KR102668680B1 - Deposition process monitoring method and device thereof, thin film transistor manufacturing method - Google Patents

Abstract

The deposition process monitoring device according to one embodiment generates a model for predicting the amount of hydrogen in the amorphous silicon layer through machine learning using the deposition variables of the set deposition equipment and the previously measured amount of hydrogen in the amorphous silicon (a-Si) layer. Prediction model generation unit; A prediction unit that predicts the amount of hydrogen in the amorphous silicon layer according to the deposition variables of the deposition equipment measured in real time, based on the hydrogen amount prediction model; and a defect determination unit that determines whether the predicted amount of hydrogen falls within a normal range.

Description

Deposition process monitoring method and device, thin film transistor manufacturing method {DEPOSITION PROCESS MONITORING METHOD AND DEVICE THEREOF, THIN FILM TRANSISTOR MANUFACTURING METHOD}

The present disclosure relates to a deposition process monitoring method and monitoring device and a thin film transistor manufacturing method.

The organic light emitting display device includes a pixel driving circuit for each pixel, and the pixel driving circuit includes a thin film transistor using silicon. Amorphous silicon (a-Si) or polycrystalline silicon (poly silicon) is used as the silicon that makes up the thin film transistor.

An amorphous silicon thin film transistor (a-Si TFT) has low electron mobility because the semiconductor active layer constituting the source, drain, and channel is amorphous silicon. Accordingly, there is a recent trend to replace the amorphous silicon thin film transistor with a polycrystalline silicon thin film transistor (polycrystalline silicon TFT: poly-Si TFT).

Since polycrystalline silicon can be fabricated by crystallizing amorphous silicon, the thin film transistor manufacturing process begins by depositing amorphous silicon on a substrate.

In addition, whether the deposition process on the substrate was properly performed can be confirmed by destruction analysis of the amorphous silicon layer after the deposition process is completed.

The embodiments are intended to confirm the state of the amorphous silicon layer during the deposition process without analyzing the destruction of the amorphous silicon layer in the thin film deposition process.

The deposition process monitoring device according to one embodiment generates a model for predicting the amount of hydrogen in the amorphous silicon layer through machine learning using the deposition variables of the set deposition equipment and the previously measured amount of hydrogen in the amorphous silicon (a-Si) layer. Prediction model generation unit; A prediction unit that predicts the amount of hydrogen in the amorphous silicon layer according to the deposition variables of the deposition equipment measured in real time, based on the hydrogen amount prediction model; and a defect determination unit that determines whether the predicted amount of hydrogen falls within a normal range.

an input unit that receives the deposition parameters of the preset deposition equipment and the previously measured amount of hydrogen in the amorphous silicon layer; And it may further include a deposition variable collection unit that collects deposition variables of the deposition equipment in real time.

The input unit may provide the preset deposition parameters of the deposition equipment and the previously measured hydrogen amount of the amorphous silicon layer to the prediction model generator.

The deposition variable collection unit may provide the collected deposition variables of the deposition equipment to the prediction unit.

If the amount of hydrogen in the amorphous silicon (a-Si) layer is 8 atm% or less before the dehydrogenation process, the defect determination unit may determine the amorphous silicon layer to be within the normal range.

If the amount of hydrogen in the amorphous silicon (a-Si) layer is 3.0 atm% or less after the dehydrogenation process, the defect determination unit may determine the amorphous silicon layer to be within the normal range.

It may further include a display unit that displays whether the amorphous silicon layer is defective or normal according to the defect determination unit.

The deposition variables of the deposition equipment include the temperature of the equipment susceptor, argon (Ar), silane (SiH ₄ ), ammonia (NH ₃ ), hydrogen (H ₂ ), nitrous oxide (N ₂ O), and nitrogen (N ₂ ) It may correspond to at least one of the gas inflow amount, and the pressure and temperature inside the chamber.

The machine learning may be performed based on at least one of a decision tree, a neural network, and a support vector machine (SVM).

The deposition process monitoring method according to one embodiment generates a model for predicting the amount of hydrogen in the amorphous silicon layer through machine learning using the deposition variables of the preset deposition equipment and the previously measured amount of hydrogen in the amorphous silicon (a-Si) layer. steps; Based on the hydrogen amount prediction model, predicting the amount of hydrogen in the amorphous silicon layer according to the deposition variables of the deposition equipment measured in real time; and determining whether the predicted amount of hydrogen falls within a normal range.

In the step of predicting the amount of hydrogen in the amorphous silicon layer according to the deposition variable of the deposition equipment measured in real time, the deposition variable of the deposition equipment is determined in the step of forming the amorphous silicon layer on the substrate and the dehydrogenation process step of the amorphous silicon layer. It can be measured in real time.

In the step of determining whether the predicted amount of hydrogen corresponds to the normal range, the normal range before the dehydrogenation process may correspond to the amount of hydrogen in the amorphous silicon (a-Si) layer of 8 atm % or less.

In the step of determining whether the predicted amount of hydrogen corresponds to the normal range, the normal range after the dehydrogenation process may correspond to the amount of hydrogen in the amorphous silicon (a-Si) layer of 3.0 atm % or less.

In the step of determining whether the predicted hydrogen amount falls within the normal range, if it falls within the normal range, the step of crystallizing the amorphous silicon into polycrystalline silicon may proceed.

In the step of determining whether the predicted amount of hydrogen is within the normal range, if it is outside the normal range, preventive maintenance (PM) of the deposition facility may be performed.

The deposition variables of the deposition equipment include the temperature of the equipment susceptor, argon (Ar), silane (SiH ₄ ), ammonia (NH ₃ ), hydrogen (H ₂ ), nitrous oxide (N ₂ O), and nitrogen (N ₂ ) It may correspond to at least one of the gas inflow amount, and the pressure and temperature inside the chamber.

The machine learning may be based on at least one of a decision tree, a neural network, and a support vector machine.

A thin film transistor manufacturing method according to an embodiment includes forming an amorphous silicon layer on a substrate through a deposition process; Crystallizing the amorphous silicon layer into a polycrystalline silicon layer; And patterning the polycrystalline silicon layer to form a semiconductor layer, and predicting the amount of hydrogen in the amorphous silicon layer according to the deposition parameters of the deposition process.

After forming the amorphous silicon layer, a dehydrogenation process to lower the hydrogen concentration of the amorphous silicon layer may be further included.

forming the semiconductor layer and forming a gate insulating film on the semiconductor layer; forming a gate electrode on the gate insulating film; forming an interlayer insulating film on the gate electrode and forming an opening to expose the source region and drain region of the semiconductor layer; and forming a source electrode and a drain electrode to be respectively connected to the source region and the drain region.

According to embodiments, defects in the amorphous silicon layer can be prevented by monitoring the amount of hydrogen in the amorphous silicon layer in real time through a hydrogen amount prediction model in the thin film deposition process.

That is, by checking defects in the amorphous silicon layer in advance during the thin film transistor manufacturing process, the time required for the thin film transistor manufacturing process can be shortened and costs can be reduced.

Figure 1 sequentially shows a method of manufacturing a thin film transistor according to an embodiment.
Figure 2 is a flowchart showing a deposition process monitoring method according to one embodiment.
FIG. 3 is a conceptual diagram illustrating a rule for deriving the amount of hydrogen in a deposition process monitoring method according to an embodiment.
FIG. 4 exemplarily illustrates a state in which a defective state of an amorphous silicon layer can be confirmed according to a deposition variable of a deposition facility among the deposition process monitoring methods according to an embodiment.
Figure 5 shows a block diagram of a deposition process monitoring device according to an embodiment.
FIG. 6 is a contour diagram illustrating the relationship between deposition variables of deposition equipment during a deposition process according to an embodiment.
FIG. 7 is an exemplary diagram illustrating the defective state of an amorphous silicon layer according to the deposition parameters of the deposition equipment during the deposition process monitoring method according to an embodiment.

Hereinafter, with reference to the attached drawings, various embodiments of the present invention will be described in detail so that those skilled in the art can easily practice the present invention. The invention may be implemented in many different forms and is not limited to the embodiments described herein.

In order to clearly explain the present invention, parts that are not relevant to the description are omitted, and identical or similar components are given the same reference numerals throughout the specification.

In addition, the size and thickness of each component shown in the drawings are arbitrarily shown for convenience of explanation, so the present invention is not necessarily limited to what is shown. In the drawing, the thickness is enlarged to clearly express various layers and regions. And in the drawings, for convenience of explanation, the thicknesses of some layers and regions are exaggerated.

Additionally, when a part of a layer, membrane, region, plate, etc. is said to be “on” or “on” another part, this includes not only cases where it is “directly above” another part, but also cases where there is another part in between. . Conversely, when a part is said to be “right on top” of another part, it means that there is no other part in between. In addition, being “on” or “on” a reference part means being located above or below the reference part, and does not necessarily mean being located “above” or “on” the direction opposite to gravity. .

In addition, throughout the specification, when a part is said to "include" a certain component, this means that it may further include other components rather than excluding other components, unless specifically stated to the contrary.

In addition, throughout the specification, when referring to “on a plane,” this means when the target portion is viewed from above, and when referring to “in cross-section,” this means when a cross section of the target portion is cut vertically and viewed from the side.

Below, we will first briefly look at the manufacturing method of the thin film transistor.

1 shows a method of manufacturing a thin film transistor according to an embodiment.

Referring to FIG. 1, step (a) is a step of preparing a substrate 100, such as glass or plastic, and cleaning the substrate 100. Step (a) is also called the substrate cleaning step. In order to remove impurities such as residual film, metal, and transparent conductive metal located on the substrate 100, the substrate 100 is first cleaned.

Step (b) is a step of forming a buffer layer 110 and an amorphous silicon (a-Si) layer 20 on the substrate 100. Step (b) is also called the deposition step. The buffer layer 110 is intended to provide smoothness and block the penetration of impurity elements, and may include silicon dioxide (SiO ₂ ) or nitrogen oxide (SiNx). The amorphous silicon layer 20 may be deposited using chemical vapor deposition (CVD) or physical vapor deposition (PVD). In this embodiment, chemical vapor deposition is used, mainly plasma enhanced chemical vapor deposition (PECVD). Additionally, in step (b), a process of lowering the hydrogen concentration may be performed by performing dehydrogenation treatment when or after forming the amorphous silicon layer 20. The dehydrogenation process will be examined in detail after looking at Figure 1.

Step (c) is a step of crystallizing amorphous silicon into polysilicon using excimer laser annealing (ELA). Step (c) is also called the crystallization step. The excimer laser annealing method is a method of crystallizing the amorphous silicon layer 20 by injecting an excimer laser into the amorphous silicon layer 20 and locally heating it to a high temperature for a very short time. Depending on the embodiment, when crystallizing amorphous silicon into polycrystalline silicon, solid phase crystallization (SPC), metal induced crystallization (MIC), or metal induced lateral crystallization (MILC) may be used. ) can be used.

Step (d) is a step of patterning the polycrystalline silicon layer 30 to form the semiconductor layer 130 of the thin film transistor. Step (d) is also called the semiconductor layer formation step. The semiconductor layer 130 can be formed by applying photoresist on the polycrystalline silicon layer 30, exposing the exposed area where the photoresist is not applied, and then etching the exposed area. Here, the semiconductor layer 130 includes a channel region 131 doped with impurities, a source region 132 not doped with impurities, and a drain region 133.

Step (e) is a step of forming the gate insulating film 140 on the entire surface of the substrate including the semiconductor layer 130. Step (e) is also called the insulating film formation step. The gate insulating film 140 may be formed of silicon oxynitride (SiO _x ), silicon nitride (SiN _x ), or a double layer containing these.

Step (f) is a step of forming the gate electrode 121 on the gate insulating film 140. Step (f) is also called the gate electrode forming step. The gate electrode 121 may include a single layer of aluminum (Al) or copper (Cu), or a multiple layer of aluminum stacked on a chromium (Cr) or molybdenum (Mo) alloy. Although not specifically shown in FIG. 1, the gate electrode 121 is formed by forming a metal layer for the gate electrode, and etching the metal layer for the gate electrode through a photolithography and etching process to form a metal layer corresponding to the channel region 131 of the semiconductor layer 130. It can be formed in part. Thereafter, an ion doping process containing impurities may be performed in the source region 132 and drain region 133 of the semiconductor layer 130.

Step (g) is a step of cleaning the gate electrode 121 and the gate insulating film 140. Step (g) is also called the electrode and insulating film cleaning step.

In step (h), an interlayer insulating film 150 is formed on the entire surface of the substrate 100 including the gate electrode 121, and the interlayer insulating film 150 and the gate insulating film 140 are etched to form the source region of the semiconductor layer 130. This is the step of forming an opening 45 that exposes 132 and drain region 133. Step (h) is also called the opening formation step. The interlayer insulating film 150 may be formed of silicon oxynitride (SiOX), silicon nitride (SiNx), or a multilayer containing these. In step (h), after forming the opening 45, an annealing process may be performed to heat the substrate 100 to remove damage to the semiconductor layer 130.

Step (i) is a step of forming the source electrode 152 and the drain electrode 153 in the opening 45 to be connected to the source region 132 and the drain region 133, respectively. Step (i) is also called the source-drain electrode forming step. The source electrode 152 and the drain electrode 153 are made of molybdenum (Mo), chromium (Cr), tungsten (W), molybdenum tungsten (MoW), aluminum (Al), aluminum-neodymium (Al-Nd), and titanium (Ti). ), titanium nitride (TiN), copper (Cu), molybdenum alloy (Mo alloy), aluminum alloy (Al alloy), and copper alloy (Cu alloy). Accordingly, the thin film transistor can be implemented as one device, including the semiconductor layer 130, gate electrode 121, source electrode 152, and drain electrode 153 described above.

In step (j), a protective film 160 is formed on the entire surface of the substrate 100 including the source electrode 152 and the drain electrode 153, and the protective film 160 is etched to form the source electrode 152 or the drain electrode 153. ) is the step of forming an opening 65 that exposes. Step (j) is also called the protective film formation step. Here, the protective film 160 is selected from silicon oxynitride (SiOX), silicon nitride (SiNx), which is an inorganic film, or polyimide, benzocyclobutene series resin, and acrylate, which are organic films. It can be done in any one way. Additionally, it may be composed of a layered structure of an inorganic film and an organic film.

Step (k) is a step of forming the anode 191 in the opening 65 to be connected to the source electrode 152 or the drain electrode 153. Step (k) is also called the anode formation step. The anode 191 may be made of a transparent conductive film made of any one of ITO, IZO, or ITZO.

Step (l) is a step of forming a partition wall 370 having an opening 85 exposing a portion of the surface of the anode 191 on the protective film 160 and the anode 191. The (ㅣ) stage is also called the partition formation stage. Although not shown in FIG. 1, an organic emission layer (not shown) may be formed on the exposed anode 191, and a cathode (not shown) may be formed on the organic emission layer and the partition wall 370. Here, the organic light-emitting layer may include one or more layers selected from the group consisting of a hole injection layer, a hole transport layer, a hole suppression layer, an electron injection layer, and an electron transport layer. The anode, organic light-emitting layer, and cathode constitute one organic light-emitting device (OLED).

In the above, we looked at a method of manufacturing a thin film transistor according to an embodiment.

The present invention is intended to predict defects in thin film transistor elements that may occur during the deposition of the amorphous silicon layer 20 during the deposition process. Hereinafter, the dehydrogenation process mentioned in step (b) of FIG. 1 will be examined in detail. .

In step (b), a process of lowering the hydrogen concentration may be performed by performing dehydrogenation treatment when or after forming the amorphous silicon layer 20 on the substrate 100.

Specifically, since the plasma chemical vapor deposition method uses silane (SiH4) as a raw material gas, a large amount of hydrogen may be contained in the amorphous silicon layer 20. If the amorphous silicon layer 20 is crystallized without removing hydrogen from the amorphous silicon layer 20, the hydrogen may roughen the surface of the polycrystalline silicon layer 30.

If the surface of the polycrystalline silicon layer 30 is formed to be rough, the movement of charges is hindered, so a thin film transistor including the polycrystalline silicon layer 30 with high surface roughness has deteriorated device characteristics. To prevent this problem, a dehydrogenation process to remove hydrogen in the amorphous silicon layer 20 is performed before irradiating the amorphous silicon layer 20 with a laser. The dehydrogenation process can be performed by heating the temperature in the chamber to about 550°C or higher. In the dehydrogenation process, the hydrogen in the amorphous silicon layer 20 breaks the silicon-hydrogen bond in a high temperature environment and evaporates, forming the amorphous silicon layer 20. The hydrogen content within decreases.

The present invention predicts the hydrogen content of the amorphous silicon layer 20 by the dehydrogenation process, so that device defects can be detected in advance before the device process of the thin film transistor is completed. Therefore, hereinafter, with reference to FIGS. 2 to 3 and step (b) of FIG. 1 described above, we will look at a method for predicting hydrogen content.

FIG. 2 is a flow chart illustrating a deposition process monitoring method according to an embodiment, and FIG. 3 is a conceptual diagram illustrating a rule for deriving the amount of hydrogen in a thin film transistor device characteristic monitoring method according to an embodiment.

First, referring to FIG. 2, the hydrogen amount of the amorphous silicon layer 20 before and after the dehydrogenation process measured under the previously set deposition process equipment parameters and corresponding equipment conditions is shown. Provided (S10), a model for predicting deposition variables and the amount of hydrogen in the amorphous silicon layer 20 is created (S20). Here, the deposition variables are the temperature of the susceptor of the deposition device, argon (Ar), silane (SiH ₄ ), ammonia (NH ₃ ), hydrogen (H ₂ ), nitrous oxide (N ₂ O), and nitrogen (N ₂₎ . ), gas inflow amount, pressure and/or temperature inside the chamber, etc.

Referring to FIG. 3, according to each of the n previously set deposition variables (x-axis value), the range (y-axis value) in which a normal range of hydrogen is detected in the chemical vapor deposition device is calculated. In other words, when all shaded box areas in the data of n deposition variables are satisfied, the amount of hydrogen in the normal range can be identified. In this embodiment, big data analysis and machine learning methods can be used to calculate the amount of hydrogen in the normal range according to the deposition variable data. Machine learning is an algorithmic technology that classifies/learns the characteristics of input data on its own, and uses various machine learning algorithms such as decision trees, neural networks, and support vector machines (SVMs).

Afterwards, during the deposition process including the dehydrogenation process, the deposition variable is measured in real time (S30), and the amount of hydrogen according to the current deposition variable is predicted in real time through a pre-generated hydrogen amount prediction model (S40).

Afterwards, the predicted amount of hydrogen is compared with the amount of hydrogen before or after the dehydrogenation revolution (S40). In general, when the amount of hydrogen in the amorphous silicon layer 20 before the dehydrogenation process is 7 to 8 atom% or less, and when the amount of hydrogen in the amorphous silicon layer 20 after the dehydrogenation process is 1.5 atom% or less, the thin film transistor has device characteristics in the normal range. will have

If it is determined (OK) that the predicted amount of hydrogen in the amorphous silicon layer 20 is within the normal range, the crystallization process from amorphous silicon to polycrystalline silicon proceeds (S50). Referring again to FIG. 1, step (c), that is, crystallizing amorphous silicon into polysilicon using excimer laser annealing (ELA), is performed.

In addition, if the predicted amount of hydrogen in the amorphous silicon layer 20 is determined to be outside the normal range (NG), deposition equipment PM (Preventive Maintenance) is performed. The operator controls the temperature of the susceptor in the deposition device, the gas inflow rates of argon (Ar), silane (SiH ₄ ), ammonia (NH ₃ ), hydrogen (H ₂ ), nitrous oxide (N ₂ O), and nitrogen (N ₂ ); Adjust the pressure and/or temperature inside the chamber or remove foreign substances from the deposition equipment. After performing facility PM, a new amorphous silicon layer manufacturing process can be started.

Although the steps after polycrystalline silicon crystallization are not shown in FIG. 2, the amorphous silicon layer corresponding to the normal range can be manufactured into a thin film transistor by sequentially proceeding steps (e) to (l) as shown in FIG. 1.

Therefore, in this embodiment, by predicting the amount of hydrogen in the amorphous silicon layer during the deposition process through monitoring the thin film process (deposition process), defects can be identified in advance without analyzing the destruction of the amorphous silicon layer after the deposition process. .

FIG. 4 exemplarily shows how a defective state of an amorphous crystal layer can be confirmed according to deposition variables in a thin film transistor device characteristic monitoring method according to an embodiment.

In the best practice shown at the top of Figure 4, the deposition variables are Vdc and the concentration of ammonia (NH ₃ ). Here, Vdc refers to the difference between the plasma potential in the deposition chamber and the potential of the substrate. The deposition variable in FIG. 4 is an example and may vary depending on the example.

In Best Practice, if Vdc is -35.64 V to -31.51 V and the concentration of ammonia (NH ₃ ) is 0.48% or less, the amount of hydrogen for each range is calculated, and if two conditions are met, it is normal (OK). It is marked as applicable. Accordingly, the operator can determine that when the thin film transistor is manufactured at Vdc and ammonia (NH ₃ ) concentration in the normal range (OK) of the hydrogen amount, it can be manufactured without defects.

In the Risk Practice shown at the bottom of FIG. 4, the deposition variables are Vdc, concentration of ammonia (NH ₃ ), temperature inside the chamber, and concentration of silane (SH ₄ ).

In Risk Practice, Vdc is -32.94 V to -27.06 V, ammonia (NH ₃ ) concentration is 0.1 % to 0.67 %, chamber internal temperature is 397.93 ℃ to 398.4 ℃, and silane (SIH ₄ ) concentration is -0.1 % to -0.05%, the amount of hydrogen for each range is calculated, and if the four conditions are met, it is indicated as being defective (NG). Accordingly, the operator can determine the range in which defective manufacturing can occur by identifying the deposition parameters of the portion corresponding to the normal range of hydrogen amount.

In each case, if all deposition variables are satisfied and the hydrogen concentration of the amorphous silicon layer is in the normal range (OK), the crystallization process proceeds to transform the amorphous silicon into polycrystalline silicon, and if it is in the defective range (NG), the facility is PM is in progress.

That is, the operator can monitor the current state of the amorphous silicon layer shown in FIG. 4 and detect device defects in advance before completing the device process of the thin film transistor.

Below, we will look at a deposition process monitoring device that can predict the amount of hydrogen in the amorphous silicon layer during the deposition process.

Figure 5 is a block diagram of a deposition process monitoring device according to one embodiment.

Referring to FIG. 5, the deposition process monitoring device according to one embodiment includes an input unit 510, a prediction model generation unit 520, a deposition variable collection unit 540, a prediction unit 530, a defect determination unit 550, and Includes a display unit 560.

The input unit 510 receives previously set deposition process equipment variables of the amorphous silicon layer 20 and data on the amount of hydrogen in the amorphous silicon layer 20 before and after the dehydrogenation process measured under the equipment conditions. . Here, the deposition variables are the temperature of the susceptor of the deposition device, argon (Ar), silane (SiH ₄ ), ammonia (NH ₃ ), hydrogen (H ₂ ), nitrous oxide (N ₂ O), and nitrogen (N ₂₎ . ), gas inflow amount, pressure and/or temperature inside the chamber, etc.

The prediction model generator 520 learns a prediction model for the amount of hydrogen in the amorphous silicon layer according to deposition variables using a plurality of data through machine learning and generates a prediction model for the amorphous silicon layer. Machine learning is an algorithmic technology that classifies and learns the characteristics of input data on its own, and various algorithms such as decision trees, neural networks, and support vector machines (SVMs) can be used. The prediction model generator 520 receives the plurality of data from the input unit 510.

The deposition variable collection unit 540 collects deposition variables of the deposition equipment in real time. In detail, in step (b) of FIG. 1 described above, deposition variables are collected as the deposition process progresses.

The prediction unit 530 predicts the amount of hydrogen in the amorphous silicon layer through a prediction model using the deposition variables provided by the deposition variable collection unit 540.

The defect determination unit 550 uses the amount of hydrogen in the amorphous silicon layer provided by the prediction unit 530 to determine whether the amorphous silicon layer currently being manufactured is defective.

For example, if the amount of hydrogen in the amorphous silicon (a-Si) layer before the dehydrogenation process is > 8 atm %, the current amorphous silicon layer is judged to be defective, and the amount of hydrogen in the amorphous silicon (a-Si) layer after the dehydrogenation process is determined to be defective. If >3.0 atm %, it is determined that the current amorphous silicon layer is defective.

Here, the prediction model generator 520, the deposition variable collection unit 540, the prediction unit 530, and the defect determination unit 550 correspond to the control unit 500 of the monitoring device, and according to the embodiment, the control unit 500 may include other parts.

The display unit 560 displays whether the amorphous silicon layer is defective as determined by the defect determination unit 550. The display unit 560 may be implemented as a display.

The algorithmic elements described herein are actually implemented (in whole or in part) as software or hardware (e.g., as part of an application-specific IC) running on a specific digital signal processor or general-purpose processor, etc.

Although not shown herein, a chemical vapor deposition device generally includes a chamber maintained in a vacuum state, a susceptor supporting a substrate, and argon (Ar), silane (SiH ₄ ), ammonia (NH ₃ ), and hydrogen ( It may include a gas supply pipe through which at least one of H ₂ ), nitrous oxide (N ₂ O), and nitrogen (N ₂ ) flows.

In this embodiment, the method of checking whether the amount of hydrogen during the deposition process is within the normal range can be directly predicted based on the deposition variables, or indirectly predicted through the relationship between the deposition variables (Vdc, temperature and pressure inside the chamber). It may be possible.

Below, with reference to FIGS. 6 and 7 , the relationship between deposition variables that affect the amount of hydrogen in the amorphous silicon layer will be examined.

FIG. 6 is a contour diagram illustrating the relationship between deposition variables during a thin film process according to an embodiment, and FIG. 7 is a method for monitoring device characteristics of a thin film transistor according to an embodiment. Device defect status according to equipment variables can be confirmed. It is shown as an example of what it looks like.

Referring to FIG. 6, a contour diagram of Vdc (difference between the plasma potential in the deposition chamber and the potential of the substrate) according to the temperature and pressure inside the chamber among the deposition variables is shown.

Specifically, the temperature inside the chamber is in the range of about 399 ℃ to 400 ℃, and the pressure inside the chamber is in the range of about 800 atm to 1000 atm.

Looking in detail, it can be seen that as the temperature inside the chamber increases and the internal pressure increases, the Vdc value corresponds to -37.5 V, the temperature inside the chamber is about 399.8 ℃ to 400.2 ℃, and the pressure is about 920 atm to 950 atm. In the case of atm, it can be seen that the Vdc value corresponds to -35 V to -32.5 V.

In other words, the prediction model generator can determine the temperature and pressure inside the chamber in the deposition device, derive Vdc, and predict the amount of hydrogen derived according to Vdc.

Referring to FIG. 7, the Vdc (performance) derived according to deposition variables such as the temperature inside the chamber, the temperature of the susceptor, the pressure inside the chamber, and the concentration of argon (Ar) and silane (SiH ₄ ) is -33.385 V. It is marked as applicable.

A scatter plot according to each deposition variable is shown, and the prediction model generator can derive a hydrogen amount that satisfies the rule shown in FIG. 4 according to each scatter plot.

Therefore, it can be confirmed that the prediction model generator can derive Vdc according to each deposition variable in the deposition equipment and predict the amount of hydrogen derived according to Vdc.

As described above, the prediction model generator generates a hydrogen amount prediction model using the deposition variables shown in FIGS. 6 and 7, and the prediction unit can predict the hydrogen amount using this hydrogen amount prediction model.

Figures 6 and 7 illustrate an embodiment, and depending on the embodiment, the method of generating a hydrogen amount prediction model using deposition variables and predicting the hydrogen amount may be modified in various ways.

Although the embodiments of the present invention have been described in detail above, the scope of the present invention is not limited thereto, and various modifications and improvements made by those skilled in the art using the basic concept of the present invention defined in the following claims are also possible. It falls within the scope of rights.

100: substrate 110: buffer layer
121: gate electrode 130: semiconductor layer
20: Amorphous silicon layer 30: Polycrystalline silicon layer
140: gate insulating film 150: interlayer insulating film
152: source electrode 153: drain electrode
510: input unit 520: prediction model generation unit
530: prediction unit 540: heat treatment variable collection unit
550: defect judgment unit 560: display unit

Claims (20)

A prediction model generator that generates a prediction model for the amount of hydrogen in the amorphous silicon layer through machine learning using the deposition variables of the preset deposition equipment and the previously measured amount of hydrogen in the amorphous silicon (a-Si) layer;
A prediction unit that predicts the amount of hydrogen in the amorphous silicon layer according to the deposition variables of the deposition equipment measured in real time, based on the hydrogen amount prediction model; and
A deposition process monitoring device including a defect determination unit that determines whether the predicted amount of hydrogen is within a normal range. According to paragraph 1,
an input unit that receives the deposition parameters of the preset deposition equipment and the previously measured amount of hydrogen in the amorphous silicon layer; and
A deposition process monitoring device further comprising a deposition variable collection unit that collects deposition variables of the deposition facility in real time. According to paragraph 2,
A deposition process monitoring device wherein the input unit provides deposition variables of the preset deposition equipment and the previously measured amount of hydrogen in the amorphous silicon layer to the prediction model generator. According to paragraph 2,
A deposition process monitoring device wherein the deposition variable collection unit provides the collected deposition variables of the deposition equipment to the prediction unit. According to paragraph 1,
The defect determination unit determines that the amorphous silicon (a-Si) layer is in the normal range when the amount of hydrogen in the amorphous silicon (a-Si) layer is 8 atm% or less before the dehydrogenation process. According to paragraph 1,
The defect determination unit determines that the amorphous silicon (a-Si) layer is within the normal range when the amount of hydrogen in the amorphous silicon (a-Si) layer is 3.0 atm% or less after the dehydrogenation process. According to paragraph 1,
A deposition process monitoring device further comprising a display unit that displays whether the amorphous silicon layer is defective or normal according to the defect determination unit. According to paragraph 1,
The deposition variables of the deposition equipment include the temperature of the equipment susceptor, argon (Ar), silane (SiH ₄ ), ammonia (NH ₃ ), hydrogen (H ₂ ), nitrous oxide (N ₂ O), and nitrogen (N ₂ ) A deposition process monitoring device corresponding to at least one of the gas inflow amount, and the pressure and temperature inside the chamber. According to paragraph 1,
A deposition process monitoring device in which the machine learning is based on at least one of a decision tree, a neural network, and a support vector machine (SVM). Generating a hydrogen amount prediction model of the amorphous silicon layer through machine learning using the deposition variables of the preset deposition equipment and the previously measured hydrogen amount of the amorphous silicon (a-Si) layer;
Based on the hydrogen amount prediction model, predicting the amount of hydrogen in the amorphous silicon layer according to the deposition variables of the deposition equipment measured in real time; and
A deposition process monitoring method comprising determining whether the predicted amount of hydrogen falls within a normal range. According to clause 10,
In the step of predicting the amount of hydrogen in the amorphous silicon layer according to the deposition variables of the deposition equipment measured in real time,
A deposition process monitoring method in which the deposition parameters of the deposition equipment are measured in real time during the step of forming the amorphous silicon layer on the substrate and the dehydrogenation process step of the amorphous silicon layer. According to clause 11,
In the step of determining whether the predicted amount of hydrogen falls within the normal range,
The normal range before the dehydrogenation process is a deposition process monitoring method in which the amount of hydrogen in the amorphous silicon (a-Si) layer is 8 atm % or less. According to clause 11,
In the step of determining whether the predicted amount of hydrogen falls within the normal range,
The normal range after the dehydrogenation process is a deposition process monitoring method in which the amount of hydrogen in the amorphous silicon (a-Si) layer is 3.0 atm % or less. According to clause 10,
In the step of determining whether the predicted amount of hydrogen falls within the normal range,
If it falls within the normal range, a deposition process monitoring method in which a step of crystallizing the amorphous silicon into polycrystalline silicon is performed. According to clause 10,
In the step of determining whether the predicted amount of hydrogen falls within the normal range,
A deposition process monitoring method in which PM (Preventive Maintenance) of the deposition equipment is performed when the normal range is exceeded. According to clause 10,
The deposition variables of the deposition equipment include the temperature of the equipment susceptor, argon (Ar), silane (SiH ₄ ), ammonia (NH ₃ ), hydrogen (H ₂ ), nitrous oxide (N ₂ O), and nitrogen (N ₂ ) A method for monitoring the deposition process corresponding to at least one of the gas inflow amount, and the pressure and temperature inside the chamber. According to clause 10,
A deposition process monitoring method in which the machine learning is based on at least one of a decision tree, a neural network, and a support vector machine. Forming an amorphous silicon layer on a substrate through a deposition process;
Crystallizing the amorphous silicon layer into a polycrystalline silicon layer; and
Patterning the polycrystalline silicon layer to form a semiconductor layer,
A thin film transistor manufacturing method for predicting the amount of hydrogen in the amorphous silicon layer according to the deposition parameters of the deposition process. According to clause 18,
After forming the amorphous silicon layer,
A thin film transistor manufacturing method further comprising performing a dehydrogenation process to lower the hydrogen concentration of the amorphous silicon layer. According to clause 18,
forming the semiconductor layer and forming a gate insulating film on the semiconductor layer;
forming a gate electrode on the gate insulating film;
forming an interlayer insulating film on the gate electrode and forming an opening to expose the source region and drain region of the semiconductor layer; and
A thin film transistor manufacturing method further comprising forming a source electrode and a drain electrode to be connected to the source region and the drain region, respectively.

Images