Film forming method and film forming apparatus

Abstract

To provide a technique of forming a silicon nitride film with suitable film quality.SOLUTION: A substrate is placed on a mount table installed in a processing chamber, and processing gas including silicon-containing gas and nitrogen-containing gas is supplied into the processing chamber. On the other hand, first power with a first frequency higher than 300 MHz and second power with a second frequency that is lower than the first frequency are applied to an electrode facing the mount table in the overlapping manner, so that the processing gas in the processing chamber is made into plasma and thus, a silicon nitride film is formed on the substrate. By this method, a silicon nitride film having high tensile stress and film density and suitable film quality can be formed.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present disclosure relates to a film formation method and a film formation apparatus. [Background technology]

[0002] In the manufacturing process of semiconductor devices, a silicon nitride film is formed on a semiconductor wafer (hereinafter also referred to as "wafer"), which is a substrate. The silicon nitride film is used, for example, as a hard mask that covers parts that will not be removed by etching when patterning a formed film. Such a silicon nitride film is formed by, for example, a chemical vapor deposition (CVD) method by supplying a gas containing a film raw material or a gas that nitrides the film raw material to the wafer.

[0003] In the film forming process, highly reactive activated species obtained by converting the gas into plasma may be used. For example, Patent Document 1 proposes a technique for applying power of two different frequencies to an upper electrode in a parallel-plate plasma CVD apparatus in order to improve the quality of an insulating film deposited on a semiconductor wafer. The document describes that one of the two different frequencies is 10 to 100 MHz, and the other is 200 to 500 kHz. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Application Laid-Open No. 2002-367986 Summary of the Invention [Problem to be solved by the invention]

[0005] The present disclosure provides a technique for forming a silicon nitride film with good film quality. [Means for solving the problem]

[0006] The present disclosure provides: placing a substrate on a stage provided in a processing chamber; supplying a process gas containing a silicon-containing gas and a nitrogen-containing gas into the process chamber; a film formation step of applying a first power having a first frequency higher than 300 MHz and a second power having a second frequency lower than the first frequency to an electrode facing the mounting table, thereby generating plasma from the processing gas in the processing chamber and forming a silicon nitride film on the substrate; Equipped with picture, the second frequency is a frequency of 100 MHz or higher in the VHF band, The first frequency is 2n times (n is an integer) the second frequency. This is a film formation method. [Effects of the Invention]

[0007] According to the present disclosure, it is possible to form a silicon nitride film with good film quality. [Brief explanation of the drawings]

[0008] [Figure 1] 1 illustrates an example of the configuration of a film forming apparatus according to the present disclosure. [Figure 2] FIG. 10 is a first characteristic diagram showing the results of stress measurement. [Figure 3] FIG. 10 is a first characteristic diagram showing the measurement results of film density. [Figure 4] FIG. 10 is a second characteristic diagram showing the results of stress measurement. [Figure 5] FIG. 10 is a second characteristic diagram showing the measurement results of film density. [Figure 6] FIG. 10 is a third characteristic diagram showing the results of stress measurement. [Figure 7] FIG. 10 is a third characteristic diagram showing the measurement results of film density. [Figure 8] FIG. 4 is a fourth characteristic diagram showing the results of stress measurement. [Figure 9] FIG. 4 is a fourth characteristic diagram showing the measurement results of film density. [Figure 10] FIG. 5 is a fifth characteristic diagram showing the results of stress measurement. [Figure 11] FIG. 5 is a fifth characteristic diagram showing the measurement results of film density. [Figure 12] FIG. 6 is a sixth characteristic diagram showing the results of stress measurement. [Figure 13] FIG. 6 is a sixth characteristic diagram showing the measurement results of film density. [Figure 14] FIG. 7 is a seventh characteristic diagram showing the results of stress measurement. [Figure 15] FIG. 7 is a seventh characteristic diagram showing the measurement results of film density. [Figure 16] FIG. 11 is an eighth characteristic diagram showing the results of stress measurement. [Figure 17] FIG. 13 is an eighth characteristic diagram showing the measurement results of film density. [Figure 18] FIG. 9 is a ninth characteristic diagram showing the results of stress measurement. [Figure 19] FIG. 9 is a ninth characteristic diagram showing the measurement results of film density. [Figure 20] FIG. 10 is a characteristic diagram showing the relationship between stress and film density. DETAILED DESCRIPTION OF THE INVENTION

[0009] (First embodiment) <Film forming equipment> A film forming apparatus 1 according to a first embodiment for forming a silicon nitride film (SiN film) on a wafer substrate will be described with reference to Fig. 1. Fig. 1 is a longitudinal sectional side view of the film forming apparatus 1 of this example. This film forming apparatus 1 is configured as an apparatus for supplying a process gas containing a silicon-containing gas and a nitrogen-containing gas to the surface of a wafer and forming a SiN film by a plasma CVD method.

[0010] The SiN film formed by plasma is often required to have relatively high tensile stress and film density. Tensile stress is the stress that causes the wafer to bend into a bowl shape when placed on a horizontal surface, i.e., the periphery of the wafer is higher than the center. However, as will be described in detail using experimental data later, conventional film formation techniques have a trade-off between tensile stress and film density, where increasing one results in decreasing the other, making it difficult to meet this requirement. While CVD can be performed by heating the wafer at a relatively high temperature without using plasma, the miniaturization of semiconductor devices necessitates the deposition of SiN films at a relatively low temperature to minimize the thermal impact on the films formed on the wafer.

[0011] The film formation apparatus 1 of the present disclosure is configured to solve the above-mentioned trade-off problem and to form a SiN film with high tensile stress and film density and good film quality. To provide an overview of the film formation apparatus 1, the apparatus is configured to apply a first power having a first frequency higher than 300 MHz and a second power having a second frequency lower than the first frequency in a superimposed manner to form a capacitively coupled plasma in a processing chamber. In this manner, the processing gas is converted into plasma, and a SiN film is formed on the wafer. By performing this superimposed film formation and further adjusting the power, the above-mentioned trade-off problem is resolved. The configuration of each component of the film formation apparatus 1 will be described in detail below.

[0012] The film forming apparatus 1 includes a grounded, metallic, and substantially cylindrical processing chamber 10. A loading / unloading port 11 that can be freely opened and closed by a gate valve 12 is formed on the side of the processing chamber 10, for loading and unloading wafers between the processing chamber 10 and a vacuum transfer chamber (not shown). In addition, a vacuum exhaust section 14 including, for example, a pressure adjustment valve and a vacuum pump is connected to the bottom of the processing vessel 10 via an exhaust path 13, and is configured to be able to reduce the pressure inside the processing vessel 10 to a preset vacuum pressure.

[0013] A mounting table 21 for holding the wafer W substantially horizontally is provided within the processing vessel 10. In this embodiment, the mounting table 21 is grounded and serves as a lower electrode for generating plasma from the processing gas for the SiN film. The mounting table 21 is supported by support columns 22 and can be raised and lowered by an elevating mechanism 23 connected to the lower end of the support columns 22. The elevating mechanism 23 is disposed below the processing vessel 10, and a cover member 24 for keeping the interior of the processing vessel 10 airtight is provided between the bottom surface of the processing vessel 10 and the elevating mechanism 23.

[0014] A heater 25 is embedded in the mounting table 21, and is capable of heating the wafer W to a set temperature. In this example, the heating temperature of the wafer W is set to, for example, 320°C within a range of 250°C to 550°C. Furthermore, inside the processing vessel 10, elevating pins (not shown) are provided for transferring the wafer W between the mounting table 21 and an external transfer mechanism (not shown). Fig. 1 shows an example of the lower electrode (mounting table 21) to which a bias high frequency power supply that supplies bias high frequency power is not connected. Note that a bias high frequency power supply may be connected to the mounting table 21 via a matching box.

[0015] Furthermore, a flat, disk-shaped shower head 3 for supplying a film forming gas toward the wafer W is attached to the ceiling surface of the processing chamber 10 via an insulating member 34. The shower head 3 constitutes an upper electrode facing the mounting table 21. The shower head 3 has a diffusion space 31 formed therein for diffusing the processing gas, and a number of discharge holes 32 are distributed on the bottom surface of this diffusion space 31 for discharging the processing gas toward the wafer W. The inter-electrode distance, which is the distance between the upper surface of the mounting table 21 and the lower surface of the showerhead 3, is adjusted by raising and lowering the mounting table 21, and is set to a distance within the range of 60 mm to 120 mm, for example, 80 mm.

[0016] One end of a power feed rod 33 is connected to the upper surface of the showerhead 3. The other end of the power feed rod 33 is connected to a first high-frequency power supply 42 via a first matching box 41, and is also connected to a second high-frequency power supply 44 via a second matching box 43. In the example shown in Fig. 1, the first matching box 41 and the second matching box 43 are provided on the upper surface side of a cover member 35 that covers the upper surface of the processing chamber 10. The first high-frequency power supply 42 and the second high-frequency power supply 44 constitute a power application unit.

[0017] The first high frequency power supply 42 and the second high frequency power supply 44 are configured to supply high frequency power for plasma generation, each having a different frequency, to the showerhead 3. The high frequency power supplied from the first high frequency power supply 42 is, for example, power with a frequency in the UHF (Ultra High Frequency) band. In this specification, the UHF band refers to a frequency range higher than 300 MHz and equal to or lower than 3 GHz.

[0018] On the other hand, as described above, the second high-frequency power supply 44 supplies high-frequency power at a frequency lower than the frequency supplied by the first high-frequency power supply 42. This high-frequency power is, for example, power with a frequency in the VHF (Very High Frequency) band. In this specification, the VHF band refers to a frequency range of 30 MHz to 300 MHz. In the following description, the frequency and power of the first high-frequency power supply 42 will be referred to as the first frequency and first power, respectively, and the frequency and power of the second high-frequency power supply 44 will be referred to as the second frequency and second power, respectively.

[0019] As described above, the film forming apparatus 1 of the present disclosure is a parallel plate type plasma processing apparatus including the shower head 3 serving as an upper electrode and the mounting table 21 serving as a lower electrode. By supplying a processing gas to the shower head 3 while applying a first power and a second power in a superimposed manner, the two high frequency waves are superimposed and radiated into the processing chamber as a composite wave, thereby ionizing the processing gas and forming plasma.

[0020] To give a more specific example of the first frequency and the second frequency, the first frequency is 360 MHz and the second frequency is 180 MHz. Therefore, the first frequency is twice the second frequency. The reason why the first and second frequencies are set in this manner is that if the first frequency were a number other than n times the second frequency (n is an integer), the composite wave generated by superimposing the respective high-frequency waves and supplied into the processing vessel 10 would be distorted, which could result in large variations in the film density and film thickness distribution within the surface of the wafer W.

[0021] From the viewpoint of suppressing process variations within the wafer W, the first frequency need only be 2n times the second frequency, and is not limited to being twice (i.e., n = 1), but may be, for example, four times (i.e., n = 2). That is, if the first frequency is 360 MHz as described above, the second frequency may be 90 MHz. However, if the second frequency is too low, sufficient film density cannot be obtained, as shown in the experimental data described below. Furthermore, if the second frequency is set lower than the VHF band, the ion energy in the plasma becomes too high, which may result in a deterioration in the quality of the SiN film formed. Therefore, it is preferable to set the frequency to two times.

[0022] Therefore, as mentioned above, it is preferable to set the first frequency to twice the second frequency. In the following experiments, the second frequency was set to 180 MHz as described above, which solved the trade-off problem and achieved sufficient tensile stress and film density. Therefore, it is believed that a frequency slightly lower than 180 MHz, for example, a frequency of 100 MHz or higher, can ensure sufficient film quality. The preferred relationship and range of the first power and the second power will be explained later along with the explanation of the experiments.

[0023] The downstream end of a gas supply path 5 is connected to the diffusion space 31 of the showerhead 3. The upstream side of this gas supply path 5 is connected to a supply source 51 of a silicon-containing gas, such as monosilane (SiH4) gas, a supply source 52 of a nitrogen-containing gas, such as ammonia (NH3) gas, and a supply source 53 of a dilution gas, such as nitrogen (N2) gas, via respective supply flow paths 54, 55, and 56. Flow rate adjusters M1, M2, and M3 and valves V1, V2, and V3 are respectively provided in the supply flow paths 54, 55, and 56.

[0024] The mixed gas of SiH4 gas, NH3 gas, and N2 gas flows into the diffusion space 31 of the shower head 3 through the gas supply path 5, passes through the outlet holes 32, and is supplied into the processing chamber 10 as a processing gas. The processing gas supply unit in this example includes a SiH 4 gas supply source 51, an NH 3 gas supply source 52, supply flow paths 54 and 55, a gas supply path 5, and a shower head 3.

[0025] The film forming apparatus 1 having the above-described configuration includes a control unit 100. The control unit 100 is configured with a computer including a storage unit, a memory, and a CPU that stores a program. The program contains instructions (steps) for executing a SiN film forming process by outputting control signals from the control unit 100 to each unit of the film forming apparatus 1, and controlling the supply and cutoff of each gas and the supply of the first power and the second power. The program is stored in a storage unit of the computer, such as a compact disc, a hard disk, an MO (magneto-optical disc), or a non-volatile memory, and is read from the storage unit and installed in the control unit 100.

[0026] <Film formation process> The operation of the film forming apparatus 1 having the above-described configuration will now be briefly described. First, gate valve 12 is opened, and a transfer mechanism installed in a vacuum transfer chamber (not shown) transfers wafer W through transfer port 11, and the wafer W is placed on mounting table 21 via lift pins (not shown) (step of placing a substrate on a mounting table). Next, gate valve 12 is closed, and vacuum exhaust unit 14 evacuates processing chamber 10, adjusting the pressure inside processing chamber 10 to a preset pressure, for example, within a range of 3 Pa to 100 Pa. Furthermore, heater 25 heats wafer W to, for example, 320°C.

[0027] Thereafter, the supply of the processing gas is started (a step of supplying the processing gas into the processing chamber), and the supply of the first power and the second power is started from the first high frequency power supply 42 and the second high frequency power supply 44. In this way, the application of power is started simultaneously from the first and second high frequency power supplies 42, 44, and the first power and the second power are superimposed and applied to the shower head 3. Applying the first power and the second power simultaneously means that the supply of these powers to the shower head 3 is started simultaneously.

[0028] By the above-described operation, a capacitively coupled plasma is formed between the shower head 3 and the mounting table 21 in the processing chamber 10, and the supplied processing gas is converted into plasma. In this manner, a SiN film is formed on the surface of the wafer W by radicals and ions contained in the plasma (a film forming process for forming a silicon nitride film on a substrate). Then, by continuing film formation using the plasmatized processing gas for a preset period, a SiN film of a desired thickness is formed. Next, the supply of the first and second powers is terminated, and the supply of the processing gas is stopped. Thereafter, the wafer W is unloaded from the processing chamber 10 in the reverse order of the loading procedure, and the next wafer W is ready to be loaded.

[0029] The SiN film formed using the film forming apparatus 1 having the above-described configuration is used as a hard mask in, for example, an etching process for performing fine processing with a high aspect ratio, and is required to have high values ​​for both tensile stress and film density. Next, various experiments conducted until the trade-off relationship between tensile stress and film density was resolved, as well as various experiments conducted to find film formation conditions under which SiN films with good film quality having high values ​​for both tensile stress and film density, will be described together with data.

[0030] <Single frequency evaluation experiment> First, an evaluation experiment was conducted by changing the parameters when power was applied from one high-frequency power source to the showerhead 3, i.e., when high-frequency power of a single frequency was applied. The parameters evaluated were "distance between electrodes," "power," "pressure inside the processing vessel (hereinafter sometimes referred to as "pressure")," and "film formation rate."

[0031] These evaluations were performed using the film formation apparatus 1 shown in Fig. 1, with power being applied to the shower head 3 only from the first high frequency power supply 42. The basic film formation conditions (reference conditions) before adjusting the parameters were as follows. Frequency: 220MHz Power: 2700W Distance between electrodes: 80mm Pressure: 80 Pa (600 mTorr) Film formation rate: 84nm / min Mounting table temperature: 320℃

[0032] <Evaluation experiment 1: electrode distance> The SiN film deposition process was performed by changing the electrode distance, and the stress (film stress) and film density of the resulting SiN film were measured. Parameters other than the electrode distance were kept the same as the reference conditions. Stress was measured by obtaining the change in warp (radius of curvature) using a laser scan with a wavelength of 680 nm, and film density was measured using X-ray diffraction.

[0033] The measurement results of stress and film density are shown in Figures 2 and 3, respectively. In Figures 2 and 3, the horizontal axis represents the refractive index at a given film thickness. In Figure 2, the vertical axis represents stress. The greater the positive stress, the greater the tensile strength, and the greater the bowl-shaped warpage. The greater the negative stress, the greater the compressive strength, and the greater the inverted bowl-shaped warpage. Therefore, the above-mentioned tensile stress is the stress that takes on a positive value in the graph. In Figure 3, the vertical axis represents film density. In the graphs of Figures 2 and 3, the given refractive index and given film density are represented as X and Y, respectively. Therefore, X and Y are positive numbers. Note that the following characteristic diagrams of the measurement results of stress and film density also use the same vertical and horizontal axes as Figures 2 and 3, respectively. Hereinafter, the refractive index at a given film thickness will be referred to simply as the refractive index. The refractive index X is, for example, in the range of X=1.95 to 2.05, and the film density Y is, for example, in the range of Y=2.70 to 2.85. In Figures 2 and 3, data for an inter-electrode distance of 60 mm is plotted with ◇, and data for an inter-electrode distance of 80 mm is plotted with ○.

[0034] In Figure 2, an approximate line, expressed as a linear function, is drawn from each plot for an 80 mm interelectrode distance. As shown, the approximate line indicates that stress increases as the refractive index decreases. When comparing points on the approximate line for the same refractive index with the plot for a 60 mm interelectrode distance, the points on the approximate line indicate higher stress. In Figure 3, the dashed predicted line represents the empirical relationship between film density and refractive index based on the plot for an 80 mm interelectrode distance. As shown, this predicted line indicates that film density increases as the refractive index increases until a predetermined value is reached. When comparing points on the predicted line for the same refractive index with the plot for a 60 mm interelectrode distance, the plot for a 60 mm interelectrode distance indicates a higher film density. This result demonstrates that adjusting the interelectrode distance does not increase both tensile stress and film density; there is a trade-off between them.

[0035] 4 to 9, which show the results of Evaluation Experiments 2 to 4 described later, show the approximated straight lines obtained from the plots as solid lines and the predicted lines as dashed lines, just as in Figures 2 and 3. We will then describe the results of a comparison of stress and film density at the same or approximately the same refractive index, which can be inferred from the plots of data obtained directly from the experiments and the points on the approximated straight lines or the points on the predicted lines.

[0036] <Evaluation Experiment 2: Electricity> The SiN film deposition process was performed by changing the power of the high frequency power supply applied to the showerhead 3, and the obtained SiN film was evaluated in the same manner as in Evaluation Experiment 1. Parameters other than the power were the same as the reference conditions. The measurement results of stress and film density are shown in Figures 4 and 5. In these figures, data for a power of 2900 W is plotted with △, data for 2700 W with ○, and data for 2000 W with ◇.

[0037] As shown in Figure 4, the stress measurement value decreases as the power increases, so a lower power is preferable, and it was also found that the stress increases as the refractive index decreases under the same conditions. On the other hand, as shown in Figure 5, the film density measurement value increases as the power increases, so a higher power is preferable. Thus, it was found that even when the power was adjusted, there was a trade-off between tensile stress and film density.

[0038] <Evaluation Experiment 3: Pressure> The pressure inside the processing chamber 10 was changed to perform the SiN film deposition process, and the obtained SiN films were evaluated in the same manner as in Evaluation Experiment 1. The parameters other than the pressure were the same as the reference conditions. The measurement results of stress and film density are shown in Figures 6 and 7, respectively. In these figures, data at a pressure of 133 Pa (1000 mTorr) is plotted as △, data at 80 Pa (600 mTorr) as ○, and data at 40 Pa (300 mTorr) as ◇.

[0039] As shown in Figure 6, the higher the pressure, the greater the measured stress, so a higher pressure is preferable, and it was also found that the smaller the refractive index under the same conditions, the greater the stress.On the other hand, as shown in Figure 7, the lower the pressure, the greater the measured film density, so a lower pressure is preferable. Thus, it was found that even when the pressure was adjusted, there was a trade-off between the tensile stress and the film density.

[0040] <Evaluation experiment 4: Film deposition rate> The deposition rate was changed to perform the SiN film deposition process, and the obtained SiN films were evaluated in the same manner as in Evaluation Experiment 1. The deposition rate was adjusted by the flow rate of the processing gas, and the parameters other than the deposition rate were the same as those in the reference conditions. The measurement results of stress and film density are shown in Figures 8 and 9, respectively. In these figures, data for a film formation rate of 117 nm / min is plotted with △, data for 84 nm / min with ○, and data for 40 nm / min with ◇.

[0041] As shown in Figure 8, a higher deposition rate is preferable because negative stress increases at 40 nm / min. Furthermore, measurements at 84 nm / min and 117 nm / min indicate that the smaller the refractive index, the greater the stress. On the other hand, as shown in Figure 9, the lower the deposition rate, the greater the measured film density, so a lower deposition rate is preferable. Thus, it was found that even when the film formation rate, i.e., the flow rate of the processing gas, was adjusted, there was a trade-off between the tensile stress and the film density.

[0042] From the above experimental data, it can be seen that under conditions where the ion energy in the plasma supplied to the wafer W is low, such as an increase in the distance between the electrodes, a decrease in the high frequency power, an increase in the pressure inside the processing chamber, and an increase in the film formation rate, the stress of the SiN film increases, the tensile stress value increases, and the film density decreases. In contrast, under the conditions of high ion energy mentioned above, such as a reduced electrode distance, increased high-frequency power, reduced pressure inside the processing chamber, and a reduced film formation rate, the stress on the SiN film decreases, it becomes more compressive, and the film density increases.

[0043] The reason why the properties of the formed film change depending on the difference in ion energy is presumed to be as follows. Under conditions of high ion energy, the severing of bonds within each molecule in SiH4 gas and NH3 gas proceeds sufficiently. The severing of bonds within the SiH4 molecule causes one or more H atoms to be desorbed from the SiH4 molecule, resulting in the plasma containing a relatively large number of activated species consisting of only Si atoms or Si atoms bonded with H atoms. Therefore, polymerization of Si atoms occurs relatively easily on the wafer W, and the formed SiN film contains a relatively large number of bonds between Si atoms. The relatively long bond length of these bonds is thought to be a factor in promoting compressive formation.

[0044] Furthermore, as mentioned above, the bonds between N and H atoms in NH3 gas are severed sufficiently. However, if these bonds between N and H atoms exist in the film, it is believed that they can cause tensile stress. For these reasons, when film formation is performed under high ion energy conditions, tensile stress is relatively low. Furthermore, when two Si atoms bonded to each other are compared with two Si atoms and an N atom bonded to each other, two Si atoms bonded to each other are heavier. As mentioned above, since the film contains many bonds between Si atoms, the film density of the SiN film is relatively high. Furthermore, the broken H recombines and becomes H2, which is exhausted. This reduction in the H element in the film is thought to increase the Si-Si and Si-N content ratios, which also contributes to the high film density.

[0045] On the other hand, under conditions where the ion energy of the plasma supplied to the wafer W is low, breaking of bonds within the molecules of the SiH gas and NH gas is suppressed, resulting in a relatively high tensile stress and a relatively low film density, in contrast to film formation under conditions where the ion energy is high. In relation to the refractive index, the lower the refractive index, the more bonds there are between N atoms and H atoms in the SiN film, which increases the tensile stress and reduces the film density.

[0046] <Evaluation Experiment 5: Frequency Dependence> Next, a SiN film was formed by changing the frequency of the applied high frequency and the pressure inside the processing chamber, and the obtained SiN film was evaluated in the same manner as in Evaluation Experiment 1. Parameters other than the frequency and pressure were the same as the reference conditions. The measurement results of stress and film density are shown in Figures 10 and 11, respectively. In these figures, the data of frequency (MHz) and pressure (Pa (mTorr)) are plotted using the following symbols. ■: 120MHz, 16Pa (120mTorr) □: 120MHz, 64Pa (480mTorr) or more ▲: 180MHz, 16Pa (120mTorr) △: 180MHz, 64Pa (480mTorr) or more ●:220MHz, 16Pa (120mTorr) ○: 220MHz, 64Pa (480mTorr) or more

[0047] As shown in Figure 10, at pressures of 64 Pa or higher, the change in the measured stress value is small even when the frequency is different, which means that the stress has little frequency dependency. On the other hand, as shown in Figure 11, the measured film density tends to increase as the frequency increases. In other words, this experiment showed that when high frequency waves in the range of 120 MHz to 220 MHz are applied, the higher the frequency, the higher the film density becomes.

[0048] Although not shown in the graph, a similar evaluation was also performed when high-frequency powers higher than 220 MHz were applied. As a result, even when high-frequency powers of 860 MHz (UHF band) or microwave bands were applied, the film density of the deposited SiN film was comparable to that when high-frequency power of 220 MHz was applied. Given these experimental results and the tendency for equipment configurations with higher applied high-frequency frequencies to increase manufacturing and operating costs, it was confirmed that setting the frequency significantly higher than 220 MHz to increase film density is not advisable. As with Evaluation Experiments 1 to 4, a trade-off between tensile stress and film density was observed when deposition was performed using high-frequency power of 860 MHz.

[0049] <Evaluation Experiment 6> Considering the results of Evaluation Experiment 5, a similar evaluation was attempted by changing the deposition rate when applying a high frequency wave in the UHF band, which is 220 MHz or higher. Specifically, a high frequency wave with a frequency of 360 MHz and a power of 2700 W was applied to perform the deposition process, and the resulting SiN film was evaluated in the same way as in Evaluation Experiment 1. The deposition rate was set to a higher rate of 64 nm / min (high DR: 76 nm / min), with the other parameters being the same as the reference conditions. Note that the "high deposition rate" in the experimental data below is approximately the same rate as in Evaluation Experiment 6.

[0050] The measurement results of stress and film density are shown in Figures 12 and 13, respectively. In these figures, data for the reference deposition rate condition is indicated by a circle, and data for high DR is indicated by a diamond. In the graphs of Figures 12 and 13, approximate lines are drawn from the data for the reference condition. Each approximate line follows the relationship between stress and refractive index and the relationship between film density and refractive index obtained from previous empirical rules. To facilitate comparison, predicted lines showing the relationship between stress and refractive index obtained from the high DR data and the relationship between film density and refractive index are shown as dashed lines in Figures 12 and 13, respectively, according to empirical rules. Because the predicted lines are drawn based on empirical rules, each predicted line is parallel to the approximate line. In the graphs of Figures 12 and 13, when comparing the points on the approximation line and the points on the predicted line at the same refractive index, the high DR is higher than the reference in terms of stress and lower in terms of film density. In other words, it was confirmed that there is still a trade-off between tensile stress and film density.

[0051] <Summary of the evaluation experiment> From the above single frequency experimental data, it was found that even if the parameters of "electrode distance", "power", "pressure inside the processing chamber", and "film formation rate" were adjusted, the trade-off relationship between the tensile stress of the SiN film and the film density could not be resolved. The stress has little frequency dependency, but the film density tends to increase as the frequency increases, and is at the same level at frequencies above 220 MHz. It was confirmed that even when the frequency was increased to 360 MHz, there was a trade-off between the tensile stress of the SiN film and the film density.

[0052] <Superposition of First Frequency and Second Frequency: Example 1> In light of these circumstances, it was determined that it would be difficult to improve both the tensile stress and film density of the SiN film using a single frequency, and evaluation was carried out focusing on applying high frequencies of different frequencies in a superimposed manner. Specifically, the first power and the second power were simultaneously applied from the first high-frequency power supply 42 and the second high-frequency power supply 44, respectively, to perform a film formation process at a high film formation rate (79.2 nm / min) equivalent to that in single-frequency evaluation experiment 6 (76.3 nm / min) (Example 1), and the obtained SiN film was evaluated in the same manner as in evaluation experiment 1.

[0053] The first frequency was set to 360 MHz in the UHF band, which is a higher frequency than 220 MHz, at which the film density was good in the evaluation experiment. The second frequency was set to 180 MHz, which is half the first frequency. The first power and second power were as follows, and the total power of the first power and second power was set to the same as in Comparative Examples 1 and 2 described below. The other parameters were set to the same as the reference conditions. First high frequency: First frequency 360MHz, First power 1350W Second high frequency: Second frequency 180MHz, Second power 1350W

[0054] Similar evaluations were also carried out when a high frequency wave of a single frequency was applied as Comparative Examples 1 and 2. The respective conditions were as follows. Comparative Example 1: Frequency 360MHz, Power 2700W Comparative Example 2: Frequency 360 MHz, power 2700 W, high deposition rate

[0055] The measurement results of stress and film density are shown in Figures 14 and 15, respectively. Data are plotted as △ for Example 1, ○ for Comparative Example 1, and ◇ for Comparative Example 2. In these figures, an approximated line shown as a solid line was created using a linear function based on multiple data from Comparative Example 1. Similar to Evaluation Experiment 6, predicted lines are also plotted in each graph of Figures 14 and 15. The dashed line is the predicted line obtained from the data from Example 1, and the dashed-dotted line is the predicted line obtained from the data from Comparative Example 2. However, the predicted line for Comparative Example 2 is not shown in Figure 15. The predicted line for Comparative Example 1 was based on empirical rules. Similar to Evaluation Experiment 6, the predicted lines for Example 1 and Comparative Example 2 were drawn parallel to the predicted line for Comparative Example 1 based on empirical rules.

[0056] As shown in FIG. 14, when comparing stress at the same refractive index, Example 1 has a higher stress than Comparative Example 1 and Comparative Example 2. When comparing Example 1 and Comparative Example 2, which are film formation conditions with a high film formation rate (high DR), Example 1 has a higher stress than Comparative Example 2. Furthermore, as shown in FIG. 15, when comparing film densities at the same refractive index, Example 1 and Comparative Example 2 have lower stress than Comparative Example 1, and are almost the same between Example 1 and Comparative Example 2, which are high film formation rates. More specifically, the plot of the data for Comparative Example 2 is located tangent to the predicted line for Example 1 in FIG. 15. As described above, the total power supplied in Example 1 and Comparative Example 2, which are high DR, is the same. With the film formation conditions aligned in this way, by changing from applying power at a single frequency (Comparative Example 2) to applying power at two different frequencies (Example 1), the tensile stress increased and the film density was maintained without decreasing. Therefore, as described in the section on film formation apparatus 1, it was found that by superimposing and applying a first power having a first frequency and a second power having a second frequency lower than the first frequency, it is possible to break away from the trade-off relationship between tensile stress and film density.

[0057] <Superposition of First Frequency and Second Frequency: Examples 2 to 4> 15, Example 1 has a smaller film density than Comparative Example 1, and therefore Examples 2 to 4 were evaluated to find conditions for improving the film density. Examples 2 to 4 are all examples in which the first and second powers were increased compared to Example 1, and the respective conditions are as follows. Note that the parameters other than the first and second powers were the same as those of Example 1. Example 2 First frequency: 360MHz, First power: 2700W Second frequency 180MHz, second power 1400W Example 3 First frequency: 360MHz, First power: 2700W Second frequency 180MHz, second power 2700W Example 4 First frequency: 360MHz, First power: 2900W Second frequency 180MHz, second power 2900W

[0058] In this way, when the first frequency is 360 MHz and the second frequency is 180 MHz, the first power is set to be equal to or greater than the second power. This is because, in Evaluation Experiment 5, in the frequency range of 120 MHz to 220 MHz, the higher the frequency, the higher the film density. In other words, if a first frequency of 360 MHz and a second frequency of 180 MHz are applied separately, the film density will be higher when the first frequency of 360 MHz is applied. In order to reliably obtain the effect of the first frequency that obtains such a high film density, the first power and the second power have the above relationship.

[0059] The measurement results of stress and film density are shown in Figures 16 and 17, respectively. In these figures, data is plotted as triangles for Example 2, squares for Example 3, and ▽s for Example 4. The diamonds in these figures represent the data for Comparative Example 2. In Figures 16 and 17, the approximate line based on the data for Example 2 is shown as a solid line, the approximate line based on the data for Example 3 as a dashed line, and the predicted line based on the data for Example 4 as a dashed-dotted line. Each approximate line was calculated using a linear function based on the respective data. In Example 4, there is only one plot, so a line parallel to this approximate line is drawn as the predicted line.

[0060] 16, it was found that, with regard to stress, Examples 2 and 3 had values ​​that were roughly the same as or slightly smaller than Comparative Example 2, while Example 4 had a lower value than Comparative Example 2. Furthermore, between Examples 2 and 3, Example 3 had a slightly higher value. 17, it was confirmed that the film density of Examples 2 to 4 was higher than that of Comparative Example 2. Furthermore, the film density of Examples 2 and 3 was higher than that of Example 4, and between Examples 2 and 3, Example 3 was slightly higher.

[0061] As described above, in Examples 2 and 3 and Comparative Example 2, which have high DR, a 360 MHz high frequency was supplied at 2700 W. By supplying a single frequency under such uniform film formation conditions (Comparative Example 2), and then superimposing the 180 MHz frequency at 1400 W (Example 2) and 2700 W (Example 3), the film density increased while the tensile stress remained constant. While Example 4 shows a trade-off between tensile stress and film density, Examples 2 and 3 eliminate this trade-off. Therefore, the experimental results show that by optimizing the relationship between the first power and the second power, the trade-off can be eliminated by superimposing and applying the first power at the first frequency and the second power at the second frequency. As one of the appropriate relationships, it can be seen from Examples 2 and 3 that when the first power is set to 2700W, the second power should be set to 1400W to 2700W.

[0062] <Summary of Examples> From the experimental data of Examples 1 to 4, it was confirmed that the above-mentioned trade-off problem can be resolved by superimposing and applying a first power at a first frequency in the UHF band and a second power at a second frequency in the VHF band. That is, it was confirmed that high values ​​can be secured for both the tensile stress and the film density of the SiN film.

[0063] To achieve a high film density, the first frequency may be higher than the UHF band. However, as explained in Evaluation Test 5, increasing the frequency beyond a certain level results in little increase in film density. Therefore, it is preferable to use a UHF band frequency as the first frequency. Furthermore, it is preferable to set the second frequency to a relatively high frequency in order to achieve a high film density and to prevent degradation of film quality due to ion energy, as described in the film formation apparatus section. In Examples 1 to 4 described above, setting the frequency in the VHF band eliminated the trade-off, so it is preferable to set the frequency in the VHF band. However, this is not limited to setting the frequency in the VHF band. For example, if the first frequency is set to a relatively high frequency in the UHF band, the second frequency may also be set in the UHF band.

[0064] As for the first and second powers, the first power is set to be equal to or greater than the second power (the second power is set to be equal to or less than the first power) in order to increase the film density, as described in Examples 2 to 4. In the examples, it has been shown that the trade-off relationship can be resolved by setting the second power / first power ratio to 1 / 2 to 1, and it is preferable to have such a relationship. Furthermore, for the first power and the second power, the trade-off relationship could not be resolved when both were set to 2900 W in Example 4, but the trade-off was resolved when they were set to 1350 W, 1400 W, and 2700 W in Examples 1 to 3. Therefore, it is preferable that the first power and the second power be set to a value greater than or equal to 1350 W and less than 2900 W, and more preferably greater than or equal to 1350 W and less than or equal to 2700 W. It is even more preferable that the second power be set to less than 2700 W. This will be explained in later experiments.

[0065] In the above description, "superimposed application of the first power and the second power to the showerhead 3" refers to both the first power and the second power being applied to the showerhead 3. This includes cases where the application of power from the first and second high-frequency power sources 42 and 44 starts simultaneously, as well as cases where the application of the first power and the application of the second power overlap even if the application start timings are different. However, if the time difference between the start timings of the application is too long, the film density of the SiN film deposited when both the first power and the second power are not being applied may decrease, and the composition of the SiN film in the depth direction may change. For this reason, it is preferable that the application of the first and second powers starts simultaneously.

[0066] <Effects> The above-described film deposition apparatus 1 has the following advantages: By applying a first power having a first frequency higher than 300 MHz and a second power having a second frequency lower than the first frequency in a superimposed manner, the process gas is converted into plasma, and therefore a SiN film with good film quality that achieves both tensile stress and film density can be deposited.

[0067] <Second embodiment> Next, a second embodiment of the present disclosure will be described. When applying superimposed high-frequency powers of different frequencies, higher values ​​may be required for the stress and film density of the SiN film. For this reason, this embodiment adjusts parameters to further improve the tensile stress and film density of the SiN film.

[0068] The film forming apparatus of this embodiment is configured to be able to supply hydrogen gas (H2 gas) to the film forming apparatus 1 of the first embodiment, and H2 gas is supplied to the shower head 3 together with the processing gas, and the film forming process of the SiN film is performed in the same manner as in the first embodiment. At this time, the stress and film density of the SiN film are improved by adjusting the parameters of "pressure inside the processing vessel," "hydrogen gas flow rate," "distance between electrodes," and "second power."

[0069] The following is a description based on experimental data. The film formation conditions (Example 10) before parameter adjustment are as follows: Pressure inside the processing chamber: 80 Pa (600 mTorr) SiH4 gas flow rate: 85sccm NH3 gas flow rate: 132 sccm H2 gas flow rate: 0 sccm Ar gas flow rate: 1500sccm Distance between electrodes: 80mm First frequency: 360MHz Primary power: 2700W Second frequency: 180MHz Secondary power: 2700W

[0070] The stress and film density of the SiN film formed under the film forming conditions of Example 10 are shown in FIGS. 18 and 19, respectively, and the data are plotted with ♦ in the figures. 18 and 19 show data obtained when each of the parameters "pressure," "hydrogen gas flow rate," "electrode distance," "N gas / NH gas flow rate ratio," and "Ar gas flow rate" was adjusted one by one from the film formation conditions of Example 10. These data were obtained by forming a SiN film under the conditions of Example 10, with only one corresponding parameter changed.

[0071] As a result, it was found that the stress was equivalent to that of Example 10 at a refractive index lower than that of Example 10, and that although there were conditions favorable for stress, negative values ​​were observed under many conditions. Furthermore, the film density was equivalent to or lower than that of Example 10. From these findings, it was found that significant improvements in stress and film density cannot be expected by adjusting each parameter individually.

[0072] Therefore, an attempt was made to simultaneously adjust four parameters: "pressure," "hydrogen gas flow rate," "electrode distance," and "second power." The film formation conditions (Example 11) after this parameter adjustment are as follows: Pressure inside the processing chamber: 86 Pa (650 mTorr) SiH4 gas flow rate: 85sccm NH3 gas flow rate: 129sccm H2 gas flow rate: 125sccm Ar gas flow rate: 1492 sccm Interelectrode distance: 70mm First frequency: 360MHz Primary power: 2700W Second frequency: 180MHz Secondary power: 2500W

[0073] Thus, in Example 11, the conditions are such that the "pressure" is increased, the "hydrogen gas flow rate" is increased, the "electrode distance" is decreased, and the "second power" is reduced compared to Example 10. Note that, although the "NH3 gas flow rate" and the "Ar gas flow rate" are slightly different between Example 10 and Example 11, the amount of change does not affect the stress or film density of the SiN film.

[0074] The stress and film density of the SiN film formed under the film forming conditions of Example 11 are shown in Figure 20, with the data plotted by ●. In Figure 20, the vertical axis is film density and the horizontal axis is tensile stress (expressed as stress). In the same figure, the data for Example 10 is plotted as ■, the data for Example 12 as ○, the data for Example 13 as □, and the data for Comparative Example 10 as ◇. The conditions for Example 12, Example 13, and Comparative Example 10 are as follows: Example 12: An example in which only the parameter "hydrogen gas flow rate" from Example 10 was changed to 199 sccm. Example 13: Example in which only the parameter "second power" from Example 10 was changed to 1400 W. Comparative Example 10: A comparative example in which the application of the second power from the second high-frequency power supply in Example 10 was stopped, and only a single frequency of a first frequency of 360 MHz and a first power of 2700 W was applied.

[0075] As a result, it was found that the film quality of Examples 10 and 13 was improved compared to Comparative Example 10, and by comparing Examples 10 and 13, it was found that the measured values ​​of stress and film density further increased in Example 10, in which the second power was increased from 1400 W to 2700 W. In Example 12, in which only the parameter "hydrogen gas flow rate" was adjusted, the stress was 675 MPa and the film density was 2.88 g / cm 3 This shows a further improvement compared to Example 10. H present in the SiN film becomes an impurity, but it is presumed that the addition of hydrogen gas causes hydrogen radicals and ions contained in the plasma to desorb H from the SiN film, which effectively improves the film quality.

[0076] Furthermore, in Example 11, in which the four parameters from Example 10 were simultaneously adjusted, the stress and film density were significantly improved compared to Examples 10 and 12, and the film density was 2.88 g / cm 3 or more, and the stress is 700 MPa or more, and the stress is 811 MPa and the film density is 2.89 g / cm 3 It was confirmed that a SiN film of this size could be formed.

[0077] In Example 11, the "pressure" is increased, the "hydrogen gas flow rate" is increased compared to Example 10, the "electrode distance" is decreased, and the "second power" is reduced, and parameter adjustments that decrease the ion energy in the plasma (pressure, second power) are combined with parameter adjustments that increase it (electrode distance). As a result, it is presumed that the interaction of the effects of these parameters allowed high values ​​to be obtained for stress and film density, resulting in a significant improvement in the film quality of the SiN film.

[0078] <Effects of the second embodiment> In Example 11, the "pressure" is set to a value greater than 80 Pa, the "hydrogen gas flow rate" is set to a value less than 199 sccm, the "electrode distance" is set to a value less than 80 mm, and the "second power" is set to a value less than 2700 W. By setting the parameters in this manner, it is possible to form a SiN film having particularly excellent properties in terms of stress and film density.

[0079] <Other embodiments> The film forming apparatus 1 for forming a SiN film is not limited to the configuration shown in FIG. 1, and instead of providing a first high-frequency power supply and a second high-frequency power supply, it may be configured to supply power of two different frequencies from a common high-frequency power supply. The composition of the process gas supplied to the process chamber 10 to form the SiN film is not limited to the above-described example. For example, the silicon-containing gas may be disilane (SiH) or a higher silane gas containing three or more silicon atoms in a molecule, in addition to SiH gas, or an organic silane gas such as an aminosilane gas.

[0080] In the above embodiment, NH gas is described as the nitrogen-containing gas, but N gas used as a dilution gas can also be converted into plasma and serve as a nitrogen source for the SiN film. In other words, N gas serves both as a carrier gas and a nitrogen-containing gas. In addition to N gas and NH gas, trimethylamine and triethylamine can also be used as nitrogen-containing gases. Furthermore, in addition to N2 gas, inert gases such as argon (Ar) gas and helium (He) gas can be used as dilution gases. Furthermore, the silicon nitride film formed by this technology is not limited to use as a hard mask, but can also be used as a barrier film to prevent the diffusion of wiring metals and as an interlayer insulating film.

[0081] The embodiments disclosed herein should be considered to be illustrative in all respects and not restrictive, and the above-described embodiments may be omitted, substituted, or modified in various ways without departing from the scope and spirit of the appended claims. [Explanation of symbols]

[0082] 1 Film deposition equipment 10 Processing container 21 Mounting table 3. Shower head (electrode) 42 First high frequency power source 44 Second high frequency power supply

Claims

1. placing a substrate on a stage provided in a processing chamber; supplying a process gas containing a silicon-containing gas and a nitrogen-containing gas into the process chamber; a film formation step of applying a first power having a first frequency higher than 300 MHz and a second power having a second frequency lower than the first frequency to an electrode facing the mounting table, thereby generating plasma from the processing gas in the processing chamber and forming a silicon nitride film on the substrate; Equipped with the second frequency is a frequency of 100 MHz or higher in the VHF band, The film forming method, wherein the first frequency is 2n times (n is an integer) the second frequency.

2. 2. The film forming method according to claim 1, wherein the 2n times is 2 times.

3. 3. The film forming method according to claim 2, wherein the first frequency is 360 MHz and the second frequency is 180 MHz.

4. 4. The film forming method according to claim 1, wherein the film forming step includes a step of simultaneously starting application of the first power and the second power to the electrode to form a film.

5. 5. The film forming method according to claim 1, wherein the second power is equal to or less than the first power.

6. 6. The film forming apparatus according to claim 5, wherein the first power and the second power are equal to or greater than 1350 W and less than 2900 W.

7. 7. The film forming method according to claim 1, wherein the second power is lower than 2700 W.

8. 8. The film forming method according to claim 1, wherein the pressure inside the processing chamber is greater than 80 Pa in the film forming step.

9. 9. The film forming method according to claim 1, wherein the processing gas contains hydrogen gas.

10. 10. The film forming method according to claim 9, wherein the hydrogen gas is supplied into the processing chamber at a flow rate lower than 199 sccm in the film forming step.

11. 11. The film forming method according to claim 1, wherein in the film forming step, the distance between the electrode and the stage is less than 80 mm.

12. 12. The film formation method according to claim 1, wherein the processing gas contains hydrogen gas, the second power is lower than 2700 W, the pressure in the processing chamber is higher than 80 Pa, the hydrogen gas is supplied into the processing chamber at a flow rate lower than 199 sccm, and a distance between the electrode and the mounting table is shorter than 80 mm, in the film formation step.

13. The film forming process is carried out with a film density of 2.88 g / cm ３ 13. The film forming method according to claim 1, wherein the silicon nitride film has a stress of 700 MPa or more.

14. a processing vessel having a stage therein on which a substrate is placed; a processing gas supply unit that supplies a processing gas containing a silicon-containing gas and a nitrogen-containing gas into the processing chamber; an electrode facing the mounting table; a power applying unit that applies a first power having a first frequency higher than 300 MHz and a second power having a second frequency lower than the first frequency to the electrode in order to form a silicon nitride film on the substrate by converting the processing gas in the processing chamber into plasma; Equipped with the second frequency is a frequency of 100 MHz or higher in the VHF band, The first frequency is 2n times (n is an integer) the second frequency.

Images