BACKGROUND
The manufacturing of the semiconductor devices with an increased device density is becoming increasingly complicated. Among the various semiconductor processing steps, planarization or polishing schemes, e.g., chemical mechanical polish (CMP) has been widely used for thinning or polishing a processed surface of the semiconductor device. The polishing is performed with the help of slurry to facilitate the polishing efficiency and performance. The performance of the polishing operation is therefore closely related to the quality of the slurry.
BRIEF DESCRIPTION OF THE DRAWINGS
Aspects of the present disclosure are best understood from the following detailed description when read with the accompanying figures. It is emphasized that, in accordance with the standard practice in the industry, various features are not drawn to scale. In fact, the dimensions of the various features may be arbitrarily increased or reduced for clarity of discussion.
FIG. 1 is a schematic drawing illustrating a chemical mechanical polish (CMP) operation system, according to a comparative embodiment in the present disclosure.
FIG. 2 is a diagram showing a detection result regarding a concentration of an additive in a slurry during a period of time, according to a comparative embodiment in the present disclosure.
FIG. 3A shows a flowchart of a method of performing a CMP operation, in accordance with some embodiments.
FIG. 3B shows a flowchart of a method of performing a CMP operation, in accordance with some embodiments.
FIG. 4A is a schematic drawing illustrating a chemical mechanical polish (CMP) operation system, in accordance with some embodiments of the present disclosure.
FIG. 4B is a schematic block diagram of a method of performing a CMP operation, in accordance with some embodiments of the present disclosure.
FIG. 5A is a cross sectional view of a CMP apparatus, according to some embodiments of the present disclosure.
FIG. 5B is a top view of a CMP apparatus, according to some embodiments of the present disclosure.
FIG. 5C is a top view of a platen of a CMP apparatus, according to some embodiments of the present disclosure.
FIG. 5D is a top view of a platen of a CMP apparatus, according to some embodiments of the present disclosure.
FIG. 5E is a top view of a platen of a CMP apparatus, according to some embodiments of the present disclosure.
FIG. 5F is a top view of a platen of a CMP apparatus, according to some embodiments of the present disclosure.
FIG. 5G is a top view of a platen of a CMP apparatus, according to some embodiments of the present disclosure.
FIG. 5H is a top view of a platen of a CMP apparatus, according to some embodiments of the present disclosure.
FIG. 6A is a schematic diagram showing a method for training a computational model, in accordance with some embodiments.
FIG. 6B shows a flowchart of a training method for enhancing a CMP operation, in accordance with some embodiments.
FIG. 7A is a schematic drawing illustrating a chemical mechanical polish (CMP) operation system, in accordance with some embodiments of the present disclosure.
FIG. 7B is a schematic block diagram of a method of performing a CMP operation, in accordance with some embodiments of the present disclosure.
DETAILED DESCRIPTION
The following disclosure provides many different embodiments, or examples, for implementing different features of the provided subject matter. Specific examples of components and arrangements are described below to simplify the present disclosure. These are, of course, merely examples and are not intended to be limiting. For example, the formation of a first feature over or on a second feature in the description that follows may include embodiments in which the first and second features are formed in direct contact, and may also include embodiments in which additional features may be formed between the first and second features, such that the first and second features may not be in direct contact. In addition, the present disclosure may repeat reference numerals and/or letters in the various examples. This repetition is for the purpose of simplicity and clarity and does not in itself dictate a relationship between the various embodiments and/or configurations discussed.
Further, spatially relative terms, such as “beneath,” “below,” “lower,” “above,” “upper” and the like, may be used herein for ease of description to describe one element or feature’s relationship to another element(s) or feature(s) as illustrated in the figures. The spatially relative terms are intended to encompass different orientations of the device in use or operation in addition to the orientation depicted in the figures. The apparatus may be otherwise oriented (rotated 90 degrees or at other orientations) and the spatially relative descriptors used herein may likewise be interpreted accordingly.
Notwithstanding that the numerical ranges and parameters setting forth the broad scope of the disclosure are approximations, the numerical values set forth in the specific examples are reported as precisely as possible. Any numerical value, however, inherently contains certain errors necessarily resulting from the deviation normally found in the respective testing measurements. Also, as used herein, the terms “about,” “substantial” or “substantially” generally mean within 10%, 5%, 1% or 0.5% of a given value or range. Alternatively, the terms “about,” “substantial” or “substantially” mean within an acceptable standard error of the mean when considered by one of ordinary skill in the art. Other than in the operating/working examples, or unless otherwise expressly specified, all of the numerical ranges, amounts, values and percentages such as those for quantities of materials, durations of times, temperatures, operating conditions, ratios of amounts, and the likes thereof disclosed herein should be understood as modified in all instances by the terms “about,” “substantial” or “substantially.” Accordingly, unless indicated to the contrary, the numerical parameters set forth in the present disclosure and attached claims are approximations that can vary as desired. At the very least, each numerical parameter should at least be construed in light of the number of reported significant digits and by applying ordinary rounding techniques. Ranges can be expressed herein as being from one endpoint to another endpoint or between two endpoints. All ranges disclosed herein are inclusive of the endpoints, unless specified otherwise.
The terms “couple,” “coupled” and “coupling” used throughout the present disclosure describe the direct or indirect connections between two or more devices or elements. In some cases, a coupling between at least two devices or elements refers to mere electrical or conductive connections between them and intervening features may be present between the coupled devices and elements. In some other cases, a coupling between at least two devices or elements may involve physical contact and/or electrical connections.
The term “parameter” used throughout the present disclosure is not limited to a single numerical value. The term “parameter” may be referred to a group of values, a series of values, value or values derived from a plurality of factors, or different levels of data (such as groups and subgroups of data).
Chemical mechanical polish (CMP) is a skill for smoothing a non-uniform surface during a fabrication operation. In order to sustain the quality of production, it is important to monitor and address the factors that could influence the outcome and reliability of CMP operations. In some cases, the variation of a slurry regarding the chemical composition, such as a chemical property or a physical property, may affect the reliability of CMP operations. The details are discussed in FIG. 1 to FIG. 2 below.
Referring to FIG. 1, FIG. 1 is a schematic drawing illustrating a chemical mechanical polish (CMP) operation system, according to a comparative embodiment in the present disclosure. A CMP operation system 100′ includes a chemical supplying apparatus 1′ and a process apparatus 2′. The chemical supplying apparatus 1′ includes a supplier 1s′ configured to provide slurry to a CMP apparatus 2a′ of the process apparatus 2′ through conduit 3′. The process apparatus 2′ has a slurry arm 14′ configured to dispense slurry that is utilized in the CMP operation of a workpiece 99. However, in some cases, the process apparatus 2′ may be far from the chemical supplying apparatus 1′ and a length of the conduit 3′ may be relatively large (e.g. apart by a distance D1 as shown in FIG. 1). In some cases, the slurry may include an additive for facilitating the CMP operation, such as H2O2 (hydrogen peroxide), which can be utilized as an oxidizer. However. H2O2 is an unstable chemical and may decompose over a period of time.
Referring to FIG. 1 and FIG. 2, FIG. 2 is a diagram showing a detection result regarding a concentration of an additive in a slurry during a period of time, according to a comparative embodiment in the present disclosure. In some cases, when transmitting the slurry with the additive (such as H2O2 or other suitable additive) through the conduit 3′, a concentration of the additive (such as H2O2) may decrease or having a decreasing trend after a certain period of time, as shown in FIG. 2. In some other cases, when the manufacturing operation utilizing the slurry with the additive (such as H2O2) is temporarily suspended, or, when a certain amount of slurry is stored but was not dispensed during idle time, a concentration of the additive (such as H2O2) may also decrease. The decrease of concentration of the additive (such as H2O2) may be an important factor that causes the decay of the slurry chemical compositions. Performing a CMP operation with such decayed slurry composition may result in lowered polishing rate or lowered removal ability, which causes longer polishing time, fall short of completing the CMP operation, or may not properly polish the target layer to comply with predetermined requirement.
Therefore, it is important to address the aforesaid issue of the decay of additive in the slurry chemical compositions, such as regarding the concentration of the additive (e.g. H2O2) in the slurry, in order to facilitate the CMP operation and maintain the reliability of manufacturing operations. However, in fabrication lab practice, the supply of the additive to the slurry is mostly performed at the chemical supplying apparatus 1′ (the delivery end) and may be conducted manually with low frequency (such as daily or twice a day), which faces the difficulty of addressing the decay issue quickly and effectively. In some cases, it is found that even by supplying the additive at the delivery end, the composition of the slurry proximal to the process apparatus 2′ (the process end) and/or in the conduit 3′ may not promptly change accordingly due to the long distance, wherein the CMP operation utilizing the decayed slurry may be undesirably performed across multiple wafers, multiple batches of wafers, or lots of wafers. The failure of recognizing the problem caused by the decay issue and other related factors (such as input parameters A-1 to A-4 as discussed with reference to FIG. 4B, FIG. 6A to FIG. 6B and FIG. 7B) may be one of the important reasons that the conventional approach in the field of CMP techniques would face the issue of non-uniform product quality. In addition, it is found difficult to accurately control the polish rate with the conventional approach due to the complexed factors influencing the slurry chemical.
Accordingly, the present disclosure provides a method of chemical mechanical polish operation and chemical mechanical polish operation system to address the aforesaid issue, which will be subsequently discussed in FIG. 3A to FIG. 7B. Specifically, the chemical mechanical polish operation system discussed in the present disclosure may help compensating for the decay of the additive in the slurry (such as the decrease of the additive concentration) in quicker manner and in a more effective way. The approach discussed in the present disclosure may be able to control the polish rate in relatively accurate manner, thereby reduce the variation regarding the manufacturing result caused by variation of a slurry chemical, especially in the scale of wafer-to-wafer, batch-to-batch and/or lot-to-lot.
Referring to FIG. 3A, FIG. 3A show a flowchart of a method 1000 of performing a CMP operation, in accordance with some embodiments. The method 1000 for performing a CMP operation includes obtaining a first parameter, wherein the first parameter is associated with an additive of a slurry (operation 1007, which can be referred to FIG. 4A to FIG. 4B or FIG. 7A to FIG. 7B), determining a second parameter associated with the process apparatus based on the first parameter, wherein the second parameter is different from the first parameter (operation 1013, which can be referred to FIG. 4B, FIG. 6A or FIG. 7B), securing a workpiece by a head over a platen in the process apparatus, supplying the slurry with the additive over the platen, and polishing a surface of the workpiece by operating the process apparatus configured with the output parameter (operation 1018, which can be referred to FIG. 4A to FIG. 4B or FIG. 7A to FIG. 7B). In some embodiments, the first parameter in FIG. 3A can be referred to as parameters A-1 and A-2 in FIG. 4A to FIG. 7B, the second parameter in FIG. 3A can be referred to as parameters B in FIG. 4A to FIG. 7B.
Referring to FIG. 3B, FIG. 3B show a flowchart of a method of performing a CMP operation, in accordance with some embodiments. The method 2000 for performing a CMP operation includes obtaining a first input parameter and a second input parameter, wherein the first input parameter is associated with an additive of a slurry, and the second input parameter is associated with a characteristic of a process apparatus (operation 2007, which can be referred to FIG. 4A to FIG. 5H or FIG. 7A to FIG. 7B), determining an output parameter based on the first input parameter and the second input parameter (operation 2013, which can be referred to FIG. 4B or FIG. 7B), and securing a workpiece by a head over a platen in the process apparatus, applying the slurry with the additive over the platen, and polishing a surface of the workpiece by operating the process apparatus configured with the output parameter (operation 2018, which can be referred to FIG. 4A or FIG. 7A).
Referring to FIG. 4A, FIG. 4A is a schematic drawing illustrating a chemical mechanical polish (CMP) operation system, in accordance with some embodiments of the present disclosure. A CMP operation system 100 includes a chemical supplying apparatus 1 and a process apparatus 2. The chemical supplying apparatus 1 includes a supplier 1s configured to provide a slurry to a CMP apparatus 2a of the process apparatus 2 through conduit 3. The supplier 1s may be: a pump, a drum pump, a storage device, a chemical cabinet, a bottle, or the like.
The process apparatus 2 includes a platen 11 and a pad 12 disposed over the platen 11. The process apparatus 2 has a slurry arm 14 configured to dispense the slurry that is utilized in the CMP operation of a workpiece 99 (such as a wafer or a substrate), which can be secured by a head 13 during such CMP operation. The platen 11 and the head 13 rotate during the CMP operation, wherein in some embodiments, the platen 11 rotates in a first direction FD, and the head 13 rotates in a second direction SD. In some embodiments, the first direction FD is identical to the second direction SD, such as both are clockwise or counterclockwise. In some alternative embodiments, the first direction FD is reverse to the second direction SD. In some embodiments, the process apparatus 2 further includes a manifold 6 configured to distribute and control the inflow of the slurry in the conduit 3, and a flow controller 7 configured to control the flow rate of the slurry in the conduit 3. The process apparatus 2 may further include a dresser 19 for conditioning the pad 12 by removing residues. The dresser 19 applies a downforce on the pad 12 to facilitate the conditioning of a top surface of the pad 12 due to greater friction.
In some of the embodiments, the slurry may include an additive for facilitating the CMP operation, such as H2O2 (hydrogen peroxide) can be utilized as oxidizer. As previously discussed in FIG. 2, a weight concentration of H2O2 may decrease over time or after being transmitted through the conduit 3. In some alternative embodiments, the slurry may include other types of additives, wherein the concentration of such additives in the slurry may vary over time or when ambient condition changes.
It should be noted that although the weight concentration of additives are utilized as an example in the present disclosure, the volume percent, molecular percentage or other expressions representative of the concentration of specific chemicals in the slurry can be applied as well.
The conduit 3 includes several portions, such as a first portion 3a in (or proximal to) the chemical supplying apparatus 1, a second portion 3b between the chemical supplying apparatus 1 and the manifold 6 of the process apparatus 2, a third portion 3c between the manifold 6 and the flow controller 7. and a fourth portion 3d between the flow controller 7 and an exit end 14E of the slurry arm 14.
Referring to FIG. 4A and FIG. 4B, FIG. 4B is a schematic block diagram of a method of performing CMP operation, in accordance with some embodiments of the present disclosure. The CMP operation system 100 may further include at least one of sensors 5a, 5b, 5c and 5d, wherein the sensor 5a is attached to the first portion 3a of the conduit 3, the sensor 5b is attached to the second portion 3b of the conduit 3, the sensor 5c is attached to the third portion 3c of the conduit 3, and the sensor 5d is attached to the fourth portion 3d of the conduit 3. The sensors 5a, 5b, 5c, 5d can be configured to obtain a parameter of the slurry, for example, a concentration of the additive (such as H2O2) in the slurry. In some alternative embodiments, in order to obtain the concentration of the additive at delivery end (e.g. a first position proximal to the chemical supplying apparatus 1) and the concentration of the additive at the process end (e.g. a second position proximal to the process apparatus 2), the CMP operation system 100 includes the sensor 5a at a first position proximal to the chemical supplying apparatus 1 and at least one of the sensors 5b, 5c, and 5d at a second position proximal to the process apparatus 2. In some other alternative embodiments, the sensors 5a, 5b, 5c and 5d can also be configured to obtain a concentration of abrasive (or certain chemical) in the slurry. In some other alternative embodiments, the sensors 5a, 5b. 5c. 5d can also be configured to obtain a temperature of the slurry or other properties of the slurry at their corresponding positions.
Alternatively stated, referring to the method 3000 shown in the schematic block diagram in FIG. 4B, in operation 3011, a value of a parameter A-1, which is associated with the slurry at a first position proximal to the chemical supplying apparatus 1 (such as a concentration of the additive (e.g. as H2O2) or certain chemical in the slurry) can be obtained by the sensor 5a. In operation 3012, a value of a parameter A-2, which is associated with the slurry at a second position proximal to the process apparatus 2 (such as a concentration of the additive (e.g. as H2O2) or certain chemical in the slurry) can be obtained by the sensors 5b, 5c, and/or 5d. In some alternative embodiments, the parameter A-1 and/or the parameter A-2 can also include (or can be) a concentration of an abrasive (or another chemical) in the slurry, a temperature of the slurry or other properties of the slurry at certain position.
The sensors 5a, 5b, 5c, 5d can be electrically connected to a controller 4, thereby allowing the sensors 5a, 5b, 5c, 5d to transmit a signal associated with the parameters A-1 and A-2 of the slurry to the controller 4. Alternatively stated, the controller 4 is configured to receive a signal associated with values of the parameters A-1 and A-2 from the sensors 5a, 5b, 5c, 5d. Herein the controller 4 in the present disclosure can be implemented by software such that the foregoing methods disclosed therein can be automatically performed. For a general purpose computer, the software routines can be stored on a storage device, such as a permanent memory. Alternately, the software routines can be machine executable instructions stored using any machine readable storage medium, such as a diskette, CD-ROM, magnetic tape, digital video or versatile disk (DVD), laser disk, ROM, flash memory, etc. The series of instructions could be received from a remote storage device, such as a server on a network. The present disclosure can also be implemented in hardware systems, microcontroller unit (MCU) modules, discrete hardware or firmware. Furthermore, the controller 4 may include a timer and/or a memory to store a preset schedule.
Referring to FIG. 4B, in operation 3013, a value of a parameter A-3 is obtained based on a condition of ambient environment (such as in fabrication lab, or at a position between the chemical supplying apparatus 1 and the process apparatus 2) by using a sensor 8a and/or in the process apparatus 2 by using a sensor 8b. In some embodiments, the parameter A-3 may include at least one of ambient temperature, ambient humidity, ambient pressure, or the like. The parameter A-3 may be measured prior to or during a CMP operation on a workpiece 99. It should be noted that the ambient factor in ambient environment or in the process apparatus 2 can change from time to time. The controller 4 is configured to receive a signal associated with a value of the parameter A-3 from the sensors 8a and/or 8b.
Still referring to FIG. 4B, in operation 3014, a value of a parameter A-4 associated with a characteristic of the chemical mechanical polish apparatus 2a is obtained and transmitted to the controller 4 (shown in FIG. 4A). Referring to FIG. 4A and FIG. 5A, FIG. 5A is a cross sectional view of a CMP apparatus, according to some embodiments of the present disclosure. The CMP apparatus 2a of the process apparatus 2 in the CMP operation system 100 further includes a first thermal meter 21, a second thermal meter 22 and/or a third thermal meter 23 to obtain a value of the aforementioned parameter A-4 discussed in FIG. 4B. The controller 4 is configured to receive a signal associated with the value of the parameter A-4 from the first thermal meter 21, the second thermal meter 22 and/or the third thermal meter 23. In some embodiments, the CMP operation system 100 may further include a chiller conduit 11CH disposed in the platen 11, wherein working fluid (such as cooling/heating chemical) can be circulated or transported in the chiller conduit 11CH. In some embodiments, a portion of the chiller conduit 11CH is routed in turns in order to facilitate the temperature control over the platen 11 and the pad 12 thereon.
In some embodiments, the head 13 includes a retaining ring 13R configured to secure the workpiece 99 during CMP operation, a top portion 13T over the position of workpiece 99 when the workpiece 99 is secured by the head 13, and a membrane 13MB between the top portion 13T and the pad 12 over the platen 11. The membrane 13MB may be made of soft material and can be configured to serve as a buffer layer to alleviate undesired damage on a first surface 99S of the workpiece 99 facing the top portion 13T of the head 13 (wherein the first surface 99S is a surface opposite to a second surface 99Q to be polished by the pad 12).
One or more first thermal meter 21 can be disposed at the top portion 13T of the head 13 of the CMP apparatus 2a, and the first thermal meter 21 may be facing toward the pad 12 and the membrane 13MB. In some embodiments, the first thermal meter 21 is configured to obtain a temperature of the membrane 13MB proximal to the top portion 13T of the head 13. In some embodiments, a detection surface 21D of the first thermal meter 21 overlaps with the membrane 13MB in vertical direction Z In some of the embodiments, the first thermal meter 21 is a non-contact temperature measurement device, such as infrared thermal meter or other suitable radiation-based sensing device. In some embodiments, a distance t21 between the first surface 99S of the workpiece 99 and the detection surface 21D of the first thermal meter 21 is less than about 5 cm. When the distance t21 is greater than 5 cm, the accuracy of temperature measurement may decrease.
In some embodiments, one or more second thermal meter 22 can be disposed at the retaining ring 13R, which is at a peripheral area of the head 13. In some of the embodiments, the second thermal meter 22 is at least partially embedded in the retaining ring 13R, and has a detecting surface 22D facing the pad 12. In some embodiments, the second thermal meter 22 is configured to obtain a temperature of the slurry on the pad 12 (or the temperature of the pad 12 in some alternative embodiments). In some of the embodiments, the second thermal meter 22 is a non-contact temperature measurement device, such as infrared thermal meter or other suitable radiation-sensing device. In some embodiments, a distance t22 between the top surface 12T of the pad 12 and the detection surface 22D of the second thermal meter 22 is less than about 5 cm. When the distance t22 is greater than 5 cm, the accuracy of temperature measurement may decrease.
One or more third thermal meter 23 can be disposed in the platen 11 and under the pad 12. In some embodiments, the third thermal meter 23 has a detection surface 23D facing the pad 12. In some embodiments, the platen 11 includes one or more channels 11C disposed therein, which can be configured to accommodate the third thermal meter(s) 23. In some embodiments, the channels 11C is an empty receptacle. In some alternative embodiments, the channels 11C can also be a receptacle filled with fillers. The third thermal meter 23 can be configured to obtain a temperature of the slurry on the pad 12 (or in some alternative embodiments, the temperature of the pad 12). In some of the embodiments, the third thermal meter 23 is a non-contact temperature measurement device, such as infrared thermal meter or other suitable radiation-sensing device. In some embodiments, at least a portion of the pad 12 (e.g. the portions directly above the third thermal meter(s) 23) is substantially transparent or semitransparent for the wavelength of the radiation utilized by the third thermal meter 23. For example, the transparency thereof may be in a range from about 10% to about 100%, thereby allowing at least a portion of the radiation to pass through the pad 12. In some of the embodiments, the transparency of the portion in the pad 12 may be in a range from 30% to about 70% for some of the materials that may be suitable to be fabricated as the pad 12. With such configuration, the radiation emitted from the third thermal meter 23 may be reflected by the head 13. The temperature obtained by the third thermal meter 23 is determined by one or more of a temperature of the head 13, a temperature of the pad 12, a temperature of the slurry, and the heat released during the CMP operation. In some alternative embodiments, the pad 12 is made of an opaque material, wherein the temperature obtained by the third thermal meter 23 is mainly determined by a temperature of the pad 12, however, herein a temperature of the pad 12 may be influenced by a temperature of the head 13, a temperature of the slurry, and the heat released during the CMP operation.
In some embodiments, the first thermal meter 21, the second thermal meter 22, and the third thermal meter 23 measure temperatures during CMP operation (starting from the initialization of the CMP operation, which is after the slurry is supplied, and end at the termination of the CMP operation), and for each type of the thermal meters, the sampling of temperature data may be conducted periodically or triggered by event, and the temperature can be obtained by averaging multiple temperature data at different sampling time points through the thermal meter, or can be substituted/supported by other specific analysis methods (such as filtering out the outlier data, sampling peak values, obtaining trending lines, et cetera).
Referring to FIG. 5A and FIG. 5B, FIG. 5B is a top view of a CMP apparatus, according to some embodiments of the present disclosure. In some embodiments, the first thermal meter 21 and the second thermal meter 22 rotate around a rotation axis 13X of the head 13 in a first direction FD. and in some of the embodiments, the head 13 may move along direction X and direction Y (wherein direction X and direction Y are orthogonal to vertical direction Z) over the pad 12. In some embodiments, the third thermal meter 23 rotates around a rotation axis 11X in a second direction SD, wherein the second direction SD is identical to the first direction FD. In some embodiments, the rotation axis 13X is offset from the rotation axis 11X. In some alternative embodiments, the second direction SD is reverse to the first direction FD.
In some embodiments, the third thermal meter 23 can be arranged in various fashions to comply with specific requirements. Several embodiments will be subsequently described as examples in FIG. 5C to FIG. 5H, wherein each of the FIG. 5C to FIG. 5H is an example of a top view of a platen of a CMP apparatus, according to some embodiments of the present disclosure. In some embodiments as shown in FIG. 5C, the third thermal meter 23 can be disposed in the platen 11. In some embodiments as shown in FIG. 5D, a plurality of third thermal meters 23 can be arranged in array. In some embodiments as shown in FIG. 5E, a plurality of third thermal meters 23 can be arranged in a line. In some embodiments as shown in FIG. 5F, a plurality of third thermal meters 23 can be arranged in two or more parallel lines. In some embodiments as shown in FIG. 5G, a plurality of third thermal meters 23 can be arranged to form one or more circles. In some embodiments as shown in FIG. 5H, a plurality of third thermal meters 23 can be arranged in one or more curves C1, such as a spiral, an Archimedean spiral, oval, or the like. In the examples discussed in FIG. 5D to FIG. 5H. having multiple third thermal meters 23 may help improving the accuracy of temperature measurement due to more data sampling points available. In some cases, the polish rate at different areas of the workpiece 99 may be different. Accordingly, some of the third thermal meters 23 may be apart from rotation axis 11X by different distances. In some embodiments, in order to further improve the accuracy of temperature measurement, a distance between two third thermal meters 23 can be less than 800 mm when a diameter of the platen 11 is from about 711.2 mm to about 747.4 mm (or from about 28 inches to 31 inches); however, the present disclosure is not limited thereto, whereas such distance may depend on the size of the platen 11.
Referring back to FIG. 4B and FIG. 5A, in operation 3021, values of the parameter A-1, parameter A-2, parameter A-3. parameter A-4, et cetera, is obtained by the controller 4, and the controller 4 further includes a processor for executing a computational model to conduct parameter determination in order to obtain a plurality of parameters B (which will be subsequently discussed in operation 3031 and operation 3032) thereby facilitating the CMP operation. The parameter determination helps forecasting the effect of input parameters (the parameter A-1, parameter A-2, parameter A-3, parameter A-4, et cetera) that influence the performance of the CMP operation, as well as obtaining the weight of each aforesaid parameters that contribute to the performance of the CMP operation. For example, such parameter determination may include utilizing neural network, machine learning, big data mining, data optimization, or other suitable computational models that can relate the inputs and outputs. In some embodiments, the controller 4 decides whether a set of values of the input parameters (the parameter A-1, parameter A-2, parameter A-3, parameter A-4, et cetera) comply with a predetermined specification. Herein the predetermined specification may associate with the evaluator parameters (Parameters C) as discussed in FIG. 6A to FIG. 6B.
In operation 3031 (as shown in FIG. 4B), a plurality of output parameters (parameters B) can be determined based on values of the input parameters (the parameter A-1, parameter A-2, parameter A-3, parameter A-4, et cetera) obtained through the controller 4. In some embodiments, the values of the output parameters (parameters B) are determined in response to a condition that a set of the input parameters (such as parameters A-1 to A-4) fails to comply with a predetermined specification. In some embodiments, the parameters B can be parameters associated with one or more parameters of the process apparatus 2 for performing the CMP operation, such as a rotational speed of the head 13 (shown in FIG. 4A and FIG. 5A), a rotational speed of the platen 11, a head pressure applied by the head 13 for pressing the workpiece 99 against the pad 12, a downforce applied by the dresser 19 against the pad 12, a temperature of working fluid in the chiller conduit 11CH disposed in the platen 11, a flow rate of working fluid in the chiller conduit 11CH disposed in the platen 11, et cetera. Herein the downforce is referred to a force that the dresser 19 applies on the pad 12 to facilitate the conditioning of a top surface of the pad 12 due to greater friction. In some embodiments, the types of the output parameters (parameters B) may be different from those of the input parameters (the parameter A-1, parameter A-2, parameter A-3, parameter A-4, et cetera). In some embodiments, conducting the adjustment on values of one or more of the aforesaid output parameters (parameters B) is more efficient and/or more effective than the adjustment merely on some of the input parameters, for example, adjusting a concentration of the additive (e.g. as H2O2) in the slurry at a position proximal to the chemical supplying apparatus 1. Therefore, the adjustment of the aforesaid output parameters (parameters B) may help addressing the variation of input parameters (the parameter A-1, parameter A-2. parameter A-3. parameter A-4. et cetera) with a timescale of wafer-to-wafer, batch-to-batch, or lot-to-lot, or even in real time in some embodiments, thus may be able to make adjustment in a timely manner to improve the stability and uniformity of CMP operation. In some embodiments, adjusting one of the parameters B may help substantially alleviating an influence on a CMP operation from the variations of parameter A-1, parameter A-2. parameter A-3, and/or parameter A-4. However, in some other embodiments, the variations of parameter A-1, parameter A-2. parameter A-3. parameter A-4, or even another parameter have a complicated interrelation with the performance of a CMP operation. In such case, adjusting a plurality of the parameters B may help compensating for the performance variation of such CMP operation from the variations of parameter A-1, parameter A-2. parameter A-3. and/or parameter A-4.
Referring to FIG. 4A and FIG. 6A. FIG. 6A is a schematic diagram showing a method for training a computational model, in accordance with some embodiments. In some embodiments, the computational model implemented by the controller 4 for the parameter determination can be obtained by using techniques such as neural network, machine learning, big data mining, data optimization, or other suitable computational models, or other suitable computational models that can relate the inputs and outputs. For example, a model type of the computational model can be an artificial neural network (ANN) or other artificial intelligence learning model. As shown in FIG. 6A, the computational model discussed in FIG. 4A to FIG. 4B is constructed by a group of nodes (neurons) interconnected through connections with respective weights. The group of nodes may form various layers, such as an input layer, an output layer, one or more cascaded hidden layers connecting between the input layer and the output layer. In some cases, the computational model 602 further includes memory nodes, convolution/pooling nodes, or kernel nodes. Parameters of the computational model 602 may associate with the number of nodes in the input layer, the output layer, and the hidden layer, as well as the interconnection topology of the connections. In some embodiments, a large amount of parameters A-1, A-2, A-3. and/or A-4 (input parameters for constructing the input nodes 601) and parameters B (output parameters for constructing the output nodes 603) are fed to a machine learning procedure. An iterative training process is performed to optimize the computational model 602, and eventually obtaining a correlation between the input parameters A-1, A-2, A-3. and/or A-4 and the output parameters B as well as the weight of each nodes.
In some embodiments, the technique of reinforcement learning is utilized in order to optimize the computational model 602, such as using an evaluator parameters (Parameters C) which are associated with a performance of a CMP operation performed on a workpiece 99 and can be obtained by measuring feature(s) of the workpiece 99 (shown in FIG. 4A) after performing the CMP operation (or alternatively, can be obtained from empirical data of the CMP operations under various conditions). In some embodiments, the evaluator parameters (parameters C) is associated to a condition of the polished surface of the workpiece 99. In some embodiments, the evaluator parameters (parameters C) may be based on a resulting thickness at one or more positions of the workpiece 99 after the CMP operation, a polished thickness at one or more positions of the workpiece 99 after the CMP operation, and/or a thickness uniformity across at least a portion of the second surface 99Q of the workpiece 99 (shown in FIG. 5A), wherein the positions can include one in a center zone of the workpiece 99, one in a middle zone of the workpiece 99, one in a peripheral zone of the workpiece 99. and/or one at a bevel of the workpiece 99. For example, parameters C can be utilized as evaluator 604 to evaluate a status of the computational model and trigger reiteration. For example, a set of values of parameters C correspond to a set of values of input parameters A-1, A-2, A-3. A-4 and the values of the output parameters B, and the set of values of parameters C is compared to a predetermined goal value or goal range. A result of such comparison triggers reiteration of adjusting the weight of each nodes and thereby causing the computational model to converge. In some embodiments, empirical data of the CMP operation can be leveraged to examine the accuracy and stability of the computational model, for example, by testing the removal rate of the CMP operations or the uniformity of polished material.
The training method discussed in the present disclosure is not limited to reinforcement learning. For example, the computational model may utilize any other suitable types of neural network architectures and learning techniques, such as Convolutional Neural Network (CNN), Recurrent Neural Network (RNN), Long/Short term Memory (LSTM), Gated Recurrent Unit (GRU), Neural Tuning Machine (NTM), Support Vector Machine (SVM), Kohonen Network (KN), Deep Residue Network (DRN). Generative Adversarial Network (GAN), Liquid State Machine (LSM), Extreme Learning Machine (ELM). Echo State Network (ESN). Deconvolutional Network (DN), Deep Convolutional Network (DCN), Deep Convolutional Inverse Graphics Network (DCIGN), Auto Encoder (AE), Variation Auto Encoder (VAE), Denoising Auto Encoder (DAE), Sparse Auto Encoder (SAE) Perception (P), Feed Forward (FF), Radial Basis Network (RBF), Deep Feed Forward (DFF). Markov Chain (MC), Hopfield Network (HN), Boltzmann Machine (BM), Restricted Boltzmann Machine (RBM), Deep Belief Network (DBN), or the like. In some alternative embodiments, the training method discussed in the present disclosure includes a combination of two or more learning techniques.
In operation 3032 (as shown in FIG. 4B), CMP operation can be performed based on values of a plurality of output parameters (parameters B), wherein the workpiece 99 is secured by the head 13 and placed over the pad 12 and the platen 11, wherein the second surface 99Q of the workpiece 99 (shown in FIG. 5A) is polished or planarized by operating the process apparatus 2 with the output parameters (parameters B) configured with the determined value. In some embodiment, the slurry is applied over the pad 12 in order to facilitate the CMP operation, wherein the slurry may include an additive, for example, H2O2. In some embodiments. H2O2 can especially be utilized as a slurry additive when polishing materials including oxide, tungsten (W), copper (Cu), aluminum (Al), polysilicon, gallium nitride (GaN), or materials utilized in micro eletro-mechanical system (MEMS) structures. In some embodiments, the additive is configured with the parameter A-2. Parameter(s) B may include one or more parameters associated with parameters of the process apparatus 2, such as a rotational speed of the head 13, a rotational speed of the platen 11, a head pressure applied by the head 13 for pressing the workpiece 99 against the pad 12, a downforce applied by the dresser 19 against the pad 12, a temperature of working fluid in the chiller conduit 11CH disposed in the platen 11, and/or a flow rate of working fluid in the chiller conduit 11CH disposed in the platen 11, et cetera.
In some embodiments, the platen 11 and/or the head 13 (shown in FIG. 4A) can be electrically connected to the controller 4. thereby the controller 4 can adjust a rotational speed of the head 13, a rotational speed of the platen 11, a head pressure applied by the head 13 for pressing the workpiece 99 against the pad 12, and/or a downforce applied by the dresser 19 against the pad 12 based on values of parameters B derived from parameters A-1 to A-4. In some embodiments, the controller 4 is configured to increase/decrease a temperature of working fluid in the chiller conduit 11CH disposed in the platen 11 and/or to increase/decrease a flow rate of working fluid in the chiller conduit 11CH disposed in the platen 11 by controlling a temperature controlling mechanism for the working fluid (such as cooler/heater disposed at the chiller conduit 11CH or at the supplier of the working fluid) and a chiller flow controller configured to control flow rate, based on values of parameters B derived from parameters A-1 to A-4. In some of the embodiments, generally, a polishing rate is positively correlated to slurry temperature.
In some of the embodiments, operation 3041 and operation 3042 as shown in FIG. 4B can be performed to further improve the CMP operation. In some embodiments, in operation 3041, evaluator parameters (parameters C) associated with a performance of a CMP operation can be obtained by measuring feature(s) of the workpiece 99 (shown in FIG. 4A) after performing a CMP operation, and in operation 3042, a signal of values of the evaluator parameters (parameters C) can be transmitted to the controller 4 in order to adjust the computational model for parameter determination conducted by the controller 4. In some embodiments, the evaluator parameters (parameters C) may be based on a resulting thickness at one or more positions of the workpiece 99 after the CMP operation, a polished thickness at one or more positions of the workpiece 99 after the CMP operation, and/or a thickness uniformity across at least a portion of the second surface 99Q of the workpiece 99 (shown in FIG. 5A), wherein the positions can include one in a center zone of the workpiece 99, one in a middle zone of the workpiece 99, one in a peripheral zone of the workpiece 99, and/or one at a bevel of the workpiece 99. In some embodiments, the aforesaid evaluator parameters (parameter C) can be obtained in a measurement phase 41 (shown in FIG. 4A), in which the measurement are performed by the sensors 5a, 5b, 5c, 5d.
Referring back to FIG. 6A, the evaluator parameters (parameters C) associated with a performance of a CMP operation is obtained by measurement, and can be input as an evaluator 604. In the cases that the evaluator 604 deviates from a predetermined goal value or is not within a predetermined goal range, reiteration is triggered and the computational model of parameter determination can be adjusted, thereby optimizing the computational model 602 in timely manner.
In some alternative embodiments, the aforesaid evaluator parameters (parameters C) can instead be obtained during the CMP operation in real time, wherein the CMP apparatus 2a includes the sensors 5a, 5b, 5c, 5d. Such configuration may further improve the accuracy of controlling the CMP operation on the workpiece 99 since the computational model of parameter determination can be adjusted in real time accordingly.
Alternatively stated, a feedback operation can be used to enhance the training result of the computational model, as shown in FIG. 6B. Referring to FIG. 6B, FIG. 6B shows a flowchart of a training method for enhancing a CMP operation, in accordance with some embodiments. The method 4000 for enhancing a CMP operation includes collecting parameters A-1, A-2, A-3. A-4. parameters B, and parameters C (operation 4011), training a computational model utilizing parameters A-1, A-2, A-3. A-4. the parameters B, and the parameters C (operation 4013), securing a workpiece by a head over a platen in a process apparatus, supplying the slurry over the platen, and polishing a surface of the workpiece by operating the apparatus that executes the computational model (operation 4015), and collecting the parameter C (operation 4017). After collecting the parameter C of a workpiece 99 in each of one or more CMP operations, the computational model can be updated based on the obtained parameters C. For example, the weight of each connections as discussed in FIG. 6A may be adjusted based on the obtained parameters C through the feedback operation. Particularly, in some cases the outcome of polished workpiece 99 may fluctuate after a period of time and may not be in compliance with the requirements. Therefore, by utilizing the feedback operation as discussed above, the computational model can be calibrated in a timely manner.
The aforesaid feedback operation may allow the controller 4 to adjust the computational model for parameter determination of the controller 4, for example, adjusting the weight of each parameter B in the computational model, wherein each of the parameters B in the current and/or the following CMP operations (i.e. on another workpiece) may be adjusted accordingly. Therefore, the determination of relation between the variations of input parameters and the output parameters can be adjusted in time when performing CMP operation on the next workpiece 99 (also can be referred to the second workpiece) or the next batch/next lot.
Referring to FIG. 7A and FIG. 7B, FIG. 7A is a schematic drawing illustrating a chemical mechanical polish (CMP) operation system, and FIG. 7B is a schematic block diagram of a method of performing CMP operation, in accordance with some embodiments of the present disclosure. The CMP operation system 200 discussed in FIG. 7A and the method 3000B shown in FIG. 7B are respectively similar to the CMP operation system 100 discussed in FIG. 4A and the method 3000A shown in FIG. 4B, however the difference resides in that the CMP operation system 200 further includes an auxiliary compensation apparatus 9 configured to at least partially compensate for the influence of parameter A-1 and parameter A-2 at the process end (e.g. the second position proximal to the process apparatus 2).
In some embodiments, the auxiliary compensation apparatus 9 may include one or more auxiliary compensation device at the process end proximal to the process apparatus 2, for example, a first auxiliary compensation device 9a disposed at the second portion 3b of the conduit 3, a second auxiliary compensation device 9b disposed at the third portion 3c of the conduit 3, and/or a third auxiliary compensation device 9c disposed at the fourth portion 3d of the conduit 3, wherein the second portion 3b is between the chemical supplying apparatus 1 and the manifold 6 of the process apparatus 2, the third portion 3c is between the manifold 6 and the flow controller 7, and the fourth portion 3d is between the flow controller 7 and an exit end 14E of the slurry arm 14. In some embodiments, the CMP operation system 200 includes one of the first auxiliary compensation device 9a. the second auxiliary compensation device 9b, and the third auxiliary compensation device 9c; while in some alternative embodiments, the CMP operation system 200 includes more than one of first auxiliary compensation device 9a, the second auxiliary compensation device 9b, and the third auxiliary compensation device 9c. In some embodiments, the auxiliary compensation apparatus 9 is electrically connected to the controller 4.
The auxiliary compensation apparatus 9 can be utilized to supply an amount of an additional additive to the slurry in the conduit 3 before the slurry is dispensed from the exit end 14E of the slurry arm 14. In some embodiments, the amount of the additive (such as H2O2) supplied by the auxiliary compensation apparatus 9 is based on the parameter A-1,which is associated with the slurry at the first position proximal to the chemical supplying apparatus 1 (such as a concentration of the additive (e.g. as H2O2) or another chemical in the slurry) and/or a parameter A-2, which is associated with the slurry at the second position proximal to the process apparatus 2 (such as a concentration of the additive (e.g. as H2O2) or another chemical in the slurry). The detailed description can be found by referring back to the discussion for FIG. 3A to FIG. 6B. In some embodiments, the auxiliary compensation apparatus 9 may directly supply a pure additive that is consisting essentially of the additive to the conduit 3. In some alternative embodiments, the auxiliary compensation apparatus 9 may supply the slurry with an additive concentration greater than the additive concentration measured at the second position at the process end (for example, the additive concentration where parameter A-2 is detected).
In some embodiments, the auxiliary compensation apparatus 9 is configured to perform operation 3015 as shown in FIG. 7B, including obtaining a first amount of additive to be added to the slurry based on parameter A-1 and/or parameter A-2, supplying the first amount of additive to slurry, and obtaining a parameter A-5 based on the first amount, the parameter A-1, and the parameter A-2. For example, the first amount of the additive can be determined by a difference between the additive concentration at the delivery end and the additive concentration at the process end, thereby the decay of the additive in the slurry can at least be partially compensated for by the first amount of the additive. Further, a signal based on a parameter A-5, which is obtained based on the first amount, the parameter A-1, and the parameter A-2, can be transmitted to controller 4 for conducting parameter determination. Specifically, the operation 3015 may relieve the extent of influence caused by the decay of the additive in the slurry, and further enable the assessment of the compensation to be conducted by adjusting the parameters B since the “first amount” of the additive is also taken under consideration for parameter determination operation in operation 3021. Alternatively stated, the additive of the slurry applied over the platen 11 during a CMP operation is configured with the input parameter, which is associated with the parameter A-1, the parameter A-2, along with a compensation from the parameter A-5.
It should be further noted that, in some embodiments, the auxiliary compensation apparatus 9 provides a storage that alleviates or reduce the decay of the additive (such as H2O2), for example, by providing a controlled environment with a constant pressure and/or a constant temperature and isolated from the influence of the ambient environment, et cetera.
The location-specific configuration of the auxiliary compensation apparatus 9 proximal to the dispense arm 14 helps addressing the issue of decay of the additive in the slurry due to the long transmission distance between the delivery end and the process end, as well as the idle time, as previously discussed in FIG. 1 and FIG. 2.
The present disclosure provides a method of chemical mechanical polish (CMP) operation and a CMP operation system to address the issue of fluctuating performance of CMP operations. The issue may stem from the unstable additive (such as H2O2) in the slurry, wherein a concentration of the additive in the slurry may decrease after being transmitted through a relatively long distance and/or over a period of idle time. The issue may also stem from other variations, such as factors of ambient environment and the conditions of CMP apparatus 2a. As discussed in FIG. 3A to FIG. 5H, in order to compensate for the influence from the above factors, sensors (5a, 5b, 5c. 5d, 8a, and/or 8b) and/or thermal meters (21. 22. and/or 23) are utilized to obtain input parameters (such as parameters A-1 to A-4). Further, output parameters (parameters B) in the CMP operation are adjusted (such as a rotational speed of the head 13 (shown in FIG. 4A and FIG. 5A), a rotational speed of the platen 11, a head pressure applied by the head 13 for pressing the workpiece 99 against the pad 12, a downforce applied by the dresser 19 against the pad 12, a temperature of working fluid in the chiller conduit 11CH disposed in the platen 11, a flow rate of working fluid in the chiller conduit 11CH disposed in the platen 11, et cetera) to improve the outcome of CMP operation. In order to determine a relation between the input parameters (such as parameters A-1 to A-4) and the output parameters (such as parameters B), a technique of parameter determination can be utilized. In some embodiments, the values of the output parameters (parameters B) are determined in response to a condition that a set of the input parameters (such as parameters A-1 to A-4) fails to comply with a predetermined specification. In some embodiments, conducting the adjustment on one or more of the aforesaid output parameters (parameters B) is more efficient and/or more effective than the adjustment merely on some of the input parameters, for example, adjusting a concentration of the additive (e.g. as H2O2) in the slurry at a position proximal to chemical supplying apparatus 1. Therefore, the adjustment of the aforesaid output parameters (parameters B) may help addressing the variation of input parameters (the parameter A-1, parameter A-2, parameter A-3, parameter A-4. et cetera) for wafer-to-wafer, batch-to-batch, or lot-to-lot removal rate variations. In some embodiments, the parameter adjustment can even be done in real time to improve the stability and uniformity of the removal rate in a timely manner. In some embodiments, neural network models, machine learning, big data mining, data optimization, or other suitable computational model can be utilized to improve the performance of the CMP operations, as discussed in FIG. 6A.
Furthermore, a feedback mechanism discussed in FIG. 6B can also be utilized to further tune the computational model for optimizing CMP operation. In some embodiments, the evaluator parameters (parameter C) may include a factor of workpiece 99 under CMP operation, such as thickness at certain positions or thickness uniformity. The feedback mechanism may help adjusting the weight of each of the parameters B in the computational model, wherein each of the parameters B in the current and/or the following CMP operations may be adjusted accordingly.
As discussed in FIG. 7A to FIG. 7B, by utilizing an auxiliary compensation apparatus 9, additional additive can be supplied to the slurry in the conduit 3 before the slurry is dispensed from the exit end 14E of the slurry arm 14. The location-specific configuration of the auxiliary compensation apparatus 9 proximal to the dispense arm 14 helps addressing the issue of decay of the additive in the slurry due to the long distance between the delivery end and the process end.
Some embodiments of the present disclosure provide a method for polishing a workpiece. The method includes obtaining a first input parameter and a second input parameter, wherein the first input parameter is associated with an additive of a slurry, and the second input parameter is associated with a characteristic of a process apparatus, determining an output parameter associated with the process apparatus based on the first input parameter and the second input parameter, securing a workpiece by a head over a platen in the process apparatus, supplying the slurry with the additive over the platen with the additive configured with the first parameter, and polishing a surface of the workpiece by operating the process apparatus configured with the output parameter.
Some embodiments of the present disclosure provide a method for polishing a workpiece. The method includes obtaining a first parameter associated with at least one of an additive of a slurry and a first characteristic of a process apparatus, and a second parameter associated with a second characteristic of the process apparatus, determining a value of the second parameter in response to determining that the first parameter fails to comply with a predetermined specification, securing a workpiece by a head over a platen in the process apparatus, supplying the slurry with the additive over the platen, and planarizing a surface of the workpiece by operating the process apparatus with the second parameter configured with the determined value.
Some embodiments of the present disclosure provide a chemical mechanical operation system. The CMP operation system includes a platen in a process apparatus, a process apparatus including a platen, a conduit, connected to a chemical supplying apparatus configured to supply a slurry, a slurry arm connected to the conduit, wherein the slurry arm is configured to supply the slurry over the platen, a sensor attached to the conduit, wherein the sensor in configured to obtain a concentration of an additive in the slurry, a controller electrically connected to the platen, configured to receive a signal associated with the concentration of the additive in the slurry.
The foregoing outlines features of several embodiments so that those skilled in the art may better understand the aspects of the present disclosure. Those skilled in the art should appreciate that they may readily use the present disclosure as a basis for designing or modifying other operations and structures for carrying out the same purposes and/or achieving the same advantages of the embodiments introduced herein. Those skilled in the art should also realize that such equivalent constructions do not depart from the spirit and scope of the present disclosure, and that they may make various changes, substitutions, and alterations herein without departing from the spirit and scope of the present disclosure.
Patent Application
20230256561
