JPS62108525A - Method and apparatus for surface treating - Google Patents

Abstract

PURPOSE:To prevent a surface to be treated from damaging, contaminating, rising at its temperature by controlling incident amount and direction to the surface with active particles as electrically neutral particles, and selecting particle seeds mainly reacted with the surface. CONSTITUTION:Active particles formed of an active particle generator 1 become electrically neutral by a neutralizer 2, are aligned in the directions by a direction controller 3, then incident to a wafer 4, and the wafer 4 is held through a temperature controller 5 at a sample base 6. A raw material gas is fed through a pipe 9 to a chamber 8, DC or high frequency voltage is applied to electrodes 10, 10' to generate a discharge. The type and the quantity of the active particles are controlled by the generating means. When the discharge gas type is varied and the type of the active particles, the pressure of the discharge gas and the incident power are altered, the quantity of the active particles can be changed. Thus, the surface of the sample can be accurately treated without damaging, contaminating rising of the temperature.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Application of the Invention] The present invention relates to a surface treatment method and apparatus for treating a solid surface, and particularly to a non-damaging, non-contaminating, highly selective and low-temperature process suitable for manufacturing semiconductor devices. The present invention relates to a surface treatment method and an apparatus therefor.

[Background of the invention]

Conventional surface treatment methods, especially dry processes in semiconductor device manufacturing, have used ion beams or plasma (edited by Takuo Kanno [Semiconductor Plasma Process Technology])
, Sangyo Tosho, 1980) However, in this method, high kinetic energy (approximately 100 eV or more ) ions, atomic two-molecule electrons, etc. are incident, and damage and contamination of the element surface are inevitably formed. Furthermore, the temperature of the element surface increases due to the incidence of such high-energy particles. As the manufacturing precision of semiconductor devices decreases (below 1 μm), such damage, contamination, and temperature rise become serious problems. In addition, such damage, contamination, and temperature rise are fatal to semiconductor devices with three-dimensional structures, which are expected to be realized in the near future.
Conventional dry process technology/equipment becomes unusable6
[Object of the invention] The object of the invention is to treat damage and contamination of the surface to be treated.

It is an object of the present invention to provide a surface treatment method that causes little temperature rise and an apparatus therefor.

[Summary of the invention]

From about 100 eV onwards, dry processes are actively used in the field of surface treatment (particularly in the field of semiconductor device manufacturing). Here, the term "dry process" refers to the previous wet process (method using a solution), and refers to a method in which the sample surface is treated in a vacuum or gas phase. Alternatively, an ion beam is used, and the kinetic energy of electrons, ions, and ion beams incident on the sample surface is several sV to 10'
eV (Figure 9), while the atomic displacement energy in the crystal (Displacem
ent energy (the energy required to displace an atom from its normal crystal lattice position), Ed is approximately 10 eV (for example, Ed = 12.9 eV in a Si crystal).

Therefore, in the conventional dry process, damage was irreversibly formed on the sample surface. In addition, the sample temperature increases due to the incidence of such high-energy particles, and in some cases, the temperature of the sample increases
The temperature has reached 00℃ or higher. Such an increase in sample temperature may cause crystal defects (damage) to be formed on the sample surface or extremely limit the range of application of the process. Furthermore, when such high-energy particles enter a surface near the sample (for example, the vacuum chamber wall or the sample stage surface), they physically and chemically sputter the substances that make up this surface, causing them to re-adhere to the sample surface. Caused surface contamination. The damage, contamination, and temperature rise described above are fatal in surface treatment technology (particularly in semiconductor device manufacturing technology).

The essential things that must be done in the dry process are removing substances on the sample surface (etching) and depositing substances on the sample surface (deposition, epitaxy) using physical and chemical reactions on the sample surface or in the gas phase. , modifying the sample surface substance (oxidation.

In this case, chemical reactions rather than physical reactions take center stage. This is because chemical reactions can achieve a greater variety of reactions.

As shown in Figure 9, the energy (
Chemical bond energy) is distributed within the range of 0.1 to 10 eV, and the energy that generates crystal defects is approximately 10 eV.
It is eV. Therefore, the optimum energy to use for carrying out the dry process is in the range of 0.1 to 10 eV. This allows dry processes to be carried out without damage, contamination, or temperature increases.

In order to supply energy of 0.1 to 10 eV, it is optimal to use electrically neutral active particles (atoms/molecules) rather than ions or ion bees 11. According to Figure 9, the incident energy of neutral particles (particles with no electric charge) on the surface to be processed is within the range of about 10-2 to 1 eV, so crystal defects do not occur when neutral particles are incident. , and the rotational, vibrational, and electronic excitation energies of atoms and molecules are also 10
It is within the range of eV or less. Therefore, it is optimal to use electrically neutral atoms and molecules that are in the ground state (a state that has only translational kinetic energy) or that are rotated, vibrated, or electronically excited.

In order to accurately perform surface treatment (especially semiconductor-related surface treatment) using these electrically neutral active particles (hereinafter simply referred to as "active particles" unless otherwise specified), the types of active particles must be selected. It is necessary to control the amount and direction of incidence on the surface to be processed.

[Embodiments of the invention]

Embodiments of the present invention will be described below with reference to the drawings.

<Example 1> The simplest example is shown in FIG. Active particle generator 1
The active particles (including charged particles) made in the above are made electrically neutral by the neutralizer 2, aligned in direction by the direction controller 3, and then incident on the wafer 4. Wafer 4 is temperature controller 5
It is held on the sample stage 6 via.

Below, active particle generator 1, neutralizer 2. The contents of the direction controller 3 will be described in detail.

Possible methods for generating active particles in the active particle generator 1 include electric discharge, light irradiation (ultraviolet, visible, infrared light, etc.), electron beam irradiation, heating (heater heating), ion beam irradiation, molecular beam irradiation, etc. . Among these, typical examples using discharge, light irradiation, electron beam irradiation, and heating are shown in FIG. Second
Figures (a) and (b) show a method for generating active particles using electrical discharge. Raw material gas is introduced into chamber 8 through pipe 9.

In Fig. 2(a), a discharge is generated by applying a DC or high frequency (about several tens of KHz to 100 MHz) voltage to the electrode 10.10'6 Fig. 2(b)
In this method, a magnetic field microwave discharge is used to generate the discharge.

The micro liquid generated by the magnetron 14 is guided through the waveguide 13 and introduced into the discharge tube 12 . At this time, by forming a magnetic field within the discharge tube by the solenoid coil 15,
Magnetic field microwave discharges are generated (in some cases, discharge is possible only by introducing microwaves, so the solenoid coil 15 is not essential), and the active particles created by these discharges enter the chamber 7 through the ejection hole 11. Inflow. FIGS. 2(c) and 2(d) show a method of generating active particles using light.

In FIG. 2(Q), a laser beam from a laser 19 is used as a light source, and in FIG. 2(d), a lamp 20, 20' is used as a light source. For laser light and lamp light, use a quartz window 18.18
' into the chamber 8 through. Especially in Figure 2 (
In c), consideration is given to multiply reflecting the laser beam using the reflecting mirrors 16 and 16' to lengthen the effective optical path length. In this embodiment, active particles are created by injecting light with a wavelength corresponding to the excitation and dissociation energy of atoms and molecules. FIG. 3(e) shows a method of generating active particles by irradiating the raw material gas introduced into the chamber 8 through the pipe 9 with electrons and an electron beam from the radiation source 21. It is desirable that the electron beam source 21 generates an electron beam with uniform energy. FIG. 2(f) shows a method of generating active particles by heating.

Above, we have described specific examples of how to generate active particles.
By using these generation means, it is possible to control the type and amount of active particles.

For example, in FIGS. 2(a) and 2(b), it is possible to change the type of active particles by changing the type of discharge gas, and the amount of active particles by changing the discharge gas pressure and incident power. Taking the example of aFx discharge, which is the main reactive species in Si etching, the neutral FyK particles have specific numerical values:
In FIG. 2(a), F in the chamber 8 generated by discharge is shown.
The atomic concentration is less than (1110"Qll-"), and in Fig. 2(b), the FZ discharge gas pressure is set to 1O-1Z~10Pa.
By changing up to 102~10"am-'
is obtained. Therefore, by changing the discharge gas pressure and the incident power, the amount of F atoms blown out from the ejection port 11 can also be changed (
(according to these concentrations).

In the case of FIG. 2(c) using light, it is possible to control the active ions using the characteristic of light, that is, energy monochromaticity. By selecting the wavelength of the laser beam or lamp light source, it is possible to select only specific bonds in a molecule or to excite atoms and molecules to a specific level. Furthermore, the amount of active particles can be controlled by changing the output of the laser beam or lamp light, or by changing the pressure of the raw material gas introduced into the chamber 8. For example, in Fig. 2(c), an ultraviolet laser beam with a wavelength of 2000 people and an output of 10 W/cd is
2 (assuming that the absorption cross section of Fz in 2000 people is 10''cd), the dissociation rate of F2 is 10%. Similarly, it is possible to control the type and amount of active particles.In particular, to give a specific example regarding the amount of active particles, Fz ~5
When introduced at less than Torr and heated to about 800° C., about 50% of F2 dissociates. By using the methods shown in FIG. 2 in the active particle generator 1, it becomes possible to control the type and amount of active particles.

Next, the contents of the neutralizer 2 will be described. As mentioned above, when ions are incident on the sample surface, damage occurs to the sample surface due to ion bombardment. In addition, low-velocity (10 eV or less) ions cause less damage, but insulators (S i Ox
, SosN+, etc.), the surface of the sample is charged, causing dielectric breakdown of the insulator (this problem is particularly serious, for example, in dry etching of the gate oxide film of a MOS).

Therefore, in order to solve the above problems, it is necessary to take care not to irradiate the sample surface with charged particles. Neutralizer 2
is installed for the purpose of removing and electrically neutralizing these charged particles.

A specific example of the neutralizer 2 is shown in FIG. Figure 3(a) shows “
Although this may seem strange when considered in terms of the meaning of "neutralization," it is included in the specific example because it is effective when considered from the purpose of the neutralizer 2, which is to remove charged particles. In FIG. 3(a), a plurality of grids 24 are provided close to the jet nozzle 11, and the potentials of these grids are independently controlled. For example, by setting the first stage grid to ground potential and the second stage to positive potential, positive ions are generated, and by setting the third stage to negative potential and the final stage grid to ground potential, negative ions and electrons are generated. can be removed respectively. As a result, the only particles passing through the grid 24 are (electrically neutral) active particles.

FIG. 3(b) shows a method in which thermionic electrons generated from the filament 27 are used to neutralize raw ions. This method cannot remove negative ions and electrons, so Figure 3 (
As in a), a plurality of grids 30 are installed at the rear. This removes negative ions and electrons. Third
The embodiment of the type shown in FIG. 3(b) is considered to be most effective when the structure shown in FIG. 3(c) is used.

For the sake of explanation, the active particle generator 1 (the part inside the vacuum chamber 8) is also shown in the figure. The method of generating active particles within the vacuum chamber 8 is the same as that shown in FIG. 2(a). The difference from FIG. 2(a) is that the electrode 10' has a pore 32 and the pore 3
2 and the spout 11 are arranged close to each other with their center lines coincident. Therefore, in addition to the (electrically neutral) active particles generated by the discharge, ions also flow into the chamber 7 through the pores 32 and the ejection port 11. chamber 7
Of the ions that have flowed into the tube, positive ions are electrically neutralized as shown in FIG. 3(b). The amount of active particles flowing into chamber 7 through grid 30 is therefore increased by this amount of neutralized ions. The method shown in Figure 3(c), compared to the method shown in Figure 3(a), is a method that attempts to neutralize and actively utilize ions, increasing the amount of active particles. This is an excellent way to

Finally, we will discuss the contents of the direction controller. In order to perform microfabrication, it is necessary to perform surface treatment (for example, etching) according to the mask. For this purpose, it is necessary to create active particles that have a velocity component in only one direction (or a velocity component in one direction that is sufficiently larger than velocity components in other directions).

The active particles flowing into the chamber 7 through the active particle generator 1 and the neutralizer 2 do not necessarily enter the wafer 4 surface perpendicularly. The direction controller 3 is.

This is a means for aligning the speed of active particles in a fixed direction.

A specific example of the direction controller 3 is shown in FIGS. 40(a) and 40(b). 4(a) shows a method in which active particles are blown out through a perforated plate 32, and FIG. 4(b) shows a method in which active particles are blown out through a cooling plate 34 connected to a cooling device 35. FIG. 4(a) is effective when using molecularly vibrationally excited particles as active particles.While the molecularly vibrationally excited particles pass through the pores 33 of the pore plate 32, the inner walls of the pores 33 collide with This collision relaxes the vibrationally excited state of one molecule, causing the particle to lose its chemically active properties.

Therefore, only the vibrationally excited particles that have passed through the pores 33 without colliding with the inner walls of the pores 33 are chemically active and react with the wafer 4 . Passing through the pore 33 without colliding with the inner wall of the pore 33 is possible only when one particle has a sufficiently large velocity component in the X direction (as shown). As a result, among the particles that have passed through the pores 33, the only particles that can react with the wafer 4 are particles that have a sufficiently large velocity component in the X direction.
The direction of the active particles is now controlled. Figure 4(a)
As shown in , if a plurality of porous plates 32 are used, the direction of the active particles can be controlled in the X direction with even greater precision. This method is effective for particles that become chemically active due to vibrational excitation. In order to control the direction of other active particles, the method shown in FIG. 4(b) is effective. In FIG. 4(b), the active particles pass through the pores 33 of the cooling plate 34, which are cooled by the cooling device 35. In FIG. The cooling plate 34 is cooled from the liquid nitrogen temperature (70°K) to the liquid helium temperature (40"K). The surface cooled to such a low temperature has the effect of adsorbing particles. Therefore, the pores are Since the active particles colliding with the inner walls of the pores 33 while passing through the pores 33 are adsorbed, the active particles passing through the pores 33 substantially have a velocity component only in the X direction.

The individual components of Example 1 have been specifically disclosed above. Embodiment 1 is constructed by combining the specific examples shown in FIGS. 2 to 4. Further, it is also possible to combine the specific examples shown in FIGS. 2 to 4 into one component. For example, the active particle generator 1 generates active particles by combining the methods shown in FIG. 2(a) and FIG. 2(c), or the direction controller 3 generates active particles as shown in FIG.
) and FIG. 5(b) may be combined. Of course, these combinations are aimed at compound effects. Further, there is no particular restriction on the size of the jet nozzle 11.

Each embodiment is sized for maximum effectiveness.

By using Example 1, it becomes possible to control the type, amount, and direction of electrically neutral active particles.

<Example 2> In Example 1, no means for separating active particles was specifically provided. Therefore, when considering controlling the type of active particles (i.e., when considering the use of specific active particles)
, the types of active particles available may be limited. Therefore, in this example, active particle separation means was added to Example 1.

FIG. 5 shows the configuration of this embodiment. The active particles generated by the active particle generator 1 are separated by an active particle separator 37,
After passing through the neutralizer 2 and direction controller 3, the wafer is irradiated. Here, active particle generator 1. Neutralizer 2. The directional brake 3 is the same as that already described in the first embodiment.

The active particle separator 37 will be explained below.

A specific example of the active particle separator 37 is shown in FIG.

FIG. 6(a) shows that the ions generated in the chamber 8 are extracted using the slit 38 and are mass-separated by the mass separator 39, so that only specific ion species are selected and sent to the neutralizer 2.
This is a method to send it to. Since the ions are turned into electrically neutral particles by the neutralizer 2, mass-separated active particles are produced as a result, and the type of active particles can be selected. Although the slit 38 is located within the chamber 8 in FIG. 6(a), it may also be located within the chamber 7. In Fig. 6(,), it does not matter where the slit 38 is installed, but the essence is to separate the generated ions by mass and send them to the neutralizer 2.
b) is a method of mass separation using rotating disks 41, 41'. The rotating disk 41 , 41 ′ is provided with one or more passage ports 42 such that the passage ports 42 periodically cross the center line of the spout 11 . rotating disk 4
A plurality of types of particles pass through the rotating disk 41 only when one passage port 42 coincides with the jet port 11. Jump speed of each particle &±
Since the particle size differs depending on the particle, the time at which the particle reaches the rotating disk 41' differs depending on the particle type.
′. In this method, even if both ions and active particles are mixed and blown out through the ejection port 11, it is also possible to separate the active particles by mass and take out only the active particles if the velocities of the two are different. In this case, the neutralizer 2 shown in FIG. 5 may be omitted, and in the system shown in FIG. It doesn't matter if you change the position of .

According to this embodiment, the types of active particles can be controlled over a wide range, so that surface treatment can be performed with higher precision than in the first embodiment.

<Example 3> As shown in FIG. 7, Example 3 is an example in which a laser generator 43 is added to Example 1 to enable active particles to be irradiated with laser light. The laser light is applied to the active particles through the quartz window 44 . The active particles can be further excited by laser light irradiation to increase their reactivity with the wafer 4. This embodiment is also applicable to the second embodiment. Similar effects can be expected by using a suitable lamp light source instead of laser light.

In this embodiment, the laser beam is irradiated after the direction controller, but it may be irradiated at any position within the chamber 7. It is also possible to use energy sources other than laser light (as long as they re-excite the active particles).

<Example 4> Amount and direction of incidence of active particles on the surface to be treated.

and to control the type of active particles, these a
Preferably, m means are installed in the device.

FIG. 8 shows an embodiment in which these observation means are added to the first embodiment. In this embodiment, a laser beam from a laser 45 is input into a chamber 7 to excite active particles, and a spectrometer 48 is used to detect fluorescence emitted when the excited particles transition to another energy level. flft! By measuring, the type and amount of active particles can be observed. Further, the progress of the surface treatment is monitored by applying a laser beam from a laser 46 to the surface to be treated and monitoring the reflected light with a detector 49. These monitor signals are sent to the controller 50
input and signal processing. In the process of this signal processing, it is possible to perform feedback control of the power source 50 and the valve 52 by making a judgment regarding the conditions that have been input to the controller 50 in advance.

The key point of this embodiment is to have means for observing the type of active particles, the amount of incidence on the surface to be treated, the direction of incidence, etc. Therefore, in addition to the configuration shown in Figure 1 and the observation device, many configurations and I
There can be ll devices.

〔Effect of the invention〕

According to the present invention, since it is possible to control the type of electrically neutral active particles, the amount of incidence on the surface to be treated, and the direction of incidence, the surface of the sample can be precisely Processing is possible.

[Brief explanation of drawings]

Fig. 1 is a block diagram of a surface treatment apparatus showing an embodiment of the present invention, Fig. 2 is a block diagram showing a specific example of an active particle generation method, and Fig. 3 is a block diagram of a loaded particle removal method or a loaded particle neutralization method. Fig. 4 is a block diagram showing a specific example of the active particle direction control system, Fig. 5 is a block diagram of a surface treatment apparatus showing another embodiment of the present invention, and Fig. 6 is a block diagram showing a specific example of the active particle direction control method. Fig. 7 is a block diagram showing a specific example of the generation method; Fig. 7 is a block diagram of a surface treatment device showing another embodiment of the present invention; Fig. 8 is a particle fiR measuring instrument sound measuring stage showing another embodiment of the present invention. A configuration diagram of the surface treatment equipment,
FIG. 9 is a characteristic diagram showing the interrelationship of energies related to chemical reactions. DESCRIPTION OF SYMBOLS 1... Active particle generator, 2... Neutralizer, 3... Direction controller, 4... Wafer, 5... Temperature controller, 6...
...Sample stand. 7... Chamber, 8... Chamber, 9... Pipe, 10°10'... Electrode, 11... Spout, 12...
...discharge tube. 13... Waveguide, 14... Magnetron, 15...
・Solenoid coil, 16.16'... anti-transparent mirror, 17
... Hole, 18.18'... Quartz window, 19... Laser, 20°20'... Lamp, 21... Electron beam source, 22... Heater, 23... Heat Shield, 24...
・Grid, 25.25'...Power source, 26...Power source, 27...Filament, 28°28'...Electrode plate, 29...Power source, 30...Grid, 31... Power supply, 32... Pore plate, 33... Pore, 34... Cooling plate, 35... Cooling device, 36... Heat shield, 3
7... Active particle separator, 38... Slit, 39...
...Mass separator, 40...Power supply, 41.41'...
Rotating disk, 42... Passing port, 43... Laser, 44
．． 44'...Quartz window, 45...Laser, 46...
・Laser, 47... Optical filter, 48... Spectrometer, 49... Detector, 2nd eye 3rd port (α) (b) Fig. 4 (α) 6th port (shi)

Claims (1)

[Claims] 1. In a surface treatment method using chemically active particles, the active particles are electrically neutral particles, and the amount and direction of incidence on the surface to be treated is controlled. A surface treatment method characterized in that the main particle species that react with the surface to be treated are selected. 2. Observing at least one of the amount of active particles incident on the surface to be treated, the direction of incidence, the main particle species reacting with the surface to be treated, and the surface treatment results during or during the surface treatment, Based on these observation results, the above incident amount,
The surface treatment method according to claim 1, characterized in that the incident direction and particle type are controlled. 3. In a surface treatment apparatus having a means for introducing a raw material gas, a means for generating active particles from the raw material gas, and a means for holding a sample to be treated, the generating means can control the amount of energy supplied to the particles constituting the raw material gas. A means for generating and removing or neutralizing loaded particles generated with the generation of active particles, and a means capable of controlling the direction of incidence of electrically neutral active particles onto the surface to be treated. A surface treatment device comprising: 4. The surface treatment apparatus according to claim 3, further comprising means capable of separating the active particles generated by the active particle generating means into specific particle types. 5. Means that can observe at least one of the following: the amount of active particles incident on the surface to be treated, the direction of incidence, the main particle species that react with the surface to be treated, and the surface treatment results during or during surface treatment. A surface treatment apparatus according to claim 3 or 4, characterized in that it has the following. 6. A means for processing the observation signal from the observation means, and based on the processed signal, determines the amount of active particles incident on the surface to be treated, the direction of incidence, and the main particle species that react with the surface to be treated. 6. The surface treatment apparatus according to claim 5, further comprising control means capable of automatically controlling at least one of the above.