Radiant energy – Photocells; circuits and apparatus – Photocell controls its own optical systems
Reexamination Certificate
1998-11-16
2001-09-04
Lee, John R. (Department: 2878)
Radiant energy
Photocells; circuits and apparatus
Photocell controls its own optical systems
C356S401000
Reexamination Certificate
active
06285033
ABSTRACT:
FIELD OF THE INVENTION AND RELATED ART
This invention relates to a positional deviation detecting method, and to an exposure apparatus or a device manufacturing method using the same. The present invention is particularly suitable for use in an exposure apparatus or a scan type exposure apparatus to be used in a lithographic process for the manufacture of microdevices such as semiconductor devices (e.g., IC or LSI), image pickup devices (e.g., CCD), display devices (e.g., liquid crystal panel) or magnetic heads, for example, for relative positioning or alignment between a first object such as a mask or reticle (hereinafter “mask”) and a second object such as a wafer when a fine pattern such as an electronic circuit pattern formed on the first object is to be lithographically transferred to the second object.
In exposure apparatuses for device manufacture, relative alignment of a mask and a wafer is an important factor in improvement of performance. For a DRAM which is a representative semiconductor integrated circuit, an overall registration precision of about {fraction (1/3+L )} to {fraction (1/4+L )} of the minimum linewidth of a resolution pattern image is required. Particularly, in recent exposure apparatuses, an alignment precision of 20 nm or less is required to meet further enlargement of integration of semiconductor chips.
For a 1-Gbit DRAM currently being developed, an overall registration precision of 40 nm to 50 nm is required and, within it, the precision to be shared for the alignment precision will be 10-15 nm.
In many exposure apparatuses, a mask and a wafer are formed with positioning marks, called alignment marks. A positional deviation between these alignment marks is optically detected and, on the basis of a detected value, the positioning (alignment) of the mask and the wafer is performed. As for detection of alignment marks, there are a method in which a mark is optically enlarged and projected on a CCD by which image processing is performed; a method in which a straight diffraction grating is used as a mark and the phase of diffractive light produced thereby is measured; and a method in which a zone plate (grating lens) is used as a mark and light diffracted by the zone plate is detected upon a predetermined plane, whereby positional deviation of the diffracted light is detected.
Among these detection methods, the methods that use a straight diffraction grating or a zone plate as an alignment mark have a feature that, in the sense that detection is less influenced by any defect or fault of the mark, it is tough on the semiconductor process and enables relatively high precision alignment.
FIG. 1A
is a schematic view of a position detecting system of a conventional type. In the drawing, parallel light emitted from a light source
72
passes through a half mirror
74
and, then, it is collected by a condenser lens
76
toward a convergence point
78
. After this, the light illuminates a mask alignment pattern
68
a
upon the surface of a mask
68
and a wafer alignment pattern
60
a
upon the surface of a wafer
60
. The alignment patterns
68
a
and
60
a
each comprises a reflection type zone plate which serves to define a spot of light convergence upon a plane which is orthogonal to the optical axis, passing through the convergence point
78
. A deviation of light convergence spot position upon that plane is detected by a detector
82
, with the light being guided thereto by the condenser lens
76
and another lens
80
.
Control circuit
84
actuates a driving circuit
64
on the basis of an output signal from the detector
82
, by which relative positioning of the mask
68
and the wafer
60
is performed.
FIG. 1B
is a schematic view for explaining the imaging relation of lights from the mask alignment pattern
68
a
and the wafer alignment pattern
60
a
, shown in FIG.
1
A.
In
FIG. 1B
, a portion of the light divergently emitted from the light convergence point
78
is diffracted by the mask alignment pattern
68
a
, by which a light convergence point
78
a
, representing the position of the mask, is defined adjacent to the point
78
. Also, another portion of the light passes through the mask
68
as zeroth order transmissive light and, with its wavefront unchanged, it impinges on the wafer alignment pattern
60
a
on the wafer
60
surface. After being diffracted by the pattern
60
a
, the light passes through the mask
68
again as zeroth order transmissive light, and it is collected in the neighborhood of the light convergence point
78
, whereby a light convergence point
78
b
representing the position of the wafer is produced.
In
FIG. 1B
, when the light diffracted by the wafer
60
defines a light convergence spot, the mask
68
serves simply as a transparent member.
The position of the light convergence point
78
b
thus produced by the wafer alignment pattern
60
a
bears a deviation &Dgr;&sgr;′ corresponding to a deviation &Dgr;&sgr; of the wafer
60
with respect to the mask
68
, in a direction (lateral direction) along the mask or wafer surface and along a plane perpendicular to the optical axis, passing through the light convergence point
78
.
The amount of this deviation &Dgr;&sgr;′ is measured with reference to an absolute coordinate system defined upon a sensor, whereby the deviation &Dgr;&sgr; is detected.
Usually, for alignment of a mask and a wafer based on detection of a positional deviation therebetween, the mask and the wafer are controlled to be placed with a mutual spacing in a predetermined range and, thereafter, they are brought into alignment on the basis of positional information obtainable from a sensor by use of alignment patterns provided on the mask and the wafer.
Such a method, however, involves a problem that Fraunhofer diffraction light from openings of mask and wafer alignment marks enter a central portion of a sensor, to cause interference with signal light that results in a decreased signal-to-noise ratio of a produced alignment signal as well as non-linearity of a signal to the mask-to-wafer relative deviation.
The influence of such Fraunhofer diffraction light may be reduced by arranging, as shown in
FIG. 1C
, a wafer alignment mark WA with eccentricity with respect to a mask alignment mark MA, in a state where there is no positional deviation between a mask circuit pattern and a wafer circuit pattern. With this arrangement, since the Fraunhofer diffraction light from the openings is spatially separated from the signal light, the influence of interference is reduced and a good signal is produced.
With this method, however, since the wafer alignment mark WA has a shape asymmetrical with respect to the alignment detecting direction, as shown in
FIG. 1C
, there is a problem that asymmetrical non-uniformness of diffraction efficiency is easily produced within the alignment mark.
If the diffraction efficiency distribution within the mark is asymmetric, a beam spot of signal light produced on a sensor becomes asymmetric. It causes a shift of the peak position and gravity center position of the spot, which leads to a detection error. The degree of influence to asymmetry of a signal light spot of the asymmetry of diffraction efficiency distribution within the mark, becomes notable with a larger deviation of the mask-to-wafer spacing (gap) with respect to a design gap, since it causes defocus of the spot.
The non-uniformness of diffraction efficiency within the mark may be attributable to the fact that: since an alignment mark has a power with respect to the alignment direction, the pitch thereof (i.e., the linewidth thereof) changes within the pattern, whereas it is difficult to control the linewidth and pattern step height (level difference) over the whole wafer surface, with respect to every linewidth and throughout the etching process, the deposition process and so on of the device process.
Particularly, due to miniaturization of a circuit pattern, the linewidth of a circuit pattern is 0.15 micron or less and the linewidth range of an alignment mark is widened, from a few microns to about 10 mic
Canon Kabushiki Kaisha
Fitzpatrick ,Cella, Harper & Scinto
Lee John R.
Pyo Kevin
LandOfFree
Positional deviation detecting method and device... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Positional deviation detecting method and device..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Positional deviation detecting method and device... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2512226