KR102517666B1 - System for cleaning semiconductor wafers - Google Patents

Abstract

A system for controlling damage when cleaning a semiconductor wafer 1010 that includes features of patterned structures includes a wafer holder 1014 that temporarily restrains the semiconductor wafer 1010 during the cleaning process; a nozzle 1012 for delivering a cleaning solution onto the surface of the semiconductor wafer 1010; To impart sonic energy to the cleaning liquid, at a first frequency (f ₁ ) and a first power level (P ₁ ) during a first predetermined period of time (τ ₁ ) and during a second predetermined period of time (τ ₂ ). A sound wave generator (25082) configured to alternately operate at a second frequency (f ₂ ) and a second power level (P ₂ ), wherein the first predetermined time period (τ ₁ ) and the second predetermined time period (τ ) ₂ ) is the sound wave generator, continuously connected to each other; and a controller programmed to provide cleaning parameters, at least one of which is determined such that a percentage of features damaged as a result of imparting sonic energy is less than a predetermined threshold.

Description

System for cleaning semiconductor wafers

The present invention relates generally to semiconductor wafer cleaning, and more particularly to wet cleaning methods and apparatus using controlled sonic energy.

Semiconductor devices are fabricated or fabricated on semiconductor wafers using a series of processing steps to form transistors and interconnect elements. These transistors are traditionally built in two dimensions, but more recently in three dimensions, such as finFET transistors. Interconnect elements include conductive (eg, metal) trenches, vias, and the like formed in dielectric materials.

In forming these transistors and interconnect devices, semiconductor wafers are subjected to a number of masking, etching and deposition processes to form the desired structures for semiconductor devices. For example, a plurality of masking and plasma etching steps are performed to form recessed regions in a dielectric layer on a semiconductor wafer that function as fins for finFET transistors and trenches and vias for interconnect devices. A wet cleaning step is required to remove particles and contaminants in the fin structures and/or trenches and vias via post etching or photoresist ashing. However, wet cleaning with chemicals can result in sidewall loss. As device fabrication nodes move beyond 14 or 16 nm, reducing sidewall losses in fins, trenches and vias becomes critical to maintaining critical dimensions. To reduce or eliminate sidewall loss, it is important to use only suitable or diluted chemicals and sometimes deionized water. However, suitable or diluted chemicals or deionized water are not efficient enough to remove particles in fin structures and/or trenches and vias. As a result, mechanical forces generated by ultrasound or megasonic energy, for example, are required to efficiently remove these particles. Ultrasound or mega sonic waves generate bubble cavitation to apply mechanical forces to the wafer structure under cleaning.

However, cavitation is a chaotic phenomenon. The initiation of cavitation bubbles and their collapse are influenced by many physical parameters. Violent cavitation, such as transit cavitation or micro jets, can damage these patterned structures (fins, trenches and vias). In conventional ultrasonic or mega sonic cleaning processes, significant particle removal efficiency (“PRE”) only occurs when the power is sufficiently high (eg, greater than 5-10 Watts). However, significant wafer damage begins to occur when the power exceeds about 2 Watts. Therefore, it is difficult to find a power window in which the wafer can be efficiently cleaned without causing significant damage. Thus, maintaining stable or controlled cavitation is the key to controlling the mechanical forces of acoustic waves below the damage limit while being able to efficiently remove debris from the patterned structures.

Accordingly, it is desirable to provide a system and method for controlling bubble cavitation generated by ultrasonic or mega sonic devices during a wafer cleaning process to efficiently remove microscopic contaminants without damaging patterned structures on the wafer.

A system for cleaning a semiconductor wafer includes a wafer holder that temporarily restrains the semiconductor wafer during a cleaning process, an inlet that delivers a cleaning liquid over a surface of the semiconductor wafer, at a first predetermined setting for a first predetermined period of time and at a second predetermined time period. a sound wave generator configured to alternately operate at a second predetermined setting for a determined period of time, and the first and second presets by the sound wave generator at the first and second predetermined settings, the first and second period of time and the sound wave generator; and a controller programmed to determine the number of alternations between the determined settings, wherein bubble cavitation in the cleaning liquid begins to increase during a first predetermined period of time and decrease during a second predetermined period of time. The first predetermined time period and the second predetermined time period successively follow one another. Thus, bubbles in the cleaning liquid can be sufficiently cooled after cleaning in each first time period to avoid damaging the wafer.

Other aspects, features and techniques will be apparent to those skilled in the art in view of the detailed description of the examples below.

The drawings accompanying and forming a part of this specification are included to illustrate certain aspects of the present disclosure. More clear concepts of the present disclosure, and components and operation of systems provided with the present disclosure, will be more readily apparent by reference to the illustrative, and thus non-limiting, embodiments shown in the drawings, and similar reference numerals. refer to the same element (if occurring in more than one figure). The present disclosure may be better understood by reference to one or more of these drawings in conjunction with the description provided herein. Features shown in the drawings are not necessarily drawn to scale.
1A and 1B show a wafer cleaning apparatus using an ultrasonic or mega sonic device according to an embodiment of the present invention.
2a-2g show various shapes of ultrasonic or mega sonic transducers.
3 shows bubble implosion during a wafer cleaning process.
4A and 4B show transit cavitation damaging patterned structures on a wafer during a wafer cleaning process.
5A-5C show the change in thermal energy within a bubble during a sonic wafer cleaning process.
6A-6C illustrate an acoustic wafer cleaning process in which microjets ultimately occur.
7A-7E illustrate a sonic wafer cleaning process according to one embodiment of the present invention.
8A-8D illustrate a sonic wafer cleaning process according to another embodiment of the present invention.
9A-9D illustrate a sonic wafer cleaning process according to another embodiment of the present invention.
10A-10C illustrate a sonic wafer cleaning process according to another embodiment of the present invention.
11A-11B illustrate a sonic wafer cleaning process according to another embodiment of the present invention.
12A-12B illustrate a sonic wafer cleaning process according to another embodiment of the present invention.
13A-13B illustrate a sonic wafer cleaning process according to another embodiment of the present invention.
14A-14B illustrate a sonic wafer cleaning process according to another embodiment of the present invention.
15A-15C illustrate stable cavitation damaging patterned structures on a wafer during an acoustic wafer cleaning process.
15D is a flow diagram illustrating an alternative wafer cleaning process according to one embodiment of the present invention.
16A-16C illustrate a wafer cleaning process according to one embodiment of the present invention.
17 shows a wafer cleaning process according to another embodiment of the present invention.
18a-18j illustrate bubble cavitation control to enhance the circulation of fresh rinse solution within vias or trenches within a wafer.
19A-19D show the change in bubble volume in response to sonic energy.
20A-20D illustrate an acoustic wafer cleaning process that effectively cleans high aspect ratio features of vias and trenches in accordance with one embodiment of the present invention.
21A-21C illustrate another cleaning process according to an embodiment of the present invention.
22A and 22B illustrate a wafer cleaning process using sonic energy according to another embodiment of the present invention.
23 illustrates an exemplary wafer cleaning apparatus for performing the wafer cleaning process shown in FIGS. 7-22 in accordance with an embodiment of the present invention.
24 is a cross-sectional view of another wafer cleaning apparatus for performing the wafer cleaning process shown in FIGS. 7-22 according to an embodiment of the present invention.
25 illustrates a control system for monitoring operating parameters of a wafer cleaning process using sonic energy according to one embodiment of the present invention.
26 is a block diagram of the detection system shown in FIG. 25 according to an embodiment of the present invention.
27 is a block diagram of the detection system shown in FIG. 25 according to another embodiment of the present invention.
28A-28C show an exemplary implementation of the voltage attenuation circuit shown in FIG. 26 according to one embodiment of the present invention.
29A-29C show an exemplary implementation of the shaping circuit shown in FIG. 26 according to one embodiment of the present invention.
30A-30C illustrate an exemplary implementation of the main controller of FIGS. 26 and 27 according to one embodiment of the present invention.
31 shows the acoustic wave power supply oscillating several cycles after the host computer shuts down the acoustic wave power supply.
32A-32C show an exemplary implementation of the amplitude detection circuit of FIG. 27 according to one embodiment of the present invention.
33 is a flowchart illustrating a wafer cleaning process according to an embodiment of the present invention.
34 is a flowchart illustrating a wafer cleaning process according to another embodiment of the present invention.

One aspect of the present disclosure relates to controlling bubble cavitation in semiconductor wafer cleaning with sonic energy. Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings.

1A and 1B show a wafer cleaning apparatus using an ultrasonic or mega sonic device according to an embodiment of the present invention. 1A shows a wafer chuck 1014 holding a wafer 1010, a rotation drive module 1016 driving the wafer chuck 1014, and a nozzle 1012 delivering a cleaning liquid 1032 to the surface of the wafer 1010. It is a cross-sectional view of a wafer cleaning apparatus having a. The cleaning liquid 1032 can be cleaning chemicals or deionized water. In addition, the wafer cleaning device includes an ultrasonic or mega sonic device 1003 positioned above the wafer 1010, so that according to the rotation of the wafer 1010 and the constant flow of the cleaning liquid 1032 from the nozzle 1012, the wafer ( 1010) and the sonic device 1003, a film of cleaning liquid 1032 having a thickness d is held. The sonic device 1003 further includes a piezoelectric transducer 1004 acoustically coupled to the resonator 1008 in contact with the cleaning fluid. The piezoelectric transducer 1004 is electrically excited and vibrates, and the resonator 1008 transmits high-frequency acoustic energy to the cleaning liquid 1032. The bubble cavitation generated by the high-frequency acoustic energy causes foreign particles, i.e., contaminants, on the surface of the wafer 1010 to vibrate and loosen them.

Referring again to FIG. 1A, the wafer cleaning apparatus further includes an arm 1007 coupled to the sonic wave apparatus 1003 to move the sonic wave apparatus 1003 in the vertical direction Z, so as to change the liquid film thickness d. let it A vertical drive module 1006 drives the vertical movement of the arm 1007 . Both the vertical drive module 1006 and the rotary drive module 1016 are controlled by the controller 1088.

Referring to FIG. 1B, which is a plan view of the wafer cleaning apparatus shown in FIG. 1A, the acoustic wave device 1003 covers only a small area of the wafer 1010, which is rotated to receive uniform sonic energy across the entire wafer 10101. Should be. Although only one such sonic device 1003 is shown in FIGS. 1A and 1B , in other embodiments, two or more sonic devices may be used simultaneously or intermittently. Similarly, two or more nozzles 1012 may be used to more uniformly deliver the cleaning liquid 1032.

2a-2g show various shapes of ultrasonic or mega sonic transducers. 2A shows a triangular or pie shape; 2b shows a rectangular shape; Figure 2c shows an octagonal shape; Figure 2d shows an elliptical shape; 2e shows a semi-circular shape; Figure 2f shows a quarter circle shape; Figure 2g shows a perfect circular shape. A sound wave transducer in each of these shapes can be used instead of the piezoelectric transducer 1004 in the sound wave device 1003 shown in FIG. 1 .

3 shows bubble implosion during a wafer cleaning process. The shape of the bubble 3012 is gradually compressed from a spherical shape (A) to an apple shape (G) as sonic energy is applied to the bubble 3012 . Finally, the bubble 3012 reaches an implosion state (I) to form a microjet. As shown in FIGS. 4A and 4B , the microjets are very vigorous (can reach thousands of atmospheric pressures and thousands of degrees C), especially when the feature size t is reduced to 70 nm or less, the wafer 4010 may damage the fine patterned structure 4034 on the image.

4A and 4B show through-cavitation damaging patterned structures on a wafer during a wafer cleaning process. Referring to FIG. 4A , bubbles 4040 , 4042 , and 4044 are formed by sonic cavitation over patterned structure 4034 on semiconductor wafer 4010 . Patterned structure 4034 includes a plurality of features that need to be cleaned, including but not limited to fins, vias, trenches, and the like. Bubbles 4044 are so violent that they are converted into micro-jets that can reach thousands of atmospheric pressures and thousands of degrees Celsius. Referring to FIG. 4B , once the micro-jet is generated, a portion of the patterned structure 4034 is blown away. This damage is more severe for wafers with device feature sizes below 70 nm.

5A-5C show the change in thermal energy within bubble 5016 during a wafer cleaning process. As the sound pressure of the sound wave acting on the bubble 5016, the bubble 5016 reduces its volume as shown in FIG. 5A. During this volume reduction process, the acoustic pressure P _M on the bubble 5016 is forced on the bubble 5016, and the mechanical work is converted into thermal energy within the bubble 5016. Accordingly, the temperature T of the gas and/or vapor inside bubble 5016 increases as shown in FIG. 5B. The relationship between the various parameters can be expressed by the following equation.

where p ₀ is the pressure in the bubble before compression, v ₀ is the initial volume of the bubble 5016 before compression, T ₀ is the temperature of the gas in the bubble before compression, p is the pressure inside the bubble during compression, and v is the volume of the bubble during compression, and T is the temperature of the gas within the bubble during compression.

To simplify the calculations, it can be assumed that the temperature of the gas does not change during compression or that the compression is very slow and that the temperature increase is offset by the liquid surrounding the bubble. Thus, the mechanical work (w _m ) caused by the sound wave pressure (P _M ) during one hour of bubble compression (volume N units to volume 1 unit, or compression ratio=N) can be expressed as:

Here, S is the cross-sectional area of the cylinder, x ₀ is the length of the cylinder, and p ₀ is the pressure of the gas inside the cylinder before compression. Since Equation (2) does not take into account the factor of temperature increase during compression, the actual pressure within the bubble will be higher due to the temperature increase. Therefore, the actual mechanical work by sound wave pressure will be greater than the value calculated by Equation (2).

Assuming that the mechanical work by sound wave pressure is partially converted into thermal energy, and the partially converted mechanical energy of the high-pressure gas and/or vapor in the bubble is assumed, this thermal energy fully contributes to the temperature increase of the gas in the bubble (energy is not transferred to the liquid molecules surrounding the bubble), and assuming that the mass of the gas inside the bubble remains constant before and after compression, the temperature increase ΔT after one hour of compression of the bubble can be expressed as there is.

where Q is the thermal energy converted from mechanical work, β is the ratio of thermal energy to total mechanical work by sound pressure, m is the mass of the gas inside the bubble, and c is the specific heat coefficient of the gas. For hydrogen gas β = 0.65, S = 1E-12 m2, x ₀ = 1000 μm = 1E-3 m (compression ratio N = 1000), p ₀ = 1 kg/cm2 = 1E4 kg/2, m = 8.9E- At 17 kg, c = 9.9E3 J /(kg °k), then ΔT = 50.9 °C.

The temperature of the gas in the bubble after the first compression (T ₁ ) can be calculated as follows.

At this high temperature, some of the liquid molecules surrounding the bubble will evaporate when the bubble reaches a minimum size of 1 micron as shown in FIG. 5B. After that, the sound wave pressure becomes negative and the bubble starts to increase its size. In this reversal process, a hot gas and/or vapor with pressure P _G acts on the surface of the surrounding liquid. At the same time, the sound wave pressure (P _M ) pulls the bubble in the expansion direction as shown in FIG. 5c. Thus, the negative acoustic pressure (P _M ) partially acts on the surrounding liquid. As a result of the joint effect, the thermal energy in the bubble cannot be fully released or converted to mechanical energy, so the gas temperature in the bubble cannot be cooled to the original gas temperature (T ₀ ) or liquid temperature. After the first cycle of cavitation, the temperature T2 of the gas and/or vapor in the bubble will be somewhere between T ₀ and T ₁ as shown in FIG. 6B. Here, T ₂ can be expressed as follows.

where δT is the temperature decrease after one hour of bubble expansion, and δT is less than ΔT.

When the second cycle of bubble cavitation reaches the minimum bubble size, the temperature of the gas and/or vapor inside the bubble (T ₃ ) is

At the end of the second cycle of bubble cavitation, the temperature of the gas and/or vapor in the bubble (T ₄ ) is

Similarly, when the nth cycle of bubble cavitation reaches the minimum bubble size, the temperature of the gas and/or vapor in the bubble (T _{2n - 1} ) is

At the end of the nth cycle of bubble cavitation, the temperature of the gas and/or vapor in the bubble (T _2n ) is

From Equation (8), the number of implosion cycles (n _i ) can be written as:

From Equation (10), the implosion time τ _i can be written as:

Here, t ₁ is the cycle period, and f ₁ is the frequency of the ultrasonic/mega sound wave.

Based on Equations (10) and (11), the number of implosion cycles (n _i ) and the implosion time (τ _i ) can be calculated. Table 1, assuming T _i = 3000 ° C, ΔT = 50.9 ° C, T ₀ = 20 ° C, and f ₁ = 500 KHz, 1 MHz or 2 MHz, the number of implosion cycles (n1), implosion time (τi) and represents the calculated relationship between (ΔT-δT).

[Table 1]

6A-6C show an acoustic wafer cleaning process in which microjets are eventually generated and the process parameters are attached to Equations (1)-(11). Referring to FIG. 6A , electric power P is continuously supplied to the sound wave device to generate bubble cavitation in the cleaning liquid. As the number of cycles n of bubble cavitation increases, the temperature of the gas and/or vapor will increase as shown in FIG. As a result, as shown in FIG. 6C, the size increases over time. Finally, during compression, the temperature inside the bubble 6082 will reach the implosion temperature T _i (typically as high as several thousand degrees Celsius), and a vigorous micro-jet 6080 will occur as shown in FIG. 6C. Therefore, to avoid damage to the patterned structure of the wafer during cleaning, stable cavitation must be maintained and bubble implosion or micro-jets must be avoided.

7A-7E illustrate a sonic wafer cleaning process according to one embodiment of the present invention. 7A shows waveforms of power output intermittently supplied to the sonic wave device to generate bubble cavitation in the cleaning liquid. Figure 7b shows the temperature curve corresponding to each cycle of cavitation. FIG. 7c shows that during each cycle of cavitation, the bubble size increases in a τ ₁ time period and decreases when the power supply is terminated in a τ ₂ time period.

Detailed processing steps for avoiding bubble implosion according to the first embodiment of the present invention are shown in Fig. 7d. The processing steps begin with step 7010 where an ultrasonic or mega sonic device is placed near the upper surface of the wafer under cleaning. At step 7020, a rinse, chemical or gas doped water is injected over the wafer to fill the gap between the wafer and the acoustic device. At step 7030, the wafer carried by the chuck begins to rotate or vibrate. In step 7040, power having a frequency f ₁ and a power level P ₁ is applied to the acoustic wave device. At step 7050, before the temperature of the gas and/or vapor within the bubble reaches the implosion temperature T _i , or before the time τ ₁ reaches τ _i as calculated by Equation (11) Before, since the power output is set to 0, the temperature of the gas and/or vapor in the bubble starts to cool because the temperature of the cleaning liquid is much lower than the gas temperature. In step 7060, after the temperature of the gas and/or vapor within the bubble has decreased to room temperature (T ₀ ) and the time duration has reached τ ₂ (during the time period τ ₂ ), the power supply output is set to zero. ), the power supply output is restored with frequency f ₁ and power level P ₁ . In step 7070, the cleanliness of the wafer is checked, and steps 7010-7060 are repeated if the wafer has not yet been cleaned to the desired degree. Alternatively, the cleanliness check may not be performed every cycle. Alternatively, the number of cycles to be used can be determined empirically beforehand using the sample wafer.

Referring again to FIG. 7D , at step 7050, the time period τ ₁ must be shorter than τ _i to avoid bubble implosion, where τ _i can be calculated using Equation (11). In step 7060, the temperature of the gas and/or vapor within the bubble need not be cooled to room temperature or to the cleaning liquid temperature. Rather, it may be a specific temperature above room temperature or the wash liquid temperature. Preferably, the temperature is significantly lower than the implosion temperature (T _i ).

According to Equations (8) and (9), when (ΔT-δT) is known, the implosion time τ _i can be calculated. However, in general, (ΔT−δT) cannot be easily calculated or measured directly. However, τ _i can be determined empirically.

7E is a flow diagram illustrating the steps for empirically determining implosion time τ _i . In step 7210, five different time periods (τ ₁ ) are illustratively selected as design of experiment (DOE) conditions based on Table 1. In step 7220, the time period τ ₂ is set at least 10 times longer than the selected time period τ ₁ , preferably 100 times longer in the first screening test. In step 7230, the power supply level is fixed at P ₀ to run at the above five DOE conditions to individually clean five different wafers having the same specific patterned structure. where P ₀ is the power level at which the patterned structure will certainly be damaged when run in continuous mode (non-pulsed mode) as shown in FIG. 6A. At step 7240, the damage condition of the five wafers is inspected by a scanning electron microscope (SEM) or a wafer pattern damage review tool such as AMAT SEM vision or Hitachi IS3000 so that the implosion time τ _i is narrowed to a specific range. can The percentage of damaged features can be calculated by dividing the total number of damaged features inspected by the SEM by the total number of features of the patterned structure. There may be other methods for determining the percentage of damaged features. For example, the final wafer yield can be used as an indication of the percentage of damaged features.

Steps 7210 to 7240 described above may be repeated to narrow the range of implosion time τ _i . After knowing the implosion time τ _i , time τ _i can be set to a value less than 0.5*τ _i to allow for a safety margin. The following paragraphs describe examples of such experiments.

Assuming that the patterned structures are formed by 55 nm poly-silicon gate lines, ultrasonic frequencies are used in gap oscillation modes to achieve uniform energy dose within the wafer and from wafer to wafer (see PCT application PCT/CN2008/073471). 1 MHz generated by an ultrasonic/mega sonic device manufactured by Prosys operating at Other experimental parameters and final pattern damage data are summarized in Table 2 as follows.

[Table 2]

In our experiments, when τ ₁ =2 ms (or 2000 cycles), the sonic cleaning process described above introduces 1216 damage sites into the patterned structure with a feature size of 55 nm. When τ ₁ =0.1 ms (or 100 cycles), the sonic cleaning process introduces zero damage sites to the same patterned structure. Therefore, τ _i is a time value between 0.1 ms and 2 ms. Additional testing with a narrower τ ₁ range may yield a narrower τ _i range.

In the experiments described above, the number of cycles depends on the ultrasonic or megasonic power density and frequency, the higher the power density, the fewer cycles, and the lower the frequency, the fewer cycles. From the above experiments, the number of cycles without damage can be predicted to be less than 2,000 when the power density of ultrasonic or mega sonic waves is greater than 0.1 watt/cm ² , and the frequency of ultrasonic or mega sonic waves is 1 MHz or less. When the frequency increases to a range greater than 1 MHz or the power density is less than 0.1 watts/cm ² , the number of cycles can be expected to increase.

After obtaining the time period τ ₁ , the time period τ ₂ can be obtained empirically based on a similar DOE method as described above. In this case, τ ₁ is fixed at a predetermined value, and τ ₂ is progressively shortened at each DOE until damage to the patterned structure is observed. As the time period (τ ₂ ) shortens, the temperature of the gas and/or vapor within the bubble cannot cool sufficiently, which gradually increases the average temperature of the gas and/or vapor within the bubble, eventually triggering implosion of the bubble. will do This trigger time is called the critical cooling time (τ _c ). Knowing the critical cooling time τ _c , the time period τ ₂ can be set to a value greater than 2*τ _c to allow a safety margin.

Accordingly, the parameters of the cleaning process can be determined such that the cleaning effect of imparting sonic energy results in a greater yield improvement than a yield decrease caused by damage as a result of imparting sonic energy. A predetermined threshold for damage percentage may also be specified by the customer, for example. Parameters of the cleaning process may be determined such that the percent damage is below a predetermined threshold, substantially zero, or even zero. The predetermined threshold may be for example 10%, 5%, 2% or 1%. If the final yield of wafer fabrication is not substantially affected by any damage caused by the cleaning process, the damage percentage is substantially zero. That is, any damage caused by the cleaning process is acceptable in terms of the overall manufacturing process. As described above, the percent damage can be determined by inspecting the sample wafer using an electron microscope.

8A-8D illustrate a sonic wafer cleaning process according to another embodiment of the present invention. In the sonic wafer cleaning process of the present invention, instead of being maintained at a constant constant level P ₁ as shown in FIG. 7A and at step 7040 in FIG. 7D the amplitude of power source P varies with time while , other aspects of the process remain the same as shown in Figs. 7a-7d. In one embodiment, as shown in FIG. 8A , the power supply amplitude P increases over a period of time τ ₁ . In another embodiment, as shown in FIG. 8B , the power supply amplitude P decreases during a time period τ ₁ . In another embodiment, as shown in FIG. 8C , the power supply amplitude P first decreases and then increases over a period of time τ ₁ . In the embodiment shown in FIG. 8D , the power supply amplitude P first increases and then decreases during a time period τ ₁ .

9A-9D illustrate a sonic wafer cleaning process according to another embodiment of the present invention. In the sonic wafer cleaning process of the present invention, instead of remaining constant f1 as shown in FIG. 7A and at step 7040 in FIG. 7D the power supply frequency varies with time, while the other The perspective remains the same as shown in Figures 7a-7d. In one embodiment, as shown in FIG. 9A , the power supply frequency is set first at f ₁ and then at f ₃ , where f ₁ is higher than f ₃ during the time period τ ₁ . As shown in FIG. 9B , in one embodiment, the power supply frequency is set first at f ₃ and then f ₁ during the time period τ ₁ . As shown in FIG. 9C , in one embodiment, the power supply frequency changes from f ₃ to f ₁ and then back to f ₃ over a period of time τ ₁ . As shown in FIG. 9D , in one embodiment, the power supply frequency changes from f ₁ to f ₃ and then back to f ₁ over a period of time τ ₁ .

Similar to the cleaning process shown in FIG. 9C , in one embodiment, the power supply frequency is set as first f ₁ , then f ₃ and finally f 4 for a period of time τ ₁ , where f ₄ is greater than f ₃ . small, and f ₃ is smaller than f ₁ .

Again similar to the cleaning process shown in FIG. 9C , in one embodiment, the power supply frequency is set as first f ₄ , then f ₃ and finally f ₄ for a period of time τ ₁ , where f ₄ is f is less than ₃ , and f ₃ is less than f ₁ .

Again similar to the cleaning process shown in FIG. 9C , in one embodiment, the power supply frequency is set as first f ₁ , then f ₄ and finally f ₃ for a period of time τ ₁ , where f ₄ is f is less than ₃ , and f ₃ is less than f ₁ .

Again similar to the cleaning process shown in FIG. 9C , in one embodiment, the power supply frequency is set as first f ₃ , then f ₄ and finally f ₁ for a period of time τ ₁ , where f ₄ is f is less than ₃ , and f ₃ is less than f ₁ .

Again similar to the cleaning process shown in FIG. 9C , in one embodiment, the power supply frequency is set as first f ₃ , then f ₁ and finally f ₄ for a period of time τ ₁ , where f ₄ is f is less than ₃ , and f ₃ is less than f ₁ .

Again similar to the cleaning process shown in FIG. 9C , in one embodiment, the power supply frequency is set as first f ₄ , then f ₁ and finally f ₃ for a period of time τ ₁ , where f ₄ is f is less than ₃ , and f ₃ is less than f ₁ .

10A-10C illustrate a sonic wafer cleaning process according to another embodiment of the present invention. Referring to FIG. 10A , similar to the cleaning process shown in FIG. 7A , a power supply having a level of P ₁ and a frequency of f ₁ is applied to the sound wave device during a time period τ ₁ . However, during the time period τ ₂ , instead of dropping to zero as shown in FIG. 7A, the power supply decreases to the level of P ₂ . As a result, the temperature of the gas and/or vapor in the bubble decreases to T ₀ +ΔT ₂ as shown in FIG. 10B.

FIG. 10C is a flow diagram illustrating the steps of the wafer cleaning process shown in FIGS. 10A and 10B. In step 10010, an ultrasonic or mega sonic device is placed near the upper surface of the wafer under cleaning. In step 10020, a rinse, chemical or gas doped water is injected over the wafer to fill the gap between the wafer and the acoustic device. At step 10030, the chuck carrying the wafer begins to rotate for the cleaning process. In step 10040, a power supply having a frequency of f ₁ and a power level P1 is applied to the acoustic wave device. In step 10050, while maintaining the frequency at f ₁ , before the temperature of the gas and/or vapor within the bubble reaches the implosion temperature T _i , or the time τ ₁ is determined by Equation (11) Before reaching τ _i as calculated, the power supply level is lowered to P2. In step 10060, the power supply level is restored to P1 when the temperature of the gas and/or vapor within the bubble is reduced to near room temperature (T ₀ ) or the time duration reaches τ ₂ . In step 10070, wafer cleanliness is checked, and steps 10010-10060 will be repeated if the wafer has not yet been cleaned to the desired degree. Alternatively, the cleanliness check may not be performed every cycle. Alternatively, the number of cycles to be used can be determined empirically beforehand using the sample wafer.

11A-11B illustrate a sonic wafer cleaning process according to another embodiment of the present invention. This sonic wafer cleaning process is similar to that shown in FIGS. 10A-10C except for the difference present only in step 10050. Instead of holding the power supply frequency at f ₁ , the wafer cleaning process shown in FIGS. 11A and 11B lowers the frequency to f ₂ for a period of time τ ₂ . The power level (P ₂ ) should be significantly less than P1, preferably less than 5 or 10 times, in order to allow the temperature of the gas and/or vapor in the bubble to drop close to room temperature (T ₀ ).

12A-12B illustrate a sonic wafer cleaning process according to another embodiment of the present invention. The only difference between this cleaning process and the process shown in FIGS. 10A-10C is in step 10050. In this wafer cleaning process, the power supply frequency is increased to f ₂ , while the power supply level P ₂ is substantially equal to P ₁ during the time period τ ₂ .

13A-13B illustrate a sonic wafer cleaning process according to another embodiment of the present invention. The only difference between this cleaning process and the process shown in FIGS. 10A-10C is in step 10050. In this wafer cleaning process, the power supply frequency is increased to f ₂ while the power supply level is lowered from P ₁ to P ₂ for a period of time τ ₂ .

14A-14B illustrate a sonic wafer cleaning process according to another embodiment of the present invention. The only difference between this cleaning process and the process shown in FIGS. 10A-10C is in step 10050. In this wafer cleaning process, the power supply frequency is increased from f ₁ to f ₂ , while the power supply level is increased from P ₁ to P ₂ for a period of time τ ₂ . Since the frequency f ₂ is higher than f ₁ , sonic energy heats the bubble less intensely, so the power supply level P ₂ can be slightly higher than P ₁ , but as shown in FIG. 14B , the time period ( It should not be too high to ensure that the temperature of the gas and/or vapor inside the bubble decreases during τ ₂ ).

15A-15C illustrate stable cavitation damaging patterned structures on a wafer during an acoustic wafer cleaning process. Referring to FIG. 15A , a patterned structure 15034 having a spacing W is formed on a wafer 15010 . Some air bubbles 15046 formed in the cavitation process are within the space of the patterned structure 15034. Referring to FIG. 15B , as bubble cavitation continues, the temperature of the gas and/or vapor inside the bubble 15048 increases, causing the bubble 15048 to increase in size. When the size of the bubble 15048 is larger than the gap W, the expansive force of the bubble cavitation may damage the pattern structure 15034 as shown in FIG. 15C. Therefore, a new wafer cleaning process is required.

As shown in FIG. 15C , the area of damage caused by bubble expansion may be smaller than the area of damage caused by bubble implosion as shown in FIG. 4B . For example, bubble expansion can result in damage areas on the order of 100 nm, whereas bubble implosion can result in damage areas as large as 1 μm.

15D is a flow diagram illustrating an alternative wafer cleaning process according to an embodiment of the present invention. An alternative wafer cleaning process begins with step 15210 where an ultrasonic or mega sonic device is placed near the top surface of the wafer under cleaning. At step 15020, a rinse, chemical or gas doped water is injected over the wafer to fill the gap between the wafer and the acoustic device. In step 15230, the wafer transported by the chuck begins to rotate or vibrate. In step 15240, a power supply having a frequency f ₁ and a power level P ₁ is applied to the acoustic wave device. In step 15250, before the size of the bubble reaches the value of gap W, the power supply output is set to zero so that the temperature of the gas and/or vapor within the bubble is reduced because the temperature of the cleaning liquid is much lower than the gas temperature. it starts to cool In step 15260, the power supply output is set to zero when the temperature of the gas and/or vapor within the bubble decreases to room temperature T ₀ or the time period reaches τ ₂ (during the time period τ ₂ ). ), the power supply output is restored with frequency f ₁ and power level P1. In step 15270, the cleanliness of the wafer is checked, and steps 15210-15260 are repeated if the wafer has not yet been cleaned to the desired degree. Alternatively, the cleanliness check may not be performed every cycle. Alternatively, the number of cycles to be used can be determined empirically beforehand using the sample wafer.

Referring again to FIG. 15D , the temperature of the gas and/or vapor within the bubble need not be cooled to room temperature (T ₀ ), but preferably should be cooled much lower than the implosion temperature (T _i ). Also, at step 15250, the size of the bubble may be slightly larger than the spacing W of the patterned structure 15034, so long as the bubble expansion force does not destroy or damage the patterned structure 15034.

Referring back to FIG. 15D , the time duration of step 15240 can be obtained empirically as τ ₁ from the procedure shown in FIG. 7E . In some embodiments, the wafer cleaning process illustrated in FIGS. 7-14 may be combined with the wafer cleaning process illustrated in FIG. 15 .

16A-16C illustrate a wafer cleaning process according to an embodiment of the present invention. This wafer cleaning process is similar to that shown in FIGS. 7A-7E except for step 7050 of FIG. 7D. This wafer cleaning process is performed before the temperature of the gas and/or vapor within the bubble reaches the implosion temperature (T _i ) or before the time duration (τ ₁ ) is reached as calculated by Equation (11) in FIG. 16a. Set the power supply output to a positive (+) DC value shown in , or a negative (-) DC value shown in FIGS. 16B and 16C. As a result, the temperature of the gas and/or vapor inside the bubble starts to decrease because the temperature of the cleaning liquid is much lower than the temperature of the gas and/or vapor. In some embodiments, the amplitude of the positive or negative DC output is greater than (not shown) the amplitude of the power supply level (P ₁ ) applied during the time period (τ ₁ ) to create bubble cavitation in the cleaning fluid ( It can be the same (as shown in FIGS. 16A and 16B) or smaller (as shown in FIG. 16C).

17 shows a wafer cleaning process according to another embodiment of the present invention. This wafer cleaning process is similar to that shown in FIGS. 7A-7E except for step 7050 of FIG. 7D. This wafer cleaning process reverses the phase of the power output while maintaining the same frequency f ₁ as applied during the time period τ ₁ , so that bubble cavitation can be quickly stopped. Consequently, since the temperature of the cleaning liquid is much lower than the temperature of the gas and/or vapor, the temperature of the gas and/or vapor within the bubble begins to decrease.

Referring again to FIG. 17 , the power supply level during the time period τ ₂ is P , which in different embodiments may be greater than, equal to, or less than the power supply level P ₁ during the time period τ ₁ . is ₂ . In one embodiment, the power supply frequency during the time period τ ₂ may differ from f ₁ as long as the phase is reversed. In some embodiments, the ultrasonic or megasonic power supply frequency f ₁ is between 0.1 MHz and 10 MHz.

18a-18j illustrate bubble cavitation control to enhance circulation of fresh rinse solution in vias or trenches within a wafer. 18A is a cross-sectional view of a plurality of vias 18034 formed in a wafer 18010. The diameter of the via opening is denoted by W1. Air bubbles 18012 created by sonic energy within vias 18034 enhance the removal of impurities such as residues and particles therefrom. FIG. 18B is a plan view of the via shown in FIG. 18A.

18C is a cross-sectional view of a plurality of trenches 18036 formed in wafer 18010. Similarly, air bubbles 18012 created by sonic energy in trench 18036 enhance removal therefrom of impurities such as residues and particles. FIG. 18D is a top view of the trench 18036 shown in FIG. 18C.

The saturation point (R _s ) is defined by the largest amount of air bubbles that can be included inside vias 18034, trenches 18036, or other recessed region features. When the amount of bubbles is over the saturation point (R _s ), the cleaning solution will be blocked by the bubbles and may barely reach the bottom of the sidewall of the via 18034 or feature of the trench 18036, so that the cleaning performance is will be negatively affected. When the amount of bubbles is below the saturation point, clear liquid will have sufficient availability within the features of via 18034 or trench 18036, so good cleaning performance can be achieved.

Below the saturation point, the ratio (R) of the total bubble volume (V _B ) to the via or trench, recessed space (V _VTR ) is:

And, at the saturation point (R _s ), the ratio (R) is as follows.

The total bubble volume in a via 18034, trench 18036 or other recessed space feature is:

where N is the number of bubbles in the feature and V _B is the average volume of a single bubble.

18E-18H, when ultrasonic or mega sonic energy is applied to the cleaning liquid, the size of the bubble 18012 gradually expands to a specific volume, resulting in a total bubble volume (V _B ) versus the via, trench, or rib. The ratio R of the volume of the accessed space V _VTR is close to or above the saturation point R _s . The expanded air bubble 18012 blocks the path of cleaning liquid exchange and impurity removal in vias or trenches. In this case, acoustic energy is efficiently transferred into the vias or trenches and cannot reach their bottom and sidewalls, while particles, residues and other impurities 18048 are trapped in the vias or trenches. This case can easily occur in advanced semiconductor processes as the critical dimension W1 becomes smaller.

As shown in FIGS. 18I-18J , the size expansion of the bubble 18012 by ultrasound or mega sonic energy is within limits, and the total bubble volume (V _B ) versus the via, trench, or recessed space (V _VTR ) The ratio of the volumes (R) is much lower than the saturation point (R _s ). The fresh cleaning liquid 18047 circulates freely in the vias or trenches due to the small bubble cavitation inside the feature so that impurities 18048 such as residues and particles can be easily forced out of the feature for good cleaning performance.

Because the total volume of bubbles within a feature of a via or trench is determined by the number and size of the bubbles, controlling bubble size expansion due to cavitation is critical to cleaning performance for wafers with high aspect ratio features.

19A-19D show the change in bubble volume in response to sonic energy. During the first cycle of cavitation, the bubble's volume compresses from V ₀ to V ₁ after a positive acoustic power cycle and expands to V ₂ after a negative acoustic power cycle. However, the temperature T ₂ in the bubble corresponding to V ₂ is higher than the temperature T ₀ corresponding to V ₀ , so the volume V ₂ is larger than the volume V ₀ as shown in FIG. 19B. This increase in volume is caused by liquid molecules surrounding the bubbles that evaporate under higher temperatures. Similarly, the volume of V ₃ after the second compression of the bubble is somewhere between V ₁ and V ₂ as shown in FIG. 19B. V ₁ , V ₂ and V ₃ can be expressed as follows.

where ΔV is the volume compression of the bubble after one compression due to the positive pressure generated by the ultrasonic/mega-sound wave, δV is the volume increase of the bubble after one expansion due to the negative pressure generated by the ultrasonic/mega-sound wave, (δV −ΔV) is the volume increase due to the temperature increase (ΔT−δT) calculated in Equation (5) after one hour cycle.

After the second cycle of bubble cavitation, the bubble expands to a larger size as the temperature continues to increase. The volume of gas and/or vapor inside the bubble (V ₄ ) will be:

After the third compression, the volume of gas and/or vapor inside the cell (V ₅ ) will be:

According to this pattern, when the nth cycle of bubble cavitation reaches the minimum bubble size, the volume of gas and/or vapor inside the bubble (V _{2n - 1} ) will be

At the end of the nth cycle of bubble cavitation, the volume of gas and/or vapor inside the bubble (V _2n ) will be

limiting the volume of the bubble to a desired volume (V _i ), which is a dimension with sufficient physical mobility or conditions below the saturation point, and to prevent blockage of the path of cleaning fluid exchange in vias, trenches or other recessed area features. For this, the number of cycles (n _i ) can be written as

From Equation (19), the desired time (τ _i ) to achieve V _i can be written as

Here, t ₁ is the cycle period, and f ₁ is the frequency of the ultrasonic/mega sound wave. Thus, the desired number of cycles (n _i ) and the desired time (τ _i ) to prevent the bubble dimension from reaching the feature cut-off level can be calculated from equations (19) and (20).

When the number of cycles n of bubble cavitation increases, the temperature of the gas and/or vapor inside the bubble will increase, so that more molecules on the bubble surface will evaporate into the bubble's interior. Accordingly, the size of the bubble 19082 will further increase and become larger than the value calculated by Equation (8). In operation, the cell size, which is influenced by evaporation of liquid or water to the cell inner surface due to temperature increase, will not be discussed in detail here theoretically, since the cell size will be determined by the experimental method described later. As the average single cell volume continues to increase, the ratio R of total cell volume (V _B ) to vias, trenches or other recessed spaces (V _VTR ) increases continuously from R ₀ as shown in FIG. 19D. do.

As the bubble volume increases, the diameter of the bubble eventually becomes the same size as or equal to the size of feature W1 of via 18034 as shown in FIGS. 18A and 18B or trench 18036 shown in FIGS. 18C and 18D . order will be reached. After that, air bubbles inside the vias 18034 and trenches 18036 will block ultrasonic/mega-sonic energy from further entering their bottoms, especially when the aspect ratio (depth/width) is greater than or equal to 3. Thus, contaminants or particles at the bottom of these deep vias or trenches cannot be effectively removed or cleaned. Accordingly, a new cleaning process is proposed to prevent bubbles from growing to critical dimensions to block the path of cleaning fluid exchange in vias or trench features.

20A-20D illustrate an acoustic wafer cleaning process for effectively cleaning high aspect ratio features of vias and trenches in accordance with one embodiment of the present invention. This wafer cleaning process limits the size of air bubbles in cavitation by sonic energy. 20A shows a waveform of the power supply output with the power level set to P ₁ for a period of time τ ₁ and turned off for a period of time τ ₂ . Figure 20b shows the bubble volume curve corresponding to each cycle of cavitation. 20c shows the bubble size expansion during each cycle of cavitation. FIG. 20D shows a curve for the ratio (R) of the total bubble volume (V _B ) to the volume of the via, trench or other recessed space (V _VTR ).

According to , the ratio (R) of the total bubble volume (V _B ) to the volume of vias, trenches or recessed spaces (V _VTR ) increases from R ₀ to R _n , where the average single bubble volume is equal to time (τ _{1 )} . ) is expanded by sonic cavitation after a certain number of cycles (n), and R _n is controlled below the saturation point (R _s ).

And, the ratio (R) of the total bubble volume (V _B ) to the volume of vias, trenches or recessed spaces (V _VTR ) decreases from R _n to R _o , where the average single bubble volume is equal to time (τ ₂ ). It returns to its original size in the cooling process in

Referring back to FIG. 20B , the bubble expands to a large volume (V _n ) under the ultrasonic wave/mega-sound wave applied to the cleaning liquid for a time period (τ ₁ ). In this state, the pathway of mass transfer is partially blocked. After that, the fresh cleaning liquid cannot completely flow into the bottom and sidewall of the via or trench. On the other hand, particles, residues and other impurities trapped in vias and trenches cannot be efficiently removed. However, this state will alternate with the next state of bubble contraction when the ultrasonic/mega-acoustic power is turned off to cool the bubble for time τ ₂ as shown in FIG. 20A. In this cooled state, fresh cleaning liquid has the opportunity to flow into the vias and trenches to clean their bottoms and sidewalls. When the ultrasonic/mega-acoustic power is turned on again in the next cycle, particles, residues and other impurities can be removed from the vias and trenches by the pulling force created by the bubble volume increase. When the two conditions alternate in a cleaning process using ultrasound/mega-acoustics, vias, trenches and other high aspect ratio features in recessed areas on a wafer substrate can be effectively cleaned.

The cooling state at time τ ₂ plays an important role in this cleaning process. A condition that limits cell size, τ ₁ < τ _i , is preferred. The following method can experimentally determine the time (τ ₂ ) to limit the cell expansion to the blocking size by contracting the cell size during the cooling state and the time (τ ₁ ). The experiment is performed by cleaning a substrate patterned with small features of vias and trenches using an ultrasonic/mega sonic device coupled with a chemical liquid, where traceable residues are present to evaluate the cleaning performance.

The first step is to choose τ ₁ long enough to cut off features that can be used to calculate τ i based on Equation (20). The second step is to choose a different time (τ ₂ ) to run the DOE. The choice of time τ ₂ is at least 10 times τ ₁ , preferably 100 times τ ₁ in the first screen test. The third step is to individually clean the substrates having a specific patterned structure by fixing the time (τ ₁ ) and fixing the power (P ₀ ) to run under at least five conditions. where P ₀ is the power at which the features of the vias or trenches on the substrate are not reliably cleaned when run in continuous mode (non-pulsed mode). The fourth step is to inspect the state of traceable residues inside the features of the vias or trenches of the above five substrates by device analyzer tools such as SEMS or EDX. The first to fourth steps described above may be repeated several times to progressively shorten the time τ ₂ until a traceable residue inside the feature of the via or trench is observed. As the time τ ₂ shortens, the volume of air bubbles cannot be sufficiently contracted, which will gradually block the feature and affect the cleaning performance. This time is called the critical cooling time (τ _c ). After obtaining the critical cooling time τ _c , the time τ ₂ can be set to a value greater than 2τ _c to have a safety margin.

A more detailed example is shown as follows, wherein the first step is τ ₁₀ , 2τ ₁₀ , 4τ ₁₀ , 8τ ₁₀ , 16τ ₁₀ , 32τ ₁₀ , 64τ ₁₀ , 128τ ₁₀ , 256τ ₁₀ , 512τ, as shown in Table 3. As the design of the experimental (DOE) conditions such as ₁₀ , 10 different times (τ1) are selected. The second step is to select at least 10 times 512τ ₁₀ , preferably 20 times 512τ ₁₀ , in the first screen test, as shown in Table 3. The third step is to fix the power (P ₀ ) to operate under the above 10 conditions to individually clean the substrates with specific patterned structures. where P ₀ is the power at which the features of the vias or trenches on the substrate are not reliably cleaned when run in continuous mode (non-pulsed mode). The fourth step is to use the conditions shown in Table 3 to process 10 substrates having features of vias or trenches after plasma etching. The reason for choosing a post plasma etched substrate is that the polymer generated during the etching process forms on the sidewalls of the trenches and vias. The polymer formed on the bottom or sidewall of the via is difficult to remove by conventional methods. The next step is to inspect the cleaning condition of the features of the vias or trenches on the 10 substrates by SEMS in cross-section of the substrates. The resulting data is shown in Table 3 below. From Table 3, the cleaning effect reaches the best point for substrate #6 at τ ₁ =32τ ₁₀ , so the optimum time (τ ₁ ) is 32τ ₁₀ .

[Table 3]

If no peak is found, the first to fourth steps described above may be repeated again with a wider time range of τ ₁ to find the time τ ₁ . After finding the initial time τ ₁ , the first and fourth steps described above may be repeated again with a narrower time range τ ₁ to narrow the range of time τ ₁ . Knowing time τ ₁ , time τ ₂ can be optimized by decreasing time τ ₂ from 512τ ₂ to a value at which the cleaning effect begins to decrease. From Table 4, the cleaning effect reaches the best point for substrate #5 at τ ₂ =256τ ₁₀ , so the optimal time (τ ₂ ) is 256τ ₁₀ .

[Table 4]

21A-21C illustrate another cleaning process according to an embodiment of the present invention. This cleaning process is similar to that shown in FIGS. 20a-20d, with the difference that the power in the current cleaning process is still turned on for a period of mτ ₁ after cavitation reaches the saturation point (R _s ). Here, m can be a number from 0.1 to 100, preferably 2, depending on the via and trench structure and the cleaning solution used. And the value of M needs to be optimized by experiments similar to the embodiment shown in Figs. 20a-20d.

22A and 22B illustrate a wafer cleaning process using sonic energy according to another embodiment of the present invention. During the time τ ₁ when sonic power P ₁ is applied to the cleaning liquid, bubble implosion starts to occur when the temperature of the first bubble reaches its implosion temperature at the point T _i , and then from T _i Some bubble implosion continues for the increasing temperature up to Tn (for a time of Δτ). After turning off the sonic power in a time period of Τ ₂ , the temperature of the bubble is cooled from T _n to the original T ₀ by the surrounding liquid. T _i is determined as the threshold of temperature for bubble implosion in features of vias and trenches, triggering the first bubble implosion.

Since the heat transfer is not precisely uniform across the features, more bubble implosion may occur after the temperature reaches T _i . The bubble implosion strength will become higher and higher as the bubble temperature (T) increases. However, cell implosion must be controlled below the implosion strength which can damage the patterned structure. Bubble implosion can be controlled by controlling the temperature (T _n ) to be less than or equal to the temperature (T _d ) by adjusting the time (Δτ), where T _n is the maximum temperature of the bubble due to sonic power applied to the cleaning solution for n cycles. where T _d is the temperature of accumulation of a certain amount of bubble implosion with high intensity (or power) that causes damage to the patterned structure. In this cleaning process, controlling the cell implosion intensity is achieved by adjusting the time Δτ after the first cell implosion starts, so as to avoid the cell implosion intensity becoming too high to cause damage to the patterned structure under cleaning while avoiding the desired cleaning. performance and efficiency are achieved.

To increase particle removal efficiency (PRE), it is desirable to have controlled pass-through cavitation in an ultrasonic or mega sonic cleaning process as shown in FIGS. 22A-22B. Controlled pass cavitation sets the acoustic power supply to power (P ₁ ) at a time interval shorter than τ ₁ , sets the acoustic power supply to power (P ₂ ) at a time interval longer than τ ₂ , and when the wafer is being cleaned. It is achieved by repeating the above steps until , where the power P ₂ is zero or much smaller than the power P ₁ , τ1 is the time interval during which the temperature within the bubble rises above the critical implosion temperature, and τ ₂ is The time interval during which the temperature within the bubble falls to a temperature well below the critical implosion temperature. Because controlled-through cavitation has specific bubble implosion in the cleaning process, controlled-through cavitation will provide higher particle removal efficiency (PRE) with minimal damage to the patterned structure. The critical implosion temperature is the lowest temperature inside the cell that will cause the first cell implosion. In order to further increase the PRE, it is necessary to further increase the temperature of the bubble, so a longer time (τ ₁ ) is required. Also, the temperature of the bubble can be increased by shortening the time of _τ2 .

The frequency of ultrasound or megasonic waves is another parameter to control the level of implosion. Maintaining controlled pass-through cavitation involves setting the acoustic power supply to a frequency f ₁ at a time interval shorter than τ ₁ , setting the acoustic wave power supply to a frequency f ₂ at a time interval longer than τ ₂ , and setting the wafer may be achieved by repeating the above steps until is cleaned, where f ₂ is much higher than f ₁ , preferably 2 or 4 times higher. Typically, the higher the frequency, the lower the level or strength of the implosion. Again, τ ₁ is the time interval during which the temperature within the bubble rises above the critical implosion temperature, and τ ₂ is the time interval during which the temperature within the bubble falls to a temperature well below the critical implosion temperature. Controlled through cavitation will provide higher particle removal efficiency (PRE) with minimal damage to the patterned structures. The critical implosion temperature is the lowest temperature inside a cell that causes first cell implosion. In order to further increase the PRE, it is necessary to further increase the temperature of the bubble, so a longer time (τ ₁ ) is required. In addition, the temperature of the bubble can be increased by shortening the time interval (τ ₂ ). In general, ultrasound or mega with a frequency between 0.1 MHz and 10 MHz can be applied to the wafer cleaning process disclosed in the present invention.

23 illustrates an exemplary wafer cleaning apparatus for performing the wafer cleaning process shown in FIGS. 7-22 in accordance with an embodiment of the present invention. The wafer cleaning apparatus includes a wafer chuck 23014 for mounting a wafer 23010 thereon. Wafer chuck 23014 along with wafer 23010 rotates during a cleaning process driven by rotation drive mechanism 23016. The wafer cleaning apparatus also includes a nozzle 23064 for delivering a cleaning chemical or cleaning liquid, such as deionized water 23060, to the wafer 23010. An ultrasonic or mega sonic device 23062 is coupled with the nozzle 23064 to impart ultrasonic or mega sonic energy to the cleaning liquid. Ultrasound or mega sonic waves generated by the ultrasonic or mega sonic wave device 23062 are transmitted from the nozzle 23064 through the cleaning liquid 23060 to the wafer 23010 .

24 is a cross-sectional view of another wafer cleaning apparatus for performing the wafer cleaning process shown in FIGS. 7-22 according to an embodiment of the present invention. The wafer cleaning apparatus includes a cleaning tank 24074 containing a body of cleaning liquid 24070 and a wafer cassette 24076 holding a plurality of wafers 24010 immersed in the cleaning liquid 24070. The wafer cleaning device further includes an ultrasonic or mega sonic wave device 24072 attached to the wall of the cleaning tank 24074 for imparting ultrasonic or mega sonic wave energy to the cleaning liquid. There is at least one inlet (not shown) for filling the cleaning tank 24074 with the cleaning liquid 24070 so that the wafer 24010 is immersed in the cleaning liquid 24070 during the cleaning process.

In the foregoing embodiments, all critical process parameters of the acoustic wave power supply, such as power level, frequency, power-on time (τ ₁ ) and residual power-off time (τ ₂ ) are monitored in real time during the wafer cleaning process by the power supply controller. , damage to the patterned structure may still occur due to some abnormal condition during the wafer cleaning process. Therefore, there is a need for an apparatus and method for real-time monitoring of the operating state of a sound wave power supply. If the parameter is not in the normal range, the sonic power supply should be cut off, and an alarm signal should be sent and reported.

25 illustrates a control system for monitoring operating parameters of a wafer cleaning process using sonic energy according to one embodiment of the present invention. The control system includes a host computer 25080, a sound wave generator 25082, a sound wave converter 1003, a detection system 25086 and a communication cable 25088. The host computer 25080 provides control commands such as power set (P ₁ ), power-on time set ( _τ1 ), power set (P ₂ ), power-off time set (τ ₂ ), frequency set, and power enable commands. Sound wave parameter settings such as are transmitted to the sound wave generator 2508. The sound wave generator 25082 generates a sound wave after receiving this command, and transmits the sound wave to the sound wave converter 1003 to clean the wafer 1010 . On the other hand, the parameter settings transmitted by the host computer 25080 and the output of the sound wave generator 25082 are read by the detection system 250886. The detection system 25086 compares the output from the sonic generator 25082 with the parameter settings transmitted by the host computer 25080, and then transmits the comparison result to the host computer 25080 via the communication cable 25088. . If the output from the sonic generator 25082 differs from the parameter settings sent by the host computer 25080, the detection system 25086 sends an alarm signal to the host computer 25080. Upon receiving the alarm signal, the host computer 25080 shuts off the sonic generator 25082 to prevent damage to the patterned structures on the wafer 1010.

26 is a block diagram of the detection system 25086 shown in FIG. 25 according to an embodiment of the present invention. The detection system 25086 illustratively includes a voltage attenuation circuit 26090, a shaping circuit 26092, a main controller 26094, a communication circuit 26096 and a power circuit 26098. The main controller 26094 may be implemented as an FPGA. Communication circuitry 26096 is implemented as an interface to host computer 25080. The communication circuit 26096 performs RS232/RS485 serial communication with the host computer 25080 to read parameter settings from the host computer 25080, and transmits the comparison result back to the host computer 25080. The power circuit 26098 is designed to convert DC 15V to target voltages of DC 1.2V, DC 3.3V, and DC 5V for the entire system.

27 is a block diagram of a detection system 25086 according to another embodiment of the present invention. The detection system 25086 illustratively includes a voltage attenuation circuit 26090, an amplitude detection circuit 27092, a main controller 26094, a communication circuit 26096 and a power circuit 26098.

28A-28C show an exemplary implementation of a voltage attenuation circuit 26090 according to one embodiment of the present invention. When the sound wave signal output from the sound wave generator 25082 is first read, it has a relatively high amplitude value as shown in Fig. 28B. The voltage attenuation circuit 26090 is designed to use two operational amplifiers 28102 and 28104 to reduce the amplitude value of the waveform as shown in FIG. 28C. The attenuation factor of the voltage attenuation circuit 26090 is set in the range of 5 to 100, preferably 20. Voltage attenuation can be expressed by the following equation.

Here, Vout is the amplitude value output by the voltage attenuation circuit 26090, Vin is the amplitude value input to the voltage attenuation circuit 26090, and R1, R2, R3, and R4 are the two operational amplifiers 28102 and 28104 is the resistance of

29A-29C show an exemplary implementation of the shaping circuit 26092 shown in FIG. 26 according to one embodiment of the present invention. Referring again to FIG. 26 , the output of voltage damping circuit 26090 is connected to shaping circuit 26092 . The waveform from the voltage attenuation circuit 26090 is input to the shaping circuit 26092 to convert the sinusoidal waveform into a square wave to be processed by the main controller 26094. The shaping circuit 26092 includes a window comparator 29102 and an OR gate 29104 as shown in FIG. 29A. Vcal _- < Vin < Vcal+, Vout = 0, Vout = 1, where Vcal _- and Vcal+ are two thresholds, then Vin is the input value of the shaping circuit and Vout is the output value of the shaping circuit. The waveform passing through the voltage attenuation circuit 2190 is a sine wave as shown in FIG. 29B. The shaping circuit 26092 converts the sine wave into a square wave as shown in Fig. 29C.

30A-30C illustrate an exemplary implementation of the main controller 26094 of FIGS. 26 and 27 according to one embodiment of the present invention. The main controller 26094 includes a pulse conversion module 30102 and a period measurement module 3104 as shown in FIG. 30A. The pulse conversion module 30102 is used to convert the pulse signal into a high level signal for a period of time τ ₁ , and to maintain the low level signal for a period of time τ2 as shown in FIGS. 30B and 30C . A circuit symbol of the pulse conversion module 30102 is shown in Fig. 30A, where Clk_Sys is a 50 MHz clock signal, Pulse_in is an input signal, and Pulse_Out is an output signal. The period measurement module 30104 is used to measure the duration of the high level and low level signals by means of a counter using the equation below.

Where Counter_H is the high level number and Counter_L is the low level number.

The main controller 26094 compares the calculated power-on time with a preset time τ ₁ . If the calculated power-on time is longer than the preset time (τ ₁ ), the main controller 26094 sends an alarm signal to the host computer 25080. Host computer 25080 shuts down sonic generator 2508 upon receiving the alarm signal. The main controller 26094 compares the calculated power-off time with a preset time τ ₂ . If the calculated power-off time is shorter than the preset time (τ ₂ ), the main controller 26094 sends an alarm signal to the host computer 25080. The host computer 25080 shuts off the sonic generator 26082 upon receiving the alarm signal. In one embodiment, main controller 26094 may be implemented using Altera Cyclone IV FPGA model number EP4CE22F17C6N.

Fig. 31 shows that the sonic power supply still vibrates several cycles after the host computer cuts off the sonic power supply due to the characteristics of the sonic device. The time period (τ ₃ ) over which the sound generator 25082 oscillates several cycles after being powered down is measured by the main controller 2609. This time period (τ ₃ ) can be obtained experimentally. Thus, the actual power-on time is equal to τ ₁ -τ ₃ , where τ is the time calculated by periodic measurement module 25104. When the actual power-on time is longer than the preset time (τ ₁ ), the main controller 26094 transmits an alarm signal to the host computer 25080.

32A-32C show an exemplary implementation of the amplitude detection circuit 27092 of FIG. 27 according to one embodiment of the present invention. The amplitude detection circuit 27092 illustratively includes a reference voltage generation circuit and a comparison circuit. As shown in FIG. 32B, the reference voltage generator circuit is a D/A converter 32118 to convert digital inputs from the main controller 26094 to analog DC reference voltages (Vref + and Vref -) as shown in FIG. 27C. ) is designed to be used. The comparison circuit is designed to use window comparator 32114 and AND gate 32116 to compare the attenuated amplitude (Vin), i.e., the output from voltage decay circuit 26090, to reference voltages (Vref + and Vref -). . When the attenuated amplitude (Vin) exceeds the reference voltage (Vref + and/or Vref -), the amplitude detection circuit 27092 sends an alarm signal to the host computer 25080. Upon receiving the alarm signal, the host computer 25080 shuts off the sonic generator 25082 to prevent damaging the patterned structures on the wafer 1010.

33 is a flow chart illustrating a wafer cleaning process according to an embodiment of the present invention. The wafer cleaning process begins at step 33010 in which a cleaning liquid is applied to the space between the wafer and ultrasonic/mega-acoustic waves. In step 33020, the ultrasonic/mega-sonic wave power supply unit is set to a frequency f ₁ and a power level P ₁ to drive the ultrasonic/mega-sonic wave device. In step 33030, the detected power-on time is compared to a preset time (τ ₁ ). If the detected power-on time is longer than τ ₁ , the power supply will be cut off and an alarm signal will also be sent. At step 33040, the ultrasonic/mega-acoustic power supply is set to zero output prior to bubble cavitation in the cleaning liquid damaging patterned structures on the wafer. In step 33050, the acoustic power supply is restored to the frequency f ₁ and power level P ₁ after the temperature within the bubble is reduced to a specific level. In step 33060, the detected power-off time is compared to a preset time (τ ₂ ). If the detected power-off time is shorter than τ ₂ , the ultrasonic/mega-acoustic power supply will be cut off and an alarm signal will also be sent. In step 33070, wafer cleanliness is checked, and steps 33010-33060 described above will be repeated if the desired cleanliness is not met. Alternatively, the cleanliness check may not be performed every cycle. Alternatively, the number of cycles to be used can be determined empirically beforehand using the sample wafer.

34 is a flowchart illustrating a wafer cleaning process according to another embodiment of the present invention. The wafer cleaning process begins at step 34010 in which a cleaning liquid is applied to the space between the wafer and the ultrasonic/mega sonic device. In step 34020, the ultrasonic/mega-sonic wave power supply unit is set to a frequency f ₁ and a power level P ₁ to drive the ultrasonic/mega-sonic wave device. In step 34030, the amplitude of the sonic power output is detected and compared to a preset value. If the detected amplitude is higher than the preset value, the power will be cut off and an alarm signal will also be sent. At step 34040, the sonic supply is set to zero power prior to bubble cavitation in the cleaning solution damaging patterned structures on the wafer. In step 31050, the sound wave power supply unit is restored to the frequency f ₁ and the power level P ₁ after the temperature inside the bubble is reduced to a specific level. In step 34060, the wafer cleanliness is checked and steps 34010-34050 will be repeated if the desired cleanliness is not met. Alternatively, the cleanliness check may not be performed every cycle. Alternatively, the number of cycles to be used can be determined empirically beforehand using the sample wafer.

In some embodiments, wafer cleaning processes shown in various figures throughout this disclosure can be combined to produce a desired cleaning result. In one embodiment, amplitude detection in step 34030 of FIG. 34 may be incorporated into the wafer cleaning process shown in FIG. 33 . In another embodiment, the amplitude detection circuit 27092 of FIG. 27 as well as the voltage attenuation 26090 and shaping circuit 26092 of FIG. 26 may be applied to implement the wafer cleaning process shown in FIGS. 33 and 34 .

The present invention provides an apparatus for cleaning a semiconductor substrate using an ultrasonic/mega-sonic wave device including a chuck, an ultrasonic/mega-sonic wave device, at least one nozzle, an ultrasonic/mega-sonic wave power supply, a host computer, and a detection system. The chuck holds the semiconductor substrate. An ultrasonic/mega sonic device is positioned adjacent to the semiconductor substrate. At least one nozzle injects a chemical liquid onto the semiconductor substrate and into a gap between the semiconductor substrate and the ultrasonic/mega sonic device. The host computer drives the ultrasonic/mega-acoustic wave device by setting the ultrasonic/mega-acoustic wave power supply to a frequency f ₁ and a power level P ₁ ; Before bubble cavitation takes place in the damaging patterned structure on the semiconductor substrate, the ultrasonic/mega-acoustic power supply is set to zero output; After the temperature in the bubble cools down to the set temperature, the ultrasonic/mega-acoustic wave power supply unit is reset to the frequency f ₁ and power level P ₁ . The detection system separately detects a power-on time and a power-off time at power (P ₁ ) and frequency (f ₁ ), and determines the detected power-on time at power (P ₁ ) and frequency (f ₁ ). Compare with the preset time (τ ₁ ). If the detected power-on time is longer than the preset time τ ₁ , the detection system sends an alarm signal to the host computer, and the host computer receives the alarm signal to cut off the ultrasonic/mega-acoustic power supply. Also, the detection system compares the detected power-off time with a preset time (τ ₂ ). When the detected power-off time is shorter than the preset time _τ2 , the detection system sends an alarm signal to the host computer, and the host computer receives the alarm signal to cut off the ultrasonic/mega-acoustic power supply.

In one embodiment, an ultrasonic/mega-acoustic device is coupled to the nozzle and positioned adjacent to a semiconductor substrate, and the energy of the ultrasonic/mega-acoustic device is transferred from the nozzle through the liquid column to the semiconductor substrate.

The present invention provides another device for cleaning a semiconductor substrate using an ultrasonic/mega sonic device including a chuck, an ultrasonic/mega sonic device, at least one nozzle, an ultrasonic/mega sonic power supply, a host computer and a detection system. . The chuck holds the semiconductor substrate. An ultrasonic/mega sonic device is positioned adjacent to the semiconductor substrate. At least one nozzle injects a chemical liquid onto the semiconductor substrate and into a gap between the semiconductor substrate and the ultrasonic/mega sonic device. The host computer drives the ultrasonic/mega sonic wave device by setting the ultrasonic/mega sonic wave power supply to the frequency f ₁ and power P ₁ ; Before bubble cavitation takes place in the patterned structure in the damaging liquid on the semiconductor substrate, the ultrasonic/mega-acoustic wave power supply is set to zero output; After the temperature in the bubble cools down to the set temperature, the ultrasonic/mega-acoustic wave power supply unit is reset to the frequency f ₁ and the power P ₁ . The detection system detects the amplitude of each waveform output by the ultrasonic/mega-acoustic wave power supply, and compares the amplitude of each detected waveform with a preset value. When the detected amplitude of any waveform is greater than the preset value, the detection system sends an alarm signal to the host computer, and the host computer receives the alarm signal and cuts off the ultrasonic/mega-acoustic wave power supply, where the preset value is greater than the waveform amplitude in normal operation.

In one embodiment, an ultrasonic/mega-acoustic device is coupled to the nozzle and positioned adjacent to a semiconductor substrate, and the energy of the ultrasonic/mega-acoustic device is transferred from the nozzle through the liquid column to the semiconductor substrate.

The present invention provides another device for cleaning a semiconductor substrate using an ultrasonic/mega sonic device including a cassette, a tank, an ultrasonic/mega sonic device, at least one inlet, an ultrasonic/mega sonic power supply, a host computer and a detection system. do. The cassette holds at least one semiconductor substrate. The tank holds the cassette. An ultrasonic/mega sonic device is attached to the outer wall of the tank. At least one inlet is used to fill a chemical liquid into the tank for immersing the semiconductor substrate. The host computer drives the ultrasonic/mega sonic wave device by setting the ultrasonic/mega sonic wave power supply to the frequency f ₁ and power P ₁ ; Before bubble cavitation takes place in the patterned structure in the liquid on the semiconductor substrate, the ultrasonic/mega-acoustic wave power supply is set to zero output; After the temperature in the bubble cools down to the set temperature, the ultrasonic/mega-acoustic wave power supply unit is reset to the frequency f ₁ and power P ₁ . The detection system separately detects a power-on time and a power-off time at power (P ₁ ) and frequency (f ₁ ), and determines the detected power-on time at power (P ₁ ) and frequency (f ₁ ). Compare with the preset time (τ ₁ ). If the detected power-on time is longer than the preset time τ ₁ , the detection system sends an alarm signal to the host computer, and the host computer receives the alarm signal to cut off the ultrasonic/mega-acoustic power supply. Also, the detection system compares the detected power-off time with a preset time (τ ₂ ). When the detected power-off time is shorter than the preset time (τ ₂ ), the detection system sends an alarm signal to the host computer, and the host computer receives the alarm signal to cut off the ultrasonic/mega-acoustic power supply.

The present invention provides another device for cleaning a semiconductor substrate using an ultrasonic/mega-acoustic device comprising a cassette, a tank, an ultrasonic/mega-sonic device, at least one inlet, an ultrasonic/mega-sonic power supply, a host computer and a detection system. to provide. The cassette holds at least one semiconductor substrate. The tank holds the cassette. An ultrasonic/mega sonic device is attached to the outer wall of the tank. At least one inlet is used to fill a chemical liquid into the tank for immersing the semiconductor substrate. The host computer drives the ultrasonic/mega sonic wave device by setting the ultrasonic/mega sonic wave power supply to the frequency f ₁ and power P ₁ ; Before bubble cavitation takes place in the patterned structure in the liquid on the semiconductor substrate, the ultrasonic/mega-acoustic wave power supply is set at zero output; After the temperature in the bubble cools down to the set temperature, the ultrasonic/mega-acoustic wave power supply unit is reset to the frequency f ₁ and power P ₁ . The detection system detects the amplitude of each waveform output by the ultrasonic/mega-acoustic wave power supply, and compares the detected amplitude of each waveform with a preset value. When the detected amplitude of any waveform is greater than the preset value, the detection system sends an alarm signal to the host computer, and the host computer receives the alarm signal to cut off the ultrasonic/mega-acoustic wave power supply, where the preset value is greater than the waveform amplitude in normal operation.

Although the present disclosure has been specifically shown and described with reference to exemplary embodiments of the present invention, those skilled in the art will appreciate that various changes may be made in form and detail without departing from the spirit of the claimed embodiments.

Claims (55)

A system for controlling damage when cleaning a semiconductor wafer comprising features of patterned structures comprising:
a wafer holder that temporarily restrains the semiconductor wafer during a cleaning process;
an inlet for delivering a cleaning solution onto the surface of the semiconductor wafer;
an acoustic wave generator configured to alternately operate at a first frequency and a first power level for a first predetermined period of time and at a second frequency and a second power level for a second predetermined period of time to impart sonic energy to the cleaning liquid; wherein the first predetermined time period and the second predetermined time period successively follow each other; and
the first and second frequencies, the first and second power levels, the first and second predetermined time periods, and the number of alternations between the first and second predetermined time periods by the sound wave generator a controller programmed to provide
At least one of the first and second predetermined time periods, the first and second power levels, and the first and second frequencies are such that a percentage of features that are damaged as a result of imparting sonic energy is less than a predetermined threshold. is determined, wherein the predetermined threshold is greater than zero;
powering at the second frequency and at the second power level for the second predetermined period of time after the ratio of total bubble volume to volume within a via, trench or recessed region on the semiconductor wafer increases to a first set value; and the first frequency and the first power level for the first predetermined time period after the ratio of the total bubble volume to the volume in a via, trench or recessed region on the semiconductor wafer decreases to a second set value. to set the power to,
system.
According to claim 1,
wherein the bubble size in the cleaning liquid increases due to sonic energy during the first predetermined period of time and decreases during the second predetermined period of time.
system.
According to claim 1,
The wafer holder is a rotary chuck,
system.
According to claim 1,
The wafer holder is a cassette immersed in a cleaning tank,
system.
According to claim 1,
The inlet has a nozzle,
system.
According to claim 1,
Further comprising a sonic transducer coupled to the sound wave generator,
system.
According to claim 6,
the sonic transducer is positioned above the semiconductor wafer having a gap between the sonic transducer and the semiconductor wafer, the gap being filled with the cleaning liquid during a cleaning process;
system.
According to claim 7,
wherein the gap changes during the cleaning process;
system.
According to claim 6,
The sound wave converter is connected to the inlet and imparts sound wave energy to the cleaning liquid flowing through the inlet.
system.
According to claim 1,
The washing liquid is selected from the group consisting of chemical solutions, deionized water, and combinations thereof.
system.
According to claim 1,
The second power level is 0,
system.
According to claim 1,
The first frequency is equal to the second frequency, all of the frequencies remain constant during the respective operating time period, while the first power level is higher than the second power level, and both of the power levels is kept constant during each operating time period,
system.
According to claim 1,
The first frequency is higher than the second frequency, both of which remain constant during each operating time period, while the first power level is higher than the second power level, and all of the power levels are respectively which remains constant over the operating time period of
system.
According to claim 1,
The first frequency is lower than the second frequency, and both of the frequencies remain constant during each operating time period, while the first power level is equal to the second power level, and all of the power levels are which remains constant during each operating time period,
system.
According to claim 1,
The first frequency is lower than the second frequency, and both frequencies remain constant during each operating time period, while the first power level is higher than the second power level, and all of the power levels are respectively which remains constant over the operating time period of
system.
According to claim 1,
The first frequency is lower than the second frequency, and both frequencies remain constant during each operating time period, while the first power level is lower than the second power level, and all of the power levels are respectively which remains constant over the operating time period of
system.
According to claim 1,
the first power level rises during the first predetermined period of time;
system.
According to claim 1,
wherein the first power level falls during the first predetermined period of time;
system.
According to claim 1,
wherein the first power level rises and falls during the first predetermined period of time;
system.
According to claim 1,
the second frequency is substantially close to zero, and the second power level remains at a constant positive value for the second predetermined period of time;
system.
According to claim 1,
wherein the second frequency is substantially close to zero and the second power level is maintained at a constant negative value for the second predetermined period of time.
system.
According to claim 1,
The sound waves from the sound wave converter in the first and second time periods have opposite phases.
system.
According to claim 1,
the first predetermined time period is shorter than 2,000 times the cycle period of the first frequency;
system.
According to claim 1,
the first predetermined period of time is shorter than ((T _i -T ₀ -ΔT)/(ΔT-δT)+ 1)/f ₁ , where T _i is the implosion temperature and T ₀ is the temperature of the cleaning liquid; ΔT is the temperature increase after one compression time, δT is the temperature decrease after one expansion time, f ₁ is the first frequency,
system.
According to claim 1,
wherein the first frequency changes from a higher value to a lower value during the first predetermined period of time.
system.
According to claim 1,
wherein the first frequency changes from a lower value to a higher value during the first predetermined period of time.
system.
According to claim 1,
the first frequency changes from a lower value to a higher value during the first predetermined period of time and then returns to the lower value;
system.
According to claim 1,
the first frequency changes from a higher value to a lower value during the first predetermined period of time and then returns to the higher value;
system.
According to claim 1,
the first frequency is first set to f ₁ , then f ₃ and finally f ₄ during the first predetermined period of time, where f ₄ is less than f 3 and f ₃ is less than f _{1 ;}
system.
According to claim 1,
the first frequency is first set to f ₄ , then f ₃ , and finally f ₁ during the first predetermined period of time, where f ₄ is less than f ₃ and f ₃ is less than f ₁ ;
system.
According to claim 1,
the first frequency is first set to f ₁ , then f ₄ and finally f ₃ during the first predetermined period of time, where f ₄ is less than f 3 and f ₃ is less than f _{1 ;}
system.
According to claim 1,
the first frequency is first set to f ₃ , then f ₄ and finally f ₁ during the first predetermined period of time, where f ₄ is less than f 3 and _{f 3 is} less than f ₁ ;
system.
According to claim 1,
the first frequency is first set to f ₃ , then f ₁ and finally f ₄ during the first predetermined period of time, where f ₄ is less than f 3 and f ₃ is less than f _{1 ;}
system.
According to claim 1,
the first frequency is first set to f ₄ , then f ₁ and finally f ₃ during the first predetermined period of time, where f ₄ is less than f 3 and f ₃ is less than f _{1 ;}
system.
According to claim 1,
further comprising a detection circuit coupled to the sonic generator for detecting an output of the sonic generator.
system.
The method of claim 35,
The detection circuit comprises a voltage attenuation circuit for attenuating the input signal.
system.
The method of claim 35,
wherein the detection circuit comprises a shaping circuit for converting a signal from a first waveform to a second waveform.
system.
38. The method of claim 37,
The first waveform is a sine wave, and the second waveform is a square wave.
system.
The method of claim 35,
The detection circuit includes an amplitude detection circuit for detecting an input signal and comparing it with a reference value, the detection circuit causing an alarm signal to be generated and the sound wave generator to be turned off when the detected amplitude exceeds the reference value.
system.
The method of claim 39,
The reference value is generated by a digital-to-analog converter (DAC),
system.
The method of claim 35,
wherein the detection circuit detects the first period of time and compares it to a predetermined value, and causes an alarm signal to be generated and the sonic generator to be turned off when the first period of time exceeds the predetermined value.
system.
The method of claim 35,
wherein the detection circuit detects the second period of time and compares it to a predetermined value, and when the second period of time exceeds the predetermined value, an alarm signal is generated and the sonic generator is turned off.
system.
According to claim 1,
wherein the first predetermined period of time is empirically determined to avoid bubble implosion in the cleaning solution by inspecting the semiconductor wafer for damage to patterned structures thereon.
system.
44. The method of claim 43,
The empirical determination is to keep the first and second frequencies and the first and second power levels unchanged, as well as to keep the second predetermined time period substantially longer than the first predetermined time period without changing. selecting a different value for the first predetermined period of time in a different cleaning process, while holding longer.
system.
According to claim 1,
wherein the first predetermined period of time is empirically determined to allow limited bubble cavitation that does not cause damage to patterned structures on the semiconductor wafer under cleaning.
system.
The method of claim 45,
The empirical determination is to keep the first and second frequencies and the first and second power levels unchanged, as well as to keep the second predetermined time period substantially longer than the first predetermined time period without changing. selecting a different value for the first predetermined period of time in a different cleaning process, while holding longer.
system.
According to claim 1,
wherein the second predetermined period of time is empirically determined such that a temperature inside bubbles in the cleaning liquid is cooled to a predetermined temperature;
system.
The method of claim 47,
wherein the predetermined temperature is substantially close to room temperature;
system.
According to claim 1,
wherein the first frequency and the first power level are empirically determined to avoid bubble implosion in the cleaning solution by inspecting the semiconductor wafer for damage to patterned structures thereon.
system.
According to claim 1,
wherein the first frequency and the first power level are empirically determined to allow limited bubble implosion that does not cause damage to patterned structures on the semiconductor wafer under cleaning.
system.
According to claim 1,
wherein the second frequency and the second power level are determined empirically such that a temperature inside bubbles in the cleaning liquid is cooled to a predetermined temperature.
system.
The method of claim 51 ,
wherein the predetermined temperature is substantially close to room temperature;
system.
According to claim 1,
wherein the number of shifts is determined empirically by examining damage to the semiconductor wafer;
system.
According to claim 1,
The cleaning effect of imparting the sonic energy causes a greater yield improvement than the yield reduction caused by damage as a result of imparting the sonic energy.
system.
delete

Images