JP6653217B2 - Quality evaluation method of laminate having protective film on surface of oxide semiconductor thin film, and quality control method of oxide semiconductor thin film - Google Patents

Description

  The present invention relates to an oxide for a semiconductor layer (hereinafter, referred to as an “oxide semiconductor thin film”) suitable for a thin film transistor (TFT) used for a display device such as a liquid crystal display or an organic EL display. The present invention relates to a method for evaluating the quality of a laminate having a protective film on the surface of the above-mentioned item) and a method for controlling the quality of the oxide semiconductor thin film. In detail, defects due to defects in the oxide semiconductor thin film or defects due to interface defects with a protective film formed on the surface of the oxide semiconductor thin film are measured by a contact method or a non-contact method. The present invention relates to a technique for evaluating the quality of a stacked body and a quality control method for an oxide semiconductor thin film having a protective film.

  An amorphous oxide semiconductor thin film has high carrier mobility, a large optical band gap, and can be formed at a low temperature. Therefore, the amorphous oxide semiconductor thin film is expected to be applied to next-generation displays that require large size, high resolution, and high-speed driving, and resin substrates with low heat resistance used for transparent displays and flexible displays.

  Among the oxide semiconductor thin films, an amorphous oxide semiconductor thin film including indium (In), gallium (Ga), zinc (Zn), and oxygen (O) (hereinafter referred to as “InGaZnO” or “IGZO”) is particularly referred to. Is very preferably used because it has a very high carrier mobility.

  However, it is known that an oxide semiconductor thin film changes its electronic state due to a film formation process and a heat treatment performed thereafter, which affects the quality of a TFT. For example, due to lattice defects generated in the film formation process and hydrogen in the film, the carrier concentration that governs the TFT characteristics greatly changes, and the TFT characteristics tend to vary. Therefore, in the manufacturing process of a display device or the like, it is important from the viewpoint of productivity improvement to evaluate the characteristics of the oxide semiconductor thin film and feed back the results to adjust the manufacturing conditions and control the TFT quality. Becomes

  As a method of evaluating the mobility of an oxide semiconductor thin film by a non-contact method without attaching an electrode, the present inventors have disclosed in Patent Document 1 a microwave photoconductive decay method (hereinafter referred to as a “μ-PCD method”). In some cases) discloses a method for qualitatively or quantitatively evaluating the mobility of an oxide semiconductor thin film.

  Meanwhile, in a TFT using an oxide semiconductor thin film, it is important to suppress not only basic mobility as a transistor characteristic but also product defects such as luminance unevenness, color skipping, and display defects. For that purpose, it is also required that they have excellent resistance to stress such as light irradiation or voltage application that causes these product defects (hereinafter, sometimes referred to as “stress resistance”). The stress tolerance means that a semiconductor element such as a transistor has good characteristics even when a stress such as continuous irradiation of light or application of a gate voltage is applied to the semiconductor element.

One of stress resistances is that a threshold voltage (V _th : Threshold Voltage) does not shift in drain current-gate voltage characteristics (hereinafter sometimes referred to as “IV characteristics”), that is, before and after stress application. the amount of change of V _th (hereinafter sometimes referred to as "[Delta] V _th"), and it is small. For example, in an organic EL display, a positive voltage (hereinafter, sometimes referred to as “positive bias”) is continuously applied to a gate electrode of a driving TFT while the organic EL element emits light. Therefore, there is a problem that V _th changes and switching characteristics change. In a liquid crystal television, light from a backlight is applied to the TFT. Therefore, when light irradiation is continued in addition to a negative voltage (hereinafter, sometimes referred to as a “negative bias”), _Vth changes, which causes a problem that switching characteristics change.

Further, it is necessary that V _th of the TFT is kept within an appropriate range as a stress resistance related to a positive bias. If the value of _Vth is a negative value, a current flows when the gate voltage is not applied, so that power consumption increases. On the other hand, if the positive value of V _th is too large, a large voltage is required for the operation of the TFT.

  Such a change in the switching characteristics causes a decrease in the reliability of the display device itself, and thus an improvement in stress resistance has been desired.

  In addition to the above, process conditions that cause a large difference in TFT characteristics have been reported. For example, in Non-Patent Document 1, the electronic state in the oxide semiconductor film after the oxide semiconductor thin film is annealed changes depending on the type of the gate insulating film used for the TFT, and as a result, the characteristics of the TFT are reduced. It is disclosed that it has a significant effect.

  Further, Patent Document 2 discloses a technique using a μ-PCD method as a quality evaluation method of an oxide semiconductor thin film.

JP 2012-33857 A WO 2016/009686 pamphlet

Journal of Applied Physics Vol. 112, 114515 (2012)

  In the development of TFT materials, not only the composition and content of the oxide semiconductor are changed according to the required characteristics, but also the film forming conditions such as the film forming temperature and atmosphere, and the processing conditions for the subsequent steps. (Hereinafter, these may be collectively referred to as “manufacturing conditions”). When examining the optimal combination, it is necessary to evaluate the stress resistance of the oxide semiconductor thin film, but conventionally, the stress resistance was evaluated after actually manufacturing a TFT. Therefore, it takes much time and cost to find out the manufacturing conditions of the oxide semiconductor thin film that affect the stress resistance. Therefore, there is a need for a technique for accurately and easily evaluating stress tolerance.

  As a quality control method of the oxide semiconductor thin film, the present inventors have measured and evaluated the electronic state of the oxide semiconductor thin film between any two steps of the TFT manufacturing process, thereby grasping the defect of the oxide semiconductor thin film. Patent Application 2 has been filed as a technique for managing the quality of an oxide semiconductor thin film. The present inventors have evaluated stress resistance more easily, and have further studied on a technology capable of stabilizing the quality of TFT characteristics.

  The present invention has been made in view of the above situation. An object of the present invention is to accurately and easily evaluate the defect of the oxide semiconductor thin film and the quality of the oxide semiconductor thin film in a laminate having a protective film on the surface of the oxide semiconductor thin film, and also to improve the quality of the TFT. The purpose is to provide a method that can be properly managed.

  The present invention which has solved the above problems is a method for evaluating the quality of a stacked body having a protective film on the surface of an oxide semiconductor thin film, wherein the electronic state of the oxide semiconductor thin film having the protective film is measured. Accordingly, the gist is to evaluate a defect caused by a defect of the oxide semiconductor thin film.

By evaluating a defect caused by the defect of the oxide semiconductor thin film, (1) when a light irradiation and a negative bias are applied to the TFT, a difference ΔV _th between threshold voltages before and after the application and (2) switching of the TFT It is also preferable to evaluate any of (3) the threshold voltage _Vth of the TFT.

  In the present invention, the electronic state of the oxide semiconductor thin film irradiates the oxide semiconductor thin film with excitation light and microwaves, and the reflected wave from the irradiated portion of the microwaves changed by the irradiation of the excitation light. After measuring the maximum value, stop the irradiation of the excitation light, reflectivity measurement to measure the temporal change of the reflectance of the reflected wave from the irradiated portion of the microwave after stopping the irradiation of the excitation light In a preferred embodiment, the measurement is performed by a non-contact method based on a step and a parameter calculation step of calculating a parameter corresponding to a slow attenuation observed after the irradiation of the excitation light is stopped from the temporal change of the reflectance. is there.

  The parameter calculating step may calculate a parameter corresponding to a slow decay seen in 0.1 to 10 μs after stopping irradiation of the excitation light from the change in the reflectance, and evaluate an electronic state of the oxide semiconductor thin film. preferable.

  The electronic state of the oxide semiconductor thin film is preferably evaluated as good when the parameter corresponding to the slow decay in the parameter calculation step is 1.5 or less.

  The present invention also relates to a quality control method for an oxide semiconductor thin film using the above-described quality evaluation method, wherein an electronic state of the oxide semiconductor thin film having the protective film is evaluated based on the quality evaluation method. It is also a preferred embodiment to adjust at least one of the conditions for forming the oxide semiconductor thin film, the conditions for heat treatment after the formation of the oxide semiconductor thin film, and the conditions for forming the protective film based on the above.

  According to the present invention, by measuring the electronic state of an oxide semiconductor thin film having a protective film, defects due to defects in the oxide semiconductor thin film, and the quality of the oxide semiconductor thin film, and the quality of the TFT can be accurately determined. It can be evaluated easily. Further, if the manufacturing conditions are controlled so that the electronic state of the oxide semiconductor thin film of the laminated body after the formation of the protective film is good based on the evaluation of the present invention, the quality of the oxide semiconductor thin film and the quality of the TFT can be appropriately adjusted. Can manage.

FIG. 1 is a graph showing measurement results by the μ-PCD method in Example 1 (1). FIG. 2 is a graph showing a relationship between a shift amount of a threshold voltage and ΔV _th with respect to light irradiation and application of a negative bias in Example 1 (2). FIG. 3 is a diagram illustrating an example of a microwave attenuation waveform. FIG. 4 is a schematic diagram showing the structure of the oxide semiconductor TFT used in the example. FIG. 5 is a schematic diagram showing an example of another configuration of the evaluation element used in the present invention. FIG. 6 is a schematic diagram illustrating an example of another configuration of the evaluation element used in the present invention. FIG. 7 is a schematic view showing an example of another configuration of the evaluation element used in the present invention. FIG. 8 is a schematic diagram of an evaluation substrate showing an example of an array configuration of evaluation elements used in the present invention. FIG. 9 is a schematic explanatory view showing an example of the evaluation device used in the present invention. FIG. 10 is a schematic explanatory view showing an example of an evaluation device having a measuring unit of the contact type method and a measuring unit of the non-contact type method. FIG. 11 is a graph showing the relationship between the oxygen concentration at the time of forming an oxide semiconductor thin film and the measurement result of the μ-PCD method. FIG. 12A is a graph showing a pre-annealing time after forming an oxide semiconductor thin film and a measurement result by a μ-PCD method. FIG. 12B is a graph showing the relationship between the pre-annealing time and ΔV _th or lifetime. FIG. 13A shows lifetime values, defects, defects due to defects in the oxide semiconductor thin film, and defects caused by the interface defects, the relationship between the electrical resistivity, the defect in the film, the relationship between the interface defects, and the hydrogen in the oxide semiconductor thin film. FIG. 3 is a schematic explanatory diagram showing the relationship between donors. FIG. 13B is a schematic explanatory view showing the relationship between mobility and stress resistance based on evaluation of an oxide semiconductor thin film. FIG. 14 is a graph showing the relationship between the parameter P measured in Example 1 (1) and the shift amount of the threshold voltage measured in (2).

  The present inventors' research results show that the following relationship exists between the electronic state of the oxide semiconductor thin film and the characteristics of the TFT.

  First, it is known that the cause of color skipping and display failure of a display device is caused by a shift of a threshold voltage due to a negative bias and light irradiation (hereinafter, sometimes referred to as “negative bias stress”). Further, it is known that uneven brightness of a display device using a TFT is caused by a shift in threshold voltage when a positive bias is applied (hereinafter, sometimes referred to as “positive bias stress”). .

  (A) The change in the threshold voltage due to the negative bias stress is mainly caused by the formation of the oxide semiconductor thin film and the subsequent annealing and the like in the manufacturing process of the TFT. This is caused by a change in the electronic state due to a defect (hereinafter, sometimes referred to as a “defect in the film”). If there is a defect in the film, color jump and display failure occur due to negative bias stress. In addition, (b) a change in threshold voltage due to a positive bias is mainly caused by a defect caused by an interface state existing at an interface between the gate insulating film and the oxide semiconductor thin film after formation of the protective film and the protective film (hereinafter, referred to as “ Interface defects). If an interface defect exists, luminance unevenness occurs due to the positive bias stress.

Further, when the above-described defect exists in the film, the defect has a property of emitting electrons serving as carriers in the oxide semiconductor film; thus, the electronic state changes. Further, when an interface defect exists, band bending occurs in the oxide semiconductor thin film. As a result, it has been found that the electronic state of the oxide semiconductor thin film changes. Then, it was found that the change in the electronic state can be measured also by the electric resistivity of the oxide semiconductor thin film. That is, the increase in ΔV _th that occurs after light irradiation and application of a negative bias or after application of a positive bias (hereinafter, these may be collectively referred to as “bias stress”) is caused by an increase in the electric resistance of the oxide semiconductor before the bias stress. Previous studies have shown a good relationship with the rate.

Therefore, in evaluating the magnitude of ΔV _th generated after the application of the bias stress, it is not necessary to measure the ΔV _th by actually manufacturing a TFT and performing a stress test in which a bias stress is applied. That is, it was found that the evaluation can be easily performed only by measuring the electronic state of the oxide semiconductor, particularly, the electrical resistivity. It was also found that ΔV _th associated with the application of a positive bias was closely correlated with the absolute value of the threshold voltage. Therefore, by measuring the electronic state of the oxide semiconductor thin film, in particular, the electrical resistivity, it is possible to evaluate (1) the difference ΔV _{th in} threshold voltage, which is an index of stress resistance due to negative bias stress. In addition, it is possible to evaluate (2) whether or not the TFT can be switched, that is, whether or not there is non-conductivity, and (3) the threshold voltage _{Vth of the} TFT, which is an index of stress tolerance due to the positive bias stress.

  Based on the above findings, the present inventors performed a first step of measuring and evaluating defects in the film and a second step of measuring and evaluating defects in the interface, so that defects in the film of the oxide semiconductor thin film were Patent Document 2 discloses a technique for grasping defects individually and controlling the quality of an oxide semiconductor thin film. As described above, the stress resistance of the oxide semiconductor thin film is affected by defects in the oxide semiconductor thin film and interface defects (hereinafter, sometimes simply referred to as “defects”). It is necessary to measure and evaluate the electronic state of the oxide semiconductor thin film at least twice in order to grasp the condition.

  As a result of further studies by the present inventors, it has been found that the step of forming a protective film on the oxide semiconductor thin film has a large effect on the quality of the oxide semiconductor thin film, and also has an effect on the characteristics of the TFT. After forming a protective film on the oxide semiconductor thin film, it was found that the characteristics of the TFT can be estimated by measuring and evaluating the electronic state of the oxide semiconductor thin film. That is, in the oxide semiconductor thin film after the formation of the protective film, in-film defects and interfacial defects are mixed. In the present invention, the oxide state is measured by measuring and evaluating the electronic state of the oxide semiconductor thin film after the formation of the protective film. It is possible to optimize the manufacturing conditions of the oxide semiconductor thin film and manage the quality of the oxide semiconductor thin film and the TFT. Therefore, according to the present invention, the defect state of the oxide semiconductor thin film can be grasped in a smaller number of times of measurement than in Patent Document 2 in which it is necessary to measure at least twice. In the present invention, “measurement” refers to indirectly or directly measuring the electronic state of an oxide semiconductor thin film by a noncontact method or a contact method. The term “evaluation” refers to a defect resulting from a defect in a film or an interface defect based on a measurement result, or the quality of the defect resulting from the defect.

  Hereinafter, the present invention will be described.

[Oxide semiconductor thin film]
In the present invention, an electronic state of an oxide semiconductor thin film having a protective film (hereinafter, sometimes referred to as a “stack”) is measured, and a defect due to a defect in the oxide semiconductor thin film is evaluated. Then, based on the evaluation, the relationship between the defect and the manufacturing condition is grasped, and the manufacturing condition for reducing the defect is predicted and determined. Then, based on the determined optimum manufacturing conditions, an oxide semiconductor thin film with reduced defects can be obtained.

  The stacked body is obtained by forming an oxide semiconductor thin film on at least a substrate and a protective film formed on a surface of the oxide semiconductor thin film. The oxide semiconductor thin film may be directly formed over the substrate, or an oxide semiconductor thin film may be formed after an arbitrary insulating film such as a gate insulating film is formed over the substrate.

  As the oxide semiconductor thin film, an amorphous oxide semiconductor thin film containing at least one element selected from the group consisting of In, Ga, Zn, and Sn is preferably used. These elements may be contained alone or in combination of two or more. Specifically, for example, an In oxide, an In-Sn oxide, an In-Zn oxide, an In-Sn-Zn oxide, an In-Ga oxide, a Zn-Ga oxide, an In-Ga-Zn oxide, Zn oxide and the like can be mentioned.

  The thickness of the oxide semiconductor thin film is not particularly limited, but is preferably 10 nm or more, more preferably 30 nm or more, preferably 500 nm or less, more preferably 200 nm or less, and still more preferably 100 nm or less.

[Protective film]
The protective film only needs to be formed on the surface of the oxide semiconductor thin film. However, if necessary, the oxide semiconductor thin film is subjected to a heat treatment (hereinafter, sometimes referred to as “pre-annealing”), and then the protective film is formed. You may. Note that the protective film is a protective film (hereinafter, sometimes referred to as an “etch stop layer” or “ESL”) for directly protecting the surface of the oxide semiconductor thin film, and is an insulating film such as SiO _2. Is preferred.

  In the present invention, the electronic state of the oxide semiconductor thin film of the stacked body is measured as described above, but the electronic state of the oxide semiconductor thin film may be measured by any of a contact method and a non-contact method. The electronic state can be measured based on the electric resistivity of the oxide semiconductor thin film. The electrical resistivity may be measured directly or indirectly. In addition, examples of the electric resistivity include sheet resistance (Ω · cm / □) and specific resistance (Ω · cm). The specific resistance is obtained by multiplying the sheet resistance by the film thickness.

[Contact method]
First, a method for measuring an electronic state of an oxide semiconductor thin film by a contact method will be described.

The contact method is a method in which a resistance measurement terminal is brought into contact with an oxide semiconductor thin film to perform measurement.
The electronic state of the oxide semiconductor thin film can be directly measured as electric resistivity by a contact method. For example, an electric resistance may be measured by forming an electrode on an oxide semiconductor film as in a four-terminal method or a four-probe method, or a measuring terminal having an electrode may be formed as in a double ring electrode method. It may be used to measure the electrical resistivity. Note that since the surface of the oxide semiconductor thin film is covered with the protective film, the protective film is removed only where the measurement terminal such as a needle or a measurement probe is in contact with the oxide semiconductor surface, and the surface of the oxide semiconductor thin film is measured. The terminals may be brought into a contactable state. Various known measurement conditions can be adopted as the measurement conditions of the electrical resistivity by the contact method.

[Non-contact method]
Next, a method for measuring an electronic state of an oxide semiconductor thin film by a noncontact method will be described.

  The non-contact method is a method of measuring an electronic state without bringing a resistance measuring terminal into contact with an oxide semiconductor thin film. The non-contact method allows non-destructive and non-contact and indirect measurement of electrical resistivity. As a non-contact method, for example, a μ-PCD method is exemplified. In the case of the μ-PCD method, the electrical resistivity cannot be directly measured. However, since the measured value of the μ-PCD method has a correlation with the electrical resistivity as described later, the electrical resistivity can be indirectly evaluated. In addition, the μ-PCD method is a simpler measurement method than the contact method because measurement can be performed without removing the protective film, and is preferable.

  The measurement method using the μ-PCD method refers to a method in which an oxide semiconductor thin film is irradiated with excitation light and microwaves, and the maximum value of a reflected wave from the oxide semiconductor thin film, which is changed by the irradiation of the excitation light, is changed. After the measurement, the irradiation of the excitation light is stopped, and the reflectance measurement step of measuring the temporal change in the reflectance of the reflected wave from the oxide semiconductor thin film of the microwave after stopping the irradiation of the excitation light And a parameter calculating step of calculating a parameter corresponding to slow attenuation observed after the stop of irradiation of the excitation light from the temporal change of the reflectance.

  The correlation between the measured value of the μ-PCD method and the electrical resistivity is as follows. The slow decay of the parameter calculation step, specifically, the irradiation of the excitation light is stopped, and the degree of microwave decay observed about 1 μs after the stop (hereinafter sometimes referred to as a “slow microwave decay waveform”) is oxidized. It is greatly affected by the defect order below the conduction band of the semiconductor thin film, that is, the electronic state. Therefore, when the signal in this region is analyzed, it has a close correlation with the stress resistance of the oxide semiconductor thin film and the like.

  Therefore, by employing the μ-PCD method, a defect caused by a defect in the oxide semiconductor thin film can be accurately and simply measured and evaluated in a non-contact manner based on a change in reflectance.

In the present specification, for example, the “parameter corresponding to slow decay observed about 1 μs after the stoppage of excitation light irradiation” (hereinafter, sometimes referred to as “B value”) is, for example, the case where the microwave reflection intensity is the maximum value. Time to 1 / e ² ; slope obtained by logarithmically converting the attenuation curve of reflected wave intensity from 1 / e of the maximum value to 1 / e ² of the maximum value, or the reciprocal of the slope The absolute value of the slope of the attenuation curve of the reflected wave intensity of about 1 μs to 2 μs after the stoppage of the excitation light irradiation, or the absolute value of the reciprocal of the slope; the reflected wave intensity of the microwave observed about 1 μs after the stoppage of the excitation light irradiation When the attenuation of the reflected wave of the microwave is represented by the sum of two exponential functions, the longer value of the obtained logarithmically transformed slope, or the absolute value of the reciprocal of the slope; For example, there is a slope represented by a power relational expression of Expression (1). Here, the “logarithmically transformed slope obtained when the attenuation of the reflected microwave wave is represented by the sum of two exponential functions” is, for example, τ ₁ described in the following equation (1) and the following equation (2 It refers to tau ₂ described).

In the above formulas (1) and (2), n ₁ is the carrier density immediately after irradiation of the fast-decay laser, n ₂ is the carrier density immediately after irradiation of the slow-decay laser, exp is an exponential function, and t is the laser. In the time immediately after irradiation, β is a relaxation factor.

Among the above parameters, a slope obtained by logarithmically converting the attenuation curve of the microwave reflected wave intensity in a certain range or an absolute value of a reciprocal of the slope is preferable. Particularly preferable parameters are a slope obtained by logarithmically converting the attenuation curve of the reflected wave intensity from 1 / e of the maximum value to 1 / e ² of the maximum value, or an absolute value of a reciprocal of the slope, and from about 1 μs to about 2 μs. Is the absolute value of the slope obtained by logarithmically converting the attenuation curve of the reflected wave intensity, or the reciprocal of the slope.

  Here, “about 1 μs” in the above parameter is not intended to be strictly limited to 1 μs, but the microwave attenuation after the reflectance attenuation after the stop of the excitation light irradiation is slow, that is, the slope of the reflectance attenuation becomes small. The meaning includes the range of the reflectance. Therefore, it is difficult to uniquely define the above-mentioned time, but for example, preferably 0.1 μs or more, more preferably 0.5 μs or more, still more preferably 1 μs or more, preferably 10 μs or less, more preferably Is 2 μs or less, more preferably 1.5 μs or less, and most preferably 1 μs or less. In the present invention, “about 1 μs” may be used to represent the above range.

  The "slow decay" will be described in more detail with reference to FIG. FIG. 3 is a diagram showing how the excess carrier density changes in the μ-PCD method. The vertical axis of FIG. 3 corresponds to the microwave reflectivity. When the oxide semiconductor thin film sample is irradiated with excitation light, the carrier is absorbed by the oxide semiconductor thin film to excite carriers, thereby generating excess carriers. At that time, as the excess carrier density increases, the disappearance speed also increases, but when the carrier injection speed and the disappearance speed become equal, the excess carrier density has a constant peak value. When the rate of generation and extinction of the excess carrier becomes equal, the carrier saturates and maintains a constant value. When the irradiation of the excitation light is stopped in this state, the excess carriers decrease due to the recombination and disappearance of the excess carriers, and eventually return to the values before the excitation light irradiation.

  As shown in FIG. 3, the reflectivity of the microwave reflected from the oxide semiconductor thin film temporarily shows a maximum value, but rapidly attenuates at the same time as the stop of the excitation light irradiation. After that, an attenuation having a certain slope is observed. However, this slope generally corresponds to the above-mentioned “parameter corresponding to slow attenuation observed in about 1 μs after stopping the irradiation of the excitation light”. , Or parameter P. The maximum value of the reflectance of the reflected wave may be referred to as a parameter Q.

  Specifically, examples of the slope include a slope between the time in the above range and the reflected wave intensity, and a slope of a value obtained by logarithmically converting the reflected wave intensity with respect to a value obtained by logarithmically converting the time in the above range. For example, the slope (-B) within the range of the measurement time x (μs) can be used as the B value in the following equation (3). Note that, as described above, the inclination also includes the inclination when the reflectance decay after the stop of the excitation light irradiation becomes slow.

  In the formula, A represents a constant.

  Hereinafter, the evaluation method by the non-contact method will be described in detail. The device used in the present invention needs to be capable of irradiating the oxide semiconductor thin film with excitation light and microwaves and detecting the intensity of the microwave reflected from the oxide semiconductor thin film, which changes due to the irradiation of the excitation light. It is. Examples of such an apparatus include an apparatus shown in FIGS. 9 and 10 described in detail below and a lifetime measuring apparatus shown in FIG. The description of the apparatus is described in detail in Japanese Patent Application Laid-Open No. H10-157, which may be referred to. However, the device used in the present invention is not limited to this.

  In the μ-PCD method, a sample in which an oxide semiconductor thin film and a protective film are formed over a substrate is irradiated with excitation light and microwaves from the side of the protective film.

  In the present invention, it is possible to determine the carrier concentration of the oxide semiconductor thin film by analyzing the change in excess carrier density, and to evaluate the quality based on the electronic state, and thus the electrical resistivity, that is, the sheet resistance or the specific resistance. It is thought to be based on the following reasons.

  Microwaves applied to the oxide semiconductor thin film are reflected by plasma oscillation caused by carriers present in the oxide semiconductor thin film. This reflectivity depends on the carrier density in the oxide semiconductor thin film. However, in the oxide semiconductor thin film in a steady state, there is no carrier number at a level at which microwave reflection can be practically observed. However, when the excitation light is irradiated, excess carriers are generated in the film, and the microwave reflectance increases due to the plasma vibration of the excess carriers. On the other hand, by stopping the excitation light irradiation, the microwave reflectivity also decreases as the number of excess carriers decreases.

In general, carriers in a silicon semiconductor or the like are generated due to a shallow donor level existing below a conduction band in an energy band. The energy level in this case is about several tens meV below the conduction band, and is almost activated near room temperature. On the other hand, it is known that carriers in the oxide semiconductor thin film in the steady state are also caused by a shallow donor level existing below the conduction band in the energy band. However, in an oxide semiconductor, the level is about 0.1 to 0.2 eV, which is relatively deep. Therefore, the excess carriers generated by the irradiation with the excitation light may not only be recombined with the excited holes and electrons but also be once captured by the donor level and re-emitted. The rate of this capture and re-emission depends on the amount of shallow donor levels present below the conduction band in the energy band. Therefore, by tracing the annihilation process observed after the stop of the excitation light, it is possible to analyze the influence of the donor level on the excess carriers generated by the excitation light irradiation. Note that the specific resistance of an oxide semiconductor thin film is represented by the product of charge, free electrons, and mobility, and the mobility of the oxide semiconductor thin film is such that the composition of a metal element included in the oxide semiconductor thin film is the same. If there is no change. For example, the mobility of IGZO is about 10 cm ² / VS. Therefore, a change in the reflectivity of microwaves observed in the μ-PCD method, that is, a change in the excess carrier density generally correlates with the carrier concentration and the electrical resistivity.

  Note that some amorphous semiconductor materials such as an oxide semiconductor have a continuous level between a conduction band and a donor level, such as amorphous silicon and IGZO. In such a case, the carrier annihilation process observed in the μ-PCD method can be understood as a superposition of individual carrier transition behavior between each level. As a result, the decay process will be observed over a somewhat longer time range than the transition between one level. The time dependency at that time has a power-of-time relationship with time.

  Therefore, after the above-described reflectance measurement step, the carrier density of the oxide semiconductor thin film is determined by calculating a parameter corresponding to the slow decay generally observed in a time range ranging from 0.1 to 10 μs. Can be. As a result, electric resistivity such as sheet resistance and specific resistance can be indirectly measured and evaluated.

  As described above, the method for measuring the defect due to the defect in the oxide semiconductor thin film by the non-contact method by the μ-PCD method has been described in detail.

  In the present invention, manufacturing conditions for reducing defects of the oxide semiconductor thin film can be predicted and set from the evaluation results of the oxide semiconductor thin film. According to the study results of the present inventors, (i) a film forming step of a gate insulating film, (ii) a film forming step of an oxide semiconductor thin film, (Iii) It has been found that there is a heat treatment (hereinafter sometimes referred to as “pre-annealing”) step after the formation of the oxide semiconductor thin film. Further, as a manufacturing process which affects stress resistance due to interface defects, (iv) a process of forming a protective film that can be formed on the surface of the oxide semiconductor thin film, and (v) a heat treatment after forming the protective film (hereinafter, referred to as "Post-annealing process"). For example, based on the above evaluation, the (ii) oxide semiconductor thin film with reduced defects in the film can be formed on the substrate by adjusting the film formation conditions of the oxide semiconductor thin film. In addition, the (iii) thermal hysteresis condition can be adjusted based on the evaluation to reduce defects in the film. Alternatively, an oxide semiconductor thin film with reduced interface defects can be formed by adjusting the manufacturing conditions of the (iv) protective film and the thermal history conditions after the formation of the (v) protective film. In addition, even if the measured oxide semiconductor thin film does not include one in which defects are sufficiently reduced, a plurality of evaluation results are plotted on a graph to find an optimum condition from the manufacturing conditions and the result of the defect. It is also possible to optimize the defect of the oxide semiconductor thin film by forming and evaluating the oxide semiconductor thin film and the protective film again based on the conditions.

  In Patent Document 2, based on the above findings, first, defects in the film of the oxide semiconductor thin film formed on the substrate are evaluated, and the manufacturing conditions (i) to (iii) are adjusted based on the evaluation to reduce the defects in the film. A reduced oxide semiconductor thin film is newly formed. Thereafter, a protective film was formed to evaluate interface defects, and based on the evaluation, the manufacturing conditions (iv) and (v) were adjusted to reduce interface defects.

  In the present invention, after a protective film is formed on the oxide semiconductor thin film, the electronic state of the oxide semiconductor thin film is measured and evaluated in a state where defects in the film and interfacial defects are mixed, and the manufacturing conditions are determined based on the evaluation result. To adjust. At that time, it was found that if the defects of the oxide semiconductor thin film after the formation of the protective film were not sufficiently reduced, it was difficult to sufficiently reduce the defects even if the manufacturing conditions in the subsequent steps were adjusted. Preferably, if the value of the parameter corresponding to the slow decay in the parameter calculation step is within the following predetermined range, it can be determined that the defects of the oxide semiconductor thin film after the formation of the protective film are sufficiently reduced.

  From the viewpoint of manufacturing a TFT having good stress resistance, if the value of the parameter P corresponding to the slow decay of the parameter calculation step obtained by applying the μ-PCD method to the oxide semiconductor thin film having the protective film is too large, the depth becomes deep. Since there are many defects in the region and recombination is dominant, it is preferably 1.5 or less, more preferably 1.2 or less, and still more preferably 1.0 or less. The lower limit is not limited, but is preferably 0.5 or more, more preferably 0.8 or more. The present inventors evaluated the stress tolerance of the oxide semiconductor thin film having the protective film manufactured under various manufacturing conditions in each of the above manufacturing steps and the stress resistance of the actually manufactured TFT, and examined the relevance. It has been found that the parameter P is desirably within the above range in order to obtain resistance. Therefore, it is desirable to appropriately adjust the manufacturing conditions, particularly the protective film formation conditions, so that the parameter P falls within the above range. Thus, an oxide semiconductor thin film having a favorable electronic state can be manufactured based on the evaluation result of the oxide semiconductor thin film. Specific means for reducing defects in the oxide semiconductor thin film will be described later.

  According to the present invention, the electronic state of an oxide semiconductor thin film can be measured directly or indirectly, and a defect resulting from a defect in an oxide semiconductor thin film can be evaluated. Therefore, according to the present invention, the quality of the laminate can be accurately and simply evaluated. In addition, by appropriately adjusting manufacturing conditions based on the above evaluation results, a stacked body suitable for a TFT can be manufactured while maintaining a favorable electronic state of the oxide semiconductor.

  With the use of the evaluation method of the present invention, electrical characteristics of an oxide semiconductor thin film can be evaluated in a short time in an in-line manner in a manufacturing process of a liquid crystal display device or the like. Further, since the electronic state can be measured in a non-contact manner according to the μ-PCD method, productivity can be improved, such as improvement in yield, by applying the evaluation method of the present invention using the μ-PCD method. In addition, the quality control of the laminate can be performed more appropriately.

[Method of predicting manufacturing conditions]
Means for reducing defects in the oxide semiconductor thin film based on the evaluation of the oxide semiconductor thin film will be described.

  In the present invention, the defects of the oxide semiconductor thin film are reduced by adjusting the manufacturing conditions (i) to (v) as described above based on the evaluation result of the oxide semiconductor thin film having the protective film as described above. It is possible to optimize the electric resistivity caused by the defect.

The present inventors have examined the relationship between the oxygen concentration at the time of forming an oxide semiconductor thin film and defects in the film by a μ-PCD method in Patent Document 2, and found that the lifetime value differs depending on the oxygen concentration as shown in FIG. Is revealed. Hereinafter, the lifetime is described using τ ₂ as an index. When oxygen is not contained, the lifetime τ ₂ becomes longer because the resistance of the film is low. While the oxygen concentration is low, the lifetime value tends to be short, that is, the slope of τ ₂ tends to be steep. As the slope of τ ₂ increases, the number of defects in the film decreases. As the oxygen concentration increases, the lifetime value becomes extremely short. In this case, τ ₂ does not occur. Therefore, the optimum oxygen concentration can be determined by forming an oxide semiconductor thin film at different oxygen concentrations and evaluating the slope of τ ₂ , which makes it possible to form an oxide semiconductor thin film with reduced defects in the film. Become.

Further, the relationship between the thermal history after the formation of the oxide semiconductor thin film and defects in the film was examined by a μ-PCD method. As shown in FIG. 12A, after the oxide semiconductor thin film was formed, pre-annealing was performed in air at 350 ° C. and the heating time was changed from 0 minute to 120 minutes, and then the lifetime was measured by a μ-PCD method. As a result, it can be seen that the slope of τ ₂ changes depending on the heating time.

When the relationship between the pre-annealing time and ΔV _th or the lifetime is plotted as shown in FIG. 12B, it can be seen that the lifetime value of τ _{2 that} changes according to the pre-annealing time is approximately proportional to ΔV _th . It is to be noted that defects in the film tend to decrease as the pre-annealing temperature increases.

  Accordingly, defects in the film can be reduced by appropriately adjusting manufacturing conditions such as a film forming condition of the oxide semiconductor thin film and a pre-annealing condition after the film formation. Note that the above-described oxygen concentration, pre-annealing temperature, and pre-annealing time vary depending on the composition of the oxide semiconductor thin film, the number of defects in the film, the thermal history temperature, and the like. Therefore, by evaluating defects for a plurality of oxide semiconductor thin films having different oxide semiconductor thin film formation conditions and pre-annealing conditions after film formation, conditions for reducing defects in the film can be selected from the evaluation results. In addition, even if the measured oxide semiconductor thin film does not include one in which the defects in the film are sufficiently reduced, if the plurality of evaluation results are plotted on a graph, the optimum condition can be found from the tendency of the results. Is also possible. For example, defects in the film can be reduced by controlling the pre-annealing temperature after forming the oxide semiconductor thin film to about 300 to 350 ° C. and the pre-annealing time in the temperature range to about 50 to 70 minutes, preferably about 60 minutes.

As in the case of the above-described defect in the film, a defect occurs at the interface of the oxide semiconductor thin film depending on manufacturing conditions, and the electronic state changes. Therefore, the slope of τ ₂ changes due to the interface defect, and the stress resistance is also affected. Therefore, interface defects can be reduced by controlling manufacturing conditions such as a protective film forming condition and a post-annealing condition after forming the protective film based on the evaluation of the oxide semiconductor thin film. It is particularly preferable to control the conditions for forming the protective film.

  In the case where the manufacturing conditions for reducing the defects in the film and reducing the interface defects based on the evaluation result of the oxide semiconductor thin film are predicted, it is preferable to perform the following procedure.

First, the electronic state of an oxide semiconductor thin film of a stacked body obtained by performing processing under different manufacturing conditions is measured, and a defect due to a defect is evaluated. As shown in FIG. 11, since the slow decay measured by the μ-PCD method differs depending on the oxygen concentration at the time of forming the oxide semiconductor film, a defect due to a defect in the oxide semiconductor thin film can be evaluated from τ ₂ . For example, the sheet resistance of each oxide semiconductor thin film processed under different manufacturing conditions is measured, and τ _{2 is} plotted on the horizontal axis and sheet resistance is plotted on the vertical axis as shown in FIG. 13A. When the relationship is plotted, when τ ₂ is long or the electric resistivity is small, the number of defects in the film tends to increase. On the other hand, when the electric resistivity is high, there are many interface defects, and when the electric resistivity is low, the number of hydrogen donors tends to increase at the interface between the oxide semiconductor thin film and the protective film. Accordingly, a plurality of oxide semiconductor thin films that have been subjected to different treatments under different manufacturing conditions are prepared, and the electronic state of the oxide semiconductor thin film is evaluated by a contact method or a non-contact method, for example, as shown by a thick arrow in FIG. 13A. It is necessary to adjust the difference between the measured resistance value and the optimum resistance value. That is, the manufacturing conditions of the oxide semiconductor thin film are determined so that τ ₂ and the electrical resistivity have appropriate values.

As a method for adjusting the defects in the film, as described above, the manufacturing conditions before forming the protective film, such as the conditions for forming the oxide semiconductor thin film and the pre-annealing conditions after forming the oxide semiconductor thin film, may be optimized. In addition, as a method for adjusting the interface defect, it is only necessary to optimize manufacturing conditions after formation of the protective film, such as conditions for forming the protective film such as an atmosphere and a gas ratio at the time of forming the protective film, and post-annealing conditions after forming the protective film. For example, as the number of interface defects increases, the depletion region on the surface of the oxide semiconductor thin film tends to increase. Therefore, the depletion region can be adjusted by controlling the post-annealing conditions. Since the accumulation region of the oxide semiconductor thin film tends to increase as the number of hydrogen donors increases, the accumulation region can be adjusted by controlling the amount of oxygen at the time of forming the protective film. By controlling the manufacturing conditions that affect the quality of the oxide semiconductor thin film in this manner, an oxide semiconductor thin film in which defects due to defects in the film and interface defects are optimized can be obtained. Since the oxide semiconductor thin film optimized based on the evaluation result has reduced defects, the value of τ ₂ and the electrical resistivity are plotted in FIG. 13A as compared with the oxide semiconductor thin film before optimization. , Shift to the optimal value.

  When the peak value of the lifetime obtained by measuring the lifetime of the oxide semiconductor thin film by the μ-PCD method and the “absolute value of the power shoulder” corresponding to the slow decay are plotted, for example, as shown in FIG. 13B. 13B, the mobility of the oxide semiconductor thin film can be predicted from the position on the vertical axis, and the stress resistance of the oxide semiconductor thin film can be predicted from the position on the horizontal axis. As the sheet resistance increases, the donor-like center below the conduction band tends to decrease, and as the absolute value of the power shoulder increases, the shallow level that creates space charge decreases. Therefore, it is preferable that the mobility is high and the absolute value of the power shoulder is large. As described above, the sheet resistance of the oxide semiconductor thin film can be improved by controlling manufacturing conditions such as heat treatment. The mobility can also be improved by controlling the manufacturing conditions. However, as for the relation between the sheet resistance and the value of the power shoulder, the absolute value of the power shoulder tends to decrease at a certain sheet resistance value. For example, in FIG. 13B, by adjusting the annealing conditions, the absolute value of the power shoulder is shifted to the left on the horizontal axis, and the peak value is shifted upward in the vertical axis to optimize the oxide semiconductor thin film. Therefore, the manufacturing conditions such as the amount of oxygen added during the formation of the protective film and the pre-annealing time are appropriately adjusted so that the peak value is high from the plotted measured values and the absolute value of the power shoulder has the maximum value. It is desirable to optimize the electronic state of the thin film. Note that it is sometimes difficult to optimize the electronic state of the oxide semiconductor thin film only by adjusting the final protective film forming step and the post-annealing step. Therefore, the above-described manufacturing steps (i) to (iv) are optimized. It is desirable.

  As described above, at least one of manufacturing conditions such as a film forming condition of an oxide semiconductor thin film when manufacturing a TFT, a pre-annealing condition after forming an oxide semiconductor thin film, or a film forming condition of a protective film and a post annealing condition. By adjusting the ratio, a stacked body of the oxide semiconductor thin film and the protective film with few defects due to defects can be obtained, and a TFT having excellent stress resistance can be manufactured.

[Quality control method]
Further, the present invention also includes a method of performing quality control of an oxide semiconductor thin film by applying the above evaluation method. As the quality control method, the evaluation method of the present invention may be applied as described above, and the evaluation result of the electronic state of the oxide semiconductor thin film may be fed back. By adjusting the manufacturing conditions as described above, specifically, at least one of the manufacturing steps (i) to (v) based on the feedback, defects in the oxide semiconductor thin film can be reduced. As a result, appropriate quality control of the oxide semiconductor thin film can be performed.

  Hereinafter, parameters are derived based on the μ-PCD method, which is a non-contact method as a typical example, and the electronic state of the oxide semiconductor thin film is evaluated based on the parameters, and the manufacturing conditions are predicted based on the evaluation. A method for controlling the quality of an oxide semiconductor thin film will be described.

  When the μ-PCD method is used, (1) the parameter P corresponding to the slow attenuation in the parameter calculation step corresponding to the reflectance measurement step, or (2) the maximum value of the reflected wave in the reflectance measurement step is used as the above parameter. Parameter Q and the corresponding parameter P can be used.

  As another evaluation item, it is also desirable to evaluate the in-plane uniformity of the oxide semiconductor thin film from the viewpoint of exhibiting good TFT performance when incorporated in a product such as a display.

  (1) A case where the parameter P is used will be described. In the present invention, a preferable value PX of the parameter P is set according to the required characteristics of the TFT. An oxide is formed based on a parameter PXn (n is an arbitrary number given for each manufacturing condition, the same applies hereinafter) which is a value obtained by actually measuring a parameter P of an oxide semiconductor thin film having a protective film manufactured under different manufacturing conditions. Evaluate the electronic state of the semiconductor thin film. Although the electronic state of the oxide semiconductor thin film, that is, the parameter PXn, varies depending on the manufacturing conditions in each manufacturing process, the manufacturing conditions are adjusted based on the evaluation to select a manufacturing condition that falls within a preferable range of the parameter PX. A transistor structure including an oxide semiconductor thin film having favorable bias stress characteristics corresponding to medium defects and interface defects can be manufactured.

  When comparing the electronic states of the oxide semiconductor thin film having a protective film, it is preferable to use a comparison result of a plurality of oxide semiconductor thin films in order to evaluate the effect of processing conditions on the electronic state of the oxide semiconductor thin film. The more comparison results, the more precise the production conditions can be predicted. For example, a plurality of samples in which an oxide semiconductor thin film is formed under the same manufacturing conditions are prepared, and then, a stacked body obtained by forming a protective film and the like under different manufacturing conditions is evaluated. Conditions may be determined. Alternatively, prepare a plurality of samples in which the oxide semiconductor thin film is formed under different manufacturing conditions, and then evaluate the oxide semiconductor thin film of the stacked body obtained by forming the protective film and the like under the same manufacturing conditions, and determine the manufacturing conditions. Is also good.

  As described above, in the present invention, the relationship between the defect and the manufacturing condition is grasped by measuring the electronic state of the oxide semiconductor thin film having the protective film as the parameter PXn, and the manufacturing condition for reducing the defect is predicted and determined. it can. By employing the manufacturing conditions determined based on these evaluations, an oxide semiconductor thin film in which both in-film defects and interface defects are reduced can be manufactured. When the oxide semiconductor thin film is used, an oxide semiconductor thin film having excellent stress resistance and a stacked body having a protective film on the surface of the oxide semiconductor thin film can be manufactured. The laminate preferably has a structure equivalent to that of a TFT. Therefore, in addition to the oxide semiconductor thin film and the protective film, the laminate may include an insulating layer such as a gate insulating film as necessary. Alternatively, it may be processed into a wiring structure necessary for the TFT.

  Next, (2) a case where the parameter Q and the parameter P are used will be described. According to the present invention, the characteristics of the transistor structure having the oxide semiconductor thin film can be grasped more precisely by grasping the change in the electronic state of the oxide semiconductor thin film during the manufacturing process based on the parameters P and Q. As disclosed in Patent Document 1 by the present inventors, the relationship between the mobility of the oxide semiconductor thin film and the maximum value of the reflectance is such that the mobility is proportional to the maximum value of the reflectance. Tends to be higher. Therefore, similarly to the parameter PX, a preferable value parameter QX is set for the parameter Q. In addition, similarly to the parameter PXn, the electronic state of the oxide semiconductor thin film is evaluated based on the actually measured parameter QXn (n is the same number as the parameter PXn). The parameter Q also fluctuates depending on the manufacturing process and manufacturing conditions as in the case of the parameter P. However, by adjusting the manufacturing conditions based on the evaluation and selecting the manufacturing conditions that fall within the preferable range of the parameter QX, defects are reduced, and A transistor structure having a simple oxide semiconductor thin film can be manufactured. In the present invention, by adjusting the manufacturing conditions in consideration of both the parameter Q and the parameter P, the quality of the oxide semiconductor thin film with respect to mobility and bias stress characteristics can be more precisely managed.

  The parameter Q and the parameter P are values of the oxide semiconductor thin film in the same step. That is, when the oxide semiconductor thin film is measured by the μ-PCD method, the maximum value of the reflected wave in the reflectance measurement step is the parameter Q, and the parameter Q is derived after measuring the maximum value. The corresponding parameter is parameter P.

  In the present invention, the parameters Q and P may be evaluated by plotting them on a graph. For example, a plurality of oxide semiconductor thin films processed under different manufacturing conditions, where the vertical axis of the graph is the parameter Q corresponding to the maximum value of the reflected wave in the reflectance measurement step and the horizontal axis is the parameter P corresponding to the slow attenuation in the parameter calculation step By plotting the above parameters, the mobility and stress tolerance can be easily evaluated. The parameters may be plotted on the same or different graphs depending on the process and processing conditions. The number of steps, the number of processing conditions, and the number of parameters to be evaluated are not particularly limited. However, as the number of evaluation objects increases, it becomes easier to track changes in the electronic state of the oxide semiconductor thin film. The larger the parameter, the easier it is to predict the manufacturing conditions for correcting a change in the electronic state of the oxide semiconductor thin film.

  Further, quality control of the oxide semiconductor thin film can be performed by comparing parameters of the oxide semiconductor thin film between manufacturing steps. For example, in (1-1) to (3-3) of FIG. 1 of Example 1 to be described later, a plurality of oxide semiconductor thin films manufactured by changing the oxygen addition amount at the time of film formation and the gas ratio at the time of forming the protective film. For each of the processes, the parameters Q and P are measured for each process.

  From the viewpoint of manufacturing a TFT having good stress resistance, the value of the parameter PX after the protective film forming step is preferably 1.5 or less, more preferably 1.2 or less, and further preferably 1. Since it is 0 or less, preferably 0.5 or more, and more preferably 0.8 or more, if the value of the parameter PXn fluctuates, the manufacturing conditions may be appropriately adjusted so as to be within the above range. desirable.

  In addition, from the viewpoint of manufacturing a TFT having good stress resistance, the value of the parameter Q or QX after the protective film forming step is preferably 300 mV or more, more preferably 500 mV or more, and still more preferably 700 mV or more. Is preferably 1800 mV or less, more preferably 1500 mV or less, and still more preferably 1200 mV or less. When the value of the parameter QXn fluctuates, it is desirable to appropriately adjust the manufacturing conditions so as to fall within the above range.

[Evaluation element]
The evaluation element means a sample used in the above evaluation method. An evaluation element suitable for measuring the electronic state of the oxide semiconductor thin film has an oxide semiconductor thin film on a substrate and a protective film on the oxide semiconductor thin film. The process corresponds to a process typified by “a process for forming a protective film that can be formed on the surface of a semiconductor thin film”.

  The configuration shown in FIGS. 5 to 7 is exemplified as an evaluation element corresponding to the above step (iv).

  FIG. 5 shows that a gate insulating film 43, a patterned oxide semiconductor thin film 20b, and an etch stop layer 45 as a patterned protective film are formed in this order on a substrate 20a.

  FIG. 6 shows that a gate insulating film 43, an oxide semiconductor thin film 20b, and an etch stop layer 45 are formed in this order on a substrate 20a.

  FIG. 7 shows a structure in which an oxide semiconductor thin film 20b is directly formed on a surface of a substrate 20a, and a protective film such as an etch stop layer 45 is formed on the surface of the oxide semiconductor thin film.

Any of the evaluation elements may have a configuration required for measurement, such as an electrode, if necessary. It is important that the evaluation element has an oxide semiconductor thin film formed directly on the surface of the substrate or the gate insulating film. That is, a metal electrode such as a gate electrode does not exist directly below the oxide semiconductor thin film. When a gate electrode or the like exists immediately below the oxide semiconductor thin film, the number of electrons that are free carriers of the gate electrode is as large as 10 ¹⁸ cm ^-3 or more, so that the influence of the gate electrode on the microwave reflectance is dominant. Because it becomes.

  Furthermore, it is also preferable to use an evaluation substrate in which a plurality of the evaluation elements described above are arranged on the substrate.

  FIG. 8 is a schematic diagram of an evaluation substrate showing an example of the arrangement of the evaluation elements. As shown in FIG. 8, a display 50 and a plurality of evaluation elements 49 are regularly arranged on a mother glass 51 such as a glass substrate used in a mass production line. By using such an evaluation substrate, the quality control of the oxide semiconductor thin film, specifically, the in-plane distribution, that is, the variation of the electrical resistivity in the plane, and the distribution between the substrates, that is, the electrical resistivity between the substrates, Can be measured.

[Evaluation device]
Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. However, the evaluation device suitable for the present invention is not limited to the following configuration, and can be appropriately modified.

  FIG. 9 is a schematic diagram illustrating an example of a configuration of an apparatus used for measuring the oxide semiconductor thin film by a noncontact method. The evaluation device shown in FIG. 9 irradiates excitation light to a measurement site of the sample 20 having the configuration of FIG. 7 in which the laminate 41 having the protective film 45 on the surface of the oxide semiconductor thin film 20b is formed on the substrate 20a. Light irradiating means 1 for generating electron-hole pairs in an oxide semiconductor thin film, microwave irradiating means 3 for irradiating a measurement site of sample 20 with microwaves, microwave changing by irradiation of excitation light A reflected microwave intensity detecting means 7 for detecting the intensity of the reflected microwave from the sample 20; and a means for evaluating the electrical resistivity of the sample 20 based on the detection data of the reflected microwave intensity detecting means 7. With this configuration, the change in reflectance and the electrical resistivity can be measured and evaluated with the same device. It is desirable to use an evaluation element having a configuration as shown in FIGS. 5, 6, and 7 as a sample, but it is not limited to this.

  The excitation light irradiating means 1 has a light source that outputs excitation light for irradiating the sample 20, and generates electron-hole pairs in the oxide semiconductor thin film by irradiation with the excitation light. Since a carrier can be efficiently generated and measurement can be performed with high sensitivity, a light source which preferably outputs energy equal to or higher than the band gap of the oxide semiconductor thin film is provided. As the excitation light irradiating means 1, for example, an ultraviolet laser may be used as a light source. More specifically, a semiconductor laser such as a pulse laser that emits pulsed ultraviolet light having a wavelength of 349 nm, a power of 1 μJ / pulse, a pulse width of about 15 ns, and a beam diameter of about 1.5 mm, for example, a YLF laser third harmonic as excitation light. And so on.

  Further, the excitation light irradiating means 1 outputs pulse light which is excitation light, triggered by input of a timing signal transmitted from the evaluation means 9 (broken line in the figure). Note that the timing signal is also transmitted to the signal processing device 8 at the same time. The output of the excitation light output from the excitation light irradiation means 1 can be adjusted by the output adjustment power monitor 16a and the output adjustment means 16b.

  Excitation light output from the excitation light irradiating means 1 is reflected by an optical path changing means (hereinafter, may be referred to as “mirror”) 12 such as a mirror, and is also condensed by a condensing means such as a condensing lens (not shown). The light is condensed by a “condenser lens”), passes through the minute opening 6 c provided in the first waveguide 6 a, and the opening of the first waveguide 6 a at the end close to the sample 20. Through 6d, the sample 20 is irradiated on a measurement site having a diameter of, for example, about 5 to 10 μm. As described above, the mirror 12 and the condenser lens collect the excitation light output from the excitation light irradiating means 1 and guide it to the measurement site of the sample 20. As a result, excitation carriers are generated in the minute excitation light irradiation area 21 which is a measurement site in the sample 20.

  The microwave irradiating unit 3 is a unit that outputs a microwave to irradiate a measurement site of the sample 20. The microwave irradiation means 3 is, for example, a microwave oscillator such as a Gunn diode having a frequency of 26 GHz.

  The directional coupler 4 splits the microwave output from the microwave irradiation means 3 into two. One of the branched output waves (hereinafter, referred to as a first microwave Op1) is transmitted to the magic T5 side, and the other output wave (hereinafter, referred to as a second microwave Op2) is a phase adjuster 4a, which detects reflected microwave intensity. It is transmitted to the LO input of the means 7. As the directional coupler 4, for example, a 10 dB coupler or the like is employed.

  The magic T5 divides the first microwave Op1 into two, and the difference signal Rt1 (hereinafter, may be referred to as a "reflected wave difference signal") of each of the two branched first microwaves with respect to the sample 20. It outputs a sum signal.

  One of the microwaves Op1 branched into two by the magic T5 (hereinafter, may be referred to as “first main microwave Op11”) is excited by the first waveguide 6a connected to the magic T5. And is radiated from the opening 6d at the tip thereof. Thereby, the first main microwave Op11 is irradiated on the measurement site of the sample 20. Further, the first waveguide 6a has a function as an antenna for radiating the first main microwave Op11 (hereinafter, sometimes referred to as a “waveguide antenna”), and also has a function of irradiating the first main microwave Op11 to the measurement site. The reflected wave of the wave Op11 is also captured by the opening 6d at the tip of the wave Op11, and also functions to guide the wave back to the magic T5.

  On the other hand, the other of the first microwave Op1 branched into two by the magic T5 (hereinafter, referred to as “first sub-microwave Op12”) is measured by the second waveguide 6b connected to the magic T5 at the measurement site of the sample 20a. , But is guided to a portion that does not include the excitation portion by the excitation light and is radiated from the opening 6e at the tip thereof. As a result, the first sub-microwave Op12 is applied to the vicinity of the measurement site of the sample 20a. Further, the second waveguide 6b functions as a waveguide antenna for radiating the first sub-microwave Op12, and also transmits the reflected wave of the first sub-microwave Op12 irradiated near the measurement site to the opening at the tip thereof. The portion 6e also captures and performs a function of leading back to the magic T5. Here, the path length of the first waveguide 6a for guiding the microwave is equal to the path length of the second waveguide 6b for guiding the microwave.

  Also, the two reflected waves guided to the magic T5 by the first waveguide 6a and the second waveguide 6b, that is, the difference signal between the two reflected first microwaves Op11 and Op12 reflected on the sample 20, That is, the reflected wave difference signal Rt1 is output by the magic T5 and transmitted to the RF input terminal of the reflected microwave intensity detecting means 7.

  The reflected microwave intensity detecting means 7 outputs a detection signal Sg1 by mixing the second microwave Op2 and the reflected wave difference signal Rt1. The detection signal Sg1 is a signal representing an example of the intensity of the reflected wave difference signal Rt1, for example, the intensity of the reflected wave of the first microwave Op1 applied to the sample 20, and is taken into the signal processing device 8. The intensity of the reflected wave difference signal Rt1 changes when the sample 20 held at a predetermined position by the substrate holding unit is irradiated with the excitation light. As described above, the reflected microwave intensity detecting means 7 detects the intensity of the reflected wave difference signal Rt1, and as the reflected microwave intensity detecting means 7, a mixer or a microwave is input to correspond to the intensity. A microwave detector (hereinafter, sometimes referred to as a “detector”) that outputs an electric signal, that is, a current or a voltage, may be provided.

  The intensity of the reflected wave difference signal Rt1 detected by the reflected microwave intensity detecting means 7 changes by irradiating the measurement site of the sample 20 with the excitation light. Specifically, the intensity of the reflected wave difference signal Rt1 is temporarily increased by irradiation with the excitation light and then attenuated. Also, as the number of impurities, defects, and the like in the measurement site increases, the peak value of the intensity of the reflected wave difference signal Rt1 decreases, and the decay time, that is, the carrier lifetime also decreases.

  Here, with respect to the intensity of the reflected wave difference signal Rt1 that changes due to the excitation light irradiation, a parameter corresponding to the slow decay observed after the excitation light irradiation is stopped after the peak value is generated is an index for evaluating the electrical resistivity of the sample 20. Becomes

  The signal processing device 8 is a device that detects the peak value Sp of the change in the intensity of the reflected wave difference signal Rt1 detected by the reflected microwave intensity detection means 7 and transmits the detection result to the evaluation means 9. More specifically, the signal processing device 8 monitors the change of the reflected wave difference signal Rt1 for a predetermined time, triggered by the input of the timing signal from the evaluation means 9, and obtains the maximum value of the level of the reflected wave difference signal Rt1 obtained during that time. Is detected as the peak value Sp of the change in the intensity of the reflected wave difference signal Rt1. Here, the signal processing device 8 includes a delay circuit that performs delay processing on the reflected wave difference signal Rt1, sequentially detects the signal strength at a predetermined sampling frequency for the signal after the delay processing, and changes the detected value. , The peak value Sp of the change in the intensity of the reflected wave difference signal Rt1 is detected.

  As the evaluation means 9, a computer having a CPU, a storage unit, an interface for input / output signals, and the like can be used. The CPU executes various programs by executing a predetermined program.

  For example, the evaluation unit 9 outputs a timing signal indicating the output timing of the excitation light to the excitation light irradiating unit 1 and the signal processing device 8 and the peak value of the reflected wave difference signal Rt1 detected by the signal processing device 8. Sp is fetched and recorded in the storage unit provided in the evaluation means 9. The recorded reflected wave difference signal Rt1 (detection data) is used for evaluating the electrical resistivity of the sample 20.

  Further, the stage controller 10 controls the XY stage 11 in accordance with a command from the evaluation means 9 to control the positioning of the measurement site on the sample 20.

  A sample stage (not shown) is provided above the XY stage 11. The sample stage is a plate-shaped conductor member made of a metal such as aluminum, stainless steel or iron, or another conductor. A substrate holder (not shown) is provided on the upper side, and the sample 20 is placed on the substrate holder. As a result, the sample stage is arranged on the side opposite to the side on which the first microwaves Op11 and Op12 are irradiated to the sample 20, that is, below the sample 20.

  The substrate holding unit is a solid dielectric fixed on the upper side of the sample stage. The substrate holder is a solid dielectric inserted between the substrate and the sample table, and is made of a material having a relatively large refractive index, such as glass or ceramic. As a result, the wavelength of the microwave using the substrate holding portion as a medium is shortened, and a thinner and lighter substrate holding portion can be adopted.

  As described above, according to the configuration for evaluating electric resistivity of the present invention, photoexcited carriers are generated in the oxide semiconductor thin film by the excitation light radiated from the excitation light irradiating means 1 and the microwave excitation means 3 The photoexcited carriers move in the irradiated microwave electric field, and the motion state is affected by the presence of impurities, defects, and the like in the semiconductor. Therefore, the reflected microwave intensity detecting means 7 detects the intensity of the reflected microwave from the sample, and analyzes the change in the excess carrier concentration as already described in the evaluation means 9, thereby obtaining the carrier of the oxide semiconductor thin film. By determining the concentration, the electrical resistivity can be evaluated indirectly from the change in the electronic state. At this time, by the evaluation means 9 controlling the position of the stage including the XY stage 11 and the like, mapping measurement for determining the electric resistivity in a predetermined range is also possible.

  Further, by providing the evaluation device of the present invention with an electric resistance measuring means, it is possible to provide not only the evaluation of the electric resistivity but also a device for in-line evaluation of the electric characteristics of the oxide semiconductor thin film in a short time. it can. In the evaluation of the electric resistivity, the electric resistivity is evaluated based on so-called slow decay. However, as a result of the study by the present inventors, measurement and evaluation based on the μ-PCD method depend on the number of defects. The electrical resistivity changes. Further, the electrical resistivity of the oxide semiconductor thin film may be different due to contamination, impurities, and the like even in the same plane, and the value varies depending on a measurement location. Therefore, in order to perform more appropriate quality control of the oxide semiconductor thin film, it is important that the measurement site by the μ-PCD method and the site by the electric resistivity measurement unit are substantially the same.

  Therefore, if the above-described evaluation device is provided with an electric resistance measuring means, it is possible to simply and accurately measure substantially the same portion simply by moving the XY stage appropriately. Therefore, if the above-described evaluation device provided with the electric resistance measuring means is used for a production line of a liquid crystal display device or the like, productivity can be greatly improved and more appropriate quality control of the oxide semiconductor thin film can be performed.

  An apparatus configuration provided with the electric resistance measuring means will be described with reference to FIG. FIG. 10 shows an apparatus provided with an electric resistance measuring means 30 in addition to the apparatus shown in FIG. 9 for measuring and evaluating a change in reflectance and an electric resistivity based on the above-described μ-PCD method. The electric resistance measuring means 30 does not necessarily have to be installed. Although the specific installation location is not limited, by moving the XY stage 11 as described above, the electrical resistivity is measured by the electrical resistance measuring means 30 at substantially the same location as the microwave photoconductive measurement location of the oxide semiconductor thin film. It is desirable to install it as much as possible. The electric resistance measuring means 30 preferably has an electric resistivity measuring head 31 and an elevating means 32 of the electric resistivity measuring head 31. The electric resistivity of the sample 20 can be measured by the electric resistance measuring means 30.

  The electric resistivity measuring head 31 is a means for measuring electric resistivity by a contact method. The electric resistivity measuring head 31 is provided with a resistance measuring terminal corresponding to the above-described resistance measuring means. As the resistance measuring terminal, there are a measuring probe such as a double ring electrode or four needle-like electrodes in a straight line. Is exemplified. The electrical resistivity of the oxide semiconductor thin film can be measured by a resistance measurement using a double ring electrode in accordance with JIS K6911 or a four-probe method in accordance with JIS K7194.

The elevating means 32 of the electric resistivity measuring head 31 is an elevating mechanism for lowering the electric resistivity measuring head to a desired position when measuring the electric resistivity of the sample 20. As a means for measuring the electric resistivity, various known electric resistivity measuring devices can be used. For example, when an electric resistance measuring device such as a high resta manufactured by Mitsubishi Chemical Analytech Co., Ltd. is used, the probe corresponding to the electric resistivity measuring head 31 is lowered by the elevating / lowering means 32 so as to contact the surface of the sample 20, and the electric resistance is measured. After measuring the ratio, the probe and the sample 20 may be raised so as to be in a non-contact state. The measured electrical resistivity can be sent to an evaluation unit (not shown) having a configuration similar to that of the evaluation unit 9 through the measured value transmission line 33 and evaluated. In addition, an electric resistance measuring device such as a probe head manufactured by JANDEL can be similarly evaluated.

[TFT manufacturing method]
Further, by applying the above evaluation method to a method for manufacturing a TFT, a TFT having an oxide semiconductor thin film in which defects of the oxide semiconductor thin film are reduced can be manufactured.

  The configuration of the TFT is not limited, and a known configuration can be employed. For example, a TFT has a gate insulating film, an oxide semiconductor thin film, a protective film formed on a surface of the oxide semiconductor thin film, and a source electrode and a drain electrode on a substrate. Then, the oxide semiconductor thin film and the protective film of the TFT may be formed under conditions determined based on the evaluation of the oxide semiconductor thin film.

  As described above, the electronic state of the oxide semiconductor thin film can be measured by any of a contact method and a noncontact method, and a defect resulting from a defect in the oxide semiconductor thin film can be evaluated. Therefore, according to the above-described method for manufacturing a TFT, manufacturing conditions that can reduce defects of the oxide semiconductor thin film as described above can be appropriately determined.

  Hereinafter, the present invention will be described more specifically with reference to Examples. However, the present invention is not limited to the following Examples, and may be appropriately modified within a range that can conform to the purpose of the preceding and the following. It is, of course, possible to implement them, and all of them are included in the technical scope of the present invention.

Example 1
In Example 1, the following experiment was performed to evaluate the correlation between the lifetime value of the oxide semiconductor thin film calculated based on the μ-PCD method and the TFT characteristics such as ΔV _th as stress resistance. In this example and the examples described later, the correlation between the lifetime value calculated by the following evaluation 1 and the threshold shift was examined.
Evaluation 1: The microwave reflectance is represented by the following equation (4), and the lifetime value when parameter fitting is performed: Parameter B

  In the equation, A is a constant, and B is a parameter corresponding to a slow decay observed about 1 μs after the stop of the irradiation of the excitation light.

  First, in this example, after a plurality of stacked bodies were manufactured by changing the amount of oxygen added in the oxide semiconductor thin film forming step and the carrier gas ratio in the protective film forming step, the oxide semiconductor thin film was formed by the μ-PCD method. Was evaluated. The temperature conditions and the processing time of the pre-annealing process were adjusted based on the evaluation results. In this example, in order to evaluate the film quality variation due to the protective film forming condition, the parameter P of the oxide semiconductor thin film after the pre-annealing process of each sample was adjusted to be in the range of 0.5 to 0.7. . As a preferable electronic state of the oxide semiconductor thin film having the protective film, the parameter PX was set to 0.8 to 1.0, and the parameter QX was set to 300 to 1500 mV.

  In the following examples, under such preconditions, the manufacturing conditions of the protective film formation step have a great effect on the quality of the oxide semiconductor thin film, and the protective film formation step is performed under appropriate manufacturing conditions. And that the evaluation by the μ-PCD method has a correlation with the characteristics of the TFT.

(1) Preparation of Sample for Measuring Lifetime Value First, a glass substrate having a diameter of 100 mm and a thickness of 0.7 mm (EAGLEX X manufactured by Corning Incorporated) was used.
On top of G), a 200-nm-thick SiO ₂ film was formed as a gate insulating film. The gate insulating film is formed by plasma CVD using a carrier gas of a mixture gas of SiH ₄ and N ₂ O, N ₂ O = 100 sccm, SiH ₄ = 4 sccm, N ₂ = 36 sccm, pressure: 200 Pa, film forming power: 300 W, The film was formed at a film temperature of 320 ° C.

[Oxide semiconductor thin film forming step: sometimes referred to as "ASDEPO"]
Next, IGZO was formed as an oxide semiconductor thin film by a sputtering method. In this example, a plurality of samples were prepared by changing the amount of added oxygen as described below.
Sputtering equipment: "CS-200" manufactured by ULVAC
Composition of sputtering target: InGaZnO ₄ [In: Ga: Zn = 1: 1: 1 (atomic ratio)]
Substrate temperature: room temperature Thickness of oxide semiconductor thin film: 40 nm
Gas pressure: 1 mTorr
_{Oxygenation: O 2 / (Ar + O} 2) = 4%, 10%, 20% ( by volume)

[Pre-annealing process: sometimes referred to as “PRE”]
Next, a pre-annealing process was performed. The pre-annealing treatment was performed at 350 ° C. for one hour in the air extracted as the optimum condition. Pre-annealing under optimum conditions reduced defects in the film.

[Protective film forming step: sometimes referred to as "ESL"]
Next, a protective film was formed over the oxide semiconductor thin film. In this example, a plurality of samples were prepared by changing the carrier gas at the time of forming the protective film as described below.
Gas pressure: 133Pa
Film formation power: 100W
Film formation temperature: 230 ° C
Film thickness: 100 nm
Carrier gas: N ₂ O = 100 sccm, SiH ₄ / N ₂ = 4/36 sccm,
Carrier gas: N ₂ O = 150 sccm, SiH ₄ / N ₂ = 6/54 sccm,
Carrier gas: N ₂ O = 250 sccm, SiH ₄ / N ₂ = 10/90 sccm

[Final protective film formation step: sometimes referred to as "PV"]
Further, a laminated film of 200 nm thick SiO ₂ and 200 nm thick SiN was formed as a final protective film. The formation of the final protective film was performed by using a plasma CVD method using "PD-220NL" manufactured by Samco. In this embodiment, after performing a plasma treatment with N ₂ O gas, SiO ₂ and SiN were sequentially formed on the etch stop layer under the following conditions. A mixed gas of N ₂ O and SiH ₄ was used for forming SiO ₂ , and a mixed gas of SiH ₄ , N ₂ , and NH ₃ was used for forming SiN. In each case, the film forming power was 100 W and the film forming temperature was 150 ° C.

(First layer): SiO ₂
Carrier gas: N ₂ O = 100 sccm, SiH ₄ / N ₂ = 4/36 sccm
Gas pressure: 133Pa
Film formation power: 100W
Film formation temperature: 150 ° C
Film thickness: 100 nm

(Second layer): SiN
Carrier gas: N ₂ O = 100 sccm, SiH ₄ = 12.5 sccm, N ₂ = 297.5 sccm
Gas pressure: 133Pa
Film formation power: 100W
Film formation temperature: 150 ° C
Film thickness: 150 nm

[Post-annealing process: sometimes referred to as "PA"]
After forming the final protective film, heat treatment was performed at 250 ° C. for 30 minutes in a nitrogen atmosphere as final annealing.

  The samples after each of the PRE, ESL, PV, and PA steps were subjected to the μ-PCD method to measure the change in reflectance. The measurement result is analyzed, and a parameter Q corresponding to the maximum value of the reflected wave and a B value corresponding to a gradient of 0.3 to 1 μsec are calculated. did.

The parameters Q and P of each process are as follows.
QX0, PX0: Parameter P and parameter Q after oxide semiconductor thin film deposition step
QX1, PX1: Parameter P and parameter Q after oxide semiconductor thin film formation and after pre-annealing process
QX2, PX2: parameters P and Q after a protective film forming step for directly protecting the surface of the oxide semiconductor thin film
QX3, PX3: Parameter P and parameter Q after final protective film formation step for further protecting the protective film surface
QX4, PX4: Parameter P and parameter Q after final protective film formation and post-annealing process
Each value was plotted in FIG. In FIG. 1, the vertical axis represents the parameter Q (unit: mV), and the horizontal axis represents the parameter P.

Parameters PX = 0.8 to 1.0 and parameters QX = 300 to 1500 mV were set as reference values. In FIG. 1, Low, Mid, and High indicate oxygen concentrations: Low = 4%, Mid = 10%, and High = 20% under sputtering conditions when an oxide semiconductor is formed. Further, SiH ₄ / N ₂ O indicates a gas ratio under the CVD conditions when forming an etch stop layer as a protective film. 1 (1-1), (2-1) and (3-1) show oxygen concentrations: Low, (1-2), (2-2) and (3-2) show oxygen concentrations: Mid and (1). (3), (2-3) and (3-3) are examples of oxygen concentration: High. (1-1) to (1-3): SiH ₄ / N ₂ O = 4/100, (2-1) to (2-3): SiH ₄ / N ₂ O = 6/150, (3-) 1) to (3-3) are examples of SiH ₄ / N ₂ O = 10/250.

(2) Preparation of TFT sample for measuring TFT characteristics and stress resistance To confirm the TFT characteristics of the sample prepared in (1) above, a TFT shown in FIG. 4 was prepared and the TFT characteristics and stress resistance were evaluated.

First, a Mo thin film as a gate electrode and a gate insulating film SiO ₂ were sequentially formed in a thickness of 100 nm on a glass substrate (EAGLE2000 manufactured by Corning) having a diameter of 100 mm and a thickness of 0.7 mm. The gate electrode was formed by a DC sputtering method using a pure Mo sputtering target. The sputtering conditions were as follows: substrate temperature: room temperature, gas pressure: 2 mTorr. The gate insulating film is formed by a plasma CVD method using a carrier gas: a mixed gas of SiH ₄ and N ₂ O (N ₂ O = 100 sccm, SiH ₄ = 4 sccm, N ₂ = 36 sccm), and a film forming power of 300 W. Temperature: 320 ° C.

  Next, IGZO was formed as an oxide semiconductor thin film by a sputtering method under the same conditions as those of the sample manufactured in the above (1).

  After the oxide semiconductor thin film was formed as described above, patterning was performed by photolithography and wet etching. As the wet etchant, "ITO-07N" manufactured by Kanto Chemical was used.

  After patterning the oxide semiconductor thin film in this manner, a pre-annealing step and a protective film forming step were performed under the same conditions as those of the sample manufactured in (1).

  Next, using pure Mo, a film was formed to a thickness of 100 nm by a DC sputtering method, and then patterned to form source / drain electrodes. The method for forming the pure Mo film and the patterning method are the same as those for the gate electrode described above, and the channel length of the TFT is 10 μm and the channel width is 200 μm.

After forming the source / drain electrodes in this manner, a final protective film for protecting the oxide semiconductor thin film was formed. As the protective film, a stacked film of SiO ₂ having a thickness of 200 nm and SiN having a thickness of 200 nm was used. The formation of the above-mentioned SiO ₂ and SiN was performed using “PD-220NL” manufactured by Samco by a plasma CVD method. In this embodiment, after performing a plasma treatment with N ₂ O gas, a SiO ₂ film and a SiN film are sequentially formed. A mixed gas of N ₂ O and SiH ₄ was used for forming the SiO ₂ film, and a mixed gas of SiH ₄ , N ₂ , and NH ₃ was used for forming the SiN film. In each case, the film forming power was 100 W and the film forming temperature was 150 ° C.

  After forming the final protective film, heat treatment was performed at 250 ° C. for 30 minutes in a nitrogen atmosphere as final annealing. Next, a contact hole for probing for evaluating transistor characteristics was formed in the protective film by photolithography and dry etching to obtain a TFT.

For each of the TFTs thus obtained, the change ΔV _th of the threshold voltage V _th before and after the application of the stress was examined as follows. The results are shown in FIG. In FIG. 2, “E” of “1.0E” means “exponential”.

(I) ΔV _th was evaluated as stress resistance. In the present embodiment, a stress application test was performed in which a gate electrode was irradiated with light while applying a negative bias to simulate an actual stress environment during panel driving. The stress application conditions are as follows. The wavelength of light was selected to be about 400 nm, which is close to the band gap of the oxide semiconductor and in which the transistor characteristics are likely to fluctuate.
Gate voltage: -20V
Substrate temperature: 60 ° C
Light stress Light source: White light source Light intensity applied to TFT as illuminance: 25,000 NIT
Light irradiation device: YSM-1410 manufactured by Yang Electronics
Stress application time: 2 hours

Regarding the parameter P, the following can be understood from FIG. 1 showing the electronic state of each sample and FIG. 2 showing the TFT characteristics of the corresponding sample. As shown in FIG. 1, the value of the parameter P1 was adjusted to about 0.5 to 0.7 for each of the oxide semiconductor thin films of PRE, but the SiH ₄ / N ₂ O gas ratio of ELS and the oxygen content of ASDEPO were adjusted. As a result, the value of the parameter PX2 fluctuated, and the range of the parameter PX defined in the present embodiment exceeded 0.8 to 1.0 depending on the manufacturing conditions. For example, under the manufacturing conditions shown in FIGS. 1A to 1C, the value of the parameter PX2 satisfies the upper limit value 1.5 of the parameter P defined in the present invention, and the good TFT characteristics shown in FIG. This is a suitable production condition with many samples. However, FIGS. 1 (1-2) to (2-1) were inferior in stress resistance as compared with FIG. 1 (1-1), but this was due to the parameter of FIGS. 1 (1-2) to (2-1). This is because PX2 is large outside the range of the parameter PX set in this embodiment. On the other hand, the parameter PX2 in FIG. 1 (1-1) is within the range of the parameter PX set in the present embodiment. As shown in FIG. 2, FIG. It can be seen that the manufacturing conditions are appropriate.

  On the other hand, under the manufacturing conditions shown in FIGS. 1 (2-2) to (3-3), many samples have the parameter PX2 exceeding 1.5, and as shown in FIG. ) Are inferior in stress resistance, indicating that the manufacturing conditions are not fully optimized.

  In addition, in FIGS. 1 (2-2) to (3-3), the values of the parameters PX3 and PX4 remained large even after performing each step of PV and PA. Similarly, in FIGS. 1 (1-2) to (2-1), the parameter PX2 and the parameters PX3 and PX4 show almost the same value, and the improvement was not possible. On the other hand, as shown in FIG. 1A, when the parameter PX2 is within the range of the set parameter PX, the parameters PX3 and PX4 substantially reduce the value of the parameter PX2 even after a post-process such as PV or PA. It has been found that a good electronic state can be maintained. From these results, it is found that the quality of the oxide semiconductor thin film is greatly affected by the manufacturing conditions in the ESL process, and the parameter PX2 after the ESL process hardly improves or changes even after the subsequent processes. Therefore, if the parameter PX2 of the oxide semiconductor thin film after the ESL step is made as close as possible to the parameter PX set as a good electronic state, a good electronic state can be maintained even after the subsequent manufacturing steps.

  1 and 2 show the following for the parameter P and the parameter Q.

  (1): As shown in FIG. 1 (3-1) to (3-3), the parameter PX2 has a very large value such that PX2> 1.5, and the parameter QX2 satisfies QX2 <300 mV. It has been found that when the TFT has a very small value, the TFT becomes conductive as shown in FIGS. 2 (3-1) to (3-3). In the case where PX2> 1.5 and QX2 <300 mV in either one of PX2 and QX2, the threshold voltage shift is large and the resistance to stress may be poor. It is important to control so that it is within the proper range.

(2): When the parameter PX2 has a very large value such that PX2> 1.5 as shown in FIGS. 1 (2-2) and (2-3) and the parameter QX2 has QX2 ≧ 300 mV. Although the TFT is not made conductive, the stress resistance accompanying light irradiation is ΔV _th ≧ 6.0 V, the threshold voltage shift is large with time, and the TFT does not have sufficient stress resistance. It can be seen that the characteristics are inferior.

(3): When the parameter PX2 is 1.0 <PX2 <1.5 and the parameter QX2 is in the range of 300 to 1800 mV as shown in FIGS. 1 (1-2) to (2-1), Good stress resistance satisfying ΔV _th ≦ 5.0 V was exhibited.

(4): 1 Parameter PX2 as shown in (1-1) is in the range of 0.8 to 1.0, and if the parameter QX2 is in the range of 300~1800mV, ΔV _th ≦ 3. 0V, indicating further excellent stress resistance.

  From the above results, it has been found that if the parameters QX2 and PX2 satisfy the preferable ranges of the parameters QX and PX, respectively, as described in (3) above, good stress resistance is obtained. In particular, it was found that when the parameter PX2 was in the range of 0.8 to 1.0 as in (4) above, it had more excellent stress resistance.

  Also, as shown in FIG. 1, as with the parameter P, even if the parameter QX1 is good, the parameter QX2 changes depending on the protective film deposition conditions and may deviate from a good electronic state as shown in FIG. . In addition, the parameters QX3 and QX4 tended to have substantially the same value as the parameter QX2 even after performing each step of PV and PA. Therefore, if the parameter QX2 is out of the range of the parameter QX, it is difficult to return the parameter QX2 to the range of the parameter QX even after a subsequent manufacturing process, and the defect of the oxide semiconductor thin film may not be sufficiently reduced. On the other hand, if the parameter QX2 is optimized by adjusting the conditions for forming the protective film, the oxide semiconductor thin film can maintain a good state even after the subsequent manufacturing process.

As is clear from a comparison between FIG. 1 and FIG. 2, by reducing defects in the oxide semiconductor thin film, a TFT having favorable TFT characteristics such as stress resistance can be manufactured. FIG. 14 in which ΔV _th and the value of the parameter P are plotted respectively shows that there is a correlation between the evaluation of the oxide semiconductor thin film and the stress resistance. That is, in the present embodiment, the closer the parameter P is to the optimum value of 0.8 to 1.0, the smaller the shift amount of the threshold voltage is, and the better the stress resistance is. Similarly, the closer the parameter Q is to the optimal value of 300 to 1500 mV, the more excellent the threshold voltage shift amount and the stress resistance are.

  From the above, the electronic state of the oxide semiconductor thin film having a protective film is measured and evaluated, and based on the evaluation, the film forming condition and the protective film forming condition of the oxide semiconductor thin film are adjusted, and the protective film after ESL is formed. By optimizing the electronic state of the oxide semiconductor thin film, defects in the oxide semiconductor thin film can be reduced, and a TFT having favorable TFT characteristics such as stress resistance can be manufactured.

DESCRIPTION OF SYMBOLS 1 Excitation light irradiation means 3 Microwave irradiation means 4 Directional coupler 4a Phase adjuster 5 Magic T
6a first waveguide 6b second waveguide 6c minute aperture 6d, 6e aperture 7 reflected microwave intensity detection means 8 signal processing device 9 evaluation means 10 stage controller 11 XY stage 12 optical path changing means 16a output adjustment Power monitor 16b Output adjusting means 20 Sample 20a Substrate 20b Oxide semiconductor thin film 21 Excitation light irradiation area 30 Electrical resistivity measuring means 31 Electrical resistivity measuring head 32 Elevating means 33 Measured value transmission line 41 Stacked body 42 Gate electrode 43 Gate insulating film 45 Etch stop layer or protective film 46a Source electrode 46b Drain electrode 47 Final protective film 48 Contact hole 49 Evaluation element 50 Display 51 Mother glass

Claims (6)

A quality evaluation method for a laminate having a protective film on the surface of an oxide semiconductor thin film,
By measuring an electronic state of the oxide semiconductor thin film having the protective film, a defect caused by a defect of the oxide semiconductor thin film is evaluated, and thereby the following (1) is evaluated. Quality evaluation method.
(1) When light irradiation and a negative bias are applied to the thin film transistor, the difference ΔVth between the threshold voltages before and after the application of the light irradiation and the negative bias The electronic state of the oxide semiconductor thin film is
After irradiating the oxide semiconductor thin film with excitation light and microwaves and measuring the maximum value of the reflected wave from the irradiated portion of the microwaves changed by the irradiation of the excitation light, stopping the irradiation of the excitation light Then, a reflectance measuring step of measuring a temporal change in reflectance of a reflected wave from the irradiated portion of the microwave after stopping irradiation of the excitation light,
A step of calculating a parameter P corresponding to a slow decay observed after the stop of the irradiation of the excitation light from a temporal change of the reflectance by a non-contact method. Quality evaluation method described.   The evaluation of the electronic state of the oxide semiconductor thin film is to evaluate the quality of (1) as good when the parameter P corresponding to slow decay in the parameter calculation step is 1.5 or less. 2. The quality evaluation method described in 2. A quality evaluation method for a laminate having a protective film on the surface of an oxide semiconductor thin film,
By measuring the electron state of the oxide semiconductor thin film having the protective film, to evaluate the defects caused by defects in the oxide semiconductor thin film, thereby state, and are used to evaluate the following (2),
The electronic state of the oxide semiconductor thin film is
After irradiating the oxide semiconductor thin film with excitation light and microwaves and measuring the maximum value of the reflected wave from the irradiated portion of the microwaves changed by the irradiation of the excitation light, stopping the irradiation of the excitation light Then, a reflectance measuring step of measuring a temporal change in reflectance of a reflected wave from the irradiated portion of the microwave after stopping irradiation of the excitation light,
From the temporal change of the reflectance, a parameter calculation step of calculating a parameter P corresponding to the slow attenuation seen after the stop of the irradiation of the excitation light, a non-contact method,
The evaluation of the electronic state of the oxide semiconductor thin film is such that when the maximum value of the reflected wave in the reflectance measurement step is a parameter Q, the parameter P corresponding to the slow attenuation in the parameter calculation step is 1.5 or less. The quality evaluation method of the laminate , wherein the quality of the following (2) is evaluated as good when the parameter Q is 300 mV or more .
(2) Whether thin film transistors can be switched The parameter calculating step calculates a parameter P corresponding to a slow decay seen in 0.1 to 10 μs after stopping the irradiation of the excitation light from the change in the reflectance, and evaluates an electronic state of the oxide semiconductor thin film. The quality evaluation method according to any one of claims 2 to 4 , wherein the quality is evaluated. A quality control method for an oxide semiconductor thin film using the quality evaluation method according to any one of claims 2 to 5 ,
The electronic state of the oxide semiconductor thin film having the protective film is evaluated based on the quality evaluation method, and based on the evaluation, a film forming condition of the oxide semiconductor thin film, a heat treatment condition after forming the oxide semiconductor thin film, And a quality control method for the oxide semiconductor thin film, wherein at least one of the film forming conditions of the protective film is adjusted.