DE102011113940A1 - Method for determining dose alterations for adapting e.g. diameter of contact holes of mask for manufacturing semiconductor component, involves determining alterations as variations in intensity values from extreme intensity value - Google Patents

Abstract

The method involves setting a reference value of structural size of a mask (5). An aerial photograph of the mask is recorded. Intensity values within the aerial photograph, at which the structural size corresponds to the reference value, are determined. Dose alterations are determined as variations in the intensity values from an extreme intensity value. Transmission or reflection of the mask is changed in accordance with the determined dose alterations. Dose is adjusted based on the determined dose alterations when performing a wafer exposure using the mask. An independent claim is also included for a mask inspection microscope.

Description

This patent relates to a method for determining dose changes for adapting feature sizes of a mask.

This patent also relates to a mask inspection microscope, which has a computing unit for performing the method.

In the manufacture of semiconductor devices, structures are formed on wafers in a lithographic process. These structures are first shown on a mask (photomask, reticle), for example by chrome on quartz glass. Other masks are also known, for example phase-shifting masks (PSM) or reflective masks, which are used in particular for short-wavelength illumination radiation in the EUV range. By exposure in a scanner, this structure is displayed on a resist-coated wafer. Subsequent development and etching yields the desired pattern on the wafer.

The smallest structure sizes that can still be represented in a mask are also called critical dimension, critical dimension. In most cases, the critical dimension of a mask refers to the line width of a structure consisting of lines and spaces (spaces and spaces). The critical dimension can for example also be related to the edge lengths of contact holes. The critical dimension is given for a mask and often serves to characterize the mask itself or to characterize wafer exposures made with the mask.

The measurement of the critical dimension constancy across the face of a mask or an exposed wafer (Critical Dimension Uniformity, CDU) is one way to characterize it.

The determination of structure sizes, for example the CDU of exposed wafers, takes place for example by scanning electron microscopy (SEM). To determine feature sizes from aerial images, a calibrated mask inspection microscope can be used.

If the deviations from structural quantities, such as the CDU, exceed a permissible error, the critical dimension must be corrected. To correct the critical dimension of a mask, a method may be used in which the transmission of a mask is changed. In areas with a critical dimension to be corrected, far-spot laser lasers produce density variations in the mask's substrate, which reduces the transmission of the area corresponding to the error. Such a method is for example in the US 7736819 disclosed. Even with reflective masks, the correction of the critical dimension by the introduction of density variations or inhomogeneities in the reflective layers of the mask by electron radiation is possible, as in DE 10 2011 080 100 disclosed. By introducing density variations or inhomogeneities, the reflectivity of the masks is reduced in the case of reflective masks. The reflectivity of a mask is the quotient of reflected and incident intensity of the illumination radiation, it is also referred to as reflectance, reflectance or reflectance.

Besides the correction of the mask, the variation of the dose of the illumination radiation during the exposure of the wafer is an additional or alternative possibility for optimizing feature sizes such as, for example, the CDU. This variation of the illumination radiation can take place, for example, by using a can mappers or a CDU optimizer from ASML.

In both methods, both in the correction of the transmission or reflection of the masks and in the variation of the dose of the illumination radiation, the determination of the dose changes for adaptation of feature sizes is applied. The dose changes to be determined can be used both to correct the transmission or reflection of the masks and to vary the dose of the illumination radiation.

The determination of the dose changes of a mask takes place by the evaluation of test measurements. Several wafer exposures, each with a different dose, can be made. Structure sizes can also be determined with a mask inspection microscope, wherein the determined dose changes can be used to correct masks, as in US Pat US 20090191475 disclosed. The dependence of the dimensions of the feature sizes on the dose change varies with the degree of coherence of the illumination radiation, ie with the illumination setting used in the exposure. Therefore, mask inspection microscopes are used. Aerial images taken with a mask inspection microscope largely show the features which are also imaged on the wafer in the case of wafer exposures. The same lighting settings are used as are used for wafer imaging with a scanner. The numerical aperture (NA) of the imaging optics of known mask inspection microscopes coincides mask-wise with that of a scanner for wafer exposure.

In the case of the methods mentioned, the change of feature sizes from the wafer exposures or the aerial images can be determined as a function of the dose change. Since the correlation is approximately linear, a factor CD _delta / A can be determined, where the dose change is denoted by A.

The determination of the dose changes with the mentioned methods is complicated and time-consuming.

The object of the invention is to provide a rapid and cost-effective method for determining the dose changes with high measurement accuracy and an apparatus for carrying out the method.

• – Vorgeben eines Sollwertes CD_s einer Strukturgröße CD der Maske,
• – Aufnehmen zumindest eines Luftbildes der Maske,
• – Ermitteln von Intensitätswerten I innerhalb des zumindest einen Luftbildes, an welchen die Strukturgröße CD dem Sollwert CD_s entspricht,
• – Ermitteln der Dosis-Änderungen A als Abweichungen der Intensitätswerte I von einem Intensitätswert T.

According to the invention, this object is achieved by a method for determining dose changes for adapting feature sizes of a mask, comprising the steps:

• Predetermining a target value CD _{s of} a feature size CD of the mask,
• Picking up at least one aerial image of the mask,
• Determining intensity values I within the at least one aerial image at which the structure variable CD corresponds to the setpoint value CD _s ,
• Determining the dose changes A as deviations of the intensity values I from an intensity value T.

In the method according to the invention, the determination of dose changes is made possible by means of a single series of aerial images for a mask. For each area to be considered, only one aerial photograph is to be taken. The dose changes are determined directly from the aerial photographs. It is not necessary to calculate the difference CD _{delta of} the structure variable CD from the nominal value CD or the factor CD _delta / A. Elaborate reference measurements are avoided.

Structure sizes can be any dimensions of structures present on the mask or of parts of these structures. These are, for example, line widths, diameters or edge lengths of contact holes. Structure sizes can also refer to the smallest representable size, the critical dimension, of a mask.

The aerial image of a mask is an image with graded intensity values, it can also be called a grayscale image. Instead of the discrete structures of the mask to be imaged, in the aerial image between light-dark transitions, i. H. at edges, continuous intensity curves shown. Only by the wafer exposure of the resist-coated wafer is a good approximation discrete image of a structure obtained after the development of the resist.

For evaluating the aerial images, intensity values (also referred to as thresholds) are determined or predefined, in which discrete values of a feature size of a respective structure shown in the aerial image are determined.

The specification of an intensity value for the evaluation is also a simple resist model. It consists of the simple assumption that areas in the resist on the mask in which the selected intensity values are exceeded or undershot during the exposure take place, the respective reactions in the resist.

When a target value of a feature size CD _s is _specified , the intensity value I at which this feature size is present is determined. In this case, an iterative method is used according to the invention. In this case, structural quantities CD are determined for given intensity values. The intensity values are then iteratively varied until the value of the structure variable CD corresponds to the predetermined target value CD _s .

As a reference value T for the determination of the dose changes A, a corresponding intensity value is predetermined or determined depending on the further use of the data. By way of example, the mean value between maximum and minimum intensity of the measured mask can be specified as reference value T. Further possible values are discussed in the measures below.

For evaluation, the average value can be formed over a complete aerial image to determine a feature size CD. It is also possible to take into account only one section (or several sections) of the at least one aerial image during the evaluation.

In a variant of the method, the intensities of the aerial images are normalized. It is advantageous to set the intensity of an unstructured mask (I _clear ) to 1 or to 100% and to relate the measured intensities to this value, ie to divide all measured values of intensity by I _clear (also referred to as clear normalization).

The dose change A can be calculated as the difference between the values I and T.

In a further embodiment of the invention, the transmission or reflectivity of the mask is changed in accordance with the determined dose changes A.

In a further embodiment of the invention, when performing a wafer exposure using the mask, the dose is adjusted in accordance with the determined dose changes A.

The two above measures have the advantage that an optimization of the structure sizes is possible in a simple manner. The measures can be applied individually or in succession. Thus, after a first determination of dose changes A, the transmission or reflectivity of the mask can be corrected. A subsequent second determination of remaining dose changes may be used to adjust the dose in a wafer exposure.

In a further embodiment of the invention, the intensity values are determined with subpixel accuracy.

Conventionally used mask inspection microscopes use aerial images with a detector of a matrix of 1000 x 1000 pixels. This resolution is not sufficient to determine the feature sizes CD. Therefore, a method for sub-pixel accurate determination of the positions of edges in aerial images is used. Such a method is in the DE 10 2009 015 594 disclosed. The subpixel-exact determination of the edge positions enables a precise determination of the dimensions of the feature sizes CD and thus a subpixel-accurate determination of the intensities to the desired values of the feature size.

In a further embodiment of the invention, an extreme value of the intensity values I is determined for the intensity value T.

This measure is particularly advantageous in connection with the abovementioned measure of changing the transmission or reflectivity of a mask. The illumination radiation used in the wafer exposure is scattered by these density variations. This results in a reduction of the transmitted or reflected illumination radiation. Only a reduction in the transmission or reflectivity of a mask is possible, but no increase. In cases where the structure size CD is determined over bright areas of the aerial image, the minimum value of the intensities I is determined as the extreme value. The minimum of the intensities I then corresponds to the largest value of the feature size CD. This can be reduced by lowering the transmission or reflection. In cases where the feature sizes are determined over dark areas, a maximum of the intensities is determined as the extreme value. The determined dose changes A are in both cases a direct measure of the necessary changes in the transmission or the reflectivity of a mask, which lead to a corrected CD.

In a further embodiment of the invention, the dose changes A are determined as A = I - T / T ,

The dose changes A are calculated in this measure as relative deviations of the intensities from the reference value T. This measure has the advantage that the values of the dose changes are available for further use independently of the absolute values of the intensities.

In a further embodiment of the invention, all structural quantities CD _T and differences CD _delta of the nominal value CD _{s are} determined for the intensity value _T.

As explained above, an advantage of the method according to the invention is that the difference CD _delta and the factor CD _delta / A need not be determined. However, it may be desirable to display the differences CD _delta graphically (so-called CDU map), so as to illustrate the constancy of the CD. The determination of the differences CD _{delta also} allows the calculation of the factor CD _delta / A.

The CD _delta and CD _delta / A factors may be required to drive a can mappers or a CDU optimizer.

If the factor CD _delta / A is known for a mask or for selected structure types of a mask, the correction of the mask itself or the correction of the dose can be carried out in a simple manner as soon as the critical dimension of a mask is known. For further measurements, it is then sufficient to determine the critical dimension CD of a similar mask with a simpler method, for example by measuring the transmission, as in the US Pat WO 2009 007977 A2 disclosed.

In a further embodiment of the invention, positions P are predefined on the mask at which measurements are to be performed within measuring ranges B.

This measure has the advantage that an efficient implementation of the method is made possible. For example, the positions of the measurement areas may be distributed over the structure of the mask in a uniform grid. However, it is also possible to select measuring ranges which are particularly suitable for measuring or characterizing a mask. These can be areas of a specific texture type, such as lines and spaces, or areas that are particularly prone to error.

The measuring ranges are to be selected in the dimensions so that the measurement can be evaluated with the desired accuracy, but the calculation time required for the evaluation does not become too great.

Within an aerial image, a single measuring range can be specified. This area may include all or part of the aerial view. Within an aerial image, several measuring ranges can also be specified.

The invention further comprises a mask inspection microscope for carrying out the method comprising imaging optics for imaging a mask, a light source, a detector for receiving at least one aerial image of the mask, a computing unit for carrying out the method according to the invention.

The invention will be described and explained in more detail below with reference to a few selected embodiments and with reference to the drawings.

Show it:

1 the construction of a mask inspection microscope;

2 : a diagram illustrating the method for determining an intensity value to a target value CD of a feature size;

3 Fig. 3 is a diagram showing an example of the method for determining dose changes A;

4 : Representation of a first aerial image of a mask;

5 : Representation of a first intensity distribution of a region of the first aerial image;

6 : Representation of a second aerial image of a mask;

7 : Representation of a second intensity distribution of a region of the second aerial image.

To carry out the process, use a mask inspection microscope 1 Aerial pictures taken.

The construction of a mask inspection microscope 1 is determined by 1 explained. The mask inspection microscope 1 has a mask holder 10 on, on which the mask to be imaged 5 rests on and designed as a CCD chip (Charged Coupled Device) detector 20 , A light source 25 Illuminates the mask 5 via a lighting optics 30 forming a pupil plane 35 having. Illumination settings can be made via a pupil filter, which is at the pupil level 35 is arranged and a polarizer 36 be set. When taking the aerial pictures of the mask 5 with the detector 20 are adapted to the structure of the adjusted illumination settings and polarization settings, which are also used in the wafer exposure with a scanner.

An aerial view of the mask 5 is about the imaging optics 15 , with the optical axis 2 , in the plane of the detector 20 generated. The numerical aperture (NA) of the imaging optics 15 On the mask side, it coincides with that of a scanner for wafer exposure, a common value being NA = 0.2325. Focusing is the imaging optics 15 in the direction perpendicular to the XY plane, as the Z direction along the optical axis 2 , designated, moved. The aerial picture is taken from the arithmetic unit 40 read out, which is designed as a computer. The aerial image initially exists as a data structure in the computer's RAM. This can be saved as a graphics file on the hard disk of the computer. The data structure or graphic file is a two-dimensional matrix, which is made up of pixels. The intensities of the pixels are represented by numerical values, for example, in a range of 0 to 65535. The image field on the mask 5 is square, with an edge length of 10 μm. The section of the recorded substructure is determined by the image field. The magnification is for example 450: 1.

At the beginning of the procedure, the necessary parameters are specified in an input file. This input file is processed by the arithmetic unit 40 and taken into account in the implementation of the procedure. The input file includes: the positions P on the mask, dimensions and orientation of the measurement areas 51 , the target value CD _{s of} the structure variable to be determined, a reference value T or the instruction for its determination, if necessary instructions for determining CD _T , CD _delta , CD _delta / A.

As structure sizes, for example, the diameter of contact holes or the width of lines can be determined.

On the mask 5 positions P are given at which structure sizes CD are to be determined. At the positions P become measuring ranges 51 predetermined, which include at least the structure size to be measured. Within these measuring ranges 51 become the aerial photos 50 evaluated. The aerial photos 50 be with the mask inspection microscope 1 recorded for each position P, so that they are the measuring ranges 51 include.

In the 4 to 7 Two example measurements and their evaluation are shown. 4 and 6 show two aerial pictures 50 with a measuring range 51 whose boundaries are shown in white lines. As position P of the measuring range 51 the coordinates of the center of the area are given. The grayscale of 4 and 6 are a measure of the intensities of the aerial image, as indicated in the scale shown. The intensities of the aerial images are clear normalized. The measured intensities of an unstructured mask (I _clear ) are normalized to 1. The higher the value of the intensity, the brighter the grayscale.

The dimensions of a measuring area 51 are chosen such that the most advantageous evaluation possible. As in 4 and 6 illustrates, in this example, measurement range 51 chosen such that the most accurate determination of the critical dimension (CD) of the aerial image 50 is possible. Aerial view 50 is the image of an area of the mask 5 with lines and spaces (lines and spaces). The longer side 52 of the measuring range 51 runs across the lines and clearances, the shorter side 53 of the measuring range 51 runs along the lines and open spaces. For evaluation, the intensity values are directed towards the shorter side 53 an average is formed. The resulting intensity distribution towards the longer side 53 is in 5 and 7 illustrated. These intensity distributions are also referred to as intensity profiles. On the x-axis of the 5 and 7 the clear-normalized intensities are plotted, on the x-axis the dimensions correspond to the 4 and 6 ,

The structure size CD is determined either over belle range or over dark areas of the aerial image. The choice depends on whether a positive or a negative resist is used in the wafer exposure to be performed using the mask being examined. The masks for different resist types are also referred to as bright field or dark field masks. If, for example, the structures in a mask are represented by a chromium layer on quartz glass (this is also referred to as a COG mask), in the resulting aerial image the areas coated with chromium are dark and the chromium-free areas are light.

The determination of the desired CD values for a given intensity value takes place in arithmetic unit 40 as explained below. Intersections of the intensity profile with the given intensity value are searched for. In a first step, the pixel of the intensity profile is selected for this, with the intensity that comes closest to the given intensity value. The CD values are calculated from the distances of the intersections. Depending on the type of mask or the resist used, the CD value is calculated from two points of intersection which comprise an intensity maximum or an intensity minimum. Within a range 51 All CD values are determined. From all CD values of a range 51 the arithmetic mean is formed and output as the CD value of the area.

In a variant of the method, the CD values are determined with subpixel accuracy. Such a method is in the DE 10 2009 015 594 disclosed.

To the given setpoint CD _s is for the measuring range 51 an intensity value I, which is also called a threshold, determined. This value is over the measuring range 51 averaged. This evaluation will be for all measuring ranges 51 carried out. The following description of the determination of feature sizes CD refers to bright field masks, but an application for dark field masks is easily possible. First, a starting value of the intensity value to be determined is determined. This is calculated as the mean between the maximum I _max and minimum I _min intensity values of the intensity profile (see 5 and 7 ) of the area 51 calculated. Then the structure size CD is determined pixel-exactly at this starting value. The values of the structure size CD are again over the range 51 averaged. If the CD value is greater than the target value CD _s , the start value of the intensity value In is increased by a correction value K, the CD value is lower than the target value CD _s , then the starting value of the intensity value is decreased by the correction value K. , Then for the new intensity value again the structure size CD of the area 51 determined pixel-precise. The correction factor K is reduced in the second iteration. The number of iterations is denoted by n.

The calculation of the intensity values I _n for an iteration n is in equations 1 and 2 given as: I _{n + 1} = I _n · (1 + K / n) if CD _n > CD _s and 1 I _{n + 1} = I _n · (1 - K / n) if CD _n <CD _s . 2

A starting value for K is, for example, K = 0.1, K = 0.05 or K = 4.01.

The value of the feature size CD thus obtained is now used as the starting value for a further run of the said iterative method, a subpixel-accurate determination of the feature sizes CD now taking place.

This method leads to a highly accurate determination of the intensity values.

The two examples of 4 to 7 show measurements of two areas 51 a mask with a nominal value of the critical dimension CD _s = 380 nm. The evaluation of the area 51 in 4 gives the intensity profile as in 5 shown. The intensity value is determined to be I _a = 0.275931. The evaluation of the area 51 in 6 gives the intensity profile as in 7 shown. The intensity value is determined to be I _b = 0.276715.

To calculate the dose changes A, a reference value T is specified. The calculation of dose changes A is given in Equation 3. A = I - T / T 3

In a variant of the method, the values of the dose changes A become the correction of the examined mask 5 used.

In the case of transmissive masks, density variations or inhomogeneities are produced by femtosecond laser for correcting the feature sizes CD in the substrate of the mask, as in the above-mentioned US7736819 disclosed.

The correction of the structure sizes CD of reflective masks takes place by introducing density variations or inhomogeneities by electron radiation, as in the above-mentioned DE 10 2011 080 100 disclosed.

The illumination radiation used in the wafer exposure is scattered by these density variations. This results in a reduction of the transmitted or reflected illumination radiation. By the method mentioned only a reduction of the transmission or reflection of a mask is possible, but no increase. Therefore, an extreme value of all intensity values I is determined as the reference value T and used to calculate the dose changes A. In cases where the structure size CD is determined over bright areas of the aerial image, the minimum value of the intensities is determined as the extreme value. In cases where the critical dimension is determined over dark areas, a maximum of intensities is determined as the extreme value.

The determined dose changes A are then a direct measure of the necessary changes in the transmission or the reflectivity of a mask, which lead to corrected feature sizes CD.

In the examples mentioned from the 4 to 7 a brightfield mask was measured. Thus the minimum of the intensity values has to be determined. Thus, T = I _a = 0.275931, which corresponds to the intensity value of the measuring range 51 of the aerial photo 4 , The dose change for the measuring range 51 out 6 is then calculated according to Equation 3: A = (0.276715 - 0.275931) / 0.275931 = 0.028413

The determined values of the dose changes A with the dimensions of the measurement areas and the coordinates of the positions P on the mask are calculated by the arithmetic unit 40 provided in an output file. This output file is read in by a device for transmission modification, for example a CDC32 from Carl Zeiss SMS Ltd. The mask is inserted in this device and the transmission changes are made.

In a variant of the method, the values of the dose changes A or the values contained in the output file are used to control a can mappers or a CDU optimizer from ASML. This allows the optimization of the critical dimension through a customized exposure. The output file may contain additional information to control a can mappers, as explained below.

As the reference value T, as in the above variant, an extreme value of all intensity values I can be used.

In a variant of the method, instead of the extreme value of the intensity values I, a value is determined for the reference value T at which the mean value of all the determined values of the critical dimension CD corresponds to the setpoint value of the critical dimension CD _s . The dose changes A are calculated according to Equation 1. The use of this reference value T for controlling the can mappers or the CDU optimizer is possible because in these cases an increase and a decrease in the dose is possible.

In a variant of the method, the structure variables CD _{T are} determined for the reference value T for all regions in a further step. This is possible, for example, by the CMi software from Mentor Graphics in an automated process as described in " Mentor-Graphics CMi-Software, Automated aerial image based CD metrology initiated by pattern marking with photomask layout data, photomask and next generation lithography mask technology. Conference, Yokohama, Japan (2007) vol. 6607 (2), pp. 66072A.1-66072A.10 ".

The differences of the measured values CD _T from the target value of the CD _s are calculated as CD _delta as indicated in Equation 4. CD _delta = CD _T - CD _s 4

From the dose changes A and CD _delta described above, the Dependence of the CD change on the dose change can be determined as CD _delta / A.

For further measurements, it is then sufficient to determine the critical dimension CD of a similar mask with a simpler method, for example by measuring transmission changes, as in the above WO 2009 007977 A2 disclosed to be determined.

QUOTES INCLUDE IN THE DESCRIPTION

This list of the documents listed by the applicant has been generated automatically and is included solely for the better information of the reader. The list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions.

Cited patent literature

• US 7736819 [0007, 0071]
• DE 102011080100 [0007, 0072]
• US 20090191475 [0010]
• DE 102009015594 [0029, 0062]
• WO 2009007977 A2 [0037, 0083]

Cited non-patent literature

• Mentor-Graphics CMi-Software, Automated aerial image based CD metrology initiated by pattern marking with photomask layout data, photomask and next generation lithography mask technology. Conference, Yokohama, Japan (2007) vol. 6607 (2), pp. 66072A.1-66072A.10 [0080]

Claims (9)

Method for determining dose changes for adapting feature sizes of a mask comprising the steps of: specifying a target value CD _{s of} a feature size CD of the mask, recording at least one aerial image of the mask, determining intensity values 1 within the at least one aerial image at which the structure size CD corresponds to the set value CD _s , - determining the dose changes A as deviations of the intensity values I from an intensity value T. The method of claim 1, wherein the transmission or reflectivity of the mask is changed in accordance with the determined dose changes A. The method of claim 1 or 2, wherein upon performing a wafer exposure using the mask, the dose is adjusted according to the determined dose changes A. Method according to one of claims 1 to 3, wherein for the intensity value T, an extreme value of the intensity values I is determined. Method according to one of claims 1 to 4, wherein the dose changes A are determined as A = I - T / T , Method according to one of claims 1 to 5, wherein the intensity values are determined subpixel exactly. Method according to one of claims 1 to 6, wherein for the intensity value T all feature sizes CD _{T are} determined and differences CD _delta of the target value CD _{s are} determined. Method according to one of claims 1 to 7, wherein on the mask positions P are given, at which measurements are to be performed within measuring ranges. Having a mask inspection microscope an imaging optics for imaging a mask, a light source, a detector for receiving at least one aerial image of the mask, an arithmetic unit for carrying out the method according to one of claims 1 to 8.

Images