JP2005057037A - Calculation method of amount of resist shrink - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To calculate a real shrink amount which is not affected by an outer environment such as length measurement SEM chamber atmosphere and the state of a substrate. <P>SOLUTION: A relation between the number of measuring times of resist pattern width and a variation of resist pattern width is guided. The relation between the number of measuring times of resist pattern width and the variation of resist pattern width by contamination is guided from the variation of resist pattern width after the prescribed number of measuring times, in which the variation of resist pattern width with respect to the number of measuring times of resist pattern width is almost constant. Then, the relation between the number of measuring times of resist pattern width and the variation of resist pattern width by resist shrink is guided from the relation between the number of measuring times of resist pattern width and the variation of resist pattern width, and the relation between the number of measuring times of resist pattern width and the variation of resist pattern width by contamination. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

  The present invention relates to a method for calculating a resist shrink amount. Specifically, the present invention relates to a method for calculating a resist shrink amount by irradiating a resist pattern with an electron beam by a length measuring SEM.

  It is known that when a resist pattern is irradiated with an electron beam by a length measuring SEM, the resist shrinks due to a cross-linking phenomenon between the resins constituting the resist. Conventionally, the same resist pattern is continuously measured and the measurement is performed. A method of obtaining a resist shrink amount corresponding to the number of times is performed.

However, when the same resist pattern is continuously measured, the resist pattern lines generated when the contamination in the SEM chamber or the contamination generated from the substrate adheres to the resist pattern simultaneously with the shrinking phenomenon of the resist. A thickening phenomenon has also occurred.
In other words, the conventional shrink amount measurement described above calculates the resist shrink amount including the line width thickening due to contamination. If an environment or material with much contamination is used, the true shrink amount is obtained. There was an inconvenience that it was difficult to find.

  The present invention was devised in view of the above points, and is a resist shrink amount calculation for obtaining a true shrink amount that is not affected by the environment such as the atmosphere of the length measurement SEM chamber or the state of the substrate. It is intended to provide a method.

  In order to achieve the above object, the resist shrink amount calculation method of the present invention measures the resist pattern width a plurality of times by a length measurement SEM, and shows the relationship between the number of times the resist pattern width is measured and the amount of change in the resist pattern width. The resist pattern width is determined by the number of times the resist pattern width is measured and the number of contaminations, and the amount of change in the resist pattern width after the predetermined number of measurements is approximately constant. From the step of deriving the relationship of the change amount of the resist pattern, the relationship between the number of times the resist pattern width is measured and the change amount of the resist pattern width, and the relationship between the number of times the resist pattern width is measured and the amount of change in the resist pattern width due to contamination Width measurement times and resist shrink resist And a step of directing the variation of the relationship between the turn width.

  Here, the resist pattern width measurement count and the resist pattern width change amount, and the resist pattern width measurement count and the resist pattern width change amount due to contamination, the resist pattern width measurement count and the resist shrink resist By deriving the relationship between the change amounts of the pattern width, it is possible to calculate an objective and accurate resist shrink amount that is not influenced by external factors.

  In the method for calculating the resist shrink amount of the present invention, the true shrink amount can be calculated without being influenced by the environment such as the atmosphere of the length measurement SEM chamber or the state of the substrate.

Hereinafter, embodiments of the present invention will be described with reference to the drawings to provide an understanding of the present invention.
FIG. 1 shows a flowchart of an example of a method for calculating a resist shrink amount to which the present invention is applied. First, a resist pattern is prepared in an example of a method for calculating a resist shrink amount to which the present invention is applied. That is, a resist pattern for calculating the shrink amount is produced on the silicon wafer.

  Here, although the resist pattern base, pattern shape, pattern formation conditions, etc. are arbitrary, the resist shrink amount has a strong correlation with the pattern formation conditions. It is preferable to combine them.

  Next, an arbitrary portion of the produced resist pattern is continuously measured using a length measuring SEM until the amount of increase in resist pattern width (hereinafter referred to as resist line width) becomes constant.

Here, the electron beam irradiation conditions, image processing conditions, length measurement value calculation method, etc. during length measurement may be arbitrarily set, but these conditions must all be performed under the same conditions during continuous length measurement. is there.
In addition, although the behavior of the resist line width due to continuous length measurement varies depending on the resist characteristics, as a common resist line width behavior, the resist line width decreases from the start of length measurement to a predetermined change point. Subsequently, the resist line width after the predetermined change point continues to increase.

  In addition, the behavior of the resist line width by performing the continuous length measurement described above is a case where a resist line is selected as the length measurement target using the length measurement SEM, and the resist space or the hole resist pattern is the length measurement target. When is selected, the resist line width continues to increase from the start of length measurement to a predetermined change point, and the resist line width after the predetermined change point continues to decrease.

Next, a graph as shown in FIG. 2 is created from the relationship between the number of measurements obtained by continuous length measurement by the length measurement SEM and the line width of the resist.
Here, it is clear that the resist shrinkage is caused by the cross-linking phenomenon between the resins constituting the resist. Since the resist shrinkage is saturated, the influence of the resist shrinkage in the region indicated by a in FIG. In addition, it can be considered that the line width of the resist is changed only by the influence of the line width thickening due to contamination.

Next, out of the length measurement values obtained by the continuous length measurement by the length measurement SEM, the length measurement value of the area indicated by the symbol a in FIG. 2 is extracted, and only the influence of the line width thickening due to contamination is used for the resist. The relationship between the number of measurements and the line width of the resist when the line width changes is derived.
Hereinafter, a specific method for deriving the relationship between the number of measurements and the line width of the resist when the line width of the resist is changed only by the influence of the line width increase due to contamination will be described.

In order to derive the relationship between the number of measurements and the line width of the resist when the line width of the resist changes only due to the influence of the line width thickening due to contamination, first, all the points after the inflection point indicated by symbol b in FIG. The first-order approximation of the measured value is performed, and the correlation function at that time is calculated. Next, a linear approximation is performed in the same manner except for one length measurement value closer to the inflection point, and a correlation function at that time is calculated. Similarly, remove the length measurement value near the inflection point by 2 points, 3 points,..., Perform a linear approximation and calculate the correlation function at that time. Guide the relationship of the number of relationships.
From the relationship between the length measurement value removal number derived as described above and the correlation coefficient, an area indicated by reference sign c in FIG. 3 in which the correlation coefficient is substantially constant is obtained, and the correlation coefficient is substantially constant. 2, that is, the region indicated by the symbol a in FIG. 2, the relationship between the number of measurements and the line width of the resist when the line width of the resist is changed only by the influence of the line width thickening due to contamination by the linear approximation method. Lead.

  Here, in the above, when the line width of the resist is changed only by the relationship between the number of times of measurement of the region indicated by symbol a in FIG. The relationship between the number of measurements and the line width of the resist is derived. However, as long as these relationships can be derived, it is not always necessary to use the first-order approximation method. .

  Thereafter, the line width thickening component due to contamination is subtracted from the length measurement value obtained by continuous length measurement using the length measurement SEM described above, and the number of measurements and the length measurement value after correction are as shown in FIG. It is possible to derive a relationship between resist shrink amounts that are not affected by line width increase due to contamination.

In an example of the method for calculating the resist shrink amount to which the present invention is applied, the resist shrink amount without the influence of the line width increase due to contamination can be calculated, and accurate resist evaluation can be performed.
In other words, the amount of resist shrink that is not affected by line width increase due to contamination that depends on the external environment such as the atmosphere of the SEM chamber and the state of the substrate can be calculated. This makes it possible to carry out resist evaluation that is objective and accurate.

  Further, by quantifying the resist shrink amount, it is possible to accurately calculate the actual finished pattern line width from the resist shrink amount of the line width monitor in the semiconductor manufacturing process. In particular, since the number of times of length measurement of the resist pattern by the length measurement SEM is usually one, it is possible to calculate the actual finished pattern line width from the resist shrink amount of the line width monitor when the length measurement number is one. it can.

  Further, the length measurement SEM apparatus can be evaluated by quantifying the resist shrink amount. That is, when the resist shrink amount is quantified, the line width thickening amount due to contamination when the resist pattern is measured by the length measurement SEM is calculated. The length measurement SEM can be evaluated based on the above.

It is a process flow of the calculation method of the resist shrink amount to which the present invention is applied. It is a graph which shows the relationship between the frequency | count of a measurement, and the line | wire width of a resist. It is a graph which shows the relationship between the measured value removal number and a correlation coefficient. It is a graph which shows the relationship between the frequency | count of a measurement, and a resist shrink amount.

Claims (3)

A step of measuring the resist pattern width a plurality of times by the length measurement SEM, and deriving a relationship between the number of times the resist pattern width is measured and the amount of change in the resist pattern width
From the amount of change in the resist pattern width after a predetermined number of times that the change amount of the resist pattern width with respect to the number of times of measurement of the resist pattern width is substantially constant, the amount of change in the resist pattern width due to the number of times of resist pattern width measurement and contamination The process of guiding the relationship;
From the relationship between the resist pattern width measurement count and the change amount of the resist pattern width and the relationship between the resist pattern width measurement count and the resist pattern width change amount due to contamination, the resist pattern width measurement count and the resist pattern by resist shrink And a step of deriving a relationship of the amount of change in width. The relationship between the resist pattern width measurement count and the amount of change in the resist pattern width due to contamination is the same as the resist pattern width measurement count after the inflection point in the relationship between the resist pattern width measurement count and the resist pattern width change amount. The method for calculating the resist shrink amount according to claim 1, wherein the long value is derived using a first-order approximation method. The method further comprises a step of deriving a change amount of the resist pattern width due to the resist shrink when the number of measurement times is 1, based on a relationship between the number of measurement times of the resist pattern width and the change amount of the resist pattern width due to the resist shrink. The method for calculating the amount of resist shrink according to claim 1 or 2.

Images