METHOD AND SYSTEM FOR THE ACOUSTIC DETECTION OF AN ARC FORMATION DURING A SEMICONDUCTOR PROCESS

Abstract

Procedure (700) with the following steps: Generating (702) audio signals from sound waves using an arrangement of microphones (108) positioned in a semiconductor processing cleanroom environment (102); Processing (704) the audio signals with a control system (110); and Detecting (706) an arc formation in the semiconductor processing cleanroom environment (102) with the control system (110) based on the sound signals, the arrangement of microphones (108) comprises the following: one or more microphones (108) configured to detect infrasound; one or more microphones (108) configured to detect audible sound; and one or more microphones (108) configured to detect ultrasound.

Description

BackgroundField of Expertise

[0001] The present invention relates to the field of semiconductor processing. In particular, the present invention relates to detecting arcing in a semiconductor process. Description of the state of the art

[0002] Integrated circuit manufacturing is typically achieved by performing a large number of processing steps on semiconductor wafers. These processing steps typically result in the fabrication of a large number of transistors in highly complex arrays in conjunction with a semiconductor substrate. These processing steps also create dielectric layers, metallic interconnects, vias, pins, and other integrated circuit structures and components.

[0003] In various semiconductor processes, electrical arcing can occur in the processing environment. Electrical arcing can damage semiconductor wafers. This may make it necessary to discard the damaged semiconductor wafers.

[0004] US6753499B1 discloses a device having a plurality of ultrasonic detectors arranged at different portions of a plasma processing chamber so that an ultrasonic wave accompanying an arc discharge can be detected by the ultrasonic detectors. JP 2003-100714 A discloses a method for determining an abnormal discharge on a wafer by efficiently supplying ultrasound to ultrasonic detectors when the abnormal discharge is generated. JP 2007-3443 A discloses a method for determining an abnormal condition of an object using a sheet-type piezoelectric ultrasonic sensor.

[0005] DE 10 2018 124 690 A1 discloses a method for operating a device for producing or analyzing semiconductor wafers. In the method, sound is detected in a process chamber of the device during an operation of the device. An electrical signal corresponding to the detected sound is obtained by a signal processor. The obtained electrical signal is processed by the signal processor. An event during the operation of the device is detected based on the processed electrical signal. The operation of the device is controlled according to the detected event. Short description of the drawings Fig. 1 is a block diagram of an arc detection system according to one embodiment. Fig. 2 is an illustration of an arc detection system according to one embodiment. Fig. 3 is an illustration of a semiconductor wafer showing locations where sustained arcing events occur, according to one embodiment. Fig. 4 shows several curves associated with audio signals generated by microphones of an arc detection system according to one embodiment. Fig. 5 is a block diagram of an audio signal output circuit of an arc detection system according to an embodiment. Fig. 6 is a block diagram of a control system of an arc detection system according to an embodiment. Fig. 7 is a flow diagram of a method for detecting arcing in a semiconductor clean room environment according to one embodiment. Fig. 8 is a flow diagram of a method for detecting arcing in a semiconductor cleanroom environment according to one embodiment. Detailed description

[0006] In the following description, numerous thicknesses and materials are described for various layers and structures of an integrated circuit die. Specific dimensions and materials are provided as examples for various embodiments. Those skilled in the art, in light of the present invention, will recognize that in many cases, other dimensions and materials may be used without departing from the scope of the present invention.

[0007] The following description provides many different embodiments or examples for implementing various features of the provided subject matter. Specific examples of components and arrangements are described below to facilitate the present invention. These are, of course, merely examples and are not intended to be limiting. For example, the fabrication of a first element over or on top of a second element in the following description may include embodiments in which the first and second elements are fabricated in direct contact, and may also include embodiments in which additional elements may be fabricated between the first and second elements such that the first and second elements are not in direct contact. Furthermore, in the present invention, reference numerals and / or letters may be repeated in the various examples.This repetition is for simplicity and clarity and does not, in itself, prescribe any relationship between the various embodiments and / or configurations discussed.

[0008] Furthermore, spatially relative terms such as "beneath," "under," "lower," "above," "upper," and the like may be used herein to conveniently describe the relationship of one element or structure to one or more other elements or structures illustrated in the figures. The spatially relative terms are intended to encompass other orientations of the device in use or operation, in addition to the orientation illustrated in the figures. The device may be oriented differently (rotated 90 degrees or in a different orientation), and the spatially relative descriptors used herein may be interpreted accordingly.

[0009] In the following description, specific details are set forth to provide a thorough understanding of various embodiments of the invention. However, one skilled in the art will recognize that the invention may be practiced without these specific details. In other instances, well-known structures associated with electronic components and manufacturing processes are not described in detail to avoid unnecessarily obscuring the description of embodiments of the present invention.

[0010] Throughout the specification and in all claims, the term "comprise" and variations thereof, such as "... comprising", "comprise", "... comprising", "having", "... having", "with", etc., are to be interpreted in an open, inclusive sense, that is, as "comprise, but not limited to", unless the context requires otherwise.

[0011] The use of ordinal numbers, such as “first”, “second”, and “third”, does not necessarily imply an ordering of ranks, but merely serves to distinguish between multiple instances of a process or structure.

[0012] Throughout the specification, reference to "one embodiment," or "an embodiment," means that a particular element, structure, or feature described in connection with the embodiment is part of at least one embodiment. Thus, the appearances of the phrase "in one embodiment" in various places throughout the specification are not necessarily all referring to the same embodiment. Moreover, the specific elements, structures, or features may be suitably combined in one or more embodiments.

[0013] As used in the specification and the appended claims, the singular forms "a" and "an" and "the" also include plural forms unless the context clearly dictates otherwise. It should also be noted that the term "or" is generally used to mean "and / or" unless the context clearly dictates otherwise.

[0014] Fig. Figure 1 is a block diagram of an arc detection system 100 according to one embodiment. The arc detection system 100 includes an array of microphones 108 positioned within a semiconductor processing cleanroom environment 102. The arc detection system 100 also includes a control system 110 communicatively coupled to the microphones 108. The microphones 108 and the control system 110 cooperate to detect arcing in the cleanroom environment 102.

[0015] The cleanroom environment 102 is used to perform semiconductor processing on a semiconductor wafer 104. The semiconductor processing results in a plurality of integrated circuits in the semiconductor wafer 104. After the semiconductor processing is complete, the semiconductor wafer 104 can be sawn to separate individual integrated circuit dies from the semiconductor wafer 104.

[0016] The semiconductor processes may include depositing or growing semiconductor layers, dielectric layers, metal layers, and other structural features on or in the semiconductor wafer 104. The semiconductor processes may include photolithographic processes for patterning structural features from the various materials deposited or grown on the semiconductor wafer 104. The semiconductor processes may include etching processes for removing material in conjunction with the photolithographic processes for patterning the various materials deposited or grown on the semiconductor wafer 104. The semiconductor processes may include ion implantation processes for implanting dopant species into the semiconductor wafer 104. The semiconductor processes may further include cleaning, planarization, and annealing steps.Other semiconductor processes may also be performed without departing from the scope of the present invention.

[0017] A semiconductor process system 106 controls the various semiconductor processes performed on the semiconductor wafer 104 in the cleanroom environment 102. The semiconductor process system 106 may include chemical vapor deposition systems, physical vapor deposition systems, wet etching systems, dry etching systems, annealing systems, ion implantation systems, photolithography systems, and other systems suitable for performing semiconductor processing on the semiconductor wafer 104. The semiconductor process system 106 may include components positioned within the cleanroom environment 102 and components positioned outside the cleanroom environment 102. The semiconductor process system 106 may include power sources that generate electric fields in the cleanroom environment 102 or apply voltages in the cleanroom environment 102. The semiconductor process system 106 may include control systems that control the semiconductor processes performed by the semiconductor process system 106.

[0018] While processing steps are being performed on the semiconductor wafer 104, electrical arcing may occur in the cleanroom environment 102. Arcing is the momentary flow of a high electrical current in the cleanroom environment 102. The arc may be caused by an electrostatic charge buildup between objects or materials in the cleanroom environment 102. Electrostatic charges may build up on the walls of the cleanroom environment 102, on process equipment in the cleanroom environment 102, in gases present in the cleanroom environment 102, or in other ways. If an electric field becomes too large due to the electrostatic charge buildup or other factors, an arc is created.

[0019] If an arc occurs near the semiconductor wafer 104, the semiconductor wafer 104 may be damaged. For example, if an arc causes large currents to flow between a portion of the semiconductor wafer 104 and another object or material in the cleanroom environment 102, the currents may damage the semiconductor wafer 104. If the semiconductor wafer 104 is damaged, some integrated circuit dies derived from the semiconductor wafer 104 may not function properly. In some cases, the semiconductor wafer 104 may be so severely damaged that the entire semiconductor wafer 104 must be discarded. This results in a significant loss of time, materials, and money.

[0020] The Arc Detection System 100 from Fig. 1 can detect and control arcing in the clean room environment 102 in real time. In the arc detection system 100 of Fig. 1, the microphones 108 and the control system 110 are used to detect and control arcing in the clean room environment 102.

[0021] Arcing leads to the creation of sound waves. Sound waves are typical of arcing. Furthermore, different types of arcing between different materials lead to different types of sounds.

[0022] Accordingly, the arc detection system 100 detects arcing based on the tones generated by the arcing.

[0023] In the arc detection system 100, an array of microphones 108 is used to detect sounds generated by arcing. In the array of microphones 108, the microphones 108 are positioned at various locations within the cleanroom environment 102. The microphones 108 are configured to enable detection of arcing sounds at various locations within the cleanroom environment 102.

[0024] In one embodiment, microphones 108 are positioned within cleanroom environment 102 to enable detection of an arcing location within cleanroom environment 102. By placing the microphones at various selected positions, the characteristics of the sound detected by each microphone can be used to determine the arcing location within cleanroom environment 102. For example, the magnitude of the arcing sound detected by each microphone, taking into account the position of each microphone within cleanroom environment 102, allows identification of the exact arcing location within cleanroom environment 102.

[0025] The microphones 108 generate audio signals in response to the sound waves impinging on the microphones 108. The audio signals correspond to electrical signals typical of the sounds received by the microphones 108.

[0026] The microphones 108 are microphones sensitive to various sound wave frequency ranges. Some of the microphones 108 are sensitive to sound waves that are perceptible to the human ear. For example, most people can hear sounds in a range of 50 Hz to 20,000 Hz. Some of the microphones 108 are sensitive to sound waves in this range or to subsets of sound waves in this range.

[0027] In one embodiment, some of the microphones 108 are sensitive to ultrasonic waves. Ultrasonic waves correspond to sound waves with frequencies higher than the frequencies normally perceived by the human ear. Some of the microphones 108 are sensitive to sound waves in a range of 20 kHz to 20 MHz. Some of the microphones 108 may be sensitive to various subsets of frequencies from 20 kHz to 20 MHz.

[0028] In one embodiment, some of the microphones 108 are sensitive to infrasonic waves. Infrasonic waves correspond to sound waves with frequencies lower than the frequencies normally perceived by the human ear. Accordingly, some of the microphones 108 are sensitive to sound waves in a range of 0 Hz to 100 Hz. Some of the microphones 108 may be sensitive to various subsets of frequencies from 0 Hz to 100 Hz.

[0029] In one embodiment, some of the microphones 108 are in contact with a semiconductor processing device located within the cleanroom environment 102. For example, some of the microphones 108 may be in contact with a carrier or other object on which the semiconductor wafer 104 is positioned during processing. These microphones may assist in determining whether the carrier or other object, or the semiconductor wafer 104 itself, has been affected by the arc.

[0030] The control system 110 is communicatively connected to the microphones 108. In particular, the control system 110 receives the audio signals from the microphones 108. The control system 110 processes the audio signals and, based on the audio signals, determines whether an arc has occurred in the cleanroom environment 102.

[0031] The control system 110 may receive the audio signals wirelessly from the microphones 108. The microphones 108 may be connected to the control system 110 using one or more wireless connection protocols, including Bluetooth, Wi-Fi, or Zigbee protocols. Other wireless connection protocols may also be used without departing from the scope of the present invention. Accordingly, the arc detection system 100 may include wireless transceivers or transmitters connected to the microphones 108 positioned within the clean room environment 102. The wireless transceivers or transmitters may transmit the audio signals to the control system 110.

[0032] In one embodiment, the arc detection system 100 includes power sources connected to the microphones 108 and, if present, to wireless communication systems connected to the microphones 108. The power sources may be batteries or wired connections to constant current sources. In one embodiment, the arc detection system 100 includes energy harvesters for recharging batteries or otherwise powering the microphones 108 and the wireless communication systems. The energy harvesters may harvest energy from sound wave vibrations or other types of motion.

[0033] In one embodiment, the microphones 108 are connected to the control system 110 via a wired connection. Accordingly, the microphones 108 can communicate with the control system 110 via a wired connection that relays signals to the control system 110.

[0034] The microphones 108 may initially generate analog audio signals. The analog audio signals may be provided to the control system 110. In some cases, the audio signals may first be converted from analog audio signals to digital audio signals before being forwarded to the control system 110. However, as set forth herein, providing audio signals to the control system 110 may include providing audio signals that have been digitized or otherwise adapted. Accordingly, providing audio signals to the control system 110 may include providing signals based on the audio signals generated by the microphones 108.

[0035] In one embodiment, upon receiving audio signals from microphones 108, control system 110 may filter the audio signals. Accordingly, control system 110 may include one or more filters. The audio signals are passed through the filters to remove background noise. The filters may also remove audio signals that fall within specific frequency bands. After the audio signals have been passed through the one or more filters, they are ready for further processing and analysis by control system 110.

[0036] In one embodiment, the audio signals may be passed through filters before being sent to the control system 110. Accordingly, the microphones 108 may be connected to one or more electrical filters or other circuitry to reduce noise before the audio signals are sent to the control system 110. This allows the control system 110 to more easily identify arcing based on the audio signals.

[0037] In one embodiment, control system 110 identifies arcing by analyzing the audio signals in a time domain. Accordingly, control system 110 performs a time-domain analysis on the audio signals to identify waveforms typical of arcing in cleanroom environment 102.

[0038] In one embodiment, the control system 110 identifies arcing by analyzing the audio signals in the frequency domain. Accordingly, the control system 110 may perform frequency-domain analysis on the audio signals. The frequency-domain analysis may include performing a Fourier transform on the audio signals. The Fourier transform takes audio signals from the time domain and provides a frequency-domain representation of the audio signals. The frequency-domain analysis may assist in identifying sounds typical of arcing in the cleanroom environment 102. The Fourier transform may include a fast Fourier transform, a discrete Fourier transform, or other types of Fourier transforms. The frequency-domain analysis may include performing a wavelet analysis on the audio signals.In one embodiment, the control system 110 uses a combination of time domain analysis and frequency domain analysis to identify arcing from the audio signals.

[0039] In one embodiment, control system 110 includes an analysis model. The analysis model is trained using a machine learning process to identify arcing in cleanroom environment 102. The analysis model can analyze the audio signals or signals based on the audio signals to identify arcing in cleanroom environment 102.

[0040] In one embodiment, the machine learning process is a supervised machine learning process. Supervised machine learning processes are typically distinguished from unsupervised machine learning processes by the use of labels in the learning process. For example, if a machine learning process trains an analytics model to classify data items as belonging to one of a plurality of possible categories, a supervised machine learning process uses labeled data items. The labeled data items are a training set for the machine learning process. The label for a data item indicates the correct category for that data item. In the supervised machine learning process, the analytics model is trained to classify the data items in a manner consistent with the labels. In this way, the machine learning process is a supervised process.

[0041] The supervised machine learning process for the analysis model of control system 110 may include generating training set data. Generating training set data may include capturing audio signals known to represent arcing events. Generating training set data may include capturing audio signals known to not be arcing sounds. The training set data may be labeled to identify audio signals that represent arcing and audio signals that do not represent arcing.

[0042] The known arc events for the training set data may include various types of arc events between different objects or locations within the cleanroom environment 102. The known arc events may also include an arc contacting the semiconductor wafer 104 or the device holding or supporting the semiconductor wafer 104.

[0043] The training set data may include audio signals in the time domain and the frequency domain. The training set data may include audio signals after various frequency-domain transforms, such as fast Fourier transforms, discrete Fourier transforms, wavelet transforms, or other types of frequency-domain transforms.

[0044] The supervised machine learning process may involve passing the training set data through the analysis model. The analysis model classifies each audio signal from the training set as either an arc flash event or a non-arc flash event. The classifications are then compared to the labels. The parameters or weights of one or more classification functions of the analysis model are adjusted, and the audio signals from the training set are then passed through the analysis model again. The classifications are then compared to the labels, and the functions are adjusted accordingly. In this process, iterations are repeated until the analysis model is able to reliably classify the audio signals in a manner that matches the labels.

[0045] In one embodiment, the audio signals are converted into a vector form before being passed to the analysis model. Accordingly, the analysis model analyzes a vectorized representation of the audio signals in the time domain and / or the frequency domain. Thus, analyzing the audio signals with the analysis model may include analyzing vectorized variants of the audio signals or other types of representations of the audio signals.

[0046] In one embodiment, the analysis model is a classifier model, a multi-class classifier model, or a binary classifier model. The analysis model may also be a classifier model based on a neural network. The classifier model may be a logistic regression model, a naive Bayes model, a stochastic gradient descent model, a k-nearest neighbor model, a decision tree model, a random forest model, or another type of classifier model. The analysis model may also include other types of classifier models or models other than the classifier models without departing from the scope of the present invention.

[0047] In one embodiment, the analytics model is an unsupervised machine learning model. Unsupervised machine learning processes typically do not use labeled data. Thus, a training set for an unsupervised machine learning process includes data elements for which no labels are available. Using unsupervised machine learning processes, an analytics model can be trained to recognize structures and similarities between data elements. This can be used to group data elements.

[0048] The analysis model of the control system 110 can be trained using an unsupervised machine learning process to detect structures in audio signals that indicate arcing in the cleanroom environment 102. After training, the analysis model can be used to classify audio signals that are similar to each other. This can be used to group audio signals into arcing and non-arcing audio signals.

[0049] When the control system 110 receives audio signals, they can be filtered or formatted and then provided to the analysis model. The analysis model analyzes the audio signals and determines whether any portion of an audio signal represents arcing in the cleanroom environment 102. The analysis model outputs an arcing or lack of arcing indication for each audio signal sample passed through the analysis model.

[0050] The control system 110 may also utilize structure recognition systems other than a machine-learning analysis model to detect arcing based on audio signals. A rule-based analysis model is coded by a person skilled in the art with rules that define how data elements should be analyzed and classified. Thus, a rule-based analysis model may be coded to classify data elements based on established rules, rather than training an analysis model with a machine learning process. The analysis model of the control system 110 may utilize a rule-based system to detect audio signals. The control system 110 may utilize rule-based analysis systems or other types of analysis systems to identify time-domain or frequency-domain structures or features indicative of arcing in the cleanroom environment 102.

[0051] In one embodiment, when control system 110 identifies arcing in cleanroom environment 102, it may take various real-time actions in response to the arcing. Control system 110 may take actions to reduce or otherwise manage arcing in cleanroom environment 102 in real time.

[0052] In one embodiment, control system 110 may determine whether or not arcing is at a level that is hazardous to semiconductor wafer 104. If arcing is at a level that is not yet hazardous to semiconductor wafer 104, control system 110 may refrain from taking corrective action. Alternatively or additionally, control system 110 may issue a message indicating non-hazardous levels of arcing in cleanroom environment 102. The message may be received and viewed by personnel monitoring semiconductor processes in cleanroom environment 102.

[0053] In one embodiment, control system 110 may determine that the level of arcing is hazardous. In this case, control system 110 may take steps to reduce or prevent further arcing in cleanroom environment 102. Control system 110 may be communicatively coupled to semiconductor process system 106 or to devices associated with semiconductor process system 106. Control system 110 may control semiconductor process system 106 to adjust parameters associated with the semiconductor process to reduce or prevent further arcing in cleanroom environment 102.

[0054] In one example, the semiconductor process is a plasma-based process, such as plasma etching, plasma-enhanced chemical vapor deposition, or plasma-enhanced physical vapor deposition. The semiconductor process system 106 may include devices that perform or facilitate the plasma-based process. The control system 110 may cause the semiconductor process system 106 to adjust parameters of the plasma-based process. For example, in plasma-based processes, high voltages are often applied in the cleanroom environment 102 or to gases introduced into the cleanroom environment 102. The control system 110 may cause a reduction in these voltages to reduce or eliminate dangerous arcing in the cleanroom environment 102.Alternatively, the control system 110 may adjust other parameters of the plasma-based semiconductor process to reduce or eliminate hazardous arcing in the clean room environment 102.

[0055] In one example, arcing may occur between a wall of the cleanroom environment 102 and devices or materials within the cleanroom environment 102. The arcing may result from a buildup of dielectric material on a wall or on the devices within the cleanroom environment 102. The control system 110 may cause an automated system to clean or otherwise remove portions of the cleanroom environment 102 associated with the arcing. This may contribute to reducing the arcing.

[0056] In one example, an arc may occur in part due to temperatures or deposition rates during various semiconductor processes. Control system 110 may reduce temperatures or adjust deposition rates in response to detecting arcing.

[0057] In one example, an arc flash may occur in part due to the concentrations of various types of gases in the cleanroom environment 102 during semiconductor processing. The control system 110 may cause the semiconductor processing system 106 to adjust the introduction of one or more gases into the cleanroom environment 102 to reduce or prevent dangerous arcing.

[0058] In one example, the control system 110 may cause the semiconductor process system 106 to completely halt a semiconductor process in response to detecting arcing in the cleanroom environment 102. To avoid severe damage to the semiconductor wafer 104, the control system 110 may, in some cases, determine that the best course of action is to completely halt the semiconductor process until other adjustments can be made to reduce or prevent arcing in the cleanroom environment 102.

[0059] Fig. Figure 2 is an illustration of an arc detection system 100 according to one embodiment. The arc detection system 100 includes a semiconductor wafer 104 positioned on a wafer carrier 116 in the cleanroom environment 102. A plurality of microphones 108 are positioned in the cleanroom environment 102. The arc detection system 100 detects arcing in the cleanroom environment 102.

[0060] The microphones 108 are positioned throughout the cleanroom environment 102. The microphones 108 are positioned laterally around the semiconductor wafer 104 and the wafer carrier 116. The microphones 108 are also positioned above the semiconductor wafer 104 and below the wafer carrier 116. Although Fig. 2, one or more of the microphones 108 may also be positioned in contact with the wafer carrier 116 or with other devices in the clean room environment 102.

[0061] In one embodiment, each microphone 108 included in Fig. 2, correspond to a group of microphones. Each group of microphones may include microphones sensitive to sounds in different frequency ranges. For example, each group of microphones may include one or more microphones sensitive to infrasound. Each group of microphones may include one or more microphones sensitive to audible sound. Each group of microphones may include one or more microphones sensitive to ultrasound. Additionally or alternatively, each group of microphones may include microphones pointing in different directions.

[0062] In the presentation of Fig. 2, an arc event 118 occurs in a specific area of ​​the semiconductor wafer 104. Each of the microphones 108 receives sound waves generated by the arc event 118. The microphones 108 each generate sound signals based on the sound waves. The microphones 108 can transmit the sound signals to a control system 110 (in Fig. 2 not shown). The control system 110 detects the arc event 118 based on the audio signals. The control system 110 can also identify the location of the arc event 118 based on the audio signals and the known positions of the microphones 108.

[0063] In the example of Fig. 2, the arc event 118 has come into contact with a specific portion of the semiconductor wafer 104. The control system 110 can identify the position of the arc event 118 on the semiconductor wafer 104. The control system 110 can determine that the arc has come into contact with the semiconductor wafer 104 because the analysis model can be trained to recognize the sound of the arc coming into contact with the semiconductor wafer 104. The control system 110 can determine the location of the arc event 118 by analyzing the audio signals from each of the plurality of microphones 108 with known positions. The position of the origination of a sound can be determined by the strength of the corresponding audio signals from each microphone 108. The position of the arc event 118 corresponds to a specific portion of the semiconductor wafer 104.The specific position corresponds to one or more integrated circuit dies that will later be separated from the semiconductor wafer 104. The control system 110 can flag the individual integrated circuits originating from this region of the semiconductor wafer 104 as potentially damaged. These integrated circuits can be discarded after singulation.

[0064] In one embodiment, the control system 110 may determine, based on the characteristics of the audio signals, that the arcing event 118 likely caused significant damage to the semiconductor wafer 104. The control system 110 may issue a warning that a technician should inspect the semiconductor wafer 104. Alternatively, the control system 110 may determine that the semiconductor wafer 104 should be discarded.

[0065] In one embodiment, the sound waves of an arc event 118 vary depending on various factors associated with the arc event 118. For example, sound waves associated with an arc event may vary due to a type of gas present in the cleanroom environment 102 during the growth or deposition of a thin film for the semiconductor wafer 104. Arc sound patterns may vary depending on process parameters, such as a voltage present in the cleanroom environment 102 or a vacuum level in the cleanroom environment 102. Arc sound patterns may also vary depending on the type of semiconductor process, which may include wet etches, dry etches, physical vapor deposition, chemical vapor deposition, or other types of processes.Arc sound patterns may also vary depending on the type of damaged material, such as an oxide layer, a photoresist layer, a metal layer, a semiconductor layer, or other types of materials associated with the semiconductor wafer 104. Arc sound patterns may also vary depending on the type of damage, such as the size or shape of the damaged area or other factors associated with damage. Arc sound patterns may also vary depending on the occurrence of chipping of the semiconductor wafer 104 or other structures or devices positioned within the cleanroom environment 102.

[0066] The control system 110 can be trained to detect tone structures associated with all the various parameters of arc events. The control system 110 can diagnose and adjust semiconductor process parameters based on the specific tone structures associated with the various types of arc events.

[0067] In one embodiment, control system 110 can detect whether an arc has come into contact with wafer carrier 116. Control system 110 can identify a location on wafer carrier 116 that has been affected by an arc event. The location can be detected based on the audio signals and the known positions of microphones 108. Control system 110 can detect arcing and the location of the arcing at other locations and relative to other materials or devices positioned within cleanroom environment 102.

[0068] Although it is in Fig. 2, each microphone 108 or group of microphones may be connected to additional circuitry including wired or wireless communication devices, power sources, energy harvesters, filters, signal processors, or other electronic circuitry that may assist in detecting arcing in the cleanroom environment 102.

[0069] Fig. 3 is a diagram of a semiconductor wafer 104 showing locations where sustained arc events occur, according to one embodiment. Control system 110 detects arc events and their locations on a semiconductor wafer 104. Control system 110 records the locations of the arc events on semiconductor wafer 104.

[0070] The Fig. The representation of the semiconductor wafer 104 shown in Figure 3 includes a plurality of raster lines. The raster lines represent dicing lines along which the semiconductor wafer 104 is diced. Accordingly, the raster lines represent individual integrated circuit dies that result when the semiconductor wafer 104 is diced.

[0071] The Fig. The representation of the semiconductor wafer 104 shown in Figure 3 includes arc indicators 120. The arc indicators 120 indicate the positions of arc events on the semiconductor wafer 104. Accordingly, the arc indicators 120 indicate portions of the semiconductor wafer 104 that have been affected by arc events.

[0072] The arc indicators 120 extend across various integrated circuit dies resulting from the singulation of the semiconductor wafer 104. Based on the positions of the arc indicators 120, the control system 110 can register which integrated circuit dies are affected by the arc events. These affected integrated circuit dies can be discarded after singulation. Alternatively, the affected integrated circuit dies can be subjected to further testing before singulation.

[0073] Fig. 4 shows a plurality of curves 130, 132, 134, and 136 associated with audio signals generated by microphones 108 of an arc detection system 100 according to one embodiment. Curve 130 is a time-domain curve of audio signals recorded by microphones 108 in a clean room environment 102 during a period in which no arcing occurred. Curve 130 is a graphical representation of the amplitude of sound waves versus time. The control system 110 is trained or otherwise configured to recognize that the waveform of curve 130 represents a period without arcing. The absence of arcing is represented by the relatively constant noise level.

[0074] Curve 132 is a time-domain curve of audio signals recorded by microphones 108 in the cleanroom environment 102 during a period in which an arc occurred in the cleanroom environment 102. Curve 132 is a graphical representation of the amplitude of sound waves as a function of time. Curve 132 has a structure different from that of curve 130. The control system 110 is trained or otherwise configured to recognize that the waveform of curve 132 corresponds to a period in which an arc occurred. The arc tones correspond to the peaks of the audio signals at time periods t1 and t2 in curve 132. In practice, arc tones may have characteristics different from those shown in curve 132.

[0075] Curve 134 is a frequency-domain curve of audio signals recorded by microphones 108 in the cleanroom environment 102 during a period in which no arcing occurred in the cleanroom environment 102. Curve 134 is a graphical representation of the amplitude of sound waves as a function of frequency. The control system 110 is trained or otherwise configured to recognize that the waveform of curve 134 represents a period without arcing. The absence of arcing is represented by the absence of pronounced frequency peaks.

[0076] Curve 136 is a frequency-domain curve of audio signals recorded by microphones 108 in the cleanroom environment 102 during a period in which an arc occurred in the cleanroom environment 102. Curve 136 is a graphical representation of the amplitude of sound waves as a function of frequency. The control system 110 is trained or otherwise configured to recognize that the waveform of curve 136 corresponds to a period in which an arc occurred. The arc tones correspond to the peaks of the audio signals in frequency ranges f1 and f2. In practice, arc tones may have characteristics different from those shown in curve 136.

[0077] Fig. 5 is a block diagram of an audio signal output circuit 148 of an arc detection system 100 according to one embodiment. In one embodiment, the audio signal output circuit 148 includes one or more microphones 108; a filter 140; a signal processor 142; a communication system 144; and a power source 146. The audio signal output circuit 148 may correspond to a circuit package positioned within a clean room environment 102.

[0078] In the Fig. 1 and Fig. 2, an arc detection system 100 includes a plurality of microphones 108 positioned within the clean room environment 102. In one embodiment, the arc detection system 100 may include a respective audio signal output circuit 148 instead of a single microphone 108 or a group of microphones 108. Thus, each microphone 108 or group of microphones 108 is part of a respective audio signal output circuit 148.

[0079] Each microphone 108 or group of microphones 108 receives audio signals, generates audio signals, and forwards the audio signals to a filter 140. The filter 140 filters the noise from the audio signals and forwards the filtered audio signals to the signal processor 142. The signal processor 142 processes the audio signals, for example, by digitizing the audio signals, and provides the audio signals to the communication system 144. The communication system 144 outputs the audio signals to the control system 110. The communication system 144 can be a wireless or a wired communication system. The power source 146 supplies power to the audio signal output circuit 148. The power source 146 can be an energy harvesting device. Although Fig. 5, portions of the control system 110 may be included in the audio signal output circuit 148. The audio signal output circuit 148 may have other components or configurations without departing from the scope of the present invention.

[0080] Fig. 6 is a block diagram of a control system 110 of an arc detection system 100 according to one embodiment. The control system 110 includes: a communication system 150; a filter 152; a signal processor 154; an analysis model 156; and a process controller 158.

[0081] The communication system 150 communicates with the microphones 108 or with the audio signal output circuit 148. The communication system 150 may be a wireless communication system that communicates wirelessly with the audio signal output circuit 148. The communication system 150 may be a wired communication system that communicates with the audio signal output circuit 148 via a wired connection. The communication system 150 receives audio signals from the microphones 108.

[0082] Communication system 150 provides the audio signals to filter 152. Filter 152 may filter background noise from the audio signals. Filter 152 may also filter out selected frequency bands from the audio signals. Control system 110 may include multiple filters 152.

[0083] The signal processor 154 receives the audio signals from the filter 152. The signal processor 154 processes the audio signals. The signal processor 154 can generate time-domain representations of the audio signals. The signal processor 154 can generate frequency-domain representations of the audio signals, as discussed above. The signal processor 154 can convert the audio signals into a format that can be used by the analysis model 156.

[0084] The analysis model 156 analyzes the audio signals and determines whether the audio signals indicate that an arc has occurred in the cleanroom environment 102. The analysis model 156 may be trained using a machine learning process, as described above with reference to Fig. 1. In the machine learning process, the analysis model 156 can be trained to identify various types of arcing in the cleanroom environment 102 based on the audio signals.

[0085] The process controller 158 communicates with and controls the semiconductor process system 106. For example, if the analytical model 156 indicates that an arc is present in the cleanroom environment 102, the process controller 158 can cause the semiconductor process system 106 to adjust a semiconductor process in real time. The process controller 158 can also cause the semiconductor process system 106 to stop the semiconductor process in response to detecting arcing in the cleanroom environment 102.

[0086] The control system 110 may include components and systems located within the cleanroom environment 102, as well as components located outside the cleanroom environment 102. One skilled in the art, in light of the present invention, will recognize that a control system 110 may include other components and systems, or other configurations of components and systems, without departing from the scope of the present invention.

[0087] Fig. 7 is a flow diagram of a method 700 for detecting arcing in a semiconductor cleanroom environment according to one embodiment. In a step 702 of the method 700, audio signals are generated from sound waves using an array of microphones positioned in a semiconductor processing cleanroom environment. An example of the array of microphones includes the microphones 108 of Fig. 1. An example of a semiconductor processing cleanroom environment is cleanroom environment 102 of Fig. 1. In a step 704, the method 700 includes processing the audio signals with a control system. An example of the control system is the control system 110 of Fig. 1. In a step 706, the method 700 includes detecting arcing in the semiconductor processing clean room environment with the control system based on the audio signals.

[0088] Fig. 8 is a flow diagram of a method 800 for detecting arcing in a semiconductor cleanroom environment according to one embodiment. In a step 802, the method 800 includes performing a semiconductor process on a semiconductor wafer in a semiconductor processing cleanroom environment. An example of the semiconductor wafer is the semiconductor wafer 104 of Fig. 1. An example of the semiconductor processing cleanroom environment is the cleanroom environment 102 of Fig. 1. In a step 804, the method 800 includes converting sounds into audio signals during the semiconductor process with a plurality of microphones positioned in the semiconductor processing cleanroom environment. An example of the arrangement of microphones includes the microphones 108 of Fig. 1. In a step 806, the method 800 includes detecting arcing in the semiconductor processing cleanroom environment by processing the audio signals with a control system. An example of the control system is the control system 110 of Fig. 1. In a step 808, the method 800 includes adjusting a semiconductor process in the semiconductor processing cleanroom environment in response to detecting arcing in the semiconductor processing cleanroom environment.

[0089] In one embodiment, a method comprises generating audio signals from sound waves using an array of microphones positioned in a semiconductor processing cleanroom environment. The method further comprises processing the audio signals with a control system. The method further comprises detecting arcing in the semiconductor processing cleanroom environment with the control system based on the audio signals.

[0090] In one embodiment, a method comprises performing a semiconductor process on a semiconductor wafer in a semiconductor processing cleanroom environment; and converting sounds into audio signals during the semiconductor process with a plurality of microphones positioned in the semiconductor processing cleanroom environment. The method further comprises detecting arcing in the semiconductor processing cleanroom environment by processing the audio signals with a control system. The method further comprises adjusting the semiconductor process in the semiconductor processing cleanroom environment in response to detecting arcing in the semiconductor processing cleanroom environment.

[0091] In one embodiment, a system includes a plurality of microphones positioned in a semiconductor processing cleanroom environment and configured to detect sounds in the semiconductor processing cleanroom environment and generate audio signals based on the sounds. The system further includes a control system communicatively coupled to the plurality of microphones and configured to detect arcing in the semiconductor processing cleanroom environment based on the audio signals.

Claims

[1] Method (700) comprising the following steps: Generating (702) audio signals from sound waves with an array of microphones (108) positioned in a semiconductor processing clean room environment (102); Processing (704) the audio signals with a control system (110); and Detecting (706) arcing in the semiconductor processing clean room environment (102) with the control system (110) based on the audio signals, wherein the arrangement of microphones (108) comprises: one or more microphones (108) configured to detect infrasound; one or more microphones (108) configured to detect audible sound; and one or more microphones (108) configured to detect ultrasound. [2] The method (700) of claim 1, wherein processing (704) the audio signals comprises performing a time domain analysis on the audio signals. [3] The method (700) of any preceding claim, wherein processing (704) the audio signals comprises performing a frequency domain analysis on the audio signals. [4] The method (700) of claim 3, wherein performing the frequency domain analysis comprises performing a Fourier transform and / or a wavelet analysis. [5] The method (700) of any preceding claim, wherein detecting (706) arcing comprises determining that the audio signals correspond to acoustic characteristics of the arcing. [6] The method (700) of any preceding claim, wherein processing (704) the audio signals comprises filtering the audio signals. [7] Method (700) according to one of the preceding claims, further comprising: Determining a position of arcing in the semiconductor processing cleanroom environment (102) based on the audio signals and positions of the microphones (108) in the semiconductor processing cleanroom environment (102). [8] The method (700) of any preceding claim, further comprising adjusting, with the control system (110), a semiconductor process in the semiconductor processing clean room environment (102) in response to detecting arcing. [9] The method (700) of claim 8, wherein adapting the semiconductor process comprises one or more of the following steps: Adjusting an inflow of one or more gases into the semiconductor processing cleanroom environment; Adjusting pressure in the semiconductor processing cleanroom environment; Adjusting a voltage or electric field used in the semiconductor processing cleanroom environment; and Stopping the semiconductor process. [10] Method (700) according to one of the preceding claims, further comprising: Training an analytical model of a computing system with a machine learning process to detect arcing in the semiconductor processing cleanroom environment; and Detecting arcing by analyzing the audio signals with the analysis model. [11] The method (700) of any preceding claim, further comprising detecting whether the arc has come into contact with a semiconductor wafer (104) in the semiconductor processing clean room environment (102). [12] The method (700) of any preceding claim, further comprising identifying a position of arcing on the semiconductor wafer (104). [13] System with: a plurality of microphones (108) positioned in a semiconductor processing clean room environment (102) and configured to detect sounds in the semiconductor processing clean room environment (102) and generate audio signals based on the sounds; and a control system (110) communicatively connected to the plurality of microphones (108) and configured to detect arcing in the semiconductor processing clean room environment (102) based on the audio signals, wherein the plurality of microphones (108) comprises: one or more microphones (108) configured to detect infrasound; one or more microphones (108) configured to detect audible sound; and one or more microphones (108) configured to detect ultrasound. [14] The system of claim 13, wherein the control system (110) includes an analysis model (156) trained with a machine learning process to detect arcing based on the audio signals. [15] The system of any one of claims 13 to 14, wherein the plurality of microphones (108) are arranged such that the control system (110) can determine a position of arcing in the semiconductor processing clean room environment (102) based on the audio signals. [16] The system of any one of claims 13 to 15, further comprising a wireless communication system (144) communicatively connecting the microphones (108) to the control system (110). [17] Method (800) comprising the following steps: Performing (802) a semiconductor process on a semiconductor wafer (104) in a semiconductor processing clean room environment (102); Converting (804) sounds into audio signals during the semiconductor process with a plurality of microphones (108) positioned in the semiconductor processing clean room environment (102); Detecting (806) arcing in the semiconductor processing clean room environment (102) by processing the audio signals with a control system (110); and Adjusting (808) the semiconductor process in the semiconductor processing cleanroom environment (102) in response to detecting arcing in the semiconductor processing cleanroom environment (102), wherein the plurality of microphones (108) comprises: one or more microphones (108) configured to detect infrasound; one or more microphones (108) configured to detect audible sound; and one or more microphones (108) configured to detect ultrasound. [18] The method (800) of claim 17, wherein processing (802) the audio signals comprises processing digital or analog signals generated from the audio signals. [19] The method (800) of claim 17 or 18, further comprising determining, with the control system (110), whether the semiconductor wafer (104) is damaged by the arcing.

Images