JPS5821330A - Positioning method for semiconductor wafer surface and exposing unit - Google Patents

Abstract

PURPOSE:To realize a precise positioning by a method wherein photoluminescent signals obtained by laser scanning of a wafer surface are observed for their intra-wafer distribution, when the mask surface for exposure and a buried hetero-structured semiconductor wafer surface are positioned together. CONSTITUTION:This relates to a method of positioning the surface of a buried hetero-structured wafer 30 consisting of an undoped InGaAsP active layer 13 with a forbidden band gap of -0.96eV, an InGaAsP electrode forming layer 17 with a forbidden band gap of -1.05eV, or the like, and the surface of a glass mask 32 for exposure. That is, an impurity diffusion preventing SiO2 film 19 and a photoresist film 31 are provided on the surface of the wafer 30, and a glass mask 32 consisting of a non-light transmitting section 321 and a light transmitting section 322 is arranged facing the rear side of the wafer 30. Parallel laser beams 341 and 342 with an internal between them are applied to the mask 32 and a change in the luminescent signal level which the laser beams generate upon arriving at the active layer 13 is detected by a light detecting unit with the intermediary of optical fibers 33 whereby the relative positions of the wafer 30 and the mask 32 are determined.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a method and an exposure apparatus for attaching an eye to an exposure master semiconductor substrate among lithography techniques widely used in the production of semiconductor devices.

Strive amount double, Hete Sonotei conductor V-ZO is well known as an excellent structure, and the channel in the Kamekomi Hete-Fall (usually abbreviated as BH) structure. Brass straight structure (
For example, Showa 10! In the 1st year patent application No. 0062 (1), a semiconductor region with a small forbidden band width, that is, an active region, is used as a conductor.
However, the active region of the tortoise has a forbidden zone area that is larger than that of a wide semiconductor, i.e., these conductor lasers. In Wear 1, the active layer has a two-dimensional or three-dimensional structure that cannot be described by thickness alone, and the two- or three-dimensional structure of the active layer is covered with tararad. In addition to the two examples mentioned above, a structure of a striped semiconductor laser having a nine active layer structure has been proposed in order to control the low threshold current and the oscillation optical state. Many wafers for semiconductor devices include a special layer with a two- or three-dimensional structure inside the crystal, and the structure of this special layer cannot be understood by surface observation.

In addition, with the progress of selective epitaxy technology, it is becoming increasingly important to improve device characteristics and to create new structures. As an example other than semiconductor lasers, there is an ij in the research stage, but it is a future high-speed field effect transistor, i.e. permable, transistor H example, TTF197.
9th grade international level.

A simple explanation of the electron device structure concept.

Tenda Technical Digest Page 384 C6
Although it has been proposed by Mr. O., Boiler et al.

What has been proposed as a parable transistor is a mesh-like structure in the middle of a semiconductor layer with a number of tens to 100 OA.
A YrII metal layer of about OA is embedded, and this metal layer is used as the base of a Schittky barrier type.

Spreading of the space charge layer in the semiconductor layer at the position corresponding to the hole 7. In this case, the field effect transistor controls the flow of current by applying a voltage to the metal film pattern: The metal film pattern is covered by the semiconductor layer stacked on top, and it is visible from the surface of the semiconductor layer in a mesh shape. Observing the structure of the metal layer, that is, the special layer, is an inversion.

Therefore, when a special layer like the one described above exists inside a semiconductor, special measures are required to arrange electrodes and the like on the semiconductor surface in alignment with the structure of the special layer.

For example, here is an example of how to form a selective impurity diffusion region for TK current injection in a BH laser structure. FIG. 1 is a conceptual diagram of the cross section of the BH laser structure groove γ which becomes the rusted surface of the Fabry-Perot resonator after reciprocating. The Bl laser shown here consists of a heterostructure made of InP or nGaAsP for long wavelengths, and consists of an a-InP buffer layer 12, an undoped InGaAsP active layer 13, and a p-IaP cladding layer on an eye-shaped IoP substrate. 14 is grown in the first continuous liquid phase using the Schall method in Evita. The first layer is then etched to leave a striped mesa portion with a width of several micrometers on both sides as shown in Figure 1. A p-1ssP current blocking layer 15, an n-IrIPJi! inlay layer 16, and an n?-InGaAsP electrode forming layer 17 are formed using the continuous liquid phase Evita-in-Shall method.

The n-InGaAsP electrode forming layer 170 thus obtained
The surface is almost smooth, and it is not possible to accurately determine the location of moles and mesas from the surface.

K to make BH laser is n-InGaAaP electrode forming layer 1
Right above the mesa part of No. 7 K Z o + Cd k Selectively diffuse and p
It is necessary to invert to the shape area 18. To do this, first 8
Amount 02 or 81. The impurity diffusion prevention film 19 such as N4 film is a-
A window 20t for selective diffusion of impurities must be formed on the InGmAiP electrode formation layer 17 by a gas phase reaction method, aligned precisely with the striped matrix GL'' portion, and made using a photolithography technique. but.

Because it is difficult to accurately determine MET4 from the surface.

The 10 sides of the wear perpendicular to the striped mesa were polished at a shallow angle of several degrees as shown in Figure 2.

When the angular deformation paper 11t7 is stained with a mixture of Elysian potassium, caustic potash, and water, the cross-sectional structure of the first surface is enlarged in the layer thickness direction and appears on the angularly polished surface 21. By aligning a glass mask for photo and IJ lithography to the mesa portion position appearing on this shallow angle polished surface 21, a window 2G for selective diffusion is formed in the impurity diffusion prevention film 19.
It is possible to accurately drill a hole directly above the t-mesa using photolithography technology. In addition to the angle polishing described above, there are other devised methods for producing striped mesa tubes, but these methods process the sides of the wafer in a direction perpendicular to the stripes of ware 1, which is the same as the angle polishing method. @0 thing. The paper is made by angle-polishing the 10 sides of the wafer, and the processing process is not only complicated in terms of polishing or grinding, but also requires additional cleaning steps to avoid stains on the wafer. This results in an increase in man-hours and a decrease in yield.

Furthermore, such an angle polishing process itself is a very laboratory-like process, which is undesirable as it increases the number of man-hours required for the production process.

An object of the present invention is to solve the above-mentioned conventional drawbacks and to provide a method and an exposure apparatus for aligning O0 to a nine-point conductor.
be.

According to the present invention, a strong rubanesense III obtained when the surface of a wafer of a wafer for semiconductor devices is optically excited.
When applied to the semiconductor wafer 1 in which a photoluminescence signal such as L+1 luminescence wavelength and luminescence spectrum shape has an in-plane distribution, the surface or surface material layer of the semiconductor wafer is applied to the semiconductor wafer surface or surface material layer with an in-plane distribution of the photoluminescence signal. The above-mentioned method of alignment when processing with a regular relationship with ? In-plane distribution t-
A method for aligning the semiconductor cube 71 is obtained, which is characterized by aligning by observing. An exposure apparatus for lithography using such in-plane distribution of photoluminescence signals as a guide includes a light source for optically exciting a semiconductor wafer to be exposed, and a photodetector for detecting luminescent light generated by the optical excitation from the exposed semiconductor wafer γ. An exposure apparatus characterized by having a tube, in particular, the relative position of the irradiation position of a semiconductor wafer to be exposed with optical excitation light and the photodetector end of a photodetector for luminescence light or a good leading waveguide for guiding luminescence light to the photodetector. It is possible to obtain an exposure apparatus equipped with a photoexcitation light source and a photodetection system in which the value is constant.

According to the present invention, the semiconductor device wear 1 includes a special layer with a two- or three-dimensional structure inside the crystal without any processing on the wafer, and the internal pattern cannot be observed from the surface shape. Accurate alignment along the internal pattern can be carried out even when using a semiconductor wafer.Next, the method and exposure apparatus for aligning a semiconductor wafer according to the present invention will be described below.1. This will be explained in detail using figures.

lI3 The figure shows the case of forming an impurity selective diffusion mask on the BH silicone wear T2:, Figure 1 and! s2 shows an embodiment in which the alignment method and exposure device AT of the present invention are applied to the semiconductor wafer 1 in place of the conventional method shown in FIG.

FIG. 3 shows a cross-sectional view of a case where a glass mask 32 for photolithography is placed adjacent to a BHH laser wafer 30. The BHH laser wafer shown in Figure 3 is T
It is made of nP and InGaAsP and is similar to the one shown in FIG. Although the explanation is omitted in the case of FIG. 1, the forbidden band width of the undoped InGaAsP active layer 13 is ~
0.96 eV, whereas the forbidden band width of the n-InGaA@P electrode forming layer 17 is ~1.05 eV. The surface of Lawe 71 for BH laser is doped with impurities such as 8301 l11
A stopper film 19 is applied, and then a photoresist film 31 is applied.

The arrow marked 341 or 342 in Figure 3 is Y.
AG laser light (wavelength 1.06J1tllkM bandwidth L1
This corresponds to 7 eV. ). now.

As shown in FIG. 3 1a), a BHH laser wafer 3 is placed directly under the light transmission N312 of the photoresist glass mask 32.
If the undoped InGaAsP active layer 13 of 0 does not exist, laser light is irradiated with the arrow of IC 341. In the case of 9, the laser light hits the non-light transmitting part 11321 of the glass mask 32? , even if it does not reach the BH laser wear Y30 and passes through the light transmitting part 321t and reaches the BH laser wear 130 as shown by the arrow 342, the Y laser light is InP.
The absorption t is almost impossible for the electrode forming layer 171C which is crystalline.

That is, the BH silage wear 130 is not excited and does not produce luminescence. - If the BH silage wear F30O undoped InGaAsP active layer 13 is placed directly under the photoresist glass mask 320 and the light transmitting part 312 as shown in 10,000@5aabs, then the laser beam is irradiated with the arrow 341. In this case, the YAG laser beam hits the non-light transmitting portion 5321 of the glass mask 32 and does not reach the BH silage wafer SOK, as in the case of FIG.
The AG laser passes through the light transmitting portion 312 of the glass mask 32 and is absorbed by the active layer 13. In other words, the active layer 13 is excited by the YAG laser beam and generates luminescence. When luminescence occurs, the luminescence light moves in synchronization with the irradiation position ffi of #1 YAG laser light (331h or 3
32 and is guided to the photodetector at G@ etc. by the optical fiber 33. In other words, the irradiation position of the KYAG laser beam in Figure 4 11t341
．． 342, when moving sequentially from left to right in Figure 3 [moves in synchronization with the YAG laser irradiation position (optical fiber 3
3. This luminescence intensity is indicated by the voltage across the series resistor connected to the Ge detector. However, when the Ge detector and the light 71
Between Iper 33 is this pigeon house, IuGaA turtle P active layer 1
Center wavelength 1.3 μ that transmits luminescence light from 3
An interference filter with a half width of 2 oeo* is attached to prevent YAG laser light from entering GC detection III. In the $4 figure (ml is the one corresponding to No. 3 II (a) K and YAG
Even if the laser beam irradiation position is changed, the optical output Fi is still observed.
), #C is the active layer 1 of the BH laser wafer 1 whose optical output is
This figure shows how this occurs in three areas. That is, luminescence occurs when the positions of the glass, master O light transmitting portion and the striped active layer 13 of the BH Silleyware F30 match, and the luminescence intensity becomes the strongest at the most matching position. This one-piece manufacturing is done automatically, and there is a method of automatically scanning the irradiation position of the YAG laser beam within the surface of the tube glass mask, and the glass and mask 32.
Alternatively, it would be possible to automatically perform in-plane direction and in-plane rotation, but in reality, it would be better to use glass.

The mask 32 can be aligned with the container by manually moving the mask 32 in the in-plane direction, horizontally and laterally, and slightly rotating it, without automating the movement of the BH laser rollers 7 and 30.

In the above explanation of the photoresist film 31t, a photoresist such as AZN is used, but if a negative resist such as 0MR is used, the light transmitting portion 32 is used as the glass mask 32.
2 (! - It is strange to use an inverted version of the non-light transmitting part 321, but in this case, in 1!, it is determined that the alignment has been carried out at t*, when no luminescence occurs, and at t*, the luminescence is at its weakest). Needless to say, it takes a long time to do this.As mentioned above, if the method for aligning the semiconductor wafer 1 and the exposure device 1 of the present invention are used, it will be difficult to The special layered tube with a structure is also extremely effective in aligning the surface of the seam.

As shown in the conventional example of eta and s2 diagram, O0 to sea urchin
FIG. 5 shows a long-wavelength plano-convex waveguide (PCW) structure in which the present invention is applied without a semiconductor laser having the above-mentioned monolithic chimanonel substrate structure.
(abbreviated as) Laser Web OIIfIMslI. P
Groove I with CW laser! 3-4 μs.

Keyaki"1ifLl~(Forbidden band width~1-13e on n-type InP substrate 51 with grooves of L8sms8m shaped like stripes.
V()a-InGaAsP guide layer 52. Forbidden layer width 2Q
, 96eV ()LssGaAsl' active layer 53, p-I
n1' cladding layer 14 and 1llllIll band width ~1.
05eVOn -I@Gamu sr electrode forming layer 5[-$1
Next liquid phase Ebikitaki is made by tatami. Of the luminescence components obtained by excitation with this laser laser for PCW lasers, J & When the ζ netsense particle is taken at the exit of the spectroscopic hQe machine between the optical fiber 33 and the G@detector, the luminescence intensity changes to j5 in FIG. Become stronger.

This is because the ratio of the recombination at the four interfaces in the thin crystal part, the YAG laser Ilb activation in the active layer s3, and the inflow of +17 becomes larger than the recombined carriers in the guide layer 52. Therefore, the luminescence intensity increases with the groove *
The O0 alignment method can be applied to the first half of the BH laser wafer 30 by utilizing things other than S.

This also applies to the case where the metal layer t-111 is present inside the metal layer as in the case of the above-mentioned bar electrodes and totsuyisu I.

A permable metal layer like a transistor? ! The sample contained inside the F1 conductor is currently obtained, and the fact that the method of the present invention can be applied to devices such as the IO with a cross-sectional structure like the anth one.
Experiments can be conducted on luminescence from .

That is, h GaAs basic K n −A I 0.30
a 0.7AS layer 61 of about 166 m@, then GaAs
1162 t #lS 7+111 grown wear T
First, a part of the surface of the GaAs layer 620 of the ware 1 was lowered by about 4 μ4 Il by etching or the like, and a sample tube was prepared with the WKku vapor deposited film 63 coated with this Iip lowered layer. Furthermore, if a GaAs layer is formed on the KAu 851M1 s s, it will have the same layer structure as the above-mentioned permable transistor, but at present, it is formed on the Au vapor deposited film aSV.
I haven't been able to form a 91kGa substrate processed in this way! - Removed nine pieces of wear with 116 figures! It is. In FIG. 6, when krypton is irradiated to the position of the sky mark 341 in the laser beam tm, the luminescence generated from the GaA@ layer 62 is krypton, and the luminescence generated from the GaA layer 62 is due to krypton. Luminescence becomes brighter when excited in areas where there is no eclipsed film, which weakens because it is covered by a space charge layer that spreads inside. However, in the 1α4GaO, 6As layer 61, krypton, laser, and luminescence particles are absorbed t+n, but the 40g0.6As layer 61
a? Ja 7 F handling round wear 1
Since it was formed as a reinforcing layer, it was originally 1kvhlllI for this experiment. be. A company B1 detection at- was used as an i-element photodetector.

In addition, the O@ alignment method of the present invention is applicable to cases where there is a metal film locally on the surface MK and an insulating film, or when there is a region doped with impurities or a conductive moa1 in some regions of the surface MK. , insulating material or doped iF! Can be applied to clothing with different conductivity types.
In wafers having good regions with different doping levels, differences in photoluminescence intensity and spectral shape occur, making the present invention more difficult to apply. In general (semiconductor table NHK: semiconductor wear Y), the present invention can be applied to all cases where there is a region in which a foreign substance is locally formed, or a region in which the luminescence intensity, wavelength, or spectral shape of the semiconductor wear 1 has changed. It's a demon.

Furthermore, in order to realize the alignment method 1 for the semiconductor wafer t of the present invention, a normal exposure apparatus is equipped with a moving mechanism for the ware 1*iwh gas master.
Exposure with a photoluminescence excitation source and a luminescence detector tube from 11! If you add the above-mentioned light source and detection AT to the conventional equipment, the conventional installation will be slowed down with just one use.

[Brief explanation of the drawings] Figure I11 is a cross-sectional structure diagram of the BHH laser, and Figure 2 is the angle grinding 1it (I#c stay) for alignment before phosphorography for manufacturing the BHH laser current injection impurity selective diffusion mask. The internal structure of the BHH laser can be observed as a surface pattern. Figure 3 is a diagram of the BHH laser's internal structure by the method using the photoluminescence viewing column of the present invention. Diagram for explaining the alignment method, Wear 1 corresponding to Fig. 4IIl#i Fig. 3
In-wafer distribution of photoluminescence intensity from [
Figure 5 shows the alignment method according to the present invention for the FipCW laser beam 1 (all (a)), and Figure 1) shows the photonescence intensity distribution (O) for Tol and Figure 6 for par. riF T's model wear t 0
When applying the present invention-011 matching method, tI! This is a heartwarming figure. In the figure 341> and 34! Conductor wear 1 while being exposed
33 shows the optical path of the excitation light source to the photodetector, and 33 shows the 0-light 7 shielder for guiding the hot light beams generated by the optical excitation from the exposed conductor to the photodetector. 1st Country 82 Figure 3

Claims (1)

[Claims] 1. A semiconductor having an in-plane distribution of photoluminescence y In the alignment method when processing the surface of the wear 10 or the nine material layers formed on the surface in a regular relationship with the in-plane distribution of the two-sense signal, the in-plane distribution of the photoluminescence signal is m- A method 1 for aligning a cow conductor rie 71, which is characterized by aligning g by
and a photodetector for detecting luminescence light due to optical excitation from the body t.
The relative position of the luminescence detection aperture of the luminescence waveguide and the luminescence detection aperture of the luminescence photodetector is constant.
Entry I! O exposure device.