Process aware, defectivity based contamination control recipe development in chip manufacturing

Abstract

The invention relates to a method for cleaning items, especially containers for transporting and / or storing wafers or reticles such as FOUPs or reticle pods, in a fab environment, comprising the steps of identification of a plurality of critical contaminants amongst the total contamination, especially in in a closed (mini) environment such as the interior of a container, after identifying the critical contaminants, choosing and / or optimizing a cleaning process with a view to suppressing at least one of the plurality of critical contaminants, especially until a predetermined first threshold level of the at least one of the critical contaminants has been reached, and after supressing the at least one of the plurality of the critical contaminants, reducing the overall contamination, especially comprising AMCs, VoCs etc, especially until a predetermined second threshold level of the overall contamination has been reached.

Description

[0001] Brooks Automation (Germany) GmbH DG1.DD174922 / 01 IDF3187 21 November 2025

[0002] 1

[0003] Process aware, defectivity based contamination control recipe development in chip manufacturing

[0004] Background

[0005] Semiconductor chip manufacturing is an extremely complicated affair, involving a multitude of complex and competing processes. Every process goes is subjected to a high level of regulation and control to be able to result in a working chip. Contamination occurring or generated within a fab poses a great threat if not addressed immediately after each process step. There are many species of contaminants that are generated in the various steps of chip manufacturing, especially in the intermediate steps, and some of these contaminant species are more harmful than others.

[0006] For example, contaminants present during semiconductor chip manufacturing, such as particulate matter, especially airborne molecular contamination (AMC), chemical residues, metal ions, and volatile organic compounds (VOCs), can significantly impact chip yield by introducing defects that degrade device performance or cause process failures. Particulate matter can interfere with photolithography and deposition processes, leading to misalignment or defects in circuit patterns. Chemical contamination, including residual etchants or solvents, may disrupt subsequent process steps like etching or doping, while metal ion contamination can induce electrical shorts or leakage currents, especially in advanced nodes with smaller feature sizes. VOCs and water impurities can also affect layer formation and cleaning processes, while fluctuations in temperature and humidity can cause alignment errors or material property variations. Human contamination, such as oils or fibres from clothing, may introduce particles that result in surface defects. Metal ion contamination is typically one of the most harmful, as even trace amounts can cause irreversible electrical failures, significantly degrading device performance and reducing yield, especially in smaller, more sensitive semiconductor nodes. AMC (Airborne Molecular Contamination) and VOC (Volatile Organic Compound) contamination are also highly harmful because they can adsorb onto wafer surfaces or react with process chemicals, leading to defects in critical layers, degradation of material properties, and interference with photolithography and etching processes.

[0007] As a further example, fluoride ions have been identified as a particularly harmful contaminant, causing bond pad detectivity in subsequent processing steps. This type of contamination can lead to electrical failure or degradation of chip performance, ultimately resulting in non-performing devices. The presence of such ionic species, if not effectively detected and removed, also poses a significant risk to yield and reliability in semiconductor manufacturing.

[0008] As semiconductor technology advances, minimizing contamination is critical to maintaining high yields, requiring stringent cleanroom protocols, precise environmental controls, and careful handling throughout the manufacturing process.

[0009] In an automated semiconductor integrated circuit fabrication facility, usually referred to as a fab, front opening unified pods, so called FOURS, in which wafers, out of which semiconductor chips are manufactured, are stored, are moved between various stations using corresponding handling systems. FOUPs must comply with specific cleanliness standards, as, otherwise, FOUP contamination can significantly reduce chip yield. Similar considerations apply to so called reticle pods, in which reticles are stored and transported within the fab environment.

[0010] It is essential to develop an effective FOUP cleaning strategy that considers the specific contaminant species and their control limits to ensure optimal performance and yield in semiconductor manufacturing. In this disclosure, we propose a "process-aware" FOUP cleaning strategy that is tailored to the unique requirements of various technologies, applications, and substrate types. Such a strategy is critical as different processes, such as low power CMOS, III- V semiconductors (e.g., SiC and GaN), memory applications (e.g., DRAM, NAND, 3D NAND), and sensor applications, each have distinct contamination control needs. This targeted approach ensures that each type of device benefits from a cleaning protocol specifically designed to address the particular challenges posed by the contaminants relevant to the respective technology or application.

[0011] Also, there are fab to fab variations for the same process and the contamination control / cleaning strategy needs to adapt according to the said process. Hence, it is imperative for the cleaning strategy to be aligned with the process requirements as a generic strategy might not perform optimally across fabs and processes.

[0012] Cleaning systems, particularly FOUP / pod cleaners such as the applicant’s PuroMaxx 800 (LEAP and XP), play a critical role in contamination control within the fab environment by effectively removing airborne molecular contaminants (AMCs) and volatile organic compounds (VOCs) generated during the chip production process. Each step in semiconductor manufacturing can produce numerous chemical species, both molecular and ionic, some of which can lead to detectivity in subsequent processing stages if not promptly addressed. Targeted detection and elimination of AMCs and VOCs — especially those known to cause such defects — by minimizing or arresting their concentration is essential for maintaining a clean mini-environment within the FOIIP. This, in turn, is a key measure for reducing detectivity and improving overall yield throughout the semiconductor manufacturing flow.

[0013] In this disclosure a method for developing a process aware solution or an optimised cleaning strategy for contamination control is proposed.

[0014] According to the invention, a method and a device for cleaning a container configured and adapted for transporting wafers or reticles in a fab environment as defined in the independent claims are provided. Advantageous embodiments are the subject matter of the dependent claims. According to a first aspect of the invention, there is provided a method for cleaning a container configured and adapted for transporting and / or storing wafers or reticles comprising the features of claim 1 .

[0015] By implementing the invention, it is possible to specifically and effectively target critical contaminants, thereby enabling a more efficient cleaning process as a whole, which can especially serve to minimise consumption of cleaning fluids and the overall cleaning time.

[0016] The invention is especially useful in case the container is provided as a FOUR or a reticle pod.

[0017] According to an embodiment of the method according to the invention, the method steps are performed in the order a) followed by b) followed by c) followed by d). Herewith, it is effectively possible to first reduce “critical” contaminants of a container used in the fab environment to a first cleanliness level, using a first cleaning method / agent, and then general contaminants to a second cleanliness level using a second cleaning method / agent. One of the advantageous effects which can be observed in this case is that by specifically targeting the critical contaminants first, the overall usage of cleaning fluid and the overall cleaning time can be effectively reduced. A further advantage of this embodiment is that a cleaning agent effective for removal of a critical contaminant is more likely to be critical itself, so that this would also be removed more effectively during a subsequent reduction of the overall contamination.

[0018] According to a further embodiment of the method according to the invention, the method steps are performed in the order a) followed by b) followed by d) followed by c). In this embodiment, the removal of the overall contamination is performed before the removal of the at least one critical contaminant. This implementation of the invention also leads to reduced overall consumption of cleaning fluids. The advantage in this embodiment is that all information relating to identification of critical contaminants has been obtained before beginning with the removal of the overall contamination. Hereby, the process for the removal of overall contamination can be optimised with a view to the subsequent process for the removal of the critical contaminants.

[0019] Advantageously, the method is performed based on contamination as measured in a closed environment, especially in the interior of the container.

[0020] According to a further aspect of the invention, there is provided a device for cleaning a container configured and adapted for transporting and / or storing wafers comprising the features of claim 6. The device is especially adapted to perform the method according to the first aspect and / or any of its embodiments.

[0021] The invention will now further be explained with reference to the appended figures. Herein,

[0022] Figure 1 shows a schematical view of an exemplary embodiment of a device according to the invention, and

[0023] Figure 2 shows a schematical process diagram illustrating an embodiment of the method according to the invention.

[0024] A combined cleaning and inspection system 400 for a container such as a FOUR 100 according to a preferred embodiment of the invention is shown in Figure 1 in a purely schematic manner. The system 400 comprises a housing 401 including an inspection unit 408 and a cleaning unit 410.

[0025] The system 400 comprises at least one load port 402, configured to introduce a container, such as FOUR 100, which is to be cleaned, into the cleaning system, an Equipment Front End Module EFEM 404 adapted to cooperate with the at least one load port 402, a gate opening section 406 connected to the EFEM 404, and a gate closing section 409 connected to the EFEM 404. The system can comprise a buffer station (not shown), for example between the gate opening section 406 and the inspection unit 408. The buffer station may be adapted to transport the FOUR 100 into the inspection unit 408. The system is provided with at least one control unit 440.

[0026] The inspection unit 408 is adapted to measure contamination of FOUR 100. It is especially adapted to identify an overall contamination and at least one critical contaminant amongst the overall contamination of the FOIIP 100. Inspection unit 408 comprises, for example, at least one sensor 420 or measurement device for measuring and identifying at least one type of contamination, especially at least one critical contaminant, of the FOIIP 100. Optionally, the inspection unit 408 may comprise one or more imaging devices 424, for example a video camera, for visually identifying contaminants and / or other features or defects of the FOIIP 100

[0027] Control unit 440 is adapted to choose and / or optimize a cleaning process to suppress or minimize the at least one critical contaminant thus identified, and to reduce the overall contamination. An input terminal 403 may be provided to allow manual input of data to the control unit 440, for example.

[0028] The cleaning unit 410 is adapted to implement the cleaning process thus chosen, initially removing the at least one critical contaminant from the FOIIP 100, until a predetermined first threshold level of the at least one critical contaminant has been reached. The cleaning unit 410 is further adapted to subsequently reduce overall contamination of the FOIIP 100 until a second threshold level relating to overall contamination has been reached. Following this reduction of overall contamination, the FOIIP 100 may be subjected to a further inspection in inspection unit 408. In case it is determined that additional cleaning, either with respect to the at least one critical contaminant, or overall contamination is necessary or expedient, the FOIIP 100 may then be subjected to an additional cleaning in cleaning unit 410. It is also possible, of course, that the FOIIP 100 thus cleaned is transferred from the cleaning unit 410 directly to gate closing section 409 without any further inspection. The inspection unit 408 can be provided with separate inspection sections, one section, for example, being adapted to perform an inspection before the cleaning, and another section adapted to perform inspection after the cleaning. It is, however, also possible to perform all inspections within the same inspection section.

[0029] The cleaning unit 410 may, for example, comprise a cleaning chamber 411 in which cleaning operations such as wet cleaning may be performed. In such a cleaning chamber 411 nozzles 412 for spraying a cleaning fluid towards the FOUR 100 to be cleaned may be provided. Additionally or alternatively, the cleaning unit 410 may comprise a vacuum chamber 413 for providing a reduced pressure inside and around the FOUR 100, for example after a wet cleaning process. This enhances evaporation of any cleaning fluids such as water and may also remove volatile organic contaminants. To this end, such a vacuum chamber 413 may be provided or connected with a vacuum pump 414. The cleaning chamber 411 and / or the vacuum chamber 413 may be provided with temperature control equipment (not depicted), such as IR-heaters, for example.

[0030] The control unit 440 may be communicatively coupled to the nozzles 412 and / or the vacuum pump 414 to control their respective operation. For example, the control unit may control the nozzles 412 to apply a certain recipe of cleaning fluid to the FOIIP 100, according to the chosen cleaning process.

[0031] The inspection unit 408 can also be provided, for example, with an imaging unit (e.g. imaging device(s) 424) for imaging of at least part of the FOIIP being inspected, for example a gasket, a pressure detection unit for monitoring a pressure within the FOIIP following a pressurization or a de-pressurization, a detection unit for monitoring a heat reaction following an exposure of the container to a heat source, and / or a humidity detection unit for monitoring humidity within the FOIIP.

[0032] Further sensors 420, 422 can be positioned throughout the system 400, especially in at least one of the load port 402, the EFEM 404, the gate opening section 406, the buffer station, the cleaning unit 410, or the gate closing section 409. Only some of these sensors 420, 422 are schematically shown in Figure 1 . These sensors can be adapted to provide various data for monitoring FOUR characteristics.

[0033] The path typically taken by the FOUR 100 in connection with inspection and cleaning as described above within cleaning system 400 is indicated by arrows in Figure 1 .

[0034] According to a further embodiment, it is, for example, possible, after the inspection unit 408 has measured contamination of FOIIP 100 and identified an overall contamination and at least one critical contaminant amongst the overall contamination of the FOIIP 100, to initially reduce overall contamination, before suppressing or minimizing the at least one critical contaminant.

[0035] According to a further embodiment, it is, for example, possible to first reduce overall contamination of FOIIP 100 in cleaning unit 410, before performing an identification of a remaining overall contamination and at least one critical contaminant in inspection unit 408, and then choosing and / or optimizing a cleaning process to suppress or minimize the at least one critical contaminant, followed by implementing this cleaning procedure as chosen or optimised.

[0036] Be it noted that the inspection unit 408 and the cleaning unit 410 may be realised as one combined unit. The description as separate units 408, 410 above is mainly for i llustrational purposes, although it is possible to provide two separate units together with a transportation device for transporting FOUPs 100 therebetween.

[0037] The method according to the invention will now be further explained with reference to Figure 2.

[0038] Before explicitly discussing Figure 2, some general observations regarding FOIIP cleaning, as an example for cleaning requirements within a fab environment, are made as follows: Currently, the cleaning recipes for FOUPs are generic, as the capability for contaminant detection and speciation, i.e. identification of critical contamination, is not present in current versions of the cleaning tools. There are existing projects, such as the applicant’s HCC (Holistic Contamination Control) project, that enable such capabilities and involve for example design and development with third party technology vendors. The present invention is designed to be used in connection with such systems.

[0039] As an example of a critical contaminant, which can be identified and then reduced or suppressed according to the invention, it is referred to flouride ions, which are the cause of bond pad, as mentioned above: If these ions are attempted to be removed from FOUPs in connection with the general cleaning step targeting the overall contamination (i.e. especially aimed at AMCs, VoCs etc), this general cleaning step will have to be executed over a much longer period of time and / or using greater amounts of cleaning agents. According to the invention, on the other hand, by first specifically targeting critical contaminants, the intensity and or duration of general cleaning step can be significantly reduced as compared to prior art solutions. Further examples of critical contaminants are copper (Cu) ions, acetic acid, nitrates, organo-metallic ions from substrates, sulfates, nitrates etc.

[0040] A bond pad is a small, flat area on the surface of an electronic component, such as an integrated circuit (IC), used to establish an electrical connection for bonding wires or other interconnects. These pads are essential in the process of connecting the internal circuitry of the IC to external components, like wires, solder balls, or leads.

[0041] Fluoride ions are difficult to remove from surfaces, especially from materials like bond pads (used in electronics), because of the strong chemical interactions they form with the surface. Fluoride (F“) is a highly electronegative ion, meaning it has a strong tendency to attract electrons. When it comes into contact with a metal or metal oxide surface, it tends to form strong ionic bonds with the metal atoms, making the interaction very stable and hard to break. Bond pads (which are typically made from materials like gold, copper, or aluminium) have a surface that can readily form these stable bonds with fluoride ions. The strength of the bond between fluoride and the metal surface makes it difficult to remove the fluoride ions without applying significant energy. When fluoride ions come into contact with the bond pad material, they can form metal fluoride compounds (such as aluminium fluoride in the case of aluminium bond pads). These metal fluoride compounds are usually stable and not easily soluble in common solvents or cleaning solutions. This creates a situation where the fluoride is essentially "locked" into a solid compound on the surface, and removing it requires breaking the bond or dissolving the fluoride compound, which is not straightforward.

[0042] In the electronics industry, common cleaning methods (such as using solvents, acids, or plasma cleaning) often struggle to effectively remove fluoride ions once they have bonded to a surface. Some methods, like acidic treatments, might not be aggressive enough to break the strong ionic bonds formed with fluoride. Additionally, the bond pad materials themselves can be sensitive to over-aggressive cleaning, risking damage to the pad or surrounding components.

[0043] When fluoride ions interact with the surface, they can also leave behind a thin residue of metal fluoride. This residue is highly resistant to removal, and if it is not cleaned properly, it can contribute to long-term issues like reduced performance or reliability of the electronic component.

[0044] To avoid the difficulty of removing fluoride, it is common to take preventative measures during the manufacturing process, such as controlling the environment to prevent fluoride contamination in the first place. Once fluoride is present, however, dealing with it effectively requires specialized cleaning techniques, often involving more aggressive or tailored methods than those typically used for other contaminants.

[0045] These explanations serve to illustrate the cleaning effort necessary to reduce critical contaminants such as fluoride contaminations. The following cleaning processes and substances are typically used: Acidic Solutions: The method utilizes acidic solutions to dissolve metal fluoride compounds and break the ionic bonds between the fluoride ions and the metal surface. The preferred acidic cleaning agent is hydrofluoric acid (HF), which is highly effective at dissolving fluoride-based residues. HF is typically used in a diluted form to ensure safe handling and to prevent damage to the bond pad surface. A suitable concentration range for HF is approximately 1-10% by volume, mixed with deionized water, depending on the specific application and desired cleaning effect.

[0046] In some situations, nitric acid (HNO3) may be used in combination with hydrofluoric acid to enhance cleaning performance, especially for removing oxidation and other residues from the bond pad surface. The nitric acid concentration may range from 1-10% by volume.

[0047] In addition to acidic solutions, ammonium hydroxide (NH40H) or other alkaline solutions may be employed, particularly in cases where a milder cleaning process is desired or when there is a need to reduce corrosion. Ammonium hydroxide, when used in combination with other agents, can facilitate the dissolution of metal fluoride residues.

[0048] After the fluoride ions are removed, organic solvents such as acetone or isopropyl alcohol (IPA) are used to remove organic contaminants, oils, or residual chemicals from the bond pad surface. These solvents assist in ensuring that the bond pad is free from any residues that could interfere with subsequent bonding or functionality.

[0049] In conclusion, the difficulty in removing fluoride ions from surfaces in a fab environment arises from the strength of the bond formed between fluoride and the surface, the stability of the metal fluoride compounds that can form, and the limitations of typical cleaning methods. Breaking this bond requires careful and precise techniques, which are often more challenging than removing other types of contaminants. Further examples of potential contaminants are copper (Cu) ions, acetic acid, nitrates, organo-metallic ions from substrates, sulphates, nitrates etc. Similarly, there are hundreds of processes that are dependent on removal of certain specific contaminants that pose a greater threat than other species (even though present in larger volume / concentration). The ability to identify all the constituents (speciation) and suppressing the most critical ones, is the key to reduce detectivity and thereby improve yield.

[0050] Referring to Figure 2, an embodiment of the method according to the invention will be further explained: In a first step, an overall contamination of a semiconductor carrier device such as a FOUR is measured (step 200). An overall contamination profile is shown in a schematic manner for illustrative purposes and designated OCP. As indicated by an arrow, this is followed by a step 210 of identification or speciation of at least one critical contaminant within the overall contamination. Be it assumed that in a subsequent step 220 one critical contaminant C has thus been identified. Be it further assumed that a generally acceptable level of overall contamination as shown in the overall contamination profile OCP can be characterized by means of a first threshold T1 . Be it assumed that his first threshold T1 is, however, not acceptable, i.e. too high, for critical contaminant C, for which a lower threshold T2 is required. As can be seen in connection with step 220, the overall contamination profile exceeds both thresholds T1 and T2. Steps 200 to 220 are especially performed in a closed (mini) environment, i.e. within the FOIIP for which contamination is to be measured. Identification of critical contaminants identifies those contaminant species that are extremely critical from the point of view of subsequent process steps.

[0051] Having identified the at least one critical contaminant C, the method proceeds to a step 230, as indicted by an arrow, to choose and / or optimize a FOIIP cleaning process in order to suppress the critical contaminant C as well as the overall contamination. Cleaning recipes used in this connection will typically comprise at least one of washing, especially using cleaning fluids, drying, especially using heat and / or streams of expedient gases, as well as megasonic treatments, nano-bubble based cleaning, chemical cleaning. These various cleaning techniques will typically be employed especially based on the nature of contaminant C and / or the cleaning efficiency or capability of the specific cleaning technique. In a subsequent step 240, this optimised cleaning process is performed to ensure that the contamination level of critical contaminant C falls below second threshold T2.

[0052] Step 240 is followed by a general cleaning step general cleaning step 250, as again indicated by an arrow. This is usually the final cleaning step, wherein the overall contamination (especially AMCs, VoCs etc) is reduced to acceptable values, as defined by first threshold T 1 . The final cleanliness state of the FOUR is achieved in step 260. Herein, as described above, the contamination level of critical contaminant C has been reduced to a value below threshold T2, and the overall contamination level to a value below (slightly higher) threshold T1 .

[0053] A number of exemplary advantages achievable in connection with the invention are listed as follows:

[0054] 1. The cleaning process is developed based on historical process knowledge and yield-limiting detectivity issues.

[0055] 2. It makes the process ‘green’ because of a focused cleaning approach, thereby reducing the consumption of cleaning agents such as critical gases, DiW (De-ionized water), N2 as well as energy.

[0056] 3. The process also improves overall throughput as the cleaning process is precise and targeted.

[0057] 4. The cleaning process is wafer yield oriented and not open ended.

Claims

Claims1 . Method for cleaning a container configured and adapted for transporting and / or storing wafers or reticles in a fab environment, comprising the steps of a) measuring an overall contamination of the container and identifying a plurality of critical contaminants amongst the overall contamination of the container, b) choosing and / or optimizing a cleaning process to reduce at least one of the plurality of critical contaminants and to reduce the overall contamination, c) implementing the cleaning process as chosen and / or optimised to reduce at least one of the plurality critical contaminants, especially until a predetermined first threshold level of the at least one of the plurality of critical contaminants has been reached, d) reducing the overall contamination, especially until a predetermined second threshold level of the overall contamination has been reached.

2. Method according to claim 1 , wherein the container is provided as a FOUR or a reticle pod.

3. Method according to claim 1 or 2, wherein steps a), b), c) and d) are performed in the order a) followed by b) followed by c) followed by d).

4. Method according to claim 1 or 2, wherein steps a), b), c) and d) are performed in the order a) followed by b) followed by d) followed by c).

5. Method according to any one of the preceding claims, wherein steps a) to d) are performed based on contamination in a closed environment, especially in the interior of the container.

6. Device for cleaning a container configured and adapted for transporting and / or storing wafers or reticles such as FOUPs or reticle pods in a fab environment, comprising an identifying unit configured and adapted for identifying a plurality of critical contaminants amongst an overall contamination, especially in in a closed environment such as the interior of the container, a choosing and / or optimizing unit configured and adapted for choosing and / or optimizing a cleaning process with a view to suppressing or minimising at least one of the plurality of critical contaminants, an implementing unit configured and adapted to implement the cleaning process as chosen or optimised, especially until a predetermined first threshold level of the at least one of the plurality of critical contaminants has been reached, and a unit configured and adapted for reducing the overall contamination, especially comprising AMCs, VoCs etc, especially until a predetermined second threshold level of the overall contamination has been reached.

7. Device according to claim 6, adapted to perform the method according to any one of claims 1 to 5.

Images