CN110398500A - Method and Experimental Device for Evaluating Wafer Cleaning Efficiency - Google Patents

Abstract

The invention relates to a method and an experimental device for evaluating wafer cleaning efficiency, which solves the problems of low accuracy, systematic error and human error in the existing method. The method is to drop the adhesive solution onto the surface of the rotating wafer sample to perform a contamination experiment, and obtain a sample of the contaminated wafer with defects on the surface; use a morphology characterization instrument to analyze the surface of the sample of the contaminated wafer, and obtain the surface of the sample of the contaminated wafer. The number of defects before cleaning; the cleaning solution was dropped into the surface of the contaminated wafer sample in the rotation for cleaning experiments to obtain the cleaned wafer sample; the morphology of the cleaned wafer sample was analyzed again using a morphology characterization instrument to obtain the surface of the contaminated wafer sample after cleaning The number of defects; the cleaning efficiency is calculated by the following formula (A), and the contamination experiment and cleaning experiment are carried out with a special experimental device. The measurement method is simple, reliable, and accurate, and can provide powerful guidance for the research on the cleaning solution formula.

Description

Method and Experimental Device for Evaluating Wafer Cleaning Efficiency

technical field

The invention relates to an experimental method and equipment, in particular to a method and an experimental device for evaluating wafer cleaning efficiency.

Background technique

As the critical dimensions of semiconductor devices continue to shrink, most of the problems that were not considered at the previous technology node have been enlarged, including small size defects. The "1/10 law" points out that defects above 1/10 of the minimum line width will largely cause device failure, so the removal of defects is particularly important to improve the yield of semiconductor device manufacturing.

Wet cleaning is an important link in the manufacturing process of semiconductor devices. The cleaning solution used is the main functional material to reduce defects. The defects that can be removed include but are not limited to particles, organic residues, metal residues, etc., which benefit from The slight etching of the wafer surface by the cleaning solution and the weakening of the adhesion of the above-mentioned pollutants.

In the conventional cleaning solution formulation research, the manufacturing machine is generally used for on-machine experiments, but the high cost of wafers and the large amount of cleaning solution used restrict the flexibility of the method and compress the experimental space of researchers. The general research method is to study the ability of the cleaning solution in the laboratory by static immersion, and at the same time combine the optical microscope to count the number of defects on small-sized wafers. Although this method is one of the most common means of screening cleaning fluids, on the one hand, it ignores the influence of physical effects in the actual cleaning process of the wafer, including the impact of the liquid in the acid tank machine, the spin in the single-wafer cleaning machine, etc. And the centrifugal force provided, and the process of drying by spin; on the other hand, the resolution of the optical microscope method used for statistics is not good, and it cannot distinguish smaller-sized defects, and pays more attention to the observation and analysis of the defect plane size. However, ignoring the relationship between the longitudinal dimension and defects leads to misjudgment, so this method will introduce too many system errors and human errors, and has little guiding significance for the research on cleaning solution formulations.

In addition, the samples in the experiment are usually small-sized wafers, which are about one-tenth the size of normal wafers. In order to carry out experiments better, corresponding experimental equipment is also required. It is not possible to just use the same proportion of related machines and other equipment. How to design a kind of experimental equipment suitable for laboratories and suitable for small-sized wafers is also a direction that those skilled in the art need to study in order to reduce the use.

Contents of the invention

The purpose of the present invention is to solve the above technical problems, to provide a simple and reliable measurement method, low experimental cost, good flexibility, effectively reduce system errors and human errors, high accuracy, and can provide powerful guidance for cleaning liquid formula research Method for evaluating wafer cleaning efficiency.

The present invention also provides an experimental device for evaluating wafer cleaning efficiency which is simple in structure, easy to manufacture, easy to disassemble, adjustable in experimental conditions and suitable for the above method.

The experimental device of the present invention comprises an integrated cleaning device, which includes:

Vacuum rotating device, used to vacuum absorb wafer samples and drive the wafer samples to rotate;

Injection device for injecting the adhesive solution or cleaning solution onto the surface of the wafer sample;

and a bracket for supporting the vacuum rotary device and the injection device.

It also includes an analysis device used in conjunction with the cleaning device to analyze sample defects before and after cleaning. Preferably, the analysis device includes a topography characterization instrument. More preferably, the topography characterization instrument is an atomic force microscope or a white light microscope. interferometer.

The vacuum rotary device includes a vacuum chuck with a vacuum groove on its surface, a rotation controller for controlling the rotary motor to drive the vacuum chuck to rotate, and a vacuum pump communicating with the bottom of the vacuum groove through a vacuum air pipe.

The vacuum chuck is located in the shielding case, and the top surface of the shielding case is provided with a glass cover.

The injection device includes a syringe and a pressurizing and pushing device that drives the piston rod of the syringe to move, and the needle of the syringe is made of non-metallic material.

The pressurizing and pushing device includes a screw-type stepping motor connected with a step controller or a telescopic cylinder, wherein the screw-type stepping motor is driven by a push rod to push the top piece to connect with the piston rod.

The vertical distance between the front end of the needle of the syringe and the surface of the wafer sample is 8-10 mm, and the diameter of the needle of the syringe is 0.5-2 mm.

A method for evaluating wafer cleaning efficiency using the above-mentioned experimental device for evaluating wafer cleaning efficiency, comprising the following steps:

(1) drip the sticky matter solution into the surface of the rotating wafer sample to carry out the contamination experiment, and obtain the contaminated wafer sample with defects on the surface;

(2) Use a morphology characterization instrument to perform morphology analysis on the surface of the contaminated wafer sample, and obtain the number of defects on the surface of the contaminated wafer sample before cleaning;

(3) The cleaning liquid is dripped into the surface of the contaminated wafer sample in the rotation to perform a cleaning experiment to obtain a cleaned wafer sample;

(4) Use a morphology characterization instrument to analyze the morphology of the cleaned wafer sample again to obtain the number of defects on the surface of the contaminated wafer sample after cleaning;

(5) Carry out the calculation of cleaning efficiency by following formula (A):

In the contamination experiment in the step (1) and the cleaning experiment in the step (3), the experimental device described in any one of claims 1-7 is used, and in the contamination experiment of the step (1), the syringe is equipped with If there is a contaminant solution, directly replace the corresponding syringe containing the cleaning solution in the cleaning experiment of step (3).

In the contamination experiment of step (1) and the cleaning experiment in step (3), the dripping speed of the sticky substance solution or cleaning solution is controlled to be 15-45ml/min, and the wafer sample or the contaminated wafer sample is rotated in two stages. Perform pre-rotation: control the speed at 500-800rpm, and the time is 60-90s; then switch to high-speed rotation, the speed is 1000-2000rpm, and the time is 100-180s.

The method for judging the defects in the steps (2) and (4) is: measure the protrusions on the surface of the wafer sample by a topography instrument, if the longitudinal height threshold of the protrusions is greater than 2nm, the protrusions are recorded as Defects, and so on, the total number of defects on the wafer sample is calculated as the number of defects.

The shape characterization instrument is an atomic force microscope or a white light interferometer, both of which can characterize the shape of the sample, form a 3D pattern by shooting or scanning, and measure the longitudinal height threshold of the protrusions on the surface of the sample. Those skilled in the art can refer to the equipment instructions and related documents.

The contaminants may be aqueous sols of silicon nanoparticles, dispersion solutions of flocculent metal complexes or aqueous solutions of organic molecules, etc.; the wafers may be silicon wafers, copper-plated silicon wafers, or silicon oxide thin film silicon wafers.

In view of the problems existing in the background technology, the inventor improved the existing experimental method. In order to simulate the use state of the cleaning solution as much as possible, the inventor first conducted a contamination experiment on the surface of the wafer sample to artificially create defects, and then cleaned the wafer. Cleaning experiments were carried out on the surface of the sample, and the cleaning efficiency of the cleaning solution can be calculated by detecting the change of the number of defects before and after cleaning, and substituting it into the formula. In the contamination experiment and cleaning experiment, soaking is no longer used as the experimental method, but the solution is added to the surface of the rotating wafer sample by dripping, which effectively simulates the working environment of the real wafer sample and reduces the experimental method. Error problem; further, the inventors found that after the contamination experiment, the particles attached to the surface of the wafer sample (forming bumps) are the main cause of affecting the function of the wafer and causing defects, such as silicon oxide particles, metal complex flocculent Residues, organic clusters, etc. These particles (also known as protrusions) usually have a certain vertical height threshold, which is not convenient to define with plane dimensions (such as diameter or circumference, etc.), so the invention changes the criteria for judging defects, by The morphology characterization instrument measures the longitudinal height threshold of the crystal surface to obtain the longitudinal heights of multiple protrusions. Combined with the roughness requirements of wafers in this field, it is set that if the longitudinal height of the protrusions is greater than 2nm, the protrusions will be recorded. is a defect, and so on, the total number of defects on the wafer sample can be calculated as the number of defects, and the determination method of defects before cleaning and after cleaning is the same. The above experimental method starts from the material characteristics that cause defects on the wafer surface in actual engineering applications, fundamentally reduces the possibility of errors, and has great guiding significance for the research and development of cleaning solutions.

Further, it is also possible to classify the longitudinal height thresholds of the protrusions judged as defects, and obtain the number of defects of each level in the wafer sample before and after cleaning respectively, and respectively substitute them into the formula (1) for calculation, which can be further analyzed to obtain The defect improvement status of each level of defects on the wafer sample surface before and after cleaning.

In the above-mentioned contamination experiment and cleaning experiment, two-stage rotation was adopted, and the low-speed pre-rotation was carried out to spread the dripped liquid evenly on the surface of the crystal sample, which played a certain wetting effect, and then switched to the high-speed rotation state, and the liquid was quickly It spreads on the surface of the crystal sample to form a thin film with uniform thickness, and drives the excess liquid from the surface of the sample by centrifugal force.

The structure of the experimental device of the present invention is simple and extremely simple. In view of the characteristics of the sample being a small-sized wafer, on the one hand, a syringe is used instead of a nozzle, and a pressurized pusher is used to drive the piston rod of the syringe to move, so that dripping or spraying water flow can be formed, and the flow rate can be flexibly adjusted. It solves the problem that the nozzle cannot be applied to the experimental state; the pressurization and pushing device is not limited to any form, such as a screw-type stepping motor or a telescopic cylinder or other structural forms, which can drive the piston rod to move downward. , by correspondingly pushing the controller to control the movement speed of the piston rod; on the other hand, the method of the present invention involves two dripping processes, one for the dripping of the contaminant solution and one for the cleaning solution, which involves the need to replace For different solutions, in order to ensure the reliability of the experiment, the container containing the clean liquid must not introduce pollutants and impurities. The existing liquid tank cannot solve the above problems. For this reason, the inventor used a syringe, which is not only simple in structure, but also can be flexibly disassembled from the bracket. By replacing the syringe with the corresponding solution, it can easily meet the experimental requirements of different stages and avoid mixing between the two solutions. Contamination, to further improve the reliability and accuracy of the experiment.

In order to simulate the rotation state of the wafer under actual working conditions, a vacuum rotation device is used. The wafer sample is placed on a vacuum chuck with a vacuum groove on the surface, and the wafer sample is vacuum adsorbed by a vacuum pump connected to the bottom of the vacuum groove through a vacuum tube. The rotation controller controls the rotating motor to drive the vacuum chuck to rotate, and finally drives the wafer sample to rotate. The rotation controller can control the rotation speed of the vacuum chuck and the rotation time at different speeds according to the needs, which is highly automated, flexible and reliable.

The experimental method of the present invention is simple and reliable, has good flexibility, effectively reduces system errors and human errors, has high accuracy, can provide powerful guidance for the research of cleaning liquid formula, and is especially suitable for the wet cleaning research of integrated circuits; the experimental device of the present invention is used In the experiment of the present invention, the structure is simple, easy to manufacture, easy to disassemble, flexible and adjustable experimental conditions, and suitable for experiments on small-sized wafer samples.

Description of drawings

Fig. 1 is a structural schematic diagram of the experimental device of the present invention.

Fig. 2 is a side view of the experimental device of the present invention.

Figure 3 is a top view of the vacuum chuck.

Among them, 1. Bracket, 2. Vacuum suction cup, 3. Rotary controller, 4. Screw type stepper motor, 5. Rotary motor, 6. Push top piece, 7. Piston rod, 8. Syringe, 9. Needle, 10. Glass cover, 11. Shielding cover, 12. Vacuum tube, 13. Step controller, 14. Vacuum pump, 15. Vacuum tank, 16. Wafer sample.

Detailed ways

Below in conjunction with accompanying drawing, experiment device of the present invention is further explained:

Referring to Fig. 1, it includes a vacuum rotating device: used for vacuum adsorption of the wafer sample and driving the wafer sample to rotate; an injection device: used for injecting the adhesive solution or cleaning solution onto the surface of the wafer sample; and bracket 1: used for installing the vacuum rotating device and Injection device. The injection part with the containing cavity in the injection device is detachable, so as to replace the contained adhesive solution or cleaning solution, and the injection part can be specifically a syringe.

It also includes an analysis device (not shown) used in conjunction with the cleaning device to analyze sample defects before and after cleaning. Preferably, the analysis device includes a topography characterization instrument. More preferably, the topography The characterization instrument is an atomic force microscope or a white light interferometer.

Described vacuum rotating device comprises the vacuum suction cup 2 that has vacuum groove 15 on the surface, the rotary motor 5 that drives vacuum suction cup 2 to rotate, the rotation controller 3 that connects rotary motor 5, and through vacuum air pipe 12 and vacuum groove 15 (this implementation Example is the vacuum pump 14 connected to the bottom of the king-shaped vacuum groove). The vacuum chuck 2 is located in the shielding case 11 , and the top surface of the shielding case is provided with a glass cover 10 .

Referring to Fig. 1 and Fig. 2, the injection device comprises a syringe 8 positioned above the vacuum chuck 2 and a pressure pushing device that drives the piston rod 7 of the syringe 8 to move, the needle 9 of the syringe 8 is made of non-metallic material, and the front end of the needle 9 is The vertical distance from the surface of the wafer sample 17 is 8-10 mm, and the aperture of the needle 9 is 0.5-2 mm.

The pressurizing and pushing device is a screw-type stepping motor or a telescopic cylinder connected with a stepping controller 13 . In this embodiment, a screw-type stepping motor 4 is adopted, which can convert the rotational motion of the motor into the linear reciprocating motion of the push rod, and the push rod drives the top piece 6 to apply the downward force to the piston rod 7 of the syringe 8 Up, the piston rod is moved downwards, so that the liquid in the syringe is dripped or flowed out from the needle 9 downwards. The screw-type stepping motor 4 can control the speed at which the push rod moves downward, thereby controlling the flow rate at which the liquid drips or flows out from the needle 9 downwards.

It should be pointed out that, in order to avoid the corrosion of metal materials by acid-base chemicals, the materials used in the present invention in contact with the cleaning liquid are all organic polymers, including but not limited to HDPE, LDPE, PP, etc.

Taking the sticking and dyeing experiment as an example, the usage method of the device of the present invention is introduced:

Put the sticky substance solution into the syringe 8, put the syringe 8 into the bracket 1, and connect the piston rod 7 to the screw type stepping motor; place the wafer sample 16 on the vacuum chuck 2, start the vacuum pump 14, and the negative pressure passes through the vacuum tube 12. The vacuum chamber 15 acts on the wafer sample 16, and the wafer sample 16 is vacuum-adsorbed on the vacuum chuck 2; start the rotating motor 5 to drive the vacuum chuck 2 to rotate, and control the speed and time of the rotation of the vacuum chuck 2 through the rotation controller 3; Start the screw-type stepping motor 4 to change the rotational force into a diameter movement, and push the top piece 6 to drive the piston rod 7 to move downward, and control the downward speed through the step controller 13, and then achieve the control of the needle 9 of the injection needle 8. The purpose of the dripping speed of the solution. After the contamination experiment is completed, take out the wafer for the next step of detection, and at the same time control the screw-type stepping motor 4 to reverse to drive the top sheet 6 back to its position, remove the syringe 8, replace the syringe 8 with cleaning solution and install it on the bracket 1 , in preparation for the next cleaning experiment. The operation process of the cleaning experiment is the same as that of the sticking experiment.

The method embodiment of evaluating wafer cleaning efficiency:

(1) use the experimental device of the present invention to drip the sticky matter solution into the surface of the wafer sample in rotation and carry out the contamination experiment, control the dripping speed of the sticky matter solution or cleaning solution to be 15 ~ 45ml/min, control the wafer sample or stain the wafer sample to carry out Two-stage rotation, first pre-rotation: the control speed is 500-800rpm, the time is 60-90s; then switch to high-speed rotation, the speed is 1000-2000rpm, the time is 100-180s, and the contaminated wafer sample with surface defects is obtained ;

(2) Use an atomic force microscope (purchased from Bruker) to analyze the morphology of the surface of the contaminated wafer sample, and obtain the number of defects on the surface of the contaminated wafer sample before cleaning; the determination method is: measure the protrusion on the surface of the wafer sample by the atomic force microscope , if the vertical height threshold of the protrusion is greater than 2nm, the protrusion is recorded as a defect, and so on, the total number of defects on the wafer sample is calculated as the number of defects before cleaning;

(3) use the experimental device of the present invention to drip the cleaning liquid into the surface of the contaminated wafer sample in the rotation and perform the cleaning experiment to obtain the cleaned wafer sample; the method of operation and control conditions are the same as in step (1), and the difference is that the injector 8 used for replacement is equipped with Washing fluid;

(4) Use an atomic force microscope (purchased from Bruker) to analyze the morphology of the wafer sample again to obtain the number of defects on the surface of the contaminated wafer sample after cleaning; the judgment method is the same as in step (3), and calculate the defects on the cleaned wafer sample The total is the number of defects after cleaning;

(5) Substitute the number of defects before cleaning and the number of defects after cleaning into the following formula (A) to calculate the cleaning efficiency:

(A);

(6) In order to further analyze and obtain the defect improvement state of the wafer sample before and after cleaning. It is also possible to classify the longitudinal height thresholds of protrusions judged as defects, and obtain the number of defects of each level in the wafer sample before and after cleaning, and respectively substitute them into formula (1) for calculation, and obtain the Defect cleaning efficiency. In this embodiment, the longitudinal height thresholds of defects are divided into three levels from low to high, namely low level: greater than 2nm to less than 5nm; middle level: greater than or equal to 5nm to less than 10nm; high level: greater than or equal to 10nm.

The atomic force microscope can scan the surface of the wafer sample. After obtaining the image data, the analysis software can convert the image into a 3D form, and then count the amount of contamination remaining on the surface, and can set the vertical height threshold in the software to filter out The number of protrusions higher than a certain vertical height can also be counted by inputting data at the vertical threshold to filter out the total number of defects of the corresponding size. Longitudinal dimensional characterization in AFM is obtained by changing the position of the probe to reflected laser light. Specifically, the probe probes on the surface of the sample without defects in the knocking mode, and the instrument will fit a reference plane, that is, the horizontal baseline of the sample. When the probe hits the defect position, the position of the baseline will change , at this time, the size change of the longitudinal dimension between the defect location and the horizontal position can be obtained from the analysis interface, and the longitudinal dimension of the defect can be judged.

In addition to the atomic force microscope, a white light interferometer (purchased from Bruker) can also be used to photograph the surface of the wafer sample, and then analyze the image to obtain the 3D topography of the surface, and express the height of the protrusions in different colors. By setting the vertical height threshold, Screen out the number of protrusions higher than a certain vertical height, and you can also input data at the vertical threshold during statistics to filter out the total number of defects of the corresponding size. Longitudinal dimension characterization with white light interferometers is achieved using visible light that produces interference fringes on the sample surface. Specifically, the magnification size is adjusted from the microscopic image of the sample surface by the white light interferometer, and a single interference fringe fills the entire microscopic field of view. According to the change of the optical signal, the instrument can distinguish the horizontal reference plane and surface protrusion of the sample, and The surface will be scanned and eventually several scanned areas can be bonded to form a large-scale microscopic pattern that gives the number of all defects on the surface.

experiment:

Below, the silicon nanoparticle aqueous sol is used as the contamination, and the copper ^- coated silicon wafer is used as the wafer. Company) cleaning efficiency measured as an example, and soaked in the same cleaning solution A after contamination, the number of defects with a size greater than 0.1 micron on the surface before and after cleaning were counted by the Compass defect analysis tool, and the cleaning efficiency calculated from this As a comparative example, the specific results are shown in Table 1.

Table 1 Cleaning Efficiency of Contaminants

It can be seen from Table 1 that the experimental method closer to the real working conditions is used in the embodiment, and the accuracy of defect detection is higher and the reliability is better. And it can also compare the cleaning effects of different levels of defects, which has better practical significance for the study of wafer cleaning and cleaning solution formulations that require higher defect rate control.

Claims (10)

1. An experimental device for evaluating wafer cleaning efficiency is characterized in that, comprising an integrated cleaning device, which includes: Vacuum rotating device, used to vacuum absorb wafer samples and drive the wafer samples to rotate; Injection device for injecting the adhesive solution or cleaning solution onto the surface of the wafer sample; and a bracket for supporting the vacuum rotary device and the injection device. 2. the experimental device of evaluating wafer cleaning efficiency as claimed in claim 1, is characterized in that, also comprises the analyzing device that is used to analyze the sample defect situation before and after cleaning in conjunction with described cleaning device, preferably, described analyzing device It includes a morphology characterization instrument, more preferably, the morphology characterization instrument is an atomic force microscope or a white light interferometer. 3. the experimental apparatus of evaluating wafer cleaning efficiency as claimed in claim 1, is characterized in that, described vacuum rotary device comprises the vacuum chuck that surface has vacuum groove, the rotation controller that controls rotation motor to drive vacuum chuck to rotate, and The vacuum pump communicates with the bottom of the vacuum tank through the vacuum air pipe. 4. The experimental device for evaluating wafer cleaning efficiency as claimed in claim 3, wherein the vacuum chuck is located in a shield, and the top surface of the shield is provided with a glass cover. 5. The experimental device for evaluating wafer cleaning efficiency as claimed in any one of claims 1 to 4, wherein the injection device includes a syringe and a pressurizing device that drives the piston rod of the syringe to move, and the needle of the syringe For non-metallic materials. 6. the experimental device of evaluating wafer cleaning efficiency as claimed in claim 5, is characterized in that, described pressurization pusher comprises the screw type stepper motor or telescopic cylinder that is connected with step controller, wherein, described The screw type stepping motor drives the push rod to connect the top piece with the piston rod. 7. The experimental device for evaluating wafer cleaning efficiency as claimed in claim 5, wherein the vertical distance between the front end of the needle of the syringe and the surface of the wafer sample is 8-10 mm, and the aperture of the needle of the syringe is 0.5-2 mm. 8. A method for evaluating wafer cleaning efficiency using the experimental device for evaluating wafer cleaning efficiency as claimed in any one of claims 1 to 7, characterized in that, comprising the following steps: (1) drip the sticky matter solution into the surface of the rotating wafer sample to carry out the contamination experiment, and obtain the contaminated wafer sample with defects on the surface; (2) Use a morphology characterization instrument to perform morphology analysis on the surface of the contaminated wafer sample, and obtain the number of defects on the surface of the contaminated wafer sample before cleaning; (3) The cleaning liquid is dripped into the surface of the contaminated wafer sample in the rotation to perform a cleaning experiment to obtain a cleaned wafer sample; (4) Use a morphology characterization instrument to analyze the morphology of the cleaned wafer sample again to obtain the number of defects on the surface of the contaminated wafer sample after cleaning; (5) Carry out the calculation of cleaning efficiency by following formula (A): In the contamination experiment in the step (1) and the cleaning experiment in the step (3), the experimental device according to any one of claims 1 to 7 is used, and in the contamination experiment in the step (1), the syringe is equipped with If there is a contaminant solution, directly replace the corresponding syringe containing the cleaning solution in the cleaning experiment of step (3). 9. the method for evaluating wafer cleaning efficiency as claimed in claim 8, is characterized in that, in the contamination experiment of step (1) and the cleaning experiment in step (3), the dripping speed of control sticky matter solution or cleaning solution is 15-45ml/min, control the wafer sample or contaminated wafer sample for two-stage rotation, first perform pre-rotation: control the rotation speed at 500-800rpm, the time is 60-90s; then switch to high-speed rotation, the rotation speed is 1000-2000rpm, the time 100-180s. 10. the method for evaluating wafer cleaning efficiency as claimed in claim 8, is characterized in that, the judging method of defect in described step (2) and step (4) is: by topography characterization instrument to the convexity of wafer sample surface The measurement is performed, and if the threshold height of the protrusion is greater than 2nm, the protrusion is recorded as a defect, and so on, and the total number of defects on the wafer sample is calculated as the number of defects.