JP3144250B2 - Automatic development method and apparatus - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an automatic developing method and apparatus, and more particularly to an automatic developing method and apparatus for a resist used as a mask for a semiconductor wafer or a glass substrate.

[0002]

2. Description of the Related Art In order to form an integrated circuit on a substrate such as a semiconductor wafer or a glass substrate, a process such as etching is performed while the surface of the substrate is covered with a patterned mask. Then, in order to form a patterned mask on the surface of the substrate, the resist film coated and formed on the substrate surface is exposed, and then developed, and if the resist is of a positive type, this exposed portion is dissolved. In the case of a negative resist, an unexposed portion is dissolved and washed away to form a patterned mask.

The above-mentioned development step is an important step for determining the pattern accuracy after development. For example, in the case of a positive resist, if the development time is long, the pattern width of the resist becomes narrow, and if the development time is short, the pattern width becomes small. Therefore, the development time must be set appropriately.

Conventionally, as shown in FIG. 5, a support 103 which is raised and lowered by a lifting device 102 is provided in a developing tank 101 filled with a developing solution 100, and a substrate 104 coated with a resist 104 and exposed is mounted thereon. Place developer solution 10
0 and the substrate 105 and the developing tank 101 from the start of development.
Of the current flowing between the electrode 106 and the electrode 106 immersed in the developing solution 100 is monitored and its BTT (Break Thr
ough Time) is determined by the BTT determination circuit 107.

[0007] Here, the change with time of the current flowing between the substrate 105 and the electrode 106 is as shown in FIG. That is, at the start of development T0, as shown in FIG.
a and the unexposed portion 104b remain in the initial state, and at the start of development T0, a current flows due to electrochemical effects due to the battery effect. Next, as shown in FIG. 3B, a current flows when the exposed portion 104a of the resist 104 is melted and the substrate 105 is exposed, and this time T1 becomes BTT.

[0006] Therefore, TDT (Total Development)
Time) calculation circuit 108 calculates the T based on the determination result of BTT determination circuit 107 and the development correction coefficient K from development correction coefficient storage circuit 110 in which development correction coefficient K is stored in advance.
Using the formula of DT = K × BTT, the proper development time T
Calculate DT. That is, as shown in FIG.
From time T2 to the appropriate development time TDT,
At time T2, as shown in FIG. 7C, the exposed portion 104a of the resist 104 is dissolved, and the unexposed portion 104b remains.

Incidentally, the development correction coefficient K in this case is
For example, in the range of 1.50 to 2.50 (this value is an example in a case where use conditions of a resist, a developer, an exposure apparatus, and the like are specified, and differs depending on use conditions), and the ratio of the exposure area to the substrate area , The development correction coefficient K decreases as the ratio of the exposure area to the substrate area increases, and the development correction coefficient K increases as the ratio of the exposure area to the substrate area decreases. This is because the particle beam (electron beam, etc.) used as an exposure light source is scattered in the resist, and as the ratio of the exposure area to the substrate area increases, the amount of scattering increases, the sensitivity of the resist increases, and the appropriate development time TDT increases. This is to shorten it.

As described above, the development correction coefficient K for the development time
Changes in accordance with the ratio of the exposure area to the substrate area, the value of the development correction coefficient K corresponding to the product pattern (exposure area) is stored in advance in the development correction coefficient storage circuit 109, and the The development correction coefficient K is selected and set, and the TDT calculation circuit 108 fetches the value of the selected and set development correction coefficient K and integrates it with the BTT obtained by the BTT determination circuit 107 to calculate an appropriate development time. Like that. Then, after the elapse of the appropriate development time TDT,
The support 103 is raised by the lifting device 102, and the substrate 1
05 is pulled up from the developer 100 to complete the development.

In order to determine the value of the development correction coefficient K stored in the development correction coefficient storage circuit 109, the value of the development correction coefficient K is set to, for example, 1.5 to 2.5 for a plurality of test samples of a certain product. The mask is developed while changing it at intervals of 0.1, and the pattern width after development is measured for each of them. The value of the development correction coefficient K at which the length measurement result becomes equal to the design value is obtained, and the product-specific development correction is performed. The coefficient is assumed to be K.
Then, a development correction coefficient K unique to the product is obtained for other products by the same method, and is selectively set when developing. Therefore, for the n kinds of products, it is necessary to perform many experiments to obtain n development correction coefficients K and store them in the development correction coefficient storage circuit 109.

[0010]

In the conventional automatic developing method described above, the development correction coefficient K corresponding to the exposure area of each product needs to be obtained in advance through a number of experiments. In the case of small-volume production, the operation of obtaining the development correction coefficient K becomes complicated, and the development correction coefficient K in the development correction coefficient storage circuit must be selectively set each time a product to be developed is changed. However, there is a problem that is reduced.

Therefore, the present invention eliminates the need to previously obtain the development correction coefficient K by experiment for each product by calculating the development correction coefficient K corresponding to the exposure area. The purpose is to significantly improve productivity.

[0012]

SUMMARY OF THE INVENTION As means for achieving the above object, an object is to provide an automatic developing method and an automatic developing apparatus described below. 1. The exposed substrate is immersed in a developing solution filled in a developing tank, and a current flowing between the substrate and an electrode immersed in the developing solution in the developing tank is detected during development, and based on the detection result, In the automatic developing method for performing development with the calculated developing time, the absorption energy distribution function E of the exposure area in the actual mask exposure represented by the equations (2) and (3) and the resist represented by the equation (5) Melting boundary energy Ek
Is determined, the value of the variable δk is obtained using the equations (2), (3) and (5), and the value of the variable δk is substituted into the equation (5) to obtain the resist melting boundary energy Ek.
Is calculated, and the value of the resist dissolution boundary energy Ek is substituted into Expression (4) to obtain a development correction coefficient K of the development time. An automatic development method comprising performing development.

(Equation 6) 2. A developing tank filled with a developing solution for immersing the exposed substrate, and an electrode immersed in the developing solution of the developing tank,
Detecting a current flowing between the substrate and the electrode during development,
In an automatic developing apparatus that performs development by calculating a development time based on the detection result, an absorption energy distribution function E of an exposure area in actual mask exposure expressed by Expressions (2) and (3) and Expression Assuming that the resist melting boundary energy Ek indicated by (5) is the same, the value of the variable δk is obtained using the equations (2), (3) and (5), and the value of the variable δk is calculated by the equation (5). A development correction coefficient calculating means for calculating the value of the resist melting boundary energy Ek by substituting the value into the equation (5), and calculating the development correction coefficient K of the development time by substituting the value of the resist melting boundary energy Ek into the equation (4); A developing correction coefficient storing means for storing the developing correction coefficient calculated by the developing correction coefficient calculating means; and a developing correction coefficient based on the developing correction coefficient supplied from the developing correction coefficient storing means. An automatic developing device provided with a developing time calculating means for correcting and outputting a time interval.

(Equation 7)

[0013]

DESCRIPTION OF THE PREFERRED EMBODIMENTS First, a method for calculating a development correction coefficient K corresponding to an exposure area will be described. Generally, when a fine pattern is formed on a substrate such as a semiconductor mask or a wafer, exposure is performed using a particle beam such as an electron beam.
The energy of the electron beam during this exposure is absorbed by the resist on the substrate. Then, when a single beam is irradiated on the mask, the absorbed energy distribution Es in the resist is expressed by the following equation.
It is generally known that it can be obtained by the energy distribution function of (1) (Ref .; Journal of Vacuum Scie
nce & Technology P.1271, Vol12, No.6 1975).

[0014]

(Equation 1)

The first term of the equation (1) indicates an energy distribution due to forward scattering of the electron beam,
The next section shows the energy distribution due to backscattering. It is generally known that the energy distribution function E of the exposure area in actual mask exposure can be obtained by Expression (2) (Reference: Journal of Vac
uumScience & Technology P.1271, Vol12, No.6 1975
). The energy distribution function E is an energy sum in which n absorption energy distributions Es are arranged two-dimensionally at a pitch equal to the diameter of the exposure beam and equal to the exposure area (design data).

[0016]

(Equation 2)

The electron beam collides with atoms in the solid when it enters the fixed state, and proceeds while changing (scattering) its traveling direction until it loses energy. At this time, 90
Those having a small scattering angle of less than degrees are called forward scattering (forward scattering), and those having a scattering angle of about 180 degrees are called back scattering (backscattering). Since the scattering angle increases as the atomic number of the solid increases, the electron beam travels straight without scattering in the resist composed of light elements such as C, H, and O (forward scattering). In a substrate containing a heavy element such as Cr, Si, or the like, the light is scattered in the lateral direction (back scattering).

In the equation (1), α, β, and δ are constants easily obtained as follows. Since α, which is the standard deviation of the beam distribution of forward scattering, is almost equal to the beam system of the electronic writing apparatus, it can be measured by a beam diameter measuring device provided in the electronic writing apparatus. The standard deviation β of the beam distribution of back scattering can be obtained by a simple experiment as described below. For example, a test pattern in which about 10 lines illuminated with an electron beam with a design width of 1 μm are arranged and the space between the lines is changed (0.5, 1..
0, 2.0, 4.0, 6.0 μm, etc.), and when the space becomes smaller, β of the irradiation beam of the adjacent line becomes smaller.
By taking advantage of the fact that the line width increases under the influence of
Let β be the space width that is a critical point that is not affected.

The back scattering coefficient δ is, for example, L1 when the space width is S1 and L2 when the space width is S2 in the test pattern described above. At this time, the line edge positions of these line widths L1 and L2 are respectively represented by x
Assuming that the energy at the line edge portion is equal to 1, x2, Es (x1) = Es (x2). Therefore, δ can be obtained by substituting this relational expression into equations (1) and (2).

As described above, α, β, and δ can be easily obtained. Α is a beam diameter, and β and δ are affected by a substrate material, a resist material, development conditions, and the like. , The same coefficient can be used even if the mask pattern is changed.
Therefore, the measurement itself can be reduced, and the number of man-hours is much smaller than the conventional case where the development correction coefficient K is obtained by experiment.

When calculating the pattern line width based on the energy distribution function E obtained from the equations (1) and (2), when the development correction coefficient K is fixed, the calculated result agrees with the actually measured value. , Good results can be obtained. However, when the development correction coefficient K is changed, the calculation result deviates from the actually measured value, and it is difficult to calculate the pattern width with high accuracy. Therefore, in order to calculate a highly accurate pattern width even when the development correction coefficient K changes, an equation used in place of equation (1) is used.
(3) is newly proposed.

[0022]

(Equation 3)

Δk in the equation (3) is a variable that changes with the development correction coefficient K, and other constants are common to the equation (1).
Then, by incorporating this variable δk, it is possible to calculate the pattern width with high accuracy as described later. Hereinafter, a description will be given including how to obtain the variable Δk.

FIGS. 2 and 3 show the energy distribution functions of the resist obtained from the equations (3) and (2). In addition, in order to simplify the display of each figure, the pattern width is displayed only in the x-axis direction, and therefore, the energy is displayed as E (x). And Figure 2 is 1um Line & 1um Space, Figure 3 is 1u
This is a test pattern of m Line & 6 um Space. The vertical axis indicates the energy absorption E (x) of the resist, and the horizontal axis indicates the distance (line width) in the x-axis direction. Line is the exposure unit,
Space means an unexposed portion, and two beams having a diameter of 0.5 μm are used to expose a 1 μm line.

The structure of the test pattern of FIG. 2 has a larger exposure area with respect to the substrate area than that of FIG. 3, so that the electron scattering amount in FIG. 2 is larger than that in FIG. It can be seen that the absorption peak of increases. In addition, the dissolution of the exposed portion of the resist in the developing solution during the development of the resist progresses from a portion having a high absorption energy distribution to a portion having a low absorption energy as the development time elapses. For example, in FIG. 2 and FIG.
Assuming that the resist dissolution boundary energy Ek corresponding to = 1.8 is 2.50, the line width at that time can be obtained as the distance between the energy distribution function and the x coordinate of the intersection of Ek = 2,50.

When the relationship between the development correction coefficient K during the development and the resist melting boundary energy Ek corresponding to the development time (K × BTT) was examined, log (Ek) and log
(K) was found to be approximately proportional. Therefore, these two variables are expressed by a relational expression such as Expression (4), and it has been found that linear regression is possible.

Log (Ek) = a × log (K) + b (a is negative and b is positive constant) Equation (4)

As is apparent from the equation (4), as the development proceeds (K increases), the resist dissolution boundary energy Ek of the dissolving portion of the resist decreases, and accordingly, the equation (3) δk is also changing.
When the relationship between Ek and .delta.k was examined, it was found that this relationship was also approximately proportional, and that linear regression was possible using a relational expression such as equation (5).

E k = c × δ k + d (c and d are positive constants) Equation (5)

When the development correction coefficient K is determined from the equation (4), the resist melting boundary energy Ek corresponding to the development correction coefficient K is obtained. Then, by substituting Ek into Equation (5), δk is obtained. By substituting δk into Equation (3), the absorption energy distribution function E (x, y) or E (x, y) is obtained from Equations (3) and (2). E (x) is obtained. Further, by setting E (x) = Ek and solving the energy distribution function, the finished pattern width can be obtained.

From the above, by combining the previously known equation (2) with the equations (3), (4) and (5) newly proposed in the present invention, the completed pattern width can be calculated from the development correction coefficient K. Calculation can be performed, or conversely, an appropriate development correction coefficient K corresponding to the exposure area (K at which the designed value and the finished pattern width become equal) can be obtained by calculation.

Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings. Here, FIG. 1 is a schematic configuration diagram of a resist automatic developing apparatus according to an embodiment of the present invention, and FIG. 4 is a graph showing a relationship between a development correction coefficient K and a pattern width. The automatic developing device 1 includes a developing tank 3 filled with a developer 2, a support 4 provided in the developing tank 3 so as to be movable up and down, and a support 4.
Platform elevating device 5 for elevating and lowering the developer,
A substrate 8 having an electrode 6 immersed therein and coated with a resist 7 and exposed is placed on a support 4 and immersed in a developer 2 to perform development. The BTT (Break Throu) is monitored by monitoring the temporal change of the current flowing between the substrate 8 and the electrode 6 immersed in the developing solution 2 in the developing tank 3 from the start of development.
gh Time) is determined by the BTT determination circuit 9.

In addition, TDT (Total Development T)
ime) The calculation circuit 10 calculates TDT = K × based on the determination result of the BTT determination circuit 9 and the development correction coefficient K from the development correction coefficient storage circuit 11 in which the development correction coefficient K is stored in advance.
This is a circuit for calculating an appropriate development time TDT using a BTT calculation formula. Then, an appropriate development time T from the start of development
After the elapse of the DT, the support platform elevating device 5 is raised to lift the substrate 8 from the developer 2.

Here, the development correction coefficient calculation circuit 12 will be described. The development correction coefficient calculation circuit 12 automatically calculates an appropriate development correction coefficient K corresponding to the exposure area or the design pattern data based on the above equations (2), (3), (4), and (5). It is constituted by a PC (Personal Computer) having a program that can calculate the data. And
Exposure area required for calculation by the development correction coefficient calculation circuit 12,
Alternatively, the design pattern data can be directly input by an operator or can be transferred from a pattern design apparatus such as a CAD (Computer Aided Design).
The appropriate development correction coefficient K obtained by the development correction coefficient calculation circuit 12 is transferred to the development correction coefficient storage circuit 11, and automatic development is performed based on the development correction coefficient K in the same manner as in the conventional example. FIG. 4 shows the result of calculation performed by the development correction coefficient calculation circuit 12 on the relationship between the development correction coefficient K and the pattern line width.

FIG. 4 shows a comparison between the relationship between the development correction coefficient K and the pattern width calculated by the apparatus of the present invention and the completed pattern width (actually measured value) of the semiconductor mask developed by the apparatus of the present invention. It is a graph. There are two types of masks, a and b. Mask a is 1um Line & 1um Space (Line is an exposed part, Space is unexposed part), and mask b is 1um Line &
A pattern of 6um Space (Line is an exposed portion, Space is an unexposed portion) is laid out on the mask. The mask line width when the development correction coefficient K for mask a changes from 1.5 to 3.7 and the mask b changes from 1.8 to 3.7 at 0.1 intervals (Line width of 1 um design) ) Obtained by calculation, and the finished line width of the mask actually processed by development with the same development correction coefficient K (Li
ne width).

From the graph of the calculation result shown in FIG.
In the case of ine & 1um Space, the appropriate development correction coefficient K (K value at which the finished line width of the designed 1um line width is 1um) is 1.9, and in the case of 1um Line & 6um Space, the appropriate development correction coefficient K is 2.3. Can be expected. Further, it can be seen that the actually measured value at the same development correction coefficient K is almost the same as the expected value obtained from the calculation. Then, the development correction coefficients K of all the calculated areas (1.5 ≦ K ≦ 3.7)
Also, the difference between the calculated value and the actually measured value is in the range of ± 0.07 μm, so the calculation result by this device is:
It is clear that the results are in good agreement with the measured values.

[0037]

As described above, according to the present invention, an appropriate development correction coefficient is obtained by calculation, so that the development correction coefficient of the development time is previously obtained for each product as in the prior art. It is possible to obtain a high-precision pattern while eliminating the need for a preparatory operation, and in particular, there is an effect that the productivity of products of many kinds and small lots is remarkably improved.

[Brief description of the drawings]

FIG. 1 is a schematic configuration diagram showing one embodiment of an automatic developing device of the present invention.

FIG. 2 is a graph showing an energy distribution function obtained by calculation.

FIG. 3 is a graph showing an energy distribution function obtained by calculation.

FIG. 4 is a graph showing a relationship between a development correction coefficient K and a pattern width.

FIG. 5 is a schematic configuration diagram showing a conventional automatic developing device.

FIG. 6 is a waveform diagram showing a current flowing between a substrate and an electrode.

7 (a) to 7 (c) are explanatory diagrams showing a development progressing process used for explaining FIG. 6;

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 ... Automatic developing device 2 ... Developer 3 ... Developing tank 4 ... Support base 5 ... Support base elevating device 6 ... Electrode 7 ... Resist 8 ... Substrate 9 ... BTT discrimination circuit 10 ... TDT calculation circuit 11 ... Development correction coefficient storage circuit 12 ... Development correction coefficient calculation circuit 13 ... CAD

Claims (2)

(57) [Claims] The method comprises the steps of: immersing an exposed substrate in a developing solution filled in a developing tank; detecting a current flowing between the substrate and an electrode immersed in the developing solution in the developing tank during development; In an automatic development method for performing development with a development time calculated based on a detection result, an absorption energy distribution function E of an exposure area in actual mask exposure, which is expressed by Expressions (2) and (3), is used in advance.
And the resist melting boundary energy Ek shown by the equation (5) is the same, the value of the variable δk is obtained using the equations (2), (3) and (5). Into the equation (5) to determine the value of the resist melting boundary energy Ek. Substituting the value of the resist melting boundary energy Ek into the equation (4) to determine the development correction coefficient K for the developing time. Wherein the image is developed with a development time corrected based on the development correction coefficient K. (Equation 4) 2. A developing tank filled with a developing solution for immersing an exposed substrate, and an electrode immersed in the developing solution in the developing tank, wherein a current flowing between the substrate and the electrode during development is supplied. In an automatic developing apparatus that detects and calculates a development time based on the detection result and performs development, an absorption energy distribution function E of an exposure area in an actual mask exposure represented by Expressions (2) and (3) is determined in advance.
And the resist melting boundary energy Ek shown by the equation (5) is the same, the value of the variable δk is obtained using the equations (2), (3) and (5), and this variable δk
Is substituted into equation (5) to determine the value of the resist melting boundary energy Ek.
And a development correction coefficient calculating means for calculating the development correction coefficient K for the development time by substituting the value of the expression (3) into the expression (4); and a development correction coefficient storing the development correction coefficient calculated by the development correction coefficient calculation means. An automatic developing apparatus comprising: a storage unit; and a development time calculation unit that corrects and outputs the development time based on the development correction coefficient supplied from the development correction coefficient storage unit. (Equation 5)