The Effects of Mask Error Factor on Process Window Capability

BACUS 1999 #3873-22

The Effects of Mask Error Factor on Process Window Capability Dan Schurz, Warren W. Flack, Simon Cohen, Tom Newman, Khiem Nguyen Ultratech Stepper, Inc. San Jose, CA 95134 In the photolithographic process, critical dimensions (CD) of exposed features in photoresist need to be controlled to within a specified tolerance related to the nominal feature size. A portion of this tolerance budget is consumed by variations in CD on the photomask. At low k1 factor, a number of parameters in the lithography system impact linearity including lens aberrations, defocus, exposure, partial coherence, and photoresist contrast. The combined effect of these parameters is that errors in the mask CDs are not transferred to the wafer in direct proportion to the optical reduction value of the lithography system. This Mask Error Factor (MEF) becomes a significant problem as it consumes a larger than anticipated portion of the CD tolerance budget. This paper will discuss experimentally evaluated MEF using a 4X i-line stepper for a range of feature sizes from subwavelength to approximately twice the exposure wavelength. A test reticle was built with isolated lines from 200 nm to 600 nm in 12.5 nm increments at 1X. CD measurements on the reticle were compared to corresponding CD measurements on the wafer in order to establish both linearity and MEF curves for the lithography system. MEF values were also determined across a process window for multiple feature sizes. The MEF was observed to be less than 1.4 for CDs greater than 330 nm (k1 = 0.5) throughout the process window. However, the MEF rises rapidly to over 3 for CD values smaller than 300 nm (k1 = 0.45) at nominal focus and exposure. Changes in exposure were not observed to have a noticeable impact on MEF while focus offsets were observed to result in significant increases in MEF. These results indicate that MEF has a much larger impact on focus latitude than on exposure latitude. As a result the process window will be compressed more in focus than in exposure. Key Words: Mask Error Factor (MEF), MEEF, Process Window, Linewidth Error, Non-linearity effect.

1.0 INTRODUCTION The semiconductor industry is aggressively pushing to produce smaller, lower k1 feature sizes from their existing base of lithography systems. This in turn drives designers and manufacturers to reduce the portion of the critical dimension (CD) error budget consumed by the reticle. Recently considerable attention has been given to the non-linear transfer of differing feature sizes within a pattern. This non-linear transfer results in a loss of the full error demagnification between reticle and wafer through reduction lithography systems at lower k1 values [1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11]. This loss of error demagnification is commonly referred to as Mask Error Factor (MEF) and is defined as: MEF = m (∂ CDwafer / ∂ CDreticle)

(1)

Schurz, Flack, Cohen, Newman and Nguyen

1

BACUS 1999 #3873-22

where ∂ CDwafer / ∂ CDreticle is the local slope of the linearity curve and m is the lens reduction ratio of the lithography system [5]. MEF has been shown to have a wide range of values depending on feature size, shape (line or contact) [4], proximity, pattern tone, mask bias [6, 7], the existence of enhancement features as well as illumination method, exposure, focus, and photoresist contrast [1, 3, 12]. As MEF values and their corresponding impact on the total CD error budget is increasingly understood, multiple sources of MEF are being investigated and methods of reducing MEF are being considered. P. Y. Yan was one of the early investigators of the non-linearity of wafer CDs relative to the photomask in low k1 lithography, and demonstrated the need for improved aerial image quality [7]. Van Schoot et. al. have suggested enhancing aerial image contrast as one way of reducing MEF. They have also shown that MEF can be reduced by changing the exposure dose and the threshold of the photoresist [3]. Wong et. al. have also shown, through simulation, that MEF can be reduced by using high contrast photoresist [1]. In addition to manipulating the light through the optical system, controlling the image focus is important in achieving high contrast. Microlithography in a production environment requires sufficient latitude to maintain statistical process control, that is, an exposure and focus window where CD control is maintained with a specified range. As a result, it becomes important to determine the extent of process window change when MEF values are effected by changes in the defining parameters. One of the conclusions of Yan’s investigation was that being centered in a process window does not always result in optimized CD control at low k1 values, especially without the application of positive, feature dependent mask biasing [7]. Wong’s simulation work showed that while a MEF of one is obtained for large feature sizes at best focus, as CDs decreased the MEF values rose to over 4 [1]. This is not surprising since the MEF values increase with focus change as defocus degrades aerial image contrast [3]. Because this degradation is non-linear, the MEF values for the smaller CDs are likely to have more of a negative impact on process window [1]. Maurer observed in simulations that increases in exposure dose reduced the slope of the pattern transfer function, which determines MEF [5]. Because the slope of the function is steeper for lower k1, the linewidth errors at the smaller CDs should be more influenced by variations in exposure dose [6]. Variations in dose will change the position of the aerial image with respect to the photoresist threshold which should in turn affect MEF [3]. Wong found this to be true for contacts, but determined little effect on line and space features [1]. Several of these studies have shown that nested lines and spaces have higher MEF than isolated lines [1, 3, 7]. Most of the experimental work has been performed on a very limited number of line sizes, typically at extremely low k1 values. Based on the difficulty of experimental verification, simulation has been used to evaluate a wider range of feature sizes. A disadvantage to an experimental study for nested lines is that multiple CD measurements must be taken at different linewidths with constant pitch to obtain a single MEF value. To perform this type of testing for a large range of nested CD values through focus and exposure variation would require an unmanageable amount of data. The alternative is to use isolated lines and determine the MEF by comparison of the CD of nearest sized features. This approach reflects the feature structure on the polysilicon level which typically has the tightest CD control requirements in an advanced CMOS technology. In this study, the intention is to determine experimentally how changes in focus and exposure dose affect the MEF of a variety of isolated line sizes from subwavelength to approximately twice the wavelength. This information will then be used to determine how MEF affects a process window.


2

BACUS 1999 #3873-22

2.0 EXPERIMENTAL METHODS 2.1 Reticle Design and Manufacture A 4X reticle was designed to determine the effect of MEF for a variety of isolated line sizes. The reticle contains nine data cells in a 20 by 20 mm field. Each data cell is composed of horizontal and vertical lines from 200 nm to 600 nm in 12.5 nm increments (at 1X) for a total of 33 unique feature sizes. These line sizes corresponds to k1 values from 0.3 to 0.9 for the optical stepper used in this study. These isolated lines were duplicated in both clear and dark tones. An alignment feature to facilitate wafer CD measurement on a KLA metrology SEM was also placed in the data cell. Figure 1a shows part of the data cell containing lines from 200 to 275 nm. The 4X reticle was written on a MEBES 4500 at DuPont Photomasks using their high-resolution process and measured on a KMS310 confocal linewidth measurement system. Figure 1b shows a linearity plot for vertical chrome lines located in the center cell of the reticle versus the CAD size in microns. The reticle CD values are scaled by the stepper reduction ratio (4:1). The dashed line with a slope of one provides a reference for the case of perfect linearity. The graph shows that the feature size is highly linear with respect to the CAD drawn size. The reticle errors can also be visualized by determining the normalized reticle CD error: Normalized CD Error = (Reticle CD / 4 - CAD Size) / CAD Size

(2)

Figure 1c shows a plot of the normalized reticle CD error in percent versus the CAD size in microns. The solid line is a regression analysis of the experimental data. There is a significant difference in the CD error as a function of CAD size. The small features near 200 nm show approximately 3.5 percent CD error while the larger features near 600 nm show CD errors of approximately 1.0 percent. The actual reticle CD size will be used in all subsequent analysis to eliminate this as an error source in the calculation of linearity and MEF.

2.2 Imaging Conditions All lithography was performed on an Ultratech Stepper XLS 7500/2955i®. The system specifications for the XLS 7500 are shown in Table 1. The stepper is a 4x reduction system using i-line illumination and having a numerical aperture of 0.55. The system is conservatively specified at 0.5 µm resolution with a 1.0 µm depth of focus. At a reasonable working k1 factor of 0.6, the achieved resolution is 0.4 µm. Multiple wafers were exposed in a focus/exposure matrix consisting of an eleven by eleven field array as shown in Figure 2. The nominal expose dose for the matrix was based on SEM measurements of a 600 nm isolated line. The exposure was varied from 185 mJ/cm2 to 285 mJ/cm2 in 10 mJ/cm2 increments while the focus was varied from -1.0 µm to +1.0 µm in 0.2 µm increments. A KLA 8100 low-voltage metrology SEM was used to measure the vertical linewidth for the 33 linesizes at each of the 121 unique fields in the matrix. All CD measurements were made at the data cell in the center of the stepper field. The SEM criteria selected for determination of the CD size is a 50 percent theshold. A total of 3872 linewidth measurements were made on each test wafer.

2.3 Processing Conditions Shipley Ultra-i 120® photoresist was selected for this study. This is a high contrast photoresist that requires bottom antireflective coating (BARC) for optimum performance. The BARC selected is Brewer Science XHRi A-16®. SEMI standard 200mm ultraflat silicon wafers were used for this work. The wafers were not vapor-primed prior to the application of the BARC as this may inhibit adhesion. The processes for the application of the BARC and the photoresist are shown in Tables 2 and 3 respectively. The final thicknesses for the XHRi A-16 BARC and the Ultra-i 120 photoresist are 160 nm and 960 nm respectively.


3

BACUS 1999 #3873-22

2.4 Data Analysis An isolated line was chosen for this study to reflect the poly gate in a typical CMOS process where CD control is critical for device performance. The metrology data was used to generate linearity curves for the wafer CDs as a function of the measured reticle CDs (at 1X) for each focus and exposure condition on the wafer. Using the linearity data it is then possible to calculate the MEF from the derivative of the linearity curve as defined in equation (1). The derivative at a given CDi is calculated as the slope of a linear regression analysis of the experimental data at CDi-1, CDi and CDi+1. As an example, the MEF of a 300 nm CD is the slope of the linearity curve at 287.5, 300 and 312.5 nm, while the MEF at 312.5 nm is the slope generated by the CD measurements at the 300, 312.5 and 325 nm lines. A least squares regression analysis is then performed on the MEF data to generate a curve representing the data. The MEF curves at each focus and exposure condition can then be used to establish the effect on nominal process window for a given CD.

3.0 RESULTS AND DISCUSSIONS 3.1 Linearity A plot of wafer CDs as a function of measured reticle CDs (at 1X) is shown for the nominal focus and exposure in Figure 3a. The dashed line with a slope of one provides a reference for the case of perfect linearity. While the 265 mJ/cm2 exposure dose is nominal for the 600 nm isolated line, the smaller features are underexposed down to approximately 270 nm (k1 = 0.41). At this CD there is a change in the slope of the curve and it drops below the dashed reference line. This CD is the linear resolution limit for this optical system and photoresist process. The effect of exposure dose on linearity is shown by comparing Figures 3e, 3a and 3d. The CDs in Figure 3e are underexposed for all line sizes down to 270 nm which is the linear resolution limit. The CDs in Figure 3d show a better nominal dose over a wider range of linewidths, but the linear resolution limit occurs at the same value of 270 nm. These three plots demonstrate that when focus is held constant, changes in exposure dose have a small overall affect on linearity. There is, however, a more noticeable impact on the smaller features, similar to Mauer’s simulation work [6]. The effect of focus on linearity is shown by comparing Figures 3b, 3a and 3c. The +0.6 µm focus in Figure 3b shows that the linear resolution limit occurs at a 330 nm CD (k1 = 0.50) which is larger than the 270 nm observed at nominal focus. The -0.6 focus in Figure 3c shows the linear resolution limit at 340 nm CD (k1 = 0.51). These three plots demonstrate that when exposure dose is held constant, changes in focus have a significant impact on the linear resolution limit.

3.2 MEF The linearity plots in Figure 3 were used to calculate the MEF as described in section 2.4. A plot of MEF as a function of reticle CD (at 1X) is shown for nominal focus and exposure in Figure 4a. The dashed line at a value of one provides a reference for the case of no MEF effect. The calculated MEF values show that there is a significant amount of variation in the data. This is not surprising since a derivative operation tends to emphasize experimental noise. Another source of variation is the fact that only three measurement points were used to calculate the slope at each CDi. However, using a larger data sampling would tend to suppress the actual MEF at small features by creating a moving average MEF versus CD. One approach that would help reduce the variation would be the use of smaller CD step increments to allow more data points to be sampled in the slope calculation. The range of the noise can be quantified by observing that for CDs larger than 300 nm the MEF varies from 0.75 to 1.25 while the expected value would be one. For smaller CD values the MEF rises rapidly to greater than 3. In


4

BACUS 1999 #3873-22

order to help visualize the MEF trend an inverse cubic function has been curve fit to the data using a least squares regression analysis. The effect of exposure dose on MEF is shown in Figures 4e, 4a and 4d. The MEF is less than 1.2 for CDs greater than 0.33 (k1 = 0.5) for all three exposure curves. At the 245 mJ/cm2 dose in Figure 4e the regression curve appears to be somewhat steeper than the regression curves at 265 and 285 mJ/cm2. However, all three curves appear very similar with MEF increasing at about the same values below 300 nm (k1 = 0.45). This is supported by the observation that the linear resolution limit in Figure 3 was the same for all three exposure doses. The effect of focus on MEF is shown in Figures 4b, 4a and 4c. The MEF is less than 1.4 for CDs greater than 0.33 (k1 = 0.5) for all three focus curves. Both the +0.6 µm focus in Figure 4b and the -0.6 µm focus in Figure 4c show a significant increase in MEF compared to the nominal focus in Figure 4a. At the +0.6 µm and -0.6 µm focus values the MEF begins to increase at larger CDs and then rises more rapidly as CDs decrease. For example, the MEF at 300 nm is greater than 2 for the +0.6 µm and -0.6 µm focus values while only 1.1 at nominal focus. This is due, in large part, to the degradation (reduced contrast) of the aerial image [1].

3.3 Process Window Analysis The effect of focus and exposure on MEF can be seen more clearly by superimposing the regression curves from Figure 4. A plot of MEF as a function of reticle CD (at 1X) for three focus values at nominal exposure dose is shown in Figure 5a. Both the +0.6 µm and -0.6 µm focus setting show an increase in MEF at CD values less than 400 nm compared with 320 nm for nominal focus. The MEF curve rises more rapidly for the +0.6 µm focus than the -0.6 µm focus. A plot of MEF as a function of reticle CD for three exposure values at nominal focus is shown in Figure 5b. At CDs smaller than 320 nm (k1 = 0.48) the MEF rises rapidly for all three exposures. The MEF rises somewhat more rapidly for the case of the 245 mJ/cm2 dose as was discussed in the previous section. However, the changes are much smaller overall than those observed by varying the focus. A contour plot of MEF for 300 nm isolated lines (k1 = 0.45) is shown in Figure 6. Here MEF is plotted as a function of focus and exposure dose. The contours tend to form a saddle where the MEF increases rapidly with focus. Note that if the contour lines were completely vertical there would be no influence on MEF by exposure variation. These results show that MEF has a much larger impact on focus latitude than exposure latitude. As a result the process window will be compressed more in focus than in exposure. Unfortunately, depth of focus is already marginal for subwavelength CDs because of the large NA required to obtain small k1 features. The results reinforce the observation that improved aerial image quality should help minimize the impact of MEF.

4.0 CONCLUSIONS This study has experimentally determined the linearity and MEF for isolated features for a range of CD sizes through focus and exposure. The conclusions are as follows: 1. The measured reticle linearity to the CAD size was excellent for larger feature sizes. However, for features approaching 200 nm the error exceeds 3.5 percent. This suggests that the use of CAD size would introduce an error source in the calculation of linearity and MEF. 2. At nominal exposure and focus the experimental linearity resolution limit for isolated lines is 270 nm (k1 = 0.41). Changes in exposure were not observed to have a significant effect on the linearity resolution limit. However, changes in focus of 0.6 µm shifted the linearity resolution limit to 340 nm (k1 = 0.51). 3. The experimentally determined MEF is less than 1.4 for CDs greater than 330 nm (k1 = 0.5) throughout a focus range of 1.2 µm and exposure range of 40 mJ/cm2.


5

BACUS 1999 #3873-22

4. The MEF was observed to rise rapidly to over 3 for CD values smaller than 300 nm (k1 = 0.45) at nominal focus and exposure. Changes in exposure were not observed to have a noticeable impact on MEF while focus offsets were observed to result in significant increases in MEF. For example, the MEF at 300 nm is greater than 2 for both the +0.6 µm and -0.6 µm focus settings, while it is only 1.1 at nominal focus. 5. Experimental variation on the order of ±0.25 was observed in the calculated MEF values. This effect is exaggerated by the use of only three data points to calculate the local slope of the linearity curve. 6. A MEF contour plot for 300 nm isolated lines (k1 = 0.45) as a function of focus and exposure dose shows that the contours tend to form a saddle where the MEF increases rapidly with focus. These results show that MEF has a much larger impact on focus latitude than on exposure latitude. As a result the process window will be compressed more in focus than in exposure.

5.0 ACKNOWLEDGEMENTS The authors would like to thank Scott Kulas and Makoto Nakamura for their assistance with exposure and photoresist processing of the wafers used in this study. We would also like to thank Sylvia White for programming and operating the metrology SEM for the CD measurements required for this study.

6.0 REFERENCES 1. A. Wong, R. Ferguson, L. Liebmann, S. Mannsfield, A. Molless, M. Neisser, “Lithographic Effects of Mask Critical Dimension Error,” Optical Microlithography XI Proceedings, SPIE 3334, 1998. 2. J. Waelpoel, J. van Schoot, A. Zanzal, “Demonstrating Next Generation CD Uniformity with Today’s Tools and Processes,” 17th Annual Symposium on Photomask Technology Proceedings, SPIE 3236, 1998. 3. J. van Schoot, J. Finders, K. van Ingen Schenau, M. Klaassen, C. Buijik, “The Mask Error Factor: Causes and Implications For Process Latitude,” Optical Microlithography XII Proceedings, SPIE 3679, 1999. 4. F. M. Schellenberg, V. Boksha, N. Cobb, J.C. Lai, C.H. Chen, C. Mack, “Impact of Mask Errors on Full Chip Budgets,” Optical Microlithography XII Proceedings, SPIE 3679, 1999. 5. W. Maurer, “Mask Specifications for 193 nm Lithography,” 16th Annual Symposium on Photomask Technology Proceedings, SPIE 2884, 1996. 6. W. Maurer, K. Satoh, D. Samuels, T. Fischer, “Pattern Transfer at k1 = 0.5: Get 0.25µm Lithography Ready for Manufacturing,” Optical Microlithography IX Proceedings, SPIE 2726, 1996. 7. P. Yan, B. Hainsey, J. Farnsworth, J. Neff, “Sub-micron Low-k1 Imaging Characteristics using a DUV Printing Tool and Binary Masks,” Optical/Laser Microlithography VIII Proceedings, SPIE 2440, 1995. 8. M. D. Levenson, “Can Phase-Shift Save the Semiconductor Industry?,” Proceedings of the Olin Interface 1998 Microlithography Symposium, 1998. 9. A. Vacca, B. Eynon, S. Yeomans, “Improving Wafer Yields at Low k1 with Advanced Photomask Defect Detection,” Solid State Technology, June 1998. 10. C. Mack, “Mask Linearity and the Mask Error Enhancement Factor,” Microlithography World, Winter 1999. 11. H. Levinson, P. Ackmann, M. Preil, B. Rericha, “The Factors Which Determine the Optimum Reduction Factor for Wafer Steppers,” Metrology, Inspection and Process Control XIII Proceedings, SPIE 3677, 1999.


6

BACUS 1999 #3873-22

12. K. Tsudaka, M. Sugawara, H. Kawahira, A. Ogura, S. Nozawa, “A New Mask Optimization Methodology Using Exposure-Defocus and Mask Fabrication Latitude,” Photomask and X-Ray Mask Technology Proceedings, SPIE 2254, 1994.

Parameter Reduction factor Wavelength (nm) Wafer Plane Irradiance (mW/cm2) Numerical Aperture (NA) Partial Coherence (σ) Field Size (mm) Feature Size (µm) Depth of Focus (µm)

Value 4X 365 400 0.55 0.54 29 Ø (20 ❑) 0.5 1.0

Table 1: System specifications for the Ultratech XLS 7500 reduction stepper.

Process Steps

Parameters

Equipment

BARC Coating

Dynamic dispense 700 rpm for 5 seconds 3500 rpm for 30 seconds

Soletic Coater

BARC Bake

175o C for 60 seconds

Soletic Hotplate

Table 2: Process conditions for Brewer Science XHRi A-16 BARC.

Process Steps

Parameters

Equipment

Photoresist Coating

Static dispense 2800 rpm for 30 seconds

MTI Coater

Softbake

90oC for 60 seconds

MTI Hotplate

Exposure

210 to 390 mJ/cm2

XLS 7500/2955i

Post Exposure Bake

110oC for 90 seconds

MTI Hotplate

Develop

Shipley CD-26 2 puddles at 30 seconds each

MTI Developer

Table 3: Process conditions for Shipley Ultra-i 120 photoresist.


7

BACUS 1999 #3873-22

10.0 um

0.225

0.20

0.25

0.275

Figure 1: (a) Partial view of the photomask data cell showing vertical and horizontal isolated features from 200 to 275 nm in 12.5 nm increments. 4.5 4.0

0.55

Reticle CD Error (%)

0.5 0.45 0.4 0.35 0.3 0.25

3.5 3.0 2.5 2.0 1.5 1.0

0.6

0.55

0.5

0.45

0.4

0.35

0.2

0.60

0.55

0.50

0.45

0.40

0.35

0.30

0.25

0.20

0.2

0.3

0.5 0.0

0.25

Reticle CD (microns)

0.6

CAD Size (microns)

CAD Size (microns)

(c) Reticle CD Error

(b) Mask Linearity

Figure 1: (b) Linearity plot of reticle CD versus CAD linewidth size. The dashed line is for reference and shows perfect linearity at a 45 degree slope. (c) Plot of reticle CD error versus CAD linewidth size showing a least squares regression analysis. The reticle CD values are scaled by the stepper reduction ratio (4:1). Individual field

Isolated lines

Focus

Exposure

Figure 2: Wafer layout for the focus and exposure test matrix. An eleven by eleven field array was exposed with focus varying in the vertical axis and exposure dose varying in the horizontal axis.


8

BACUS 1999 #3873-22

0.65

Wafer CD (microns)

0.6 0.55 0.5 0.45 0.4 0.35 0.3 0.25

0.55 0.60 0.65

0.45 0.50

0.30 0.35 0.40

0.20 0.25

0.2




mJ/cm2


mJ/cm2

(e) Exposure = 245 Focus = 0.0 microns

(a) Exposure = 265 Focus = 0.0 microns

(d) Exposure = 285 mJ/cm2 Focus = 0.0 microns

0.65 0.6 0.55 0.5 0.45 0.4 0.35 0.3 0.25 0.2 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65

Wafer CD (microns)

0.65 0.6 0.55 0.5 0.45 0.4 0.35 0.3 0.25 0.2 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65

0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65

Wafer CD (microns)

0.65 0.6 0.55 0.5 0.45 0.4 0.35 0.3 0.25 0.2

0.65

0.55 0.60

0.45 0.50

Wafer CD (microns)

0.65 0.6 0.55 0.5 0.45 0.4 0.35 0.3 0.25 0.2 0.20 0.25 0.30 0.35 0.40

Wafer CD (microns)

(b) Exposure = 265 mJ/cm2 Focus = 0.6 microns


(c) Exposure = 265 mJ/cm2 Focus = -0.6 microns Figure 3: Wafer linearity plots for the Shipley Ultra-i 120 photoresist with BARC on the XLS 7500 stepper. The focus and exposure is listed under each plot. The dashed line is for reference and shows perfect linearity at a 45 degree slope. The reticle CD values are scaled by the stepper reduction ratio (4:1).


9

BACUS 1999 #3873-22

4

Mask Error Factor

3.5 3 2.5 2 1.5 1 0.5

0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65

0


4 3.5

2.5 2 1.5 1

3 2.5 2 1.5 1

3 2.5 2 1.5 1

0

0



mJ/cm2


(d) Exposure = 285 mJ/cm2 Focus = 0.0 microns

mJ/cm2

(e) Exposure = 245 Focus = 0.0 microns

0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55

0

0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65

0.5

0.55 0.60 0.65

0.5

0.45 0.50

0.5

0.60 0.65

3

Mask Error Factor

4 3.5

Mask Error Factor

4 3.5

0.20 0.25 0.30 0.35 0.40

Mask Error Factor

(b) Exposure = 265 mJ/cm2 Focus = 0.6 microns

(a) Exposure = 265 Focus = 0.0 microns 4

Mask Error Factor

3.5 3 2.5 2 1.5 1 0.5

0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65

0


(c) Exposure = 265 mJ/cm2 Focus = -0.6 microns Figure 4: Plots of MEF for isolated lines as a function of Reticle CD using Shipley Ultra-i 120 photoresist with BARC on the XLS 7500 stepper. The focus and exposure is listed under each plot. The dashed line is for reference and shows a MEF of 1. The reticle CD values are scaled by the stepper reduction ratio (4:1).


10

BACUS 1999 #3873-22

4

4 MEF F = -0.6

MEF 245 mJ/cm2

3.5

3.5 MEF F = 0

MEF 265 mJ/cm2

3



(a) Exposure = 265 mJ/cm2

(b) Focus = 0.0 microns

0.65

0.60

0.55

0.50

0.65

0.60

0.55

0.50

0.45

0.40

0 0.35

0 0.30

0.5 0.25

0.5

0.45

1

0.40

1

1.5

0.35

1.5

2

0.30

2

MEF 285 mJ/cm2

2.5

0.25

2.5

0.20

Mask Error Factor

MEF F = 0.6

0.20

Mask Error Factor

3

Figure 5: Plots of MEF for isolated lines as a function of linesize using Shipley Ultra-i 120 photoresist with BARC on the XLS 7500 stepper. Figure (a) shows the effect of focus offset on MEF. Figure (b) shows the effect of exposure dose on MEF. These plots are based on the regression curves shown in Figure 4. The reticle CD values are scaled by the stepper reduction ratio (4:1).

Exposure (mJ/cm2)

245

MEF 2.4 2.2 2.0 1.8 1.6 1.4 1.2 1.0

265

285 -0.6

0 Focus (microns)

0.6

Figure 6: Contour plot of MEF as a function of exposure dose and focus offset for a 300 nm isolated line.


11

The Effects of Mask Error Factor on Process Window Capability

Recommend Documents