wafers W, thus ensuring a guided movement across the entire surface of the polishing cloth 20 (not shown). The flow rate of the polishing agent, the polishing pressure, and the speed of the carriers 13 influence the material removal rate.

[0241] Because of the remaining residues of the oxidizing reagents of the polishing compound in the polishing cloth, it is necessary to perform a mechanical dressing of the polishing cloth between the individual polishing processes.

[0242] Fig. 2 shows a device for planarizing wafers on one side. Only one polishing plate (the lower polishing plate 15) is present, on which working discs 21 and wafers W are arranged. The wafers are bonded to the working discs 21. An internal drive disc 22 drives the working discs 21, causing them to rotate about their own axis, while the lower polishing plate 15 also rotates in the opposite direction. Polishing medium inlets 17 are also visible. Furthermore, pressure-generating devices 19 are shown, which press the working disc 21 with the wafers W against the lower polishing plate 15.

[0243] Fig. 3 shows a diagram illustrating the probability of a wafer being located in specific areas of the polishing cloth. Each diagram depicts a view of the polishing cloth plane. The path of a point on the wafer relative to the polishing cloth is simulated as an example.

[0244] The carriers with the wafers (in the corresponding holes) constantly change their positions. Structuring the polishing cloth helps to make the polishing removal more uniform. Figure 4 schematically shows the parameter of the global thickness variation W of a wafer. Here, a highest elevation 2 and a lowest elevation 3 are shown, each relative to a reference surface (here: the back surface of the wafer). The difference between these two elevations is the global thickness variation (also called TTV).

[0245] Figure 5 schematically illustrates the so-called taper value W of a wafer. A reference plane 1, parallel to the back side of the wafer, is used. Furthermore, an average plane (best-fit plane) 4 is used. The taper value is the difference in parallelism between the reference plane 1 (parallel to the back side of the wafer, passing through the lowest point of the average plane (best-fit plane) 4) and the reference plane 1 (parallel to the back side of the wafer). The numerical value given is the amount of the slope in the average plane (best-fit plane) 4 across the entire surface of the wafer.

[0246] Figure 6 shows the TIR value of a wafer W. A focal plane 7 is defined. A highest elevation 2 and a lowest elevation 3 are shown, representing the greatest distance 8 of the focal plane 7 to the highest elevation 2 and the greatest distance 9 of the focal plane 7 to the lowest elevation 3. The TIR value is the sum of the greatest distance 8 (of the focal plane 7 to the highest elevation 2) and the greatest distance 9 (of the focal plane 7 to the lowest elevation 3).

[0247] Fig. 7 shows a deviation of the focal plane (also called FPD) of a wafer W. The focal plane is indicated by the reference symbol 7; it again has a highest elevation 2 and a lowest elevation 3. The distance of the highest elevation 2 from the focal plane 7 is thus the greatest distance 8 above the focal plane, and the distance between the focal plane 7 and the lowest elevation 3 is the greatest distance 9 below the focal plane. The FPD value is therefore the greater of the greatest distance 8 above the focal plane and the greater of the greatest distance 9 below the focal plane. Fig. 8 shows the local thickness variation of a wafer W in several measurement fields F. Here it is evident that there can be different values ​​depending on the measurement field F. In each measurement field F, there is a highest elevation 2 and a lowest elevation 3 (shown in the left measurement field, but applicable to all measurement fields). Furthermore, a virtual reference plane is shown as a dashed line for each measurement field F.The distance from the lowest elevation 3 to the reference plane is marked with the letter A, and the distance from the highest elevation 2 to the reference plane is marked with the letter B. The sum of these two distances gives the local thickness variation, which, in other words, is the distance from the highest elevation 2 to the lowest elevation 3. This varies depending on the measurement field.

[0248] Fig. 9 shows the LTIR value in several measurement fields F of a wafer W, where the LTIR value in one measurement field F is schematically represented: There is a shifted focal plane 7a, a maximum distance 8a above the shifted focal plane 7a, and a maximum distance 9a below the shifted focal plane. The LTIR value is the sum of the maximum distance above the shifted focal plane 7a and the maximum distance below the shifted focal plane. The shifted focal plane 7a is obtained by shifting the focal plane 7a such that the local average plane coincides with the shifted focal plane at the center of the measurement field.

[0249] Fig. 10 shows the Bow value of a wafer W. Here, the distance between the wafer W and the focal plane 7 at the center M of the wafer W is shown, which is the Bow value.

[0250] Fig. 11 schematically shows the warp value of a wafer W. The focal plane 7 is also shown; the warp value is the sum of the greatest distance 8 above the focal plane and the greatest distance 9 below the focal plane of the respective wafer surfaces.

[0251] Fig. 12 schematically shows the Sori value of a wafer W. Here, a best-fit plane 4 is shown, and there exists a point with the greatest distance 6 below the best-fit plane 4, and there also exists a point with the greatest distance 5 above the best-fit plane 4. The Sori value is the largest of these two.

[0252] Fig. 13 shows measured values ​​of the global thickness variation (TTV), taper value, TIR value, and focal plane deviation (FPD) – all global flatness parameters of six wafer groups. Batch A is a group of wafers (with a diameter of 76.2 mm) produced according to the prior art manufacturing process. Batch B is a wafer in which the polishing agent according to the invention (with α-hydroxycarboxylic acid polishing agent) was used. Batches C and D also used α-hydroxycarboxylic acid as a polishing agent, but additionally a structured polishing cloth was used. Batch E, like batch D, has a wafer diameter of 100 mm. Batch F has a wafer diameter of 150 mm. It is evident that a significant improvement occurs between Batch A and Batch B in all the values ​​considered (i.e.,(significantly lower flatness deviations), batches C and D show a further improvement. Batches E and F also perform better than batch B in most of the parameters considered here, although usually not with as large an improvement as batches C and D.

[0253] The box plots are presented as follows: The box contains the values ​​from the first to the third quartile of all measurements, excluding outliers. The marker inside the box indicates the mean. The horizontal lines (whiskers) above and below the box indicate the minimum and maximum values, excluding outliers. Outliers are shown as individual points above and below the horizontal lines (whiskers).

[0254] Fig. 14 shows the local thickness variation (LTV), the local Tir value (LTIR), the local deviation from the focal plane (LFPD), and the local deviation from the focal plane (LFPD) in a measurement field of 20 mm x 20 mm (LFPD20x20) – all local flatness parameters for six wafer groups. Similar results are seen here as in Fig. 14; however, when comparing batches A and B, the use of the polishing compound according to the invention in batch B results in a significant improvement in the individual values. Even better local flatness parameters can be achieved with additional polishing cloth structuring (batches C, D, E, and F).

[0255] Figure 15 shows the warp value, bow value, and sori value of four or five wafer groups, respectively. The sori value for batch E, however, was not measured. Regarding the aforementioned wafer values, there is a slight improvement between batches A and B, but there are outliers in the warp and bow values. There is also some variation in the sori value.

[0256] In a method according to the invention for polishing III / -semiconductor wafers, the semiconductor wafer is brought into contact with a polishing cloth at least on one side to be polished, wherein the semiconductor wafer and the polishing cloth perform a rotational movement relative to each other. A polishing agent is added between the wafer and the polishing cloth, comprising: deionized water, particles, and at least one α-hydroxycarboxylic acid. The polishing agent thus contains a polishing chemistry (active chlorine), a polishing mechanism (particles), and a complexing chemistry (α-hydroxycarboxylic acid).

[0257] Surprisingly, it was found that an α-hydroxycarboxylic acid increases the removal rate on the wafer surface, but still allows for the flattest possible wafer surface, especially in a defined measurement field.

[0258] The α-hydroxycarboxylic acid is preferably selected from citric acid, tartaric acid, glycolic acid, lactic acid, malic acid and mandelic acid.

[0259] a-Hydroxycarboxylic acids have proven to be particularly efficient here, as they act as complexing agents and simultaneously serve to adjust the pH value of the polishing solution.

[0260] Complexing agents can also lead to flocculation in wastewater, as they form complexes with wastewater treatment additives, such as flocculants, thereby rendering these additives inactive. However, within the scope of the present invention, it was surprisingly discovered that lactic acid does not negatively affect wastewater treatment, because the lactic acid molecule can be easily cleaved, and the products then no longer have a complexing effect in wastewater treatment.

[0261] Therefore, an α-hydroxycarboxylic acid as a component of the polishing agent for wafers offers further advantages.

[0262] Preferably, the polishing compound contains between 0.1 vol% and 1.0 vol% of the substance.

[0263] containing α-hydroxycarboxylic acid, preferably at least 0.4 vol%.

[0264] The active complexing range in the polishing solution begins at 0.1 vol%, meaning that complexes can form in sufficiently high concentrations. However, above 1.0 vol%, inhomogeneous reactivity of the polishing solution is to be expected, meaning that complexes are no longer formed selectively.

[0265] Preferably, the polishing compound contains between 1 vol.% and 10 vol.% particles, and more preferably between 6 vol.% and 8 vol.% particles.

[0266] Preferably, the main fraction of particles lies in a range of 10 nm to 30 nm particle diameter, with the specific surface area of ​​the particles further preferably in a range of 100 m². 2 / g up to 150 m 2 / g lies.

[0267] The particles preferably contain SiO2, TiÜ2, Al2O3, MnÜ2 and / or CeO2, further preferably SiÜ2.

[0268] In a chemical-mechanical polishing process, particles are preferably selected such that they can remove the softer reaction product from the wafer surface, but cannot attack or scratch the unaffected semiconductor material. SiÜ2 has very preferably suitable properties for InP, as precisely this condition is met.

[0269] Preferably, the polishing agent also contains active chlorine, preferably between 1 g / l and 10 g / l active chlorine. Active chlorine comprises various chlorine species that have an oxidizing effect. Active chlorine describes those chlorine species that can stimulate and increase the oxidation of metals, especially non-ferrous metals. These are highly reactive chlorine compounds that can increase oxide formation during chemical-mechanical polishing. They act as follows:

[0270] Surface atoms are oxidized; unlike non-oxidized semiconductor material, these oxides can be mechanically removed from the wafer surface using a polishing cloth and abrasive particles.

[0271] Sodium dichloroisocyanurate (Na-DCC) serves as a starting material for the formation of active chlorine species. When solid sodium dichloroisocyanurate is dissolved in water, hypochlorous acid is formed in the solution.

[0272] Sodium dichloroisocyanurate + water

[0273]

[0274] hypochlorous acid + sodium cyanurate

[0275]

[0276] Hypochlorous acid is a strong oxidizing agent and decomposes into hydrochloric acid and highly reactive oxygen radicals:

[0277] (1) HCIO - HCl + O (oxygen radical)

[0278] In acidic conditions, hypochlorous acid decomposes into molecular chlorine and water:

[0279] (2) 2 HCIO + 2 H + - CI2T + 2 H2O

[0280] For example, at a pH of 2, hypochlorous acid and molecular chlorine are present in a ratio of 1:1.

[0281] The value of active chlorine indicates the amount of oxidatively active species and can be determined indirectly using iodometry.

[0282] This takes advantage of the fact that the oxidizing species oxidize iodide ions to molecular iodine:(3) 2 HCIO + 2 K + + 2 I' l2+ H2O + 2 K + + 2Ch

[0283] (4) CI2 + 2 K + + 2 I' - I2 + 2 K + +2 Ch

[0284] The iodine formed is determined by means of a redox titration with sodium thiosulfate solution, and conclusions can be drawn about the amount of the originally oxidatively active chlorine content.

[0285] (5) 2 S2O3 2 - + I2 - S4O6 2 - + 2 1-

[0286] Preferably, the polishing agent has a pH value between 2 and 4.

[0287] Surprisingly, it was found that this range is a particularly favorable operating range for the formation of readily soluble oxide groups of In and P. Furthermore, it was found within the scope of the present invention that the pH range of 2 to 4 is ideal with respect to the ratio of material removal to surface roughness.

[0288] The insolubility of indium reaction products in the alkaline pH range necessitates the use of polishing chemistry in the acidic pH range.

[0289] Preferably, several semiconductor wafers with at least one side to be polished are brought into contact with a polishing cloth, wherein several semiconductor wafers are inserted into a carrier, and the carrier moves on a circular path and rotates simultaneously, wherein the polishing cloth preferably also rotates about its central axis.

[0290] In such a multi-wafer process, several wafers are polished, leading to increased product throughput. Furthermore, a group of wafers with very similar flatness properties can be generated.

[0291] Preferably, at least one area of ​​the polishing cloth is structured, wherein the thickness of the polishing cloth is 500–800 pm less in structured areas than in unstructured areas, and / or the polishing cloth is hydrophilized in structured areas. Preferably, the total area fraction of the structured areas of the polishing cloth is between 1% and 18% of the total area of ​​the polishing cloth, particularly preferably between 2% and 12%. It can also be approximately 7.1%, preferably 7.1%.

[0292] The polishing action differs between the outer and inner areas of the polishing cloth, i.e., closer to the polishing machine's axis of symmetry. This can be attributed, among other things, to the fact that the polishing fluid flows from the inside out, moving from an area of ​​small volume per segment to an area of ​​larger volume per segment. Furthermore, the polishing fluid is already used by the polishing process further inward before it reaches the polishing area further out. An additional asymmetry is created by the rotation of the polishing carrier around its own axis when the polishing cloth rotates around its own axis.

[0293] The polishing cloth is structured to compensate for the different polishing abrasion in different areas of the polishing cloth.

[0294] The purpose of the structuring is to reduce the polishing attack on the wafer.

[0295] One option is to make the surface of the polishing cloth hydrophilic, thus reducing the amount of polishing fluid present in that area. It is also possible to reduce the thickness of the polishing cloth in desired areas. This is achieved by ensuring the underside of the polishing cloth, where it is attached to the rotating part of the polishing machine, lies flat, while a step is created on the top side of the polishing cloth, the side facing the wafer. This step can be created by grinding the polishing cloth or by cutting out individual layers. Another possibility is to densify the polishing cloth in the areas to be textured by applying heat.

[0296] This creates areas where less or no pressure is exerted on the polishing agent or the wafer surface.

[0297] An IIIA / -wafer according to the invention has two opposing surfaces and further has at least one of the following properties: • in a measurement field on at least one surface of the wafer, the distance between the highest and lowest point of the surface is a maximum of 1.067 pm;

[0298] • In a measurement field on at least one surface of the wafer, the maximum distance between the highest or lowest point above or below the shifted focal plane and the shifted focal plane is a maximum of 0.544 pm.

[0299] In photolithography, smaller height deviations are advantageous in the exposed area to ensure homogeneous exposure across the surface. This allows for the fabrication of homogeneous structures (gate widths in transistors, for example). Such local flatness parameters lead to uniform growth of an epitaxial layer on the wafer surface and thus to homogeneous exposure in a photolithography process.

[0300] Surprisingly, it was found within the scope of the present invention that such local flatness parameters can be achieved with the new planarization method according to the invention.

[0301] Preferably, the measuring field has a size of 20 mm x 20 mm, and the maximum distance between the highest or lowest point above or below the shifted focal plane and the shifted focal plane is a maximum of 0.367 pm.

[0302] In this measurement field size, this value leads in particular to a uniform growth of an epitaxial layer on a surface of the wafer and thus to a homogeneous exposure in a photolithography process.

[0303] Preferably, an Ill / V wafer has the following local flatness properties: • in a measurement field on at least one surface of the wafer, the distance between the highest and lowest point of the surface is a maximum of 0.754 pm;

[0304] • In a measurement field on at least one surface of the wafer, the maximum distance between the highest point above the shifted focal plane and the lowest point below the shifted focal plane is a maximum of 0.48 pm;

[0305] • In a measurement field on at least one surface of the wafer, the maximum distance between the highest or lowest point above or below the shifted focal plane and the shifted focal plane is a maximum of 0.313 pm.

[0306] Such local flatness parameters lead to a particularly uniform growth of an epitaxial layer on a surface of the wafer and thus to a particularly homogeneous exposure in a photolithography process.

[0307] Preferably, the measuring field has a size of 20 mm x 20 mm, and the maximum distance between the highest or lowest point above or below the shifted focal plane and the shifted focal plane is a maximum of 0.305 pm.

[0308] In this measurement field size, this value leads in particular to a particularly uniform growth of an epitaxial layer on a surface of the wafer and thus to a particularly homogeneous exposure in a photolithography process.

[0309] A IIIA / -wafer, which has two opposing surfaces, exhibits at least one of the following properties:

[0310] The taper value is a maximum of 0.588 pm; preferably a maximum of 0.554 pm, more preferably a maximum of 0.43 pm; the distance between the highest point of the wafer above the focal plane and the lowest point of the wafer below the focal plane is a maximum of 1.535 pm; preferably a maximum of 1.028 pm, and even more preferably 0.984 pm;

[0311] • The distance between the point with the greatest distance to the focal plane and the focal plane is a maximum of 0.996 pm; preferably a maximum of 0.649 pm;

[0312] • The distance between the highest and lowest point of a surface is a maximum of 1.419 pm; preferably a maximum of 1.047 pm, further preferably a maximum of 10.964 pm.

[0313] These global flatness parameters on at least one wafer side ensure that the entire wafer is as flat as possible, so that it is as well prepared as possible for the subsequent finalization (final polishing) process - enabling a uniform polish.

[0314] Furthermore, the global flatness of a wafer is important in a subsequent epitaxial process for the homogeneous epitaxial deposition of atoms across the entire surface. Global flatness is also crucial for a lithography process, as flatness deviations would interfere with focusing.

[0315] Surprisingly, it was found within the scope of the present invention that such global flatness parameters can be achieved with the new planarization method according to the invention.

[0316] Preferably the Hl element is selected from Ga, AI and In, and the V element is selected from P and As, wherein the Hl element is preferably In, and the V element is preferably P.

[0317] Preferably, the wafer diameter is between 75 and 155 mm, and the wafer thickness is 575 pm to 650 pm. It has therefore been shown that the inventive method works for a wide range of common wafer diameters and wafer thicknesses, and that wafers exhibit improved flatness parameters compared to the prior art.

[0318] A group of III / V wafers according to the invention, which have two opposing surfaces, has at least one of the following properties:

[0319] • In a measurement field on at least one surface of the wafer, the mean distance between the highest and lowest point of the surface is a maximum of 0.894 pm;

[0320] • In a measurement field on at least one surface of the wafer, the mean value of the maximum distance between the highest point above the shifted focal plane and the lowest point below the shifted focal plane is a maximum of 0.573 pm;

[0321] • In a measurement field on at least one surface of the wafer, the mean value of the maximum distance between the highest or lowest point above or below the shifted focal plane and the shifted focal plane is a maximum of 0.445 pm.

[0322] It has been demonstrated that particularly advantageous flatness parameters can be achieved for a group of wafers that are simultaneously planarized in a single machine as part of a multi-wafer process. This has the advantage that multiple wafers of the same quality can be produced, which is beneficial for setting up equipment for a subsequent photolithography process on a fully developed epitaxial layer.

[0323] Preferably, for a group of wafers in a measurement field of 20 mm x 20 mm on at least one surface of the wafer, the mean value of the maximum distance between the highest or lowest point above or below the shifted focal plane and the shifted focal plane is a maximum of 0.279 pm. This is particularly advantageous for subsequent growth of epitaxial layers and subsequent photolithography processes.

[0324] The use of an α-hydroxycarboxylic acid in a polishing solution for semiconductor wafers according to the invention leads to the prevention of flocculation in the wastewater. EXAMPLES

[0325] Wafer

[0326] The wafers used for the examples are InP wafers with nominal diameters of 76.2 mm ± 0.5 mm, 100 mm ± 0.5 mm, and 150 mm ± 0.5 mm, and a target thickness of 625 pm ± 25 pm (for nominal diameters of 76.2 mm and 100 mm) or 675 pm ± 25 pm (for a nominal diameter of 150 mm). The wafers were sulfur-doped and are n-type semiconducting. The charge carrier concentration is in the range of 2–8 × 10⁻⁸ cm⁻¹. -3 , the specific resistance is in the range of 5E-4 to 2E-3 Ohm*cm and the mean dislocation density is less than 500 cm' 2 .

[0327] The crystals were produced using a VGF (Vertical Gradient Freeze) process. In this process, the seed crystal was not moved; instead, the crystal grew upwards in a crucible from the top of the seed crystal upwards (see, for example, M. Jurisch et al. in "Handbook of Crystal Growth Bulk Crystal Growth: Basic Techniques" VOLUME II, Part A, Second Edition, Chapter 9 "Vertical Bridgman Growth of Binary Compound Semiconductors", 2015). The crucible itself was not moved; rather, the temperature field above the seed crystal exhibited a positive temperature gradient and was shifted upwards by appropriate heater control.

[0328] The produced single crystal was removed from the crucible. By removing the conical and end sections, a cylinder with an irregular outer surface was obtained. This was ground on the outer surfaces to ensure a constant cylinder diameter along its entire length. Wire lapping was then used to separate the crystals into wafers (see, for example, H.-J. Möller in "Handbook of Crystal Growth Bulk Crystal Growth: Basic Techniques" VOLUME II, Part A, Second Edition, Chapter 18 "Wafer Processing", 2015).

[0329] Alternatively, crystals could also be produced using the VG (Vertical Bridgman) method. In this method, the crucible containing the seed is moved vertically relative to the heater assembly. Combinations of the VGF and VB methods are also possible (see, for example, M. Jurisch et al. in "Handbook of Crystal Growth Bulk Crystal Growth: Basic Techniques" VOLUME II, Part A, Second Edition, Chapter 9 "Vertical Bridgman Growth of Binary Compound Semiconductors", 2015).

[0330] Alternatively, crystals can also be produced using the LEC (Liquid Encapsulated Czochralski) method. For this, a seed crystal is immersed from above into a melt and, while rotating around its own axis, withdrawn from the melt so that a single crystal grows downwards from this seed (see, for example, JB Mullin in "Handbook of Crystal Growth Bulk Crystal Growth: Basic Techniques" VOLUME II, Part A, Second Edition, Chapter s "Liquid Encapsulation and Related Technologies for the Czochralski Growth of Semiconductor Compounds", 2015).

[0331] Planarization process

[0332] A multi-wafer planarization process was used.

[0333] Non-woven polyurethane polishing cloths were used, with a Asker (Scale C) hardness of 42, a compressibility between 5 and 10%, and a density of 0.326 g / cm³. 3 and a thickness of 1.3 mm.

[0334] Five groups (batches) of wafers were planarized, as shown in Table 1.

[0335] This emerges. Batch planarizing process diameter using a polishing agent

[0336] according to the state of the art,

[0337] A 76.2 mm

[0338] no structuring of the

[0339] Polishing cloth

[0340] Use of lactic acid as

[0341] B Complexing agents, no 76.2 mm

[0342] Structuring the polishing cloth

[0343] Use of lactic acid as

[0344] C Complexing agent, structuring 76.2 mm

[0345] of the polishing cloth

[0346] Use of lactic acid as

[0347] D Complexing agent, structuring 76.2 mm

[0348] of the polishing cloth

[0349] Use of lactic acid as

[0350] E Complexing agent, structuring 100.0 mm

[0351] of the polishing cloth

[0352] Use of lactic acid as

[0353] E Complexing agent, structuring 150.0 mm

[0354]

[0355] of the polishing cloth

[0356] Table 1: Overview of the groups of wafers examined

[0357] Batches C and D had the same parameters; only a second group of wafers was measured (which were each planarized in a separate process) to show that the values ​​were nearly constant and reproducible.

[0358] The polishing solution for batch A contained the following components:

[0359] • Solution in deionized water with an active chlorine content of 4 g / l (with dichloroisocyanuric acid from starting material)

[0360] • SiO2 particles with a content of 1.2 vol%

[0361] The polishing solution for batches B, C, D, E and F contained the following components:

[0362] • Solution in deionized water with an active chlorine content of 4 g / l

[0363] • Lactic acid with a content of 0.4% by volume

[0364] • SiO2 particles with a content of 1.2 vol%

[0365] The pH value was between 2 and 4. An α-hydroxycarboxylic acid, such as lactic acid, acts here as a complexing agent, whose function can be explained as follows:

[0366] An α-hydroxycarboxylic acid acts here as a polydentate ligand with two binding sites; that is, it possesses both a hydroxyl group (-OH) and a carboxyl group (-COOH), thus enabling the formation of coordination compounds. This allows for the formation of complexes with indium, phosphorus, and also mixed complexes. Due to the property of the α-hydroxycarboxylic acid to offer multiple binding sites, particularly stable complex compounds (chelate complexes) can be formed.

[0367] For batches C, D, E and F, the polishing cloth was textured - i.e., a surface area of ​​7.1% was treated with a material removal rate of between 500 pm and 800 pm.

[0368] The wafers were planarized on both sides. A Speedfam 12B machine was used for this purpose.

[0369] For wafers with a diameter of 76.2 mm, 24 wafers were planarized simultaneously. For wafers with a diameter of 100 mm, 18 wafers were planarized simultaneously, and for wafers with a diameter of 150 mm, 6 wafers were planarized simultaneously.

[0370] The polishing was done with a pressure of approximately 100 g / cm². 2 0.95 liters of polishing compound were added per minute and flowed through the polishing machine. The upper disc rotated at 13 revolutions per minute, the lower disc rotated at

[0371] 40 revolutions per minute, the ring gear rotated at 9 revolutions per minute, and the planetary gear rotated at 15 revolutions per minute.

[0372] Approximately 25 pm of material was removed from each side of the wafer.

[0373] The liquid polishing compound (i.e., the polishing solution) was supplied via the machine's media supply. The carriers were placed symmetrically distributed into the polishing system and then lightly moistened with deionized water.

[0374] A maximum of 6 carriers with a different number of working holes, depending on the wafer diameter, could be inserted on a tool with an 830 mm disc diameter.

[0375] The wafers were placed face down into the carriers with the lower cloth facing down and rotated slightly. Once all wafers were in the polishing carriers, the upper polishing plate was lowered onto the lower polishing plate.

[0376] A corresponding program stored in the polishing system was started, taking into account the wafer thickness, minimum material removal and material removal rate.

[0377] After the polishing time had elapsed, the polished wafers were placed into a transport tray using vacuum tweezers.

[0378] A one-sided finalization step was then performed (the front side of the wafer was mirror-polished), although this did not change the flatness parameters. The wafers were then washed and rinsed in deionized water. Finally, they were dried using the spin-rinse-dry method. Flatness measurements were then taken.

[0379] Measurement methods

[0380] measuring device

[0381] The geometry measurements were performed using a Troptel® Ultrasort™ interferometer from Corning Tropei Corporation. The underlying physical principle is based on the interference of monochromatic light. Inside the interferometer, a helium-neon laser beam (with a wavelength of 321 nm) is split into a bundle of parallel beams. At the specially coated underside of the prism, these beams are split into a component reflected by the prism and a component reflected by the sample: the reference beam and the test beam. The test beam exits the prism and grazes the sample surface of the wafer at a shallow angle, is reflected from there, and re-enters the prism, where it interferes with the reference beam. This creates the characteristic interference rings, the spacing of which represents a measure of topographical differences on the wafer surface. Software analysis yielded the various geometric data. The accuracy of the instrument is 0.05 pm.

[0382] Geometry measurements

[0383] Definitions of wafer layers

[0384] The following are the relevant definitions for the relevant layers of a wafer:

[0385] Focal plane:

[0386] The focal plane is the plane defined by three points on the wafer's front surface. These three points are described as follows: For a wafer with radius R, the three points are located at a radius of 0.97 R and are 120° apart, with one point situated on the wafer's front surface (180° from the wafer's orientation feature, the main flat or notch). The focal plane can be determined for both aspirated and non-aspirated wafers. In each case, the resulting focal planes are typically different.

[0387] Shifted focal plane:

[0388] The shifted focal plane is the plane that results when the focal plane is shifted perpendicular to the back of the wafer such that the local average plane coincides with the shifted focal plane at the center of the measurement field.

[0389] Each measurement field has its own shifted focal plane.

[0390] The center point of the measuring field is the intersection of the diagonals of the measuring field.

[0391] The local best fit plane is defined by fitting a plane to the wafer front in the measurement field as follows:

[0392] A plane is defined by the equation z = A + Bx + Cy (Equation 1). The least squared plane fit is determined by minimizing the sum of the squares of the deviations of the real data from the plane. This method yields the least squares of A, B, and C.

[0393] The matrices are determined as follows:

[0394]

[0395] When this equation is solved for A, B and C, the least squares fitting is complete.

[0396] Average plane (best fit plane):

[0397] The average plane (best fit plane) of the wafer front is defined by fitting a plane to the wafer front as follows:

[0398] A plane is defined by the equation z = A + Bx + Cy (Equation 1 above). The least squared plane fit is determined by minimizing the sum of the squares of the deviations of the real data from the plane. This method yields the least squares of A, B, and C. The matrices are determined as follows:

[0399] When the equation for A, B and C is solved, the least squares fitting is complete.

[0400] Specific levels are used as a basis for each parameter (see Table 2):

[0401] Parameter Underlying Plane Wafer Back TTV None Suctioned Taper Average Plane (best fit plane) Suctioned TIR Focal Plane Suctioned FPD Focal Plane Suctioned LTV None Suctioned LTIR Shifted Focal Plane Suctioned LFPD Shifted Focal Plane Suctioned LFPD20x20 Shifted Focal Plane Suctioned Warp Focal Plane Free

[0402]

[0403] Bow Focal Plane Free | Sori _ | _ Average Plane (best fit plane) _ | Free _ |

[0404] Table 2: Measured parameters and underlying layers, as well as the status of the wafer backside during measurement

[0405] Global flatness parameters

[0406] Global Thickness Variation (TW)

[0407] All measurements are performed using a common reference surface. This is defined by the surface of the chuck (i.e., the clamping device) and, due to the influence of the vacuum, is identical to the back side of the wafer. The wafer is held against the chuck by vacuum suction.

[0408] The global thickness variation is the difference between the highest and lowest elevation of the front surface of the sample relative to the rear surface.

[0409] To determine the global thickness variation, an edge exclusion of 3 mm is used, i.e., the area of ​​the wafer that is 3 mm or less from the outermost edge of the wafer is not taken into account.

[0410] Taper

[0411] All measurements are performed using a common reference surface.

[0412] The surface of the chuck (the plate on which the wafer is mounted for measurement) is polished to be perfectly flat. The chuck on which the wafer rests is therefore considered ideally flat. However, it can be tilted, or inclined, relative to the horizontal. This means that only the inclination or taper of the chuck relative to the horizontal needs to be known to correctly align the front and back surfaces of the wafer. The UltraSort calibrates this automatically by performing a wafer reversal during a measurement. In this process, the wafer is measured in its standard orientation and then rotated 180°. The average taper (or inclination) of these two measurements is the taper of the chuck, i.e., the narrowing between the surface of the chuck and the prism of the measuring instrument.

[0413] The taper is the deviation in parallelism between the back of the wafer and the best-fit plane of the front. The specified numerical value is the amount of the rise in the best-fit plane across the entire surface of the wafer.

[0414] To determine the taper, an edge exclusion of 3 mm is used, i.e., the area of ​​the wafer that is 3 mm or less from the outermost edge of the wafer is not taken into account.

[0415] The average plane (best fit plane) of the wafer front is defined by fitting a plane to the wafer front as follows:

[0416] A plane is defined by equation 1 (see above) z = A + Bx + Cy. The least squared plane fit is determined by minimizing the sum of the squares of the deviations of the real data from the plane. This method yields the least squares of A, B, and C.

[0417] The matrices are determined as follows:

[0418] When this equation is solved for A, B and C, the least squares fitting is complete.

[0419] TIR (Total Indicated Reading)

[0420] The TIR value describes the variation in the flatness of the wafer with respect to the focal plane.

[0421] In other words, the TIR value is the difference between the highest point above the focal plane and the lowest point below the focal plane of the wafer's front surface. The TIR value is always positive.

[0422] To determine the TIR value, an edge exclusion of 3 mm is used; that is, the area of ​​the wafer that is 3 mm or less from the outermost edge of the wafer is not considered. The wafer is measured in the vacuum-sealed state. Focal plane deviation - FPD (Focal Plane Deviation)

[0423] Focal plane deviation is the greater distance of the wafer front either above or below the focal plane. FPD values ​​are always positive.

[0424] To determine the focal plane deviation, a 3 mm edge exclusion is used; that is, the area of ​​the wafer that is 3 mm or less from the outermost edge of the wafer is not considered. The wafer is measured in the suction state.

[0425] Local

[0426]

[0427] Each of the measurement fields (e.g., 15 mm x 15 mm, 20 mm x 20 mm) yields a value for the respective measurement parameter. The highest value from each field is used as the characteristic value for the wafer. This applies to all measurement parameters that represent local flatness parameters. For example, the LFPD value is actually the maximum value of the LFPD values ​​of the individual measurement fields i = 1 ... n, where n represents the number of measurement fields.

[0428] Local Thickness Variation

[0429]

[0430] Local thickness variation is the distance between the highest and lowest points of the surface within a measurement field of a specific size (usually 15 mm x 15 mm). Distance refers to the magnitude of the component of the vector from the highest to the lowest point that is perpendicular to the wafer support. The component of the vector parallel to the wafer support is not considered here. The values ​​for local thickness variation discussed here refer to the maximum value of the local thickness variation for all measured 15 mm x 15 mm fields.

[0431] Measurements are evaluated across localized zones on the wafer to assess the wafer in each zone and to simulate the imaging process using a lithography stepper system. No back-tilting is performed, so the tapering is included in the measurement. The wafer is measured in the suction state.

[0432] LTIR (Local Total Indicated

[0433]

[0434] The LTIR value is the difference between the highest point above the shifted focal plane and the lowest point below the shifted focal plane for the measurement field. The numerical value for LTIR is always positive. The value discussed here is the highest value of all measured points.

[0435] The LTIR values ​​discussed here refer to the maximum LTIR value for all measured measurement fields of 15 mm x 15 mm.

[0436] The measurement is taken while the wafer is suctioned on.

[0437] The shifted focal plane is the plane that results when the focal plane is shifted perpendicular to the back of the wafer such that the local average plane coincides with the shifted focal plane at the center of the measurement field.

[0438] Each measurement field has its own shifted focal plane.

[0439] The center point of the measuring field is the intersection of the diagonals of the measuring field.

[0440] The local best fit plane is defined by fitting a plane to the wafer front in the measurement field as follows:

[0441] A plane is defined by equation 1 (see above) z = A + Bx + Cy. The least squared plane fit is determined by minimizing the sum of the squares of the deviations of the real data from the plane. This method yields the least squares of A, B, and C.

[0442] When this equation is solved for A, B and C, the least squares fitting is complete.

[0443] Local

[0444]

[0445] - LFPD (Local Focal Plane

[0446]

[0447] The local focal plane deviation is the largest distance from either above or below to the shifted focal plane in a measurement field of a specific size. The measurement is performed with the wafer held in a vacuum.

[0448] The LFPD values ​​discussed here refer to the maximum LFPD value for all measured measurement fields of 15 mm x 15 mm.

[0449] Local focal plane deviation with a measuring field of 20 mm x 20 mm - LFPD20x20

[0450] The values ​​discussed here for LFPD20x20 refer to a measurement area (i.e., a measurement field) of 20 mm x 20 mm. LFPD20x20 specifies the maximum value for all measurement fields considered on a wafer.

[0451] Bow deflection

[0452] Deflection is the difference between the focal plane and the surface height of the front face of the unsupported, or in other words, unclamped (free) sample at its center point. Deflection measurements can have positive or negative values ​​and are reported as such. If the center point of the wafer's front face lies below the focal plane, the measurement value is negative; if the center point lies above the focal plane, the value is positive.

[0453] The focal plane is the plane defined by three points on the wafer's front surface. These three points are described as follows: For a wafer with radius R, the three points lie at a radius of 0.97 R and are 120° apart, with one point located on the front of the wafer (180° from the wafer's orientation feature, the main flat or notch).

[0454] Curvature (Warp): The curvature is the sum of the maximum positive and the amount of the maximum negative deviation of the wafer front from the focal plane, measured in the non-suctioned state.

[0455] Curvature is the maximum distance between the highest and lowest points on the surface of an unclamped (free) sample from the focal plane. The value for curvature measurements is always positive.

[0456] Sori value

[0457] SORI is the maximum distance between the highest and lowest points on the surface of an unclamped (free) sample from the average (best-fit) plane of the wafer face. The value for SORI measurements is always positive.

[0458] The least squares flatness measurements performed without clamping all share a common reference surface, defined by the least squares fit to the unclamped wafer. The reference point is determined by subtracting the least squares plane (see below) from the data of an unclamped (free) wafer.

[0459] The least squares plane is defined as follows: A plane is determined by equation 1 (see above) z = A + Bx + Cy. The least squares fit of the plane is determined by minimizing the sum of the squares of the deviations of the real data from the plane. This method yields the least squares values ​​of A, B, and C.

[0460] Solving this equation for A, B, and C completes the least squares fitting.

[0461] Measurement results

[0462] Reference is made here to Figures 13, 14, and 15. The representation of the box plots is described below. The box contains the values ​​from the first to the third quartile of all measurements, excluding outliers. The marker inside the box indicates the mean. The horizontal lines (whiskers) above and below the box indicate the minimum and maximum values, excluding outliers. Outliers are shown as individual points above and below the horizontal lines (whiskers).

[0463] The following values ​​were determined:

[0464] • Maximum (highest value)

[0465] • Minimum (lowest value)

[0466] • Mean (average value)

[0467] • Median (the median divides the sample so that 50% of the data are below and 50% are above this value)

[0468] • Q1 - first quartile (the value below which 25% of the data lie)

[0469] • Q3 - third quartile (the value below which 75% of the data lie)

[0470] • Standard deviation (this indicates how far the individual values ​​deviate on average from the mean)

[0471] • Relative standard deviation with respect to the median (here the standard deviation is represented as a percentage of the median value; this measure describes the variability of the data with respect to the median)

[0472] The standard deviation of a series of values ​​x is calculated as follows, where x represents the mean of all values ​​x and n is the number of values ​​x:

[0473] Standard deviation =

[0474] IZ(x - x)2

[0475] y

[0476]

[0477] ] (n - 1)Local flatness parameters

[0478] LTV

[0479] Table 3 shows the measured LTV values ​​(local thickness variation). All values ​​are measured in pm; only the relative standard deviation with respect to the median is dimensionless. These are also shown in Fig. 14.

[0480] Batch

[0481] ABCDEF

[0482] Maximum 1.8010 1.0670 0.4950 0.4910 0.7540 0.442

[0483] Q3 1.6100 0.9543 0.4460 0.4528 0.6700 0.439 Mean 1.5338 0.8944 0.4116 0.4023 0.6534 0.429 Median 1.5035 0.8955 0.4120 0.3975 0.6540 0.430

[0484] Ql 1.4493 0.8193 0.3850 0.3745 0.6170 0.419 Minimum 1.2830 0.7450 0.3240 0.3040 0.5780 0.411 Standard deviation 0.1155 0.0882 0.0425 0.0525 0.0444 0.011

[0485] Relative standard deviation with respect to

[0486]

[0487] Median 0.0768 0.0985 0.1032 0.1320 0.0678 0.026 Table 3: Local flatness parameters (LTV) for 6 batches of InP wafers

[0488] For batch A (see Table 1 above), a maximum LTV value of 1.801 pm and a minimum LTV value of 1.283 pm were measured. The mean is 1.5338 pm, and the median is 1.5035 pm. The third quartile is 1.61 pm, and the first quartile is 1.4493 pm. The standard deviation is 0.1155 pm, and the relative standard deviation with respect to the median is 0.0768.

[0489] For batch B (see Table 1 above), a maximum LTV value of 1.067 pm and a minimum LTV value of 0.745 pm were measured. The mean is 0.8944 pm, and the median is 0.8955 pm. The third quartile is 0.9543 pm, and the first quartile is 0.81933 pm. The standard deviation is 0.0882 pm, and the relative standard deviation with respect to the median is 0.0985.

[0490] For batch C (see Table 1 above), a maximum LTV value of 0.495 pm and a minimum LTV value of 0.324 pm were measured. The mean is 0.4116 pm, and the median is 0.4120 pm. The third quartile is 0.446 pm, and the first quartile is 0.385 pm. The standard deviation is 0.0425 pm, and the relative standard deviation with respect to the median is 0.1032.

[0491] For batch D (see Table 1 above), a maximum LTV value of 0.491 pm and a minimum LTV value of 0.304 pm were measured. The mean is 0.4032 pm, and the median is 0.3975 pm. The third quartile is 0.4528 pm, and the first quartile is 0.3745 pm. The standard deviation is 0.0525 pm, and the relative standard deviation with respect to the median is 0.1320.

[0492] For batch E (see Table 1 above), a maximum LTV value of 0.754 pm and a minimum LTV value of 0.578 pm were measured. The mean is 0.6534 pm, and the median is 0.654 pm. The third quartile is 0.67 pm, and the first quartile is 0.617 pm. The standard deviation is 0.0444 pm, and the relative standard deviation with respect to the median is 0.0678.

[0493] For batch F (see Table 1 above), a maximum LTV value of 0.442 pm and a minimum LTV value of 0.439 pm were measured. The mean is 0.429 pm, and the median is 0.430 pm. The third quartile is 0.439 pm, and the first quartile is 0.419 pm. The standard deviation is 0.011 pm, and the relative standard deviation with respect to the median is 0.026.

[0494] LTIR

[0495] Table 4 shows the measured LTIR (Local Total Indicated Reading) values. All values ​​are measured in pm; only the relative standard deviation with respect to the median is dimensionless. These are also shown in Fig. 14.

[0496] Batch

[0497] ABCDEF

[0498] Maximum 0.9270 0.6600 0.4800 0.3730 0.4060 0.407

[0499] Q3 0.7813 0.6225 0.3580 0.3388 0.3908 0.355 Mean 0.7535 0.5732 0.3363 0.3133 0.3699 0.316 Median 0.7375 0.5715 0.3250 0.3125 0.3675 0.296

[0500] Ql 0.7075 0.5270 0.3020 0.2930 0.3590 0.281 Minimum 0.6520 0.4780 0.2360 0.2510 0.3340 0.279 Standard deviation 0.0680 0.0521 0.0522 0.0339 0.0203 0.050

[0501] Relative standard deviation with respect to

[0502]

[0503] Median 0.0922 0.0911 0.1606 0.1084 0.0552 0.168 Table 4: Local flatness parameters (LTIR) for 6 batches of InP wafers

[0504] For batch A, a maximum LTIR value of 0.9270 pm and a minimum LTIR value of 0.6520 pm were measured. The mean is 0.7535 pm, and the median is 0.7375 pm. The third quartile is 0.7813 pm, and the first quartile is 0.7075 pm. The standard deviation is 0.0680 pm, and the relative standard deviation with respect to the median is 0.0922.

[0505] For batch B, a maximum LTIR value of 0.6600 pm and a minimum LTIR value of 0.4780 pm were measured. The mean is 0.5732 pm, and the median is 0.5715 pm. The third quartile is 0.6225 pm, and the first quartile is 0.5270 pm. The standard deviation is 0.0521 pm, and the relative standard deviation with respect to the median is 0.0911.

[0506] For batch C, a maximum LTIR value of 0.4800 pm and a minimum LTIR value of 0.2360 pm were measured. The mean is 0.3363 pm, and the median is 0.3250 pm. The third quartile is 0.3580 pm, and the first quartile is 0.3020 pm. The standard deviation is 0.0522 pm, and the relative standard deviation with respect to the median is 0.1606.

[0507] For batch D, a maximum LTIR value of 0.3730 pm and a minimum LTIR value of 0.2510 pm were measured. The mean is 0.3133 pm, and the median is 0.3125 pm. The third quartile is 0.3388 pm, and the first quartile is 0.2930 pm. The standard deviation is 0.0339 pm, and the relative standard deviation with respect to the median is 0.1084.

[0508] For batch E, a maximum LTIR value of 0.4060 pm and a minimum LTIR value of 0.3340 pm were measured. The mean is 0.3699 pm, and the median is 0.3675 pm. The third quartile is 0.3908 pm, and the first quartile is 0.3590 pm. The standard deviation is 0.0203 pm, and the relative standard deviation with respect to the median is 0.0552. For batch F, a maximum LTIR value of 0.407 pm and a minimum LTIR value of 0.279 pm were measured. The mean is 0.316 pm, and the median is 0.296 pm. The third quartile is 0.355 pm, and the first quartile is 0.281 pm. The standard deviation is 0.050 pm, the relative standard deviation with respect to the median is 0.168.

[0509] LFPD

[0510] Table 5 shows the measured LFPD values ​​(Local Focal Plane Deviation). All values ​​are measured in pm; only the relative standard deviation with respect to the median is dimensionless. These are also shown in Fig. 14.

[0511] Batch

[0512] ABCDEF

[0513] Maximum 0.7010 0.5440 0.3070 0.2750 0.3130 0.235

[0514] Q3 0.6125 0.4700 0.2580 0.2553 0.2838 0.229 Mean 0.5846 0.4454 0.2376 0.2295 0.2702 0.199 Median 0.5805 0.4405 0.2310 0.2250 0.2690 0.198

[0515] Ql 0.5518 0.4033 0.2180 0.2043 0.2630 0.172 Minimum 0.4990 0.3800 0.1760 0.1910 0.2290 0.158 Standard deviation 0.0493 0.0421 0.0314 0.0275 0.0227 0.030

[0516] Relative standard deviation with respect to

[0517]

[0518] Median 0.0849 0.0955 0.1359 0.1221 0.0843 0.151 Table 5: Local flatness parameters (LFPD) for 6 batches of InP wafers

[0519] For batch A, a maximum LFPD value of 0.7010 pm and a minimum LFPD value of 0.4990 pm were measured. The mean is 0.5846 pm, and the median is 0.5805 pm. The third quartile is 0.6125 pm, and the first quartile is 0.5518 pm. The standard deviation is 0.0493 pm, and the relative standard deviation with respect to the median is 0.0849.

[0520] For batch B, a maximum LFPD value of 0.5440 pm and a minimum LFPD value of 0.3800 pm were measured. The mean is 0.4454 pm, and the median is 0.4405 pm. The third quartile is 0.4700 pm, and the first quartile is 0.4033 pm. The standard deviation is 0.0421 pm, and the relative standard deviation with respect to the median is 0.0955.

[0521] For batch C, a maximum LFPD value of 0.3070 pm and a minimum LFPD value of 0.1760 pm were measured. The mean is 0.2376 pm, and the median is 0.2310 pm. The third quartile is 0.2580 pm, and the first quartile is 0.2180 pm. The standard deviation is 0.0314 pm, and the relative standard deviation with respect to the median is 0.1359.

[0522] For batch D, a maximum LFPD value of 0.2750 pm and a minimum LFPD value of 0.1910 pm were measured. The mean is 0.2295 pm, and the median is 0.2250 pm. The third quartile is 0.2553 pm, and the first quartile is 0.2043 pm. The standard deviation is 0.0275 pm, and the relative standard deviation with respect to the median is 0.1221.

[0523] For batch E, a maximum LFPD value of 0.3130 pm and a minimum LFPD value of 0.2290 pm were measured. The mean is 0.2702 pm, and the median is 0.2690 pm. The third quartile is 0.2838 pm, and the first quartile is 0.2630 pm. The standard deviation is 0.0227 pm, and the relative standard deviation with respect to the median is 0.0843.

[0524] For batch F, a maximum LFPD value of 0.235 pm and a minimum LFPD value of 0.158 pm were measured. The mean is 0.199 pm, and the median is 0.198 pm. The third quartile is 0.229 pm, and the first quartile is 0.172 pm. The standard deviation is 0.030 pm, and the relative standard deviation with respect to the median is 0.151.

[0525] LFPD20x20

[0526] Table 6 shows the measured LFPD20x20 values ​​(Local Focal Plane Deviation with a measurement field of 20 mm x 20 mm). All values ​​are measured in pm; only the relative standard deviation with respect to the median is dimensionless. They are also shown in Fig. 14.

[0527] ABCDEF

[0528] Maximum 0.7530 0.3670 0.2850 0.2290 0.3050 0.251

[0529] Q3 0.6895 0.2948 0.2080 0.1948 0.2778 0.244 Mean 0.6371 0.2787 0.1843 0.1714 0.2581 0.217 Median 0.6315 0.2650 0.1780 0.1635 0.2550 0.222

[0530] Ql 0.6013 0.2478 0.1550 0.1503 0.2390 0.187 Minimum 0.5190 0.2250 0.1300 0.1160 0.2220 0.176 Standard deviation 0.0602 0.0419 0.0398 0.0300 0.0242 0.030

[0531] Relative standard deviation with respect to

[0532]

[0533] Median 0.0953 0.1580 0.2237 0.1837 0.0948 0.137 Table 6: Local flatness parameters (LFPD20x20) for 6 batches of InP wafers

[0534] For batch A, a maximum LFPD value of 0.7530 pm was measured with a 20 x 20 mm measurement area, and a minimum LFPD value of 0.5190 pm with a 20 x 20 mm measurement area. The mean is 0.6371 pm, and the median is 0.6315 pm. The third quartile is 0.6895 pm, and the first quartile is 0.6013 pm. The standard deviation is 0.0602 pm, and the relative standard deviation with respect to the median is 0.0953.

[0535] For batch B, a maximum LFPD value of 0.3670 pm was measured with a 20 x 20 mm measurement area, and a minimum LFPD value of 0.2250 pm was measured with a 20 x 20 mm measurement area. The mean is 0.2787 pm, and the median is 0.2650 pm. The third quartile is 0.2948 pm, and the first quartile is 0.2478 pm. The standard deviation is 0.0419 pm, and the relative standard deviation with respect to the median is 0.1580.

[0536] For batch C, a maximum LFPD value of 0.2850 pm was measured with a 20 x 20 mm measurement area, and a minimum LFPD value of 0.1300 pm was measured with a 20 x 20 mm measurement area. The mean is 0.1843 pm, and the median is 0.1780 pm. The third quartile is 0.2080 pm, and the first quartile is 0.1550 pm. The standard deviation is 0.0398 pm, and the relative standard deviation with respect to the median is 0.2237. For batch D, a maximum LFPD value of 0.2290 pm was measured with a 20 x 20 mm measurement area, and a minimum LFPD value of 0.1160 pm was measured with a 20 x 20 mm measurement area. The mean is 0.1714 pm, the median is 0.1635 pm. The third quartile is 0.1948 pm, the first quartile is 0.1503 pm. The standard deviation is 0.0300 pm, the relative standard deviation with respect to the median is 0.1837.

[0537] For batch E, a maximum LFPD value of 0.3050 pm was measured with a 20 x 20 mm measurement area, and a minimum LFPD value of 0.2220 pm was measured with a 20 x 20 mm measurement area. The mean is 0.2581 pm, and the median is 0.2550 pm. The third quartile is 0.2778 pm, and the first quartile is 0.2390 pm. The standard deviation is 0.0242 pm, and the relative standard deviation with respect to the median is 0.0948.

[0538] For batch F, a maximum LFPD value of 0.251 pm was measured with a 20 x 20 mm measurement field, and a minimum LFPD value of 0.176 pm was measured with a 20 x 20 mm measurement field. The mean is 0.217 pm, and the median is 0.222 pm. The third quartile is 0.244 pm, and the first quartile is 0.187 pm. The standard deviation is 0.030 pm, and the relative standard deviation with respect to the median is 0.137.

[0539] Global flatness parameters

[0540] TTV (Total Thickness Variation)

[0541] Table 7 shows the measured TTV values ​​(global thickness variation). All values ​​are measured in pm; only the relative standard deviation with respect to the median is dimensionless. These are also shown in Fig. 13.

[0542] Batch

[0543] ABCDEF

[0544] Maximum 2.6670 1.4190 0.9530 0.9640 1.0470 1.014

[0545]

[0546] Q3 2.4883 1.2183 0.7950 0.8208 0.9850 1.009Mean 2.3205 1.1893 0.7145 0.7602 0.9299 0.891 Median 2.3170 1.1875 0.6970 0.7850 0.9365 0.896

[0547] Ql 2,2018 1.1488 0.6510 0.6775 0.8738 0.781 Minimum 1.9680 1.0290 0.5580 0.5350 0.8100 0.743 Standard deviation 0.1940 0.0839 0.1040 0.1034 0.0681 0.111

[0548] Relative standard deviation with respect to

[0549]

[0550] Median 0.0837 0.0706 0.1493 0.1317 0.0727 0.124 Table 7: Global flatness parameters (TTV) for 6 batches of InP wafers

[0551] For batch A, a maximum TTV value of 2.6670 pm and a minimum TTV value of 1.9680 pm were measured. The mean is 2.3205 pm, and the median is 2.3170 pm. The third quartile is 2.4883 pm, and the first quartile is 2.2018 pm. The standard deviation is 0.1940 pm, and the relative standard deviation with respect to the median is 0.0837.

[0552] For batch B, a maximum TTV value of 1.4190 pm and a minimum TTV value of 1.0290 pm were measured. The mean is 1.1893 pm, and the median is 1.1875 pm. The third quartile is 1.2183 pm, and the first quartile is 1.1488 pm. The standard deviation is 0.0839 pm, and the relative standard deviation with respect to the median is 0.0706.

[0553] For batch C, a maximum TTV value of 0.9530 pm and a minimum TTV value of 0.5580 pm were measured. The mean is 0.7145 pm, and the median is 0.6970 pm. The third quartile is 0.7950 pm, and the first quartile is 0.6510 pm. The standard deviation is 0.1040 pm, and the relative standard deviation with respect to the median is 0.1493.

[0554] For Batch D, a maximum TTV value of 0.9640 pm and a minimum TTV value of 0.5350 pm were measured. The mean is 0.7602 pm, and the median is 0.7850 pm. The third quartile is 0.8208 pm, and the first quartile is 0.6775 pm. The standard deviation is 0.1034 pm, and the relative standard deviation with respect to the median is 0.1317. For Batch E, a maximum TTV value of 1.0470 pm and a minimum TTV value of 0.8100 pm were measured. The mean is 0.9299 pm, and the median is 0.9365 pm. The third quartile is 0.9850 pm, and the first quartile is 0.8738 pm. The standard deviation is 0.0681 pm, the relative standard deviation with respect to the median is 0.0727.

[0555] For batch F, a maximum TTV value of 1.014 pm and a minimum TTV value of 0.743 pm were measured. The mean is 0.891 pm, and the median is 0.896 pm. The third quartile is 1.009 pm, and the first quartile is 0.781 pm. The standard deviation is 0.111 pm, and the relative standard deviation with respect to the median is 0.124.

[0556] TIR (Total Indicated Reading)

[0557] Table 8 shows the measured TIR values. All values ​​are measured in pm; only the relative standard deviation with respect to the median is dimensionless. These are also shown in Fig. 13.

[0558] Batch

[0559] ABCDEF

[0560] Maximum 2.5850 1.5350 0.9840 0.7450 1.0280 1.125

[0561] Q3 2.3603 1.3145 0.7980 0.6688 0.9428 1.046 Mean 2.2247 1.2577 0.7299 0.6271 0.9198 0.936 Median 2.2310 1.2470 0.7120 0.6310 0.9150 0.918

[0562] Ql 2.0713 1.2003 0.6300 0.5950 0.8848 0.829 Minimum 1.8880 1.0270 0.6020 0.4910 0.8490 0.795 Standard deviation 0.1819 0.1018 0.0978 0.0627 0.0439 0.121

[0563] Relative standard deviation with respect to

[0564]

[0565] Median 0.0815 0.0816 0.1374 0.0994 0.0479 0.132 Table 8: Global flatness parameters (TIR) ​​for 6 batches of InP wafers

[0566] For batch A, a maximum TIR value of 2.5850 pm and a minimum TIR value of 1.8880 pm were measured. The mean is 2.2247 pm, and the median is 2.2310 pm. The third quartile is 2.3603 pm, and the first quartile is 2.0713 pm. The standard deviation is 0.1819 pm, and the relative standard deviation with respect to the median is 0.0815.

[0567] For batch B, a maximum TIR value of 1.5350 pm and a minimum TIR value of 1.0270 pm were measured. The mean is 1.2577 pm, and the median is 1.2470 pm. The third quartile is 1.3145 pm, and the first quartile is 1.2003 pm. The standard deviation is 0.1018 pm, and the relative standard deviation with respect to the median is 0.0816.

[0568] For batch C, a maximum TIR value of 0.9840 pm and a minimum TIR value of 0.6020 pm were measured. The mean is 0.7299 pm, and the median is 0.7120 pm. The third quartile is 0.7980 pm, and the first quartile is 0.6300 pm. The standard deviation is 0.0978 pm, and the relative standard deviation with respect to the median is 0.1374.

[0569] For batch D, a maximum TIR value of 0.7450 pm and a minimum TIR value of 0.4910 pm were measured. The mean is 0.6271 pm, and the median is 0.6310 pm. The third quartile is 0.6688 pm, and the first quartile is 0.5950 pm. The standard deviation is 0.0627 pm, and the relative standard deviation with respect to the median is 0.0994.

[0570] For batch E, a maximum TIR value of 1.0280 pm and a minimum TIR value of 0.8490 pm were measured. The mean is 0.9198 pm, and the median is 0.9150 pm. The third quartile is 0.9428 pm, and the first quartile is 0.8848 pm. The standard deviation is 0.0439 pm, and the relative standard deviation with respect to the median is 0.0479.

[0571] For batch F, a maximum TIR value of 1.125 pm and a minimum TIR value of 0.795 pm were measured. The mean is 0.9360 pm, and the median is 0.918 pm. The third quartile is 1.046 pm, and the first quartile is 0.829 pm. The standard deviation is 0.121 pm, and the relative standard deviation with respect to the median is 0.132. FPD (Focal Plane Deviation)

[0572] Table 9 shows the measured FPD values ​​(Global Focal Plane Deviation). All values ​​are measured in pm; only the relative standard deviation with respect to the median is dimensionless. These are also shown in Fig. 13.

[0573] Batch

[0574] ABCDEF

[0575] Maximum 1.7210 0.9960 0.6490 0.5290 0.5930 0.628

[0576] Q3 1.5648 0.7833 0.5330 0.4830 0.5408 0.621 Mean 1.4609 0.7207 0.5024 0.4410 0.5066 0.547 Median 1.4500 0.7205 0.5120 0.4465 0.4965 0.560

[0577] Ql 1.3893 0.6413 0.4320 0.4055 0.4648 0.483 Minimum 1.2060 0.5190 0.3780 0.3410 0.4560 0.411 Standard deviation 0.1297 0.1077 0.0676 0.0504 0.0436 0.083

[0578] Relative standard deviation with respect to

[0579]

[0580] Median 0.0895 0.1495 0.1320 0.1128 0.0879 0.148 Table 9: Global flatness parameters (FPD) for 6 batches of InP wafers

[0581] For batch A, a maximum FPD value of 1.7210 pm and a minimum FPD value of 1.2060 pm were measured. The mean is 1.4609 pm, and the median is 1.4500 pm. The third quartile is 1.5648 pm, and the first quartile is 1.3893 pm. The standard deviation is 0.1297 pm, and the relative standard deviation with respect to the median is 0.0895.

[0582] For batch B, a maximum FPD value of 0.9960 pm and a minimum FPD value of 0.5190 pm were measured. The mean is 0.7207 pm, and the median is 0.7205 pm. The third quartile is 0.7833 pm, and the first quartile is 0.6413 pm. The standard deviation is 0.1077 pm, and the relative standard deviation with respect to the median is 0.1495.

[0583] For batch C, a maximum FPD value of 0.6490 pm and a minimum FPD value of 0.3780 pm were measured. The mean is 0.5024 pm, and the median is 0.5120 pm. The third quartile is 0.5330 pm, and the first quartile is 0.4320 pm. The standard deviation is 0.0676 pm, and the relative standard deviation with respect to the median is 0.1320.

[0584] For batch D, a maximum FPD value of 0.5290 pm and a minimum FPD value of 0.3410 pm were measured. The mean is 0.4410 pm, and the median is 0.4465 pm. The third quartile is 0.4830 pm, and the first quartile is 0.4055 pm. The standard deviation is 0.0504 pm, and the relative standard deviation with respect to the median is 0.1128.

[0585] For batch E, a maximum FPD value of 0.5930 pm and a minimum FPD value of 0.4560 pm were measured. The mean is 0.5066 pm, and the median is 0.4965 pm. The third quartile is 0.5408 pm, and the first quartile is 0.4648 pm. The standard deviation is 0.0436 pm, and the relative standard deviation with respect to the median is 0.0879.

[0586] For batch F, a maximum FPD value of 0.628 pm and a minimum FPD value of 0.411 pm were measured. The mean is 0.547 pm, and the median is 0.560 pm. The third quartile is 0.621 pm, and the first quartile is 0.483 pm. The standard deviation is 0.083 pm, and the relative standard deviation with respect to the median is 0.148.

[0587] Taper

[0588] Table 10 shows the measured taper values ​​(slope). All values ​​are measured in pm; only the relative standard deviation with respect to the median is dimensionless. These are also shown in Fig. 13.

[0589] Batch

[0590] ABCDEF

[0591] Maximum 0.8000 0.5880 0.5540 0.4290 0.4300 0.510

[0592] Q3 0.7015 0.5018 0.2870 0.3468 0.4080 0.499 Mean 0.5810 0.4413 0.2307 0.2805 0.3706 0.263 Median 0.6285 0.4430 0.1890 0.2775 0.3775 0.162

[0593]

[0594] Ql 0.4973 0.3803 0.1370 0.2445 0.3390 0.136 Minimum 0.0380 0.2860 0.0330 0.0790 0.3050 0.104 Standard deviation 0.1736 0.0818 0.1336 0.0816 0.0396 0.187 Relative standard deviation with respect to

[0595]

[0596] Median 0.2762 0.1847 0.7068 0.2941 0.1048 1.158 Table 10: Global flatness parameters (taper) for 6 batches of InP wafers

[0597] For batch A, a maximum taper value of 0.8000 pm and a minimum taper value of 0.0380 pm were measured. The mean is 0.5810 pm, and the median is 0.6285 pm. The third quartile is 0.7015 pm, and the first quartile is 0.4973 pm. The standard deviation is 0.1736 pm, and the relative standard deviation with respect to the median is 0.2762.

[0598] For batch B, a maximum taper value of 0.5880 pm and a minimum taper value of 0.2860 pm were measured. The mean is 0.4413 pm, and the median is 0.4430 pm. The third quartile is 0.5018 pm, and the first quartile is 0.3803 pm. The standard deviation is 0.0818 pm, and the relative standard deviation with respect to the median is 0.1847.

[0599] For batch C, a maximum taper value of 0.5540 pm and a minimum taper value of 0.0330 pm were measured. The mean is 0.2307 pm, and the median is 0.1890 pm. The third quartile is 0.2870 pm, and the first quartile is 0.1370 pm. The standard deviation is 0.1336 pm, and the relative standard deviation with respect to the median is 0.7068.

[0600] For batch D, a maximum taper value of 0.4290 pm and a minimum taper value of 0.0790 pm were measured. The mean is 0.2805 pm, and the median is 0.2775 pm. The third quartile is 0.3468 pm, and the first quartile is 0.2445 pm. The standard deviation is 0.0816 pm, and the relative standard deviation with respect to the median is 0.2941.

[0601] For batch E, a maximum taper value of 0.4300 pm and a minimum taper value of 0.3050 pm were measured. The mean is 0.3706 pm, and the median is 0.3775 pm. The third quartile is 0.4080 pm, and the first quartile is 0.3390 pm. The standard deviation is 0.0396 pm, and the relative standard deviation with respect to the median is 0.1048.

[0602] For batch F, a maximum taper value of 0.510 pm and a minimum taper value of 0.104 pm were measured. The mean is 0.263 pm, and the median is 0.162 pm. The third quartile is 0.499 pm, and the first quartile is 0.136 pm. The standard deviation is 0.187 pm, and the relative standard deviation with respect to the median is 1.158.

[0603] Bow, Warp and Sori

[0604] Bow, warp, and sori also represent global flatness parameters that are measured on the wafer in the un-suctioned state. The wafers can therefore relax, and residual stresses lead to deformation of the wafer compared to its shape in the suctioned state.

[0605] Since the wafers are pressed onto a flat surface during the planarization process, these parameters cannot be significantly influenced by the planarization process, but are essentially determined by the upstream processes.

[0606] Warp

[0607] Table 11 shows the measured warp values. All values ​​are measured in pm; only the relative standard deviation with respect to the median is dimensionless. These are also shown in Fig. 15.

[0608] Batch

[0609] ABCDE

[0610] Maximum 3.5510 2.6870 2.2110 5.1360 5.8570

[0611] Q3 3.2883 2.3960 1.7450 3.7028 1.7700

[0612] Average 3.0208 2.1557 1.4820 2.9568 1.9437

[0613] Median 2.9970 2.1555 1.3380 2.5925 1.4970

[0614] Ql 2.7903 2.0813 1.2280 2.1065 1.3008

[0615]

[0616] Minimum 2.3850 1.2250 0.9170 1.7120 1.1440 Standard deviation 0.3223 0.3810 0.3564 1.0232 1.3781 Relative standard deviation with respect to

[0617]

[0618] Median 0.1075 0.1768 0.2663 0.3947 0.9205

[0619] Table 11: Warp values ​​for 5 batches of InP wafers

[0620] For Batch A, a maximum warp value of 3.5510 pm and a minimum warp value of 2.3850 pm were measured. The mean is 3.0208 pm, and the median is 2.9970 pm. The third quartile is 3.2883 pm, and the first quartile is 2.7903 pm. The standard deviation is 0.3223 pm, and the relative standard deviation with respect to the median is 0.1075.

[0621] For Batch B, a maximum warp value of 2.6870 pm and a minimum warp value of 1.2250 pm were measured. The mean is 2.1557 pm, and the median is 2.1555 pm. The third quartile is 2.3960 pm, and the first quartile is 2.0813 pm. The standard deviation is 0.3810 pm, and the relative standard deviation with respect to the median is 0.1768.

[0622] For batch C, a maximum warp value of 2.2110 pm and a minimum warp value of 0.9170 pm were measured. The mean is 1.4820 pm, and the median is 1.3380 pm. The third quartile is 1.7450 pm, and the first quartile is 1.2280 pm. The standard deviation is 0.3564 pm, and the relative standard deviation with respect to the median is 0.2663.

[0623] For batch D, a maximum warp value of 5.1360 pm and a minimum warp value of 1.7120 pm were measured. The mean is 2.9568 pm, and the median is 2.5925 pm. The third quartile is 3.7028 pm, and the first quartile is 2.1065 pm. The standard deviation is 1.0232 pm, and the relative standard deviation with respect to the median is 0.3947.

[0624] For Batch E, a maximum warp value of 5.8570 pm and a minimum warp value of 1.1440 pm were measured. The mean is 1.9437 pm, and the median is 1.4970 pm. The third quartile is 1.7700 pm, and the first quartile is 1.3008 pm. The standard deviation is 1.3781 pm, and the relative standard deviation with respect to the median is 0.9205.

[0625] Bow

[0626] Table 12 shows the measured Bow values. All values ​​are measured in pm; only the relative standard deviation with respect to the median is dimensionless. These are also shown in Fig. 15.

[0627] Batch

[0628] ABCDE

[0629] Maximum 1.7770 1.4080 1.4950 1.1070 0.2920

[0630] Q3 1.5793 0.9885 0.5920 0.4023 0.1665

[0631] Average 1.2875 0.4691 0.2313 0.1835 -0.0230

[0632] Median 1.3565 0.7050 0.2180 0.2340 0.1005

[0633] Ql 0.9640 -0.4250 -0.1360 -0.2058 -0.0220

[0634] Minimum 0.6050 -0.6360 -0.4810 -0.5990 -1.0220 Standard deviation 0.3553 0.7013 0.4805 0.4756 0.3629

[0635] Relative standard deviation with respect to

[0636] Median 0.2619 0.9948 2.2043 2.0323 3.6112

[0637]

[0638] Table 12: Bow values ​​for 5 batches of InP wafers

[0639] For batch A, a maximum Bow value of 1.7770 pm and a minimum Bow value of 0.6050 pm were measured. The mean is 1.2875 pm, and the median is 1.3565 pm. The third quartile is 1.5793 pm, and the first quartile is 0.9640 pm. The standard deviation is 0.3553 pm, and the relative standard deviation with respect to the median is 0.2619.

[0640] For Batch B, a maximum Bow value of 1.4080 pm and a minimum Bow value of -0.6360 pm were measured. The mean is 0.4691 pm, and the median is 0.7050 pm. The third quartile is 0.9885 pm, and the first quartile is -0.4250 pm. The standard deviation is 0.7013 pm, and the relative standard deviation with respect to the median is 0.9948. For Batch C, a maximum Bow value of 1.4950 pm and a minimum Bow value of -0.4810 pm were measured. The mean is 0.2313 pm, and the median is 0.2180 pm. The third quartile is 0.5920 pm, and the first quartile is -0.1360 pm. The standard deviation is 0.4805 pm, the relative standard deviation with respect to the median is 2.2043.

[0641] For batch D, a maximum Bow value of 1.1070 pm and a minimum Bow value of -0.5990 pm were measured. The mean is 0.1835 pm, and the median is 0.2340 pm. The third quartile is 0.4023 pm, and the first quartile is -0.2058 pm. The standard deviation is 0.4756 pm, and the relative standard deviation with respect to the median is 2.0323.

[0642] For batch E, a maximum Bow value of 0.2920 pm and a minimum Bow value of -1.0220 pm were measured. The mean is -0.0230 pm, and the median is 0.1005 pm. The third quartile is 0.1665 pm, and the first quartile is -0.0220 pm. The standard deviation is 0.3629 pm, and the relative standard deviation with respect to the median is 3.6112.

[0643] Sori

[0644] Table 13 shows the measured Sori values. All values ​​are measured in pm; only the relative standard deviation with respect to the median is dimensionless. These are also shown in Fig. 15.

[0645] Batch

[0646] ABCD

[0647] Maximum 4.2610 3.0910 2.2630 3.8060

[0648] Q3 3.7575 2.7950 1.5980 2.9778

[0649] Average 3.4165 2.3851 1.3864 2.5320

[0650] Median 3.4000 2.4380 1.3400 2.3830

[0651] Ql 2.9670 2.2383 1.0930 2.0693

[0652] Minimum 2.7940 1.0740 0.9010 1.9410

[0653] Standard deviation 0.4312 0.5312 0.3552 0.5564

[0654] Relative standard deviation with respect to

[0655]

[0656] Median 0.1268 0.2179 0.2651 0.2335 Table 13: Sori values ​​for 4 batches of InP wafers

[0657] For batch A, a maximum Sori value of 4.2610 pm and a minimum Sori value of 2.7940 pm were measured. The mean is 3.4165 pm, and the median is 3.4000 pm. The third quartile is 3.7575 pm, and the first quartile is 2.9670 pm. The standard deviation is 0.4312 pm, and the relative standard deviation with respect to the median is 0.1268.

[0658] For batch B, a maximum Sori value of 3.0910 pm and a minimum Sori value of 1.0740 pm were measured. The mean is 2.3851 pm, and the median is 2.4380 pm. The third quartile is 2.7950 pm, and the first quartile is 2.2383 pm. The standard deviation is 0.5312 pm, and the relative standard deviation with respect to the median is 0.2179.

[0659] For batch C, a maximum Sori value of 2.2630 pm and a minimum Sori value of 0.9010 pm were measured. The mean is 1.3864 pm, and the median is 1.3400 pm. The third quartile is 1.5980 pm, and the first quartile is 1.0930 pm. The standard deviation is 0.3552 pm, and the relative standard deviation with respect to the median is 0.2651.

[0660] For batch D, a maximum sori value of 3.8060 pm and a minimum sori value of 1.9410 pm were measured. The mean is 2.5320 pm, and the median is 2.3830 pm. The third quartile is 2.9778 pm, and the first quartile is 2.0693 pm. The standard deviation is 0.5564 pm, and the relative standard deviation with respect to the median is 0.2335.

[0661] W Wafer

[0662] M Wafer center

[0663] F measuring field

[0664] 1 Reference surface

[0665] 2 highest elevation

[0666] 3 lowest survey

[0667] 4 Average level (Best-Fit level)

[0668] 5 largest gap above average level

[0669] 6 largest gap below average level

[0670] 7 Focal plane

[0671] 7a shifted focal plane

[0672] 8 greatest distance across focal plane

[0673] 8a greatest distance above shifted focal plane 9 greatest distance below focal plane

[0674] 9a greatest distance under shifted focal plane 10 Multiwafer polishing machine

[0675] 11 outer gear ring

[0676] 12 inner gear ring

[0677] 13 carriers

[0678] 14 upper polishing plate

[0679] 15 lower polishing plate

[0680] 16 Polishing compound distribution device

[0681] 17 Polishing agent inlet

[0682] 18 Pressure roller

[0683] 19 Printing device (printing cylinder) 20 Polishing cloth

[0684] 21 working disc

[0685] 22 Drive pulley

Claims

REQUIREMENTS 1. Method for polishing III / V semiconductor wafers (W), in which the semiconductor wafer (W) is brought into contact with a polishing cloth (P) with at least one side to be polished, wherein the semiconductor wafer (W) and the polishing cloth (P) perform a rotational movement relative to each other, wherein a polishing agent (M) is added between the wafer (W) and the polishing cloth (P), comprising: deionized water, particles, and at least one α-hydroxycarboxylic acid.

2. Method according to claim 1, wherein the α-hydroxycarboxylic acid is selected from the group consisting of citric acid, tartaric acid, glycolic acid, lactic acid, malic acid and mandelic acid.

3. Method according to claim 1 or 2, wherein the polishing agent contains between 0.1 vol% and 1.0 vol% α-hydroxycarboxylic acid, preferably at least 0.4 vol%.

4. Method according to one of the preceding claims, wherein the polishing agent contains between 1 vol.% and 10 vol.% particles, preferably between 6 vol.% and 8 vol.% particles, and / or wherein the particles contain SiO2, TiO2, Al2O3, MnÜ2 and / or CeO2, preferably SiO2.

5. A method according to any of the preceding claims, wherein the main fraction of particles lies in a range of 10 nm to 30 nm particle diameter, wherein the specific surface area of ​​the particles is preferably in a range of 100 m². 2 / g up to 150 m 2 / g lies.

6. A method according to any one of the preceding claims, wherein the polishing agent (M) further contains active chlorine, preferably between 1 g / l and 10 g / l active chlorine.

7. A method according to any one of the preceding claims, wherein the polishing agent (M) has a pH value between 2 and 4.

8. Method according to one of the preceding claims, wherein several semiconductor wafers (W) with at least one side to be polished are brought into contact with a polishing cloth (P), wherein several semiconductor wafers (W) are inserted into a carrier (T), wherein the carrier (T) moves on a circular path and rotates simultaneously, wherein the polishing cloth (P) preferably also rotates about its central axis.

9. Method according to claim 8, wherein at least one area of ​​the polishing cloth (P) is structured, wherein in structured areas the thickness of the polishing cloth (P) is 500 pm - 800 pm less than in non-structured areas, and / or the polishing cloth (P) is hydrophilized in structured areas.

10. III / V wafer, which has two opposing surfaces and which has at least one of the following properties: • In a measurement field on at least one surface of the wafer, the distance between the highest and lowest point of the surface is a maximum of 1.067 pm; • In a measurement field on at least one surface of the wafer, the maximum distance between the highest or lowest point above or below the shifted focal plane and the shifted focal plane is a maximum of 0.544 pm.

11. III / V wafer according to claim 10, wherein the measurement field has a size of 20 mm x 20 mm, and the maximum distance between the highest or lowest point above or below the shifted focal plane and the shifted focal plane is a maximum of 0.367 pm.

12. IIIAA wafer according to claim 10 or 11, which has at least one of the following features: • In a measurement field on at least one surface of the wafer, the distance between the highest and lowest point of the surface is a maximum of 0.754 pm; • In a measurement field on at least one surface of the wafer, the maximum distance between the highest point above the shifted focal plane and the lowest point below the shifted focal plane is a maximum of 0.48 pm; • In a measurement field on at least one surface of the wafer, the maximum distance between the highest or lowest point above or below the shifted focal plane and the shifted focal plane is a maximum of 0.313 pm.

13. IIIA / -wafer according to one of claims 11 or 12, wherein the measurement field has a size of 20 mm x 20 mm, and the maximum distance between the highest or lowest point above or below the shifted focal plane and the shifted focal plane is a maximum of 0.305 pm.

14. IIIA / -wafer, which has two opposing surfaces and which has at least one of the following properties: • The taper value is a maximum of 0.588 pm; • The distance between the highest point of the wafer above the focal plane and the lowest point of the wafer below the focal plane is a maximum of 1.535 pm; • The maximum distance of the point furthest from the focal plane is 0.996 pm; • The maximum distance between the highest and lowest points of a surface is 1.419 pm.

15. III / V wafer according to one of claims 10 to 14, wherein the Hl element is selected from Ga, AI and In, and the V element is selected from P and As, wherein the Hl element is preferably In, and the V element is preferably P.

16. III-V wafer according to any one of claims 10 to 15, wherein the wafer diameter is between 75 and 155 mm, and the thickness of the wafer is 575 pm to 650 pm.

17. Group of III / V wafers having two opposing surfaces, wherein the group has at least one of the following properties: • In a measurement field on at least one surface of the wafer, the mean distance between the highest and lowest point of the surface is a maximum of 0.894 pm; • In a measurement field on at least one surface of the wafer, the mean value of the maximum distance between the highest point above the shifted focal plane and the lowest point below the shifted focal plane is a maximum of 0.573 pm; • In a measurement field on at least one surface of the wafer, the mean value of the maximum distance between the highest or lowest point above or below the shifted focal plane and the shifted focal plane is a maximum of 0.445 pm.

18. Group of IIIA / -wafers according to claim 17, wherein in a measurement field of 20 mm x 20 mm on at least one surface of the wafer the mean value of the maximum distance between the highest or lowest point above or below the shifted focal plane and the shifted focal plane is a maximum of 0.279 pm.

19. Use of an α-hydroxycarboxylic acid in a polishing solution for semiconductor wafers to prevent flocculation in wastewater.

Images