JPH01304648A - Convergence ion beam processor - Google Patents

Abstract

PURPOSE:To make it possible to perform machining with focused ion beam while optically monitoring the machining by arranging at least one of light correcting radiating/detection systems on a light axis of ion beam convergence optical system. CONSTITUTION:A reflection mirror 23 having an aperture at the center is arranged on the optical axis of ion beam under a deflector electrode 8. The ion beam passes through this aperture to irradiate a sample therewith. Light from a lamp 19 for observation illumination passing through a half mirror 20, reflection mirror 22 and aperture 18 enters a vacuum chamber 17 and is collected on a sample 10 through the mirror 23. The reflection light from the sample passing through the same route reversely to form image, which is displayed on a monitor 27 through a main controller 25. As a result, an SIM image and optical microscope image are simultaneously detected. It is possible to shorten the operation distance of focusing lens 4 so as to improve the focusing characteristic by providing parts equipped with functions of both systems in consideration of the functions of parts of both ion beam and light optical systems.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a focused ion beam processing apparatus and method thereof, and is particularly suitable for processing semiconductor devices such as VLSI as a workpiece or processing such as CUP. , relates to a focused ion beam processing device.

[Conventional technology]

A conventional focused energy beam device is a focused ion beam processing device, IMA (Ion Micro).

Ana17zer) + electron beam lithography equipment, etc.
In these methods, the only means for observing the workpiece is a secondary particle image obtained by detecting secondary particles emitted from the surface of the workpiece. Here, the secondary particles include secondary electrons, secondary ions, etc.

As an example of a conventional device, the configuration of a focused ion beam processing device is shown in FIG. An ion beam 2 extracted from an ion source 1 is focused by a focusing lens 4 and irradiated onto a sample 10 for processing. Additionally, CVD is performed at the same time as ion beam irradiation.
Gas is supplied from the nozzle 16 to perform local film formation. At this time, the blanking controller 12 controls the ion beam
N.

The deflector controller 13 controls the deflection of the ion beam. Secondary electrons or secondary ions generated from the sample 10 with the ion beam irradiation are detected by a secondary particle detector 9 (e.g. Eva multi-channel plate), and SIM (Scanni
ng Ion Microscopes: Obtain an image. The sample surface is observed using this SIM image. Incidentally, related to this type of focused ion beam processing apparatus is, for example, Japanese Patent Application Laid-Open No. 61-245.
No. 553 is mentioned.

Further, as an example of an electron beam apparatus in which a secondary electron image and an optical microscope image can be switched to make it possible to observe both, there is Japanese Utility Model Publication No. 59-10687.

[Problem to be solved by the invention]

In the conventional focused energy beam apparatus described above, only secondary particle images are used for sample observation. Here, the secondary particle image is an image that detects secondary particles generated from the sample surface in synchronization with the deflection scanning of a focused beam, and expresses the difference in the generation rate of secondary particles depending on the location as a change in brightness. It is. What can be obtained from the secondary particle image is information about the unevenness of the surface of the sample and the material, but it is not possible to know, for example, the internal structure of the sample.

-Currently, semiconductor devices such as LSIs are highly integrated.

In order to improve functionality, wiring and elements are becoming more multilayered. When processing a multilayered semiconductor device as a workpiece, the following new requirements have arisen.

0) Position the desired wiring etc. in the lower layer and perform processing while monitoring the positional deviation.

(2) Control the machining in the depth direction and complete the machining without damaging the lower layer of the workpiece.

As an observation method, conventional focused energy beam equipment that uses secondary particle images cannot meet these demands.In other words, it is not possible to observe the lower layer using secondary particle images, so it is impossible to position the processing to the lower layer. It is. Also,
It is possible to detect secondary ions as secondary particles and obtain information in the depth direction, but when processing multi-layered semiconductor devices with a high aspect ratio, detection of secondary ions itself becomes difficult and accuracy becomes difficult. Good depth control is not possible.

On the other hand, if optical observation means is used, the underlying structure can be observed through the transparent insulating layer of a multilayered semiconductor device. In addition, using optical interference, processing depth,
Information on the direction of the insulating layer, such as its thickness, can be obtained.

Here, as an apparatus capable of observing both a secondary electron image and an optical microscope image, there is an apparatus described in Japanese Utility Model Publication No. 10687/1987. However, this device uses a system that switches between secondary electron image observation and optical image observation by moving a prism for optical image observation in and out under the optical axis of the electron beam, so observation using both is performed simultaneously. Therefore, when applied to a focused energy beam device, it is impossible to perform processing by beam irradiation while optically monitoring the positional deviation and processing depth with respect to the underlying layer at the same time.

In addition, as mentioned above, considering ion beam processing for LSIs, which will become increasingly miniaturized and multilayered in the future, high-speed and highly accurate processing will be required, and in order to finely focus a large current beam, the focusing performance of the lens will be improved. becomes even more important. Therefore, the above-mentioned decrease in focusing performance due to the increase in working distance becomes an important problem in putting the device into practical use.

An object of the present invention is to provide a focused ion beam processing apparatus that can perform processing using a focused ion beam while optically monitoring the processing and has high ion beam focusing performance.

[Means to solve the problem]

The above object is achieved by shortening the working distance of an ion beam focusing lens in a focused ion beam and light optical system that allows focused light and a focused ion beam to be irradiated extremely close to a sample at the same time.

[Effect]

First, the focused light irradiation means and the detection means are installed so that a part of their optical axes coincide with the optical axis of the focused ion beam (focused energy beam), and the focused light and the focused ion beam are simultaneously applied to the sample. It is possible to irradiate the top.

Specifically, an optical component with a hole in the center through which the focused ion beam can physically pass is installed on the original axis of the focused ion beam, and while processing is performed using the focused ion beam that has passed through the hole in the optical component. At the same time, these optical components are used to optically monitor the processing state.

At this time, several parts of the focused ion beam element system and the focused light optical system must be installed between the ion beam focusing lens sample stage, so the working distance would become extremely long if left as is. . Therefore, by considering the functions of the components of both optical systems to be installed and providing a component that has the functions of both optical systems at the same time, the working distance can be shortened and the lens focusing performance can be improved. I can do it.

That is, the focused light irradiation means and the detection means are provided so that a part of their original axis coincides with the optical axis of the focused ion beam (specifically, a hole in the center through which the focused ion beam can physically pass) is provided. If it is possible to irradiate the sample with the focused light and the focused ion beam at the same time, it is possible to process the sample with the focused ion beam while optically monitoring the process. It can be carried out.

As described above, according to the present invention, the SI of a desired location on a sample can be
Since the M image and the optical microscope image can be detected simultaneously, it is possible to perform position detection and processing alignment using the optical microscope image even for lower layer wiring that cannot be detected using only the SIM image. Furthermore, since the processing status can be constantly monitored (via optical microscope images) even during processing, processing with high positional accuracy can be performed while monitoring the processing position shift with respect to the lower layer. Furthermore, when processing a window to the lower wiring, it is possible to detect the end point of processing by using the fact that when the At wiring is exposed, the intensity of reflected light changes due to the high reflectance of A/=. can.

In addition, by replacing the optical microscope image detection system with a laser interference optical system, it is possible to perform processing with high depth accuracy while monitoring the depth of the machined hole and the interlayer insulation film thickness using optical interference. . Similarly, by replacing the optical microscope image detection system with a laser scanning microscope image detection system, the laser scanning microscope image, which has a higher resolution and a deeper focal depth than a normal optical microscope image, can be used to detect processing conditions. Since machining can be carried out while monitoring, machining can be carried out with even higher positional and depth accuracy.

However, if a reflecting mirror and a reflecting objective lens are simply inserted on the original axis of a focused ion beam optical system, the distance from the focusing lens to the sample (working distance) becomes extremely long. Here, an example of the relationship between the working distance and the focusing performance of the optical system will be explained using FIGS. 4 and 5. As shown in FIG. 4, the distance from the ion source 1 to the focusing lens 4 is D, the length of the focusing lens is 40-5, and the working distance is D2.

A simulation of the focusing performance was performed by changing the and D2. As a result, when D = D2, the lens aberration is minimized and the focusing performance is the best, so next, a simulation was performed with D = D2, and the results are shown in FIG. Here, the calculation parameters are ion acceleration voltage 2, which is a typical value for focused ion beam processing equipment.
0Kv, ion extraction voltage 10KV, beam current α1
nA was used.

According to FIG. 5, the focused beam diameter rapidly expands as the working distance D2 increases. Applying this result, for example, if the working distance increases by about 70 m by inserting a reflecting mirror and a reflecting objective lens, and the working distance changes from 50 m to about 200 m, the focused beam diameter will be 2
However, according to the present invention, it is possible to finely focus a large-current ion beam, and simultaneously optically monitor processing such as processing or CVD for multilayered micro LSIs. Do the effects that can be done.

〔Example〕

The present invention will be specifically described below based on examples. First, FIG. 3 shows an embodiment of a focused ion beam processing apparatus. That is, it is necessary to install several parts of the focused ion beam optical system and the focused light optical system on the original axis of the focused ion beam optical system between the focusing lens 4 and the sample 10. In other words, the components of the focused ion beam system include the beam limiting aperture 5,
These are a blanking electrode 6, a blanking aperture, a deflector stigma electrode 8, and a secondary particle detector 9. The components of the optical system for focusing light are a reflecting mirror 23 and a reflecting objective lens 24. In the device shown in FIG. 3 in which the above components are coaxially arranged, the working distance of the focusing lens 4 is very long, and the focusing performance of the focusing lens 4 is degraded.
An ion beam 2 extracted from an ion source 1 is focused by a focusing lens 4 and irradiated onto a sample 10 for processing.

At the same time as the ion beam irradiation, CVD gas is supplied from the nozzle 16 to perform local film formation. At this time, the blanking control 12 controls the on and off of the ion beam, and the deflector controller 13 controls the deflection of the ion beam. do.

Further, the flow rate of the CVD gas from the CvD gas cylinder 15 is controlled using the flow rate adjusting pulp 14.

Secondary electrons or secondary ions generated from the sample 10 along with the ion beam irradiation are detected by the secondary particle detector 9 to obtain a SIM image. The SIM image is sent to the main controller 25 and displayed on the monitor 26. In this apparatus, since the sample is irradiated with focused light at the same time as the ion beam, reflected light, scattered light, etc. are generated. Therefore, as the secondary particle detector 9, one that uses light for detection (such as a combination of a scintillator and a 7-meter) is inappropriate, and one that directly multiplies secondary particles as electrons (multi-channel plate, channel Tron, etc.).

A reflecting mirror 23 with a hole in the center is installed on the optical axis of the ion beam below the deflector electrode 8. The ion beam passes through this central hole and irradiates the sample 10. On the other hand, the light from the lamp 19 for observation illumination enters the vacuum chamber 17 through the half mirror 20, the reflecting mirror 22, and the window 18, and after its optical path is bent by the reflecting mirror 23, it passes through the reflecting objective lens 24. The light is focused on the sample 10. The reflected light from the sample passes through the same path in reverse and forms an image on the camera 21,
The obtained optical microscope image is sent to the main controller 25 and displayed on the monitor 27.

Since the SIM image and optical microscope image of a desired location on the sample can be detected simultaneously, even for lower layer wiring that cannot be detected with the SIM image alone, the position can be detected using the optical microscope image and processing alignment can be performed. be able to. Further, even during processing, the processing state can be constantly monitored using optical microscope images, so processing with high positional accuracy can be performed while monitoring the processing position shift with respect to the lower layer. Furthermore, when machining a window to the lower wiring, the end point of the machining can be detected by using the fact that when the M wiring is exposed, the intensity of reflected light changes due to the high reflectance of M.

Furthermore, in the configuration shown in FIG. 3, by replacing the optical microscope image detection system with a laser interference optical system, the depth of the machined hole and the interlayer insulation film thickness can be monitored using optical interference.
Machining with good depth accuracy can be performed. At the same time, by replacing the optical microscope image detection system with a laser scanning microscope image detection system, we can use laser scanning microscope images with higher resolution and a deeper depth of focus than normal optical microscope images to determine the processing state. Since machining can be carried out while monitoring, machining can be carried out with even higher positional and depth accuracy.

Therefore, the present invention takes into account the functions of the components of both the ion beam and light optical systems, and shortens the working distance of the focusing lens 4 by providing components that simultaneously have the functions of both optical systems. , which improves focusing performance.

Next, a first embodiment of the present invention will be described.

That is, FIG. 1 shows the device configuration of the first embodiment of the present invention. In this embodiment, a reflecting mirror 23 also functions as a blanking aperture, and a reflecting objective lens 28 provided with an 8-focus deflector stigma electrode is used. In addition, the working distance of the focusing lens 4 is shortened so that the focusing of the light and the deflection of the ion beam are performed simultaneously.

Second, the specific structure of the reflective objective lens 28 provided with the 8-focus deflector stigma electrode is shown in the sixth factor and FIG. 7. In FIG. 6, the optical path of the light is shown by a chain line, and the light incident on the reflective objective lens is successively reflected by the reflective convex mirror 29 and the reflective concave mirror 30, and is focused on the sample. At this time, since the effective reflective surface of the single reflective film mirror 29 is concentrated around it, a hole is provided in the unnecessary central portion to allow the ion beam indicated by the solid arrow to pass through. An eight-electrode deflector stigma electrode 32 is provided in a blind spot on the optical path of the light, and a voltage is applied from a deflector controller to adjust the beam shape and deflect the ion beam. The AA′ cross section in FIG. 6 is shown in FIG.
Each electrode 51a to 31 of the single-pole deflector stigma
Wiring is performed to h by screws 33a to 53h, and a voltage is applied.

Next, a second embodiment of the present invention will be described.

That is, FIG. 8 shows the apparatus configuration of a second embodiment of the present invention. In this embodiment, in place of the reflective objective lens 28 of the first embodiment, a transmission objective lens 55 provided with an 81-pole deflector stigma electrode is used. , the working distance of the focusing lens 4 is shortened so that the focusing of the light and the deflection of the ion beam are performed simultaneously.

Next, the specific structure of the transmission objective lens 35 provided with an 81-pole deflector stigma electrode is shown in FIGS. 9 and 10. The light that has passed around the lens is focused onto the sample, and the ion beam I passes through a hole provided in the center and is irradiated onto the sample. 81-pole deflector stigma electrodes 56a to 56h are provided on the inner wall of the center hole, and a voltage is applied from the deflector controller to adjust the beam shape.
Deflect the ion beam. Wiring to each electrode is first done to blocks 38a to 38h using screws 39a to 39h, and wiring from each block to the electrode is done using a transparent conductive film (In
2O5H5n02, etc.) 37a to 57h, since current hardly flows through each electrode, a thickness of several tens of OA is sufficient for the transparent conductive film, but by adjusting the thickness of the transparent conductive film, The transmittance of light passing through can be adjusted.

Next, a third embodiment of the present invention will be described.

That is, FIG. 11 shows the apparatus configuration of the third embodiment of the present invention. In this embodiment, a reflecting mirror 40 provided with a deflector electrode bends the optical path of the light and deflects the ion beam at the same time, and also functions as a blanking aperture, so that the working distance of the focusing lens 4 is shortened. It is.

Next, the specific structure of the reflecting mirror 40 provided with a deflector electrode is shown in FIGS. 12 and 13. The upper block 42 and the lower block 43 are roughly composed of two transparent blocks, and the boundary surfaces of the upper block 42 and the lower block 43 are coated with a reflective film, and the reflective film coating surface 46 reflects the original light and bends the optical path. The upper block 42 and the lower block 43 are provided with a square hole in the center through which the ion beam passes, and the inner walls thereof are provided with deflector electrodes 44a to 44d in directions perpendicular to each other on the upper and lower sides. At this time, since light enters the lower block 43 from the right direction, the deflector electrodes 44o and 44d are provided in directions that do not block the light.

Each electrode is wired using transparent conductive films 45 to 45d, and a voltage is applied from a deflector controller to deflect the ion beam in the X and Y directions when passing through the upper and lower blocks. By adjusting the thickness of the transparent conductive film used for the wiring, optimum transmittance for passing light can be obtained. Now, as shown in Figure 13, among the reflective parties,
The light that crosses the hole in the center is L2, and the other light is Ll.
Then, an optical path difference of (n-1)t occurs between L and L2. Here, the refractive index of the lower block 43 is nl, and the length of one side of the central hole is t. Due to the above optical path difference, L and L
2, the focal point shifts, but since the proportion of the light amount of L2 is small, it has almost no effect on the resolution of the obtained optical microscope image.

However, if the decrease in resolution affects the position accuracy, it is possible to change the orientation of the deflector electrodes in the upper and lower blocks.This will block all the light from L2, and as a result,
Although the total amount of light decreases slightly, all reflected light can be focused on the same point.

Next, a fourth embodiment of the present invention will be described. That is, FIG. 14 shows the device configuration of the fourth embodiment of the present invention. Ion source 1
The well-extracted ion beam 2 is focused by a focusing lens 4 and irradiated onto a sample 10 for processing. Further, at the same time as the ion beam irradiation, CVD gas is supplied from the nozzle 16 to perform local film formation. At this time, the blanking controller 12 controls turning on and off of the ion beam, and the deflector controller 13 controls the deflection of the ion beam. Further, the flow rate of the CVD gas from the CVD gas cylinder 15 is controlled using the flow rate adjusting pulp 14. The secondary particle detector 9! detects secondary electrons or secondary ions generated from the sample 1° upon irradiation with the ion beam. /c to obtain a SIM image.

The SIM image is sent to the main controller 25 and further displayed on the monitor 26. Here, in the present invention, since the sample is irradiated with focused light at the same time as the ion beam, reflected light,
Scattered light etc. are generated. Therefore, the secondary particle detector 9
Detection using light (such as a combination of a scintillator and a 7-meter) is inappropriate, and it is necessary to choose a type that directly multiplies secondary particles as electrons (such as a multi-channel plate or channeltron). be.

A reflecting mirror 25.40 having a hole in the center is installed on the i-axis of the ion beam of the deflector electrode 28.

The ion beam passes through this central hole and irradiates the sample 1°. The light from the observation illumination lamp 19 enters the vacuum chamber 17 through the half mirrors 20, 22 and the window 18, is reflected by the reflector 23°40, and is reflected by the objective lens 28, 30, 35. The light is focused by 41 and irradiated onto the sample 10. The reflected light from the sample passes through the same path in reverse and forms an image on the camera 21, and the obtained optical microscope image is sent to the main controller 25 and displayed on the monitor 27.

Here, the objective condensing system for illumination light and the like has been explained in FIGS. 6 and 7, factors 9 and 10, and FIGS. 12 and 13.

In the present invention, by simultaneously irradiating the sample with an ion beam and focused light, a SIM image and an optical microscope image of a desired location on the sample can be detected simultaneously.

Therefore, a method for processing a multilayer sample using simultaneous observation of a SIM image and an optical microscope will be explained using FIG. 15.

Here, an example is shown in which a window opening is processed to expose the lowest layer wiring 50 in an LSI having a three-layer wiring structure. When the glabellar insulating film is flattened to form multilayer wiring, the unevenness information indicating the position of the underlying wiring no longer appears on the LSI surface. Therefore, in the SIM image in FIG. 16, the position of the lowest layer wiring 70 is cannot be detected, and positioning at the start of machining is impossible. On the other hand, in the optical microscope image detected at the same time, the bottom layer wiring 70
The position of can be detected. Therefore, processing and positioning of the lowest layer wiring 70 can be performed by the following procedure. First, the magnifications of the SIM image and the optical microscope image are made to match in advance. This is easily possible by adjusting the deflection voltage of the ion beam with a deflector controller and finely adjusting the magnification of the SIM image. The positional coordinates of the images are made to correspond to each other by using a reference that does not cause positional deviation in the images (for example, the center position of the through hole 73 in FIG. 15). Finally, the position of the processing hole 72 is set with respect to the lowest layer wiring 70 detected by the optical microscope image, the set position is made to correspond to the SIM image, and the ion beam is irradiated to the processing hole position on the SIM image for processing. Start.

Machining is started as described above, and according to the present invention, while machining is being carried out, the machining position can be constantly observed using an optical microscope image and the machining state can be monitored. Here, in the conventional ion beam machining apparatus, the observation means is only a SIM image, and only a SIM image of the inside of the machined hole can be obtained during processing, so that positional deviation etc. due to charge-up of the machined hole itself cannot be detected. Although it is conceivable that SIM images of a wide area are intermittently detected in order to observe the position of the machined hole, it is impossible to detect positional deviation in real time with this method. On the other hand, according to the present invention, observation using an optical microscope image can be performed in real time during machining, so that machining with high positional accuracy can be performed while monitoring the positional deviation of the machined hole.

Furthermore, in the conventional device, there was no effective means for monitoring the completion of the window opening process for the M wiring.

For example, even if the machined part is observed using a 31M image, it becomes difficult for secondary particles to come out from the bottom of the machined hole as the process progresses.
As shown in the figure, the machined hole 72 only appears dark, and it is not possible to determine whether or not the lowest layer wiring 70 is exposed.On the other hand, if you observe the machined part using an optical microscope image, you can see that the machined hole 72 The exposure of the M wiring can be determined by the change in the brightness within. In other words, while the 5102 film between the eyebrows remains on the M wiring, the intensity of the reflected light is weak due to multiple interference and refraction caused by the thin film, and when the M wiring is exposed, the intensity of the reflected light is strong due to the high reflectance of M. Therefore, the exposure of the M wiring can be determined by detecting the change in the reflected light intensity as a change in the brightness of the processed hole. According to the invention, therefore:
Since the optical microscope image can be observed in real time during processing, processing can be performed while constantly monitoring the exposure of the At wiring, and it is possible to process window openings with high precision without excess or deficiency.

Next, a laser interferometer, which is another optical monitoring means for processing in this embodiment, will be explained. In FIG. 14, nine laser beams oscillated from a laser oscillator 50 pass through a shutter 51, have their beam diameter expanded by an optical path extender 52, and then pass through a variable transmittance filter 53. Here, the laser beam is turned on and off by the shutter 51, and the intensity of the laser beam is adjusted by the variable transmittance filter 53.The optical path of the laser beam is divided into two by the beam splitter 4B. 18 and enters the vacuum chamber 17. The incident laser beam bends its optical path through the reflecting mirrors 23 and 40 and the objective lenses 28, 35, and 41C, and is condensed and irradiated onto the sample 10. At this time, in order to reduce the percentage of laser light intensity lost due to the hole in the center of the reflecting mirror 23, the beam diameter is expanded by the optical path expander 52. Sample 10
The reflected light passes through the same path in the opposite direction, has its optical path bent by the beam splitter 48 , and enters the 7-meter 49 . Further, the other laser beam whose optical path was first split by the beam splitter 48 is reflected by the reflecting mirror 59, passes through the beam splitter 48, and enters the photomultiple 49. As described above, the two lights reflected by the sample 10 and the reflecting mirror 59 interfere, and the intensity of the interference light is reduced to 7.
Measured using Otomal 49. At this time, according to instructions from the main controller 25, the hiezo element 60 is used to reflect f.
Finely move the i59 to finely adjust the optical path difference between the two lights. A method for measuring the machining depth using the Michelson interferometer configured as described above will be explained with reference to FIGS. 16 to 19.

First, FIG. 16 shows a schematic diagram of an interferometer using a laser with a single wavelength λ. The laser beam irradiated onto the sample 10 is the 16th laser beam.
As shown in the figure, the light should always be reflected back from the bottom of the machined hole. Since the reflective surface at the bottom of the machined hole recedes during processing, the optical path difference between the two lights changes and the intensity of the interference light changes. Therefore, by measuring the change in the interference light intensity using the photomultiplier 49 and monitoring the optical path difference between the two lights, the depth of the bottom of the machined hole can be determined. Here, if a material with equal transmittance and reflectance is selected for the beam splitter 48, the amplitudes of the two lights incident on the photomultiplier 49 are proportional to the reflectances of the sample 10 and the reflector 590, respectively. For example, if the reflectances of the sample 10 and the reflector 590 are r and 1, then 2
The amplitude of the two lights is rao at the time of incidence on the photomultiple 49.
It can be expressed as l 8q. Therefore, if the phase difference between the two lights is δ, the amplitude a of the interference light is aj = aO+rageδ: (11).
Finely adjust the position of the reflecting mirror 59 so that the value becomes 0). Then, the phase difference δ can be expressed by the following equation from the machining depth dVc.

δ=μ・2π (2) λ (1) From equation (2), the interference light intensity factor is determined as follows.

■, =: l aj 12 = ao'(1+r2+2r0Tπ...(3) For example, L
In SI processing, when processing an M wiring or an 81 board, the reflectance r can be considered to be approximately constant.

At this time, r in equation (3) becomes a constant, and the interference source strength changes as the machining depth d changes as shown in the graph shown in FIG. Therefore, from the start of processing, interference light intensity was worked on.

By tracking the change in , the machining depth d can be monitored from the relationship shown in FIG. Conversely, the reflecting mirror 59 is moved so that the maximum value of the interference light intensity 11 does not change, that is, so as to cancel out the retreat of the reflecting surface due to processing on the sample 10 side. Then, the machining depth d can be directly read from the amount of movement of the reflecting mirror 59.

At this time, the resolution of the movement amount of the hiezo element used for fine movement of the reflecting mirror 59 with respect to the entire stroke is about 1/1000, and for example, when processing up to a machining depth of 20 μm,
The depth reading accuracy is 102 μm, which is sufficient accuracy.

On the other hand, in LSI processing, when processing the 5i02 film of the insulating layer, the light irradiated onto the sample 10 causes multiple interference with the 5102g. At this time, the reflectance r changes depending on the 5102 film thickness, that is, the processing depth, and can no longer be treated as a constant.

Therefore, when the depth monitor using interference described above is applied, the interference light intensity I.

This may result in complex changes and lower monitoring accuracy. In this case, it would be better to measure only the intensity of the reflected light from the sample 10 and monitor the machining depth using changes in the reflected light intensity as the machining depth changes, that is, the 5102 film thickness. improves.

This monitoring method will be described later. Note that in the interferometer of this embodiment as well, by blocking the reflected light from the reflecting mirror 59 using a light absorber or the like, it is possible to easily switch to measuring the intensity of the reflected light from the sample 10 alone. . Therefore, when processing the 8102 layer and the M layer sequentially in a multilayer LSI, switching to depth monitoring using reflected light intensity measurement and interference light intensity measurement for the 8102 layer and the M layer, respectively, will improve the overall performance. Depth monitoring can be performed with high accuracy.

Next, a schematic diagram of an interferometer using lasers with multiple wavelengths is shown in the first part.
It is shown in Figure 8. The structure of the interferometer section is the same as that shown in FIG. 16.

The illumination light oscillation unit includes laser oscillators A61, B62 . C
The respective wavelengths λ1. λ2. The laser beam of λ3 is sent to the shutter 64. Shutter 65. Shutter 65. The shutter 66 is configured to switch between ON and OFF to irradiate each laser beam. The interference source intensity is measured by irradiating a laser beam. Figure 19 shows the change in the intensity of the interference source with the machining depth d when the reflectance r of the sample 100 is constant (when processing A1 or Si). show. Here, the interference light intensity when using laser oscillators A, B, and C is I
2A I I2B I I2° binding, which are shown by solid lines, -dot-dashed lines, and dotted lines in FIG. 19, respectively. Each interference light intensity change is similar to that shown in FIG. 17, and changes at a period of 1/2 of the wavelength. However, the wavelength λ1. λ2. Since λ3 is different, the value of I2A t I21112G changes as the depth d increases. For example, as lasers A, B, and C, Ar laser blue, Ar laser green + H
When using e Ne laser λ1=488nm+
λ2=515nm.

λ, = 633 nm and %E2A I I2B
Calculated depth d is 511 for t I2C to match again.
It will be around 3rd grade. Actual processing (number of 10 μm)
If we consider, E2A + I2B, I,. will never match again. Therefore, I2A I I2B I
By measuring the intensity of the three I2C interference lights, the depth d at that point can be determined from the relationship shown in FIG.

In depth monitoring using an interferometer as shown in Figure 16, which uses a laser with a single wavelength, the initial value is set so that the intensity of the interference light is maximized at each start of each process, even when multiple processes are performed on the same sample. need to be adjusted.

On the other hand, in the depth monitor using an interferometer shown in Fig. 18, which uses lasers with multiple wavelengths, at the reference height, 3
It is only necessary to adjust it once so that the intensity of the two interference lights becomes maximum. At the start of each process, I2A l 1231
First, the relative height of the machining start surface with respect to the reference height is determined from the value of tI2C, and then the relationship shown in Figure 19 is used to monitor the machining to the desired depth. This monitor is particularly effective when performing multi-stage machining as shown in FIG.

Next, a local film forming method using the apparatus of this embodiment will be explained. In FIG. 14, the laser beam oscillated from the laser oscillator 50 is irradiated onto the sample 10 through the aforementioned path. The transmittance variable filter 53 adjusts the laser beam intensity to a sufficient level, and (4D gas is supplied to the nozzle 16
Then, local film formation is performed by laser CVD. Further, at the same time as ion and partial irradiation, CVD gas is supplied to perform local film formation by ion beam CVD. In this device, the ion beam and laser beam can be turned on and off as desired using a blanking controller 12 and a shutter 51, respectively.
Can be FF. Therefore, while supplying CVD gas from the nozzle 16, the ion beam and laser beam are turned on.
Ion beam CVD by controlling OF'F
, laser CVD, and simultaneous film formation using both methods can be arbitrarily selected.

FIG. 20 shows an example of local wiring formation using this device. In laser CVD, the sample is locally heated using a laser beam, and the thermal energy decomposes the CVD gas molecules near the sample surface to form a film.If the laser beam intensity is adjusted to a sufficient level, the crystal structure can be Although it has the advantage of being able to form high-quality low-resistance wiring at high speed, it has the disadvantage that it is difficult to form thin wiring due to heat diffusion. Further, in ion beam CVD, film formation is performed by decomposing CVD gas molecules using the physical energy of irradiated ions. Ion beams have the advantage of being able to form fine wiring because they can be focused finely and have little energy dispersion to the surrounding area, but the resistivity is high because the crystal structure is difficult to prepare, and the low energy content makes it difficult to form films at high speed. There are some difficult drawbacks.

FIG. 20 shows an example of wiring formed by combining the advantages of both of these CVD wirings. Lower N at contact hole 86
jA1 wiring 85, M pad 83 and M pad 84
Only the ion beam is irradiated while passing through the narrow gap between the two, forming fine ion beam CVD wiring 87. After drawing out the wiring to a position where it will not affect the surrounding area, a laser CVD wiring 88 with low resistance is formed at high speed by irradiating the wiring with a heavy laser beam.

Next, laser CVD wiring formation on an LSI using this apparatus will be explained using FIG. 21. The surface of most of the LSI is covered with an SiO□ insulating layer and an M wiring layer therebelow. When a laser beam is irradiated on an LSI, 810
It is transmitted through the second layer, reflected on the surface of the M layer, and only a small portion of the laser beam energy is absorbed. Therefore, when forming laser CVD wiring on the surface of an LSI, simply irradiating the laser beam with the laser beam will reduce the intensity of the laser beam. It becomes necessary to make it extremely strong, and damage to the element becomes a problem. On the other hand, according to the apparatus of this embodiment, as shown in FIG. 21, the ion beam 89 and the laser beam 90 can be superimposed and irradiated simultaneously. When CvD wiring is formed using this superimposed beam, an ion beam CVD wiring 87 is first formed, and this wiring efficiently absorbs the laser beam 90 to form a laser CVD wiring 88. At this time, since the intensity of the laser beam 900 can be kept low, problems such as damage to the element can be prevented. Note that by using this method, local film formation using laser CVD becomes possible not only for LSIs but also for materials with considerably high transmittance and reflectance.

FIG. 22 shows an apparatus configuration of another embodiment of the present invention. The lens barrel section of the focused ion beam optical system and the observation section for optical microscope images are the same as those in the above embodiment.In this embodiment, a laser scanning microscope is added as an optical measurement means. The laser beam oscillated from the laser oscillator 50 is transmitted through a shutter 51,
The light passes through the variable transmittance filter 55 and is once focused by the focusing lens 91, and then enters the XY scanner 93. XY
The laser that has passed through the scanner 93 enters the vacuum chamber 17 through the window 18, and after its optical path is bent by the reflecting mirrors 23 and 40, it is focused onto the sample 10 by the objective lens 28° 35.41. be done. The reflected light from the sample 10 passes through the same path under reverse pressure, has its optical path bent by the half mirror 92, forms an image at the pinhole 94, and enters the 7-axis 96. Here, the laser beam focused by the focusing lens 91 is used as a point light source, and the image formation position (pinhole position) of this point light source and the reflected light form a confocal position relationship, and 1
The laser beam reciprocates through the two objective lenses 28, 35, and 41, making the entire optical system a confocal type. As described above, the spot area 1 of the focused laser beam on the sample 10 is
The intensity of reflected light from the point is detected by a photomultiplier 95. Using the XY scanner 93, the focused laser beam is applied to the sample 10.
The intensity of the reflected light is detected by a photomultiplier 95 in synchronization with the scanning, and a laser scanning microscope image is obtained.

Laser scanning microscope images have high resolution, and unnecessary scattered light is removed by pinholes and is not affected by it at all, so a clear image with high contrast can be obtained. Furthermore, even if the same objective lens is used, the depth of focus as an image becomes deeper, so it becomes an optimal optical observation means especially for a deep sample such as a multilayer LSI.

Next, a processing depth monitor using reflected light intensity measurement will be explained. As mentioned in Example 1, when a transparent insulating film (S10□ film, etc.) of an LSI is irradiated with light, multiple interference occurs, and the intensity of reflected light changes depending on the film thickness. Therefore, when processing a window in the insulating film of a multilayer LSI, a focused laser beam is irradiated onto the processed material at the same time as the processing, and changes in the thickness of the insulating film due to processing are detected as changes in reflected light intensity. Monitors the machining depth.

Figure 23 shows a schematic diagram of multiple interference of light due to a transparent film.
The sample is an LSI, and a transparent insulating layer 10 is placed on the M wiring 102.
A thin film of No. 1 was formed. The incident light Lo is repeatedly reflected between the thick insulating layers 71 and passes through the first . Second. Third.・・・
・Reflected light L1+ L2 +L5. ... will occur.

Ll+L2+L5+...The interference light intensity in which all the reflected lights interfere is detected as the actual reflected light intensity. Transmittance 9 reflectance from vacuum to insulating layer 101 is t, 1
rl 1 The transmittance 9 reflectance from the insulating layer 101 to the M wiring 102 is t2 Hr2 + the transmittance 1 reflectance from the insulating layer 101 to vacuum is t'.

r/, and the incident light. The amplitude of a. , the phase difference between adjacent reflected lights (for example, and L2) is a, and the insulating layer 71
If the light absorption rate of is α, then the amplitude of the repeatedly reflected interference light @a
5 beams, , L2. Add all L,... to a3
=aqr1+a(1t1t'1r2e a+aot
1t'1r4r22e-'αD, 12δ+0. .

(Here r'1 ''-r 1 + fa, t'1 =
1 r2. The phase difference δ obtained by i can be expressed by the following equation using the thickness of the insulating layer 101 and the refractive index n.

(From Equation (5), the intensity of the repeatedly reflected interference light can be found as follows.

15 =la312 (6) Next, in this example, a schematic diagram of the device configuration for realizing depth monitoring by measuring the intensity of reflected light is shown in FIG. A laser with multiple wavelengths is used as a light source. Laser oscillator AlO3, B104. Each wavelength λ4 oscillated from ClO3. λ2. The laser beam of λ3 is sent to the shutter 106. Shutter 1o7° ON, OFF'F by shutter 108
Then, each laser beam is switched and irradiated. Laser light is irradiated and the reflected light intensity I is measured, but when the sample is an LSI, α kg is obtained in equation (6), and the change in I due to film thickness is generally as shown in the 25th factor.

Here, the reflected light intensity when using laser oscillators A, B, and C is I3A l km I Is, respectively. The binding is shown in FIG. 25 by a solid line, a two-dot chain line, and a two-dot chain line, respectively. The intensity of each reflected light changes at a period of 1/2n of the wavelength, but since the three wavelengths are different, the intensity of each reflected light shifts with the value of the film thickness. * Since rxc will never match, by measuring the five reflected light intensities, the thickness of the insulating film can be determined from the values. To actually open the window, at the same time as processing, three laser beams are repeatedly irradiated onto the processed material, the film thickness is determined from the intensity of the reflected light, and processing is performed until the value of D becomes zero. .

Note that it is clear that the laser illumination optical system that irradiates light onto the sample may be configured to provide oblique illumination.

〔Effect of the invention〕

As explained above, according to the present invention, processing can be performed using a focused energy beam while optically monitoring the processing state using focused light. Even in such cases, processing can be performed with high positional accuracy and depth accuracy, and the focusing performance of the energy beam can be improved and a large current beam can be focused finely, so there is an effect that processing with high plane processing accuracy can be performed at high speed.

[Brief explanation of the drawing]

FIG. 1 is an apparatus configuration diagram showing a first embodiment of the present invention, and FIG.
The figure is a block diagram of a conventional device, Figure 3 is a block diagram of a device having a coaxial optical system for a focused ion beam and light, and Figures 4 and 5.
The figure shows an example of the relationship between working distance and focusing performance. The sixth factor and FIG. 7 are schematic diagrams of a reflective objective lens equipped with an eight-electrode deflector stigma electrode. FIG. 8 shows a second embodiment of the present invention. 9 and 10 are schematic diagrams of a transmission objective lens provided with an 8-electrode deflector stigma electrode, and FIG. 11 is a diagram of an apparatus configuration showing a fifth embodiment of the present invention, and Fig. 13 is a schematic diagram of a reflecting mirror provided with a deflector electrode, Fig. 14 is an apparatus configuration diagram showing the fourth embodiment of the present invention, Fig. 15 is an explanatory diagram of a SIM image and an optical microscope image, Fig. 16 is an explanatory diagram showing the principle of depth measurement using light interference. Fig. 17 is a diagram showing an example of the relationship between depth and interference light intensity, Fig. 18 is an explanatory diagram explaining the principle of depth measurement by light interference using multiple wavelength illumination, and Fig. 19 is a diagram showing an example of the relationship between depth and interference light intensity. A diagram showing an example of the relationship between depth and interference light intensity when using illumination, Figure 20 is a schematic diagram of CVD wiring performed using the apparatus shown in Figure 14, and Figure 21 is a diagram showing CVD wiring using the apparatus shown in Figure 14.
FIG. 22 is a diagram illustrating the wiring formation method, FIG. 22 is an apparatus configuration diagram showing an embodiment of the present invention different from FIG. FIG. 25, which is a diagram explaining the principle of measurement, is a diagram showing an example of the relationship between the intensity of reflected interference light with repeated changes in film thickness. 1...Ion source, 2...Ion beam, 3...
- Extraction electrode, 4... Focusing lens, 5... Beam limiting aperture, 6... Blanking electrode. 7...Blanking aperture, 8...Deflector stigma electrode, 9...Secondary particle detector, 1
2...Blanking controller, 13...Deflector stigma controller, 14...Flow rate adjustment valve, 15...CVD gas cylinder, 16...Nozzle) L/, t9---Lamp, 2l--- TV camera, 24-- Reflective objective lens, 25... Main controller, 28... Reflective objective lens provided with an 8-electrode deflector stigma electrode, 29... Reflective convex mirror,
30... Reflective concave mirror, 31.32... 8-electrode deflector stigma electrode, 34... Insulator, 35
...Transmissive objective lens provided with an 8-electrode deflector stigma electrode, 36...8 electrode deflector stigma electrode, 37...Transparent conductive film, 40...Reflector provided with a deflector electrode, 44...Deflector electrode, 4
5...Transparent conductive film, 46...Reflective film coating surface. Figure 3 Figure 40 Figure 50 Figure 60 Figure '7 Figure 8L\1 Figure 90\10 M11 Figure 1z Figure 11' +611\1 170 Depth to 20 Figure 21 Figure 24121 Figure 2511 Gland Thickness D

Claims (1)

[Scope of Claims] 1. In a focused ion beam processing apparatus that irradiates an ion beam in a focused manner and performs processing or CVD film formation on a substrate having wiring covered with a protective film, a light condensed irradiation/detection optical system is used. A focused ion beam processing device, characterized in that at least a portion of the ion beam is disposed on the optical axis of the ion beam focusing optical system. 2. A deflection electrode of the ion beam focusing optical system and a small convex reflecting mirror and a concave reflecting mirror of the light focusing irradiation/detection optical system are arranged adjacent to the lower part of this deflection electrode, and the center portions of both are hollow. Claim 1 characterized in that it is integrally formed with
The focused ion beam processing device described. 3. The lens of the light condensing irradiation/detection optical system is integrally arranged inside the deflection electrode of the ion beam focusing optical system, and the deflection electrode is placed in the hole drilled in the center of the lens. 2. The focused ion beam processing apparatus according to claim 1, wherein wiring is provided from the ion beam.