A method for testing the cleanliness of semiconductor pipes and a method for cleaning the pipes

Abstract

The present invention relates to the field of related components in the semiconductor production process, specifically a method for testing the cleanliness of semiconductor pipes and a method for cleaning the pipes. The method comprises: washing the pipes with ultrapure water to remove impurities attached to the pipes; selecting electronic-grade hydrofluoric acid with a small molecular weight and extremely strong permeability as a pickling material, and performing deep immersion cleaning on the pipes through the pickling process to fully remove metal ions, anions, particles and other impurities in the pipes. After reaching the predetermined cleaning judgment standard, a cleanliness test is performed. The test process is to fill a certain amount of electronic-grade hydrofluoric acid, adopt a placement and interval test method for characterization, calculate the precipitation amount of impurities in the pipes based on the test data, and propose a cleaning method for semiconductor fluoropolymer pipes based on the above test method, which realizes flexible adjustment of cleaning cost and time and meets the use requirements of high-level wet chemicals / photoresists.

Description

Technical Field

[0001] The present invention relates to the field of related components in a semiconductor production process, and in particular to a method for testing the cleanliness of semiconductor pipes and a method for cleaning the pipes. Background Art

[0002] Currently, semiconductor manufacturing processes are advancing rapidly. As line widths become narrower and chip sizes shrink, the cleanliness requirements for wet chemicals and photoresists used in the manufacturing process are becoming increasingly stringent. Currently, mainstream 12-inch wafer foundries and other manufacturers have reached G5 cleanliness standards for wet chemicals. High-end photoresists are technically complex and require significant R&D investment. The manufacturing process requires strict control of impurities such as metal ions to ensure product stability. Product quality must meet at least G4 cleanliness standards. The following table lists the impurity requirements for the G5 / G4 cleanliness standards.

[0003] Table 1 Impurity requirements for G5 / G4 grade standards

[0004]

[0005] The pipes used in the production of wet electronic chemicals and photoresists are basically made of high-end fluoropolymer materials represented by HDPE, PTFE, and PFA. During the production and processing, impurities such as metal ions, anions, and particles are introduced. Metal ion impurities can easily cause chip short circuits, and particulate matter can easily cause chip defects. The degree of impurity precipitation in polymer materials directly affects the quality of wet electronic chemicals and photoresists. Therefore, the preparation of delivery pipes for wet electronic chemicals and high-end photoresists has also become a "bottleneck" in the semiconductor industry chain. At present, my country's low-end fluoropolymer production capacity is in excess and competition is fierce, but the high-end production capacity is insufficient and is monopolized by foreign countries. The breakthrough in the high-end transformation of fluoropolymers is the way out for the domestic industry. At present, many domestic companies have begun to get involved in high-end transformation.

[0006] During the development and production of tubing, characterization of impurity release is crucial, making tubing cleanliness testing methods crucial. However, domestic standards for tubing cleanliness testing are currently lacking, and there is a lack of sound cleanliness testing methods and corresponding tubing cleaning procedures. Therefore, it is crucial to design a rational and optimized cleanliness testing method for semiconductor tubing, integrating the characteristics of high-end wet electronic chemicals and photoresists, and, based on this foundation, to develop a comprehensive and optimized tubing cleaning methodology. Summary of the Invention

[0007] In view of the problem that a complete cleanliness test method for semiconductor pipes has not yet been established in the industry, the present invention provides a cleanliness test method for semiconductor pipes and a pipe cleaning method. It mainly includes: using ultrapure water to wash the pipes to remove impurities attached to the pipes; selecting electronic-grade hydrofluoric acid with a small molecular weight and extremely strong permeability as the pickling material, and performing deep immersion cleaning on the pipes through the pickling process to fully remove metal ions, anions, particles and other impurities in the pipes. After reaching the predetermined cleaning judgment standard, a cleanliness test is performed. The test process is to fill a certain amount of electronic-grade hydrofluoric acid, adopt the placement and interval testing method to characterize, and calculate the precipitation amount of impurities in the pipes based on the test data. Based on the above test method, a cleaning method for fluoropolymer pipes for semiconductors is proposed, which realizes the flexible adjustment of cleaning cost and time, and meets the use requirements of high-level wet chemicals / photoresists.

[0008] The complete technical solution of the present invention includes:

[0009] A method for testing the cleanliness of semiconductor pipes, characterized by comprising the following steps:

[0010] (1) Sample pre-processing and cleaning

[0011] (2) Soak to obtain the test solution

[0012] (3) Sampling analysis and statistics.

[0013] The sample pre-processing includes installing a valve or a sampling port on the tubular sample.

[0014] The sample cleaning includes water washing and acid washing.

[0015] The water washing uses ultrapure water with a metal ion content of less than 3 ng / kg, and the acid washing uses 49% electronic grade hydrofluoric acid with a metal ion content of less than 3 ng / kg.

[0016] Furthermore, the sample cleaning is specifically as follows:

[0017] Water washing: Fill the pipe sample with ultrapure water containing less than 3ng / kg of metal ions, tap and shake the pipe gently to soak for 2 minutes, then drain the ultrapure water, and repeat the above operation 10 times.

[0018] Acid cleaning: After water washing, replace ultrapure water with 49% electronic grade hydrofluoric acid with metal ions less than 3ng / kg to fill the pipe sample, tap and shake the pipe for 2 minutes, soak for about 2 hours, drain the hydrofluoric acid, and repeat the above operation twice.

[0019] After the pickling is completed, refill the pipe with 49% electronic grade hydrofluoric acid with a metal ion content of less than 3ng / kg, tap and shake the pipe for 2 minutes, drain the hydrofluoric acid, and refill the pipe with 49% electronic grade hydrofluoric acid with a metal ion content of less than 3ng / kg, tap and shake the pipe for 2 minutes.

[0020] The soaking includes soaking in 49% electronic grade hydrofluoric acid having a metal ion content of less than 3 ng / kg.

[0021] The soaking times were 2 hours, 1 day, 3 days, 5 days and 7 days respectively.

[0022] The sampling and analysis uses ultra-pure PFA bottles with metal ion precipitation less than 1 ppt to analyze metal ions, anions, and particulate impurities.

[0023] The cleanliness standard is determined as follows:

[0024] β=η / V

[0025] β: Cleaning degree of polymer tubing, unit: μg / L;

[0026] η: the amount of impurities precipitated in the cleaning solution, in μg;

[0027] V: The volume of the pipe in contact with the cleaning fluid, unit: L.

[0028] A semiconductor pipe cleaning method based on the test method includes:

[0029] 1) Calculate the precipitation rate of various ions based on the precipitation amount of all metal ions and anions in each test solution obtained:

[0030]

[0031] Among them, v k is the ion precipitation rate in the kth immersion time period, σ k is the amount of ions released during the kth immersion period, t k is the soaking time of the kth soaking period;

[0032] Compare the variation ranges of the precipitation rates of various ions, and define the anions and cations with the smallest variation ranges as the benchmark anions and benchmark cations;

[0033] 2) cleaning the pipes using different water wash times, water wash times, acid wash times, and acid wash times, and soaking the cleaned pipes in 49% electronic-grade hydrofluoric acid with a metal ion concentration of less than 3 ng / kg for 3 days, sampling and analyzing the precipitation amounts of reference cations and reference anions; establishing a data set, wherein each data item includes: data ID, water wash times, water wash times, acid wash times, acid wash times, reference cation precipitation amounts, and reference anion precipitation amounts; performing regression analysis on the data to obtain the relationship between the reference cations and anions and the cleaning parameters;

[0034] 3) Based on the ion precipitation requirements for different usage scenarios, the cleaning parameters are determined according to the relationship between the above-mentioned benchmark cations and anions and the cleaning parameters.

[0035] The pipe is a fluoropolymer pipe for semiconductors.

[0036] Compared with the prior art, in the current field of pipe cleanliness testing, only some high-end manufacturers use ultrapure water to detect changes in water impurities after cleaning, which cannot fully characterize the impurity precipitation of the pipe. The present invention is aimed at the problem that the radius of the fluoride ion in hydrofluoric acid is very small, even smaller than the oxygen ion, which causes it to have strong permeability and can penetrate the gaps in the surface molecular chains of polymers such as HDPE, PTFE, and PFA better than other chemicals. At the same time, hydrofluoric acid can strongly corrode metals, glass and silicon-containing objects, and can fully dissolve impurities such as dust attached to the pipe. Based on the above testing method, a cleaning method for fluoropolymer pipes for semiconductors is proposed. According to specific needs, a reasonable number and time of water washing / acid washing are selected for pipe cleaning, thereby realizing flexible adjustment of cleaning cost and time. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 Flow chart of the testing method of the present invention. DETAILED DESCRIPTION

[0038] The present invention is described in detail below with reference to the embodiments and accompanying drawings. However, it should be understood that the embodiments and accompanying drawings are merely exemplary descriptions of the present invention and do not constitute any limitation on the scope of protection of the present invention. All reasonable variations and combinations within the scope of the inventive concept of the present invention fall within the scope of protection of the present invention.

[0039] like Figure 1 As shown, the present invention first discloses a method for testing the cleanliness of semiconductor pipes. Based on the characteristics of hydrofluoric acid, a complete testing process for the cleanliness of semiconductor pipes is creatively proposed, including preliminary treatment of pipe samples, determination of cleaning standards and specific testing methods.

[0040] The test method specifically includes the following steps:

[0041] (1) Sample pre-processing and cleaning

[0042] Pre-process the sample according to its shape. For example, a pipe-shaped sample needs to be equipped with a valve and sampling port, and the sample is cleaned with ultrapure water and electronic-grade hydrofluoric acid. In this embodiment, the unified cleaning process for subsequent testing is as follows:

[0043] Water washing: Fill the pipe sample with ultrapure water containing less than 3ng / kg of metal ions, tap and shake the pipe gently to soak for 2 minutes, then drain the ultrapure water, and repeat the above operation 10 times.

[0044] Acid cleaning: After water washing, replace ultrapure water with 49% electronic grade hydrofluoric acid with metal ions less than 3ng / kg to fill the pipe sample, tap and shake the pipe for 2 minutes, soak for about 2 hours, drain the hydrofluoric acid, and repeat the above operation twice.

[0045] After the pickling is completed, refill the pipe with 49% electronic grade hydrofluoric acid with a metal ion content of less than 3ng / kg, tap and shake the pipe for 2 minutes, drain the hydrofluoric acid, and refill the pipe with 49% electronic grade hydrofluoric acid with a metal ion content of less than 3ng / kg, tap and shake the pipe for 2 minutes to complete the cleaning process.

[0046] (2) Soak to obtain the test solution

[0047] After cleaning, the sample was immersed in 49% electronic grade hydrofluoric acid with metal ions less than 3 ng / kg, respectively: 2 hours to obtain test solution 1, 1 day to obtain test solution 2, 3 days to obtain test solution 3, 5 days to obtain test solution 4, and 7 days to obtain test solution 5.

[0048] (3) Sampling analysis and statistics

[0049] Each test liquid was sampled using an ultra-pure PFA bottle (metal ion precipitation less than 1 ppt), and the precipitated metal ion, anion and particle content in each test liquid was analyzed.

[0050] To statistically analyze the data, first use the following cleaning standard determination formula to calculate the amount of impurities precipitated during each interval:

[0051] β=η / V

[0052] β: Cleaning degree of polymer tubing, unit: μg / L;

[0053] η: the amount of impurities precipitated in the cleaning solution, in μg;

[0054] V: The volume of the pipe in contact with the cleaning fluid, unit: L.

[0055] For metal ions and anions, the qualification criteria are: less than the ion precipitation content requirements in Table 2 and Table 3. If there are specified or other cleaning requirements, they shall be implemented accordingly.

[0056] Table 2 Anion requirements in test solution

[0057] Anions Test Method Test solution anion precipitation content requirement (μg / L) chloride JY / T 0575-2020 ≤20 Nitrates JY / T 0575-2020 ≤10 Phosphate JY / T 0575-2020 ≤10 sulfate JY / T 0575-2020 ≤20

[0058] Table 3 Requirements for cations in test solution

[0059]

[0060]

[0061] Currently, semiconductor manufacturing focuses on impurities in wet electronic chemicals / photoresists, primarily focusing on metal ions, particles, and anions. The impurity detection parameters selected for this method, listed in Table 4, meet industry requirements. The instrumentation used in this experiment has a detection limit of less than 0.1 ppt for metal ions and less than 0.1 ppb for anions, covering a particle size range of 30 nm to 0.5 μm.

[0062] Table 4 Instrument detection limit and impurity detection items

[0063]

[0064] In summary, this test method utilizes G5-level electronic-grade hydrofluoric acid, a relatively small molecular weight and highly permeable wet electronic chemical, as the soaking agent for fluoropolymer tubing. This soaking process maximizes the precipitation of impurities contained in the tubing. Interval testing is then used to analyze the degree of impurity precipitation in the tubing. The precipitation amount is calculated according to the precipitation determination formula, thereby determining the appropriate grade of electronic chemicals / photoresist for the tubing.

[0065] The present invention also discloses a method for cleaning semiconductor fluoropolymer pipes based on the above test method, comprising the following steps:

[0066] 1) For the test solutions 1-5 obtained in step (2) at different immersion time periods, sampling and analysis were performed using the method described in step (3), and the precipitation rates of various ions were calculated for all metal ions and anions precipitated in each test solution obtained:

[0067]

[0068] Among them, v k is the ion precipitation rate in the kth immersion time period, σ k is the amount of ions released during the kth immersion period, t kis the soaking time of the kth soaking period, where t1 = 2 h, t2 = 24 h, t3 = 72 h, t4 = 120 h, and t5 = 168 h.

[0069] For each type of ion, the precipitation rate of 5 time periods is obtained. Under normal circumstances, the precipitation rate will decrease with the increase of immersion time. On this basis, the change range of the precipitation rate of each ion is compared, and the anion and cation with the smallest change range are defined as the benchmark anion and benchmark cation. The above method is to determine the difficulty and speed of various metal ions, anions, and particles through the degree of change in the precipitation rate, and the ion that precipitates the slowest is used as the benchmark ion. In the subsequent data statistical process, the study of the precipitation process of all ions can be completed by only examining the benchmark ion as a representative, which significantly reduces the required amount of data and saves costs.

[0070] The amplitude of the change can be calculated in two ways. The first is to simply use Δv=v5-v1 and take the ion with the smallest Δv as the reference ion. The second is to calculate Where, is the average precipitation rate of the ion, and S 2 The smallest ion is used as the reference ion.

[0071] 2) Cleaning is performed by water washing + pickling, wherein the water washing parameters include the number of water washings, the time of each water washing, the number of picklings, and the time of each pickling. The pipes are cleaned with different number of water washings, the time of each water washing, the number of picklings, and the time of each pickling. The cleaned pipes are soaked in 49% electronic grade hydrofluoric acid with a metal ion content of less than 3ng / kg for 3 days, and the precipitation amount of the benchmark cations and benchmark anions is sampled and analyzed. The above data are used to establish a data set, and each data item includes: data ID, number of water washings, the time of each water washing, the number of picklings, the time of each pickling, the precipitation amount of the benchmark cations, and the precipitation amount of the benchmark anions. The data are subjected to regression analysis to obtain the relationship between the benchmark cations and anions and each parameter:

[0072] Y p =β p0 +β p1 X1+β p2 X2+β p3 X3+β p4 X4+∈ p

[0073] Y N =β N0 +β N1 X1+β N2 X2+β N3 X3+β N4 X4+∈ pN

[0074] Where Y represents the amount of reference ion precipitation, ∈ is the error term, X1 is the number of water washes, X2 is the time for each water wash, X3 is the number of acid washes, X4 is the time for each acid wash, β0 is the intercept term, β1~β4 are the coefficients of each parameter, subscript p represents the reference cation, and subscript N represents the reference anion.

[0075] Generally speaking, increasing the number of water washing and pickling and the time of each cleaning will help remove more impurities and ions and improve the cleanliness of the pipe, but it will significantly increase time and cost, leading to unnecessary waste of resources. Therefore, this embodiment obtains the above relationship through data statistics. It can be based on different semiconductor application scenarios, such as the use requirements of different grades of electronic chemicals / photoresists, to define the required cation and anion precipitation amounts. Since the reference ion is the slowest ion, the cleaning process that meets its precipitation requirement can be considered to be in compliance with all ions. And for this specific ion precipitation amount requirement, the number and time of reasonable water washing / pickling are selected to carry out pipe cleaning, which can flexibly adjust the cleaning cost and time under the premise of meeting the ion precipitation amount requirement.

[0076] The above applications are only some embodiments of the present application. For those skilled in the art, without departing from the inventive concept of the present application, several modifications and improvements can be made, which all fall within the scope of protection of the present application.

Claims

1. A semiconductor pipe cleaning method based on a semiconductor pipe cleanliness test method, characterized in that: The testing method comprises the following steps: (1) Preliminary sample processing and cleaning; (2) soaking to obtain a test solution; the soaking comprises soaking the sample in 49% electronic grade hydrofluoric acid having a metal ion content of less than 3 ng / kg, wherein the sample is soaked for 2 hours to obtain test solution 1, for 1 day to obtain test solution 2, for 3 days to obtain test solution 3, for 5 days to obtain test solution 4, and for 7 days to obtain test solution 5; (3) Sampling analysis and statistics: The sampling analysis uses ultra-pure PFA bottles with metal ion precipitation less than 1 ppt to sample each test liquid and analyze the content of metal ions, anions and particles precipitated in each test liquid; The cleaning method comprises: 1) For the test solutions 1-5 obtained in step (2) at different immersion time periods, sampling and analysis were performed using the method described in step (3), and the precipitation rates of various ions were calculated for all metal ions and anions precipitated in each test solution obtained: Among them, v k is the ion precipitation rate in the kth immersion time period, σ k is the amount of ions released during the kth immersion period, t k is the soaking time of the kth soaking period, where t1 = 2 h, t2 = 24 h, t3 = 72 h, t4 = 120 h, and t5 = 168 h; Compare the variation range of each ion precipitation rate, and define the anion and cation with the smallest variation range as the benchmark anion and benchmark cation; the variation range is calculated by calculating Where, is the average precipitation rate of the ion, and S 2 The smallest ion is used as the reference ion; 2) cleaning the pipes using different water wash times, water wash times, acid wash times, and acid wash times, and soaking the cleaned pipes in 49% electronic-grade hydrofluoric acid with a metal ion concentration of less than 3 ng / kg for 3 days, sampling and analyzing the precipitation amounts of reference cations and reference anions; establishing a data set, wherein each data item includes: data ID, water wash times, water wash times, acid wash times, acid wash times, reference cation precipitation amounts, and reference anion precipitation amounts; performing regression analysis on the data to obtain the relationship between the reference cations and anions and the cleaning parameters; Y p =b p0 +b p1 X1+β p2 X2+β p3 X3+β p4 X4+∈ p Y N =b N0 +b N1 X1+β N2 X2+β N3 X3+β N4 X4+∈ pN Where Y represents the amount of reference ion precipitation, ∈ is the error term, X1 is the number of water washes, X2 is the time for each water wash, X3 is the number of acid washes, X4 is the time for each acid wash, β0 is the intercept term, β1~β4 are the coefficients of each parameter, subscript p represents the reference cation, and subscript N represents the reference anion; 3) According to the ion precipitation requirements of different usage scenarios, water washing + acid washing is used for cleaning, where the cleaning parameters include the number of water washings, the time of each water washing, the number of acid washings, the time of each acid washing, and the cleaning parameters are determined based on the relationship between the above-mentioned benchmark cations and anions and the cleaning parameters.

2. The semiconductor pipe cleaning method according to claim 1, wherein: The sample pre-processing includes installing a valve or a sampling port on the tubular sample.

3. The semiconductor pipe cleaning method according to claim 2, wherein: Sample cleaning includes water washing and acid washing.

4. The semiconductor pipe cleaning method according to claim 3, wherein: The water washing uses ultrapure water with a metal ion content of less than 3 ng / kg, and the acid washing uses 49% electronic grade hydrofluoric acid with a metal ion content of less than 3 ng / kg.

5. The semiconductor pipe cleaning method according to claim 4, wherein: The pipe is a fluoropolymer pipe for semiconductors.

Images